Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Prenatal and lifestyle predictors of metabolic health and neurocognition during childhood
(USC Thesis Other)
Prenatal and lifestyle predictors of metabolic health and neurocognition during childhood
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
2020 Alves, Jasmin Marie
PRENATAL AND LIFESTYLE PREDICTORS OF METABOLIC HEALTH
AND NEUROCOGNITION DURING CHILDHOOD
by
Jasmin Marie Alves
This Dissertation is Presented to the
FACULTY OF THE UNIVERSITY OF SOUTHERN CALIFORNIA GRADUATE SCHOOL
In Partial Fulfillment of the Requirements for the Degree
DOCTOR OF PHILOSOPHY
(MEDICAL BIOLOGY PROGRAM)
August 2020
ii
This dissertation is dedicated to my father, who taught me to never give up no matter
what life throws your way.
iii
Acknowledgements
They say it takes a village to raise a child. Over the past 6 years, it most certainly took a
village to get me to this point of finishing my PhD. I am so incredibly grateful for
everyone in my life who supported me along the journey.
Words cannot express my gratitude to my mentor, Dr. Page. Thank you for your
constant support and inspiration. Your tenacity and passion for science is contagious. It
has been a privilege to work with someone as smart and dedicated as you are. You
have instilled in me a genuine zeal for human health and scientific innovation that I will
forever be grateful for.
I feel truly fortunate to have had such a supportive dissertation committee. You
all went above and beyond to mentor me along the years. Thank you, Dr. Watanabe
and Dr. Monterosso, for always being available for guidance and being there in my
many moments of sheer distress. Thank you, Dr. Herting and Dr. Xiang, for all the
informative meetings and all the incredibly useful knowledge you have given me over
the years. I have learned so much from all of you to be the scientist I am today, and I
can only hope I made you proud.
I can honestly say I have the best lab mates. It was truly a blessing to work with
such a passionate and caring team. Thank you, Ana, Shan, Brendan, Alex, Alexis, and
Sabrina. You all have been instrumental in helping me complete my PhD. We really
have become a family over the years. And I have loved working with you all.
I also want to acknowledge the incredibly selfless support of my husband, Cesar
Nuño. Thank you for being ridiculously supportive and loving me through it all. Cesar did
everything possible to help me succeed and for that I am exponentially thankful. To
Cesar’s family and my own, Anthony and Selena, you have all been so patient as I have
endured the life of being a perpetual student. Thank you for supporting me even when I
missed social obligations or was too stressed to help with mom. And thank you, dad, for
always encouraging me to reach for the stars, even on the days when I felt the world
had defeated me. Your hard-working spirit continues to leave me in awe. While he is no
longer here, I also want to thank my incredibly hard-working grandfather, Manual Alves.
He always encouraged me to study more than I thought I needed to.
I also want to thank all of my amazing and supportive friends, Kendra, Rachel,
Neya, Jesse, Kristine, Mary and the Pasadena City Church community for their prayers
and support.
Thank you also to the PIBBS program, Bami, Drs Joyce and Ite. You all helped
me with whatever I needed to finish this PhD, so from the bottom of my heart, thank
you.
Lastly, I want to thank the fantastic medical doctors and nurse practioners at
Keck Medicine of USC. If it wasn’t for these individuals, I do not think I would be as
healthy and capable as I am today. They really did give me my quality of life back.
iv
Table of Contents
Title Page i.
Dedications ii.
Acknowledgements iii.
Table of Contents iv.
List of Tables vi.
List of Figures xi.
Project Abstract xiii.
Chapter One: Introduction to Dissertation 1
Part One: Metabolic and Neurocognitive Consequences of in utero 1
Exposure to Gestational Diabetes Mellitus or Maternal Obesity
Part Two: Cognition and the Hungry, Hungry Hippocampus 7
Part Three: An Overview of Hippocampal Structural Development 30
Chapter Two: Sex differences in the Association between Prenatal Exposure 32
to Maternal Obesity or Gestational Diabetes and Hippocampal Volume in Children
Part One: Maternal Obesity 32
Part Two: Gestational Diabetes 51
Chapter Three: Contributions of Prenatal Exposures and Child Lifestyle 64
to Insulin Sensitivity and Adiposity during Childhood
Chapter Four: Impact of prenatal exposure to Maternal Obesity and 83
v
Child Physical Activity Levels on Neurocognition
Chapter Five: Concluding Remarks and Future Directions 95
Cumulative Methods 98
References 108
Appendix of Figures and Tables 1 to 40
vi
List of Tables
Table 1. MRI data Characteristics of the 88 Child Participants and Appendix page 5
their Mothers
Table 2. Characteristics of the 99 Child Participants and Appendix page 6
their Mothers for Cognitive data
Table 3. Regression coefficients between Maternal Pre-pregnancy Appendix page 8
BMI per 5-unit increments and Total Hippocampal Volume and
Hippocampal Subfield Volumes in Boys (N=37)
Table 4. Regression coefficients between Maternal Pre-pregnancy Appendix page 9
BMI per 5-unit increments and Total Hippocampal Volume and
Hippocampal Subfield Volumes in Girls (N=51)
Table 5. Regression coefficients between Maternal Pre-pregnancy Appendix page 10
BMI per 5-unit increments and Total Hippocampal Volume and
Hippocampal Subfield Volumes (N=88)
Table 6. Correlation Matrix for Hippocampal-Related Memory Appendix page 11
Table 7. Association between maternal pre-pregnancy BMI and Appendix page 11
hippocampal-dependent memory (Relational Memory),
hippocampal-independent memory (Item-familiarity), and
picture-sequence memory task, and full-scale IQ
vii
Table 8. In girls, the association between maternal pre-pregnancy Appendix page 12
BMI and hippocampal-dependent memory (Relational Memory),
hippocampal-independent memory (Item-familiarity), and
picture-sequence memory task, and full-scale IQ
Table 9. In boys, the association between maternal pre-pregnancy Appendix page 12
BMI and hippocampal-dependent memory (Relational Memory),
hippocampal-independent memory (Item-familiarity), and
picture-sequence memory task, and full-scale IQ
Table 10. The association between relational memory and Appendix page 13
hippocampal volume. ICV-adjusted*
Table 11. The association between episodic memory and Appendix page 14
hippocampal volume. ICV-adjusted*
Table 12. The association between item-familiarity performance Appendix page 15
and hippocampal volume. ICV-adjusted*
Table 13. LSmeans of Hippocampal Volume stratified by GDM- Appendix page 16
exposure in Boys
Table 14. LSmeans of Hippocampal Subfield Volume stratified Appendix page 16
by GDM-exposure in Boys
Table 15. Effect size of GDM-exposure as Cohen’s D in Boys Appendix page 17
Table 16. LSmeans of hippocampal volume stratified by GDM- Appendix page 18
exposure in Girls
viii
Table 17. LSmeans of Hippocampal Subfield Volume stratified Appendix page 18
by GDM-exposure in Girls
Table 18. Effect size of GDM-exposure as Cohen’s D in Girls Appendix page 19
Table 19. LSmeans of hippocampal-dependent memory Appendix page 20
(relational memory), hippocampal-independent memory
(item familiarity), picture-sequence memory task and full-scale IQ
in Boys
Table 20. Effect size of GDM-exposure on cognition Appendix page 21
Cohen’s D, in Boys
Table 21. LSmeans of hippocampal-dependent memory Appendix page 22
(relational memory), hippocampal-independent memory
(item familiarity), picture-sequence memory task and full-scale IQ
in Girls
Table 22. Effect size of GDM-exposure on cognition Appendix page 23
Cohen’s D, in Girls
Table 23. Summary of Demographics for BrainChild Cohort Appendix page 25
Table 24. Chapter Two: Participant Demographics (N=91) Appendix page 26
Table 25. Summary of unadjusted and adjusted linear regression Appendix page 27
models, with Matsuda insulin sensitivity index as the outcome variable
Table 26. Summary of unadjusted and adjusted linear regression Appendix page 29
models, with child BMI z-score as the outcome variable
Table 27. Summary of unadjusted and adjusted linear regression Appendix page 30
ix
models, with child total percent body fat as the outcome variable
Table 28. Summary of unadjusted and adjusted linear regression Appendix page 31
models, with child waist to height ratio as the outcome variable
Table 29. Summary of unadjusted and adjusted logistic regression Appendix page 32
models, with odds of being overweight/obese as the outcome variable
Table 30. Summary of unadjusted and adjusted linear regression Appendix page 33
models, with maternal pre-pregnancy BMI as the predictor variable
Table 31. Summary of unadjusted and adjusted linear regression Appendix page 34
models, with GDM exposure as the predictor variable
Table 32. Participant and Mother’s Demographics (N=100) Appendix page 35
Table 33. The Relationship between Maternal Pre-Pregnancy BMI Appendix page 36
and Physical Activity with Child IQ
Table 34. The Relationship between Maternal Pre-Pregnancy BMI Appendix page 37
and Physical Activity with Global FA
Table 35. Tracts with significant FA clusters for children above the Appendix page 38
median reported time spent in VPA, before adjusting for maternal
pre-pregnancy BMI, at a threshold of p<0.05 and controlling for
multiple comparisons
Table 36. Tracts with significant FA clusters for children above the Appendix page 38
median reported time spent in VPA, after adjusting for maternal
pre-pregnancy BMI, at a threshold of p<0.05 and controlling for multiple comparisons
Table 37. Tracts with significant FA clusters for the interaction Appendix page 38
between maternal weight status and vigorous physical activity levels,
at a threshold of p<0.05 and controlling for multiple comparisons.
x
Table 38. Tracts with significant FA clusters for the interaction Appendix page 39
between maternal weight status and vigorous physical activity levels,
at a threshold of p<0.01 and controlling for multiple comparisons.
xi
List of Figures
Figure 1. Formula for Matsuda insulin sensitivity index. FPG, Appendix page 1
fasting plasma glucose. FPI, fasting plasma insulin. OGTT,
oral glucose tolerance test
Figure 2. FreeSurfer Subcortical Reconstruction Pipeline Appendix page 1
Figure 3. Hippocampal sub-regions denoted by different colors. Appendix page 2
RED: CA1, GREEN: CA2/3, TURQUISE: Dentate Gyrus,
BLUE: subiculum, GREY: CA4
Figure 4. Summary of Types of Memory Appendix page 3
Figure 5. Example of Relational Memory Task Appendix page 4
Figure 6. Sex stratified scatter plots between maternal Appendix page 7
pre-pregnancy BMI and child hippocampal volume (A) and volumes of
hippocampal subfields (B), N=88.
Figure 7. Enrollment Flowchart BrainChild Appendix page 24
Figure 8. LSmeans of Insulin Sensitivity by Terciles of Moderate Appendix page 28
to Vigorous Physical Activity
Figure 9. LSmean IQ scores of children above and below Appendix page 39
median reported time spent in VPA.
Figure 10. LSmean global FA of children above and below Appendix page 40
median reported time spent in VPA.
xii
Figure 11. Axial view of TBSS results for children above the Appendix page 41
median reported time spent in VPA compared to children below the
median
xiii
Abstract
This dissertation project had two study aims. The first aim of this dissertation project
was to determine how prenatal exposure to gestational diabetes mellitus or maternal
obesity impacts metabolic health and neurocognition during childhood. The second aim
of this project was to determine how modifiable lifestyle factors such as physical activity
levels contribute to metabolic health and neurocognition during childhood. Electronic
medical records were used to determine mother’s pre-pregnancy BMI (kg/m
2
) and
gestational diabetes mellitus status. Children came in for two study visits. During the
first study visit, they completed an oral glucose tolerance test, their anthropometric
measurements were captured, diet and physical activity levels were assessed, and
cognitive assessments were completed. From the oral glucose tolerance test, Matsuda
insulin sensitivity index was calculated to assess insulin sensitivity. The 3-day physical
activity recall was used to assess sedentary time, moderate to vigorous physical activity
(MVPA) and vigorous physical activity (VPA), using metabolic equivalents (METS).
METS>1.1 and <1.5 determined sedentary time. METS>3 determined moderate to
vigorous physical activity and METS>6 determined vigorous physical activity. The
Weschler abbreviated scale of intelligence, 2
nd
edition was used to assess intelligence
quotient (IQ). The NIH toolbox, picture-sequence memory task was used to assess
episodic memory and a visuospatial memory task was used to assess hippocampal-
dependent, relational memory. During the second study visit, participants completed
magnetic resonance imaging (MRI) and a dietary recall. Functional and structural scans
were collected. For this dissertation, the T1-structural images and diffusion weighted
images (DWI) were used to analyze hippocampal grey matter volume and global white
xiv
matter. FreeSurfer was used to calculate hippocampal grey matter volume. FMRIB
software library (FSL) was used to calculate global fractional anisotropy (FA). Tract-
based spatial statistics was used to compare significant clusters of FA across the brain
between children who engaged in above or below the median time spent in VPA.
Compared to prenatal exposures and dietary components, MVPA was the only predictor
of child insulin sensitivity suggesting that engaging in moderate to vigorous physical
activity during childhood is beneficial for insulin sensitivity and may ameliorate future
risk for metabolic disease. Additionally, we found that there was a significant interaction
between boys and girls in the relationship between maternal pre-pregnancy BMI and
hippocampal grey matter volume. Compared to girls, boys exposed to higher levels of
maternal pre-pregnancy BMI, had significantly reduced hippocampal grey matter
volume. Interestingly, both boys and girls exposed to higher levels of maternal pre-
pregnancy BMI had reduced IQ scores. In contrast, higher IQ scores and greater global
FA was observed in participants who engaged in VPA for at least 10 minutes a day, and
this was independent of maternal obesity. Additionally, there was a significant
interaction between VPA and maternal obesity exposure on both child IQ and global FA.
VPA was also associated with greater FA in the superior longitudinal fasciculus, and
anterior thalamic radiation (ATR). Further, there was a significant interaction between
maternal pre-pregnancy BMI category and VPA on FA in left forceps major, left ATR,
right cingulate gyrus, right inferior frontal-occipital fasciculus (IFOF), and cingulum
(hippocampal portion). These findings suggest that engaging in VPA may be particularly
beneficial for children exposed to maternal obesity in utero. Collectively, two important
findings were a result of this dissertation. It is pivotal to consider sex differences in the
xv
effects of in utero exposure to maternal obesity and secondly, engaging in physical
activity during childhood is important for metabolic health and neurocognition and may
be protective for children exposed to maternal obesity in utero. Future longitudinal and
intervention studies are needed to confirm these findings.
1
Introduction Chapter
Part One: Metabolic and Neurocognitive Consequences of in utero Exposure to
Gestational Diabetes Mellitus and/or Maternal Obesity.
Since the Dutch famine in 1944-1945, where several pregnant women were
severely malnourished, many informative studies have illuminated the importance of
developmental programming in utero
1–3
. Developmental programming can best be
summarized as the environmental stimuli that the fetus is exposed to in utero that can
impact the trajectory of offspring development. This concept has been hypothesized as
the Developmental Origins of Health and Disease Hypothesis”
4,5
. For example, studies
in rodents have found that an offspring’s preference for “junk” food can be programmed
in utero
6–8
. Rat offspring exposed to mothers who consumed a “junk-food” diet during
pregnancy and lactation, were more likely to prefer junk food to normal chow. Further,
they were more likely to be obese and engage in hyperphagia of “junk-food”. Other
studies have found similar findings, that independent of energy expenditure, rats still
over-consume a high fat diet compared to chow when exposed to mothers on a “junk
food” diet during pregnancy and lactation
6–8
.
An obesogenic diet during pregnancy and a mother’s obesity status prior to the
onset of pregnancy can both influence an offspring’s adiposity. Maternal obesity is
defined as a pre-pregnancy BMI greater than 30 that occurs prior to the onset of
pregnancy and remains during pregnancy, it is also described as pre-gravid obesity
9
.
Approximately 30% of women of childbearing age are obese
10
; and offspring are
2
therefore exposed to an obesogenic environment in utero. Consequently, increasing
rates of in utero exposure to maternal obesity is the result of the increased obesity
pandemic currently affecting developed nations across the world. It is a pertinent public
health concern because in addition to the risks posed on the mother, the child is at
increased risk for various health concerns. For example, studies in children exposed to
maternal obesity in utero, have shown increased obesity risk during childhood
compared to children whose mothers were normal weight during pregnancy
11–14
. One
study by Reilly et al.
12
, showed that at 7 years old, the risk of being obese was four
times higher among children exposed to maternal obesity in utero. A similar study by
Salsberry & Reagan
13
, showed that children exposed to maternal obesity in utero, were
almost three times as likely to be obese compared to children whose mothers were
normal weight during pregnancy.
As obesity becomes a public health concern that continues to affect more and
more people unequivocally, grievously, the number of children affected by obesity
continues to rise
15
. With obesity, comes a slew of health concerns including, Type II
diabetes mellitus (T2DM), metabolic syndrome (METS), cancer and cardiovascular
disease during adulthood and more recently shown, obesity induced memory
impairments
15–22
. These various health ailments occur in comorbidity with obesity due
to excessive body fat being stored on vital organs
23
. Obesity during childhood is
defined using a BMI percentile growth chart created by the CDC
24
. Because children
are still growing, the BMI percentile chart is based on growth curves developed by the
CDC. A child in the 95th percentile is considered obese, while a child in the 85th
percentile is considered overweight. Recent reports from the CDC show that
3
approximately 18% of children and adolescents are obese
25
. Further, almost 1 in 3
children and adolescents are overweight or obese
26
. Therefore, based on rising rates of
obesity, it is pertinent to understand the role of prenatal exposures in offspring
metabolic disease risk.
In utero exposure to gestational diabetes mellitus (GDM) is another risk factor for
child obesity
27,28
. GDM is defined as diabetes that is first diagnosed during the time
period when a woman is pregnant
29
. It is characterized by abnormally high levels of
blood glucose in the mother. As a result of high blood glucose levels in the mother, the
developing fetus is exposed to hyperglycemia
30–32
. Approximately 1 in 10 children are
exposed to mothers who develop GDM, and there has been a significant rise in GDM
incidence over the last decade
33–35
. Page et al.
32
found that children exposed to GDM
in utero, had an increased BMI percentile compared to non-exposed children. Along this
line of research, Kubo et al.
31
found that children ages 6 to 8 years old exposed to
GDM in utero were more likely to have larger BMI’s and a greater total percent body fat.
Further, this study found that children exposed to mothers who had both GDM and pre-
gravid obesity, were most likely to be overweight and have a higher waist to hip ratio.
Additional to obesity risk, studies in offspring exposed to GDM in utero, have found, that
offspring also have an increased risk for T2DM and insulin resistance
27,30,36–38
.
Similarly, children born to mothers with obesity during pregnancy are also at
higher risk for insulin resistance and T2DM
9,39–42
. Prior studies have observed an
association between in utero exposure to maternal obesity and reduced insulin
sensitivity in offspring at birth
9,40
and during adulthood
39,42,43
. Interestingly, studies
4
during childhood have found that offspring exhibit insulin resistance, however it is yet to
be determined if this is independent of child adiposity
41,44,45
.
There are other in utero exposures associated with an increased risk for
childhood obesity and insulin resistance such as mothers who smoke during pregnancy
or mothers who are undernourished during pregnancy
2,13,46,47
. However, the focus of
this dissertation will be on in utero exposure to maternal obesity and in utero exposure
to GDM.
As introduced, in utero exposure to maternal obesity or GDM are both risk factors
for childhood obesity and insulin resistance. In addition to metabolic risks, these in utero
exposures are also associated with an increased risk of mental health problems, and
mild cognitive deficits, that are independent of a child’s environment
48–61
. For example,
a study by Casas et al.
52
, showed that infants exposed to maternal obesity in utero, had
delayed cognitive development. Interestingly, paternal BMI was unrelated to the infant’s
cognitive development. While animal studies have shown that in utero exposure to
maternal obesity impacts brain development and leads to cognitive impairments
8,62,63
,
there are limited studies in humans.
The fetal brain begins to develop as early as the second trimester of pregnancy,
therefore it has been hypothesized that an obesogenic environment in utero, contributes
to altered pathways in the brain which in turn are associated with obesity related
behaviors
5,64
. In animal models, exposure to maternal obesity in utero, is associated
with altered brain development in the hypothalamus and hippocampus along with
altered gene expression in the forebrain
8,62,65,66
. Behaviorally, in utero exposure to
5
maternal obesity is associated with hyperphagia, cognitive impairments and increased
anxiety
63,67–69
.
While there are numerous studies in rodents exemplifying the effects of maternal
obesity exposure in utero on offspring brain development, there are limited studies in
humans. To date, only a handful of studies in humans have recapitulated findings in
rodent studies
70,71
. For example, our group found that maternal obesity exposure in
utero was associated with an altered hypothalamic response to glucose in children ages
7 to 11 years old
71
. Additionally, our group has shown that maternal pre-pregnancy BMI
was negatively associated with offspring hippocampal grey matter volume
70
. To
conclude, similar to rodent studies, we have found that maternal obesity exposure in
utero is associated with alterations in the hypothalamus and hippocampus. Given the
compelling findings in animal studies, our lab has focused on the hypothalamus and
hippocampus, however there may be other brain regions impacted by in utero exposure
to maternal obesity. For the purpose of this dissertation, the focus will be on the
hippocampus. As a brief introduction, the hippocampus is a vital brain region involved in
memory, cognition, emotional processing, and appetite regulation
72–77
. In depth
functionalities of the hippocampus will be detailed in Part Three of this Introduction
chapter.
Although human studies about maternal obesity and offspring brain development
are limited, there have been many epidemiological studies that have found an
association between maternal obesity exposure in utero and offspring cognition. For
example, a study in 30,212 mother and child pairs, found that a higher maternal pre-
pregnancy BMI was associated with reduced IQ scores in children who were 7 years old
6
78
. Further studies have found, maternal obesity was associated with worse
performance on tests of verbal and nonverbal ability, deficits in working memory, and
impairments in reading and math skills in offspring
49,58,79–81
. These findings were
irrespective of the child’s BMI.
Similar to studies about in utero exposure to maternal obesity and child cognition,
some studies have eluded to GDM exposure contributing to mild cognitive deficits and
altered brain development
50,82,83
. For example, compelling studies in animals have
found that young rats exposed to GDM in utero have impaired hippocampal structure
and function, including neuronal loss in the CA1/CA3 subfields and alterations in cell
signaling
84–88
. In line with these findings, a study by Bolanos et al.
50
, showed that
GDM-exposed children aged 7-9, had specific deficits on the working memory subscale
of the Weschler Intelligence Scale for children IV (WISC-IV). Additionally, 12-month old
infants, exposed to GDM in utero exhibited impaired recognition and declarative
memory when compared to non-exposed age-matched controls
89,90
.
Collectively, research suggests that in utero exposure to either GDM or maternal
obesity not only increases the risk for a child being overweight or obese, but also
increases the risk for mild cognitive impairments. Several other studies have found that
GDM/maternal obesity exposure in utero is associated with neurodevelopmental
disorders such as Autism spectrum disorder and Attention Deficit Hyperactive Disorder
54,56,61,91
. It has been proposed that the altered development in utero, (i.e., GDM and
maternal obesity) contributes to the development of neurodevelopmental disorders
during childhood. However, the neural correlates of these findings have yet to be
determined.
7
To summarize, in utero exposure to GDM/maternal obesity, are both associated
with increased risk for childhood obesity, insulin resistance, mental health problems,
mild cognitive deficits, neurodevelopmental disorders, and differences in brain,
specifically impacting the hippocampus and hypothalamus. Altogether, evidence
suggests, that an altered metabolic environment in utero contributes to differences in
hippocampal structure and function.
Concluding notes: As the reader assesses this dissertation for clarity, the focus will
mainly be on how prenatal exposures impact child metabolic health and neurocognition.
While a large focus of this dissertation will detail the negative consequences of prenatal
exposures, the benefits of physical activity during childhood will also be explored as a
potential therapeutic recommendation.
Part Two: Cognition and the Hungry, Hungry Hippocampus
Assessing the Human Hippocampus
There have been many scientific advances in the field of neuroscience to better
understand the hippocampus. The hippocampus is an intricate brain region within the
medial temporal lobe. It has many sub-regions, best described as the cornu ammonis
subfields, the dentate gyrus region, and the subicular complex
92,93
. The subfields
include the CA1, CA2, CA3 and CA4. The subicular complex includes the presubiculum,
parasubiculum, and subiculum. And the dentate gyrus region includes the molecular
8
layer, granule cell layer, and polymorphic layer. Altogether, these sub-regions compose
the hippocampal formation.
Fortunately, the hippocampus is a rather homogenous structure to other species,
which has furthered our knowledge of its elaborate composition
92
. Primate and rodent
studies have been informative to the functions and structural integrity of the
hippocampus. Additional to animal studies, have been the case studies of patients with
severe hippocampal damage
77,94
. The most classic case being of HM (Henry
Molaison), an individual with such severe epilepsy, the only practical treatment, was to
remove his medial temporal lobe, where the seizures were evolving from
94,95
. While his
seizures improved remarkably, he was left with several memory deficits, including the
ability to remember if he had finished a meal.
Other important methods for studying the hippocampus include the use of
neuroimaging techniques. These include structural and functional measures
96
. While
the focus of this dissertation will be on the use of structural methods to assess the
hippocampus, a brief introduction into functional imaging techniques will be
summarized. The two main functional imaging techniques utilized via magnetic
resonance imaging, include, blood oxygen-level dependent (BOLD) response and
arterial spin labeling (ASL)
96
. Another field of study includes the use of positron
emission tomography (PET) imaging. But this technique is mostly used for clinical
purposes, due to the fact that radioactive isotopes are injected into the individual in
order to assess brain metabolism and therefore function. There are limited studies using
this technique in children, given the inherent risk involved.
9
The BOLD technique is common when assessing brain function concurrently with
the introduction of a stimulus i.e. a functional task. This task can include the introduction
of some sort of stimulus such as images of food, and then measuring the brain’s
hemodynamic response to this stimulus
96
.
ASL as a functional method is commonly used to measure the change in cerebral
blood flow from a baseline state to an experimental state
96
. Or rather, how cerebral
blood flow changes following the introduction of a stimulus. The Page lab uses ASL to
measure changes in the brain’s response to glucose ingestion. Children’s resting blood
flow is measured via an ASL sequence, then the children are given a glucose drink to
consume, and their brain’s blood flow is once again measured via an ASL sequence.
The change from before the drink is consumed compared to after, is used to derive how
certain regions are responding to a glucose load.
While both of these functional techniques provide key information about the
functionality of the brain, structural information is commonly derived from magnetic
resonance imaging sequences known as T2-weighted image, T1-weighted image and
diffusion tensor imaging (DTI)
96
. The T1-weighted image is used to quantify grey matter
volume of brain regions such as the hippocampus. T2-weighted images and DTI can
provide important information about other aspects of brain structure. But because
hippocampal structure will be the focus of this dissertation, the technique used to derive
grey matter volume of the hippocampus will be discussed. This technique utilizes the
T1-weighted image.
The T1-weighted image is collected during a magnetic resonance imaging scan
to give an anatomical outline of the brain. It is also used to differentiate gray and white
10
matter using contrasting intensity values from the image. T1-weighted images are
composed of voxels rather than pixels.
Different methods can be used to identify anatomical landmarks, either using
manual segmentation or automatic segmentation software. Given the large sample size
of the cohort utilized in this study, automatic segmentation software was used. There
are many different types of automatic segmentation software, but a software shown to
be reliable for the hippocampus and the sub-regions of the hippocampus includes
FreeSurfer 6.0 (http://surfer.nmr.mgh.harvard.edu/). How FreeSurfer works is the T1-
weighted image is inputted into the FreeSurfer software, and processed, then an output
with the grey matter volumes of the hippocampus and hippocampal sub-regions is
created
97,98
. The processing steps include first creating a reconstruction of all the
subcortical regions and differentiating them from cortical regions. This is done by first
removing all non-brain tissue from the T1-weighted image such as skull, dura mater and
the neck, this step is known as skull stripping. The next step is known as volumetric
labeling, which is the labeling of the subcortical regions which occurs using a
probabilistic atlas created by FreeSurfer that allows for identification of regional
boundaries. The probabilistic atlas is used on a voxel by voxel basis. After each
subcortical region is identified, intensity normalization is completed, this step involves
differentiating between grey and white matter based on intensity values. The white
matter is then separated from the image. Then the output of the subcortical regions with
grey matter values in millimeter (mm
3
) is created. This file is known as aseg.stats (see
Figure 2 for summary). The aseg.stats file also contains measurements such as
intracranial volume, and total grey matter. Intracranial volume (ICV) is the complete
11
measure of brain volume. It consists of grey matter, white matter and cerebral spinal
fluid. ICV is used to control for individual differences in brain volume.
After the subcortical regions have been processed, an additional reconstruction
flag in FreeSurfer is used to process the sub-regions of the hippocampus. These sub-
regions include the presubiculum, CA1, CA2/3, CA4, subiculum, dentate gyrus, tail of
the hippocampus, hippocampal fissure and the fimbria. The sub-regions are identified
during this processing step using an atlas created by FreeSurfer that can properly
differentiate hippocampal sub-region borders. The atlas was generated from 15 post-
mortem brain samples at 0.13 mm isotropic resolution and also from an in vivo dataset
at 1 mm resolution. After the regions are identified, algorithms based on Bayesian
inference are used to quantify grey matter volume in the various sub-regions of the
hippocampus. The output created after the hippocampal subfield segmentation
procedure is then available for the left hippocampus and the right hippocampus in text
files that contain the grey matter volume of the various sub-regions. Additionally, an
image is created that contains the hippocampal sub-regions distinguished as different
colors (Figure 3). This image consists of a series of slices through the brain with
hippocampal sub-regions.
In order to ensure proper segmentation was achieved, quality checks of each
participant’s hippocampal subfields and hippocampal formation was done, using the
hippocampal subfield reconstruction and the subcortical reconstruction images overlaid
each participant’s skull stripped T1 image. The quality of each image was then rated
using the technique described by Backhausen and Herting et al.
99
and also in the
methods section. Before quality checks were completed on the reconstructions, I was
12
extensively trained in hippocampal anatomy by Dr. Megan Herting. With her assistance,
I properly learned how to visually assess the hippocampus on a T1-weighted image, and
I also learned how to identify when the hippocampus was improperly segmented by
FreeSurfer. Additionally, she taught me how to rate the reconstructions to better ensure
quality data was used in the analyses completed for this dissertation.
To further distinguish FreeSurfer methodology, FreeSurfer has been shown to be
as reliable as manual segmentation methods, while proving to be more efficient
100–103
.
Based on advances in identifying hippocampal sub-region boundaries with improved
accuracy by Iglesias et al.
104
, FreeSurfer 6.0 is commonly being used to assess the
hippocampus. Additionally, FreeSurfer has been shown to accurately segment the
hippocampus and subfields of the hippocampus in children aged 7 to 11 years
105,106
.
It is important to note, there are potential limitations in using FreeSurfer 6.0. Such
as, manual tracing of brain regions to quantify grey matter is considered the gold
standard. Consequently, at this time, there are limited studies validating FreeSurfer
hippocampal subfield segmentation with manual tracing methods
107
. However, Schmidt
et al.
102
did recently find that FreeSurfer 6.0 is a scientifically comparable method to
manual tracing for the hippocampus and hippocampal subfields
102
. To ensure validity,
they did recommend manually reviewing the output. Additionally, as mentioned before,
the methodology created by FreeSurfer to classify the hippocampal sub-regions is
based on an atlas created from ex vivo images with 0.13 isotropic resolution. This atlas
helps define clearer hippocampal sub-regions boundaries. To conclude, due to the
complex boundaries of the hippocampus, FreeSurfer 6.0 automated hippocampal
13
segmentation, is a useful technique for accuracy and practicality and therefore was
used to for this dissertation research.
There may be an important question on the reader’s mind if they are unfamiliar
with why grey matter volume is measured in a region of interest. To begin to answer this
question, it is important that the reader understands the different types of tissue present
in the brain. As mentioned briefly before, the brain is composed of grey and white
matter tissue
96
. White matter tissue consists of bundles of axons that connect brain
regions together. Axons are long fibers covered in myelin sheath; a type of fatty tissue
made up of glial cells. Their function is to relay messages across the brain through
electrical impulses.
Grey matter tissue generally consists of a specific type of cell, known as neurons.
Neurons contain a cell body and dendrites
96
. When the brain has endured trauma,
there may be increased rates of neuroapoptosis (neuronal death). Or, if the trauma
occurs while the brain is developing, there may be lower levels of neurogenesis, the
creation of neurons. This is why grey matter is often studied, to see if there is either a
loss of neurons, or less neurons that were created. For example, it is well known that
babies born extremely premature, have less grey matter present following birth
108,109
.
This reduction in grey matter can even persist through childhood
110
. Further, some
studies have found that this decrease in grey matter is associated with reduced
cognitive function
110–112
. On the opposite side of the spectrum, exercise intervention
studies have found aerobic exercise can increase grey matter volume in the
hippocampus
68,113
. Various antidepressants also promote neurogenesis in the
14
hippocampus and therefore lead to increased grey matter volume following the use of
antidepressants
114–116
.
To summarize, measuring grey matter volume is theoretically an indirect way to
assess neuronal volume. It is also in vivo method to infer if levels of neuroapoptosis or
neurogenesis differ due to some experimental factor (i.e., premature birth, increased
levels of physical activity or use of antidepressants).
What can’t we do without the hippocampus: An overview of hippocampal function.
The hippocampus is a part of the brain that has a role in memory, spatial navigation,
learning, emotional processing and even appetite regulation. These brain functions are
all aspects of cognition, more specifically, intellectual thinking that involves information
processing in some form. While it would be impossible to detail the many aspects of
cognition that involve the hippocampus, a few well known, and well researched aspects
will be introduced. I will first briefly summarize the well-known aspects of cognition that
the hippocampus contributes to. And then, I will introduce cognitive functions that are
considered hippocampal-dependent.
Learning
There are many brain regions and cognitive processes that contribute to the process of
learning. The concept of learning itself is a broad term that can encompass reward-
based learning, semantic learning, or even learning in an academic setting. The
15
definition used here will consist of applying one’s experience to instigate a change in
behavior. As Benjamin Franklin once said, “Tell me and I forget, teach me and I may
remember, involve me and I learn”. The concept here, is that learning results in a
retention of information that can later be recalled in order to guide behavior. Learning
goes beyond memory processing, however memory most assuredly plays a role.
Molecularly, there has been a large emphasis on the role that the neurotrophic
factor, brain-derived neurotrophic factor (BDNF) has in the process of learning
117
.
BDNF is largely found in the hippocampus
117
. It also contributes to the process known
as long-term potentiation which is a molecular process that occurs at the synapse of a
neuron as a part of synaptic plasticity
118
. Synaptic plasticity is critical to consolidating
information during learning and memory
118
. For example, in a study by Mu et al.
119
using rats, BDNF production was blocked and as a result, LTP was limited.
Consequently, the rats were significantly impaired in learning a spatial memory task.
Further, they had significant memory deficits on the spatial memory task. Several other
animal studies have found similar results, blocking BDNF, reduces LTP as well as
learning
117
.
In humans, a handful of studies have found that aerobic exercise is associated
with improved learning, and is associated with increased volume in the hippocampus
113,120,121
. Interestingly, animal studies have found that increased neurogenesis and
BDNF levels in the dentate gyrus of the hippocampus after exercise, is associated with
improved learning of a hippocampal-related memory task
122–124
. In contrast, factors
known to negatively influence hippocampal neurogenesis such as prenatal stress lead
to a reduction in learning following a reduction in hippocampal neurogenesis
125,126
. In
16
humans, a reduction in hippocampal volume and learning a memory task has been
reported in individuals with post-traumatic stress disorder
127–130
. Collectively, there is
evidence that the hippocampus has an important role in learning. However, there are
many other brain regions that contribute to learning such as regions in the striatum and
prefrontal cortex. In a study by Pasupathy and Miller
131
, monkeys that performed an
associative learning task had activation in both the dorsal prefrontal cortex and caudate
nucleus during different phases of the task. Additionally, it is well known that aging can
affect hippocampal integrity but interestingly, learning specific tasks that are more
focused on fronto-striatal regions is unaffected in individuals with evidence of
hippocampal atrophy
132
. Therefore, while the hippocampus contributes to aspects of
learning, there are other brain regions that play an important role in the process of
learning.
Emotion Regulation
Emotion regulation is the cognitive process that occurs when an individual encounters
emotional cues in their environment and regulates their emotional response
133,134
.
Often, emotional dysregulation arises when an individual is unable to control their
emotional response, such is the case in various affective disorders such as major
depressive disorder, post-traumatic stress disorder, bipolar disorder, anxiety-based
disorders and even schizophrenia
135
. These are but a few examples of disorders
characterized by emotion dysregulation. While another region in the temporal lobe
contributes mostly to emotion regulation (the amygdala), it is also the prefrontal cortex
17
(PFC) and the hippocampus that contribute to the process of an individual regulating
their emotions
133,135
.
Historically, the hippocampus has been known for its role in various memory
processes, but animal research has helped elucidate a duality of the hippocampus that
contributes to both memory and emotions
135–137
. The posterior hippocampus, or in
rodents, the dorsal hippocampus, is best known for contributing to memory performance
and learning, and other cognitive processes
136,138
. The anterior hippocampus, which is
the ventral hippocampus in rodents, contributes to emotion regulation, and fear-based
responses
136,138,139
. Selective damage to the ventral hippocampus in rodents results in
decreased anxiety in situations previously shown to elicit anxiety, decreased fear
response and decreased cortisol secretion
136
. Interestingly, memory performance is
intact, while selective damage to the dorsal hippocampus results in memory
impairments
136
.
In humans, it is more challenging to recapitulate selective damage to the anterior
compared to the posterior hippocampus. However, hippocampal atrophy is often
associated with affective disorders where emotion dysregulation is the primary symptom
such as observed in post-traumatic stress disorder (PTSD), major depressive disorder
(MDD), bipolar disorder (BD) and anxiety-based disorders
129,140,141
. MDD is a classic
example of emotion dysregulation in which the hippocampus is thought to contribute to.
MDD is defined as a relentless sadness, and hopelessness with a host of
accompanying debilitating symptoms such as fatigue, decreased/increased appetite,
insomnia and anhedonia
142
. This emotional dysregulation can be improved via the use
of selective serotonin reuptake inhibitors, thought to target brain regions such as the
18
hippocampus
115,116,143
. For example, in a study of patients with MDD, larger
hippocampal volume was observed in patients who were responsive to SSRI’s
compared to healthy controls, while patients with treatment-resistant MDD had smaller
hippocampal volumes
116
.
Collectively, there is evidence that the hippocampus contributes to emotion
regulation, however several studies have suggested that the hippocampus is only a part
of a network of regions that controls emotion regulation, which include the amygdala
and the prefrontal cortex
133–135
. For example, in a study of amnesiac patients with
selective hippocampal damage, greater recall of emotionally-based memories
compared to neutral memories was similar between patients and controls
144
. This
finding suggests that, processing of emotional cues is not entirely dependent on the
hippocampus. Further, some researchers have found that impairments in the
connections between the hippocampus and other brain regions such as the amygdala
and PFC is associated with emotion dysregulation
134,135
. For example, another affective
disorder hallmarked by emotion dysregulation is schizophrenia
135,141
. One interesting
study found that, reduced white matter integrity between the prefrontal cortex, and
hippocampus was not only present in schizophrenic patients, but also their healthy
siblings. Consequently, compared to the general population, siblings of schizophrenic
patients are at a 9-fold increased risk for developing schizophrenia. Other studies have
found in schizophrenic patients, functional responses in the tracts that connect the PFC
and hippocampus are also impaired during a working memory task as well as when the
participant is in default mode
135,145,146
. Altogether, these studies provide evidence that it
19
is not only the hippocampus that contributes to emotion regulation, but also the PFC
and the amygdala as well as the connections between these regions.
Appetite Regulation
A lesser known but important function that the hippocampus contributes to is, appetite
regulation. For example, humans with hippocampal damage are more likely to consume
a second or even third meal close to their previously consumed meal mainly because,
they forgot when the last time they ate was and how much they ate
74,147
. One study
found, that healthy participants who recall their previous meal are more likely to eat
smaller quantities for their next meal
148
. Because the hippocampus is important for
episodic memory, or rather the memory of recent events, recalling previous meals is
important for food intake behavior and thus contributes to appetite regulation
149
. To
elaborate, imagine you have had a busy day and realize by 4 in the afternoon, that you
forgot to eat lunch. Now, this information may influence your next meal choice, because
your hunger may be greater based on the information that you have just now realized
you have consumed inadequate meals for the day. Because of the hippocampus, you
are able to recall that you did not eat lunch. There may be other physiological hunger
cues, such as a growling stomach, a burning headache or a slight fatigue. But it is the
higher order thinking process that allows you to recall that you have no memory of
eating lunch. Given that it is unlikely, you succumbed to hippocampal damage and
became an amnesiac for your lunch hour, you can conclude, your hippocampus
20
consolidated the day’s events and properly determined, there was no lunch hour and
that is why your stomach is growling.
There is another important aspect to the role of the hippocampus in appetite
regulation. It is not just the recall of past foods; it is the associations between our
environment and our internal hunger cues that can influence appetite regulation. This is
another important function of the hippocampus, known as relational memory, or the
forming of associations between two previously unrelated items. Referring back to the
example of skipping lunch, it is the association between an internal cue, and a memory
that helps formulate the realization that a meal was not consumed. Further, external
cues may also form an association with meal consumption, such as, the smell of food
that arises around noon and the association of it being lunchtime, may be enough to
instigate meal consumption. The hippocampus has many connections to olfactory
regions, that play a role in olfactory-based cues such as taste, and smell related to food
149
.
The concept of associations being important for appetite regulation can also be
exemplified through the viewing of food-cues. One study found that children who
watched unhealthy food cues in a controlled setting, were more likely to consume more
food following the viewing of the unhealthy food cues compared to children who
watched neutral cues
150
. The children were primed to eat more due to the viewing of
food-related stimuli. Or rather, their hippocampus formed an association between food
intake and the food-related cues and therefore consumed more food. Along this line of
research, several neuroimaging studies have found that the hippocampus is activated
during the viewing of food cues
151–153
.
21
Additionally, there are other known regions that contribute to appetite regulation.
These regions include the hypothalamus, the prefrontal cortex, reward regions such as
the nucleus accumbens and the ventral tegmental area, see review by Ferrario et al.
154
for more information. These regions are also frequently activated during the viewing of
food cue images
151,152,155
. Moreover, one study found that, individuals with hippocampal
damage still experience sensory specific satiety, an internal cognitive process, where an
individual reports decreased pleasantness as the meal is consumed
156,157
. Sensory-
specific satiety is important for appetite regulation because the desire to continue eating
the same food, should diminish after a meal is consumed, as the individual is satiated.
Without sensory-specific satiety, an individual may continue to eat that same food in a
larger quantity. Interestingly, the hippocampal-damaged participants reported no
change in hunger and fullness after consuming a meal, while their report of
pleasantness of the food, decreased as the meal was consumed, which suggests
another brain region contributes to sensory-specific satiety
156
. This study provides
evidence that an intact mechanism related to appetite regulation exists in the brain of
individuals with hippocampal damage, while also illustrating specific deficits related to
appetite regulation are present as well.
Similarly, animals with hippocampal damage/manipulation, also experience
changes to appetite regulation. Several studies by Dr. Scott Kanoski and colleagues
have illuminated the increasing role that the hippocampus has in appetite regulation
through his articulate mechanism-based studies. One study for example, showed that
stimulating glucagon-like peptide-1 (GLP-1) receptors in the hippocampus of rodents,
resulted in decreased food intake and weight loss
158
. While other rodent studies have
22
found, manipulating ghrelin receptors (growth hormone secretagogue receptor) in the
hippocampus can increase frequency of eating, along with the quantity of food eaten
149
.
Importantly, other regions where ghrelin related receptors and GLP-1 receptors are
manipulated such as in the hypothalamus, can have similar appetite regulation related
behaviors
149,158,159
. The hippocampus is a region with many connections to other brain
regions involved in appetite regulation, such as the hypothalamus, the amygdala, and
prefrontal regions
149
. Therefore, this network of regions contributes to appetite
regulation in various ways, but importantly the hippocampus has a pivotal role in
aspects of appetite regulation.
Hippocampal-Dependent Memory
While the hippocampus is broadly involved in many cognitive processes, one thing that
we have learned from patients with hippocampal damage, is certain types of memory
processes are non-existent by the ablation of the hippocampus. This notion is also clear
in animal literature, where controlled removal of the hippocampus is feasible
160
. The
functions that substantial research supports to be dependent on the hippocampus
include: aspects of declarative memory
161,162
, aspects of spatial memory
161
, and most
notably, relational memory
77,163
. It is important to note, declarative memory and spatial
memory also require other brain regions such as prefrontal and medial temporal lobe
structures. Therefore, while aspects of declarative memory and spatial memory are
hippocampal-dependent, in general, these functions are considered related to
23
hippocampal function rather than dependent entirely on the hippocampus. This concept
will be elaborated on shortly.
Declarative Memory
Declarative memory is a broad term for memory that is essentially the ability to recall
events and facts learned in the past. The role of the hippocampus in declarative
memory has been well researched from the time of the patient HM
94,160,161
. The most
obvious impairment he suffered was an impairment in declarative memory. There are
two branches of declarative memory, episodic memory and semantic memory (See
Figure 4 for summary). For this dissertation, episodic memory will be the focus.
Episodic memory is the ability to recall memories of objects, places and experiences
from one’s past. An example of a task, that tests episodic memory, was actually created
by the National Institutes of Health, as a part of an extensive cognitive battery known
NIH toolbox. In this task, participants are shown a series of pictures that tell a story. The
participant is shown the order of the pictures, and then the pictures disappear and
reappear in random order and the participant is instructed to reconstruct the series of
pictures they were previously shown. This task is known as the picture-sequence
memory task and will be one of the cognitive tasks that will be briefly mentioned in the
coming chapters of this dissertation. One aspect of episodic memory is autobiographical
memory. Autobiographical memory is the ability to recall specifically, one’s personal
experiences. Interestingly, in patients with hippocampal damage, aspects of
autobiographical memory are unaffected. For example, a patient may not be able to
24
recall when their last meal was, but they might be able to recall their childhood home or
a distant memory of a Thanksgiving meal
161
.
Recalling detailed autobiographical memories is considered hippocampal-
dependent
161
. For example, a study by Kurczek et al.
164
, found that patients with
bilateral hippocampal lesions were not only impaired in their ability to reconstruct past
events, but also could not accurately recall details about past events. In another study
by Viard et al.
165
, when healthy participants recalled details about a past event during a
functional MRI, whether the detail was related to a recent or remote memories, the
hippocampus continuously was activated. Further, successful recall of details about a
past event prior to the scan predicted hippocampal activation during the functional MRI
task. Other studies have found that lesions specific to the hippocampus result in a
retrograde amnesia that prevents recall of details of autobiographical memories as well
as semantic memories from the past three years
162
. To summarize, there is substantial
evidence that recalling details about past personal events appears to be a hippocampal-
dependent function.
Spatial Memory
Another aspect of memory thought to strongly involve the hippocampus is spatial
memory. Since the discovery of place cells in the hippocampus and their role in the
creation of cognitive maps of an environment, the hippocampus has strongly been
implicated in spatial memory
161,166
. A rather simplified definition of spatial memory is,
the consolidation of an environment based on cues in the environment in order to locate
25
something or navigate the environment. The ability of an organism to navigate its
environment has proven to be an important component to spatial navigation
167
. Some
studies have found that navigating an environment based on recently learned spatial
cues is considered to be hippocampal-dependent when the cue involves spatial
reasoning or picturing one’s self
162,168
. Or rather, the aspect of spatial memory that is
impaired following hippocampal damage, is the ability to navigate an environment based
on recently learned spatial cues that involve some sort of spatial reasoning or one’s
personal memory of a place. For example, in a study with a patient with bilateral
hippocampal lesions, the patient’s ability to navigate his environment using roads not
frequently used, was severely compromised
169
. The patient found himself lost on
several accounts when he had to navigate using roads that were not frequently used in
his route of being a taxi driver
169
. An interesting finding here, was that on the main
roads, which are well known and frequently used, he was unimpaired. The authors
hypothesized that some roads became procedural over time, therefore eventually not
being hippocampal-dependent. But, recalling routes not frequently used, still is
considered hippocampal-dependent because the driver has to use learned cues to
navigate his surroundings as opposed to a subconscious procedural memory.
Another study by Broadbent et al.
170
, found that rats with as much of 70% of
their hippocampus intact, still exhibited deficits on a spatial memory task, taking
significantly longer to find the platform in a pool of water. These same rats however had
no deficits on a recognition memory task, highlighting the importance of the
hippocampus in navigating one’s environment. Another study in rats with lesioned
hippocampi, found that across all time intervals spanning 12 months, rats with lesions
26
were significantly impaired on a spatial memory task that required learning relational
cues
171
.However, while spatial memory involves the hippocampus, other brain regions
contribute to intact spatial memory such as the parahippocampal gyrus
162
. There are
still obvious aspects of spatial memory that seem to rely upon the hippocampus, given
the deficits observed in humans with hippocampal lesions and rats with more localized
hippocampal lesions. This suggests the hippocampus has a fundamental role in
consolidating information, or rather creating relations of disparate elements for recall
later. This important concept will be introduced shortly.
Relational Memory
Several lines of research suggest that the most relevant function of the hippocampus is
relational memory
77,163,168,171
. Relational memory is the binding of two disparate
elements to form a rememberable association. These two elements can be a person
and a place, or the temporal order of objects seen or a location with a landmark. It is
arguably the premise of all memory functions, given the foundational nature of relational
memory
77,163,168,171–173
.
Extensive work by Dr. Neal Cohen, has elucidated the critical role the
hippocampus has in relational memory
174–177
. To start with, his work with patients with
hippocampal damage and reported memory loss, found that in a relational memory task,
performance is significantly compromised
172,177
. Further, patients with hippocampal
damage were unable to complete certain aspects of the relational memory task. The
patients who completed this relational task, consisted of three patients who had a
27
traumatic brain injury that specifically affected their hippocampus and/or the medial
temporal lobe. Two patients had prior episodes of anoxia, that resulted in bilateral
hippocampal atrophy. The last patient experienced an episode of herpes simplex
encephalitis that left the individual with severe damage to the medial temporal lobe,
bilaterally.
The hippocampal-dependent relational memory task that was used, is also
known as a visuospatial task. Participants are shown a series of items that appear in a
specific pattern. The items are then removed, and the participant is instructed to
reconstruct the image based on pattern of items previously seen. The number of items
shown, are slowly increased, i.e. two items, then four items. The task is then scored
based on a composite score that has four components. The first being, how many times
two item’s locations were swapped with each other, this is also known as the swap rate.
The second being, the distance of each item from its original location. The third is the
distance between each object to another, and lastly is the participant’s ability
reconstruct the overall shape of the items, i.e. a circle or a square configuration (see
Figure 5 for example of a child’s version of task). Patients with hippocampal damage
scored significantly worse on the composite score than healthy, matched controls and
also had significantly greater swap errors. No matter how many items were shown (i.e.
two items), the patients swapped the two item’s locations. Moreover, patients made 40
times more swap errors than control participants who had very low swap rate.
Based upon their findings, the authors hypothesized, that these increased swap
errors were due to the brain’s inability to properly create relations among the different
items, therefore frequently swapping item’s locations. Or in other words, if the purple
28
monster was below the striped orange monster, they were severely compromised in
their ability to recall this relationship, and therefore would put the purple monster into
the striped orange monster’s location and visa versa. To summarize, in patients with
hippocampal damage, the patients’ performance on a hippocampal-dependent memory
task that assesses relational memory was severely compromised, and they were unable
to perform the task without making swap errors. This same task was the foundation for
a child’s version, that was used for the studies that will soon be described in the
research portion of this dissertation.
In addition to studies in patients with severe hippocampal damage, various
groups have found that the hippocampus is strongly activated during a functional MRI
scan where a relational memory task is used to elicit a response
173,178,179
. Preston et al.
178
found that during the trials where the participant was required to recall formed
relations among two elements, the hippocampus had a much greater activation
compared to trials where recognition memory was assessed. While functional imaging
studies exemplify the role of the hippocampus in relational memory, structural imaging
studies have found that hippocampal grey matter volume is strongly associated with
performance on a relational task
163,180
. For example, in a study by Monti et al.
180
, older
adult participants with a history of mild traumatic brain injury performed significantly
worse on a relational task, and their performance was associated with reduced
hippocampal volume.
Along this line of research, several studies have found that the hippocampus is a
region of the brain that is exceptionally sensitive to environmental insults, and often the
hippocampus is one of the first regions to undergo atrophy during healthy aging
29
47,128,180–183
. Collectively, there is evidence that even a mild brain injury can result in
atrophy and reduced function of the hippocampus.
Now that the findings from clinical, functional and structural studies have
illustrated the critical role of the hippocampus in relational memory, developmental
evidence will be briefly introduced. The human hippocampus is one of the few brain
regions with protracted development
168
. While the subiculum complex, and the CA2 are
relatively formed by birth, other subfields continue to undergo development, specifically,
the CA3, and dentate gyrus (DG)
168
. Importantly, these subfields contribute to relational
memory, a memory function that begins to develop around year one, and matures
during childhood
168
.
Some researchers have suggested that it is the combination of connections
between the CA1, CA3 and DG subfields with other brain regions as well as the growth
of these subfields that contributes to more complex forms of relational memory during
childhood
168,184,185
. For example, in a study by Ribordy et al.
185
, children of varying
ages were asked to complete an allocentric spatial relational task that had different
levels of recall. Prizes were hidden in 18 locations, some with cues, and some without
cues. Children who were two years old, could only recall the location of up to three
prizes, and only locations with cues. Children who were 4-5 years old, were able to find
prizes for all 18 locations, and further could find prizes without previously learned cues.
Interestingly, from the ages of one to four years old, the CA3 and DG subfields undergo
rapid growth, almost paralleling the increase in complex relational memory function
168
.
By age four, the CA1 subfields is almost 100% of its adult volume, while the CA3 and
DG continue to develop past four years old
168
.
30
To summarize, there are some rather specific brain functions that depend on the
hippocampus, while other cognitive functions appear to build upon functions that are
dependent on the hippocampus. It is however evident, that the hippocampus has many
roles and is a part of several networks that contribute to successful brain functioning.
Moreover, it is the combination of connections between the hippocampus with other
brain regions, as well as hippocampal integrity that is vital to many aspects of brain
function.
Part Three: An Overview of Hippocampal Structural Development
The brain begins to develop as early as the first trimester of pregnancy
186
. First, the
stem cells meant to form the central and peripheral nervous system known as neural
progenitor cells, begin to form the neural tube. The neural tube eventually develops into
the central and peripheral nervous system. At about 9 weeks, the hippocampus begins
to form
187
. The first subregions that begin to develop are the CA1 and subiculum. The
CA2-4 begin to develop shortly thereafter. The dentate gyrus begins to develop during
week 15-19. The CA1, CA2 and subiculum are among the first few structures to mature
in neuronal characteristics, followed by the CA3-4 and dentate gyrus. By birth, most
subregions are relatively mature, with the exception being the dentate gyrus, and the
CA3 which have protracted growth
168
.
During infancy, hippocampal gray matter volume continues to increase, as does,
synaptic density
188
Additionally, various white matter pathways within the hippocampus
are strengthened via a mechanism known as myelination
101,187
. Connections between
31
the hippocampus and other structures are also strengthened. It is a time of learning and
increased functionality, as children learn concepts, language, and constructs that set
the foundation for the rest of their life
168
. By age 5, the brain has developed to 90% of
its full volume
188
. And at about 11 years of age in girls and 14 years in boys, the brain
has reach its total cerebral volume
188
. The hippocampus reaches its peak volume at
different time points based on the various subregions
93,101
. The exact timepoint at
which total hippocampal volume and each subregion reach its peak volume is still a
topic of speculation
93,101,188,189
. Studies by Tamnes et al.
93
and Krogsrud et al.
101
,
suggest that total hippocampal volume reaches its peak volume by early adolescence.
In contrast, Tamnes et al.
93
and Daugherty et al.
190
found that the CA2-4, and dentate
gyrus decrease in volume from early childhood to adulthood. While Tamnes et al.
93
and
Krogsrud et al.
101
found that the CA1 appears to increase with age until early
adolescence. The subiculum fluctuates in increases/decreases in volume in a non-linear
fashion as reported by Tamnes et al.
93
. Collectively, evidence suggests that different
subregions increase as well as decrease at different rates from childhood to
adolescence, however the exact pattern is still uncertain. Future research with similar
hippocampal segmentation protocols and study design are needed to find a general
consensus on growth trajectories of the various subfields.
In contrast, there is a general consensus that there are sex differences in total
hippocampal volume
93,191,192
. Boys on average have larger hippocampal volumes
93,191,192
. Another consistent finding, is that total hippocampal volume increases until
early adolescence, at which point hippocampal volume slowly begins to decrease in
volume
93,191–193
. The rate at which hippocampal volume increases and decreases can
32
be influenced by a host of factors such as in utero exposures, neuropsychiatric
conditions, metabolic conditions, chronic stress, traumatic brain injury, lifestyle factors
such as physical activity and diet, and many other factors which are beyond the scope
of this dissertation. Altogether, the hippocampus is a heterogenous structure from which
its subregions contribute to memory, learning, emotion regulation and even appetite
regulation. Hippocampal integrity is pivotal to these various cognitive processes. In the
coming chapters, the relationship between hippocampal-dependent memory function,
hippocampal volume and prenatal exposures will be assessed.
Chapter Two: Sex differences in the Association between Prenatal
Exposure to Maternal Obesity or GDM and Hippocampal Volume in
Children
Part One: Maternal Obesity
Introduction
Almost two thirds of women, that are of childbearing age in the U.S. are
overweight or obese
194
. Consequently, obesity rates for women in this age range,
continue to rise, posing risks for themselves, and their offspring
25
. As previously
mentioned, children exposed to maternal obesity in utero, are at increased risk for
obesity themselves, as well as at increased risk for developing neurobehavioral
disorders and cognitive impairments. While the focus of epidemiological research has
33
focused on cognition and increased rates of neurobehavioral disorders among offspring
exposed to maternal obesity in utero, there have been limited studies investigating
potential mechanisms in the brain, that contribute to these observed cognitive deficits
and neurobehavioral disorders. A breadth of animal literature suggests that the
hippocampus has an important role in the etiology of these observed deficits. Additional
studies in animals have observed offspring exposed to maternal obesity in utero, have
decreased memory performance and alterations to hippocampal structure. This chapter
introduction will summarize findings in animals, as well as lay the foundation for the
novelty of the study design found in this chapter.
It has long been known that an obesogenic environment in utero contributes to
changes in the brain. While the majority of research has focused on the hypothalamus,
a brain region pivotal for appetite regulation and body weight modulation, studies over
the past decade, have also included the hippocampus. Studies by Niculescu & Lupu,
62
and Tozuka et al.
8
, were among the first to find that prenatal exposure to maternal
obesity leads to altered hippocampal development in mice offspring. To elaborate, they
showed that offspring exposed to maternal obesity exhibited reductions in hippocampal
neurogenesis and increased apoptosis within the dentate gyrus hippocampal subfield
8,195
. Moreover, another study found, that offspring had decreased levels of hippocampal
brain-derived neurotrophic factor (BDNF), and had abnormal dendritic differentiation of
new hippocampal neurons during early postnatal development
63
. Numerous studies
have found that an obesogenic environment in utero increases systemic inflammation
and thus contributes to increased inflammation in the hippocampus
67,196–198
. The
hippocampus is not only sensitive to an altered metabolic environment, but is also
34
selectively vulnerable to in utero insults such as inflammation
67,199–201
. To elaborate,
prior work has shown that prenatal exposure to maternal obesity is associated with
increased microglial cells and inflammatory markers such as toll-like receptor 4, within
the hippocampus of offspring
67,199,202,203
. In addition to increased inflammation, another
study found that there were, sex-specific effects of in utero exposure to maternal obesity
such that, male, but not female, offspring had alterations in fatty acid metabolites within
the hippocampus and prefrontal cortex
204
. Collectively, there is a wealth of data in
animals, that demonstrate that prenatal exposure to maternal obesity promotes altered
hippocampal development in offspring. To date, there had been no studies in humans
examining if the hippocampus is altered in offspring exposed prenatally to maternal
obesity.
Many of the previously mentioned studies showing altered hippocampal
development in rodent offspring exposed to maternal obesity in utero have also found
that these offspring exhibited hippocampal-related memory deficits. For example, Zhu et
al.
204
, and Tozuka et al.
63
, found that mice offspring had altered hippocampal
development, and were also impaired on a spatial memory task. Interestingly, Bilbo et
al.
67
, also found that offspring were impaired on a spatial memory task and had
increased inflammation in the hippocampus. Similarly, another study found that
offspring were also impaired on a hippocampal-related memory task, but after an
exercise intervention, the rat offspring not only had increased volume in the dentate
gyrus subfield of the hippocampus but also had markedly improved memory
performance
68
. Altogether, there is substantial evidence in the animal literature that
35
hippocampal structure and function is altered due to in utero exposure to maternal
obesity.
To date, there have been limited in humans that have investigated the impact of
maternal obesity on offspring brain development
205,206
. One study found that various
white matter pathways were altered in newborn offspring exposed to maternal obesity in
utero
205
. It is yet to be determined if these alterations are present during childhood and
extend to the hippocampus. Our study is the first to investigate these potential
relationships. Importantly, no one has reported how hippocampal function may be
impaired in offspring exposed to maternal obesity in utero. Interestingly, several
epidemiological studies have found strong associations between cognitive performance
and maternal pre-pregnancy BMI
49,58,80,207,208
. Some of these studies have found,
these associations remain even after accounting for genetic and environmental factors.
For example, one study found that child cognitive performance was unrelated to
paternal BMI, but related to the mother’s pre-pregnancy BMI
52
. Another study found
that, the association between a child’s IQ and the mother’s pre-pregnancy BMI,
remained even after accounting for the sibling’s characteristics, i.e., mother’s pre-
pregnancy BMI with sibling
78
. While many of these studies have investigated more
global measures of cognition, I was interested if there were also more specific deficits
observed in offspring exposed to maternal obesity in utero. Because the hippocampus
is involved in many aspects of cognition, a better understanding of whether
hippocampal-specific deficits are present, is of much importance. Especially,
considering the animal literature that suggests exercise can ameliorate hippocampal
function following prenatal exposure to maternal obesity. Perhaps, higher levels of
36
physical activity may improve a child’s cognitive performance, and hippocampal function
following exposure to maternal obesity in utero. As previously mentioned, there are
many children whose mothers are obese prior to pregnancy. These children are already
at substantial risk for obesity themselves. This study will also elucidate if children
exposed to maternal obesity in utero, also face the risk for cognitive impairments, and
hippocampal-dependent memory deficits. It is worth noting, altered hippocampal
development may in turn lead to difficulties with academic performance, emotional
regulation and importantly cognitive performance. Therefore, this research will be an
important first step in determining if a maternal obesity impacts a brain region that has
so many pivotal roles. Using a hippocampal-dependent memory task, that has
previously been validated in patients with hippocampal damage, I assessed the
relationship between child memory performance, and maternal pre-pregnancy BMI.
Structural MRI was used to determine if maternal pre-pregnancy BMI is associated with
altered hippocampal structure in children. Given the growing number of children
exposed to maternal obesity in utero, this work provides critical information on the
effects on brain development and cognitive processing in offspring. Additionally, this
work could lead to a number of downstream paths, including intervention strategies to
prevent adverse health effects of maternal obesity on the developing brain.
Statistical Analysis
Hippocampal Measures
37
Exposure to maternal obesity in utero was indexed by maternal pre-pregnancy BMI
obtained from EMR. The primary outcome measures were child’s bilateral hippocampal
volume and the volume of each hippocampal subfield. Diagnostic analyses were
completed to assess if there were any influential points or potential outliers. Additionally,
visual assessments, and the feature, proc univariate in SAS, were used to determine if
variables were normally distributed. Apriori covariates to control for potential
confounding were child age, BMI z-score, total intracranial volume (ICV), gestational
diabetes mellitus (GDM) exposure obtained from EMR (dummy coded), and
socioeconomic status (SES), which included household income at birth estimated based
on census tract of residence, and maternal education at birth, which was extracted from
birth certificates in EMR. These covariates were previously shown to influence
hippocampal volume, thus were added to the model to control for potential confounding
101,209–213
. Since the majority (93%) of children were pre-pubertal (Tanner Stage <2),
Tanner Stage was not adjusted in the regression models. There was also no difference
in Tanner Stage categories between boys and girls (χ
2
=1.84, P=0.40). Further, Tanner
Stage was unrelated to hippocampal volume (R=-0.10, p=0.33).
Left and right hippocampal volumes were combined because correlations
between the left and right hippocampus were high (r=0.91, unadjusted P<0.001).
Relationships between maternal obesity and hippocampal volume were consistent
between the left (correlation coefficient r=-0.13, unadjusted P=0.25) and right
hippocampus (correlation coefficient r=-0.14, unadjusted P=0.19). First, linear
regression was used to assess if there was a significant interaction between child sex
and maternal pre-pregnancy BMI in the relationship with total hippocampal volume and
38
the hippocampal subfields. Based upon a significant interaction, multivariable linear
regression was used separately in boys and girls to investigate the relationship between
maternal pre-pregnancy BMI and child hippocampal volume and subfields. The linear
regression models were unadjusted models (Model 1), and models adjusting for ICV
(Model 2), child age (Model 3), and with additional adjustments for SES and exposure to
maternal GDM (Model 4), and child BMI z-score (Model 5). For total hippocampal
volume, a significance level of P <0.05 was used. To control for false discovery rate
(FDR) in the multiple comparisons of subfields, a FDR method based on the Benjamini-
Hochberg procedure was used to assess significance within each model
214
. Each p-
value was ranked and compared to the critical value with an overall false discovery rate
at 5%.
Cognitive Measures
The primary outcome measures were, accurate single item placement (relational
memory task), and accuracy (item-familiarity task). There were no outliers and the
distributions were approximately normally distributed. Multiple linear regression was
used to analyze relationships between cognitive measures and maternal-pre-pregnancy
BMI. Apriori determined covariates included age, sex, tanner stage, income and
maternal education at birth (Socioeconomic (SES) status at birth), GDM status, BMI z-
score and IQ. Sex and SES, were accounted for, because these are two variables
previously shown to influence hippocampal-dependent memory
215–217
. Additionally,
statistically significant Pearson’s correlations existed for IQ, Tanner stage, and child age
39
with accurate single-item placement (see correlation matrix Table 6). These same
covariates were also used to assess the relationship between maternal pre-pregnancy
BMI and episodic memory performance (picture-sequence memory task). Because of
sample size constraints, IQ was not used as a covariate for this analysis. Not all
participants completed both the picture-sequence memory task and IQ testing. The NIH
toolbox (picture-sequence memory task) has multiple scoring outputs to potentially use.
The scoring metric chosen for this dissertation was the fully corrected T-score. This
scoring metric takes into account the participant’s age, gender, race and the parent’s
education level. The scoring norm for these covariates (age, gender, race and the
parent’s education level) was generated from a previously sampled population, with a
mean of 50 and a standard deviation of 10. As an effect size measure, partial omega
2
,
was used to assess the overall effect of maternal pre-pregnancy BMI. Partial omega
2
was used because it is considered a less biased estimate than eta
2
, and also shows the
effect of a predictor variable on an outcome variable after accounting for the variance
attributed to the outcome variable due to confounding variables
218–221
. Interpretation of
partial omega
2
is as follows, 0 to 0.01 as small, 0.06 as moderate and 0.138 as large
218
.
SAS 9.4 statistical software (SAS Institute, Cary, NC USA) was used for all statistical
analyses with a p<0.05 significance level.
Results
Participants Characteristics
40
Of the 117 children enrolled into the study, 99 of these children completed MRI scans.
Ten children were excluded due to excessive motion and one was excluded due to
incidental brain findings leaving a total of 88 children in the MRI analyses. The mean ±
SD age was 8.4 ± 0.9 years old, 93% of the children were pre-pubertal (Tanner Stage
<2), and 58.0% were girls (Table 1). Maternal pre-pregnancy BMI ranged from 19.0 to
50.4 kg/m
2
. Overall, 22.7% of mothers were normal-weight, 36.4% were overweight and
40.9% were obese prior to pregnancy. Children’s BMI ranged from 13.6 to 34.0 kg/m
2
;
BMI percentiles ranged from 5.3 to 99.6; BMI z-scores ranged from -1.78 to 2.64. Based
on CDC standards, 61% of children were classified as healthy-weight, 15% were
classified as overweight, and 21% of children were classified as obese
222
.
Of the 117 children, 99 completed the cognitive tasks. Some of the participants
completed these tasks at their two-year visit and were on average slightly older. The
mean ± SD age was 8.9 ± 1.4 years old, 83% of the children were pre-pubertal (Tanner
Stage <2), and 60.0% were girls (Table 2). Maternal pre-pregnancy BMI ranged from
19.0 to 54.0 kg/m
2
, 25 (25%) were normal weight, 31 (31%) were overweight, and 46
(46%) were obese. Children’s BMI ranged from 13.9 to 34.0 kg/m
2
; BMI percentiles
ranged from 7.5 to 99.6. Their BMI z-scores ranged from -1.47 to 2.64 and 55 (56%) of
children were healthy-weight, 14 (14%) of children were overweight, and 30 (30%) were
obese.
41
Interaction between Maternal Pre-pregnancy BMI and Sex on Child Hippocampal
Volume
A significant interaction between pre-pregnancy BMI and sex on child hippocampal
volume was observed in unadjusted model (p<0.001) as well as after adjusting for child
and maternal covariates including child intracranial volume (ICV), age, socioeconomic
status (SES), maternal gestational diabetes mellitus (GDM) status and child BMI z-
score (p=0.024). Figure 6A depicts the scatter plot of the data for boys and girls for total
hippocampal volume. When stratified by sex, a negative relationship between maternal
pre-pregnancy BMI and hippocampal volume was observed in boys (correlation
coefficient r=-0.39, p=0.018) but not in girls (r=0.11, p=0.45). The significant negative
relationship in boys remained after adjusting for child and maternal covariates (b=-
126.98, p=0.01; Table 3). Maternal pre-pregnancy BMI had a large effect on the
hippocampal volume of boys with a partial omega
2
of 0.144.
Significant interactions were also observed in the CA1 (p=0.008), CA2/3
(p=0.016), CA4 (p=0.002), DG (p<0.001) and subiculum (p<0.001) hippocampal
subfields (unadjusted interaction test); all remained significant after applying Benjamini-
Hochberg procedure to control for false discovery rate (FDR) among the multiple
subfields. Figure 6B depicts the scatter plot of the data for boys and girls in each
subfield. When stratified by sex, a negative relationship between maternal pre-
pregnancy BMI and the volume of CA1, CA2/3, CA4, and DG hippocampal subfields
was observed in boys (CA1: r=-0.36, p=0.029; CA2/3: r=-0.42, p=0.010; CA4: r=-0.37,
p=0.025; DG: r=-0.36, p=0.030) (Figure 6B; Table 3) but not in girls (CA1: r=0.11,
42
p=0.47; CA2/3: r=-0.06, p=0.66; CA4: r=-0.03, p=0.84; DG: r=-0.03, p=0.82) (Figure 6B;
Table 4). After adjusting child and maternal covariates, the negative relationship in boys
remained significant in the CA2/3 (b=-14.19, p=0.006), CA4 (b=-10.34, p=0.021), and
DG (b=-11.31, p=0.027) subfields (Table 3); all survived FDR correction. In both the
unadjusted and fully-adjusted models, maternal pre-pregnancy BMI had the largest
effect on the CA2/3 subfield. In the fully-adjusted model, maternal pre-pregnancy BMI
explained 17.8% of the variance in the CA2/3 volume of boys.
In unadjusted and unstratified analyses, maternal pre-pregnancy BMI was not
significantly associated with child’s hippocampal volume (b=-46.97, p=0.30) (Table 5).
However, in models adjusted for child ICV, age, sex, and the interaction between
maternal pre-pregnancy BMI and child sex, we observed a significant association
between maternal pre-pregnancy BMI and total hippocampal volume in children
(b=106.98, p=0.008).
Associations between pre-pregnancy BMI and Memory Performance
One of the many unique advantages of the BrainChild cohort, is the levels of memory
assessed using various cognitive assessments. Children completed a hippocampal-
dependent memory task (relational memory), a hippocampal-independent memory task
(item-familiarity) and lastly, a hippocampal-related memory task (episodic memory
assessed via the picture-sequence memory task). Mechanistically, we could assess if
potential deficits related to in utero exposure to maternal obesity were related to the
hippocampus, or more global. However, there were no statistically significant
43
associations or trending relationships between any of the memory tasks with maternal
pre-pregnancy BMI (Table 7). When stratified by sex, there were also no associations
between any of the memory tasks and maternal pre-pregnancy BMI (Tables 8-9). The
effect sizes further support the lack of association. Interestingly, in the fully-adjusted
model, full-scale IQ was marginally associated with maternal pre-pregnancy BMI
(Beta=-0.35, p=0.094) (Table 7). With a partial omega
2
of 0.018, pre-pregnancy BMI
explains 1.8% of the variance in full-scale IQ. After adjusting for age, sex, and BMI z-
score, maternal pre-pregnancy BMI had a moderate effect on child IQ, with a partial
omega
2
of 0.036. In girls, in the unadjusted model, pre-pregnancy BMI was also
marginally associated with full-scale IQ (Beta=-0.31, p=0.055; Table 8), and 4.6% of the
variance in full-scale IQ can be explained by pre-pregnancy BMI.
Relationships between Cognitive Assessments and Hippocampal Volume
Linear regression was used to determine relationships between performance on
cognitive measures and hippocampal volume as well as the hippocampal sub-regions.
To assess significance, a<.05 was used. The cognitive measures assessed include the
picture-sequence memory task (PSMT) (episodic memory), the iposition task (relational
memory, hippocampal-dependent), using accurate single item placement as a scoring
metric, and lastly, the creature-scene task (item-familiarity, hippocampal-independent),
using accuracy as a scoring metric. Sixty-six participants completed the iposition task
and the MRI visit. Eighty participants completed the PSMT and the MRI visit. And 59
participants completed the creature-scene task and the MRI visit.
After controlling for intracranial volume, there was no significant association
between relational memory and hippocampal volume or the CA1, CA2/3, CA4,
subiculum, and DG sub-regions (Table 10, P>0.10). However, in boys, there was a
significant association between relational memory performance and the subiculum
44
(b=0.006, P=0.012). In boys, there was a marginal association between relational
memory performance and the left hippocampal volume (b=0.0001, P=0.055). There was
no significant or trending relationship in girls.
After controlling for intracranial volume, there was a statistically significant
positive association between episodic memory and the left hippocampus (b=0.012,
P=0.040) (Table 11). There was also a statistically significant positive association
between episodic memory and the left CA4 (b=0.157, P=0.026) and the left CA2/3
(b=0.136, P=0.030). There was a marginal association between episodic memory and
the DG (b=0.114, P=0.062). In boys, there was a statistically significant positive
association between episodic memory and the right CA4 (b=0.303, P=0.038). In girls,
there was a marginal association between episodic memory the left hippocampus
(b=0.012, P=0.078). There was also a marginal association between episodic memory
and the CA2/3 (b=0.134, P=0.053), the left CA4 (b=0.137, P=0.091) and the DG
(b=0.125, P=0.077).
As hypothesized, the hippocampal-independent task (item-familiarity) was
unrelated to hippocampal volume or the hippocampal sub-regions (Table 12).
Discussion
To date, no study in humans has reported any associations between maternal
pre-pregnancy BMI and hippocampal-related memory. Interestingly, several studies in
animal models of in utero exposure to maternal obesity have observed, hippocampal
structural and functional deficits
8,62,63,67,68,198
. While there have been no reported
memory deficits in children exposed to maternal obesity, several epidemiological
studies have found associations between maternal obesity and child IQ, even after
accounting for confounding variables
49,52,78,81,223,224
. Similarly, there was a marginal
45
association between maternal pre-pregnancy BMI and child full-scale IQ in the
BrainChild cohort. However, there was no association between pre-pregnancy BMI and
memory performance.
Maternal pre-pregnancy BMI was unrelated to hippocampal-dependent memory,
hippocampal-independent memory, and hippocampal-related memory. Because of the
marginal association between pre-pregnancy BMI and full-scale IQ, it may appear that
in utero exposure to maternal obesity does not impact hippocampal function, but rather
global cognition. But it is important to keep in mind, our sample size is considerably
smaller than previous epidemiological studies. Therefore, while the goal of this
dissertation was to look for specific memory deficits, it is possible, deficits would only be
observed in a much larger sample size. For example, Casas
52
, found that in a large
cohort of 1507 children, children exposed to maternal obesity in utero had significantly
worse psychomotor development than unexposed children. This association remained
after adjustment for paternal BMI, and SES. Interestingly, they found that in a much
smaller cohort of 310 children, this same relationship was attenuated after controlling for
paternal BMI. This finding suggests, that the effects of maternal obesity in utero on
certain aspects of cognition, may only be observed in a larger sample size. Further, in a
recent study, with a sample size of 331, they did not observe any differences in
executive functioning unless children were exposed to mothers with pre-pregnancy
BMIs greater than 35
225
. So additionally, the impact of maternal obesity may also be
dose dependent, such that only extreme obesity or a much large sample size of
mothers with pre-pregnancy BMI greater than 30, would be associated with
hippocampal-related deficits.
46
While no memory deficits were observed, we observed a significant interaction of
pre-pregnancy BMI and sex on hippocampal volume, where only boys showed a
significant negative relationship between pre-pregnancy BMI and hippocampal volume.
These sex-specific effects were consistently observed across the various hippocampal
subfields. Our findings suggest that boys may be more vulnerable to maternal obesity
induced altered hippocampal development than girls. These findings are consistent with
studies in animals.
Moreover, a number of animal studies have revealed that prenatal exposure to
maternal obesity is linked to abnormal hippocampal development, including reductions
in neurogenesis, decreased levels of BDNF (a neurotrophin involved in neural
differential and survival), and abnormal dendritic differentiation of new neurons
8,62,63
.
Several lines of evidence in rodent models suggest potential mechanisms by which
maternal obesity influences the development of the fetal hippocampus. First, exposure
to maternal obesity may elevate levels of inflammation in the fetus, and the
hippocampus is vulnerable to a pro-inflammatory environment
67,198,199
. Prenatal
exposure to an aberrant inflammatory environment is associated with increased
microglial cells and inflammatory markers (i.e., toll-like receptor 4) within the
hippocampus of offspring
67,199,202,203
. Second, maternal obesity exposes the fetus to an
excessive nutrient supply, which may in turn stimulate fetal hyperinsulinemia
226,227
.
Fetal hyperinsulinemia has been shown to alter hippocampal development via impaired
insulin signaling and reduced neurogenesis in the hippocampus
228
. Notably, insulin
signaling in the hippocampus is important for both neurogenesis as well as learning and
memory
229–232
.
47
Interestingly, prior evidence suggests that the brain development of males may
be more susceptible to metabolic and environmental perturbations encountered in utero
when compared to female offspring
65,233–235
. For example, prenatal exposure to
maternal obesity resulted in alterations in fatty acid metabolites within the hippocampus
of male but not female offspring
204
. Studies investigating the role of prenatal exposures
to stress and alcohol have also found sex differences in the vulnerability of the
hippocampus to alterations in the intrauterine environment, with more pronounced
effects in males than females
234–236
. Moreover, male offspring exposed to obese
mothers in utero were found to have abnormal gene expression within the
hypothalamus and forebrain, whereas female offspring did not
65
. Motivated by these
compelling findings in animal studies
65,66,204,233
, we examined for the first time in
humans the sex-specific effects of exposure to maternal obesity in utero on
hippocampal development in children.
Our results are also in line with prior studies in humans that have observed sex
differences in the effects of prenatal exposures on brain development
1,236
. Evans et al.
237
showed that male fetuses of obese mothers have higher levels of reactive oxidative
species compared to female fetuses
237
. It is possible that girls may be more protected
by adverse events in utero due to the anti-inflammatory and neuroprotective properties
of estrogen
238–240
. Interestingly, boys are also more likely to be diagnosed with
neurodevelopmental disorders such as autism spectrum disorder (ASD) and attention
deficit hyper-activity disorder (ADHD)
56,241
, and epidemiological studies have found that
maternal obesity is associated with an increased risk for developing ASD and ADHD
54,56,59
. One study found that children diagnosed with ASD or ADHD were not only more
48
likely to be male but were also more likely to have been exposed to maternal obesity in
utero
56
. It is possible that the observed sex specific effects of prenatal exposure to
maternal obesity on hippocampal development in our study could be associated with
increased risk of neurodevelopmental disorders in boys compared with girls, however
future studies are needed to address this possibility.
Additionally, while we observed significantly reduced hippocampal volume in
boys exposed to maternal obesity in utero, these boys did not perform worse on any of
the hippocampal memory tasks. Moreover, the effect sizes of maternal obesity on child
IQ were similar for boys and girls, suggesting that there were only sex-specific
differences in hippocampal structure. While there were structural changes associated
with prenatal exposure to maternal obesity, it is evident that these changes did not
coincide with sex-specific functional deficits. Because childhood is an important time for
changes in the brain, from synaptic pruning, to myelination and increases in neuronal
density, longitudinal studies are needed to determine if hippocampal memory deficits
become apparent with age.
Another possibility to consider is that memory deficits may not arise due to the
highly plastic nature of the brain. Neural plasticity is a pivotal mechanism by which the
brain attempts to repair itself following injury by strengthening white matter pathways or
increasing neuronal volume in other brain regions needed for optimal brain function
242–
244
. Several studies have shown, that during early development, the brain is able to
remodel and modify to compensate for traumatic injuries
245,246
. For example, in a study
of children with a history of hypoxia/ischemic attack that preferentially impacted the
hippocampus, they were unimpaired in scene construction for imaginary scenarios, a
49
hippocampal function that is impaired in adults with the same severe hippocampal
damage
245
. More importantly, these children used alternative strategies to accurately
recall their previously imagined scenarios, which suggests, that during childhood, other
connections are formed to compensate for a potential deficit in hippocampal-dependent
function. A similar study in animals found that, following juvenile ischemia, hippocampal
long-term potentiation (LTP) and memory function were rescued after 30 days, and
were associated with increased BDNF expression
247
. Interestingly, hippocampal CA1
volume was not increased following the ischemic attack, thus suggesting hippocampal
connections are remodeled to compensate for neuronal injury. Further, in a study of
adults with post-traumatic stress disorder (PTSD), while they had reduced hippocampal
volume, and reduced cerebral blood flow in the hippocampus, their explicit memory
performance was no different compared to individuals without PTSD
248
. Collectively,
these studies illustrate the brain’s ability to reorganize and compensate for hippocampal
volume reductions. Additionally, there have also been several documented cases where
children’s brains undergo rapid changes following the surgical removal of part of their
hemispheres due to intractable epilepsy
243,244
. These same children impressively are
minimally impaired. To summarize, while reduced hippocampal volume was associated
with prenatal exposure to maternal obesity, other brain regions or white matter
connections may have been remodeled to compensate for potential memory deficits
given the highly plastic nature of the human brain.
Strengths and Limitations
50
While our study had many strengths, such as an objective measure of maternal pre-
pregnancy BMI assessed from electronic records and a thorough assessment of
memory function, there were some limitations. Due to our study design, we were unable
to investigate potential underlying mechanisms by which maternal obesity affects fetal
hippocampal development. Further, due to the cross-sectional nature of this study, it is
unknown whether the relationship observed predominately in boys will persist
throughout life. Because the growth trajectory for peak hippocampal volume in girls can
occur at an earlier age than boys, it is possible that we are observing a delay in brain
maturation in boys and missed the relationship in girls
101,188
. Future studies should
consider a longitudinal assessment to determine if this relationship is due to delayed
maturation or if reduced hippocampal volume persists among children exposed to
maternal obesity. Additionally, it is possible, that at this timepoint, we did not capture
hippocampal-memory deficits, and memory deficits may become more apparent with
age. For example, a study in animals found that that, hippocampal volume loss
preceded memory deficits
249
. While volume loss was evident at three days, memory
deficits were not apparent until weeks later.
Lastly, although manual tracing is considered the gold standard for hippocampal
segmentation, recent advances in automated segmentation methods, such as the newly
developed FreeSurfer 6.0, provide a rigorous and practical method to quantify
hippocampal volume overall and within each subfield
100,102,103
. FreeSurfer 6.0 has
increased accuracy and reliability making it comparable to manual segmentation while
being more efficient
98,100,102,103
. However, the anatomically defined boundaries created
51
by FreeSurfer 6.0’s ultra high-resolution scan may be more precise than captured by
our high resolution T1 image.
Conclusion
In summary, we found a significant interaction of maternal pre-pregnancy BMI
and sex on child hippocampal volume. Boys but not girls showed a significant negative
correlation between pre-pregnancy BMI and hippocampal volume, suggesting that boys
may be more vulnerable than girls to the adverse effects of exposure to maternal
obesity on hippocampal development. These results call for more attention to
considering sex differences on the effects of prenatal exposure to maternal obesity on
brain development during childhood. Additionally, we found that prenatal exposure to
maternal obesity was not associated with memory deficits at this age but was
associated with reduced IQ scores. Future longitudinal studies are needed to determine
the functional consequences of reduced hippocampal volume in children exposed to
maternal obesity in utero.
Part Two: Gestational Diabetes
Introduction
As previously mentioned, in utero exposure to GDM is associated with various
metabolic consequences. Additional studies suggest that exposure to GDM in utero has
52
adverse effects on fetal brain development leading to increased risk for cognitive
impairments and neurobehavioral disorders during infancy and childhood
50,61,83,89,90
.
Prior studies in animals have found that the hippocampus is preferentially impacted by
in utero exposure to GDM
85,86,250–252
.
The hippocampus is a brain region of interest for studying the effects of GDM,
because the hippocampus has been previously shown to be sensitive to in utero insults,
particularly fluctuations in circulating glucose levels
182,253
. Additionally, the
hippocampus has a high metabolic demand which may place it at risk due to the
fluctuations in glucose and other important nutrients such as iron
90
. Further, the
hippocampus is a brain region with projected growth compared to other brain regions
and continues to develop even after birth
181,254
. Consequently, during GDM, the fetus is
exposed to fluctuating nutrients during a vital period of brain development. Despite, the
well-known fact that GDM occurs during this important period of brain development, little
research has investigated the neurocognitive consequences of GDM exposure.
Previous studies have eluded to the hippocampus being impacted by in utero exposure
to GDM, however this is one of the first studies to assess hippocampal function and
hippocampal structure in a cohort of children exposed to GDM in utero.
Gestational Diabetes Mellitus and Offspring Hippocampal Function and Structure
Prior studies in infants exposed to diabetic mothers in utero has shown that infants
exhibit recognition memory impairments and explicit memory impairments
83,89,90
. Using
an elicited imitation paradigm to assess hippocampal-mediated non-verbal recall, they
53
showed that GDM-exposed infants performed significantly worse compared to non-
exposed offspring.
83,89
. They also found that offspring exhibited reduced cognitive
development using the Bayley scales of infant development. Along this line of research,
Siddappa et al.
90
, used event-related potentials to test auditory recognition memory in
infants of diabetic mothers. They not only observed that infants of diabetic mothers had
recognition memory impairments, but also iron levels of the diabetic mothers had an
important role in the degree of recognition memory impairments
90
. While these studies
hypothesized that the hippocampus contributed to the observed memory deficits, it has
yet to be determined if GDM exposure in utero is associated with reduced hippocampal
function and structure.
To date, there have only been a handful of studies investigating if GDM-induced
memory impairments are present during childhood
50,82,255
. While Bolanos et al.
50
,
observed that GDM-exposed offspring had reduced IQ’s and working memory
impairments, Veena et al.
255
, observed no deficits in GDM-exposed offspring.
Interestingly, Jabes et al.
82
also observed no deficits in IQ and recognition memory in
offspring of diabetic mothers. In a large epidemiological study including 723,775 men,
Fraser et al.
256
, found that the observed association between reductions in IQ and
memory performance in offspring of diabetic mothers, was attenuated once siblings’ IQ
was accounted for. These siblings were all born before the mothers had developed
diabetes. There are two important points here that should be emphasized. The first is,
none of these studies investigated hippocampal-dependent memory. The second point
is, the studies that have found null results were in women with pre-gestational diabetes,
and these women may have had their diabetes better managed than women who
54
develop GDM during pregnancy. Therefore, children exposed to gestational diabetes in
utero may be at greater risk for memory impairments.
While findings in humans are not yet certain of whether GDM-exposed offspring
exhibit hippocampal-memory deficits during childhood, findings in animals paint a
clearer picture. In rat offspring, intrauterine exposure to gestational diabetes leads to a
loss of hippocampal neurons, various altered cell signaling cascades in the
hippocampus and reduced performance on spatial memory tasks
85–87,251,253,257,258
.
Importantly, many of these studies have found changes in insulin signaling and
inflammatory pathways in the hippocampus
86,257,258
. Additionally, many of these studies
have observed decreased neurogenesis and increased apoptosis in the CA1, and CA3
subfields of the hippocampus
85,251,253
. Altogether, studies in GDM-exposed rat
offspring, provide strong evidence that GDM affects hippocampal development.
In this chapter, my goal was to investigate whether children exposed to GDM in
utero, exhibited hippocampal-dependent memory deficits, and reduced hippocampal
grey matter volume. I used a relational memory task that has been shown to be
hippocampal-dependent in a group of patients with severe hippocampal damage to
examine hippocampal-dependent memory function in GDM-exposed and non-exposed
children. In addition to using a hippocampal-dependent memory task, a hippocampal-
independent memory task that assesses item-familiarity was used. Both of these tasks
have been validated in children
16,17
. I hypothesized that children exposed to gestational
diabetes in utero would display hippocampal-related memory deficits based on previous
studies in infants and animal models, additionally, that children exposed to GDM would
not exhibit group differences in hippocampal-independent memory function. Based on
55
studies in rat offspring, I hypothesized that GDM-exposed offspring would have reduced
hippocampal volume, specifically in the left CA1 and CA3 subfields.
Statistical Analysis
The relational task has 1 scoring metric, accurate single item placement, that
was scored as a continuous variable and compared between GDM-exposed and non-
exposed children. The item-familiarity task scored on accuracy was scored as a
continuous variable. Group differences were tested using analysis of covariance
(ANCOVA) with the covariate variables including age, sex, tanner stage, income and
maternal education at birth (socioeconomic (SES) status at birth), pre-pregnancy, BMI
z-score and IQ. Sex and SES, were accounted for, because these are two variables
previously shown to influence hippocampal-dependent memory
215–217
. Additionally,
statistically significant Pearson’s correlations existed for IQ, Tanner stage, and child age
with accurate single-item placement (see correlation matrix Table 6). Statistical analysis
software (SAS) was used for the analyses with a p<0.05 significance level.
While it may be apparent to the reader, chapter 2 assessed total hippocampal
volume because both the left and right hippocampus were highly correlated and there
was not sufficient evidence for why the left hippocampus would differ from the right
hippocampus in relationship to maternal pre-pregnancy BMI. However, functionally,
research suggest that the left hippocampus differs from the right
259
. The left
hippocampus contributes to declarative memory while the right hippocampus
contributes more to spatial memory
165,259
. And further research in animals suggests that
56
GDM exposure impacts the left hippocampus preferentially due to the increased density
of insulin receptors in the left hippocampus and specifically impacts the CA1/CA3
subfields
85,86,251
. Therefore, the left vs. the right hippocampus were investigated
separately, and I hypothesize that the left hippocampus will be impacted preferentially.
Additionally, based on the animal literature, as well as to reduce multiple comparisons,
the CA1/CA3 subfields were investigated. Analysis of covariance was used to analyze
group differences in left and right hippocampal gray matter volume, along with left and
right CA1/CA2/3 gray matter volume. Apriori covariates to control for potential
confounding were child age, BMI z-score, total intracranial volume (ICV), pre-pregnancy
BMI obtained from EMR, and socioeconomic status (SES), which included household
income at birth estimated based on census tract of residence, and maternal education
at birth, which was extracted from birth certificates in EMR. These covariates were
previously shown to influence hippocampal volume, thus were added to the model to
control for potential confounding variance
101,209–213
. Additionally, false discovery rate in
SAS was used to control for multiple comparisons, with a critical value of q<.05.
Cohen’s D was used to compare GDM-exposed to non-exposed children. Interpretation
was as follows, less than 0.20 was a small effect size, around 0.50 was a moderate
effect size, and around 0.80 was a large effect size
260
. Additionally, visual
assessments, and the feature, proc univariate in SAS, were used to determine if the
potential confounding independent variables, SES, BMI z-score, maternal pre-
pregnancy BMI, and age were normally distributed. This same approach was applied to
the dependent variables, left and right hippocampal volume, IQ, accurate single-item
placement, item-familiarity, and picture-sequence episodic memory. Upon visual
57
inspection, and proc univariate diagnostics, all variables were approximately normally
distributed.
Results
Participant Demographics: See Chapter Two, Part One.
GDM Exposure and Hippocampal Structure
Based on prior results (Chapter Two: Part One), there was a sufficient rationale to
stratify by sex and investigate sex differences in the effect of GDM-exposure.
Additionally, in both the left and right hippocampus, there was a marginal interaction
with sex and GDM exposure (p<0.10). In the left (p=0.004) and right CA1 (p=0.029),
there was a significant interaction between sex and GDM-exposure. There was not a
significant interaction between sex and GDM-exposure in the CA2/3.
In boys, there was no significant difference in either left or right hippocampal
volume between GDM-exposed and non-exposed children (Table 13). After controlling
for multiple comparisons using FDR, there were no significant differences in the
hippocampal subfields. There was a marginal difference in left CA1 volume between
GDM-exposed and non-exposed boys, with non-exposed boys having larger left CA1
volume in both the unadjusted and adjusted models (Unadjusted: GDM-exposed
LSmean=655.59, SE=12.65; Non-exposed LSmean=701.19, SE=14.14; p=0.022, Table
58
14). With a Cohen’s D of 0.79, GDM-exposure had a large impact on left CA1 volume
(Table 15). There was a moderate effect on the right CA1. There was little to no effect
of GDM-exposure on the left or right CA2/3.
In girls, there was no significant difference in left or right hippocampal volume
between GDM-exposed and non-exposed children (Table 16). After controlling for
multiple comparisons using false discovery rate (FDR), there was no significant
difference between GDM-exposed and non-exposed girls in left or right hippocampal
subfields (Table 17). However, there was a marginal difference in adjusted left CA2/3
volume, with GDM-exposed girls having a larger volume than non-exposed girls (GDM-
exposed LSmean=198.87, SE=4.18; Non-exposed LSmean=184.24, SE=5.96;
p=0.045). GDM-exposure had the largest effect on the left CA2/3 subfield with a
Cohen’s D of -0.59 (Table 18). There was a small to moderate effect of GDM-exposure
on all the other subfields.
GDM exposure and Cognitive Function
There was a significant interaction between GDM-exposure and sex in the PSMT
(p=0.023). GDM-exposed boys performed worse on the picture-sequence memory task
than non-exposed boys (Table 19) while GDM-exposed girls performed significantly
better than non-exposed girls on the PSMT (Table 21). In boys, there was a small effect
of GDM-exposure on PSMT performance, with a Cohen’s D of 0.22 (Table 20). The
mean (SE) PSMT performance for GDM-exposed girls was 52.04 (1.89), and for non-
59
exposed girls was 42.25 (2.20), p=0.001. In girls, GDM-exposure had a large effect on
PSMT performance, Cohen’s D=-0.98 (Table 22). Additionally, GDM-exposure had a
small effect on IQ scores and had little to no effect on accurate single-item placement.
Interestingly, GDM-exposure had a moderate effect on item-familiarity in girls, with a
Cohen’s D of 0.48 in the model adjusting for age and IQ, suggesting that GDM-exposed
girls had worse item-familiarity performance.
Discussion
Based on animal studies and prior studies in infants, I investigated if GDM-
exposure was associated with changes in hippocampal memory and hippocampal
structure during childhood. Similar to animal models of maternal obesity, I also found
that the relationship between GDM-exposure and hippocampal structure/function
differed between boys and girls. Overall, GDM-exposure had no effect on hippocampal
structure, hippocampal function or global cognitive function (data not shown). However,
when stratified by sex, GDM-exposed girls exhibited larger left CA2/3 volumes, and
significantly better performance on the hippocampal-related task (PSMT). They had
slightly better performance on the hippocampal-dependent memory task (relational
memory), worse performance on the hippocampal-independent memory task (item-
familiarity and slightly higher IQ scores. GDM-exposed boys had smaller left CA1
volume but did not perform significantly worse on any of the cognitive measures.
However, GDM-exposed boys performed slightly worse on the hippocampal-related task
(PSMT).
60
Ordinarily, the left and right hippocampus are asymmetrical in volume, with the
right being bigger than the left
259,261,262
. These differences in volumes reflect differing
functions related to memory and spatial navigation, the left contributing more to
declarative memory (episodic memory) functioning, and the right being involved in
spatial memory
165,259
. Based on these differing functions, the number of synapses and
connections differ between the left and right hippocampus. For example, animal studies
have shown that the density of insulin receptors is greater in the left hippocampus
compared to the right
86,263
. Interestingly, some studies suggest that insulin resistance in
the brain is an important component to the pathophysiology of Alzheimer’s disease
264,265
. And further, episodic memory impairments are a notable feature of Alzheimer’s
disease. Surprisingly, GDM-exposed girls performed significantly better on the episodic
memory task and had increased left CA2/3 volume. Collectively, these findings could be
related to differences in insulin signaling. Overall, when stratified by sex, it was apparent
that GDM-exposure impacted the left hippocampus preferentially, with the left CA1
being marginally reduced in GDM-exposed boys, and the left CA2/3 being marginally
increased in the GDM-exposed girls. Moreover, animal studies have shown that the
CA1/CA3 subfields are preferentially impacted by in utero exposure to maternal
diabetes
85,251,252
.
While my initial hypothesis, that the left CA1 and left CA3 would be preferentially
impacted by GDM exposure was accurate, in girls the directionality was the opposite of
my hypothesis. GDM-exposed girls had larger left CA2/3 volumes and performed
significantly better on a hippocampal-related memory task than non-exposed girls.
There are a few potential explanations for these opposing findings. The first being that
61
this may be a compensatory mechanism of GDM-exposure. Or, it is possible, the
diagnosis of GDM-exposure instigated lifestyle changes in the mothers with GDM that
beneficially impacted GDM-exposed girls only, due to sex differences in utero. Lastly,
there is a possibility, the sub-group of GDM-exposed girls, had a potential confounding
variable in common that non-exposed girls did not have. But because the CA3 has an
important role in episodic memory, the observation of increased CA3 volume and
episodic memory performance in GDM-exposed girls is suspicious
184,266,267
.
While animal models have provided sufficient rationale for why hippocampal
structure and hippocampal memory may be impacted by GDM-exposure, additional
studies in infants exposed to GDM in utero, have also exhibited memory deficits
83,89,90
.
As previously introduced, the findings in GDM-exposed girls, may be the result of a
compensatory mechanism from having potential memory deficits during infancy. To
explain further, it is possible that the children in our cohort exposed to GDM in utero,
had memory deficits during infancy and the brain compensated for these deficits by
allocating resources to the hippocampus. And as a result, neurogenesis could have
been increased in the hippocampus, specifically the CA3 subfield and as a result,
episodic memory function was significantly improved. Given the protracted growth of the
CA3 subfield, infancy is a plausible time frame for increased neurogenesis.
Unfortunately, only longitudinal studies could determine if this is possible. Interestingly,
CA3 development coincides with episodic memory performance. Some studies have
suggested that episodic memory performance is based on CA3 development
168,184,266
.
Therefore, this potential explanation has some plausibility.
62
In contrast, if a compensatory mechanism were true, why wouldn’t GDM-exposed
boys also have increased CA2/3 volume and increased episodic memory performance.
It is important to consider; boys may not have the same compensatory mechanism
potentially due to sex differences in brain resiliency. For example, if the reader recalls
from chapter 2: part one, in utero exposures impact boys differently than girls. For
example, in utero exposure to maternal obesity or maternal stress in animals causes
decreased cognitive function/decreased hippocampal volume in boys while it is
associated with increased anxiety in girls but overall girls were less impacted
65,66,234,235,268
. Similarly, pediatric cognitive studies have shown similar patterns with girls
being less impacted by traumatic brain injuries than boys
269–271
. Collectively, there is
evidence to suggest, girls may be more resilient to brain insults, however sex
differences in brain resiliency are beyond the scope of this dissertation.
As previously suggested, GDM-exposed girls may have inadvertently benefited
from mothers diagnosed with GDM who altered their lifestyle based on the doctor’s
recommendation of diet and physical activity. It is worth noting, about a fourth of
mothers with GDM had BMIs less than 25. Moreover, the treatment for women with
GDM is either lifestyle modification (diet and exercise) or medication. Interestingly,
physical activity during pregnancy can benefit offspring cognition
272–274
. Contrarily
though, Esteban-Cornejo
272
, found that boys benefited more from maternal physical
activity during pregnancy than girls. However, they did not take into account maternal
pre-pregnancy BMI which may have confounded their findings. Conversely, it is possible
that mothers with GDM who were advised to change their diets and increase their
physical activity levels, gained less weight following their GDM diagnosis compared to
63
mothers who were not exposed to GDM. This leads to the potential third explanation of
findings. There may have been a confounding variable not accounted for in the girls
exposed to GDM. Importantly, there were many lifestyle and prenatal exposures
accounted for in these analyses and post-hoc analyses revealed that controlling for
child physical activity, maternal race, maternal weight gain during pregnancy and length
of breast feeding did not attenuate group differences. However, there may have been
an additional factor not accounted for. For example, maternal stress during pregnancy
was not accounted for. Potentially, mothers without diabetes during pregnancy may
have had greater levels of stress that negatively impacted the hippocampal-related
memory of non-exposed girls more than GDM-exposed girls. Altogether, it is not likely
that GDM-exposure had a beneficial impact on hippocampal development in girls but
rather, there was a confounding variable not accounted for. To address the possibility
that prenatal stress is a confounding variable, a modified version of the Cohen
Perceived Stress Scale could be used
275
. Similar to other groups, the length of time
assessed in the questionnaire can be modified to reflect study needs
276,277
.
Conclusion
While memory deficits have been reported in infants exposed to maternal
diabetes, currently there has only been two studies investigating memory performance
in a cohort of school-aged children
50,82,83,89,90
. This is the first study to investigate
hippocampal memory performance and hippocampal structure in a cohort of school-
aged children exposed in utero, to mothers with well-controlled and well-documented
64
gestational diabetes. There were many strengths of this study design, however a
potential caveat of this study was the initial cross-sectional design. It still remains vital
for future studies to investigate if the memory deficits observed by infants exposed to
maternal diabetes persist into childhood and more importantly as they progress through
puberty. Ornoy et al.
278
studied a cohort of children aged 5 to 12 years old exposed to
maternal diabetes in utero and found only that the younger children, 5-8 years old,
exhibited cognitive impairments
278
. Additionally, Jabes et al.
82
, found no differences in
hippocampal volume in a cohort of children aged 10 years old
82
. Therefore, future
studies could address this concern by following children exposed to gestational diabetes
through early childhood to adolescence. Based on a fairly healthy childhood, the
memory deficits observed during infancy may not persist into adolescence, as
evidenced by our study.
Chapter Three: Contributions of Prenatal Exposures and Child
Lifestyle to Insulin Sensitivity and Adiposity during Childhood
Introduction
Over the past few decades, the number of youth impacted by type 2 diabetes,
has increased substantially
279
. Since 2001, the prevalence of type 2 diabetes among
youth has increased by 31%
279
and over a third of children are now considered
overweight or obese
194
. Growing evidence suggests that prenatal exposure to
gestational diabetes mellitus (GDM) and/or maternal obesity are contributing factors to
the rising rates of type 2 diabetes and obesity in children and adolescents
30,39,280
. One
65
of the major determinants of type 2 diabetes is reduced insulin sensitivity, which
requires a compensatory increase in insulin secretion in order to maintain normal
glucose levels
281
. Inadequate β-cell compensation for insulin resistance results in
impaired β-cell function and the progression to type 2 diabetes
282–284
. Determining
which factors most strongly influence child insulin sensitivity and child adiposity may
help to identify potential targets for type 2 diabetes disease and obesity prevention
strategies.
Prior studies have shown that offspring exposed to GDM in utero exhibit reduced
insulin sensitivity as early as birth
285,286
. Additional studies have observed that reduced
insulin sensitivity in GDM-exposed offspring is present during late childhood
27,44,287
and
adulthood
288
. Children born to mothers with obesity during pregnancy are also at higher
risk for insulin resistance and type 2 diabetes
289–291
. While some studies have observed
an association between in utero exposure to maternal obesity and reduced insulin
sensitivity in offspring at birth
290,292
and during adulthood
289
, only a few have examined
associations between in utero exposure to maternal obesity and insulin sensitivity
during childhood, and they have found conflicting results
27,41,44
.
In addition to prenatal exposure factors, lifestyle factors also play an important
role in insulin sensitivity during childhood
293–296
. Higher energy intake, particularly high
amounts of added sugar is associated with reduced insulin sensitivity during
adolescence, particularly among overweight/obese youth, and increased type 2
diabetes risk
293,294,297–299
. Sedentary lifestyles during childhood may also be related to
reduced insulin sensitivity
295
, and longitudinal studies have shown that children who
66
maintain high levels of physical activity have improved insulin sensitivity and reduced
risk of developing type 2 diabetes
296,300
.
Prenatal exposures and lifestyle behaviors are also important predictors of child
adiposity. Many studies have shown that prenatal exposure to maternal obesity and/or
GDM is associated with increased birth weight
40,301–305
, higher BMI percentiles during
childhood
11,14,27,28,32,306,307
and BMI during adulthood
42,43,289,308,309
. Studies have also
found that central adiposity is higher in children exposed prenatally to maternal obesity
and/or GDM
306,307,310–312
. Alternatively, several postnatal factors have been shown to
influence child obesity risk such as physical activity levels, television time, energy intake
and consumption of sugar-sweetened beverages
313–318
. Additionally, poor diet quality
and low physical activity levels/high sedentary time, are also associated with greater
central adiposity during childhood
319–323
. Interestingly, only a handful of studies have
looked at the contributions of both prenatal and postnatal factors on child adiposity
307,319,324,325
. Moreover, these studies did not look at the independent contributions of
both prenatal exposures and postnatal factors on child adiposity.
This is the first study to investigate how both prenatal factors and early childhood
lifestyle factors influence insulin sensitivity in a well-characterized cohort of healthy,
typically developing elementary-aged children. Therefore, in addition to investigating the
impact of in utero exposure to GDM or maternal obesity on child insulin sensitivity and
child adiposity, an additional goal was to investigate the role that modifiable lifestyle
factors, including physical activity levels and energy intake, particularly dietary added
sugar, contribute to insulin sensitivity and adiposity at an early age when potential
interventions could be targeted.
67
Statistical Analysis
Participant descriptive statistics (means, frequencies) were performed. Linear
regression was used to analyze relationships between the predictor variables [MVPA
and SB (min/day), maternal pre-pregnancy BMI (kg/m
2
), maternal GDM status (yes/no),
daily EI (kJ), and dietary added sugar] and the outcome variable, ISI. Additionally, linear
regression was used to analyze relationships between prenatal factors (i.e., maternal
pre-pregnancy BMI and maternal GDM status) as the predictor variables with postnatal
factors, MVPA, SB, daily EI, and dietary added sugar as the outcome variables. For
continuous predictors, standardized regression coefficients were reported such that the
regression coefficients represent change in ISI per standard deviation of change in each
predictor variable. The standardized regression coefficients are unit free and
directly comparable across different predictors with different measurement units.
Time in MVPA and ISI were not normally distributed, and a square-root transformation
normalized the distribution and was applied for the regression analysis. A priori
covariates included in each linear regression analysis were child age in years, sex, and
BMI z-score. Each linear regression model also simultaneously included the other
exposure variables of interest (i.e., time spent in MVPA, time spent in SB, maternal pre-
pregnancy BMI, maternal GDM, daily EI and dietary added sugar), with the exception of
models containing time spent in MVPA and time spent in SB. Time spent in MVPA and
time spent in SB were highly inversely correlated with a spearman rank of 0.62
(p<0.0001), therefore these variables were not modeled together.
68
For analyses in which child adiposity was the outcome variable, the same a priori
covariates were included (child age and sex). Similar to the prior analyses, each
analysis included a model mutually adjusting for the other predictor variables. The
adiposity outcome variables of interest were, BMI z-score, total percent body fat, waist
to height ratio, and BMI percentile categories. For BMI z-score, total percent body fat,
and waist to height ratio, multiple linear regression was used. For BMI percentile
categories, children were classified as either healthy weight (BMI percentile<85), or
overweight/obese (BMI percentile>=85). Multiple logistic regression was used to assess
the odds of being overweight/obese based on prenatal (maternal obesity/GDM) and
postnatal factors (MVPA and SB (min/day), daily EI (kcal), and dietary added sugar
(%)). Additionally, for continuous predictors, standardized regression coefficients were
reported such that the regression coefficients represent change in the adiposity
measures per standard deviation of change in each predictor variable. P-values <0.05
were interpreted as statistically significant. SAS 9.4 statistical software (SAS Institute,
Cary, NC USA) was used for all statistical analyses.
Results
Participants Characteristics
Of the 114 children enrolled into the study at the time of this dissertation, 91 children
completed the OGTT (Figure 7). The demographics of participants who did not
complete the OGTT did not differ significantly from participants who did (Table 23). For
69
participants who completed the OGTT, the mean ± SD age was 8.34 ± 0.87 years old,
93% of the children were pre-pubertal (Tanner Stage <2), and 57% were girls (Table
24). Mean ISI was 10.58 ± 5.92. Mean time spent in MVPA was 134.94 ± 94.40
minutes. Mean time spent in SB was 588.15 ± 130.55 minutes. Mean daily EI was
1791.19 ± 446.65 kcal, and mean dietary added sugar was 14.94% ± 6.56%. Children’s
BMI ranged from 13.62 to 34.01 kg/m
2
; BMI percentiles ranged from 5.28 to 99.58; BMI
z-scores ranged from -1.78 to 2.64. According to CDC standards, 60 (66%) children
were healthy-weight, 12 (13%) children were overweight and 19 (21%) children were
obese. Maternal pre-pregnancy BMI ranged from 18.97 to 50.38 kg/m
2
, 24 (26%) were
normal weight, 29 (32%) were overweight, and 38 (42%) were obese. Lastly, 54 (59%)
mothers had GDM.
Association between Matsuda Insulin Sensitivity Index and Predictor Variables
There was a statistically significant positive association between time spent in MVPA
and ISI (b=0.29; p=0.005) (Table 25). Adjusting for child age, sex, maternal pre-
pregnancy BMI, maternal GDM status, daily EI, and dietary added sugar did not change
this association (b=0.29; p=0.005). The association remained significant after further
adjusting for child BMI z-score (b=0.28; p=0.002). Additionally, to visually demonstrate
this significant association, least-square means (LSmean) of ISI by terciles of time spent
in MVPA were compared using analysis of covariance, adjusted for child age, sex,
maternal pre-pregnancy BMI, maternal GDM status, daily EI, dietary added sugar and
child BMI z-score (Figure 8). Children in the upper tercile of reported time spent in
70
MVPA had significantly better insulin sensitivity than children in the lower tercile
(LSmean Upper Tercile: 12.65 SE=1.07; LSmean Lower Tercile: 9.20, SE=1.08;
p=0.027). The LSmean ISI of children in the middle tercile did not differ significantly
from children in the lower tercile (LSmean Middle Tercile: 10.70, SE=0.77; p=0.27).
Although time spent in SB was not significantly associated with ISI, the relationship was
in the expected direction (unadjusted, b=-0.13; P=0.23; fully-adjusted, b=-0.14; p=0.13).
Daily EI was not associated with ISI (unadjusted, b=0.13; p=0.24; fully-adjusted, b=0.04;
p=0.64), and neither was dietary added sugar (unadjusted, b=0.002; p=0.98; fully-
adjusted, b=0.01; p=0.90).
There was a negative relationship between maternal pre-pregnancy BMI and
child ISI (b=-0.17; p=0.090) (Table 25). However, this relationship was weakened after
adjusting for child BMI z-score (b=-0.02; p=0.83). Before and after adjusting for
covariates, GDM exposure was not significantly associated with ISI (unadjusted,
b=0.07; p=0.75; fully-adjusted, b=0.23; p=0.22).
Association between Child BMI z-score and Predictor Variables
Time spent in MVPA was not associated with child BMI z-score (unadjusted, b=-0.02;
p=0.86; fully-adjusted, b=-0.02; p=0.82) (Table 26). Time spent in SB was also not
associated with child BMI z-score (unadjusted, b=-0.04; p=0.69; fully-adjusted, b=-0.05;
p=0.62). EI was also not significantly associated with child BMI z-score (unadjusted, b=-
0.11; p=0.28; fully-adjusted, b=-0.14; p=0.14), nor was dietary added sugar (unadjusted,
b=0.001; p=0.99; fully-adjusted, b=-0.06; p=0.58). Maternal pre-pregnancy BMI was
71
significantly associated with child BMI z-score (unadjusted, b=0.32; p=0.001; fully
adjusted, b=0.31; p=0.002). GDM exposure was not significantly associated with a
higher BMI z-score (unadjusted, b=0.32; p=0.12; fully adjusted, b=0.25; p=0.22).
Association between Child Total Percent Body Fat and Predictor Variables
Time spent in MVPA was not associated with child total percent body fat (unadjusted,
b=-0.07; p=0.53; fully-adjusted, b=-0.07; p=0.51) (Table 27). Time spent in SB was also
not associated with child total percent body fat (unadjusted, b=0.04; p=0.72; fully-
adjusted, b=-0.01; p=0.89). EI was also not significantly associated with child total
percent body fat (unadjusted, b=0.02; p=0.88; fully-adjusted, b=0.01; p=0.93), nor was
dietary added sugar (unadjusted, b=0.001; p=0.99; fully-adjusted, b=-0.06; p=0.58).
Maternal pre-pregnancy BMI was significantly associated with total percent body fat
(unadjusted, b=0.29; p=0.006; fully adjusted, b=0.27; p=0.011). Before adjusting for any
covariates, GDM-exposure was significantly associated with a higher total percent body
fat (b=0.43; p=0.041). However, in the fully-adjusted model, this association was no
longer significant (b=0.36; p=0.11).
Association between Child Waist to Height Ratio and Predictor Variables
Time spent in MVPA was not associated with child waist to height ratio (unadjusted, b=-
0.07; p=0.50; fully-adjusted, b=-0.09; p=0.38) (Table 28). Time spent in SB was also not
associated with child waist to height ratio (unadjusted, b=-0.05; p=0.66; fully-adjusted,
72
b=-0.07; p=0.48). EI was also not significantly associated child waist to height ratio
(unadjusted, b=0.10; p=0.34; fully-adjusted, b=0.08; p=0.44), nor was dietary added
sugar (unadjusted, b=-0.09; p=0.41; fully-adjusted, b=-0.15; p=0.13). Maternal pre-
pregnancy BMI was significantly associated with child waist to height ratio (unadjusted,
b=0.43 p<0.0001; fully-adjusted, b=0.41; p<0.0001). Before adjusting for any covariates,
GDM-exposure was significantly associated with a higher waist to height ratio (b=0.58;
p=0.007). This significant association remained even after adjusting for covariates
(b=0.46; p=0.026).
Odds of Overweight/obese and Predictor Variables
Time spent in MVPA/SB did not increase the odds of a child being overweight/obese
(Table 29). EI and dietary added sugar also did not increase the odds of the child being
overweight/obese. In the unadjusted model, maternal pre-pregnancy BMI significantly
increased the odds of a child being overweight/obese (Odds ratio=1.07, Wald 95%CI:
1.01~1.14). Also, in the fully adjusted model, maternal pre-pregnancy BMI significantly
increased the odds of a child being overweight/obese (Odds ratio=1.08, Wald 95%CI:
1.01~1.15). After adjusting for other lifestyle and prenatal factors, for each standard
deviation increase in maternal pre-pregnancy BMI, a child was 1.08 times more likely to
be classified as overweight/obese. GDM-exposure did not significantly increase the
odds of a child being overweight/obese (unadjusted, odds ratio=2.14, Wald 95%CI:
0.85~5.40; fully-adjusted, odds ratio=2.23, Wald 95%CI: 0.79~6.23).
73
Associations between Prenatal and Postnatal Factors
Maternal pre-pregnancy BMI was not statistically significantly associated with daily EI
(unadjusted, b=0.03; p=0.80). After fully adjusting for maternal GDM status, child age,
sex, dietary added sugar, time in MVPA, ISI, and BMI z-score, there also was no
association between maternal pre-pregnancy BMI and daily EI. (b=0.008; p=0.94)
(Table 30). There was also no statistically significant relationship between maternal pre-
pregnancy BMI and dietary added sugar (unadjusted, b=0.06; p=0.59; fully-adjusted,
b=0.11; p=0.33). Maternal pre-pregnancy BMI was not associated with time spent in
MVPA (unadjusted, b=-0.06; p=0.55; fully-adjusted, b=-0.05; p=0.67) or time spent in
SB (unadjusted, b=0.02; p=0.84; fully-adjusted, b=0.02; p=0.84).
Before adjusting for any covariates, there was not a significant association
between GDM exposure and daily energy intake (b=0.39; p=0.070) (Table 31). After
adjusting for covariates, this relationship was significant (b=0.45; p=0.044). On average,
GDM-exposed children had a higher energy intake than non-exposed children (mean ±
SD, 1861.43, 468.19 calories/day vs. 1688.67, 397.28 calories/day), respectively. There
was no statistically significant association between GDM exposure and dietary added
sugar (unadjusted, b=0.19; p=0.37; fully adjusted, b=0.21; p=0.38). GDM exposure was
not significantly associated with time spent in MVPA (unadjusted b=0.12; p=0.59; fully-
adjusted, b=-0.04; p=0.87), or time spent in SB (unadjusted, b=0.19; p=0.37; fully-
adjusted, b=0.27; p=0.25).
Discussion
74
To our knowledge, this is the first study to investigate the unique contributions of
both prenatal exposures and postnatal lifestyle factors to insulin sensitivity in healthy
children. Additionally, we looked at the contributions of prenatal exposures and lifestyle
factors to child adiposity. Time spent in MVPA was the only predictor significantly
associated with child insulin sensitivity at this early age, and its contribution to insulin
sensitivity was independent of other prenatal and postnatal factors. We observed a
negative relationship between maternal pre-pregnancy BMI and child insulin sensitivity;
however, this was not independent of child BMI z-score. Our findings therefore indicate
that physical activity, a modifiable lifestyle factor, had stronger associations with
children’s insulin sensitivity compared to in utero programming mechanisms of maternal
obesity and GDM, dietary factors (energy intake, added sugar intake), and time spent
sedentary.
In contrast, prenatal exposure to maternal obesity and/or GDM were significantly
associated with greater child adiposity. Prenatal exposure to maternal obesity was
associated with greater child BMI z-score, total percent body fat, waist to height ratio
and increased odds of offspring being overweight/obese, independent of GDM
exposure, and lifestyle factors. Independent of maternal pre-pregnancy BMI and lifestyle
factors, GDM exposure was significantly associated with greater waist to height ratio.
Similar to other studies, time spent in MVPA was more strongly associated with
child insulin sensitivity than time spent sedentary
326
. Several studies in adults have
illuminated the benefits of engaging in physical activity for improving insulin sensitivity
(see review by Bird & Hawley, 2017). Potential mechanisms by which engaging in
75
physical activity positively impacts insulin sensitivity include facilitating greater glucose
transporter type 4 (GLUT4) receptor translocation to the cell surface for glucose uptake
into muscles
328
, increasing skeletal muscle capillary recruitment
329
, and decreasing
circulating tumor-necrosis factor-α concentrations and other inflammatory cytokines
330
.
Our results suggest that the benefits of MVPA on insulin sensitivity are significant during
childhood, and our study adds to the existing literature by showing an independent
contribution of MVPA on insulin sensitivity from an array of prenatal and postnatal
factors.
Interestingly, time spent in MVPA and time spent sedentary were both not
significantly associated with child adiposity. While some studies have shown that
increased sedentary time is associated with increased risk of being obese during
childhood
318,321
, some studies have found no association
322,331,332
. As demonstrated by
Falbe et al.
333
, the type of sedentary activity is important for predicting adiposity rather
than total sedentary time. To elaborate, they found that time spent watching television,
was the strongest predictor of child adiposity. Therefore, it is likely, if we assessed time
spent watching television and other social media use with child adiposity rather than
total time spent doing sedentary activities, we may observe similar relationships.
Surprisingly, time spent in MVPA was not associated with child adiposity or
reduced risk of being overweight/obese. Several prior studies have shown that greater
participation in physical activity during childhood is associated with reduced risk of
obesity
315,318,332,334,335
. However, it is possible, children who engaged in more physical
activity also spent a lot of time sedentary, which may have negated any benefits for
reduced adiposity. Coincidently, Herman et al.
335
, found that children who engaged in
76
more than the recommended time of physical activity yet reported higher levels of
sedentary time, had no difference in central and global adiposity compared to children
who did not meet the recommended time in physical activity but had lower sedentary
time. Alternatively, the children who were most physically active and had the lowest time
spent sedentary had the lowest central and global adiposity measurements. The
authors referred to the phenomenon where a child who meets the recommended
amount of physical activity but spends the rest of their time sedentary, as an “active
couch potato”. There is a possibility, that we have a cohort of active coach potatoes.
Additionally, there may be a dose-dependent relationship with adiposity and physical
activity. For example, the level of physical activity required to have an effect on
adiposity may be greater than the level needed to have an effect on insulin sensitivity.
Interestingly, some studies have shown that higher levels of only vigorous physical
activity was associated with reduced child adiposity while moderate physical activity
was not
332,336
. Future studies should consider how vigorous physical activity specifically
relates to child adiposity.
While time spent in MVPA was associated with higher ISI, we did not observe an
association between daily energy intake and insulin sensitivity. Animal studies have
shown that excessive energy intake can lead to insulin resistance
337
, however, findings
in humans are mixed. For example, Donin et al.
294
, observed that energy intake was
associated with insulin resistance during childhood, whereas other studies have found
null associations
338
. Moreover, while previous studies have observed an association
between added sugar consumption and insulin resistance during adolescence
293,298
, it
is possible that these studies were confounded by pubertal status. In our cohort, 93% of
77
children were prepubescent, and there is a possibility that the association between high
consumption of added sugar and insulin resistance becomes more prominent during
adolescence, when large changes in body composition and insulin resistance typically
occurs (see review by Kelsey and Zeitler 2016). It is also possible that other
components of the participants’ diet, such as a higher fiber content, may have had
beneficial effects on insulin sensitivity
340
.
In a similar manner, total energy intake and dietary added sugar were not
associated with greater adiposity or increased odds of being overweight/obese.
Although some studies have shown an association between higher energy intake and
greater adiposity during childhood
313,323
, some studies have shown that overall protein
intake or consumption of ultra-processed food contributes more to greater adiposity
319,341
. Similarly, dietary added sugar was not associated with adiposity. In contrast,
several studies have found that higher consumption of sugar-sweetened beverages is
associated with higher central and global adiposity in children as young as 2 years old
314,317,320
. A possible explanation for why we did not observe a similar relationship is that
participants in our cohort may have not consumed the necessary number of sugar-
sweetened beverages to have an impact on child adiposity. Interestingly, Nasreddine et
al.
322
, also found that sugar consumption from food/beverages was not associated with
adiposity during childhood. Coincidentally, they reported similar amounts of added
sugar intake to our cohort and to nationally reported averages
342
. Further detailed
nutritive studies are needed to determine if increased energy intake, specific
macronutrients, and/or overall diet quality are associated with decreased insulin
sensitivity or increased adiposity during childhood.
78
An increasing number of studies have implicated the role of developmental
programming on metabolic disease risk
27,288,289,291
. However, the mechanisms behind
this relationship are still speculative. Rodent studies provide evidence that alterations in
satiety signaling in the brain may contribute to obesogenic eating behaviors, such as
greater food intake in GDM-exposed offspring
343
. Recent studies in humans also
suggest that alterations in the hypothalamus, a brain region that contributes notably to
satiety signaling, may be involved in obesogenic eating behaviors in GDM-exposed
offspring
344
. Interestingly, we found that GDM-exposed offspring reported greater
consumption of energy intake than non-exposed offspring. Similarly, other studies have
found that GDM-exposed offspring exhibited increased obesogenic eating behaviors
345
.
Future studies should consider if altered hypothalamic satiety signaling predicts greater
caloric intake or obesogenic eating behaviors in GDM-exposed offspring.
Furthermore, a number of prior studies have shown that children exposed to
GDM in utero have an increased risk of developing obesity, insulin resistance, and type
2 diabetes
32,44,288,291,346
. Interestingly, studies have shown an age-related difference in
the relationship between GDM exposure and child insulin sensitivity whereby the
relationship between GDM exposure and child insulin resistance does not appear to
manifest until late childhood
346
. Therefore, it is possible that we did not observe an
association between GDM exposure and child insulin sensitivity due to the relatively
young age range of our cohort, and this relationship may become more apparent with
age. Similarly, age may also be a factor in the lack of evidence for an association
between maternal pre-pregnancy BMI and child insulin sensitivity. Several studies have
found an association between maternal pre-pregnancy BMI and offspring insulin
79
resistance at birth
290
and during adulthood
289
. However, there are limited studies that
have observed this relationship during childhood
41,44,346
. Further, it is unclear whether
the relationship between maternal pre-pregnancy BMI and offspring insulin resistance is
independent of the offspring’s adiposity
41,44,346
. Similar to Sauder et al.
44
, we also
observed that the relationship between child insulin sensitivity and maternal pre-
pregnancy BMI was weakened after controlling for child adiposity. Based on these
results, it is evident that more longitudinal studies are needed to determine how prenatal
exposure to maternal obesity and/or GDM influence offspring insulin sensitivity across
the lifespan.
In contrast, maternal pre-pregnancy BMI was significantly associated with greater
child adiposity, independent of maternal GDM, and lifestyle factors. Several other
studies have observed similar findings with increased adiposity during childhood
11,14,45,306,346
. Interestingly, maternal pre-pregnancy BMI was more strongly associated
with central adiposity rather than total adiposity measures. However, due to sample size
constraints, we did not control for the mother’s current BMI, therefore it is possible that
the association between maternal pre-pregnancy BMI and child adiposity is driven by
genetic factors. A larger sample size would help address this possibility.
Additionally, we observed that GDM-exposure was associated with central
adiposity but not with global adiposity. Children exposed to GDM in utero, had a
significantly greater waist to height ratio, even after controlling for maternal pre-
pregnancy BMI and other lifestyle factors. These findings suggest, independent of a
child’s diet and physical activity patterns, prenatal exposure to GDM is associated with
increased central adiposity. Interestingly, several other studies have observed a similar
80
relationship between GDM-exposure and greater central adiposity but not global
adiposity
307,310,347
. Alternatively, Sauder et al.
324
, found that GDM-exposed offspring
who ate a healthier diet, or engaged in vigorous physical activity, did not have increased
central adiposity compared to non-exposed offspring. These findings suggest that more
rigorous lifestyle modifications are necessary to influence central adiposity accumulation
in GDM-exposed offspring. Moreover, we initially observed that GDM-exposed children
had greater total percent body fat but after controlling for lifestyle factors and maternal
pre-pregnancy BMI, this relationship was no longer significant. Intriguingly, other studies
have also observed an attenuation of global adiposity and GDM-exposure following
adjustment for child lifestyle and maternal pre-pregnancy BMI
11,310,324,347,348
. The
mechanism by which GDM-exposure influences central adiposity rather than global
adiposity remains to be determined. There may be genetic or epigenetic mechanisms at
play. Future studies should consider if GDM-exposed children exhibit increased single
nucleotide polymorphisms (SNP) uniquely related to central adiposity rather than global
adiposity. For example, recent Mendelian randomization analyses have identified 24
SNPs related to central adiposity rather than global adiposity
349
. Further, these SNPs
have been implicated in a causal role with type 2 diabetes. Consequently, the presence
of these SNPs in GDM-exposed offspring may increase their likelihood to store fat in the
abdomen and also contribute to their risk of developing type 2 diabetes later on in life.
Additionally, there may be epigenetic processes that are altered as a result of
GDM-exposure that predispose offspring to greater central adiposity. Compelling
epigenetic studies in both humans and mice have shown that GDM-exposed offspring
exhibit hypomethylation of genes important for body weight regulation
350–352
. Therefore,
81
genetic or epigenetic modifications may both be factors contributing to greater central
adiposity in GDM-exposed offspring. Importantly, future studies are needed to test these
hypotheses.
Strengths and Limitations
Our unique cohort had several strengths. In addition to a thorough assessment of
the lifestyle behaviors of our participants, we had objective measures of the mother’s
GDM status and pre-pregnancy BMI, which were both obtained from electronic medical
records. We performed OGTTs and used the Matsuda ISI to estimate insulin sensitivity,
which is strongly correlated with the gold standard method using the hyperinsulinemic-
euglycemic insulin clamp technique
353,354
. However, our study had some limitations.
While the 3DPAR is a valid measurement for assessing time spent in MVPA and SB
during childhood
355
, there is the possibility that participants over or underreported their
physical activity levels. Children in our cohort reported greater time spent in MVPA than
the average of 88 minutes per a day reported by the National Health and Nutritional
Examination Survey (NHANES) for elementary-aged children
356
. Future studies should
incorporate accelerometer devices to provide an objective measurement of time spent
in MVPA and time spent sedentary. Further, although the 24-hour dietary recall method
is considered one of the gold standard methods to obtain detailed information about
dietary intake, including total energy intake and added sugar consumption
357
, the data
82
are subject to recall bias. However, children in our cohort reported consuming a similar
energy intake as well as dietary added sugar intake to nationally reported averages
342
.
Additionally, we were unable to decipher if greater waist to height ratio is
associated with increased visceral adipose tissue. Future studies should consider more
rigorous approaches to quantify central adiposity during childhood. Lastly, due to the
relatively small size of our cohort, we were limited in investigating potential interactions
between prenatal and postnatal factors, which a larger sample size could potentially
address.
Conclusions
Time spent in MVPA was the strongest predictor of child insulin sensitivity, and
its effects remained significant after adjusting for child’s dietary intake, BMI z-score, and
prenatal exposures to GDM and maternal obesity. These findings are encouraging and
suggest that engaging in MVPA during childhood may be beneficial for insulin
sensitivity, independent of diet, BMI, or prenatal exposures to maternal obesity or GDM.
Thus, engaging in MVPA during childhood may be one strategy to mitigate the adverse
effects of prenatal exposures to maternal obesity and/or GDM on future risk for insulin
resistance and type 2 diabetes. Longitudinal studies and intervention trials are
necessary to test this possibility.
In contrast, we found that lifestyle factors were not associated with child adiposity
measures. However, both prenatal exposure to maternal obesity and GDM were both
independently associated with greater waist to height ratio, even after controlling for
83
lifestyle factors. Earlier accumulation of central adiposity may increase the likelihood of
type 2 diabetes progression in the future and thus reiterates the importance of more
aggressive lifestyle modifications to increase insulin sensitivity and decrease adiposity
in children exposed to maternal obesity and/or GDM in utero.
Chapter Four: Impact of prenatal exposure to Maternal Obesity and
Child Physical Activity Levels on Neurocognition
Introduction
Almost two-thirds of women of childbearing age in the U.S. are overweight or
obese, and the prevalence and severity of obesity during pregnancy continues to rise
194
posing significant health threats to both mothers and their children
25
. Children exposed
to maternal obesity in utero are at increased risk of developing obesity and insulin
resistance
11,41,358
. Additionally, recent evidence from epidemiological studies suggests
that exposure to maternal obesity negatively influences cognitive development in
children
49,52,78,80,81,224
. For example, the US Collaborative Perinatal Project found lower
intelligence quotient (IQ) scores in 7 year-old children born to obese mothers compared
to mothers with normal weight during pregnancy
78
. Further studies have found that
prenatal exposure to maternal obesity is associated with worse academic achievement
scores
52,81,224
. Interventions that could mitigate the adverse effects of prenatal exposure
to maternal obesity on child cognitive development could have a significant impact on
84
public health. Promising results from studies in rodents suggest that engaging in
physical activity improves metabolic health
359,360
and rescues cognitive performance
and hippocampal volume in offspring exposed to maternal obesity in utero
68
; however,
this has yet to be observed in human studies.
A number of studies have illuminated the metabolic and cognitive benefits of
engaging in PA during childhood
334,361–368
. Greater participation in physical activity
during childhood was associated with better performance on intelligence assessments
369,370
. Intervention studies have shown that children who engage in a PA intervention
have significant improvements on academic achievement scores and intelligence
assessments
361–365,367
. Additionally, they have shown that children have greater white
matter integrity after participating in an exercise intervention
371,372
.
Further, the cognitive benefits of PA can be derived even during fetal
development. For example, prior studies have shown that children whose mothers
engaged in physical activity during pregnancy perform better on various cognitive
assessments compared to children born to mothers who did not engage in physical
activity during pregnancy
272,273,373,374
.
Given prior evidence showing links between maternal obesity and poor cognitive
outcomes in children as well as promising results in rodent models suggesting beneficial
effects of PA on neurocognition in offspring exposed to maternal obesity, we aimed to
determine if more time spent engaging in physical activity is associated with higher IQ
scores and increased white matter microstructure in children exposed in utero to varying
levels of maternal obesity.
85
Statistical Analysis
Participant descriptive statistics such as means and frequencies were performed.
Linear regression was used to analyze relationships between the predictor variables,
maternal pre-pregnancy BMI and time spent in MVPA/VPA with the outcome variables,
IQ and global FA. Time in MVPA and time in VPA were not normally distributed, and a
square-root transformation normalized the distribution and was applied for the
regression analysis. A priori covariates included in each linear regression analysis were
child age in years, sex, BMI z-score, and socioeconomic status (SES), which was
assessed using household income at birth, estimated based on census tract of
residence and expressed as a continuous variable, and maternal education at birth,
which was extracted from birth certificates in the EMR as a categorical variable with the
following categories: “high-school or some high-school”, “some college” and “college
and post-education”. Because Tanner stage of puberty was not associated with
outcome variables, IQ (p=0.65) and global FA (p=0.96), Tanner stage was not included
as a covariate. For predictor variable, maternal pre-pregnancy BMI, gestational diabetes
mellitus (GDM) exposure obtained from EMR (dummy coded) was used as an additional
covariate, given that GDM can be comorbid with obesity
375
. Additionally, time in VPA
was modeled as a covariate for the predictor variable pre-pregnancy BMI. Similarly, pre-
pregnancy BMI was modeled as a covariate for the predictor variables, time in VPA and
time in MVPA, to determine if the relationship between physical activity and child
IQ/global FA, is independent of maternal pre-pregnancy BMI.
86
Separate tract-based spatial statistic (TBSS) analyses showed significant
clusters with child age and sex, while SES and child BMI z-score had no significant
clusters, therefore, each contrast, comparing significant clusters of FA included child
age and sex as potential confounding covariates. Maternal pre-pregnancy BMI and
above or below median time spent in VPA were used to test for significant clusters of
FA.
Results
Participants Characteristics
One-hundred children participated in this study. The mean ± SD age was 8.51 ± 1.00
years old, 91% of the children were pre-pubertal (Tanner Stage <2), and 59% were girls
(Table 32). Mean time spent in MVPA was 132.83 ± 96.05 minutes. Mean time spent in
VPA was 20.20 ± 34.76 minutes. The percentage of children who reported above the
median time spent in VPA was 39%. Children’s BMI z-score ranged from -1.78 to 2.64.
According to CDC standards, 60 (60%) of children were healthy-weight, 16 (16%) of
children were overweight, and 24 (24%) were obese. Maternal pre-pregnancy BMI
ranged from 18.97~50.38 kg/m
2
, 24 (24%) were normal weight, 35 (35%) were
overweight, and 41 (41%) were obese.
Association between Maternal pre-pregnancy BMI and Child IQ and Global FA
87
In the unadjusted model, maternal pre-pregnancy BMI was marginally associated with
child IQ (b =-0.31; p=0.10) (Table 33). After adjusting for child age, sex, and BMI z-
score, there was a significant negative association between maternal pre-pregnancy
BMI and child IQ (b =-0.42 p=0.0.037), with 4% of the variance in child IQ being
explained by maternal pre-pregnancy BMI. This relationship became trending after
adjusting for socioeconomic status and maternal GDM status (b=-0.39; p=0.066). There
was no association between maternal pre-pregnancy BMI and child global FA, with
maternal pre-pregnancy BMI having an effect size of zero (Table 34).
Associations between Physical Activity and Child IQ
There was a statistically significant positive association between time spent in VPA and
child IQ (b=7.05; p=0.003), with 8% of the variance in child IQ being described by time
spent in VPA (Table 33). Adjusting for child age, sex, BMI z-score, and SES did not
change this association (b=5.79; p=0.021). The association remained significant after
further adjusting for maternal pre-pregnancy BMI (b=5.03; p=0.047). Additionally, to
visually demonstrate this significant association, least-square means (LSmean) of IQ
scores below and above the median of time spent in VPA were compared using
analysis of covariance, adjusted for child age, sex, BMI z-score, SES, and pre-
pregnancy BMI (Figure 9). Children above the median of reported time spent in VPA
had significantly higher IQ scores than children below the median of reported time spent
in VPA (LSmean above the median: 110.05, SE=2.18; LSmean below the median:
88
103.39, SE=1.85; p=0.023). Their IQ scores were 6% higher than children who reported
less than 10 minutes of VPA.
There was a significant interaction between children exposed to maternal
overweight/obesity in utero and high physical activity levels with child IQ (p=0.013).
Stratified by maternal overweight/obese and normal-weight exposed, children exposed
to maternal overweight/obese who reported above the median time of VPA, had
significantly higher IQ scores than children exposed to maternal overweight/obese who
reported below the median (LSmean above the median: 111.21, SE=2.81; LSmean
below the median: 102.77, SE=2.11; p=0.021), while children whose mothers were
normal-weight before pregnancy who reported above or below the median time of VPA,
had no significant difference in IQ scores (LSmean above the median: 109.73, SE=3.47;
LSmean below the median: 103.97, SE=4.54; p=0.29).
Before adjusting for any covariates, there was not a significant association
between time spent in MVPA and child IQ (b=2.63; p=0.16). After adjusting for child
age, sex, BMI z-score, SES and maternal pre-pregnancy BMI, there was a significant
association between time spent in MVPA and child IQ (b=3.90; p=0.035) (Table 33).
Associations between Physical Activity and Global FA
There was a statistically significant positive association between time spent in VPA and
global FA (b=0.005; p=0.017), with 5% of the variance in global FA related to time spent
in VPA (Table 34). Adjusting for child age, sex, BMI z-score, and SES did not change
this association (b=0.006; p=0.005). The association remained significant after further
89
adjusting for maternal pre-pregnancy BMI (b=0.006; p=0.005). Additionally, to visually
demonstrate this significant association, least-square means (LSmean) of global FA
below and above the median of time spent in VPA were compared using analysis of
covariance, adjusted for child age, sex, BMI z-score, SES, and pre-pregnancy BMI
(Figure 10). Children above the median of reported time spent in VPA had significantly
greater global FA than children below the median of reported time spent in VPA
(LSmean above the median: 0.423, SE=0.002; LSmean below the median: 0.416,
SE=0.002; p=0.019). There was a significant interaction between children exposed to
maternal overweight/obesity in utero and high physical activity levels with global FA
(p=0.015). Stratified by maternal overweight/obese and normal-weight exposed,
children exposed to maternal overweight/obese who reported above the median time of
VPA, had significantly greater global FA than children who reported below the median
time of VPA (LSmean above the median: 0.425, SE=0.002; LSmean below the median:
0.415, SE=0.002; p=0.002), while children whose mothers were normal-weight before
pregnancy who reported above or below the median reported time of VPA, had no
significant difference in global FA (LSmean above the median: 0.417, SE=0.005;
LSmean below the median: 0.419, SE=0.002; p=0.66). Before and after adjusting for
covariates, time spent in MVPA was not significantly associated with global FA p>0.05
(Table 34).
Results of TBSS Analyses
90
In children who reported above the median of time in VPA, there were significant
clusters of greater FA in the left and right superior longitudinal fasciculus (SLF), right
anterior thalamic radiation (ATR), left inferior fronto-occipital fasciculus (IFOF) and left
inferior longitudinal fasciculus (ILF) (Table 35). After controlling for maternal pre-
pregnancy BMI, significant clusters remained in the left and right SLF and right ATR
(Figure 11; Table 36). There were no significant clusters associated with maternal pre-
pregnancy BMI. There was a large significant cluster that covered the majority of the
skeleton where there was an interaction between pre-pregnancy BMI weight categories
and vigorous physical activity above and below the median (Table 37). After increasing
the significance threshold to p<0.01, significant clusters remained in the left forceps
major, left ATR, right cingulate gyrus, right IFOF, and cingulum (hippocampal portion)
(Table 38).
Discussion
This is the first study in humans to demonstrate the relationship between child
physical activity levels with neurocognition in a cohort of children exposed to varying
levels of maternal obesity. Independent of maternal pre-pregnancy BMI, time spent in
VPA was significantly associated with higher IQ scores and greater global FA in
children. Additionally, there were significant interactions between maternal weight status
and vigorous physical activity levels on both child IQ, global FA and various white
matter tracts. These findings are hopeful and suggest that children who are physically
91
active can potentially be less impacted by the effects of prenatal exposure to maternal
obesity.
Similar to epidemiological studies, we found that children exposed to higher
levels of maternal obesity, had lower IQ scores
78,80
. This is the first study to
demonstrate this trend in a cohort with well characterized maternal pre-pregnancy BMI
and a thorough assessment of IQ scores. While we found a marginal relationship
between maternal pre-pregnancy BMI and child IQ, a larger sample size is needed to
confirm these findings.
Conversely, we found that children who engaged in both moderate and vigorous
physical activity, had significantly higher IQ scores, independent of maternal pre-
pregnancy BMI. Moreover, children who reported above 10 minutes of vigorous physical
activity, had significantly higher IQ scores, even after adjusting for several confounding
variables. These findings suggest that an average of 10 minutes of vigorous physical
activity a day could potentially be associated with better cognitive performance.
However, these findings should be interpreted with caution, given the correlational
nature of this study.
We also found that children who reported more than 10 minutes of vigorous
physical activity also had significantly greater global FA and significant clusters of FA in
the left and right superior longitudinal fasciculus and right anterior thalamic radiation,
independent of maternal pre-pregnancy BMI. Similarly, other studies have shown that
physical activity is associated with greater FA in the superior longitudinal fasciculus
372,376
. The superior longitudinal fasciculus is an important white matter tract known for
many aspects of cognition such as attention, IQ, and language abilities
377–379
.
92
Therefore, strengthening of the superior longitudinal fasciculus, particularly the arcuate
fasciculus, may contribute to increased IQ scores in children who engage in vigorous
physical activity. Future studies with a larger sample size should explore if greater FA in
the SLF contributes to increased IQ as a result of vigorous physical activity.
A wealth of data has found that physical activity intervention studies can
effectively improve cognition across the lifespan and strengthen white matter pathways
specifically during childhood
113,361–365,367,371,372,380
. Potential mechanisms by which
physical activity improves cognition include, increasing brain-derived neurotrophic factor
(BDNF) levels and neurogenesis in the hippocampus
68,113,381–384
, a brain region
important for many aspects of learning and memory, and through strengthening of white
matter pathways
371,372
. Indirect physiological pathways that possibly contribute to
improved cognition include, decreasing circulating inflammatory cytokines
330
, and
increasing insulin sensitivity both peripherally and in the brain
368,385
. The increased
presence of pro-inflammatory cytokines has been shown to impact cognition even
during childhood
386,387
. Additionally, optimal insulin signaling has a pivotal role in brain
function
231,388
.
Interestingly, a handful of studies has found that physical activity contributes to
whole-body insulin sensitivity and improved neurocognition in a dose-dependent
manner with vigorous physical activity having larger effects
372,383,389–393
. Our results
also suggest that physical activity intensity is related to neurocognition during childhood.
However, the majority of studies in children have assessed moderate to vigorous
physical activity rather than vigorous physical activity specifically
363,371,394,395
. Therefore,
93
future studies are needed to confirm the type of physical activity most beneficial for
neurocognition during childhood.
Moreover, we observed a significant interaction between maternal weight status
and vigorous physical activity levels. In stratified analyses, we found that children
exposed to maternal overweight/obesity in utero who engaged in more than 10 minutes
of vigorous physical activity had significantly higher IQ scores, and greater global FA
compared to children exposed to maternal overweight/obesity in utero who reported
less than 10 minutes of vigorous physical activity. Children exposed to normal-weight
mothers had a similar trend but no significant group differences. Additionally, a large
interaction was seen in various white matter pathways, particularly the hippocampal
portion of the cingulum. Interestingly, prior studies have shown that prenatal exposure
to maternal obesity is associated with reduced hippocampal volume in both human and
rodent studies
8,70,396
. Therefore, differences in white matter properties in the
hippocampal portion of the cingulum may be related to prenatal exposure to maternal
obesity. Given the significant interaction observed between maternal obesity and
vigorous physical activity levels, future longitudinal studies or intervention studies are
needed to determine if children exposed to maternal obesity in utero may particularly
benefit from engaging in vigorous physical activity.
Strengths and Limitations
Our study design had many strengths, however there were some limitations to
keep in mind. While we used the gold standard for assessing IQ, and an objective
94
measure of maternal pre-pregnancy BMI from EMR, we used a self-report for assessing
physical activity rather than an objective measure such as accelerometers. As
previously mentioned, the self-report is subject to bias and it is possible, participants
underreported or over reported their PA levels. Children in our cohort reported similar
time spent in VPA to the average of 13 minutes per a day reported by NHANES
397
. A
strength of using self-report to assess VPA, is that children who participate in team
sports or water activities can report their time spent in these vigorous physical activities.
To elaborate, an objective wrist accelerometer may need to be removed during team
sports for safety concerns, and therefore, these children’s vigorous physical activity
would not be accounted for. In contrast, a limitation of this study design, is that our
findings were correlational between physical activity levels and neurocognition. An
intervention or longitudinal study is needed to support the evidence that physical activity
can benefit neurocognition in children exposed to maternal obesity in utero.
Conclusions
Time spent in VPA was associated with higher IQ scores and greater global FA,
independent of maternal pre-pregnancy BMI. Additionally, there was a significant
interaction between maternal weight status and physical activity levels on child IQ,
global FA and various white matter pathways. These findings suggest that physical
activity may potentially be protective against the effects of prenatal exposure to
maternal obesity on child cognition and white matter myelination. This study is an
important first step in determining if physical activity is a potential strategy to ameliorate
95
the adverse consequences of prenatal exposure to maternal obesity on child
neurocognition. Future studies are needed to determine the mechanism by which
physical activity promotes better cognitive performance and to determine if improved
white matter myelination contributes to higher IQ scores.
Chapter Five: Concluding Remarks and Future Directions
This dissertation project started out to recapitulate findings in animals that
demonstrated that in utero exposure to gestational diabetes preferentially impacted
hippocampal structure and function. However, in our cohort, in utero exposure to GDM
did not negatively influence offspring, but rather prenatal exposure to maternal obesity.
Moreover, two themes became clear from this project. The first being that there are sex
differences in the associations of prenatal exposures, potentially through developmental
programming mechanisms that were beyond the scope of this dissertation. The second
being that physical activity has the potential to be a therapeutic target for protecting
against the deleterious effects of maternal obesity on child metabolic health and
neurocognition. And lastly, we found that similar to prior studies, prenatal exposure to
maternal obesity was associated with increased child adiposity, reduced insulin
sensitivity (not independent of child adiposity), and reduced IQ scores.
The novel findings were that prenatal exposure to maternal obesity was
associated with reduced hippocampal volume (driven by boys), physical activity was
associated with greater IQ scores, greater global FA, greater FA in various white matter
tracts and the relationship between physical activity and neurocognition was
96
significantly different between children exposed to mothers who were overweight/obese
during pregnancy compared to mothers who were normal-weight. And lastly, compared
to prenatal exposures (maternal obesity and GDM), physical activity has a stronger
impact on insulin sensitivity during childhood. Importantly, we also found that prenatal
exposure to maternal obesity or GDM is not associated with hippocampal-related
memory deficits.
From this dissertation and the publications in the process of being disseminated
to the scientific community
70,398
, the goal was to communicate the importance of
considering sex differences in addressing research questions, secondly, the importance
of studying prenatal exposures (maternal obesity/GDM) on child metabolic health and
neurocognition and lastly, the therapeutic potential of physical activity to ameliorate the
effects of prenatal exposures to maternal obesity/GDM. To further test this possibility,
longitudinal studies/intervention studies are needed.
Additionally, some future statistical analyses would be warranted with a larger
sample size. To start with, it would be important to consider if the relationship between
physical activity and insulin sensitivity differs among prenatal exposure levels. If there
was a significant interaction and the relationship was stronger among children exposed
to maternal obesity or GDM, this would suggest that these children may particularly
benefit from engaging in physical activity.
As previously mentioned, there was a significant difference between GDM-
exposed offspring and non-exposed offspring in central adiposity measurements and
caloric intake. It would be interesting to determine if greater caloric intake is related to
greater central adiposity in GDM-exposed children. This would suggest that GDM-
97
exposed children could benefit from nutritional counseling and downstream this may
help reduce their risk of Type 2 diabetes given the health implications of excess caloric
intake and central adiposity.
Another analysis to consider would be a mediation analysis to determine if
greater FA in the SLF mediates the relationship between physical activity and greater IQ
scores. This would provide a mechanistic link into how greater physical activity
contributes to higher IQ scores. The last analysis to consider would be how paternal
BMI relates to child IQ. This may provide further evidence of developmental
programming if paternal BMI is unrelated to child IQ. While we do not have paternal BMI
during the time of pregnancy beyond self-report, many fathers to do accompany their
children to the BrainChild visit. Therefore, we have been able to collect anthropometric
measurements from a subset of fathers.
Overall, the BrainChild study has instigated some very exciting changes to
address some of the initial limitations presented in these chapters. During my time in the
lab, we started using accelerometers as an objective measure to assess physical
activity and time spent in sedentary behavior. This is important because accelerometers
allow us to look at the relationship between physical activity and child IQ from a different
perspective. Accelerometers capture bodily movements while the 3DPAR captures
perceived physical exertion. We additionally started measuring central adiposity using
the gold standard of magnetic resonance imaging. This will help determine if GDM-
exposed children have increased visceral adipose tissue or simply increased
subcutaneous adipose tissue compared to nonexposed children. And importantly,
children have started to come in for multiple longitudinal visits. Longitudinal timepoints
98
are critical during childhood when both metabolic and neurocognitive changes occur
exponentially, especially as children transition through puberty. Not only do longitudinal
timepoints provide evidence of behaviors predictive of numerous health outcomes, they
allow us to ascertain whether relationships observed at baseline will persist. I am
incredibly grateful to have been a part of a project that is doing so many amazing things,
and I am excited to see what is yet to come.
Methods
Overview of BrainChild Participants
Children between the ages of 7 to 11 years old were recruited from Kaiser Permanente
Southern California (KPSC), to participate in the BrainChild Study. KPSC is a large
health care organization that uses an integrated electronic medical record (EMR)
system. BrainChild studies the impact of intrauterine exposure to metabolic disorders on
brain pathways during childhood. Additionally, the BrainChild cohort is a unique cohort
of children whose mothers had well-documented glucose levels during their
pregnancies. All children were born at KPSC. A subset of participants completed
cognitive measures and a magnetic resonance imaging (MRI) during visit 2. Children
with the following characteristics were excluded: neurological, psychiatric, or other
significant medical disorders, including diabetes; use of medications known to alter
metabolism (i.e., glucocorticoids), and who had contraindications to MRI (i.e.,
permanent metal, claustrophobia, or left-handedness. Each participating Institutional
99
Review Board approved this study (University of Southern California # HS-14-00034
and KPSC # 10282). Participants’ parents gave written informed consent and children
provided written informed assent.
Anthropometric measures and Body Composition during Study Visit
During the first study visit, each child’s height and weight was measured to the nearest
0.1 cm and 0.1 kg, using a portable stadiometer and medical scale. BMI was calculated
using the standard formula, weight in kilograms divided by height in meters squared
(m
2
). BMI z-scores and BMI percentiles were calculated using the age and sex-specific
Centers for Disease Control and Prevention growth charts
24
. Waist and hip
circumferences were measured to the nearest 0.1 cm. Waist circumference was
measured at the midpoint between the iliac crest and lower costal margin in the
midaxillary line. Hip circumference was measured at the maximum width of the
buttocks. Waist-to-height-ratio (WHtR) was calculated by dividing waist circumference
by height. In place of using a computerized tomography scanner or magnetic resonance
imaging, to measure central body fat distribution, WHtR was used. While WHtR is not
the standard measure for determining central body fat distribution, it is a valid and
practical surrogate measure of central body fat distribution in children, and has been
shown to adequately predict levels of adipose tissue in the abdominal region
399,400
.
Total percent body fat was measured using bio-electrical impedance analysis (BIA),
which is another practical alternative to whole body imaging
401
. BIA was completed
using Tanita Body Composition Analyser SC 331S. The Tanita scale uses the frequency
100
current, 50 kHz, 90 μA. BIA has been validated with other referenced methodologies to
assess total percent body fat in children
401
. Further, it has been used in children
151,152,401
. Additionally, children also completed an oral glucose tolerance test (OGTT)
during this visit. Plasma glucose and insulin were assayed from blood samples collected
at 0, 30, 60, 90 and 120 minutes before and after the participant consumed a glucose
load (1.75 g/kg; maximum dose of 75 grams) (Glucola, Azer Scientific, Morgantown,
PA). Plasma glucose was measured using the enzymatic electrode YSI analyzer
machine, YSI 2300 STAT PLUS (YSI incorporated, Yellow Springs, OH). Plasma insulin
was measured using a human insulin ELISA kit (Millipore, Billerica, MA). Insulin
sensitivity was estimated using the Matsuda insulin sensitivity index (Figure 1)
354,402
.
Physical Activity Assessment
Physical activity was assessed using the self-reported, three-day physical activity report
(3DPAR)
403
. A trained staff member asked participants, with the input of their parents,
to recall their activities from 7:00am to 12:00am in 30-minute increments for the
previous three days. Activities were classified based on a reference sheet with 73
activities and classified based on the activities that best matched the participant’s
response. The participant then was asked to rate the intensity of each activity; the
intensity levels ranged from light, moderate, hard, to very hard. The activities were
categorized as either moderate to vigorous physical activity (MVPA), vigorous physical
activity (VPA) or sedentary behavior (SB) based on metabolic equivalents (MET)
determined by the compendium of energy expenditure
403,404
. Activities with METs ≥ 3
101
were classified as MVPA, METs ≥ 6 were classified as VPA. Non-sleep activities with
METs ≤ 1.5 and > 1.0 were classified as SB. Time spent sleeping or napping, was
categorized as METs = 1.0. Examples of moderate to vigorous physical activities
include bike riding or swimming. Examples of SB included watching television or playing
video games. The final output was the number of 30-minute increments spent in either
MVPA or in SB converted to average minutes per a day spent in MVPA or SB.
Diet Intake Assessment
Diet was assessed using two dietary recalls, which were performed during in-person
visits as part of the BrainChild Study. Using the multi-pass method for dietary recall, a
trained staff member asked the participant to recall what food and beverages they had
consumed over a 24-hour time period with the input of the participant’s parent. The
trained staff member then went through 3 other “passes” to complete quantity of
food/beverages consumed as well as to include missing or forgotten food/beverages.
Use of the multi-pass 24-hour dietary recall method is a valid method to assess energy
intake in children
405
. Once the dietary recalls were collected, the recalls were analyzed
using the Nutritional Data System for Research software v.2018 developed by the
Nutrition Coordinating Center (NCC), University of Minnesota, Minneapolis, MS
406
. The
variables used were daily energy intake (EI) in calories and percent calories from added
sugar by available carbohydrate.
Assessment of Pubertal Development
102
Each child was given the option of having their pubertal Tanner stage of development
assessed by a physical exam
407,407
and/or by a validated sex-specific assessment
questionnaire for children and parents, which contains both illustrations and explanatory
text about the varying Tanner stages
408
.
Cognitive Measures
Hippocampal-Dependent Memory Task
A modified version of the relational task created by Watson et al. (2013) was used to
assess hippocampal-dependent relational memory
177
. This relational task has also
been used in children ages 7-11
16,17
. During the relational memory task, stimuli are
presented on a computer screen and the children are instructed to pay attention to the
location of the stimuli. The stimuli consist of five child-friendly animated creatures (see
Figure 5). The creatures then disappear and reappear at the top of the screen. They
are then instructed to reconstruct the image they were previously shown of the
creatures, using a mouse to click on and drag each creature back to its original location.
The children have as much time as they need to reconstruct the image they were
previously shown. After each trial, they press the space bar to continue to the next trial.
The children have 3 practice trials before proceeding to the experimental trials. In the
experimental trials, there are 2 blocks. Each block has 12 trials. Each trial shows the
creatures for 2000 MS and has a waiting period of 4000 MS before the creatures
103
appear on the top of the screen. Presentation v15.1 software (NeuroBS) is used to
operate the memory task on a laptop.
The task is scored using the metric, “accurate object-location binding”. This
metric is calculated from algorithms created by Horecka et al.
172
. A simplified
explanation is as follows: The global error is subtracted from a composite misplacement
measure to identify placement of an object within a predetermined accuracy “window”.
Misplacement is the distance from where a creature was placed during the
reconstruction and its original location. Accurate object-location binding thus is a
measure of general accuracy of the object being placed in the original study location
accounting for global error attributed to the overall shape difference from the original
study location. For example, if all the objects were shifted in the x-axis, the global error
would account for this and not use it as a penalty if overall the shape was slightly
different from the original study location. Further, a higher score, indicates better
accuracy. Please see Horecka et al.
172
for more details.
Hippocampal-independent Memory Task
The hippocampal-independent memory task also known as the item-familiarity task
measures accuracy in item-familiarity, a function independent of the hippocampus
17,409
.
In part one of the task, the children are shown a series of child-friendly creatures and
asked to decide if each creature is a boy or a girl. In part two, they are shown a series of
sets of three creatures. Two creatures are novel, and one they have seen before. They
are then asked to recall which creature they were previously shown.
104
Picture-sequence Memory Task
The picture-sequence memory task is a 5-minute episodic memory task completed on
an iPad tablet. It is a part of the National Institutes of Health cognitive battery toolbox.
The NIH toolbox is a highly validated instrument for assessing cognition across the
lifespan from ages 3 to 85
410
. During the picture-sequence memory task, participants
are shown a series of images that depict a story. After being shown the story, the
participant is asked to reconstruct the story to the best of their ability. It is scored on
accuracy of reconstruction. There is both an age adjusted score, and a demographically
adjusted score. The age adjusted score has a normative mean of 100 and a standard
deviation of 15, meaning in the prior validating sample, the average score was 100 after
accounting for age differences. The demographically adjusted score accounts for race,
gender, age, and family educational attainment, with a normative mean of 50 and a
standard deviation of 10.
General Cognitive Function
Considering performance on memory tasks can also be the result of baseline
Intelligence Quotient (IQ), the shortened 2
nd
edition of the Wechsler Abbreviate Scale
for Intelligence for children (WASI-II) was used to assess full-scale IQ and general
cognitive abilities
411
. The WASI-II is well established for measuring full scale IQ and is
highly validated for ages 6 to 80
411
. It consists of two parts, vocabulary and matrix
105
reasoning. The vocabulary part asks participants to describe words to the best of their
ability. The matrix reasoning consists of a series that the participant is asked to
complete.
MRI Methods
During the second visit, after a training session in the mock scanner, an MRI was
performed using a Siemens MAGNETOM Prisma
fit
3T MRI scanner (Siemens Medical
Systems). The MRI session started with a localizer scan. A high-resolution magnetic
resonance imaging scan was then acquired using a T1-weighted three-dimensional
magnetization prepared rapid gradient echo (MP-RAGE) sequence with the following
parameters: 256 x 256 x176 matrix size with 1 x 1 x 1 mm
3
resolution; inversion
time=900 ms; repetition time (TR)=1950 ms; echo time (TE)=2.26 ms; flip angle=90°.
The total scan duration was 4 minutes and 14 seconds.
A diffusion weighted image was acquired using a dual spin echo, single shot,
pulsed gradient, echo planar imaging sequence in 64 diffusion sensitized gradient
directions with the following parameters: TR=8100 ms; TE=69 ms; flip angle=90°; 70
axial slices; 2 x 2 x 2-mm
3
voxel size; FOV=256 mm; b value=1000s/mm
2
; Total scan
duration was 9 minutes and 29 seconds.
MRI Analysis
106
The T1 MP-RAGE structural image was imported into the automated segmentation
software, FreeSurfer version 6.0 (http://surfer.nmr.mgh.harvard.edu/) in order to
examine total hippocampal grey matter volume and total grey matter volume in the
hippocampal sub-regions known as subfields (mm
3
)
97,98
. The processing steps that
FreeSurfer uses to calculate grey matter volume (mm
3
) involve, first removing non-brain
tissue from the T1 image. Then, an atlas generated by FreeSurfer is used to identify the
subcortical regions of the brain. The grey matter is then differentiated from white matter
tissue using intensity values provided by the T1 image. The white matter is then
segmented from the grey matter tissue. After, the surface tissue is labeled based on a
surface atlas. It is then extracted from the subcortical regions. Then lastly, the gyri are
labeled and two files with the subcortical and cortical measurements are generated. An
additional processing step is used to calculate grey matter volume in each hippocampal
subfield. This process includes using a probabilistic atlas created from ex-vivo tissue
captured at 0.13 isotropic resolution to identify and segment hippocampal subfield
boundaries based on Bayesian inference. The corresponding output, FS60, was used.
It is a hippocampal proper parcellation with no head/body subdivisions. Twelve subfield
volumes are generated by FreeSurfer 6.0, however, I only included subfields that have
been shown to be preferentially affected by prenatal exposures (i.e. gestational
diabetes, prenatal stress) including the CA1, CA2/3, CA4, DG (granule cell layer) and
subiculum
8,85,195,199,201,412–415
. After each T1 image was processed through FreeSurfer,
manual quality check of the automated hippocampal segmentation was performed for
each participant following an existing protocol
99
. The segmentation of the hippocampus
was visually assessed and then given a rating of “pass”, “pass on condition” and “fail”.
107
Images that failed to have defined landmarks due to motion artifacts or segmentation
error were excluded. Additionally, FreeSurfer has been shown to accurately segment
the hippocampus and hippocampal subfields, in children aged 7 to 11 years
105,106
.
Tract-Based Spatial Statistics (v1.2)
416
, a part of FSL (FMRIB Software Library)
417
was used to complete a voxel-wise analysis of fractional anisotropy data. Diffusion-
weighted images were fitted to create fractional anisotropy (FA) images using the tensor
model function, FDT. FA images were then skull-stripped using the brain extraction tool,
BET
418
, motion corrected and registered to the T1 image using FSL’s FLIRT
419
. A
mask was created of combined FA images, and then a mask with a skeleton of voxels
with FA values >0.2 was created. Mean FA for each subject was then extracted to
compare global FA across subjects. And lastly, voxel-wise statistics were completed to
compare each subject’s aligned mask to the skeletonized mask to determine significant
clusters of voxels using FSL’s randomise tool which corrects for multiple comparisons
using the threshold-free cluster enhancement option with 5000 permutations. The Johns
Hopkins University white matter tractography atlas was used to identify the location of
significant clusters
420
.
Prenatal Exposures: Maternal pre-pregnancy BMI and Gestational Diabetes Mellitus
Maternal pre-pregnancy BMI (kg/m
2
) was calculated from maternal height (cm) and
weight (kg) measurements closest to last menstrual period within 180 days from the
electronic medical record (EMR). Also using the EMR, each mother’s GDM status was
extracted. Diagnosis of GDM was based on laboratory glucose values confirming a
108
plasma glucose level ≥ 200mg/dl from a 50-g glucose challenge tests or at least 2
plasma glucose values meeting or exceeding the following values on the 100-g or 75-g
oral glucose tolerance test: fasting, 95 mg/dL; 1 hour, 180 mg/dL; 2 hours, 155 mg/dL;
and 3 hours, 140 mg/dL.
References
1. de Rooij SR, Caan MWA, Swaab DF, et al. Prenatal famine exposure has sex-
specific effects on brain size. Brain. 2016;139(8):2136-2142.
doi:10.1093/brain/aww132
2. Lumey L, Stein AD, Kahn HS, Romijn J. Lipid profiles in middle-aged men and
women after famine exposure during gestation: the Dutch Hunger Winter Families
Study. Am J Clin Nutr. 2009;89(6):1737-1743. doi:10.3945/ajcn.2008.27038
3. Roseboom T, de Rooij S, Painter R. The Dutch famine and its long-term
consequences for adult health. Early Hum Dev. 2006;82(8):485-491.
doi:10.1016/j.earlhumdev.2006.07.001
4. Heindel JJ, Vandenberg LN. Developmental Origins of Health and Disease: A
Paradigm for Understanding Disease Etiology and Prevention. Curr Opin Pediatr.
2015;27(2):248-253. doi:10.1097/MOP.0000000000000191
5. Rinaudo P, Wang E. Fetal Programming and Metabolic Syndrome. Annu Rev
Physiol. 2012;74(1):107-130. doi:10.1146/annurev-physiol-020911-153245
6. Bayol SA, Farrington SJ, Stickland NC. A maternal “junk food” diet in pregnancy
and lactation promotes an exacerbated taste for “junk food” and a greater
propensity for obesity in rat offspring. Br J Nutr. 2007;98(4):843-851.
doi:10.1017/S0007114507812037
7. Ong ZY, Muhlhausler BS. Maternal “junk-food” feeding of rat dams alters food
choices and development of the mesolimbic reward pathway in the offspring.
FASEB J. 2011;25(7):2167-2179. doi:10.1096/fj.10-178392
8. Tozuka Y, Wada E, Wada K. Diet-induced obesity in female mice leads to
peroxidized lipid accumulations and impairment of hippocampal neurogenesis
during the early life of their offspring. FASEB J. 2009;23(6):1920-1934.
doi:10.1096/fj.08-124784
9. Catalano PM, Farrell K, Thomas A, et al. Perinatal risk factors for childhood obesity
and metabolic dysregulation. Am J Clin Nutr. 2009;90(5):1303-1313.
doi:10.3945/ajcn.2008.27416
109
10. Ogden CL, Carroll MD, Curtin LR, Lamb MM, Flegal KM. Prevalence of high body
mass index in US children and adolescents, 2007-2008. Jama. 2010;303(3):242–
249.
11. Bider-Canfield Z, Martinez MP, Wang X, et al. Maternal obesity, gestational
diabetes, breastfeeding and childhood overweight at age 2 years: Maternal
exposures and childhood overweight. Pediatr Obes. 2017;12(2):171-178.
doi:10.1111/ijpo.12125
12. Reilly JJ, Armstrong J, Dorosty AR, et al. Early life risk factors for obesity in
childhood: cohort study. BMJ. 2005;330(7504):1357.
doi:10.1136/bmj.38470.670903.E0
13. Salsberry PJ. Dynamics of Early Childhood Overweight. PEDIATRICS.
2005;116(6):1329-1338. doi:10.1542/peds.2004-2583
15. Ludwig DS. Epidemic Childhood Obesity: Not Yet the End of the Beginning.
Pediatrics. 2018;141(3):e20174078. doi:10.1542/peds.2017-4078
16. Hassevoort KM, Khazoum SE, Walker JA, et al. Macular Carotenoids, Aerobic
Fitness, and Central Adiposity Are Associated Differentially with Hippocampal-
Dependent Relational Memory in Preadolescent Children. J Pediatr.
2017;183:108-114.e1. doi:10.1016/j.jpeds.2017.01.016
17. Khan NA, Baym CL, Monti JM, et al. Central Adiposity Is Negatively Associated
with Hippocampal-Dependent Relational Memory among Overweight and Obese
Children. J Pediatr. 2015;166(2):302-308.e1. doi:10.1016/j.jpeds.2014.10.008
18. Must A, Strauss R. Risks and consequences of childhood and adolescent obesity.
Published online 1999:10.
19. Pan L, McGuire LC, Blanck HM, May-Murriel AL, Grummer-Strawn LM.
Racial/Ethnic Differences in Obesity Trends Among Young Low-Income Children.
Am J Prev Med. 2015;48(5):570-574. doi:10.1016/j.amepre.2014.11.009
20. Peeters A, Barendregt JJ, Willekens F, et al. Obesity in Adulthood and Its
Consequences for Life Expectancy: A Life-Table Analysis. Ann Intern Med.
2003;138(1):24. doi:10.7326/0003-4819-138-1-200301070-00008
21. Spruijt-Metz D. Etiology, Treatment, and Prevention of Obesity in Childhood and
Adolescence: A Decade in Review: ETIOLOGY, TREATMENT, AND
PREVENTION OF OBESITY IN CHILDHOOD AND ADOLESCENCE. J Res
Adolesc. 2011;21(1):129-152. doi:10.1111/j.1532-7795.2010.00719.x
22. Yau PL, Kang EH, Javier DC, Convit A. Preliminary evidence of cognitive and brain
abnormalities in uncomplicated adolescent obesity: Brain Alterations in
Adolescent Obesity. Obesity. 2014;22(8):1865-1871. doi:10.1002/oby.20801
110
23. Herrera BM, Lindgren CM. The Genetics of Obesity. Curr Diab Rep.
2010;10(6):498-505. doi:10.1007/s11892-010-0153-z
24. CDC. About Child & Teen BMI | Healthy Weight. Published October 24, 2018.
Accessed November 27, 2018.
https://www.cdc.gov/healthyweight/assessing/bmi/childrens_bmi/about_childrens_
bmi.html
25. Hales CM, Fryar CD, Carroll MD, Freedman DS, Ogden CL. Trends in Obesity and
Severe Obesity Prevalence in US Youth and Adults by Sex and Age, 2007-2008
to 2015-2016. JAMA. 2018;319(16):1723. doi:10.1001/jama.2018.3060
26. Taveras EM, Gillman MW, Kleinman KP, Rich-Edwards JW, Rifas-Shiman SL.
Reducing Racial/Ethnic Disparities in Childhood Obesity: The Role of Early Life
Risk Factors. JAMA Pediatr. 2013;167(8):731.
doi:10.1001/jamapediatrics.2013.85
28. Tam WH, Ma RCW, Ozaki R, et al. In Utero Exposure to Maternal Hyperglycemia
Increases Childhood Cardiometabolic Risk in Offspring. Diabetes Care.
2017;40(5):679-686. doi:10.2337/dc16-2397
29. Buchanan TA, Xiang AH, Page KA. Gestational diabetes mellitus: risks and
management during and after pregnancy. Nat Rev Endocrinol. 2012;8(11):639-
649. doi:10.1038/nrendo.2012.96
30. Clausen TD, Mathiesen ER, Hansen T, et al. High Prevalence of Type 2 Diabetes
and Pre-Diabetes in Adult Offspring of Women With Gestational Diabetes Mellitus
or Type 1 Diabetes: The role of intrauterine hyperglycemia. Diabetes Care.
2008;31(2):340-346. doi:10.2337/dc07-1596
31. Kubo A, Ferrara A, Windham GC, et al. Maternal Hyperglycemia During Pregnancy
Predicts Adiposity of the Offspring. Diabetes Care. 2014;37(11):2996-3002.
doi:10.2337/dc14-1438
32. Page KA, Romero A, Buchanan TA, Xiang AH. Gestational Diabetes Mellitus,
Maternal Obesity, and Adiposity in Offspring. J Pediatr. 2014;164(4):807-810.
doi:10.1016/j.jpeds.2013.11.063
33. Ferrara A. Increasing Prevalence of Gestational Diabetes Mellitus: A public health
perspective. Diabetes Care. 2007;30(Supplement 2):S141-S146.
doi:10.2337/dc07-s206
34. Steculorum SM, Bouret SG. Maternal Diabetes Compromises the Organization of
Hypothalamic Feeding Circuits and Impairs Leptin Sensitivity in Offspring.
Endocrinology. 2011;152(11):4171-4179. doi:10.1210/en.2011-1279
111
35. Xiang AH, Li BH, Black MH, et al. Racial and ethnic disparities in diabetes risk after
gestational diabetes mellitus. Diabetologia. 2011;54(12):3016-3021.
doi:10.1007/s00125-011-2330-2
36. Dabelea D. The Predisposition to Obesity and Diabetes in Offspring of Diabetic
Mothers. Diabetes Care. 2007;30(Supplement 2):S169-S174. doi:10.2337/dc07-
s211
37. Gillman MW, Oakey H, Baghurst PA, Volkmer RE, Robinson JS, Crowther CA.
Effect of Treatment of Gestational Diabetes Mellitus on Obesity in the Next
Generation. Diabetes Care. 2010;33(5):964-968. doi:10.2337/dc09-1810
38. Sauder KA, Hockett CW, Ringham BM, Glueck DH, Dabelea D. Research:
Epidemiology Fetal overnutrition and offspring insulin resistance and β-cell
function: the Exploring Perinatal Outcomes among Children (EPOCH) study.
Diabet Med J Br Diabet Assoc. 2017;34(10):1392-1399. doi:10.1111/dme.13417
40. HAPO Study Cooperative Research Group. Hyperglycaemia and Adverse
Pregnancy Outcome (HAPO) Study: associations with maternal body mass index:
HAPO - BMI and perinatal outcomes. BJOG Int J Obstet Gynaecol.
2010;117(5):575-584. doi:10.1111/j.1471-0528.2009.02486.x
41. Maftei O, Whitrow MJ, Davies MJ, Giles LC, Owens JA, Moore VM. Maternal body
size prior to pregnancy, gestational diabetes and weight gain: associations with
insulin resistance in children at 9–10 years. Diabet Med. 2015;32(2):174-180.
doi:10.1111/dme.12637
42. Mingrone G, Manco M, Mora MEV, et al. Influence of Maternal Obesity on Insulin
Sensitivity and Secretion in Offspring. Diabetes Care. 2008;31(9):1872-1876.
doi:10.2337/dc08-0432
43. Hochner H, Friedlander Y, Calderon-Margalit R, et al. Associations of Maternal Pre-
Pregnancy Body Mass Index and Gestational Weight Gain with Adult Offspring
Cardio-Metabolic Risk Factors: The Jerusalem Perinatal Family Follow-up Study.
Circulation. 2012;125(11):1381-1389.
doi:10.1161/CIRCULATIONAHA.111.070060
45. Smith J, Cianflone K, Biron S, et al. Effects of Maternal Surgical Weight Loss in
Mothers on Intergenerational Transmission of Obesity. J Clin Endocrinol Metab.
2009;94(11):4275-4283. doi:10.1210/jc.2009-0709
46. Ravelli AC, van der Meulen JH, Osmond C, Barker DJ, Bleker OP. Obesity at the
age of 50 y in men and women exposed to famine prenatally. Am J Clin Nutr.
1999;70(5):811-816. doi:10.1093/ajcn/70.5.811
47. Rees S, Harding R, Walker D. The biological basis of injury and neuroprotection in
the fetal and neonatal brain. Int J Dev Neurosci. 2011;29(6):551-563.
doi:10.1016/j.ijdevneu.2011.04.004
112
48. Adane AA, Mishra GD, Tooth LR. Maternal pre-pregnancy obesity and childhood
physical and cognitive development of children: a systematic review. Int J Obes.
2016;40(11):1608-1618. doi:10.1038/ijo.2016.140
49. Basatemur E, Gardiner J, Williams C, Melhuish E, Barnes J, Sutcliffe A. Maternal
Prepregnancy BMI and Child Cognition: A Longitudinal Cohort Study.
PEDIATRICS. 2013;131(1):56-63. doi:10.1542/peds.2012-0788
50. Bolaños L, Matute E, Ramírez-Dueñas M de L, Zarabozo D. Neuropsychological
Impairment in School-Aged Children Born to Mothers With Gestational Diabetes. J
Child Neurol. 2015;30(12):1616-1624. doi:10.1177/0883073815575574
51. Cai S, Qiu A, Broekman BFP, et al. The Influence of Gestational Diabetes on
Neurodevelopment of Children in the First Two Years of Life: A Prospective
Study. Baradaran HR, ed. PLOS ONE. 2016;11(9):e0162113.
doi:10.1371/journal.pone.0162113
52. Casas M, Chatzi L, Carsin A-E, et al. Maternal pre-pregnancy overweight and
obesity, and child neuropsychological development: two Southern European birth
cohort studies. Int J Epidemiol. 2013;42(2):506-517. doi:10.1093/ije/dyt002
53. Hinkle SN, Schieve LA, Stein AD, Swan DW, Ramakrishnan U, Sharma AJ.
Associations between maternal prepregnancy body mass index and child
neurodevelopment at 2 years of age. Int J Obes 2005. 2012;36(10):1312-1319.
doi:10.1038/ijo.2012.143
54. Krakowiak P, Walker CK, Bremer AA, et al. Maternal Metabolic Conditions and
Risk for Autism and Other Neurodevelopmental Disorders. PEDIATRICS.
2012;129(5):e1121-e1128. doi:10.1542/peds.2011-2583
55. Laura Contu, Cheryl Hawkes. A Review of the Impact of Maternal Obesity on the
Cognitive Function and Mental Health of the Offspring. Int J Mol Sci.
2017;18(5):1093. doi:10.3390/ijms18051093
56. Li M, Fallin MD, Riley A, et al. The Association of Maternal Obesity and Diabetes
With Autism and Other Developmental Disabilities. PEDIATRICS.
2016;137(2):e20152206-e20152206. doi:10.1542/peds.2015-2206
57. Linder K, Schleger F, Kiefer-Schmidt I, et al. Gestational Diabetes Impairs Human
Fetal Postprandial Brain Activity. J Clin Endocrinol Metab. 2015;100(11):4029-
4036. doi:10.1210/jc.2015-2692
58. Mina TH, Lahti M, Drake AJ, et al. Prenatal exposure to maternal very severe
obesity is associated with impaired neurodevelopment and executive functioning
in children. Pediatr Res. 2017;82(1):47-54. doi:10.1038/pr.2017.43
113
59. Musser ED, Willoughby MT, Wright S, et al. Maternal prepregnancy body mass
index and offspring attention-deficit/hyperactivity disorder: a quasi-experimental
sibling-comparison, population-based design. J Child Psychol Psychiatry.
2017;58(3):240-247. doi:10.1111/jcpp.12662
60. Perna R, Loughan AR, Le J, Tyson K. Gestational Diabetes: Long-Term Central
Nervous System Developmental and Cognitive Sequelae. Appl Neuropsychol
Child. 2015;4(3):217-220. doi:10.1080/21622965.2013.874951
61. Xiang AH, Wang X, Martinez MP, et al. Association of Maternal Diabetes With
Autism in Offspring. JAMA. 2015;313(14):1425. doi:10.1001/jama.2015.2707
62. Niculescu MD, Lupu DS. High fat diet-induced maternal obesity alters fetal
hippocampal development. Int J Dev Neurosci Off J Int Soc Dev Neurosci.
2009;27(7):627-633. doi:10.1016/j.ijdevneu.2009.08.005
63. Tozuka Y, Kumon M, Wada E, Onodera M, Mochizuki H, Wada K. Maternal obesity
impairs hippocampal BDNF production and spatial learning performance in young
mouse offspring. Neurochem Int. 2010;57(3):235-247.
doi:10.1016/j.neuint.2010.05.015
64. Muhlhausler BS, Ong ZY. The fetal origins of obesity: early origins of altered food
intake. Endocr Metab Immune Disord Drug Targets. 2011;11(3):189-197.
65. Dearden L, Balthasar N. Sexual Dimorphism in Offspring Glucose-Sensitive
Hypothalamic Gene Expression and Physiological Responses to Maternal High-
Fat Diet Feeding. Endocrinology. 2014;155(6):2144-2154. doi:10.1210/en.2014-
1131
66. Edlow AG, Guedj F, Pennings JLA, Sverdlov D, Neri C, Bianchi DW. Males are
from Mars, and females are from Venus: sex-specific fetal brain gene expression
signatures in a mouse model of maternal diet-induced obesity. Am J Obstet
Gynecol. 2016;214(5):623.e1-623.e10. doi:10.1016/j.ajog.2016.02.054
67. Bilbo SD, Tsang V. Enduring consequences of maternal obesity for brain
inflammation and behavior of offspring. FASEB J Off Publ Fed Am Soc Exp Biol.
2010;24(6):2104-2115. doi:10.1096/fj.09-144014
68. Kim T-W, Park H-S. Physical exercise improves cognitive function by enhancing
hippocampal neurogenesis and inhibiting apoptosis in male offspring born to
obese mother. Behav Brain Res. 2018;347:360-367.
doi:10.1016/j.bbr.2018.03.018
69. Nivoit P, Morens C, Van Assche FA, et al. Established diet-induced obesity in
female rats leads to offspring hyperphagia, adiposity and insulin resistance.
Diabetologia. 2009;52(6):1133-1142. doi:10.1007/s00125-009-1316-9
114
70. Alves JM, Luo S, Chow T, Herting M, Xiang AH, Page KA. Sex differences in the
association between prenatal exposure to maternal obesity and hippocampal
volume in children. Brain Behav. Published online January 5, 2020:e01522.
doi:10.1002/brb3.1522
71. Page KA, Luo S, Wang X, et al. Children Exposed to Maternal Obesity or
Gestational Diabetes Mellitus During Early Fetal Development Have Hypothalamic
Alterations That Predict Future Weight Gain. Diabetes Care. 2019;42(8):1473-
1480. doi:10.2337/dc18-2581
72. Brunstrom JM, Burn JF, Sell NR, et al. Episodic Memory and Appetite Regulation in
Humans. Morrison C, ed. PLoS ONE. 2012;7(12):e50707.
doi:10.1371/journal.pone.0050707
73. Davachi L, Wagner AD. Hippocampal Contributions to Episodic Encoding: Insights
From Relational and Item-Based Learning. J Neurophysiol. 2002;88(2):982-990.
doi:10.1152/jn.2002.88.2.982
74. Rozin P, Dow S, Moscovitch M, Rajaram S. What Causes Humans to Begin and
End a Meal? A Role for Memory for What Has Been Eaten, as Evidenced by a
Study of Multiple Meal Eating in Amnesic Patients. Psychol Sci. 1998;9(5):392-
396. doi:10.1111/1467-9280.00073
75. Schumann CM, Hamstra J, Goodlin-Jones BL, Kwon H, Reiss AL, Amaral DG.
Hippocampal size positively correlates with verbal IQ in male children.
Hippocampus. 2007;17(6):486-493. doi:10.1002/hipo.20282
76. Sweatt JD. Hippocampal function in cognition. Psychopharmacology (Berl).
2004;174(1). doi:10.1007/s00213-004-1795-9
78. Huang L, Yu X, Keim S, Li L, Zhang L, Zhang J. Maternal prepregnancy obesity
and child neurodevelopment in the Collaborative Perinatal Project. Int J
Epidemiol. 2014;43(3):783-792. doi:10.1093/ije/dyu030
79. Hinkle SN, Sharma AJ, Kim SY, Schieve LA. Maternal prepregnancy weight status
and associations with children’s development and disabilities at kindergarten. Int J
Obes 2005. 2013;37(10):1344-1351. doi:10.1038/ijo.2013.128
80. Neggers YH, Goldenberg RL, Ramey SL, Cliver SP. Maternal prepregnancy body
mass index and psychomotor development in children. Acta Obstet Gynecol
Scand. 2003;82(3):235-240.
81. Tanda R, Salsberry PJ, Reagan PB, Fang MZ. The impact of prepregnancy obesity
on children’s cognitive test scores. Matern Child Health J. 2013;17(2):222-229.
doi:10.1007/s10995-012-0964-4
115
82. Jabès A, Thomas KM, Langworthy S, Georgieff MK, Nelson CA. Functional and
Anatomic Consequences of Diabetic Pregnancy on Memory in Ten-Year-Old
Children: J Dev Behav Pediatr. 2015;36(7):529-535.
doi:10.1097/DBP.0000000000000203
83. Riggins T, Bauer PJ, Georgieff MK, Nelson CA. Declarative memory performance
in infants of diabetic mothers. Adv Child Dev Behav. 2010;38:73-110.
doi:10.1016/B978-0-12-374471-5.00004-0
84. Bierhaus A, Nawroth PP. Multiple levels of regulation determine the role of the
receptor for AGE (RAGE) as common soil in inflammation, immune responses
and diabetes mellitus and its complications. Diabetologia. 2009;52(11):2251-2263.
doi:10.1007/s00125-009-1458-9
85. Golalipour MJ, Kafshgiri SK, Ghafari S. Gestational diabetes induced neuronal loss
in CA1 and CA3 subfields of rat hippocampus in early postnatal life. Folia
Morphol. 2012;71(2):7.
86. Hami J, Sadr-Nabavi A, Sankian M, Balali-Mood M, Haghir H. The effects of
maternal diabetes on expression of insulin-like growth factor-1 and insulin
receptors in male developing rat hippocampus. Brain Struct Funct.
2013;218(1):73-84. doi:10.1007/s00429-011-0377-y
87. Kinney BA, Rabe MB, Jensen RA, Steger RW. Maternal Hyperglycemia Leads to
Gender-Dependent Deficits in Learning and Memory in Offspring. Exp Biol Med.
2003;228(2):152-159. doi:10.1177/153537020322800204
89. DeBoer T, Wewerka S, Bauer PJ, Georgieff MK, Nelson CA. Explicit memory
performance in infants of diabetic mothers at 1 year of age. Dev Med Child
Neurol. 2007;47(8):525-531. doi:10.1111/j.1469-8749.2005.tb01186.x
90. Siddappa AM, Georgieff MK, Wewerka S, Worwa C, Nelson CA, Deregnier R-A.
Iron Deficiency Alters Auditory Recognition Memory in Newborn Infants of
Diabetic Mothers. Pediatr Res. 2004;55(6):1034-1041.
doi:10.1203/01.pdr.0000127021.38207.62
91. Buss C, Entringer S, Davis EP, et al. Impaired Executive Function Mediates the
Association between Maternal Pre-Pregnancy Body Mass Index and Child ADHD
Symptoms. PLOS ONE. 2012;7(6):e37758. doi:10.1371/journal.pone.0037758
92. Insausti R, Amaral DG. Hippocampal Formation. In: The Human Nervous System.
Elsevier; 2012:896-942. doi:10.1016/B978-0-12-374236-0.10024-0
93. Tamnes CK, Bos MGN, van de Kamp FC, Peters S, Crone EA. Longitudinal
development of hippocampal subregions from childhood to adulthood. Dev Cogn
Neurosci. 2018;30:212-222. doi:10.1016/j.dcn.2018.03.009
116
94. Corkin S. What’s new with the amnesic patient H.M.? Nat Rev Neurosci.
2002;3(2):153-160. doi:10.1038/nrn726
95. Ogden J. HM, the Man with No Memory. Psychology Today. Published 2012.
Accessed December 27, 2018. http://www.psychologytoday.com/blog/trouble-in-
mind/201201/hm-the-man-no-memory
96. Scott Huettel. Functional Resonance Imaging. Third. Sinauer Associates Inc.;
2014.
97. Fischl B. FreeSurfer. NeuroImage. 2012;62(2):774-781.
doi:10.1016/j.neuroimage.2012.01.021
98. Iglesias JE, Augustinack JC, Nguyen K, et al. A computational atlas of the
hippocampal formation using ex vivo , ultra-high resolution MRI: Application to
adaptive segmentation of in vivo MRI. NeuroImage. 2015;115:117-137.
doi:10.1016/j.neuroimage.2015.04.042
99. Backhausen LL, Herting MM, Buse J, Roessner V, Smolka MN, Vetter NC. Quality
Control of Structural MRI Images Applied Using FreeSurfer—A Hands-On
Workflow to Rate Motion Artifacts. Front Neurosci. 2016;10.
doi:10.3389/fnins.2016.00558
100. Cover KS, van Schijndel RA, Bosco P, Damangir S, Redolfi A. Can measuring
hippocampal atrophy with a fully automatic method be substantially less noisy
than manual segmentation over both 1 and 3 years? Psychiatry Res
Neuroimaging. 2018;280:39-47. doi:10.1016/j.pscychresns.2018.06.011
101. Krogsrud SK, Tamnes CK, Fjell AM, et al. Development of hippocampal subfield
volumes from 4 to 22 years: Development of Hippocampal Subfield Volumes.
Hum Brain Mapp. 2014;35(11):5646-5657. doi:10.1002/hbm.22576
102. Schmidt MF, Storrs JM, Freeman KB, et al. A comparison of manual tracing and
FreeSurfer for estimating hippocampal volume over the adult lifespan. Hum Brain
Mapp. 2018;39(6):2500-2513. doi:10.1002/hbm.24017
103. Whelan CD, Hibar DP, van Velzen LS, et al. Heritability and reliability of
automatically segmented human hippocampal formation subregions. NeuroImage.
2016;128:125-137. doi:10.1016/j.neuroimage.2015.12.039
105. Al-Amin M, Zinchenko A, Geyer T. Hippocampal subfield volume changes in
subtypes of attention deficit hyperactivity disorder. Brain Res. 2018;1685:1-8.
doi:10.1016/j.brainres.2018.02.007
106. Tamnes CK, Walhovd KB, Engvig A, et al. Regional Hippocampal Volumes and
Development Predict Learning and Memory. Dev Neurosci. 2014;36(3-4):161-174.
doi:10.1159/000362445
117
107. Foo H, Mak E, Chander RJ, et al. Associations of hippocampal subfields in the
progression of cognitive decline related to Parkinson’s disease. NeuroImage Clin.
2017;14:37-42. doi:10.1016/j.nicl.2016.12.008
108. Boardman JP, Counsell SJ, Rueckert D, et al. Abnormal deep grey matter
development following preterm birth detected using deformation-based
morphometry. NeuroImage. 2006;32(1):70-78.
doi:10.1016/j.neuroimage.2006.03.029
109. Srinivasan L, Dutta R, Counsell SJ, et al. Quantification of Deep Gray Matter in
Preterm Infants at Term-Equivalent Age Using Manual Volumetry of 3-Tesla
Magnetic Resonance Images. Pediatrics. 2007;119(4):759-765.
doi:10.1542/peds.2006-2508
111. Peterson BS, Anderson AW, Ehrenkranz R, et al. Regional Brain Volumes and
Their Later Neurodevelopmental Correlates in Term and Preterm Infants.
Pediatrics. 2003;111(5):939-948. doi:10.1542/peds.111.5.939
112. Soria-Pastor S, Padilla N, Zubiaurre-Elorza L, et al. Decreased Regional Brain
Volume and Cognitive Impairment in Preterm Children at Low Risk. Pediatrics.
2009;124(6):e1161-e1170. doi:10.1542/peds.2009-0244
113. Erickson KI, Voss MW, Prakash RS, et al. Exercise training increases size of
hippocampus and improves memory. Proc Natl Acad Sci U S A.
2011;108(7):3017-3022. doi:10.1073/pnas.1015950108
114. Leuner B, Gould E. Structural Plasticity and Hippocampal Function. Annu Rev
Psychol. 2010;61(1):111-140. doi:10.1146/annurev.psych.093008.100359
115. Rebai R, Jasmin L, Boudah A. The antidepressant effect of melatonin and
fluoxetine in diabetic rats is associated with a reduction of the oxidative stress in
the prefrontal and hippocampal cortices. Brain Res Bull. 2017;134:142-150.
doi:10.1016/j.brainresbull.2017.07.013
116. Sämann PG, Höhn D, Chechko N, et al. Prediction of antidepressant treatment
response from gray matter volume across diagnostic categories. Eur
Neuropsychopharmacol. 2013;23(11):1503-1515.
doi:10.1016/j.euroneuro.2013.07.004
117. Tyler WJ, Alonso M, Bramham CR, Pozzo-Miller LD. From Acquisition to
Consolidation: On the Role of Brain-Derived Neurotrophic Factor Signaling in
Hippocampal-Dependent Learning. Learn Mem Cold Spring Harb N.
2002;9(5):224-237. doi:10.1101/lm.51202
118. Lu B, Nagappan G, Lu Y. BDNF and Synaptic Plasticity, Cognitive Function, and
Dysfunction. In: Lewin GR, Carter BD, eds. Neurotrophic Factors. Handbook of
Experimental Pharmacology. Springer; 2014:223-250. doi:10.1007/978-3-642-
45106-5_9
118
119. Mu J-S, Li W-P, Yao Z-B, Zhou X-F. Deprivation of endogenous brain-derived
neurotrophic factor results in impairment of spatial learning and memory in adult
rats. Brain Res. 1999;835(2):259-265. doi:10.1016/S0006-8993(99)01592-9
120. Hassevoort KM, Khan NA, Hillman CH, Cohen NJ. Childhood Markers of Health
Behavior Relate to Hippocampal Health, Memory, and Academic Performance:
Childhood Markers of Health Behavior Relate to Hippocampal Health, Memory,
and Academic Performance. Mind Brain Educ. 2016;10(3):162-170.
doi:10.1111/mbe.12108
121. Herting MM, Nagel BJ. Aerobic fitness relates to learning on a virtual Morris Water
Task and hippocampal volume in adolescents. Behav Brain Res.
2012;233(2):517-525. doi:10.1016/j.bbr.2012.05.012
122. Farmer J, Zhao X, van Praag H, Wodtke K, Gage FH, Christie BR. Effects of
voluntary exercise on synaptic plasticity and gene expression in the dentate gyrus
of adult male Sprague-Dawley rats in vivo. Neuroscience. 2004;124(1):71-79.
doi:10.1016/j.neuroscience.2003.09.029
123. Praag H van, Shubert T, Zhao C, Gage FH. Exercise Enhances Learning and
Hippocampal Neurogenesis in Aged Mice. J Neurosci Off J Soc Neurosci.
2005;25(38):8680-8685. doi:10.1523/JNEUROSCI.1731-05.2005
124. Vaynman S, Ying Z, Gomez-Pinilla F. Hippocampal BDNF mediates the efficacy
of exercise on synaptic plasticity and cognition. Eur J Neurosci.
2004;20(10):2580-2590. doi:10.1111/j.1460-9568.2004.03720.x
125. Lemaire V, Koehl M, Le Moal M, Abrous DN. Prenatal stress produces learning
deficits associated with an inhibition of neurogenesis in the hippocampus. Proc
Natl Acad Sci U S A. 2000;97(20):11032-11037.
126. Naninck EFG, Hoeijmakers L, Kakava-Georgiadou N, et al. Chronic early life
stress alters developmental and adult neurogenesis and impairs cognitive function
in mice. Hippocampus. 2015;25(3):309-328. doi:10.1002/hipo.22374
127. Bremner JD, Vythilingam M, Vermetten E, et al. MRI and PET study of deficits in
hippocampal structure and function in women with childhood sexual abuse and
posttraumatic stress disorder. Am J Psychiatry. 2003;160(5):924-932.
doi:10.1176/appi.ajp.160.5.924
128. Kim EJ, Pellman B, Kim JJ. Stress effects on the hippocampus: a critical review.
Learn Mem Cold Spring Harb N. 2015;22(9):411-416. doi:10.1101/lm.037291.114
129. Kitayama N, Vaccarino V, Kutner M, Weiss P, Bremner JD. Magnetic resonance
imaging (MRI) measurement of hippocampal volume in posttraumatic stress
disorder: a meta-analysis. J Affect Disord. 2005;88(1):79-86.
doi:10.1016/j.jad.2005.05.014
119
130. Levy-Gigi E, Kéri S, Myers CE, et al. Individuals with posttraumatic stress disorder
show a selective deficit in generalization of associative learning.
Neuropsychology. 2012;26(6):758-767. doi:10.1037/a0029361
131. Pasupathy A, Miller EK. Different time courses of learning-related activity in the
prefrontal cortex and striatum. Nature. 2005;433(7028):873-876.
doi:10.1038/nature03287
132. Krishna R, Moustafa AA, Eby LA, Skeen LC, Myers CE. Learning and
generalization in healthy aging: implication for frontostriatal and hippocampal
function. Cogn Behav Neurol Off J Soc Behav Cogn Neurol. 2012;25(1):7-15.
doi:10.1097/WNN.0b013e318248ff1b
133. Frank DW, Dewitt M, Hudgens-Haney M, et al. Emotion regulation: Quantitative
meta-analysis of functional activation and deactivation. Neurosci Biobehav Rev.
2014;45:202-211. doi:10.1016/j.neubiorev.2014.06.010
134. Hartley CA, Phelps EA. Changing Fear: The Neurocircuitry of Emotion Regulation.
Neuropsychopharmacology. 2010;35(1):136-146. doi:10.1038/npp.2009.121
135. Godsil BP, Kiss JP, Spedding M, Jay TM. The hippocampal–prefrontal pathway:
The weak link in psychiatric disorders? Eur Neuropsychopharmacol.
2013;23(10):1165-1181. doi:10.1016/j.euroneuro.2012.10.018
136. Fanselow MS, Dong H-W. Are The Dorsal and Ventral Hippocampus functionally
distinct structures? Neuron. 2010;65(1):7. doi:10.1016/j.neuron.2009.11.031
137. Son Y, Yang M, Wang H, Moon C. Hippocampal dysfunctions caused by cranial
irradiation: A review of the experimental evidence. Brain Behav Immun.
2015;45:287-296. doi:10.1016/j.bbi.2015.01.007
138. Bannerman DM, Rawlins JNP, McHugh SB, et al. Regional dissociations within
the hippocampus—memory and anxiety. Neurosci Biobehav Rev. 2004;28(3):273-
283. doi:10.1016/j.neubiorev.2004.03.004
139. Abdallah CG, Wrocklage KM, Averill CL, et al. Anterior hippocampal
dysconnectivity in posttraumatic stress disorder: a dimensional and multimodal
approach. Transl Psychiatry. 2017;7(2):e1045. doi:10.1038/tp.2017.12
140. Henje Blom E, Han LKM, Connolly CG, et al. Peripheral telomere length and
hippocampal volume in adolescents with major depressive disorder. Transl
Psychiatry. 2015;5:e676. doi:10.1038/tp.2015.172
141. Harrisberger F, Smieskova R, Schmidt A, et al. BDNF Val66Met polymorphism
and hippocampal volume in neuropsychiatric disorders: A systematic review and
meta-analysis. Neurosci Biobehav Rev. 2015;55:107-118.
doi:10.1016/j.neubiorev.2015.04.017
120
142. Depression Definition and DSM-5 Diagnostic Criteria. PsyCom.net - Mental Health
Treatment Resource Since 1986. Published 2018. Accessed January 22, 2019.
https://www.psycom.net/depression-definition-dsm-5-diagnostic-criteria/
143. Boldrini M, Underwood MD, Hen R, et al. Antidepressants increase neural
progenitor cells in the human hippocampus. Neuropsychopharmacology.
2009;34(11):2376-2389. doi:10.1038/npp.2009.75
144. Hamann SB, Cahill L, Squire LR. Emotional perception and memory in amnesia.
Neuropsychology. 1997;11(1):104-113. doi:10.1037/0894-4105.11.1.104
145. Meyer-Lindenberg AS, Olsen RK, Kohn PD, et al. Regionally specific disturbance
of dorsolateral prefrontal-hippocampal functional connectivity in schizophrenia.
Arch Gen Psychiatry. 2005;62(4):379-386. doi:10.1001/archpsyc.62.4.379
146. Wolf RC, Vasic N, Sambataro F, et al. Temporally anticorrelated brain networks
during working memory performance reveal aberrant prefrontal and hippocampal
connectivity in patients with schizophrenia. Prog Neuropsychopharmacol Biol
Psychiatry. 2009;33(8):1464-1473. doi:10.1016/j.pnpbp.2009.07.032
147. Tracy AL, Jarrard LE, Davidson TL. The hippocampus and motivation revisited:
appetite and activity. Behav Brain Res. 2001;127(1-2):13-23. doi:10.1016/S0166-
4328(01)00364-3
148. Higgs S. Memory for recent eating and its influence on subsequent food intake.
Appetite. 2002;39(2):159-166. doi:10.1006/appe.2002.0500
149. Kanoski SE, Grill HJ. Hippocampus Contributions to Food Intake Control_
Mnemonic, Neuroanatomical, and Endocrine Mechanisms.
doi:10.1016/j.biopsych.2015.09.011
150. Sadeghirad B, Duhaney T, Motaghipisheh S, Campbell NRC, Johnston BC.
Influence of unhealthy food and beverage marketing on children’s dietary intake
and preference: a systematic review and meta-analysis of randomized trials:
Meta-analysis of unhealthy food and beverage marketing. Obes Rev.
2016;17(10):945-959. doi:10.1111/obr.12445
151. Luo S, Alves J, Hardy K, et al. Neural processing of food cues in pre-pubertal
children: Neural food cue reactivity in children. Pediatr Obes. Published online
July 17, 2018. doi:10.1111/ijpo.12435
152. Luo S, Romero A, Adam TC, Hu HH, Monterosso J, Page KA. Abdominal fat is
associated with a greater brain reward response to high-calorie food cues in
hispanic women: Food Images Stimulate Brain and Appetite. Obesity.
2013;21(10):2029-2036. doi:10.1002/oby.20344
121
153. Mestre ZL, Bischoff-Grethe A, Eichen DM, Wierenga CE, Strong D, Boutelle KN.
Hippocampal atrophy and altered brain responses to pleasant tastes among
obese compared with healthy weight children. Int J Obes. 2017;41(10):1496-1502.
doi:10.1038/ijo.2017.130
154. Ferrario CR, Labouèbe G, Liu S, et al. Homeostasis Meets Motivation in the Battle
to Control Food Intake. J Neurosci. 2016;36(45):11469-11481.
doi:10.1523/JNEUROSCI.2338-16.2016
155. Kaag AM, Schluter RS, Karel P, et al. Aversive Counterconditioning Attenuates
Reward Signaling in the Ventral Striatum. Front Hum Neurosci. 2016;10:418.
doi:10.3389/fnhum.2016.00418
156. Higgs S, Williamson AC, Rotshtein P, Humphreys GW. Sensory-Specific Satiety Is
Intact in Amnesics Who Eat Multiple Meals. Psychol Sci. 2008;19(7):623-628.
doi:10.1111/j.1467-9280.2008.02132.x
157. Wilkinson LL, Brunstrom JM. Sensory specific satiety: More than ‘just’
habituation? Appetite. 2016;103:221-228. doi:10.1016/j.appet.2016.04.019
158. Hsu TM, Hahn JD, Konanur VR, Lam A, Kanoski SE. Hippocampal GLP-1
receptors influence food intake, meal size, and effort-based responding for food
through volume transmission. Neuropsychopharmacol Off Publ Am Coll
Neuropsychopharmacol. 2015;40(2):327-337. doi:10.1038/npp.2014.175
159. Abegg K, Bernasconi L, Hutter M, et al. Ghrelin receptor inverse agonists as a
novel therapeutic approach against obesity-related metabolic disease. Diabetes
Obes Metab. 2017;19(12):1740-1750. doi:10.1111/dom.13020
160. Eichenbaum H. Hippocampus. Neuron. 2004;44(1):109-120.
doi:10.1016/j.neuron.2004.08.028
161. McCormick C, Ciaramelli E, De Luca F, Maguire EA. Comparing and Contrasting
the Cognitive Effects of Hippocampal and Ventromedial Prefrontal Cortex
Damage: A Review of Human Lesion Studies. Neuroscience. 2018;374:295-318.
doi:10.1016/j.neuroscience.2017.07.066
162. Moscovitch M, Cabeza R, Winocur G, Nadel L. Episodic Memory and Beyond:
The Hippocampus and Neocortex in Transformation. Annu Rev Psychol.
2016;67(1):105-134. doi:10.1146/annurev-psych-113011-143733
163. Monti JM, Cooke GE, Watson PD, Voss MW, Kramer AF, Cohen NJ. Relating
Hippocampus to Relational Memory Processing across Domains and Delays. J
Cogn Neurosci. 2015;27(2):234-245. doi:10.1162/jocn_a_00717
164. Kurczek J, Wechsler E, Ahuja S, et al. Differential contributions of hippocampus
and medial prefrontal cortex to self-projection and self-referential processing.
Neuropsychologia. 2015;73:116-126. doi:10.1016/j.neuropsychologia.2015.05.002
122
165. Viard A, Piolino P, Desgranges B, et al. Hippocampal Activation for
Autobiographical Memories over the Entire Lifetime in Healthy Aged Subjects: An
fMRI Study. Cereb Cortex. 2007;17(10):2453-2467. doi:10.1093/cercor/bhl153
166. Shapiro M. Plasticity, Hippocampal Place Cells, and Cognitive Maps. Arch Neurol.
2001;58(6):874. doi:10.1001/archneur.58.6.874
167. Eichenbaum H. The role of the hippocampus in navigation is memory. J
Neurophysiol. 2017;117(4):1785-1796. doi:10.1152/jn.00005.2017
168. Jabès A, Nelson CA. 20 years after “The ontogeny of human memory: A cognitive
neuroscience perspective,” where are we? Int J Behav Dev. 2015;39(4):293-303.
doi:10.1177/0165025415575766
169. Maguire EA, Nannery R, Spiers HJ. Navigation around London by a taxi driver
with bilateral hippocampal lesions. Brain. 2006;129(11):2894-2907.
doi:10.1093/brain/awl286
170. Broadbent NJ, Squire LR, Clark RE. Spatial memory, recognition memory, and the
hippocampus. Proc Natl Acad Sci. 2004;101(40):14515-14520.
doi:10.1073/pnas.0406344101
171. Winocur G, Moscovitch M, Caruana DA, Binns MA. Retrograde amnesia in rats
with lesions to the hippocampus on a test of spatial memory. Neuropsychologia.
2005;43(11):1580-1590. doi:10.1016/j.neuropsychologia.2005.01.013
172. Horecka KM, Dulas MR, Schwarb H, Lucas HD, Duff M, Cohen NJ.
Reconstructing relational information. Hippocampus. 2018;28(2):164-177.
doi:10.1002/hipo.22819
173. Sullivan Giovanello K, Schnyer DM, Verfaellie M. A critical role for the anterior
hippocampus in relational memory: Evidence from an fMRI study comparing
associative and item recognition. Hippocampus. 2004;14(1):5-8.
doi:10.1002/hipo.10182
174. Cohen NJ, Poldrack RA, Eichenbaum H. Memory for Items and Memory for
Relations in the Procedural/Declarative Memory Framework. Memory. 1997;5(1-
2):131-178. doi:10.1080/741941149
175. Cohen NJ, Ryan J, Hunt C, Romine L, Wszalek T, Nash C. Hippocampal system
and declarative (relational) memory: Summarizing the data from functional
neuroimaging studies. Hippocampus. 1999;9(1):83-98. doi:10.1002/(SICI)1098-
1063(1999)9:1<83::AID-HIPO9>3.0.CO;2-7
176. Ryan JD, Althoff RR, Whitlow S, Cohen NJ. Amnesia is a Deficit in Relational
Memory. Psychol Sci. 2000;11(6):454-461. doi:10.1111/1467-9280.00288
123
177. Watson PD, Voss JL, Warren DE, Tranel D, Cohen NJ. Spatial reconstruction by
patients with hippocampal damage is dominated by relational memory errors.
Hippocampus. 2013;23(7):570-580. doi:10.1002/hipo.22115
178. Preston AR, Shrager Y, Dudukovic NM, Gabrieli JDE. Hippocampal contribution to
the novel use of relational information in declarative memory. Hippocampus.
2004;14(2):148-152. doi:10.1002/hipo.20009
179. Prince SE. Neural Correlates of Relational Memory: Successful Encoding and
Retrieval of Semantic and Perceptual Associations. J Neurosci. 2005;25(5):1203-
1210. doi:10.1523/JNEUROSCI.2540-04.2005
180. Monti JM, Voss MW, Pence A, McAuley E, Kramer AF, Cohen NJ. History of mild
traumatic brain injury is associated with deficits in relational memory, reduced
hippocampal volume, and less neural activity later in life. Front Aging Neurosci.
2013;5. doi:10.3389/fnagi.2013.00041
181. Andersen SL. Trajectories of brain development: point of vulnerability or window
of opportunity? Neurosci Biobehav Rev. 2003;27(1-2):3-18. doi:10.1016/S0149-
7634(03)00005-8
182. Avishai-Eliner S. Stressed-out, or in (utero)? Trends Neurosci. 2002;25(10):518-
524. doi:10.1016/S0166-2236(02)02241-5
183. Gianaros PJ, Jennings JR, Sheu LK, Greer PJ, Kuller LH, Matthews KA.
Prospective reports of chronic life stress predict decreased grey matter volume in
the hippocampus. NeuroImage. 2007;35(2):795-803.
doi:10.1016/j.neuroimage.2006.10.045
184. Lavenex P, Banta Lavenex P. Building hippocampal circuits to learn and
remember: Insights into the development of human memory. Behav Brain Res.
2013;254:8-21. doi:10.1016/j.bbr.2013.02.007
185. Ribordy F, Jabès A, Banta Lavenex P, Lavenex P. Development of allocentric
spatial memory abilities in children from 18 months to 5 years of age. Cognit
Psychol. 2013;66(1):1-29. doi:10.1016/j.cogpsych.2012.08.001
186. Stiles J, Jernigan TL. The Basics of Brain Development. Neuropsychol Rev.
2010;20(4):327-348. doi:10.1007/s11065-010-9148-4
187. Arnold SE, Trojanowski JQ. Human fetal hippocampal development: I.
Cytoarchitecture, myeloarchitecture, and neuronal morphologic features. J Comp
Neurol. 1996;367(2):274-292. doi:10.1002/(SICI)1096-
9861(19960401)367:2<274::AID-CNE9>3.0.CO;2-2
188. Lenroot RK, Giedd JN. Brain development in children and adolescents: Insights
from anatomical magnetic resonance imaging. Neurosci Biobehav Rev.
2006;30(6):718-729. doi:10.1016/j.neubiorev.2006.06.001
124
189. Gogtay N, Nugent TF, Herman DH, et al. Dynamic mapping of normal human
hippocampal development. Hippocampus. 2006;16(8):664-672.
doi:10.1002/hipo.20193
190. Daugherty AM, Bender AR, Raz N, Ofen N. Age differences in hippocampal
subfield volumes from childhood to late adulthood. Hippocampus. 2016;26(2):220-
228. doi:10.1002/hipo.22517
191. Herting MM, Johnson C, Mills KL, et al. Development of subcortical volumes
across adolescence in males and females: A multisample study of longitudinal
changes. NeuroImage. 2018;172:194-205. doi:10.1016/j.neuroimage.2018.01.020
193. Lee JK, Ekstrom AD, Ghetti S. Volume of hippocampal subfields and episodic
memory in childhood and adolescence. NeuroImage. 2014;94:162-171.
doi:10.1016/j.neuroimage.2014.03.019
194. Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of childhood and adult
obesity in the United States, 2011-2012. JAMA. 2014;311(8):806-814.
doi:10.1001/jama.2014.732
196. Green HF, Nolan YM. Inflammation and the developing brain: Consequences for
hippocampal neurogenesis and behavior. Neurosci Biobehav Rev. 2014;40:20-34.
doi:10.1016/j.neubiorev.2014.01.004
197. van der Burg JW, Sen S, Chomitz VR, Seidell JC, Leviton A, Dammann O. The
role of systemic inflammation linking maternal BMI to neurodevelopment in
children. Pediatr Res. 2016;79(1):3-12. doi:10.1038/pr.2015.179
198. White CL, Pistell PJ, Purpera MN, et al. Effects of high fat diet on Morris maze
performance, oxidative stress, and inflammation in rats: Contributions of maternal
diet. Neurobiol Dis. 2009;35(1):3-13. doi:10.1016/j.nbd.2009.04.002
199. Graf AE, Haines KM, Pierson CR, et al. Perinatal inflammation results in
decreased oligodendrocyte numbers in adulthood. Life Sci. 2014;94(2):164-171.
doi:10.1016/j.lfs.2013.11.015
200. Van den Hove DLA, Steinbusch HWM, Scheepens A, et al. Prenatal stress and
neonatal rat brain development. Neuroscience. 2006;137(1):145-155.
doi:10.1016/j.neuroscience.2005.08.060
201. Zhu Z, Li X, Chen W, et al. Prenatal stress causes gender-dependent neuronal
loss and oxidative stress in rat hippocampus. J Neurosci Res. 2004;78(6):837-
844. doi:10.1002/jnr.20338
202. Ornellas F, Mello VS, Mandarim-de-Lacerda CA, Aguila MB. Sexual dimorphism
in fat distribution and metabolic profile in mice offspring from diet-induced obese
mothers. Life Sci. 2013;93(12-14):454-463. doi:10.1016/j.lfs.2013.08.005
125
204. Zhu C, Han T-L, Zhao Y, et al. A mouse model of pre-pregnancy maternal obesity
combined with offspring exposure to a high-fat diet resulted in cognitive
impairment in male offspring. Exp Cell Res. 2018;368(2):159-166.
doi:10.1016/j.yexcr.2018.04.019
205. Ou X, Thakali KM, Shankar K, Andres A, Badger TM. Maternal adiposity
negatively influences infant brain white matter development. Obes Silver Spring
Md. 2015;23(5):1047-1054. doi:10.1002/oby.21055
206. Verdejo-Román J, Björnholm L, Muetzel RL, et al. Maternal prepregnancy body
mass index and offspring white matter microstructure: results from three birth
cohorts. Int J Obes. 2019;43(10):1995-2006. doi:10.1038/s41366-018-0268-x
209. Bauer CCC, Moreno B, González-Santos L, Concha L, Barquera S, Barrios FA.
Child overweight and obesity are associated with reduced executive cognitive
performance and brain alterations: a magnetic resonance imaging study in
Mexican children: Child BMI, brain structure and function. Pediatr Obes.
2015;10(3):196-204. doi:10.1111/ijpo.241
210. Hair NL, Hanson JL, Wolfe BL, Pollak SD. Association of Child Poverty, Brain
Development, and Academic Achievement. JAMA Pediatr. 2015;169(9):822.
doi:10.1001/jamapediatrics.2015.1475
213. Uematsu A, Matsui M, Tanaka C, et al. Developmental Trajectories of Amygdala
and Hippocampus from Infancy to Early Adulthood in Healthy Individuals. Krueger
F, ed. PLoS ONE. 2012;7(10):e46970. doi:10.1371/journal.pone.0046970
214. Benjamini Y, Hochberg Y. Controlling the False Discovery Rate: A Practical and
Powerful Approach to Multiple Testing. J R Stat Soc Ser B Methodol.
1995;57(1):289-300. doi:10.1111/j.2517-6161.1995.tb02031.x
215. Farah MJ, Shera DM, Savage JH, et al. Childhood poverty: Specific associations
with neurocognitive development. Brain Res. 2006;1110(1):166-174.
doi:10.1016/j.brainres.2006.06.072
216. Gur RC, Richard J, Calkins ME, et al. Age group and sex differences in
performance on a computerized neurocognitive battery in children age 8–21.
Neuropsychology. 2012;26(2):251-265. doi:10.1037/a0026712
217. Noble KG, McCandliss BD, Farah MJ. Socioeconomic gradients predict individual
differences in neurocognitive abilities. Dev Sci. 2007;10(4):464-480.
doi:10.1111/j.1467-7687.2007.00600.x
218. Kotrlik J, Williams H, Jabor K. Reporting and Interpreting Effect Size in
Quantitative Agricultural Education Research. J Agric Educ. 2011;52(1):132-142.
doi:10.5032/jae.2011.01132
126
219. Okada K. Negative estimate of variance-accounted-for effect size: How often it is
obtained, and what happens if it is treated as zero. Behav Res Methods.
2017;49(3):979-987. doi:10.3758/s13428-016-0760-y
220. Olejnik S, Algina J. Generalized Eta and Omega Squared Statistics: Measures of
Effect Size for Some Common Research Designs. Psychol Methods.
2003;8(4):434-447. doi:10.1037/1082-989X.8.4.434
221. Salkind N. Encyclopedia of Research Design. SAGE Publications, Inc.; 2010.
doi:10.4135/9781412961288
222. Defining Childhood Obesity | Overweight & Obesity | CDC. Published 2018.
Accessed January 10, 2018. https://www.cdc.gov/obesity/childhood/defining.html
223. Casas M, Forns J, Martínez D, et al. Maternal pre-pregnancy obesity and
neuropsychological development in pre-school children: a prospective cohort
study. Pediatr Res. 2017;82(4):596-606. doi:10.1038/pr.2017.118
224. Widen EM, Kahn LG, Cirillo P, Cohn B, Kezios KL, Factor-Litvak P. Prepregnancy
overweight and obesity are associated with impaired child neurodevelopment.
Matern Child Nutr. 2018;14(1). doi:10.1111/mcn.12481
225. Fuemmeler BF, Zucker N, Sheng Y, et al. Pre-Pregnancy Weight and Symptoms
of Attention Deficit Hyperactivity Disorder and Executive Functioning Behaviors in
Preschool Children. Int J Environ Res Public Health. 2019;16(4).
doi:10.3390/ijerph16040667
226. Lecoutre S, Deracinois B, Laborie C, et al. Depot- and sex-specific effects of
maternal obesity in offspring’s adipose tissue. J Endocrinol. 2016;230(1):39-53.
doi:10.1530/JOE-16-0037
227. Murabayashi N, Sugiyama T, Zhang L, et al. Maternal high-fat diets cause insulin
resistance through inflammatory changes in fetal adipose tissue. Eur J Obstet
Gynecol Reprod Biol. 2013;169(1):39-44. doi:10.1016/j.ejogrb.2013.02.003
228. Schmitz L, Kuglin R, Bae-Gartz I, et al. Hippocampal insulin resistance links
maternal obesity with impaired neuronal plasticity in adult offspring.
Psychoneuroendocrinology. 2018;89:46-52. doi:10.1016/j.psyneuen.2017.12.023
229. Kern W, Peters A, Fruehwald-Schultes B, Deininger E, Born J, Fehm HL.
Improving Influence of Insulin on Cognitive Functions in Humans.
Neuroendocrinology. 2001;74(4):270-280. doi:10.1159/000054694
230. Hui L, Pei D-S, Zhang Q-G, Guan Q-H, Zhang G-Y. The neuroprotection of insulin
on ischemic brain injury in rat hippocampus through negative regulation of JNK
signaling pathway by PI3K/Akt activation. Brain Res. 2005;1052(1):1-9.
doi:10.1016/j.brainres.2005.05.043
127
231. Benedict C. Intranasal insulin improves memory in humans.
Psychoneuroendocrinology. 2004;29(10):1326-1334.
doi:10.1016/j.psyneuen.2004.04.003
232. Park C, Seeley R, Craft S, Woods S. Intracerebroventricular insulin enhances
memory in a passive-avoidance task. Physiol Behav. 2000;68(4):509-514.
doi:10.1016/S0031-9384(99)00220-6
233. Argente-Arizón P, Díaz F, Ros P, et al. The Hypothalamic Inflammatory/Gliosis
Response to Neonatal Overnutrition Is Sex and Age Dependent. Endocrinology.
2018;159(1):368-387. doi:10.1210/en.2017-00539
234. Schulz KM, Pearson JN, Neeley EW, et al. Maternal stress during pregnancy
causes sex-specific alterations in offspring memory performance, social
interactions, indices of anxiety, and body mass. Physiol Behav. 2011;104(2):340-
347. doi:10.1016/j.physbeh.2011.02.021
235. Zuena AR, Mairesse J, Casolini P, et al. Prenatal restraint stress generates two
distinct behavioral and neurochemical profiles in male and female rats. PloS One.
2008;3(5):e2170. doi:10.1371/journal.pone.0002170
236. Treit S, Chen Z, Zhou D, et al. Sexual dimorphism of volume reduction but not
cognitive deficit in fetal alcohol spectrum disorders: A combined diffusion tensor
imaging, cortical thickness and brain volume study. NeuroImage Clin.
2017;15:284-297. doi:10.1016/j.nicl.2017.05.006
237. Evans L, Myatt L. Sexual dimorphism in the effect of maternal obesity on
antioxidant defense mechanisms in the human placenta. Placenta. 2017;51:64-
69. doi:10.1016/j.placenta.2017.02.004
238. Shivers K-Y, Amador N, Abrams L, Hunter D, Jenab S, Quiñones-Jenab V.
Estrogen alters baseline and inflammatory-induced cytokine levels independent
from hypothalamic–pituitary–adrenal axis activity. Cytokine. 2015;72(2):121-129.
doi:10.1016/j.cyto.2015.01.007
239. Toung TJ, Traystman RJ, Hurn PD. Estrogen-mediated neuroprotection after
experimental stroke in male rats. Stroke. 1998;29(8):1666-1670.
240. Gillies GE, Murray HE, Dexter D, McArthur S. Sex dimorphisms in the
neuroprotective effects of estrogen in an animal model of Parkinson’s disease.
Pharmacol Biochem Behav. 2004;78(3):513-522. doi:10.1016/j.pbb.2004.04.022
241. Gillberg C, Cederlund M, Lamberg K, Zeijlon L. Brief Report: “The Autism
Epidemic”. The Registered Prevalence of Autism in a Swedish Urban Area. J
Autism Dev Disord. 2006;36(3):429-435. doi:10.1007/s10803-006-0081-6
242. Cramer SC, Sur M, Dobkin BH, et al. Harnessing neuroplasticity for clinical
applications. Brain. 2011;134(6):1591-1609. doi:10.1093/brain/awr039
128
243. Govindan RM, Chugani HT, Makki MI, Behen ME, Dornbush J, Sood S. Diffusion
tensor imaging of brain plasticity after occipital lobectomy. Pediatr Neurol.
2008;38(1):27-33. doi:10.1016/j.pediatrneurol.2007.08.004
244. Graveline CJ, Mikulis DJ, Crawley AP, Hwang PA. Regionalized sensorimotor
plasticity after hemispherectomy fMRI evaluation. Pediatr Neurol. 1998;19(5):337-
342. doi:10.1016/S0887-8994(98)00082-4
245. Cooper JM, Vargha-Khadem F, Gadian DG, Maguire EA. The effect of
hippocampal damage in children on recalling the past and imagining new
experiences. Neuropsychologia. 2011;49(7):1843-1850.
doi:10.1016/j.neuropsychologia.2011.03.008
246. Meoded A, Faria AV, Hartman AL, et al. Cerebral Reorganization after
Hemispherectomy: A DTI Study. Am J Neuroradiol. 2016;37(5):924-931.
doi:10.3174/ajnr.A4647
247. Dietz RM, Orfila JE, Rodgers KM, et al. Juvenile cerebral ischemia reveals age-
dependent BDNF–TrkB signaling changes: Novel mechanism of recovery and
therapeutic intervention. J Cereb Blood Flow Metab. 2018;38(12):2223-2235.
doi:10.1177/0271678X18766421
248. Shin LM, Shin PS, Heckers S, et al. Hippocampal function in posttraumatic stress
disorder. Hippocampus. 2004;14(3):292-300. doi:10.1002/hipo.10183
249. Rahman MM, Callaghan CK, Kerskens CM, Chattarji S, O’Mara SM. Early
hippocampal volume loss as a marker of eventual memory deficits caused by
repeated stress. Sci Rep. 2016;6:29127. doi:10.1038/srep29127
250. Hami J, Karimi R, Haghir H, Gholamin M, Sadr-Nabavi A. Diabetes in Pregnancy
Adversely Affects the Expression of Glycogen Synthase Kinase-3β in the
Hippocampus of Rat Neonates. J Mol Neurosci. 2015;57(2):273-281.
doi:10.1007/s12031-015-0617-3
251. Lotfi N, Hami J, Hosseini M, Haghir D, Haghir H. Diabetes during pregnancy
enhanced neuronal death in the hippocampus of rat offspring. Int J Dev Neurosci.
2016;51:28-35. doi:10.1016/j.ijdevneu.2016.04.009
252. Sadeghi A, Asghari H, Hami J, et al. Volumetric investigation of the hippocampus
in rat offspring due to diabetes in pregnancy–A stereological study. J Chem
Neuroanat. 2019;101:101669. doi:10.1016/j.jchemneu.2019.101669
253. Guleria RS, Pan J, DiPette D, Singh US. Hyperglycemia Inhibits Retinoic Acid-
Induced Activation of Rac1, Prevents Differentiation of Cortical Neurons, and
Causes Oxidative Stress in a Rat Model of Diabetic Pregnancy. Diabetes.
2006;55(12):3326-3334. doi:10.2337/db06-0169
129
254. Deng W, Aimone JB, Gage FH. New neurons and new memories: how does adult
hippocampal neurogenesis affect learning and memory? Nat Rev Neurosci.
2010;11(5):339-350. doi:10.1038/nrn2822
255. Veena SR, Krishnaveni GV, Srinivasan K, et al. Childhood Cognitive Ability:
Relationship to Gestational Diabetes Mellitus in India. Diabetologia.
2010;53(10):2134-2138. doi:10.1007/s00125-010-1847-0
256. Fraser A, Almqvist C, Larsson H, Långström N, Lawlor DA. Maternal diabetes in
pregnancy and offspring cognitive ability: sibling study with 723,775 men from
579,857 families. Diabetologia. 2014;57(1):102-109. doi:10.1007/s00125-013-
3065-z
257. Chandna AR, Kuhlmann N, Bryce CA, Greba Q, Campanucci VA, Howland JG.
Chronic maternal hyperglycemia induced during mid-pregnancy in rats increases
RAGE expression, augments hippocampal excitability, and alters behavior of the
offspring. Neuroscience. 2015;303:241-260.
doi:10.1016/j.neuroscience.2015.06.063
258. Vuong B, Odero G, Rozbacher S, et al. Exposure to gestational diabetes mellitus
induces neuroinflammation, derangement of hippocampal neurons, and cognitive
changes in rat offspring. J Neuroinflammation. 2017;14. doi:10.1186/s12974-017-
0859-9
259. Shipton OA, El-Gaby M, Apergis-Schoute J, et al. Left–right dissociation of
hippocampal memory processes in mice. Proc Natl Acad Sci U S A.
2014;111(42):15238-15243. doi:10.1073/pnas.1405648111
260. Lakens D. Calculating and reporting effect sizes to facilitate cumulative science: a
practical primer for t-tests and ANOVAs. Front Psychol. 2013;4.
doi:10.3389/fpsyg.2013.00863
261. Teicher MH, Anderson CM, Polcari A. Childhood maltreatment is associated with
reduced volume in the hippocampal subfields CA3, dentate gyrus, and subiculum.
Proc Natl Acad Sci. 2012;109(9):E563-E572. doi:10.1073/pnas.1115396109
262. Zach P, Mrzílková J, Stuchlík A, Valeš K. Delayed Effects of Elevated
Corticosterone Level on Volume of Hippocampal Formation in Laboratory Rat.
2010;59:12.
263. Hami J, Kheradmand H, Haghir H. Sex Differences and Laterality of Insulin
Receptor Distribution in Developing Rat Hippocampus: an Immunohistochemical
Study. J Mol Neurosci. 2014;54(1):100-108. doi:10.1007/s12031-014-0255-1
264. Blázquez E, Velázquez E, Hurtado-Carneiro V, Ruiz-Albusac JM. Insulin in the
Brain: Its Pathophysiological Implications for States Related with Central Insulin
Resistance, Type 2 Diabetes and Alzheimer’s Disease. Front Endocrinol. 2014;5.
doi:10.3389/fendo.2014.00161
130
265. Talbot K, Wang H-Y, Kazi H, et al. Demonstrated brain insulin resistance in
Alzheimer’s disease patients is associated with IGF-1 resistance, IRS-1
dysregulation, and cognitive decline. J Clin Invest. 2012;122(4):1316-1338.
doi:10.1172/JCI59903
266. Daugherty AM, Flinn R, Ofen N. Hippocampal CA3-dentate gyrus volume uniquely
linked to improvement in associative memory from childhood to adulthood.
NeuroImage. 2017;153:75-85. doi:10.1016/j.neuroimage.2017.03.047
267. Dimsdale-Zucker HR, Ritchey M, Ekstrom AD, Yonelinas AP, Ranganath C. CA1
and CA3 differentially support spontaneous retrieval of episodic contexts within
human hippocampal subfields. Nat Commun. 2018;9(1):294. doi:10.1038/s41467-
017-02752-1
268. Bock J, Murmu MS, Biala Y, Weinstock M, Braun K. Prenatal stress and neonatal
handling induce sex-specific changes in dendritic complexity and dendritic spine
density in hippocampal subregions of prepubertal rats. Neuroscience.
2011;193:34-43. doi:10.1016/j.neuroscience.2011.07.048
269. Donders J, Hoffman NM. Gender differences in learning and memory after
pediatric traumatic brain injury. Neuropsychology. 2002;16(4):491-499.
doi:http://dx.doi.org.libproxy1.usc.edu/10.1037/0894-4105.16.4.491
270. Papero PH, Prigatano GP, Snyder HM, Johnson DL. Children’s adaptive
behavioural competence after head injury. Neuropsychol Rehabil. 1993;3(4):321-
340. doi:10.1080/09602019308401445
271. Scott C, McKinlay A, Link to external site this link will open in a new window, et al.
A comparison of adult outcomes for males compared to females following
pediatric traumatic brain injury. Neuropsychology. 2015;29(4):501-508.
doi:http://dx.doi.org.libproxy1.usc.edu/10.1037/neu0000074
272. Esteban-Cornejo I, Martinez-Gomez D, Tejero-González CM, et al. Maternal
physical activity before and during the prenatal period and the offspring’s
academic performance in youth. The UP&DOWN study. J Matern-Fetal Neonatal
Med Off J Eur Assoc Perinat Med Fed Asia Ocean Perinat Soc Int Soc Perinat
Obstet. 2016;29(9):1414-1420. doi:10.3109/14767058.2015.1049525
273. Labonte-Lemoyne E, Curnier D, Ellemberg D. Exercise during pregnancy
enhances cerebral maturation in the newborn: A randomized controlled trial. J Clin
Exp Neuropsychol. 2017;39(4):347-354. doi:10.1080/13803395.2016.1227427
274. Robinson AM, Bucci DJ. Physical Exercise During Pregnancy Improves Object
Recognition Memory in Adult Offspring. Neuroscience. 2014;256.
doi:10.1016/j.neuroscience.2013.10.012
275. Cohen S, Kamarck T, Mermelstein R. A Global Measure of Perceived Stress. J
Health Soc Behav. 1983;24(4):385-396. doi:10.2307/2136404
131
276. Keim SA, Daniels JL, Dole N, Herring AH, Siega-Riz AM, Scheidt PC. A
prospective study of maternal anxiety, perceived stress, and depressive
symptoms in relation to infant cognitive development. Early Hum Dev.
2011;87(5):373-380. doi:10.1016/j.earlhumdev.2011.02.004
277. Lennon SL, Heaman M. Factors associated with family resilience during
pregnancy among inner-city women. Midwifery. 2015;31(10):957-964.
doi:10.1016/j.midw.2015.05.007
278. Ornoy A, Ratzon N, Greenbaum C, Peretz E, Soriano D, Dulitzky M.
Neurobehaviour of school age children born to diabetic mothers. Arch Dis Child
Fetal Neonatal Ed. 1998;79(2):F94-F99.
279. Dabelea D, Mayer-Davis EJ, Saydah S, et al. Prevalence of Type 1 and Type 2
Diabetes Among Children and Adolescents From 2001 to 2009. JAMA.
2014;311(17):1778-1786. doi:10.1001/jama.2014.3201
280. Dabelea D, Mayer-Davis EJ, Lamichhane AP, et al. Association of Intrauterine
Exposure to Maternal Diabetes and Obesity With Type 2 Diabetes in Youth: The
SEARCH Case-Control Study. Diabetes Care. 2008;31(7):1422-1426.
doi:10.2337/dc07-2417
281. Festa A, Williams K, D’Agostino R, Wagenknecht LE, Haffner SM. The natural
course of beta-cell function in nondiabetic and diabetic individuals: the Insulin
Resistance Atherosclerosis Study. Diabetes. 2006;55(4):1114-1120.
doi:10.2337/diabetes.55.04.06.db05-1100
282. Bergman RN, Phillips LS, Cobelli C. Physiologic evaluation of factors controlling
glucose tolerance in man: measurement of insulin sensitivity and beta-cell glucose
sensitivity from the response to intravenous glucose. J Clin Invest.
1981;68(6):1456-1467. doi:10.1172/JCI110398
283. Buchanan TA, Metzger BE, Freinkel N, Bergman RN. Insulin sensitivity and B-cell
responsiveness to glucose during late pregnancy in lean and moderately obese
women with normal glucose tolerance or mild gestational diabetes. Am J Obstet
Gynecol. 1990;162(4):1008-1014. doi:10.1016/0002-9378(90)91306-W
285. Zhang F, Xiao X, Liu D, Dong X, Sun J, Zhang X. Increased cord blood
angiotensin II concentration is associated with decreased insulin sensitivity in the
offspring of mothers with gestational diabetes mellitus. J Perinatol. 2013;33(1):9-
14. doi:10.1038/jp.2012.40
286. Luo Z-C, Delvin E, Fraser WD, et al. Maternal Glucose Tolerance in Pregnancy
Affects Fetal Insulin Sensitivity. Diabetes Care. 2010;33(9):2055-2061.
doi:10.2337/dc10-0819
132
287. Lowe WL, Scholtens DM, Kuang A, et al. Hyperglycemia and Adverse Pregnancy
Outcome Follow-up Study (HAPO FUS): Maternal Gestational Diabetes Mellitus
and Childhood Glucose Metabolism. Diabetes Care. 2019;42(3):372-380.
doi:10.2337/dc18-1646
289. Eriksson JG, Sandboge S, Salonen MK, Kajantie E, Osmond C. Long-term
consequences of maternal overweight in pregnancy on offspring later health:
Findings from the Helsinki Birth Cohort Study. Ann Med. 2014;46(6):434-438.
doi:10.3109/07853890.2014.919728
290. Group HSCR. Hyperglycaemia and Adverse Pregnancy Outcome (HAPO) Study:
associations with maternal body mass index: HAPO - BMI and perinatal
outcomes. BJOG Int J Obstet Gynaecol. 2010;117(5):575-584.
doi:10.1111/j.1471-0528.2009.02486.x
292. Catalano PM, Presley L, Minium J, Hauguel-de Mouzon S. Fetuses of Obese
Mothers Develop Insulin Resistance in Utero. Diabetes Care. 2009;32(6):1076-
1080. doi:10.2337/dc08-2077
293. Bremer AA, Auinger P, Byrd RS. Relationship Between Insulin Resistance-
Associated Metabolic Parameters and Anthropometric Measurements With Sugar-
Sweetened Beverage Intake and Physical Activity Levels in US Adolescents. Arch
Pediatr Adolesc Med. 2009;163(4):328-335. doi:10.1001/archpediatrics.2009.21
294. Donin AS, Nightingale CM, Owen CG, et al. Dietary Energy Intake Is Associated
With Type 2 Diabetes Risk Markers in Children. Diabetes Care. 2014;37(1):116-
123. doi:10.2337/dc13-1263
295. Nightingale CM, Rudnicka AR, Donin AS, et al. Screen time is associated with
adiposity and insulin resistance in children. Arch Dis Child. 2017;102(7):612-616.
doi:10.1136/archdischild-2016-312016
297. Welsh Jean A., Sharma Andrea, Cunningham Solveig A., Vos Miriam B.
Consumption of Added Sugars and Indicators of Cardiovascular Disease Risk
Among US Adolescents. Circulation. 2011;123(3):249-257.
doi:10.1161/CIRCULATIONAHA.110.972166
298. Davis JN, Alexander KE, Ventura EE, et al. Associations of dietary sugar and
glycemic index with adiposity and insulin dynamics in overweight Latino youth. Am
J Clin Nutr. 2007;86(5):1331-1338. doi:10.1093/ajcn/86.5.1331
299. Wang JW, Mark S, Henderson M, et al. Adiposity and glucose intolerance
exacerbate components of metabolic syndrome in children consuming sugar-
sweetened beverages: QUALITY cohort study. Pediatr Obes. 2013;8(4):284-293.
doi:10.1111/j.2047-6310.2012.00108.x
133
300. Savoye M, Caprio S, Dziura J, et al. Reversal of Early Abnormalities in Glucose
Metabolism in Obese Youth: Results of an Intensive Lifestyle Randomized
Controlled Trial. Diabetes Care. 2014;37(2):317-324. doi:10.2337/dc13-1571
301. Casey BM, Lucas MJ, McIntire DD, Leveno KJ. Pregnancy Outcomes in Women
With Gestational Diabetes Compared With the General Obstetric Population.
Obstet Gynecol. 1997;90(6):869-873. doi:10.1016/S0029-7844(97)00542-5
303. Kim SY, Sharma AJ, Sappenfield W, Wilson HG, Salihu HM. Association of
Maternal Body Mass Index, Excessive Weight Gain, and Gestational Diabetes
Mellitus With Large-for-Gestational-Age Births. Obstet Gynecol. 2014;123(4):737-
744. doi:10.1097/AOG.0000000000000177
304. Sandler V, Reisetter AC, Bain JR, et al. Maternal BMI and insulin resistance
associations with the maternal metabolome and newborn outcomes. Diabetologia.
2017;60(3):518-530. doi:10.1007/s00125-016-4182-2
305. Wahlberg J, Ekman B, Nyström L, Hanson U, Persson B, Arnqvist HJ. Gestational
diabetes: Glycaemic predictors for fetal macrosomia and maternal risk of future
diabetes. Diabetes Res Clin Pract. 2016;114:99-105.
doi:10.1016/j.diabres.2015.12.017
306. Fraser A, Tilling K, Macdonald-Wallis C, et al. Association of maternal weight gain
in pregnancy with offspring obesity and metabolic and vascular traits in childhood.
Circulation. 2010;121(23):2557-2564.
doi:10.1161/CIRCULATIONAHA.109.906081
307. Zhao P, Liu E, Qiao Y, et al. Maternal gestational diabetes and childhood obesity
at age 9-11: results of a multinational study. Diabetologia. 2016;59(11):2339-
2348. doi:10.1007/s00125-016-4062-9
308. Lawlor DA, Lichtenstein P, Långström N. Association of Maternal Diabetes
Mellitus in Pregnancy With Offspring Adiposity Into Early Adulthood: Sibling Study
in a Prospective Cohort of 280 866 Men From 248 293 Families. Circulation.
2011;123(3):258-265. doi:10.1161/CIRCULATIONAHA.110.980169
309. Li S, Zhu Y, Yeung E, et al. Offspring risk of obesity in childhood, adolescence
and adulthood in relation to gestational diabetes mellitus: a sex-specific
association. Int J Epidemiol. 2017;46(5):1533-1541. doi:10.1093/ije/dyx151
310. Chandler-Laney PC, Bush NC, Granger WM, Rouse DJ, Mancuso MS, Gower BA.
Overweight status and intrauterine exposure to gestational diabetes are
associated with children’s metabolic health. Pediatr Obes. 2012;7(1):44-52.
doi:10.1111/j.2047-6310.2011.00009.x
134
311. Nehring I, Chmitorz A, Reulen H, von Kries R, Ensenauer R. Gestational diabetes
predicts the risk of childhood overweight and abdominal circumference
independent of maternal obesity. Diabet Med J Br Diabet Assoc.
2013;30(12):1449-1456. doi:10.1111/dme.12286
312. Reynolds RM, Osmond C, Phillips DIW, Godfrey KM. Maternal BMI, Parity, and
Pregnancy Weight Gain: Influences on Offspring Adiposity in Young Adulthood. J
Clin Endocrinol Metab. 2010;95(12):5365-5369. doi:10.1210/jc.2010-0697
313. Beyerlein A, Uusitalo UM, Virtanen SM, et al. Intake of Energy and Protein is
Associated with Overweight Risk at Age 5.5 Years: Results from the Prospective
TEDDY Study. Obes Silver Spring Md. 2017;25(8):1435-1441.
doi:10.1002/oby.21897
314. DeBoer MD, Scharf RJ, Demmer RT. Sugar-Sweetened Beverages and Weight
Gain in 2- to 5-Year-Old Children. Pediatrics. 2013;132(3):413-420.
doi:10.1542/peds.2013-0570
315. Drake KM, Beach ML, Longacre MR, et al. Influence of Sports, Physical
Education, and Active Commuting to School on Adolescent Weight Status.
Pediatrics. 2012;130(2):e296-e304. doi:10.1542/peds.2011-2898
316. Emmett PM, Jones LR. Diet, growth, and obesity development throughout
childhood in the Avon Longitudinal Study of Parents and Children. Nutr Rev.
2015;73(Suppl 3):175-206. doi:10.1093/nutrit/nuv054
317. Malik VS, Pan A, Willett WC, Hu FB. Sugar-sweetened beverages and weight
gain in children and adults: a systematic review and meta-analysis123. Am J Clin
Nutr. 2013;98(4):1084-1102. doi:10.3945/ajcn.113.058362
318. Stettler N, Signer TM, Suter PM. Electronic Games and Environmental Factors
Associated with Childhood Obesity in Switzerland. Obes Res. 2004;12(6):896-
903. doi:10.1038/oby.2004.109
319. Costa CS, Rauber F, Leffa PS, Sangalli CN, Campagnolo PDB, Vitolo MR. Ultra-
processed food consumption and its effects on anthropometric and glucose
profile: A longitudinal study during childhood. Nutr Metab Cardiovasc Dis.
2019;29(2):177-184. doi:10.1016/j.numecd.2018.11.003
320. de Ruyter JC, Olthof MR, Seidell JC, Katan MB. A trial of sugar-free or sugar-
sweetened beverages and body weight in children. N Engl J Med.
2012;367(15):1397-1406. doi:10.1056/NEJMoa1203034
321. Herman KM, Sabiston CM, Mathieu M-E, Tremblay A, Paradis G. Sedentary
behavior in a cohort of 8- to 10-year-old children at elevated risk of obesity. Prev
Med. 2014;60:115-120. doi:10.1016/j.ypmed.2013.12.029
135
322. Nasreddine L, Naja F, Akl C, et al. Dietary, Lifestyle and Socio-Economic
Correlates of Overweight, Obesity and Central Adiposity in Lebanese Children
and Adolescents. Nutrients. 2014;6(3):1038-1062. doi:10.3390/nu6031038
323. Schröder H, Mendez MA, Gomez SF, et al. Energy density, diet quality, and
central body fat in a nationwide survey of young Spaniards. Nutr Burbank Los
Angel Cty Calif. 2013;29(11-12):1350-1355. doi:10.1016/j.nut.2013.05.019
324. Sauder KA, Bekelman TA, Harrall KK, Glueck DH, Dabelea D. Gestational
diabetes exposure and adiposity outcomes in childhood and adolescence: An
analysis of effect modification by breastfeeding, diet quality, and physical activity
in the EPOCH study. Pediatr Obes. Published online July 5, 2019:ijpo.12562.
doi:10.1111/ijpo.12562
325. Zhang T, Wang P, Liu H, et al. Physical Activity, TV Watching Time, Sleeping, and
Risk of Obesity and Hyperglycemia in the Offspring of Mothers with Gestational
Diabetes Mellitus. Sci Rep. 2017;7(1):1-9. doi:10.1038/srep41115
326. Ekelund U, Luan J, Sherar LB, et al. Moderate to Vigorous Physical Activity and
Sedentary Time and Cardiometabolic Risk Factors in Children and Adolescents.
JAMA. 2012;307(7):704-712. doi:10.1001/jama.2012.156
327. Bird SR, Hawley JA. Update on the effects of physical activity on insulin sensitivity
in humans. BMJ Open Sport — Exerc Med. 2017;2(1). doi:10.1136/bmjsem-2016-
000143
328. Thorell A, Hirshman MF, Nygren J, et al. Exercise and insulin cause GLUT-4
translocation in human skeletal muscle. Am J Physiol-Endocrinol Metab.
1999;277(4):E733-E741. doi:10.1152/ajpendo.1999.277.4.E733
329. Rattigan S, Wallis MG, Youd JM, Clark MG. Exercise Training Improves Insulin-
Mediated Capillary Recruitment in Association With Glucose Uptake in Rat
Hindlimb. Diabetes. 2001;50(12):2659-2665. doi:10.2337/diabetes.50.12.2659
330. Bradley RL, Jeon JY, Liu F-F, Maratos-Flier E. Voluntary exercise improves
insulin sensitivity and adipose tissue inflammation in diet-induced obese mice. Am
J Physiol - Endocrinol Metab. 2008;295(3):E586-E594.
doi:10.1152/ajpendo.00309.2007
331. Bickham DS, Blood EA, Walls CE, Shrier LA, Rich M. Characteristics of Screen
Media Use Associated With Higher BMI in Young Adolescents. Pediatrics.
2013;131(5):935-941. doi:10.1542/peds.2012-1197
332. Ortega FB, Ruiz JR, Sjöström M. Physical activity, overweight and central
adiposity in Swedish children and adolescents: the European Youth Heart Study.
Int J Behav Nutr Phys Act. 2007;4:61. doi:10.1186/1479-5868-4-61
136
333. Falbe J, Rosner B, Willett WC, Sonneville KR, Hu FB, Field AE. Adiposity and
Different Types of Screen Time. Pediatrics. 2013;132(6):e1497-e1505.
doi:10.1542/peds.2013-0887
334. Henderson M, Benedetti A, Barnett TA, Mathieu M-E, Deladoëy J, Gray-Donald K.
Influence of Adiposity, Physical Activity, Fitness, and Screen Time on Insulin
Dynamics Over 2 Years in Children. JAMA Pediatr. 2016;170(3):227.
doi:10.1001/jamapediatrics.2015.3909
335. Herman K, Chaput J-P, Sabiston C, Mathieu M-E, Tremblay A, Paradis G.
Combined Physical Activity/Sedentary Behavior Associations With Indices of
Adiposity in 8- to 10-Year-Old Children. J Phys Act Health. 2014;12.
doi:10.1123/jpah.2013-0019
336. Dencker M, Thorsson O, Karlsson MK, Lindén C, Wollmer P, Andersen LB. Daily
physical activity related to aerobic fitness and body fat in an urban sample of
children. Scand J Med Sci Sports. 2008;18(6):728-735. doi:10.1111/j.1600-
0838.2007.00741.x
337. Wang J, Obici S, Morgan K, Barzilai N, Feng Z, Rossetti L. Overfeeding Rapidly
Induces Leptin and Insulin Resistance. Diabetes. 2001;50(12):2786-2791.
doi:10.2337/diabetes.50.12.2786
338. Casazza K, Dulin-Keita A, Gower BA, Fernández JR. Relationships between
reported macronutrient intake and insulin dynamics in a multi-ethnic cohort of
early pubertal children. Int J Pediatr Obes. 2009;4(4):249-256.
doi:10.3109/17477160902763366
339. Kelsey MM, Zeitler PS. Insulin Resistance of Puberty. Curr Diab Rep.
2016;16(7):64. doi:10.1007/s11892-016-0751-5
340. Chen AK, Roberts CK, Barnard RJ. Effect of a short-term diet and exercise
intervention on metabolic syndrome in overweight children. Metabolism.
2006;55(7):871-878. doi:10.1016/j.metabol.2006.03.001
341. Durão C, Oliveira A, Santos AC, et al. Protein intake and dietary glycemic load of
4-year-olds and association with adiposity and serum insulin at 7 years of age:
sex-nutrient and nutrient-nutrient interactions. Int J Obes 2005. 2017;41(4):533-
541. doi:10.1038/ijo.2016.240
342. Ervin RB, Ogden CL. Trends in intake of energy and macronutrients in children
and adolescents from 1999-2000 through 2009-2010. NCHS Data Brief.
2013;(113):1-8.
345. Shapiro ALB, Sauder KA, Tregellas JR, et al. Exposure to maternal diabetes in
utero and offspring eating behavior: The EPOCH study. Appetite. 2017;116:610-
615. doi:10.1016/j.appet.2017.05.005
137
346. Boerschmann H, Pfluger M, Henneberger L, Ziegler A-G, Hummel S. Prevalence
and Predictors of Overweight and Insulin Resistance in Offspring of Mothers With
Gestational Diabetes Mellitus. Diabetes Care. 2010;33(8):1845-1849.
doi:10.2337/dc10-0139
347. Crume TL, Ogden L, West NA, et al. Association of exposure to diabetes in utero
with adiposity and fat distribution in a multiethnic population of youth: the
Exploring Perinatal Outcomes among Children (EPOCH) Study. Diabetologia.
2011;54(1):87-92. doi:10.1007/s00125-010-1925-3
348. Gillman MW, Rifas-Shiman S, Berkey CS, Field AE, Colditz GA. Maternal
gestational diabetes, birth weight, and adolescent obesity. Pediatrics.
2003;111(3):e221-226. doi:10.1542/peds.111.3.e221
349. Emdin CA, Khera AV, Natarajan P, et al. Genetic Association of Waist-to-Hip
Ratio With Cardiometabolic Traits, Type 2 Diabetes, and Coronary Heart Disease.
JAMA. 2017;317(6):626-634. doi:10.1001/jama.2016.21042
350. Allard C, Desgagné V, Patenaude J, et al. Mendelian randomization supports
causality between maternal hyperglycemia and epigenetic regulation of leptin
gene in newborns. Epigenetics. 2015;10(4):342-351.
doi:10.1080/15592294.2015.1029700
351. Hajj NE, Pliushch G, Schneider E, et al. Metabolic Programming of MEST DNA
Methylation by Intrauterine Exposure to Gestational Diabetes Mellitus. Diabetes.
2013;62(4):1320-1328. doi:10.2337/db12-0289
352. Kozak LP, Newman S, Chao P-M, Mendoza T, Koza RA. The Early Nutritional
Environment of Mice Determines the Capacity for Adipose Tissue Expansion by
Modulating Genes of Caveolae Structure. PLOS ONE. 2010;5(6):e11015.
doi:10.1371/journal.pone.0011015
354. Henderson M, Rabasa-Lhoret R, Bastard J-P, et al. Measuring insulin sensitivity in
youth: How do the different indices compare with the gold-standard method?
Diabetes Metab. 2011;37(1):72-78. doi:10.1016/j.diabet.2010.06.008
356. Belcher BR, Berrigan D, Dodd KW, Emken BA, Chou C-P, Spuijt-Metz D. Physical
Activity in US Youth: Impact of Race/Ethnicity, Age, Gender, & Weight Status.
Med Sci Sports Exerc. 2010;42(12):2211-2221.
doi:10.1249/MSS.0b013e3181e1fba9
357. Johnson RK, Driscoll P, Goran MI. Comparison of multiple-pass 24-hour recall
estimates of energy intake with total energy expenditure determined by the doubly
labeled water method in young children. J Am Diet Assoc. 1996;96(11):1140-
1144. doi:10.1016/S0002-8223(96)00293-3
138
359. Bahari H, Caruso V, Morris MJ. Late-Onset Exercise in Female Rat Offspring
Ameliorates the Detrimental Metabolic Impact of Maternal Obesity. Endocrinology.
2013;154(10):3610-3621. doi:10.1210/en.2013-1059
360. Sun B, Liang N-C, Ewald ER, et al. Early postweaning exercise improves central
leptin sensitivity in offspring of rat dams fed high-fat diet during pregnancy and
lactation. Am J Physiol-Regul Integr Comp Physiol. 2013;305(9):R1076-R1084.
doi:10.1152/ajpregu.00566.2012
361. Bunketorp Käll L, Malmgren H, Olsson E, Lindén T, Nilsson M. Effects of a
Curricular Physical Activity Intervention on Children’s School Performance,
Wellness, and Brain Development. J Sch Health. 2015;85(10):704-713.
doi:10.1111/josh.12303
362. Davis CL, Tomporowski PD, McDowell JE, et al. Exercise improves executive
function and achievement and alters brain activation in overweight children: a
randomized, controlled trial. Health Psychol Off J Div Health Psychol Am Psychol
Assoc. 2011;30(1):91-98. doi:10.1037/a0021766
363. Have M, Nielsen JH, Ernst MT, et al. Classroom-based physical activity improves
children’s math achievement – A randomized controlled trial. PLOS ONE.
2018;13(12):e0208787. doi:10.1371/journal.pone.0208787
364. Huang T, Tarp J, Domazet SL, et al. Associations of Adiposity and Aerobic
Fitness with Executive Function and Math Performance in Danish Adolescents. J
Pediatr. 2015;167(4):810-815. doi:10.1016/j.jpeds.2015.07.009
365. Makharia A, Nagarajan A, Mishra A, Peddisetty S, Chahal D, Singh Y. Effect of
environmental factors on intelligence quotient of children. Ind Psychiatry J.
2016;25(2):189-194. doi:10.4103/ipj.ipj_52_16
366. Raine LB, Khan NA, Drollette ES, Pontifex MB, Kramer AF, Hillman CH. Obesity,
Visceral Adipose Tissue, and Cognitive Function in Childhood. J Pediatr.
2017;187:134-140.e3. doi:10.1016/j.jpeds.2017.05.023
367. Reed JA, Einstein G, Hahn E, Hooker SP, Gross VP, Kravitz J. Examining the
impact of integrating physical activity on fluid intelligence and academic
performance in an elementary school setting: a preliminary investigation. J Phys
Act Health. 2010;7(3):343-351.
369. El-Kholy T, Elsayed E. Association of physical activity and health status with
intelligence quotient of high school students in Jeddah. J Phys Ther Sci.
2015;27(7):2039-2043. doi:10.1589/jpts.27.2039
370. Killgore WDS, Schwab ZJ. Sex Differences in the Association between Physical
Exercise and IQ. Percept Mot Skills. 2012;115(2):605-617.
doi:10.2466/06.10.50.PMS.115.5.605-617
139
371. Chaddock-Heyman L, Erickson KI, Kienzler C, et al. Physical Activity Increases
White Matter Microstructure in Children. Front Neurosci. 2018;12.
doi:10.3389/fnins.2018.00950
372. Krafft CE, Schaeffer DJ, Schwarz NF, et al. Improved Frontoparietal White Matter
Integrity in Overweight Children Is Associated with Attendance at an After-School
Exercise Program. Dev Neurosci. 2014;36(1):1-9. doi:10.1159/000356219
373. Clapp JF. Morphometric and neurodevelopmental outcome at age five years of
the offspring of women who continued to exercise regularly throughout pregnancy.
J Pediatr. 1996;129(6):856-863. doi:10.1016/s0022-3476(96)70029-x
374. Domingues MR, Matijasevich A, Barros AJD, Santos IS, Horta BL, Hallal PC.
Physical Activity during Pregnancy and Offspring Neurodevelopment and IQ in the
First 4 Years of Life. PLOS ONE. 2014;9(10):e110050.
doi:10.1371/journal.pone.0110050
375. Baeten JM, Bukusi EA, Lambe M. Pregnancy complications and outcomes among
overweight and obese nulliparous women. Am J Public Health. 2001;91(3):436-
440.
376. Chaddock-Heyman L, Erickson KI, Holtrop JL, et al. Aerobic fitness is associated
with greater white matter integrity in children. Front Hum Neurosci. 2014;8.
doi:10.3389/fnhum.2014.00584
377. Lebel C, Beaulieu C. Lateralization of the arcuate fasciculus from childhood to
adulthood and its relation to cognitive abilities in children. Hum Brain Mapp.
2009;30(11):3563-3573. doi:10.1002/hbm.20779
378. Schmithorst VJ, Wilke M, Dardzinski BJ, Holland SK. COGNITIVE FUNCTIONS
CORRELATE WITH WHITE MATTER ARCHITECTURE IN A NORMAL
PEDIATRIC POPULATION: A DIFFUSION TENSOR MR IMAGING STUDY. Hum
Brain Mapp. 2005;26(2):139-147. doi:10.1002/hbm.20149
379. Urger SE, De Bellis MD, Hooper SR, Woolley DP, Chen SD, Provenzale J. The
Superior Longitudinal Fasciculus in Typically Developing Children and
Adolescents: Diffusion Tensor Imaging and Neuropsychological Correlates. J
Child Neurol. 2015;30(1):9-20. doi:10.1177/0883073813520503
380. Lautenschlager NT, Cox KL, Flicker L, et al. Effect of Physical Activity on
Cognitive Function in Older Adults at Risk for Alzheimer Disease: A Randomized
Trial. JAMA. 2008;300(9):1027-1037. doi:10.1001/jama.300.9.1027
381. Bechara RG, Kelly ÁM. Exercise improves object recognition memory and induces
BDNF expression and cell proliferation in cognitively enriched rats. Behav Brain
Res. 2013;245:96-100. doi:10.1016/j.bbr.2013.02.018
140
382. Chaddock L, Erickson KI, Prakash RS, et al. A neuroimaging investigation of the
association between aerobic fitness, hippocampal volume, and memory
performance in preadolescent children. Brain Res. 2010;1358:172-183.
doi:10.1016/j.brainres.2010.08.049
383. Etnier JL, Wideman L, Labban JD, et al. The Effects of Acute Exercise on Memory
and Brain-Derived Neurotrophic Factor (BDNF). J Sport Exerc Psychol.
2016;38(4):331-340. doi:10.1123/jsep.2015-0335
384. Moon HY, Becke A, Berron D, et al. Running-induced systemic Cathepsin B
secretion is associated with memory function. Cell Metab. 2016;24(2):332-340.
doi:10.1016/j.cmet.2016.05.025
385. Muller AP, Gnoatto J, Moreira JD, et al. Exercise increases insulin signaling in the
hippocampus: Physiological effects and pharmacological impact of
intracerebroventricular insulin administration in mice. Hippocampus.
2011;21(10):1082-1092. doi:10.1002/hipo.20822
386. Jung Y-H, Shin NY, Jang JH, et al. Relationships among stress, emotional
intelligence, cognitive intelligence, and cytokines. Medicine (Baltimore).
2019;98(18):e15345. doi:10.1097/MD.0000000000015345
387. Tung SEH, Mohd Nasir MT, Chin YS, Zalilah MS, Zubaidah JO, Yim HS.
Psychological Factors and Cardiovascular Disease Risk Factors as Mediators of
the Relationship between Overweight/Obesity and Cognitive Function among
School Children in Kuala Lumpur, Malaysia. Child Obes. Published online October
19, 2018. doi:10.1089/chi.2018.0066
388. Watson KT, Wroolie TE, Tong G, et al. Neural correlates of liraglutide effects in
persons at risk for Alzheimer’s disease. Behav Brain Res. 2019;356:271-278.
doi:10.1016/j.bbr.2018.08.006
389. Aadland E, Kvalheim OM, Anderssen SA, Resaland GK, Andersen LB. The
multivariate physical activity signature associated with metabolic health in
children. Int J Behav Nutr Phys Act. 2018;15. doi:10.1186/s12966-018-0707-z
390. Cockcroft EJ, Williams CA, Tomlinson OW, et al. High intensity interval exercise is
an effective alternative to moderate intensity exercise for improving glucose
tolerance and insulin sensitivity in adolescent boys. J Sci Med Sport.
2015;18(6):720-724. doi:10.1016/j.jsams.2014.10.001
391. Esteban-Cornejo I, Tejero-Gonzalez CM, Sallis JF, Veiga OL. Physical activity
and cognition in adolescents: A systematic review. J Sci Med Sport.
2015;18(5):534-539. doi:10.1016/j.jsams.2014.07.007
392. Jelleyman C, Edwardson CL, Henson J, et al. Associations of Physical Activity
Intensities with Markers of Insulin Sensitivity. Med Sci Sports Exerc.
2017;49(12):2451-2458. doi:10.1249/MSS.0000000000001381
141
393. Rynders CA, Weltman JY, Jiang B, et al. Effects of exercise intensity on
postprandial improvement in glucose disposal and insulin sensitivity in prediabetic
adults. J Clin Endocrinol Metab. 2014;99(1):220-228. doi:10.1210/jc.2013-2687
394. Hillman CH, Pontifex MB, Castelli DM, et al. Effects of the FITKids randomized
controlled trial on executive control and brain function. Pediatrics.
2014;134(4):e1063-1071. doi:10.1542/peds.2013-3219
395. Rodriguez-Ayllon M, Derks IPM, van den Dries MA, et al. Associations of physical
activity and screen time with white matter microstructure in children from the
general population. NeuroImage. 2020;205:116258.
doi:10.1016/j.neuroimage.2019.116258
398. Alves JM, Zink J, Chow T, et al. Contributions of Prenatal Exposures and Child
Lifestyle to Insulin Sensitivity. J Clin Endocrinol Metab. Published online April 17,
2020:dgaa201. doi:10.1210/clinem/dgaa201
399. Ball GDC, Huang TT-K, Cruz ML, Shaibi GQ, Weigensberg MJ, Goran MI.
Predicting abdominal adipose tissue in overweight Latino youth. Int J Pediatr
Obes. 2006;1(4):210-216. doi:10.1080/17477160600913578
400. Brambilla P, Bedogni G, Heo M, Pietrobelli A. Waist circumference-to-height ratio
predicts adiposity better than body mass index in children and adolescents. Int J
Obes. 2013;37(7):943-946. doi:10.1038/ijo.2013.32
401. Chula de Castro JA, Lima TR de, Silva DAS. Body composition estimation in
children and adolescents by bioelectrical impedance analysis: A systematic
review. J Bodyw Mov Ther. 2018;22(1):134-146. doi:10.1016/j.jbmt.2017.04.010
402. Matsuda M, DeFronzo RA. Insulin sensitivity indices obtained from oral glucose
tolerance testing: comparison with the euglycemic insulin clamp. Diabetes Care.
1999;22(9):1462-1470.
403. Ainsworth BE, Haskell WL, Herrmann SD, et al. 2011 Compendium of Physical
Activities: a second update of codes and MET values. Med Sci Sports Exerc.
2011;43(8):1575-1581. doi:10.1249/MSS.0b013e31821ece12
404. Pate RR, Ross R, Dowda M, Trost SG, Sirard JR. Validation of a 3-Day Physical
Activity Recall Instrument in Female Youth. Pediatr Exerc Sci. 2003;15(3):257-
265. doi:10.1123/pes.15.3.257
406. Schakel SF, Buzzard IM, Gebhardt SE. Procedures for Estimating Nutrient Values
for Food Composition Databases. J Food Compos Anal. 1997;10(2):102-114.
doi:10.1006/jfca.1997.0527
407. Marshall WA, Tanner JM. Variations in the pattern of pubertal changes in boys.
Arch Dis Child. 1970;45(239):13-23.
142
408. Rasmussen AR, Wohlfahrt-Veje C, Tefre de Renzy-Martin K, et al. Validity of Self-
Assessment of Pubertal Maturation. PEDIATRICS. 2015;135(1):86-93.
doi:10.1542/peds.2014-0793
409. Diana RA, Yonelinas AP, Ranganath C. Imaging recollection and familiarity in the
medial temporal lobe: a three-component model. Trends Cogn Sci.
2007;11(9):379-386. doi:10.1016/j.tics.2007.08.001
410. Gershon RC, Wagster MV, Hendrie HC, Fox NA, Cook KF, Nowinski CJ. NIH
Toolbox for Assessment of Neurological and Behavioral Function. Neurology.
2013;80(11 Supplement 3):S2-S6. doi:10.1212/WNL.0b013e3182872e5f
411. Irby SM, Floyd RG. Test Review: Wechsler Abbreviated Scale of Intelligence,
Second EditionWechslerD.Wechsler Abbreviated Scale of Intelligence, Second
Edition. 2011; San Antonio, TX: Pearson. Can J Sch Psychol. 2013;28(3):295-
299. doi:10.1177/0829573513493982
412. Florian ML, Nunes ML. Effects of intra-uterine and early extra-uterine malnutrition
on seizure threshold and hippocampal morphometry of pup rats. Nutr Neurosci.
2011;14(4):151-158. doi:10.1179/147683010X12611460764804
414. Wang J, Zhang Y-W, Zhang J-Q, et al. Memory Dysfunction in Type 2 Diabetes
Mellitus Correlates with Reduced Hippocampal CA1 and Subiculum Volumes.
Chin Med J (Engl). 2015;128(4):465. doi:10.4103/0366-6999.151082
415. Zhou Y, Li X-L, Xie H-L, et al. Voxel-based morphology analysis of STZ-induced
type 1 diabetes mellitus rats with and without cognitive impairment. Neurosci Lett.
2018;684:210-217. doi:10.1016/j.neulet.2018.08.017
416. Smith SM, Jenkinson M, Johansen-Berg H, et al. Tract-based spatial statistics:
voxelwise analysis of multi-subject diffusion data. Neuroimage. 2006;31(4):1487–
1505.
417. Smith SM, Jenkinson M, Woolrich MW, et al. Advances in functional and structural
MR image analysis and implementation as FSL. NeuroImage. 2004;23:S208-
S219. doi:10.1016/j.neuroimage.2004.07.051
418. Smith SM. Fast robust automated brain extraction. Hum Brain Mapp.
2002;17(3):143-155. doi:10.1002/hbm.10062
419. Jenkinson M, Bannister P, Brady M, Smith S. Improved Optimization for the
Robust and Accurate Linear Registration and Motion Correction of Brain Images.
NeuroImage. 2002;17(2):825-841. doi:10.1006/nimg.2002.1132
420. Hua K, Zhang J, Wakana S, et al. Tract Probability Maps in Stereotaxic Spaces:
Analyses of White Matter Anatomy and Tract-Specific Quantification.
NeuroImage. 2008;39(1):336-347. doi:10.1016/j.neuroimage.2007.07.053
1
Appendix of Figures and Tables
10, 000
$
(&'(×&'*)×(,-./ 1(22 345678-(.9 0, 30, 60, 90, 120 >?/8)
×(,-./ 1(22 ?/854?/ (.9 0, 30, 60, 90, 120 >?/8)
Figure 2. FreeSurfer Subcortical Reconstruction Pipeline.
Figure 1. Formula for Matsuda insulin sensitivity index. FPG, fasting plasma glucose.
FPI, fasting plasma insulin. OGTT, oral glucose tolerance test.
2
Figure 3. Hippocampal sub-regions denoted by different colors. RED: CA1, GREEN:
CA2/3, TURQUOISE: Dentate gyrus, BLUE: subiculum, GREY: CA4.
3
Figure 4. Summary of Types of Memory.
HD: Hippocampal dependent function HR: Hippocampal-related (hippocampus
contributes to function and aspects are hippocampal-dependent).
4
Figure 5. Example of Relational Memory Task.
Four components of composite scoring metric. 1: Swap rate, 2: Distance from original
location, 3: Distance between objects, 4: Distance from overall shape.
5
Table 1. MRI Data Characteristics of the 88 Child Participants and their Mothers.
Child Characteristics Mean (SD) or N (%) Range
Age, years 8.37 (0.89) 7.33~11.23
Body mass index (BMI),
kg/m
2
18.68 (3.97) 13.62~34.01
BMI percentile 68.62 (27.52) 5.28~99.58
BMI z-score 0.73 (1.09) -1.78~2.64
Sex Boys: 37 (42%)
Girls: 51 (58%)
Tanner Stage of Pubertal
Development
Tanner stage 1: 82 (93%)
Tanner stage 2: 5 (6%)
Tanner stage 3: 1 (1%)
Maternal Characteristics
Maternal pre-pregnancy
BMI, kg/m
2
29.86 (6.90) 18.97~50.38
Maternal education Missing: 2 (2%)
a
<=High school: 23 (26%)
a
Some college: 17 (19%)
a
College and post: 46 (52%)
a
Family income Missing: 2 (2%)
0<=income <30 000: 7 (8%)
30000<=income <50 000: 22 (25%)
50000<=income <70 000: 30 (34%)
70000<=income <90 000: 14 (16%)
90000>=income: 13 (15%)
Mother’s race/ethnicity Hispanic: 49 (56%)
Black: 10 (11%)
Non-Hispanic White: 19 (22%)
Other: 10 (11%)
6
Table 2. Characteristics of the 99 Child Participants and their Mothers for Cognitive
data.
Child Characteristics Mean (SD) or N (%) Range
Age, years 8.86 (1.38) 7.34~13.27
Body mass index (BMI),
kg/m
2
19.39 (4.27) 13.94~34.01
BMI percentile 70.07 (28.36) 7.51~99.58
BMI z-score 0.80 (1.09) -1.47~2.64
Sex Girls: 59 (60%)
Boys: 40 (40%)
Tanner Stage of Pubertal
Development
Stage 1: 82 (83%)
a
Stage 2: 8 (8%)
a
Stage 3: 5 (5%)
a
Stage 4: 3 (3%)
a
Maternal Characteristics
Maternal pre-pregnancy
BMI, kg/m
2
30.24 (7.48) 18.97~53.96
Maternal education <=High school: 22 (22%)
a
Some college: 31 (31%)
a
College and post: 46 (46%)
a
Family income 0<=income <30 000: 12 (12%)
a
30000<=income <50 000: 30 (3
a
0%)
a
50000<=income <70 000: 31 (31%)
a
70000<=income <90 000: 15 (15%)
a
90000>=income: 11 (11%)
a
7
5000
5500
6000
6500
7000
7500
8000
8500
9000
15 20 25 30 35 40 45 50 55
Hippocampal Volume (mm
3
)
Pre-pregnancy BMI (kg/m
2
)
Maternal Pre-pregnancy BMI x Child Sex on Child
Hippocampal Volume
Boys Girls
200
400
600
800
1000
1200
1400
1600
1800
15 20 25 30 35 40 45 50 55
Subfields Volume (mm
3
)
Pre-pregnancy BMI (kg/m
2
)
Maternal Pre-pregnancy BMI x Sex on Child
Hippocampal Subfields
B
A
CA1
Subiculum
Dentate Gyrus
CA4
CA2/3
For A) Boys depicted as “turquoise circles”. Girls depicted as “purple squares”; for B)
Boys depicted as “circles”. Girls depicted as “squares”. CA1 subfield denoted in blue,
Subiculum denoted in pink, Dentate Gyrus denoted in red, CA4 denoted in green,
CA2/3 denoted in grey.
Figure 6. Sex stratified scatter plots between maternal pre-pregnancy
BMI and child’s total hippocampal volume (A) and volumes of
hippocampal subfields (B), N=88.
8
Table 3. Regression coefficients between Maternal Pre-pregnancy BMI per 5 unit
increments and Total Hippocampal Volume and Hippocampal Subfield Volumes in
Boys, (N=37).
Model 1: unadjusted
Model 2: adjusted for ICV
Model 3: adjusted for covariates in Model 2 + child age
Model 4: adjusted for covariates in Model 3 + SES + maternal Gestational Diabetes Mellitus (GDM) status
Model 5: adjusted for covariates in Model 4 + BMI z-score
*Denotes a significance level at P<.05.
a
Denotes significance remained after FDR correction for multiple subfields at a threshold of q=.05.
Region
Hippocampus Model 1 Model 2 Model 3 Model 4 Model 5
Beta -134.97 -103.84 -103.55 -113.05 -126.98
95% CI (-245.31, -24.64) (-189.08, -18.60) (-187.05, -20.06) (-199.80, -26.31) (-219.60, -34.36)
Partial Omega
2
0.123 0.122 0.117 0.130 0.144
P-value 0.018 0.018* 0.021* 0.016* 0.012*
CA1 Model 1 Model 2 Model 3 Model 4 Model 5
Beta (SE) -26.88 -21.67 -21.19 -21.57 -23.58
95% CI (-50.89, -2.86) (-42.93, -0.41) (-41.69, -0.70) (-42.88, -0.27) (-46.53, -0.64)
Partial Omega
2
0.101 0.082 0.078 0.074 0.076
P-value 0.029
a
0.046 0.051 0.057 0.054
CA2/3 Model 1 Model 2 Model 3 Model 4 Model 5
Beta (SE) -13.20 -11.05 -11.26 -12.44 -14.19
95% CI (-22.00, -3.41) (-19.69, -2.42) (-19.57, -2.95) (-21.19, -3.68) (-23.46, -4.91)
Partial Omega
2
0.149 0.007 0.141 0.154 0.178
P-value 0.010
a
0.014 0.012 0.009
a
0.006
a
CA4 Model 1 Model 2 Model 3 Model 4 Model 5
Beta (SE) -10.98 -8.38 -8.55 -9.63 -10.34
95% CI (-20.46, -1.50) (-15.87, -0.89) (-15.77, -1.34) (-17.31, -1.95) (-18.61, -2.06)
Partial Omega
2
0.109 0.101 0.106 0.120 0.120
P-value 0.025
a
0.029 0.026 0.020
a
0.021
a
Dentate Gyrus Model 1 Model 2 Model 3 Model 4 Model 5
Beta (SE) -12.30 -9.16 -9.29 -10.36 -11.31
95% CI (-23.34, -1.26) (-17.64, -0.69) (-17.53, -1.05) (-19.22, -1.51) (-20.83, -1.78)
Partial Omega
2
0.100 0.094 0.095 0.103 0.107
P-value 0.030
a
0.035 0.034 0.029
a
0.027
a
Subiculum Model 1 Model 2 Model 3 Model 4 Model 5
Beta (SE) -13.98 -10.49 -10.17 -11.80 -12.51
95% CI (-29.74, 1.77) (-24.34, 3.35) (-23.51, 3.16) (-25.54, 1.95) (-27.35, 2.34)
Partial Omega
2
0.057 0.036 0.032 0.047 0.045
P-value 0.080 0.13 0.14 0.10 0.11
9
Table 4. Regression coefficients between Maternal Pre-pregnancy BMI per 5-unit
increments and Total Hippocampal Volume and Hippocampal Subfield Volumes in Girls,
(N=51).
Model 1: unadjusted
Model 2: adjusted for ICV
Model 3: adjusted for covariates in Model 2 + child age
Model 4: adjusted for covariates in Model 3 + SES + Gestational Diabetes Mellitus (GDM) status
Model 5: adjusted for covariates in Model 4 + BMI z-score
Region
Hippocampus Model 1 Model 2 Model 3 Model 4 Model 5
Beta (SE) 45.51 54.26 51.66 61.39 28.78
95% CI (-73.68,
164.70)
(-42.17,
150.70)
(-44.61,
147.93)
(-41.17,
163.95)
(-79.70, 137.26)
Partial Omega
2
0 0.006 0.002 0.008 0
P-value 0.45 0.26 0.30 0.25 0.61
CA1 Model 1 Model 2 Model 3 Model 4 Model 5
Beta (SE) 10.83 13.07 11.54 14.37 7.56
95% CI (-20.13, 41.79) (-10.97, 37.11) (-12.35, 35.42) (-11.13, 39.87) (-19.67, 34.78)
Partial Omega
2
0 0.004 -0.002 0.005 0
P-value 0.49 0.28 0.35 0.28 0.59
CA2/3 Model 1 Model 2 Model 3 Model 4 Model 5
Beta (SE) -2.53 -1.88 -0.37 0.25 -1.44
95% CI (-14.98, 9.92) (-12.16, 8.41) (-10.33, 9.58) (-10.45, 10.94) (-13.01, 10.14)
Partial Omega
2
0 0 0 0 0
P-value 0.68 0.72 0.94 0.96 0.81
CA4 Model 1 Model 2 Model 3 Model 4 Model 5
Beta (SE) -1.04 -0.39 0.01 0.55 -1.59
95% CI (-12.29, 10.21) (-9.88, 9.10) (-9.44, 9,47) (-9.21, 10.31) (-12.08, 8.89)
Partial Omega
2
0 0 0 0 0
P-value 0.85 0.93 >0.99 0.91 0.77
Dentate Gyrus Model 1 Model 2 Model 3 Model 4 Model 5
Beta (SE) -1.53 -0.61 -0.50 0.40 -2.53
95% CI (-15.02, 11.95) (-14.54, 11.63) (-11.73, 10.74) (-11.27, 12.07) (-15.02, 9.95)
Partial Omega
2
0 0 0 0 0
P-value 0.82 0.91 0.93 0.95 0.69
Subiculum Model 1 Model 2 Model 3 Model 4 Model 5
Beta (SE) 2.72 3.30 2.35 2.97 -0.61
95% CI (-13.37, 18.80) (-11.09, 17.69) (-11.94, 16.63) (-12.52, 18.46) (-17.23, 16.01)
Partial Omega
2
0 0 0 0 0
P-value 0.74 0.65 0.75 0.71 0.94
10
Table 5. Relationships between Maternal Pre-pregnancy BMI (5-unit increments) and
Total Hippocampal Volume and Hippocampal Subfield Volume (N=88).
Region
Hippocampus Model 1 Model 2 Model 3 Model 4 Model 5
Beta (SE) -46.97 -46.14 -106.98 -113.52 -130.13
95% CI (-136.13,
42.19)
(-109.40,
17.11)
(-184.20, -
29.76)
(-193.42, -
33.62)
(-210.24, -
50.02)
Partial Omega
2
0 0.012 0 0 0.016
P-value 0.30 0.16 0.008* 0.007* 0.002*
CA1 Model 1 Model 2 Model 3 Model 4 Model 5
Beta (SE) -8.34 -8.14 -20.86 -21.11 -24.32
95% CI (-28.89, 12.21) (-23.73, 7.44) (-40.10, -1.63) (-41.76, -1.84) (-44.66, -3.98)
Partial Omega
2
0 0.001 0 0 0
P-value 0.42 0.31 0.037
a
0.036
a
0.022
a
CA2/3 Model 1 Model 2 Model 3 Model 4 Model 5
Beta (SE) -8.34 -7.47 -10.89 -11.97 -13.10
95% CI (-15.97, -0.70) (-13.87, -1.08) (-18.91, -2.88) (-20.23, -3.71) (-21.50, -4.70)
Partial Omega
2
0.041 0.046 0.028 0.029 0.044
P-value 0.033
0.025 0.009
a
0.006
a
0.003
a
CA4 Model 1 Model 2 Model 3 Model 4 Model 5
Beta (SE) -5.90 -5.65 -8.78 -9.66 -10.82
95% CI (-13.33, 1.52) (-11.42, 0.12) (-16.01, -1.55) (-17.19, -2.12) (-18.46, -3.17)
Partial Omega
2
0.017 0.030 0.015 0.017 0.034
P-value 0.12 0.058 0.020
a
0.014
a
0.007
a
Dentate Gyrus Model 1 Model 2 Model 3 Model 4 Model 5
Beta (SE) -6.51 -6.26 -9.61 -10.48 -11.99
95% CI (-15.37, 2.35) (-12.96, 0.44) (-18.02, -1.20) (-19.24, -1.73) (-20.84, -3.14)
Partial Omega
2
0.013 0.026 0.013 0.014 0.033
P-value 0.15 0.071 0.028
a
0.022
a
0.010
a
Subiculum Model 1 Model 2 Model 3 Model 4 Model 5
Beta (SE) 5.00 -5.38 -11.00 -11.71 -13.37
95% CI (-17.08, 7.07) (-15.00, 4.23) (-23.01, 1.03) (-23.97, 0.55) (-25.85, -0.90)
Partial Omega
2
0 0.002 0 0 0.006
P-value 0.41 0.28 0.077 0.065 0.039
a
Model 1: unadjusted
Model 2: adjusted for ICV + child age + sex
Model 3: adjusted for ICV + child age + sex + interaction of pre-pregnancy BMI and sex
Model 4: adjusted for ICV + child age + sex + interaction of pre-pregnancy BMI and sex + SES + GDM
Model 5: adjusted for ICV + child age + sex + interaction of pre-pregnancy BMI and sex + SES + GDM + BMI z-score
*Denotes a significance level at P<0.05.
a
Denotes significance remained after FDR was used for multiple comparisons within each model.
11
Table 6. Correlation Matrix for Hippocampal-Related Memory.
Variables Pearson R p-value
Child age and Relational Memory 0.42 <0.0001*
Child age and Item-familiarity 0.29 0.016*
Child IQ and Relational Memory 0.28 0.009*
Child IQ and Item-familiarity 0.29 0.013*
Tanner Stage and Relational Memory 0.23 0.034*
Tanner Stage and Item-familiarity 0.18 0.13
*Denotes p-value <0.05.
Table 7. Association between maternal pre-pregnancy BMI and hippocampal-
dependent memory (relational memory), hippocampal-independent memory (item-
familiarity), picture-sequence memory task, and full-scale IQ.
Dependent
Variable
Beta (95%CI) p-value Partial
Omega
2
Covariates
Relational
Memory
-0.006 (-0.022, 0.009) 0.44 0 Unadjusted
-0.002 (-0.016, 0.012) 0.79 0 Full-scale IQ, age
0.002 (-0.015, 0.019) 0.82 0 +sex, SES, GDM, and BMI z-score
0.002 (-0.015, 0.018) 0.85 0 +Tanner stage
Item-
familiarity
-0.002 (-0.008, 0.004) 0.47 0 Unadjusted
-0.001 (-0.006, 0.005) 0.81 0 Full-scale IQ, age
0 (-0.006, 0.006) 0.98 0 +sex, SES, GDM, and BMI z-score
0 (-0.007, 0.006) 0.94 0 +Tanner stage
Episodic
Memory
(PSMT)
0.045 (-0.278, 0.369) 0.78 0 Unadjusted
0.070 (-0.286, 0.426) 0.70 0 +Age, sex, and BMI z-score
-0.057 (-0.44, 0.326) 0.77 0 +SES, GDM, and Tanner stage
Full-scale
IQ
-0.306 (-0.672, 0.06) 0.105
0.017 Unadjusted
-0.425 (-0.81, -0.04) 0.033* 0.036 +Age, sex, and BMI z-score
-0.352 (-0.76, 0.056) 0.094
#
0.018 +SES, GDM, and Tanner
SES (college education and household income at birth). Picture-sequence memory task (PSMT). *Denotes a
significance level at P<0.05.
#
Denotes a trending p-value.
12
Table 8. In girls, the association between maternal pre-pregnancy BMI and
hippocampal-dependent memory (Relational Memory), hippocampal-independent
memory (Item-familiarity), Picture-sequence Memory Task, and Full-scale IQ.
Dependent
Variable
Beta (95%CI) p-value Partial
Omega
2
Covariates
Relational
Memory
-0.009 (-0.032, 0.014) 0.44 0 Unadjusted
-0.004 (-0.025, 0.016) 0.68 0 Full-scale IQ, age
-0.002 (-0.024, 0.02) 0.89 0 +SES, GDM, and BMI z-score
-0.002 (-0.024, 0.02) 0.86 0 +Tanner stage
Item-
familiarity
-0.004 (-0.013, 0.005) 0.41 0 Unadjusted
-0.002 (-0.011, 0.007) 0.62 0 Full-scale IQ, age
-0.003 (-0.012, 0.006) 0.49 0 +SES, GDM, and BMI z-score
-0.003 (-0.011, 0.006) 0.54 0 +Tanner stage
Episodic
Memory
(PSMT)
-0.061 (-0.581, 0.46) 0.82 0 Unadjusted
0.071 (-0.493, 0.635) 0.81 0 +Age, BMI z-score
0.125 (-0.451, 0.701) 0.67 0 +SES, GDM, and Tanner Stage
Full-scale
IQ
-0.523 (-1.045, 0) 0.055
#
0.046 Unadjusted
-0.515 (-1.075, 0.046) 0.077
#
0.037 +Age, BMI z-score
-0.356 (-0.938, 0.226) 0.24 0.007 +SES, GDM, and Tanner Stage
SES (college education and household income at birth). Picture-sequence memory task (PSMT).
#
Denotes a trending
p-value.
Table 9. In boys, the association between maternal pre-pregnancy BMI and
hippocampal-dependent memory (Relational Memory), hippocampal-independent
memory (Item-familiarity), Picture-sequence Memory Task, and Full-scale IQ.
Dependent
Variable
Beta (95%CI) p-value Partial
Omega
2
Covariates
Relational
Memory
-0.003 (-0.024, 0.018) 0.79 0 Unadjusted
0 (-0.021, 0.021) 0.99 0 +Full-scale IQ, age
0.007 (-0.021, 0.036) 0.62 0 +SES, GDM, and BMI z-score
-0.002 (-0.035, 0.03) 0.88 0 +Tanner stage
Item-
familiarity
-0.001 (-0.008, 0.007) 0.84 0 Unadjusted
0.001 (-0.007, 0.008) 0.87 0 +Full-scale IQ, age
0.003 (-0.007, 0.012) 0.56 0 +SES, GDM, and BMI z-score
0.002 (-0.01, 0.013) 0.75 0 +Tanner stage
Episodic
Memory
(PSMT)
0.082 (-0.351, 0.516) 0.71 0 Unadjusted
0.061 (-0.409, 0.53) 0.80 0 +Age, BMI z-score
-0.025 (-0.631, 0.581) 0.94 0 +SES, GDM, and Tanner Stage
Full-Scale
IQ
-0.117 (-0.633, 0.398) 0.66 0 Unadjusted
-0.393 (-0.899, 0.113) 0.14 0.032 +Age, BMI z-score
-0.39 (-1.042, 0.263) 0.25 0.031 +SES, GDM, and Tanner Stage
SES (college education and household income at birth). Picture-sequence memory task (PSMT).
13
Table 10. The association between relational memory performance and hippocampal
volume. ICV-adjusted*.
Iposition Task (Relational Memory) Beta (95%CI) P-value N
Left Hippocampus -0.0001 (-0.0006, 0.0004) 0.765
66
Boys 0.0008 (0, 0.0016) 0.055
#
28
Girls -0.0005 (-0.0012, 0.0001) 0.105 38
Right Hippocampus -0.0001 (-0.0007, 0.0005) 0.683
67
Boys 0.0002 (-0.0006, 0.0011) 0.603
29
Girls -0.0005 (-0.0014, 0.0003) 0.230 38
Left CA1 -0.0008 (-0.0027, 0.0011) 0.410
66
Boys 0.0016 (-0.0021, 0.0053) 0.416
28
Girls -0.0020 (-0.0043, 0.0004) 0.115 38
Right CA1 -0.0001 (-0.0022, 0.002) 0.927
67
Boys 0.0012 (-0.002, 0.0045) 0.460
29
Girls -0.0016 (-0.0046, 0.0014) 0.314 38
Left CA2/3 -0.0009 (-0.0059, 0.0041) 0.714
66
Boys 0.0027 (-0.0047, 0.0101) 0.483
28
Girls -0.0038 (-0.0105, 0.0029) 0.270 38
Right CA2/3 0.0013 (-0.0048, 0.0074) 0.682
67
Boys 0.0068 (-0.0022, 0.0158) 0.152
29
Girls -0.0023 (-0.0107, 0.0061) 0.594 38
Left CA4 0.0001 (-0.0057, 0.0059) 0.969
66
Boys 0.0055 (-0.0036, 0.0146) 0.249
28
Girls -0.0033 (-0.0111, 0.0044) 0.407 38
Right CA4 -0.0013 (-0.0078, 0.0053) 0.702
67
Boys 0.0021 (-0.009, 0.0131) 0.717
29
Girls -0.0035 (-0.0124, 0.0053) 0.440 38
Left Subiculum 0.002 (-0.0016, 0.0056) 0.274
66
Boys 0.0057 (0.0016, 0.0098) 0.012*
28
Girls -0.0024 (-0.0087, 0.004) 0.470 38
Right Subiculum -0.0007 (-0.0046, 0.0032) 0.726
67
Boys 0.0008 (-0.0044, 0.0061) 0.752
29
Girls -0.0031 (-0.0093, 0.0032) 0.341 38
Left Dentate Gyrus 0.0004 (-0.0048, 0.0056) 0.868
66
Boys 0.0062 (-0.0021, 0.0145) 0.153
28
Girls -0.003 (-0.01, 0.004) 0.410 38
Right Dentate Gyrus -0.0006 (-0.0064, 0.0051) 0.829
67
Boys 0.0018 (-0.0078, 0.0115) 0.711
29
Girls -0.0027 (-0.0108, 0.0054) 0.520 38
14
Table 11. The association between episodic memory performance and hippocampal
volume. ICV-adjusted*.
PSMT (Episodic Memory) Beta (95%CI) P-value N
Left Hippocampus 0.0119 (0.0008, 0.0231) 0.040* 80
Boys 0.0114 (-0.0121, 0.0348) 0.350 34
Girls 0.0115 (-0.0010, 0.0239) 0.078
#
46
Right Hippocampus 0.0072 (-0.0060, 0.0203) 0.288 81
Boys 0.0032 (-0.0192, 0.0256) 0.781 35
Girls 0.0097 (-0.0062, 0.0255) 0.238 46
Left CA1 0.0341 (-0.0091, 0.0774) 0.126 80
Boys 0.0024 (-0.0856, 0.0904) 0.959 34
Girls 0.0409 (-0.0072, 0.089) 0.103 46
Right CA1 0.0118 (-0.0363, 0.0598) 0.633 81
Boys -0.0177 (-0.0928, 0.0574) 0.647 35
Girls 0.031 (-0.0316, 0.0936) 0.337 46
Left CA2/3 0.1359 (0.0153, 0.2565) 0.030* 80
Boys 0.1015 (-0.1332, 0.3362) 0.403 34
Girls 0.1339 (0.0019, 0.2658) 0.053
#
46
Right CA2/3 0.1064 (-0.0233, 0.2361) 0.112 81
Boys 0.2522 (0.0285, 0.4759) 0.034* 35
Girls 0.0379 (-0.1105, 0.1863) 0.619 46
Left CA4 0.1572 (0.0211, 0.2932) 0.026* 80
Boys 0.1462 (-0.1215, 0.414) 0.293 34
Girls 0.137 (-0.0185, 0.2924) 0.091
#
46
Right CA4 .0695 (-0.0642, 0.2032) 0.3114 81
Boys 0.3033 (0.0284, 0.5782) 0.038* 35
Girls 0.0072 (-0.1344, 0.1488) 0.921 46
Left Subiculum 0.0442 (-0.0317, 0.12) 0.257 80
Boys 0.0626 (-0.0611, 0.1864) 0.329 34
Girls 0.0292 (-0.0744, 0.1329) 0.583 46
Right Subiculum 0.029 (-0.0553, 0.1134) 0.502 81
Boys 0.0103 (-0.1277, 0.1483) 0.885 35
Girls 0.0439 (-0.0679, 0.1556) 0.446 46
Left Dentate Gyrus 0.1143 (-0.0038, 0.2323) 0.062
#
80
Boys 0.0507 (-0.1834, 0.2847) 0.674 34
Girls 0.125 (-0.0101, 0.2602) 0.077
#
46
Right Dentate Gyrus 0.0641 (-0.053, 0.1812) 0.2867 81
Boys 0.2243 (-0.0158, 0.4644) 0.076
#
35
Girls 0.0166 (-0.1098, 0.1431) 0.798 46
15
Table 12. The association between item-familiarity performance and hippocampal
volume. ICV-adjusted*.
Creature-scene task (Item-familiarity) Beta (95%CI) P-value N
Left Hippocampus 0 (-0.0002, 0.0002) 0.835 59
Boys 0 (-0.0003, 0.0004) 0.897 27
Girls 0 (-0.0002, 0.0002) 0.922 32
Right Hippocampus 0 (-0.0002, 0.0002) 0.902 60
Boys 0.0001 (-0.0003, 0.0004) 0.704 28
Girls 0 (-0.0003, 0.0003) 0.827 32
Left CA1 0 (-0.0007, 0.0007) 0.929 59
Boys 0.0002 (-0.0013, 0.0016) 0.799 27
Girls -0.0001 (-0.001, 0.0007) 0.747 32
Right CA1 0.0001 (-0.0007, 0.0009) 0.804 60
Boys 0.0002 (-0.001, 0.0015) 0.717 28
Girls 0 (-0.0011, 0.001) 0.950 32
Left CA2/3 0.0001 (-0.0018, 0.0019) 0.937 59
Boys 0.0013 (-0.0017, 0.0042) 0.411 27
Girls -0.0007 (-0.0032, 0.0017) 0.567 32
Right CA2/3 0.0007 (-0.0015, 0.003) 0.529 60
Boys 0.0013 (-0.002, 0.0047) 0.432 28
Girls 0.0003 (-0.003, 0.0037) 0.849 32
Left CA4 -0.0004 (-0.0026, 0.0017) 0.692 59
Boys 0.0008 (-0.0028, 0.0045) 0.666 27
Girls -0.0012 (-0.0039, 0.0015) 0.400 32
Right CA4 0.0003 (-0.0021, 0.0027) 0.805 60
Boys 0.0004 (-0.0036, 0.0044) 0.851 28
Girls 0.0003 (-0.0028, 0.0035) 0.833 32
Left Subiculum 0.0008 (-0.0004, 0.0021) 0.205 59
Boys 0.0004 (-0.0015, 0.0023) 0.705 27
Girls 0.0016 (-0.0005, 0.0036) 0.150 32
Right Subiculum -0.0001 (-0.0016, 0.0014) 0.892 60
Boys 0.0005 (-0.0016, 0.0026) 0.660 28
Girls -0.0007 (-0.0029, 0.0014) 0.517 32
Left Dentate Gyrus -0.0003 (-0.0022, 0.0016) 0.756 59
Boys 0.001 (-0.0024, 0.0044) 0.555 27
Girls -0.0011 (-0.0035, 0.0014) 0.399 32
Right Dentate Gyrus 0.0002 (-0.0019, 0.0023) 0.858 60
Boys 0.0002 (-0.0033, 0.0038) 0.897 28
Girls 0.0002 (-0.0026, 0.0031) 0.877 32
16
Table 13. LSmeans of Hippocampal Volume stratified by GDM-exposure in Boys.
Model Group
Left Hippocampus Right Hippocampus
Mean (SE) P-value LSmean (SE) P-value
1
Non-exposed 3668.94 (71.03) 0.22
3700.09 (79.14) 0.29
GDM-exposed 3549.90 (63.54) 3586.51 (70.79)
2
Non-exposed 3650.85 (55.81) 0.26
3678.69 (59.45) 0.35
GDM-exposed 3564.38 (49.89) 3603.63 (53.15)
3
Non-exposed 3661.83 (58.33) 0.20
3676.51 (62.61) 0.42
GDM-exposed 3555.59 (51.78) 3605.38 (55.58)
4
Non-exposed 3638.30 (58.67) 0.25
3666.39 (64.25) 0.46
GDM-exposed 3547.00 (51.49) 3603.03 (56.39)
5
Non-exposed 3642.28 (59.30) 0.22
3674.75 (62.95) 0.34
GDM-exposed 3542.18 (52.22) 3592.89 (55.43)
Model 1: unadjusted
Model 2: adjusted for ICV
Model 3: adjusted for covariates in Model 2 + child age
Model 4: adjusted for covariates in Model 3 + SES + maternal pre-pregnancy BMI
Model 5: adjusted for covariates in Model 4 + BMI z-score
*Denotes a significance level at P<.05.
Table 14. LSmeans of Hippocampal Subfield Volume stratified by GDM-exposure in
Boys.
Model Group
Left CA1 Right CA1 Left CA2/3 Right CA2/3
LSmean
(SE)
P-value LSmean
(SE)
P-
value
LSmean
(SE)
P-
value
LSmean
(SE)
P-
value
1
Non-
exposed
701.19
(14.14) 0.022
#
711.68
(17.39) 0.15
209.72
(7.28) 0.84
227.82
(7.20) 0.67
GDM-
exposed
655.59
(12.65)
677.58
(15.56)
211.69
(6.51)
223.68
(6.44)
2
Non-
exposed
698.28
(12.34) 0.021
#
708.01
(15.05) 0.18
208.24
(6.38) 0.59
226.50
(6.51) 0.84
GDM-
exposed
657.92
(11.03)
680.52
(13.45)
212.87
(5.71)
224.74
(5.82)
3
Non-
exposed
698.91
(12.99) 0.027
#
705.00
(15.73) 0.32
211.06
(6.46) 0.96
226.63
(6.85) 0.83
GDM-
exposed
657.42
(11.53)
682.93
(13.96)
210.62
(5.73)
224.63
(6.08)
4
Non-
exposed
694.54
(13.19) 0.033
#
699.53
(16.96)
0.35
210.93
(6.78) 0.79
229.46
(7.04) 0.97
GDM-
exposed
655.26
(11.58)
678.37
(14.89)
213.40
(5.95)
229.12
(6.18)
5
Non-
exposed
695.30
(13.38) 0.030
#
700.96
(17.04)
0.29
211.05
(6.92) 0.81
230.58
(6.75) 0.76
GDM-
exposed
654.32
(11.78)
676.65
(15.01)
213.25
(6.10)
227.76
(5.94)
#
Denotes marginal significance after controlling for multiple comparisons.
17
Table 15. Effect size of GDM-exposure as Cohen’s D in Boys.
Mode
l Group
L.
Hippocampu
s
R.
Hippocampu
s
L. CA1 R. CA1 L. CA2/3 R. CA2/3
Cohen’s D Cohen’s D Cohen’s
D
Cohen’s
D
Cohen’s
D
Cohen’s
D
1 Non-
expose
d 0.41 0.35 0.79 0.48 -0.07 0.14
GDM-
expose
d Reference Reference
Referenc
e
Referenc
e
Referenc
e
Referenc
e
2 Non-
expose
d 0.38 0.31 0.81 0.45 -0.18 0.07
GDM-
expose
d Reference Reference
Referenc
e
Referenc
e
Referenc
e
Referenc
e
3 Non-
expose
d 0.45 0.28 0.79 0.35 0.02 0.07
GDM-
expose
d Reference Reference
Referenc
e
Referenc
e
Referenc
e
Referenc
e
4 Non-
expose
d 0.39 0.24 0.74 0.31 -0.09 0.01
GDM-
expose
d Reference Reference
Referenc
e
Referenc
e
Referenc
e
Referenc
e
5 Non-
expose
d 0.42 0.32 0.76 0.35 -0.08 0.10
GDM-
expose
d Reference Reference
Referenc
e
Referenc
e
Referenc
e
Referenc
e
Model 1: unadjusted
Model 2: adjusted for ICV
Model 3: adjusted for covariates in Model 2 + child age
Model 4: adjusted for covariates in Model 3 + SES + maternal pre-pregnancy BMI
Model 5: adjusted for covariates in Model 4 + BMI z-score
18
Table 16. LSmeans of Hippocampal Volume stratified by GDM-exposure in Girls.
Model 1: unadjusted
Model 2: adjusted for ICV
Model 3: adjusted for covariates in Model 2 + child age
Model 4: adjusted for covariates in Model 3 + SES + maternal pre-pregnancy BMI
Model 5: adjusted for covariates in Model 4 + BMI z-score
Table 17. LSmeans of Hippocampal Subfields stratified by GDM-exposure in Girls.
Model Group
Left CA1 Right CA1 Left CA2/3 Right CA2/3
LSmean
(SE)
P-value LSmean
(SE)
P-value LSmean
(SE)
P-value LSmean
(SE)
P-value
1
Non-
exposed
596.35
(12.80) 0.10
615.97
(15.75) 0.098
187.89
(5.53) 0.065
210.44
(6.19) 0.18
GDM-
exposed
622.69
(9.32)
648.81
(11.46)
200.79
(4.02)
220.88
(4.51)
2
Non-
exposed
598.15
(10.32) 0.12
618.42
(12.65) 0.12
188.55
(5.09) 0.072
211.40
(4.93) 0.24
GDM-
exposed
618.27
(7.58)
643.66
(9.34)
200.19
(3.76)
218.72
(3.64)
3
Non-
exposed
598.15
(10.32) 0.12
618.15
(12.64) 0.11
188.69
(5.05) 0.075
211.58
(4.80) 0.24
GDM-
exposed
618.35
(7.62)
643.80
(9.33)
200.12
(3.73)
218.62
(3.54)
4
Non-
exposed
594.28
(11.72) 0.10
617.30
(14.97) 0.14
184.88
(5.82) 0.053
211.88
(5.69) 0.36
GDM-
exposed
617.52
(8.25)
644.13
(10.54)
198.49
(4.10)
218.12
(4.01)
5
Non-
exposed
595.44
(12.00) 0.14
619.70
(15.24) 0.21
184.24
(5.96) 0.045
#
213.25
(5.71) 0.55
GDM-
exposed
616.85
(8.41)
642.73
(10.68)
198.87
(4.18)
217.31
(4.00)
#
Denotes marginal significance after controlling for multiple comparisons.
Group Left Hippocampus Right Hippocampus
Model Mean (SE) P-value LSmean (SE) P-value
1
Non-exposed 3255.94 (50.79) 0.17
3308.35 (57.60) 0.099
GDM-exposed 3342.67 (36.95) 3428.75 (42.54)
2
Non-exposed 3263.68 (42.23) 0.22
3318.28 (45.62) 0.070
GDM-exposed 3329.15 (31.17) 3423.33 (33.68)
3
Non-exposed 3263.45 (42.65) 0.22
3317.69 (45.90) 0.070
GDM-exposed 3329.28 (31.48) 3423.66 (33.88)
4
Non-exposed 3244.28 (48.70) 0.19
3319.72 (53.93) 0.11
GDM-exposed 3319.88 (34.29) 3423.69 (37.97)
5
Non-exposed 3252.21 (49.54) 0.29
3331.58 (54.37) 0.19
GDM-exposed 3315.24 (34.72) 3416.75 (38.11)
19
Table 18. Effect size of GDM-exposure as Cohen's D in Girls.
Mode
l Group
L.
Hippocampu
s
R.
Hippocampu
s
L. CA1 R. CA1 L. CA2/3 R. CA2/3
Cohen’s D Cohen’s D Cohen’s
D
Cohen’s
D
Cohen’s
D
Cohen’s
D
1 Non-
expose
d
-0.41 -0.49 -0.49 -0.50 -0.55 -0.40
GDM-
expose
d
Reference Reference
Referenc
e
Referenc
e
Referenc
e
Referenc
e
2 Non-
expose
d
-0.37 -0.54 -0.46 -0.47 -0.54 -0.35
GDM-
expose
d
Reference Reference
Referenc
e
Referenc
e
Referenc
e
Referenc
e
3 Non-
expose
d
-0.36 -0.54 -0.46 -0.48 -0.53 -0.35
GDM-
expose
d
Reference Reference
Referenc
e
Referenc
e
Referenc
e
Referenc
e
4 Non-
expose
d
-0.37 -0.47 -0.48 -0.43 -0.56 -0.26
GDM-
expose
d
Reference Reference
Referenc
e
Referenc
e
Referenc
e
Referenc
e
5 Non-
expose
d
-0.31 -0.38 -0.43 -0.37 -0.59 -0.17
GDM-
expose
d
Reference Reference
Referenc
e
Referenc
e
Referenc
e
Referenc
e
Model 1: unadjusted
Model 2: adjusted for ICV
Model 3: adjusted for covariates in Model 2 + child age
Model 4: adjusted for covariates in Model 3 + SES + maternal pre-pregnancy BMI
Model 5: adjusted for covariates in Model 4 + BMI z-score
20
Table 19. LSmeans of hippocampal-dependent memory (Relational Memory),
hippocampal-independent memory (Item-familiarity), Picture-sequence Memory Task
and Full-Scale IQ, in boys.
Non-
exposed
GDM-
exposed P-
value Covariates controlled for Mean (SE) Mean (SE)
Accurate Single
Item Placement
(Relational
Memory)
1.89 (0.13) 1.87 (0.13)
0.90 unadjusted
1.84 (0.14) 1.88 (0.13) 0.83 +Age IQ
1.85 (0.17) 1.90 (0.15) 0.84 +Age, IQ, SES, pre-pregnancy BMI
1.86 (0.17) 1.89 (0.15) 0.92
+Age, IQ, SES, pre-pregnancy BMI, BMI
z-score, tanner
Item-familiarity
Performance
0.76 (0.05) 0.71 (0.05) 0.42 unadjusted
0.73 (0.05) 0.72 (0.05) 0.89 +Age IQ
0.75 (0.06) 0.72 (0.05) 0.70 +Age, IQ, SES, pre-pregnancy BMI
0.75 (0.06) 0.72 (0.05) 0.72
+Age, IQ, SES, pre-pregnancy BMI, BMI
z-score, tanner
Picture-
sequence
Episodic
Memory Task
50.89 (2.50) 48.37 (2.58) 0.50 Unadjusted
50.96 (2.70) 48.30 (2.63) 0.50 +Age, BMI z-score
52.54 (3.10) 47.59 (2.95)
0.28
+SES, pre-pregnancy BMI, and Tanner
stage
Full-Scale IQ
106.94
(3.59)
108.17
(3.39)
0.81 unadjusted
106.27
(3.03)
109.47
(2.88)
0.46 +Age, BMI z-score
106.00
(3.89)
108.06
(3.31)
0.70 +Age, SES, pre-pregnancy BMI, BMI z-
score, tanner
21
Table 20. Effect size of GDM-exposure on cognition as Cohen’s D, in boys.
Model Group
Relational
Memory
Item-familiarity
Performance
Episodic
Memory
Full-scale IQ
Cohen’s D Cohen’s D Cohen’s D Cohen’s D
1 Non-
exposed 0.04 0.28 0.22 -0.09
GDM-
exposed Reference
Reference Reference Reference
2 Non-
exposed -0.08 0.05 0.23 -0.26
GDM-
exposed Reference
Reference Reference Reference
3 Non-
exposed -0.08 0.14 0.20 -0.14
GDM-
exposed Reference
Reference Reference Reference
4 Non-
exposed -0.04 0.13
GDM-
exposed Reference
Reference
22
Table 21. LSmeans of hippocampal-dependent memory (Relational Memory),
hippocampal-independent memory (Item-familiarity), Picture-sequence Memory Task
and Full-Scale IQ, in girls.
Non-
exposed
GDM-
exposed P-
value Covariates controlled for Mean (SE) Mean (SE)
Accurate Single
Item Placement
(Relational
Memory)
1.69 (0.11)
1.78 (0.12)
0.57
unadjusted
1.76 (0.10)
1.70 (0.10) 0.67
+Age IQ
1.72 (0.10) 1.67 (0.11) 0.73 +Age, IQ, SES, pre-pregnancy BMI
1.72 (0.11) 1.66 (0.11) 0.70
+Age, IQ, SES, pre-pregnancy BMI, BMI
z-score, tanner
Item-familiarity
Performance
0.71 (0.05) 0.64 (0.05) 0.30 unadjusted
0.73 (0.04) 0.63 (0.04) 0.14 +Age IQ
0.69 (0.05) 0.60 (0.05) 0.17 +Age, IQ, SES, pre-pregnancy BMI
0.69 (0.04) 0.60 (0.05) 0.18
+Age, IQ, SES, pre-pregnancy BMI, BMI
z-score, tanner
Picture-
sequence
Episodic
Memory Task
42.25 (2.20) 52.04 (1.89) 0.002 Unadjusted
42.01 (2.19) 52.21 (1.88) 0.001 +Age, BMI z-score
42.11 (2.78) 52.58 (2.12)
0.003
+SES, pre-pregnancy BMI, and Tanner
stage
Full-Scale IQ
102.65
(2.78)
106.88
(2.90)
0.30 unadjusted
103.94
(2.73)
106.73
(2.59)
0.47 +Age, BMI z-score
103.23
(2.88)
105.27
(2.99)
0.61 +Age, SES, pre-pregnancy BMI, BMI z-
score, tanner
23
Table 22. Effect size of GDM-exposure on cognition as Cohen’s D, in Girls.
Model Group
Relational
Memory
Item-familiarity
Performance
Episodic
Memory
Full-scale IQ
Cohen’s D Cohen’s D Cohen’s D Cohen’s D
1 Non-
exposed -0.17 0.33 -0.98 -0.30
GDM-
exposed Reference
Reference Reference Reference
2 Non-
exposed 0.13 0.48 -1.04 -0.21
GDM-
exposed Reference
Reference Reference Reference
3 Non-
exposed 0.10 0.42 -0.89 -0.14
GDM-
exposed Reference
Reference Reference Reference
4 Non-
exposed 0.11 0.41
GDM-
exposed Reference
Reference
24
Figure 7. Enrollment Flowchart BrainChild.
114 participants enrolled
91 participants completed OGTT
23 Participants unable to complete OGTT
16 Unwilling to do OGTT
6 Catheter difficulties
1 Partial OGTT
1 Refused to drink Glucola
25
Table 23. Summary of Demographics for BrainChild Cohort.
Variable N(%) or Mean (SD) of N=23
uncompleted OGTT
N(%) or Mean (SD) of
N=91
Completed OGTT
p-
value*
Age (years) 8.14 (0.48) 8.34 (0.87) 0.13
Sex Girls: 17 (74%)
Boys: 6 (26%)
Girls: 52 (57%)
Boys: 39 (43%)
0.14
BMI z-score 0.89 (1.19) 0.64 (1.07) 0.28
BMI category
Healthy-weight: 11 (48%)
Overweight: 3 (13%)
Obese: 9 (39%)
Healthy-weight: 60 (66%)
Overweight: 12 (13%)
Obese: 19 (21%)
0.18
Tanner stage Stage 1: 21 (92%)
Stage 2: 1 (4%)
Stage 3: 1 (4%)
Stage 1: 85 (93%)
Stage 2: 5 (6%)
Stage 3: 1 (1%)
0.56
Daily energy intake
(kcal)
1759.40 (368.60) 1791.19 (446.65) 0.77
Dietary added sugar
(%)
15.05 (6.38) 14.94 (6.56) 0.94
Time in MVPA
(min/day)
145.00 (118.3) 134.94 (94.40) 0.64
Time in SB
(min/day)
585.70 (143.30) 609.48 (122.6) 0.45
Maternal pre-
pregnancy BMI
(kg/m
2
)
31.30 (7.11) 29.95 (7.16) 0.42
GDM exposure GDM-exposed: 11 (48%)
Unexposed: 12 (52%)
GDM-exposed: 54 (59%)
Unexposed: 37 (40%)
0.53
*Unpaired t tests and chi-square tests were used to compare group means and frequencies of
participants who completed OGTT and participants who did not complete OGTT.
26
Table 14. Chapter Two: Participant Demographics (N=91).
Variable N(%) or Mean (SD) Range
Age (years) 8.34 (0.87) 7.19~11.23
Sex Girls: 52 (57%)
Boys: 39 (43%)
BMI z-score 0.64 (1.07) -1.78~2.64
BMI category Healthy-weight: 60 (66%)
Overweight: 12 (13%)
Obese: 19 (21%)
Tanner stage Stage 1: 85 (93%)
Stage 2: 5 (6%)
Stage 3: 1 (1%)
Matsuda insulin
sensitivity index
10.58 (5.92) 0.94~26.20
Daily energy intake (kcal) 1791.19 (446.65) 825.09~3006.12
Dietary added sugar (%) 14.94 (6.56) 2.65~39.90
Time in MVPA (min/day) 134.94 (94.40) 0~420.00
Time in SB (min/day) 609.48 (122.60) 220.00~830.00
Maternal pre-pregnancy
BMI (kg/m
2
)
29.95 (7.16) 18.97~50.38
GDM exposure GDM-exposed: 54 (59%)
Unexposed: 37 (40%)
27
Table 25. Summary of unadjusted and adjusted linear regression models, with Matsuda
insulin sensitivity index as the outcome variable.
Predictor Variables Beta (95% CI) p-value Covariates
Time in MVPA
0.29
a
(0.09, 0.49) 0.005* Unadjusted
0.29
a
(0.09, 0.48) 0.005*
Child age, sex, pre-pregnancy BMI,
EI, GDM exposure, added sugar
0.28
a
(0.10, 0.45) 0.002* +BMI z-score
Time in SB
-0.13
a
(-0.33, 0.08) 0.23 Unadjusted
-0.12
a
(-0.32, 0.09) 0.26
Child age, sex, pre-pregnancy BMI,
EI, GDM exposure, added sugar
-0.14
a
(-0.32, 0.04) 0.13 +BMI z-score
Daily energy intake
0.13
a
(-0.08, 0.33) 0.24 Unadjusted
0.02
a
(-0.19, 0.23) 0.85
Child age, sex, pre-pregnancy BMI,
GDM exposure, added sugar, MVPA
0.04
a
(-0.14, 0.23) 0.64 +BMI z-score
Dietary added
sugar
0.002
a
(-0.207, 0.212) 0.98 Unadjusted
0.08
a
(-0.11, 0.28) 0.42
Child age, sex, pre-pregnancy BMI,
EI, GDM exposure, MVPA
0.01
a
(-0.16, 0.19) 0.90 +BMI z-score
Maternal pre-
pregnancy BMI
-0.16
a
(-0.37, 0.05) 0.13 Unadjusted
-0.17
a
(-0.37, 0.02) 0.090
Child age, sex, EI, GDM exposure,
added sugar, MVPA
-0.02
a
(-0.20, 0.16) 0.83 +BMI z-score
GDM exposure
0.07 (-0.35, 0.49) 0.75 Unadjusted
0.11 (-0.30, 0.52) 0.60
Child age, sex, EI, pre-pregnancy
BMI, added sugar, MVPA
0.23 (-0.14, 0.60) 0.22 +BMI z-score
*Denotes p-value<0.05. MVPA=moderate to vigorous physical activity. SB=sedentary behavior. EI=daily
energy intake. GDM=gestational diabetes mellitus.
a
Standardized regression coefficient.
28
Figure 8. LSmeans of Insulin Sensitivity by Terciles of Moderate to Vigorous Physical
Activity.
Visual demonstration of the significant relationship between insulin sensitivity
index and time spent in MVPA. Adjusted for child age, sex, dietary added
sugar, daily energy intake, maternal pre-pregnancy BMI, GDM status, and child
BMI z-score. *Denotes p-value<0.05.
29
Table 26. Summary of unadjusted and adjusted linear regression models, with child
BMI z score as the outcome variable.
*Denotes p-value<0.05.
a
Standardized regression coefficient. MVPA=moderate to vigorous physical
activity. SB=sedentary behavior. EI=daily energy intake. GDM=gestational diabetes mellitus.
Predictor
Variables
Beta (95%CI) p-value Covariates
Time in MVPA
-0.04
a
(-0.24, 0.16) 0.86 Unadjusted
-0.03
a
(-0.23, 0.17) 0.68 Child age, sex, EI, added sugar
-0.05
a
(-0.24, 0.14) 0.82 +Pre-pregnancy BMI, GDM exposure
Time in SB
-0.04
a
(-0.24, 0.16) 0.69 Unadjusted
-0.03
a
(-0.23, 0.17) 0.75 Child age, sex, EI, added sugar
-0.05
a
(-0.24, 0.14) 0.61 +Pre-pregnancy BMI, GDM exposure
Daily energy
intake
0.07
a
(-0.12, 0.27) 0.46 Unadjusted
0.09
a
(-0.11, 0.3) 0.38 Child age, sex, added sugar, MVPA
0.05
a
(-0.15, 0.25) 0.64 pre-pregnancy BMI, GDM exposure
Dietary added
sugar
-0.11
a
(-0.3, 0.09) 0.28 Unadjusted
-0.11
a
(-0.32, 0.09) 0.28 Child age, sex, EI, GDM exposure,
MVPA
-0.14
a
(-0.34, 0.05) 0.14 +Pre-pregnancy BMI, GDM exposure
Maternal pre-
pregnancy BMI
0.32
a
(0.13, 0.5) 0.001* Unadjusted
0.31
a
(0.12, 0.5) 0.002* Child age, sex, EI, GDM exposure,
added sugar
0.31
a
(0.12, 0.5) 0.002* +MVPA
GDM exposure
0.32 (-0.08, 0.71) 0.12 Unadjusted
0.25 (-0.15, 0.65) 0.22 Child age, sex, EI, pre-pregnancy BMI,
added sugar
0.25 (-0.15, 0.66) 0.22 +MVPA
30
Table 27. Summary of unadjusted and adjusted linear regression models, with child
total percent body fat as the outcome variable.
Predictor
Variables
Beta (95%CI) p-
value
Covariates
Time in
MVPA
-0.07
a
(-0.27, 0.14) 0.53 Unadjusted
-0.09
a
(-0.3, 0.13) 0.43 Child age, sex, EI, added sugar
-0.07
a
(-0.27, 0.13) 0.51 +Pre-pregnancy BMI, GDM exposure
Time in SB
0.04
a
(-0.17, 0.25) 0.72 Unadjusted
0.03
a
(-0.18, 0.25) 0.76 Child age, sex, EI, added sugar
0.01
a
(-0.19, 0.22) 0.89 +Pre-pregnancy BMI, GDM exposure
Daily energy
intake
0.02
a
(-0.19, 0.22) 0.88 Unadjusted
0.07
a
(-0.15, 0.28) 0.55 Child age, sex, added sugar, MVPA
0.01
a
(-0.2, 0.22) 0.93 pre-pregnancy BMI, GDM exposure
Dietary
added sugar
0.001
a
(-0.207,
0.21)
0.99 Unadjusted
-0.03
a
(-0.24, 0.19) 0.82 Child age, sex, EI, GDM exposure, MVPA
-0.06
a
(-0.26, 0.15) 0.58 +Pre-pregnancy BMI, GDM exposure
Maternal pre-
pregnancy
BMI
0.29
a
(0.09, 0.49) 0.006* Unadjusted
0.28
a
(0.07, 0.48) 0.009* Child age, sex, EI, GDM exposure, added
sugar
0.27
a
(0.07, 0.48) 0.011* +MVPA
GDM
exposure
0.43 (0.02, 0.84) 0.04* Unadjusted
0.35 (-0.08, 0.78) 0.11 Child age, sex, EI, pre-pregnancy BMI, added
sugar
0.36 (-0.07, 0.79) 0.11 +MVPA
*Denotes p-value<0.05.
a
Standardized regression coefficient. MVPA=moderate to vigorous physical
activity. SB=sedentary behavior. EI=daily energy intake. GDM=gestational diabetes mellitus.
31
Table 28. Summary of unadjusted and adjusted linear regression models, with child
waist to height ratio as the outcome variable.
*Denotes a significant odds ratio.
a
Standardized regression coefficient. MVPA=moderate to vigorous
physical activity. SB=sedentary behavior. EI=daily energy intake. GDM=gestational diabetes mellitus.
Predictor
Variables
Beta (95%CI) p-value Covariates
Time in MVPA
-0.07
a
(-0.28, 0.14) 0.50 Unadjusted
-0.11
a
(-0.32, 0.1) 0.31 Child age, sex, EI, added sugar
-0.09
a
(-0.27, 0.1) 0.38 +Pre-pregnancy BMI, GDM exposure
Time in SB
-0.05
a
(-0.26, 0.16) 0.66 Unadjusted
-0.04
a
(-0.25, 0.17) 0.70 Child age, sex, EI, added sugar
-0.07
a
(-0.26, 0.12) 0.48 +Pre-pregnancy BMI, GDM exposure
Daily energy
intake
0.1
a
(-0.11, 0.31) 0.34 Unadjusted
0.16
a
(-0.06, 0.37) 0.16 Child age, sex, added sugar, MVPA
0.08
a
(-0.12, 0.28) 0.44 pre-pregnancy BMI, GDM exposure
Dietary added
sugar
-0.09
a
(-0.3, 0.12) 0.41 Unadjusted
-0.1
a
(-0.32, 0.11) 0.34 Child age, sex, EI, GDM exposure,
MVPA
-0.15
a
(-0.34, 0.04) 0.13 +Pre-pregnancy BMI, GDM exposure
Maternal pre-
pregnancy BMI
0.43
a
(0.24, 0.62) <0.001 Unadjusted
0.41
a
(0.23, 0.6) <0.001 Child age, sex, EI, GDM exposure,
added sugar
0.41
a
(0.22, 0.59) <0.001 +MVPA
GDM exposure
0.58 (0.17, 0.98) 0.007 Unadjusted
0.46 (0.06, 0.86) 0.026 Child age, sex, EI, pre-pregnancy BMI,
added sugar
0.46 (0.06, 0.86) 0.026 +MVPA
32
Table 29. Summary of unadjusted and adjusted logistic regression models, with odds of
being overweight/obese as the outcome variable.
*Denotes a significant odds ratio. MVPA=moderate to vigorous physical activity. SB=sedentary behavior.
EI=daily energy intake. GDM=gestational diabetes mellitus.
Predictor
Variables
Odds Ratio Wald 95% CI Covariates
Time in MVPA
1.21 0.78~1.87 Unadjusted
1.17 0.75~1.85 Child age, sex, EI, added sugar
1.24 0.76~2.00 +Pre-pregnancy BMI, GDM exposure
Time in SB
0.76 0.49~1.18 Unadjusted
0.77 0.50~1.21 Child age, sex, EI, added sugar
0.71 0.44~1.15 +Pre-pregnancy BMI, GDM exposure
Daily energy
intake
1.25 0.81~1.94 Unadjusted
1.19 0.76~1.88 Child age, sex, added sugar, MVPA
1.07 0.66~1.74 pre-pregnancy BMI, GDM exposure
Dietary added
sugar
0.84 0.54~1.31 Unadjusted
0.86 0.55~1.37 Child age, sex, EI, GDM exposure,
MVPA
0.78 0.48~1.28 +Pre-pregnancy BMI, GDM exposure
Maternal pre-
pregnancy BMI
1.07 1.01~1.14* Unadjusted
1.07 1.00~1.15* Child age, sex, EI, GDM exposure,
added sugar,
1.08 1.01~1.15* +MVPA
GDM exposure
2.14 0.85~5.40 Unadjusted
2.20 0.79~6.10 Child age, sex, EI, pre-pregnancy BMI,
added sugar
2.23 0.79~6.23 +MVPA
33
Table 30. Summary of unadjusted and adjusted linear regression models, with maternal
pre-pregnancy BMI as the predictor variable.
*Denotes p-value<0.05. †Denotes p-value<0.10. MVPA=moderate to vigorous physical activity.
SB=sedentary behavior. GDM=gestational diabetes mellitus. ISI=Matsuda insulin sensitivity index.
‡
Standardized regression coefficient.
Predictor
Variable
Daily energy intake
Beta (95% CI) p-
value
Covariates
Pre-pregnancy
BMI
0.03
a
(-0.18,
0.24)
0.80
Unadjusted
0.03
a
(-0.18,
0.24)
0.79
GDM exposure, child age, sex, pre-pregnancy
BMI, added sugar, MVPA, ISI
0.008
a
(-0.209,
0.225)
0.94
+BMI z-score
Dietary added sugar
Beta (95% CI) p-
value
Covariates
Pre-pregnancy
BMI
0.06
a
(-0.15,
0.27)
0.59
Unadjusted
0.07
a
(-0.14,
0.29)
0.50
GDM exposure, child age, sex, pre-pregnancy
BMI, EI, MVPA, ISI
0.11
a
(-0.11,
0.34)
0.33
+BMI z-score
Time in MVPA
Beta (95% CI) p-
value
Covariates
Pre-pregnancy
BMI
-0.06
a
(-0.27,
0.14)
0.55
Unadjusted
-0.01
a
(-0.22,
0.20)
0.93
GDM exposure, child age, sex, pre-pregnancy
BMI, EI, added sugar, ISI
-0.05
a
(-0.26,
0.17)
0.67
+BMI z-score
Time in SB
Beta (95% CI) p-
value
Covariates
Pre-pregnancy
BMI
0.02
a
(-0.19,
0.23)
0.84
Unadjusted
-0.01
a
(-0.23,
0.21)
0.90
GDM exposure, child age, sex, pre-pregnancy
BMI, EI, added sugar, ISI
0.02
a
(-0.20,
0.25)
0.84
+BMI z-score
34
Table 31. Summary of unadjusted and adjusted linear regression models, with GDM
exposure as the predictor variable.
*Denotes p-value<0.05. †Denotes p-value<0.10. MVPA=moderate to vigorous physical activity.
SB=sedentary behavior. GDM=gestational diabetes mellitus. ISI=Matsuda insulin sensitivity index.
Predictor
Variable
Daily energy intake
Beta (95% CI) p-value Covariates
GDM exposure
0.39 (-0.03,
0.80) 0.070
Unadjusted
0.48 (0.06,
0.90) 0.029*
Pre-pregnancy BMI, child age, sex, pre-
pregnancy BMI, added sugar, MVPA, ISI
0.45 (0.02,
0.88) 0.044*
+BMI z-score
Dietary added sugar
Beta (95% CI) p-value Covariates
GDM exposure
0.19 (-0.23,
0.61) 0.37
Unadjusted
0.16 (-0.29,
0.61) 0.50
Pre-pregnancy BMI, child age, sex, pre-
pregnancy BMI, EI, MVPA, ISI
0.21 (-0.25,
0.66) 0.38
+BMI z-score
Time in MVPA
Beta (95% CI) p-value Covariates
GDM exposure
0.12 (-0.3, 0.53) 0.59 Unadjusted
-0.003 (-0.433,
0.426) 0.99
Pre-pregnancy BMI, child age, sex, pre-
pregnancy BMI, EI, added sugar, ISI
-0.04 (-0.47,
0.4) 0.87
+BMI z-score
Time in SB
Beta (95% CI) p-value Covariates
GDM exposure
0.19 (-0.23,
0.61) 0.37
Unadjusted
0.224 (-0.23,
0.677) 0.34
Pre-pregnancy BMI, child age, sex, pre-
pregnancy BMI, EI, added sugar, ISI
0.27 (-0.19,
0.72) 0.25
+BMI z-score
35
Table 32. Participant and Mother’s Demographics (N=100).
Variable N(%) or Mean (SD) Range
Age (years) 8.51 (1.00) 7.33~11.34
Sex Girls: 59 (59%)
Boys: 41 (41%)
BMI z-score 0.75 (1.09) -1.78~2.64
Child BMI Percentile Category Healthy-weight: 60 (60%)
Overweight: 24 (24%)
Obese: 24 (24%)
Tanner Stage Stage 1: 91 (91%)
Stage 2: 6 (6%)
Stage 3: 3 (3%)
Time in MVPA (min/day) 132.2 (96.05) 0~430
Time in VPA (min/day) 20.20 (34.76) 0~270
Median VPA Category <=10 minutes: 61 (61%)
>10 minutes: 39 (39%)
WASI IQ Scores 106.4 (14.0) 76~150
Maternal pre-pregnancy BMI (kg/m
2
) 30.08 (7.11) 18.97~50.38
Maternal pre-pregnancy Category Normal-weight: 24 (24%)
Overweight: 35 (35%)
Obese: 41 (41%)
Maternal Education <=High school: 21 (21%)
Some college: 32 (32%)
College and post: 47 (47%)
Family Income 0<=income <30 000: 15 (15%)
30000<=income <50 000: 30 (30%)
50000<=income <70 000: 30 (30%)
70000<=income <90 000: 14 (14%)
90000>=income: 11 (11%)
36
Table 33. The Relationship between Maternal Pre-Pregnancy BMI and Physical Activity
with Child IQ.
Predictor Variables Beta (95% CI) Partial
Omega
2
p-value Covariates
Maternal pre-
pregnancy BMI
-0.31 (-0.67, 0.06) 0.017 0.10 Unadjusted
-0.42 (-0.81, -0.03 0.036 0.037* age, sex, BMI z-score
-0.39 (-0.80, 0.02) 0.024
0.066
#
age, sex, BMI z-score,
SES, GDM status
-0.27 (-0.69, 0.15) 0.010
0.21 age, sex, BMI z-score,
SES, GDM status, VPA
Time in MVPA
2.67 (-0.98, 6.24) 0.010 0.16 Unadjusted
2.98 (-0.65, 6.60) 0.016 0.11 age, sex, BMI z-score
3.94 (0.32, 7.56) 0.035 0.036* age, sex, BMI z-score, SES
3.90 (0.33, 7.48) 0.035 0.035* age, sex, BMI z-score,
SES, pre-pregnancy BMI
Time in VPA
7.05 (2.49, 11.60) 0.077 0.003* Unadjusted
6.30 (1.57, 11.02) 0.068 0.011* age, sex, BMI z-score
5.79 (0.98, 10.60) 0.044 0.021* age, sex, BMI z-score, SES
5.03 (0.13, 9.94) 0.030 0.047* age, sex, BMI z-score,
SES, pre-pregnancy BMI
*Denotes p-value<0.05.
#
Denotes p-value<0.10. Age, child age. Tanner, Tanner stage of pubertal
development. SES, family income and maternal education. GDM status, maternal gestational diabetes
mellitus status. Time in VPA, time in vigorous physical activity.
37
Table 34. The Relationship between Maternal Pre-Pregnancy BMI and Physical Activity
with Child Global FA.
Predictor Variables Beta (95% CI) Partial
Omega
2
p-value Covariates
Maternal pre-
pregnancy BMI
-0.00002
(-0.0004, 0.0003)
0
0.92 Unadjusted
-.000002
(-0.0004, 0.0004)
0 0.99 age, sex, BMI z-score
-0.00004
(-0.0004, 0.0003)
0 0.83 age, sex, BMI z-score,
SES, GDM status
0.0001
(-0.0003, 0.0004)
0 0.79 age, sex, BMI z-score,
SES, GDM status, VPA
Time in MVPA
-0.001
(-0.005, 0.002)
0 0.42 Unadjusted
-0.001
(-0.005, 0.002)
0 0.44 age, sex, BMI z-score
-0.001
(-0.005, 0.002)
0 0.52 age, sex, BMI z-score, SES
-0.001
(-0.005, 0.002)
0 0.50 age, sex, BMI z-score,
SES, pre-pregnancy BMI
Time in VPA
0.005
(0.001, 0.009)
0.047 0.017* Unadjusted
0.006
(0.002, 0.01)
0.060 0.008* age, sex, BMI z-score
0.006
(0.002, 0.01)
0.069
0.005* age, sex, BMI z-score, SES
0.006
(0.002, 0.01)
0.068 0.005* age, sex, BMI z-score,
SES, pre-pregnancy BMI
*Denotes p-value<0.05. Age, child age. SES, family income and maternal education. GDM status,
maternal gestational diabetes mellitus status. Time in VPA, time in vigorous physical activity.
38
Table 35. Tracts with significant FA clusters for children above the median reported
time spent in VPA, before adjusting for maternal pre-pregnancy BMI, at a threshold of
p<0.05 and controlling for multiple comparisons.
Anatomic Tract (JHU White Matter
Tractography Atlas)
MNI Maxima
Coordinates (X, Y, Z)
Voxels p-value
Superior Longitudinal Fasciculus (R.) 29 -5 41 2372 0.031
31 -3 17 176 0.046
46 -11 26 111 0.047
32 9 26 5 0.050
Anterior Thalamic Radiation (R.) 6 -20 -18 961 0.043
22 13 21 233 0.045
Inferior fronto-occipital fasiculus (L.) -34 -10 -6 858 0.041
Superior Longitudinal Fasciculus (L.) -36 -4 22 737 0.036
Inferior Longitudinal Fasiculus (L.) -32 -2 -27 25 0.049
Table 36. Tracts with significant FA clusters for children above the median reported
time spent in VPA, after adjusting for maternal pre-pregnancy BMI, at a threshold of
p<0.05 and controlling for multiple comparisons.
Anatomic Tract (JHU White Matter
Tractography Atlas)
MNI Maxima
Coordinates (X, Y, Z)
Voxels p-value
Superior Longitudinal Fasciculus (R.) 30 -31 41 1763 0.030
Superior Longitudinal Fasciculus (L.) -36 -3 22 128 0.045
-23 -4 35 53 0.049
Anterior Thalamic Radiation (R.) 30 17 26 86 0.049
23 17 20 46 0.048
Table 37. Tracts with significant FA clusters for the interaction between maternal weight
status and vigorous physical activity levels, at a threshold of p<0.05 and controlling for
multiple comparisons.
Anatomic Tract (JHU White Matter
Tractography Atlas)
MNI Maxima
Coordinates (X, Y, Z)
Voxels p-value
Forceps Major (L.) -19 -52 21 25663 0.004
Inferior fronto-occiptial fasciculus (R.) 26 24 3 1453 0.031
39
Table 38. Tracts with significant FA clusters for the interaction between maternal weight
status and vigorous physical activity levels, at a threshold of p<0.01 and controlling for
multiple comparisons.
Anatomic Tract (JHU White Matter
Tractography Atlas)
MNI Maxima
Coordinates (X, Y, Z)
Voxels p-value
Forceps Major (L.) -19 -52 21 796 0.004
Anterior Thalamic Radiation (L.) -21 17 10 597 0.007
-28 -34 16 34 0.010
Cingulate Gyrus (R.) 18 -53 34 208 0.008
Inferior fronto-occiptial fasciculus (R.) 41 -20 -11 80 0.010
Cingulum (hippocampal) 19 -49 23 60 0.010
Figure 9. LSmean IQ scores of children above and below the median reported time
spent in VPA.
LSMeans adjusted for child age, sex, BMI z-score, family income, maternal education and maternal pre-pregnancy
BMI.
103.39
110.05
96
98
100
102
104
106
108
110
112
114
1 2
WASI IQ Scores
VPA≤10 minutes VPA>10 minutes
40
Figure 10. LSmean global FA of children above and below the median reported time
spent in VPA.
LSMeans adjusted for child age, sex, BMI z-score, family income, maternal education and maternal pre-pregnancy
BMI
0.416
0.423
0.408
0.410
0.412
0.414
0.416
0.418
0.420
0.422
0.424
0.426
VPA≤10 minutes VPA>10 minutes
Fractional Anisotropy (White Matter)
41
Figure 11. Axial view of TBSS results for children above the median reported time
spent in VPA compared to children below the median.
L. SLF
R.SLF
R.
ATR
In red/orange, significant FA clusters overlaid the mean FA skeleton (green) and a T1-weighted
image, after adjusting for maternal pre-pregnancy BMI, at a threshold of p<0.05, and controlling for
multiple comparisons. R. SLF, right superior longitudinal fasciculus. L. SLF, left superior longitudinal
fasciculus. R. ATR, right anterior thalamic radiation.
Abstract (if available)
Abstract
This dissertation project had two study aims. The first aim of this dissertation project was to determine how prenatal exposure to gestational diabetes mellitus or maternal obesity impacts metabolic health and neurocognition during childhood. The second aim of this project was to determine how modifiable lifestyle factors such as physical activity levels contribute to metabolic health and neurocognition during childhood. Electronic medical records were used to determine mother’s pre-pregnancy BMI (kg/m²) and gestational diabetes mellitus status. Children came in for two study visits. During the first study visit, they completed an oral glucose tolerance test, their anthropometric measurements were captured, diet and physical activity levels were assessed, and cognitive assessments were completed. From the oral glucose tolerance test, Matsuda insulin sensitivity index was calculated to assess insulin sensitivity. The 3-day physical activity recall was used to assess sedentary time, moderate to vigorous physical activity (MVPA) and vigorous physical activity (VPA), using metabolic equivalents (METS). METS>1.1 and <1.5 determined sedentary time. METS>3 determined moderate to vigorous physical activity and METS>6 determined vigorous physical activity. The Weschler abbreviated scale of intelligence, 2nd edition was used to assess intelligence quotient (IQ). The NIH toolbox, picture-sequence memory task was used to assess episodic memory and a visuospatial memory task was used to assess hippocampal-dependent, relational memory. During the second study visit, participants completed magnetic resonance imaging (MRI) and a dietary recall. Functional and structural scans were collected. For this dissertation, the T1-structural images and diffusion weighted images (DWI) were used to analyze hippocampal grey matter volume and global white matter. FreeSurfer was used to calculate hippocampal grey matter volume. FMRIB software library (FSL) was used to calculate global fractional anisotropy (FA). Tract-based spatial statistics was used to compare significant clusters of FA across the brain between children who engaged in above or below the median time spent in VPA. Compared to prenatal exposures and dietary components, MVPA was the only predictor of child insulin sensitivity suggesting that engaging in moderate to vigorous physical activity during childhood is beneficial for insulin sensitivity and may ameliorate future risk for metabolic disease. Additionally, we found that there was a significant interaction between boys and girls in the relationship between maternal pre-pregnancy BMI and hippocampal grey matter volume. Compared to girls, boys exposed to higher levels of maternal pre-pregnancy BMI, had significantly reduced hippocampal grey matter volume. Interestingly, both boys and girls exposed to higher levels of maternal pre-pregnancy BMI had reduced IQ scores. In contrast, higher IQ scores and greater global FA was observed in participants who engaged in VPA for at least 10 minutes a day, and this was independent of maternal obesity. Additionally, there was a significant interaction between VPA and maternal obesity exposure on both child IQ and global FA. VPA was also associated with greater FA in the superior longitudinal fasciculus, and anterior thalamic radiation (ATR). Further, there was a significant interaction between maternal pre-pregnancy BMI category and VPA on FA in left forceps major, left ATR, right cingulate gyrus, right inferior frontal-occipital fasciculus (IFOF), and cingulum (hippocampal portion). These findings suggest that engaging in VPA may be particularly beneficial for children exposed to maternal obesity in utero. Collectively, two important findings were a result of this dissertation. It is pivotal to consider sex differences in the effects of in utero exposure to maternal obesity and secondly, engaging in physical activity during childhood is important for metabolic health and neurocognition and may be protective for children exposed to maternal obesity in utero. Future longitudinal and intervention studies are needed to confirm these findings.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Native American ancestry among Hispanic Whites is associated with higher risk of childhood obesity: a longitudinal analysis of Children’s Health Study data
PDF
Associations of cumulative pollution burden and environmental health vulnerabilities with gestational weight gain in a cohort of predominantly low-income Hispanic women
PDF
Prenatal air pollution exposure, newborn DNA methylation, and childhood respiratory health
Asset Metadata
Creator
Alves, Jasmin Marie
(author)
Core Title
Prenatal and lifestyle predictors of metabolic health and neurocognition during childhood
School
Keck School of Medicine
Degree
Doctor of Philosophy
Degree Program
Medical Biology
Publication Date
07/12/2020
Defense Date
04/01/2020
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
Child development,developmental programming,hippocampal structure,metabolic health,OAI-PMH Harvest,physical activity,white matter
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Watanabe, Richard (
committee chair
), Herting, Megan (
committee member
), Monterosso, John (
committee member
), Page, Kathleen (
committee member
), Xiang, Anny (
committee member
)
Creator Email
jalves@usc.edu,jazalves13@gmail.com
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c89-328048
Unique identifier
UC11663453
Identifier
etd-AlvesJasmi-8658.pdf (filename),usctheses-c89-328048 (legacy record id)
Legacy Identifier
etd-AlvesJasmi-8658.pdf
Dmrecord
328048
Document Type
Dissertation
Rights
Alves, Jasmin Marie
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
developmental programming
hippocampal structure
metabolic health
physical activity
white matter