University of Southern California Dissertations and Theses

Prenatal and brain factors shape appetite regulation and weight from childhood onwards

(USC Thesis Other)

Prenatal and brain factors shape appetite regulation and weight from childhood onwards

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Prenatal and Brain Factors Shape Appetite Regulation and
Weight from Childhood Onwards
by
Sandhya Prathap Chakravartti
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(NEUROSCIENCE)
August 2024
Copyright © 2024 Sandhya Prathap

ii
Epigraph
Somewhere, something incredible is waiting to be known.
- Carl Sagan

iii
Dedication
I dedicate this thesis to my family. Thank you all for being my wings. I would not have
been able to fly this high without your unconditional love and support.

iv
Acknowledgements
There are a number of people to whom I owe a great deal of thanks and gratitude. First, I would
like to thank my research mentors at the University of Illinois at Urbana-Champaign, Stanford
University, and The University of Southern California: Dr. Neal Cohen, Dr. Monica Fabiani, Dr.
Vinod Menon, Dr. Jack Van Horn, Dr. Meredith Braskie, and Dr. Yonggang Shi. Thank you all for
the incredible opportunities you have given me, and for your continued support.
I would also like to thank my graduate and post-doctoral mentors: Dr. John Walker, Dr. Shaozheng
Qin, and Dr. Aarthi Padmanabhan, Dr. Teresa Iuculano, and Dr. Tanya Evans. Many thanks for
your guidance, and for helping me to build my foundational knowledge in research.
Next, I would like to thank the Neuroscience Graduate Program and my Dissertation Committee
members: Dr. Mara Mather, Dr. Kay Jann, Dr. John Monterosso, and Dr. Anny Xiang. It has been
a dream come true to learn from all of you. Thank you for your support, invaluable feedback, and
enthusiasm for my project.
Next, I owe a many thanks to the members of the Branch lab, past and present: Ana, Brendan,
Patrick, Alex, Jasmin, Hilary, Alex Y., Shan, all of the undergraduate students, and members of
the DORI team. This dissertation would not exist without your hard work and efforts in collecting
this incredible dataset. I have learned so much from all of you.
Finally, I would like to express my deepest gratitude to my mentor, Dr. Katie Page. I feel incredibly
lucky and honored to have you as my mentor. I deeply cherish and will miss our scientific
discussions, which always sparked my curiosity and made me think about science in new ways.
Your guidance has been instrumental to my development as a scientist, and I am very grateful for
your unwavering support and positivity. Thank you for being an incredible role model for me and
others to follow.
I would also like to thank my friends and family for your support, love, and encouragement. Dora,
I couldn’t have made it this far without your support, thank you for being an amazing friend.
I dedicate this dissertation to my family. Thank you, Badri uncle, for always believing in me and
encouraging me to reach for the stars. I would not have come this far without your inspiration and
support. To my Father and Grandparents who are no longer with me, you are my motivation and
inspiration to become a scientist. To my Husband and Mother, I am forever grateful for all the
sacrifices you have made so that I could pursue my dream. Thank you both for being my pillars
of strength, and the wind beneath my wings. To my beautiful Daughter, you are the light of my life,
I am so blessed to be your Mom.
This research was supported by the Ruth L. Kirschstein Predoctoral National Research Service
Award (F31DK137584), the National Institute of Diabetes and Digestive and Kidney Diseases, &
the American Diabetes Association Pathway Award (R01DK116858, R01DK134079,
R01DK102794), & the American Diabetes Association Pathway Award (1-14-ACE 36).

v
Table of Contents
Epigraph ....................................................................................................................................................... ii
Dedication.................................................................................................................................................... iii
Acknowledgements ..................................................................................................................................... iv
List of Tables............................................................................................................................................... vii
List of Figures ............................................................................................................................................ viii
List of Abbreviations .................................................................................................................................... ix
Abstract ........................................................................................................................................................x
Chapter 1: Introduction.................................................................................................................................1
Dissertation Aims and Experiments .........................................................................................................2
Hypothalamic and Neuroendocrine Mechanisms of Appetite Regulation and Glucose Metabolism........3
The Hypothalamus and Homeostasis..................................................................................................3
The Anatomy of the Hypothalamus......................................................................................................4
Functional Subdivisions of the Hypothalamus Related to Feeding Behavior.......................................4
Functional Neuroimaging Tools to Study the Hypothalamic Response to Glucose: a Biomarker of
Satiety and Weight Change......................................................................................................................6
Insights from Lesion Studies in Animal Models....................................................................................6
Functional Neuroimaging Techniques..................................................................................................7
Insights from Rodent Neuroimaging Studies .......................................................................................8
History of Human Functional Imaging Studies of the Brain’s Response to Glucose ...........................9
Prenatal and Childhood Metabolic Programming of Obesity and Type 2 Diabetes – Insights from
Multimodal Neuroimaging Studies .........................................................................................................10
Classifying Childhood Obesity ...........................................................................................................10
Gestational Diabetes Mellitus Pathophysiology.................................................................................11
Understanding Gestational Diabetes Mellitus and links to Adiposity of the Offspring........................12
Maternal Diabetes associated with Altered Fetal Brain Programming- Evidence from Animal
Studies...............................................................................................................................................13
Altered Neural Programming Links Gestational Diabetes Mellitus Exposure, and Adiposity in
Offspring: Evidence from Multimodal Human Imaging Studies..........................................................13
Chapter 2: Non-Caloric Sweetener Effects on Brain Appetite Regulation in Individuals Across Varying
Body Weights .............................................................................................................................................15
Introduction ............................................................................................................................................16
Methods .................................................................................................................................................17
Results ...................................................................................................................................................23
Discussion..............................................................................................................................................26
Chapter 3: Differential Effects of Timing and Weight Status on Hypothalamic Response to Sucrose........40
Introduction ............................................................................................................................................41

vi
Methods .................................................................................................................................................42
Results ...................................................................................................................................................45
Discussion..............................................................................................................................................46
Chapter 4: Impact of Gestational Diabetes Exposure on Developing Brain and Adiposity Trajectories:
A 6 Year Longitudinal Study........................................................................................................................51
Introduction ............................................................................................................................................52
Methods .................................................................................................................................................53
Results ...................................................................................................................................................58
Discussion..............................................................................................................................................61
General Discussion ....................................................................................................................................74
References .................................................................................................................................................80
Appendix.....................................................................................................................................................95

vii
List of Tables
TABLE 1. [CHAPTER 2] PARTICIPANT CHARACTERISTICS..................................................................30
TABLE 2, SUPPLEMENTAL TABLE 1. HYPOTHALAMIC RESPONSE TO SUCRALOSE COMPARED
TO SUCROSE AND WATER .............................................................................................................35
TABLE 3, SUPPLEMENTAL TABLE 2. DRINK COMPARISONS OF HYPOTHALAMIC RESPONSE
STRATIFIED BY WEIGHT STATUS ..................................................................................................35
TABLE 4, SUPPLEMENTAL TABLE 3. DIFFERENTIAL EFFECTS OF DRINKS ON PERIPHERAL
GLUCOSE LEVELS...........................................................................................................................36
TABLE 5, SUPPLEMENTAL TABLE 4. DIFFERENTIAL EFFECTS OF DRINKS ON CHANGES IN
HUNGER ...........................................................................................................................................36
TABLE 6, SUPPLEMENTAL TABLE 5. ASSOCIATIONS BETWEEN PERIPHERAL GLUCOSE
LEVELS AND HYPOTHALAMIC RESPONSE TO SUCROSE CONSUMPTION..............................36
TABLE 7, SUPPLEMENTAL TABLE 6. ASSOCIATIONS BETWEEN PERIPHERAL GLUCOSE
AND MEDIAL HYPOTHALAMIC BLOOD FLOW RESPONSE TO SUCROSE, STRATIFIED BY
WEIGHT STATUS..............................................................................................................................36
TABLE 8, SUPPLEMENTAL TABLE 7. ASSOCIATIONS BETWEEN HYPOTHALAMIC RESPONSE
AND HUNGER...................................................................................................................................37
TABLE 9, SUPPLEMENTAL TABLE 8. DRINKS COMPARISONS BY TIME FOR LATERAL
HYPOTHALAMIC RESPONSE, PERIPHERAL GLUCOSE LEVELS, AND HUNGER RATINGS.....39
TABLE 10, [CHAPTER 3] TABLE 1. PARTICIPANT CHARACTERISTICS................................................48
TABLE 11, TABLE 2. CHANGE HYPOTHALAMIC BLOOD FLOW AT 10 MINUTES STRATIFIED BY
BMI STATUS......................................................................................................................................49
TABLE 12, TABLE 3. CHANGE HYPOTHALAMIC BLOOD FLOW AT 35 MINUTES STRATIFIED BY
BMI STATUS......................................................................................................................................49
TABLE 13, TABLE 4. PRE-DRINK HYPOTHALAMIC BLOOD FLOW STRATIFIED BY BMI STATUS......49
TABLE 14, TABLE 5. MEAN VALUES OF HYPOTHALAMIC AND WHOLE BRAIN CEREBRAL
BLOOD FLOW AT BASELINE STRATIFIED BY WEIGHT STATUS..................................................49
TABLE 15, [CHAPTER 4] TABLE 1. PARTICIPANT CHARACTERISTICS AT BASELINE ........................63
TABLE 16, TABLE 2. PARTICIPANT COUNTS BY STUDY VISIT .............................................................63
TABLE 17, TABLE 3. ASSOCIATIONS BETWEEN GDM EXPOSURE AND ADIPOSITY
MEASUREMENTS AT BASELINE.....................................................................................................63
TABLE 18, TABLE 4. ASSOCIATIONS BETWEEN GDM EXPOSURE AND BRAIN VOLUME
MEASUREMENTS AT BASELINE.....................................................................................................64
TABLE 19, TABLE 5. ASSOCIATIONS BETWEEN ADIPOSITY AND BRAIN VOLUME
MEASUREMENTS AT BASELINE.....................................................................................................64
TABLE 20, TABLE 6. SUMMARY OF GAMM MODELS: IMPACT OF GDM ON ADIPOSITY
TRAJECTORIES................................................................................................................................65
TABLE 21, TABLE 7. PREDICTED GEOMETRIC MEANS MEASURING EFFECT OF AGE ON
OUTCOMES USING GAMM MODELS .............................................................................................66
TABLE 22, TABLE 8. SUMMARY OF GAMM MODELS: IMPACT OF GDM ON GRAY MATTER
TRAJECTORIES................................................................................................................................67
TABLE 23, TABLE 9. PREDICTED GEOMETRIC MEAN OF AGE ON OUTCOMES IN GDM
EXPOSED CHILDREN USING GAMM MODELS .............................................................................68
TABLE 24, TABLE 10. PREDICTED GEOMETRIC MEAN OF AGE ON OUTCOMES IN
UNEXPOSED CHILDREN USING GAMM MODELS ........................................................................68
TABLE 25, TABLE 11. ASSOCIATIONS BETWEEN ADIPOSITY MEASUREMENTS AND BRAIN
VOLUME MEASUREMENTS ACROSS TIME...................................................................................68
TABLE 26, APPENDIX TABLE 1: ANATOMICAL INPUTS, OUTPUTS, AND FUNCTIONS OF THE
MEDIAL AND LATERAL HYPOTHALAMUS......................................................................................95

viii
List of Figures
FIGURE 1: [CHAPTER 2] PARTICIPANT ENROLLMENT FLOWCHART FOR THE RANDOMIZED
CROSSOVER BRAIN RESPONSE TO SUGAR II TRIAL AND FINAL ANALYSIS............................30
FIGURE 2: SCHEMATIC OF STUDY DESIGN ..........................................................................................31
FIGURE 3: DIFFERENTIAL HYPOTHALAMIC RESPONSE TO DRINK COMPARISONS .......................32
FIGURE 4: DIFFERENCE IN HYPOTHALAMIC RESPONSE TO DRINKS BY WEIGHT STATUS...........32
FIGURE 5: DIFFERENTIAL FUNCTIONAL CONNECTIVITY FROM HYPOTHALAMUS SEED
REGION AFTER SUCRALOSE INGESTION RELATIVE TO SUCROSE AND WATER ...................33
FIGURE 6: ASSOCIATIONS BETWEEN PERIPHERAL GLUCOSE, HUNGER, AND CHANGES IN
MEDIAL HYPOTHALAMIC BLOOD FLOW .......................................................................................34
FIGURE 7, SUPPLEMENTAL FIGURE 1: VISUAL DISPLAY OF HYPOTHALAMIC ROI.........................37
FIGURE 8, SUPPLEMENTAL FIGURE 2: VISUAL DISPLAY OF CHANGES IN HYPOTHALAMIC,
PERIPHERAL GLUCOSE, AND HUNGER RESPONSES OVER TIME............................................38
FIGURE 9, [CHAPTER 3] FIGURE 1: CHANGE IN HYPOTHALAMIC RESPONSE TO SUCROSE
BY BMI GROUP.................................................................................................................................50
FIGURE 10, [CHAPTER 4] FIGURE 1: ADIPOSITY TRAJECTORIES BY AGE.......................................70
FIGURE 11, FIGURE 2: BRAIN TRAJECTORIES BY AGE ......................................................................71
FIGURE 12, FIGURE 3: ADIPOSITY TRAJECTORIES STRATIFIED BY GDM EXPOSURE ..................72
FIGURE 13, FIGURE 4: TOTAL SUBCORTICAL BRAIN VOLUME TRAJECTORIES STRATIFIED
BY GDM EXPOSURE........................................................................................................................73
FIGURE 14, APPENDIX FIGURE 1: VISUAL DISPLAY OF HYPOTHALAMIC NUCLEI ...........................95

ix
List of Abbreviations
BMI: Body Mass Index
BOLD: Blood Oxygen Level Dependent
CBF: Cerebral Blood Flow
EMR: Electronic Medical Records System
ICC: Intraclass Correlation
ICV: Intracranial Volume
GDM: Gestational Diabetes Mellitus
LH: Lateral Hypothalamus
MRI: Magnetic Resonance Imaging
MH: Medial Hypothalamus
pASL: Pulsed Arterial Spin Labeling
sMRI: Structural Magnetic Resonance Imaging
ROI: Region of Interest

x
Abstract
This dissertation aims to characterize the neurohumoral and temporal mechanisms of
hypothalamic signaling across various nutrients in adults of varying weight status, and evaluate
the link between in utero exposure to Gestational Diabetes Mellitus (GDM) and changes in
adiposity and brain volume from childhood to adolescence.
The first study examined the effects of sucralose, a non-caloric sweetener, compared to caloric
sugar (sucrose) and water on hypothalamic, glycemic, and hunger responses using imaging
techniques. Results showed that sucralose significantly altered hypothalamic responses and
connectivity to somatosensory processing regions, highlighting differences in how nonnutritive
sweeteners affect appetite regulation. The second study examined the temporal mechanism of
hypothalamic response to sucrose ingestion in adults of varying weight statuses (healthy weight,
overweight, and obesity) at 10- and 35-minutes post-ingestion. Sucrose ingestion elicited a
sustained reduction in hypothalamic blood flow in healthy-weight individuals, while individuals with
obesity and overweight individuals displayed an attenuated reduction, suggesting that the
hypothalamic response to sucrose is dynamically modulated by weight status.
The third study investigated how in utero exposure to GDM affects adiposity, and brain volume
trajectories in children aged 7 to 16. Results showed that GDM exposure predicted higher
adiposity trajectories, which continued to diverge throughout adolescence in GDM-exposed
children. This study is also the first to reveal reduced growth of subcortical brain volume
trajectories in GDM compared to control children, suggesting a link between GDM exposure, and
adverse metabolic and neural outcomes.

1
Chapter 1: Introduction
Over the past three decades, worldwide obesity rates have significantly increased, particularly in
industrialized countries and urban areas of lower-income nations. This rise is largely attributed to
the westernization of dietary and behavioral lifestyles1
, characterized by high-fat and high-sugar
diets2
, ultra-processed foods, and sedentary behaviors. Childhood obesity, affecting over 330
million children globally3
, is a disease of energy imbalance marked by excess adiposity that
impairs physical health. It is the leading risk factor for numerous diseases later in life, including
diabetes, cardiovascular disease, and Alzheimer's 4-7
. Given that childhood is a sensitive period
hallmarked by rapid growth and development of the peripheral body and brain8
, children are more
vulnerable to excessive weight gain compared to adults9,10. A growing body of evidence has linked
imbalances in feeding behavior, and increased sugar consumption with obesity risk. An enhanced
understanding of how glucose regulation mechanisms go awry is essential for devising strategies
to combat the obesity epidemic.
Appetite regulation is a critical aspect of human physiology and weight regulation, governed by a
complex interplay of prenatal, neural, hormonal, and environmental factors. Feeding behavior is
regulated by negative and positive feedback mechanisms associated with homeostasis, the
body’s ability to actively maintain a steady internal state despite external disruptions. The
exquisite mechanism of homeostasis is governed by the hypothalamus, a tiny brain structure
located at the conjunction between the lower brainstem and upper cortex. In comparison with
other regulatory behaviors such as thermogenesis, homeostatic mechanisms of feeding behavior
are more circuitous. This is primarily because 1) it is considerably easier to gain then it is to lose
weight and 2) once weight is lost, it is difficult to keep weight off. It is theorized that the ability to
gain weight quickly may be due to an evolutionary adaptation to defend against starvation. By this
same mechanism, the inability for the body to maintain weight changes long-term may be a result

2
of humans not being biologically equipped to defend from the upper limits of adiposity presented
in modern obesogenic environments replete with high fat, sugar, and calorie dense foods 11. In
such environments, energy balance via food intake is strongly influenced by goal directed seeking
of hedonic or pleasurable foods in the absence of caloric need. The influence of hedonic control
i.e., “liking” and “wanting” to eat to satisfy reward/punishment balance overrides homeostatic
signaling of eating to restore energy balance, thus creating an imbalance between hedonic versus
homeostatic control, a hallmark of obesity. The relationship is akin to a tango, where each
partner's movements are intricately coordinated to maintain a seamless and dynamic balance.
Behavioral and neural insights from human MRI studies over the decade have provided valuable
information about how reward or hedonic feeding centers and networks regulate appetite,
behavior, and weight status. While homeostatic pathways have been extensively mapped in the
brains of animal models, understanding how homeostatic regulatory mechanisms, particularly
those involving the hypothalamus, contribute to the intricate tango of energy balance remains
understudied in human imaging studies.
Dissertation Aims and Experiments
The primary objective of this dissertation is to characterize how the hypothalamic response to
sugar (a purported biomarker of satiety) regulates appetite and weight in young adults and
children and evaluate changes in adiposity and brain development over time in children exposed
to gestational diabetes in utero. The study described in Chapter 2 investigates how the
hypothalamus responds to various nutrients including sucralose, a non-nutritive sweetener,
sucrose, and water utilizing Arterial Spin Labeling Perfusion MR Imaging in a sample of young
adults. Follow up experiments examined whether the drink pairs (e.g., sucralose vs sucrose,
sucralose vs water, sucrose vs. water) elicited different peripheral glucose and hunger responses,
and whether weight status played a role in differential drink responses. An exploratory analysis
was run to investigate whether differences in drink pairs induced differences in whole brain

3
functional connectivity. Extending the findings from Chapter 2, Chapter 3 aims to further explore
the temporal mechanisms associated with hypothalamic response to sucrose in young adults of
varying weight status (obese, overweight, and healthy weight). Finally, to better understand the
epigenetic model of obesity and diabetes transmission Chapter 4 examines the influence of in
utero exposure to gestational diabetes, to predict adiposity and brain volume trajectories in a
group of 7–8-year-old children as they transition from childhood to adolescence.
The following introductory sub-chapters will outline key topics to provide a foundational
understanding of the concepts discussed in this dissertation. First, a summary of our current
understanding of the complex interplay between the hypothalamus and peripheral glucose
signaling mechanisms based on relevant anatomy and animal work will be discussed. This will be
followed by a review on the topic of recent advances in the field of functional neuroimaging which
enable the capture of the neurohumoral response to glucose ingestion within the hypothalamus.
Finally, the use of multimodal neuroimaging tools to study the impact of prenatal contributions
towards obesity development during childhood will be discussed.
Hypothalamic and Neuroendocrine Mechanisms of Appetite Regulation and Glucose
Metabolism
The Hypothalamus and Homeostasis
Hedonic feeding associated with weight gain involves an imbalance of homeostatic and nonhomeostatic mechanisms. The homeostatic weight regulatory system is located primarily in the
hypothalamus. It is a key brain region uniquely responsible for regulating food intake, energy
homeostasis, glucose metabolism, and body weight management via regulation of neurohumoral
responses (cite) via afferent periphery signals such as those from adipose tissue and efferent
signals from the sympathetic/parasympathetic nervous system. In this way the hypothalamus is
considered to be a central link between the nervous and endocrine system.

4
The Anatomy of the Hypothalamus
The hypothalamus is part of the diencephalon, and is located under the thalamus and directly
above the pituitary gland. In a sagittal section, it extends from the optic chiasm and anterior
commissure rostrally to the cerebral peduncle caudally. The third ventricle's cavity lies in the
midline. In a coronal section, the hypothalamus has four surfaces: lateral (adjacent to the
thalamus and internal capsule), medial (extending to the third ventricle wall), superior (separated
from the thalamus), and inferior (continuous with the third ventricle floor)12 (Appendix Figure 1).
Despite only occupying approximately 2% of brain volume, no other brain structure contains so
many specialized cell groups 13. These specialized cell groups, housed within histologically
distinct nuclei display exquisite control of the regulation of a variety of homeostatic behaviors
including appetite regulation and feeding.
Functional Subdivisions of the Hypothalamus Related to Feeding Behavior
The structure of the hypothalamus can be classified into 11 distinct nuclei. In 1939 Corby and
Woodburne first recognized that the structure can be divided into three zones: the paraventricular,
lateral, and medial zones14. The paraventricular zone nuclei, closest to the 3rd ventricle, receives
afferent inputs from many nuclei centers within the hypothalamus including the arcuate nucleus,
and generally acts as an interface between the endocrine system, autonomic and somatic
systems as it relates to feeding behavior and other homeostatic processes 15.
The medial (adjacent to the periventricular nuclei zone), and lateral zones regulate the autonomic
and somatic behaviors and are crucially involved in feeding behaviors and glucose regulation.
Seminal lesion studies in the 1940’s uncovered dramatically distinct functions between the medial
and lateral nuclei with respect to feeding behavior and weight 16-22. In these studies, the lesion of
the ventromedial hypothalamus nucleus or the “satiety center”, induced an abnormal increase in

5
appetite and food intake resulting in obesogenic state. This is in stark contrast to lesions to the
lateral hypothalamus, the “feeding center”, which result in to a reduction in hunger, food intake,
motivation for pleasurable stimulation, ultimately leading to anorexia. Stimulation of the lateral
hypothalamus resulted in increased feeding as well as reward and motivating behavior (Appendix
Table 1).
The ventromedial region of the hypothalamus located dorsally to the periventricular zone, also
contains the arcuate nucleus (ARH) which plays a critical role in maintaining energy homeostasis
by integrating rapid nutrient sensing with endocrine signals to control homeostasis. The ARH
contains two antagonistic cell groups: anorexigenic [i.e. rise before meals; declines following
meals] pro-opiomelanocortin (POMC) neurons (a precursor protein), which decrease food intake
and increase energy expenditure via activation of melanocortin-4 receptors (MC4R) (peptide
cleaved from POMC), and orexigenic [i.e. rise after meals] Agouti-related protein/ Neuropeptide
Y (AgRP/NPY) neurons, which increase appetite and feeding23,24. These neurons also respond
to hunger and satiety hormones, such as leptin, insulin, PYY, and ghrelin, as well as nutrients like
plasma fatty acids, to regulate feeding behavior and energy balance25.
Mechanisms of Glucose-Mediated Appetite Control in the Hypothalamus
Glucose, a primary energy substrate for the brain, plays a significant role in regulating appetite
through neuroendocrine mechanisms involving the ventromedial hypothalamus (VMH), lateral
hypothalamus (LH), and arcuate nucleus (ARC). Peripheral glucose homeostasis maintains
plasma glucose levels within a range of 70-110 mg per dL. Levels are monitored by tissues such
as the pancreas, liver, and gastrointestinal tract, which communicate with the hypothalamus to
regulate energy balance. Neurons within the hypothalamus express specific components
including glucose transporters (GLUT2, GLUT3), enzyme glucokinase (Hexokinase IV) which
enable the neurons to directly sense changes in blood glucose levels and, thereby modulating

6
their activity and appetite26. During a state of hunger, AgRP neurons inhibit the anorexigenic
effects of POMC neurons by blocking alpha-melanocyte-stimulating hormone (alpha-MSH
activation) of MC4R and release NPY to promote feeding27,28. By contrast, increased extracellular
glucose levels activate glucose-sensing neurons in the VMH and ARC while inhibiting feedingpromoting neurons in the LH26. Within the ARC, high glucose levels activate POMC neurons and
inhibit AgRP/NPY neurons, promoting satiety and reducing hunger25.
Dysregulation of Glucose-Mediated Appetite Control in Obesity: Insights from Animal Models
Recent evidence has shown that genetic mutations affecting the POMC gene or MC4R result in
obesity in both humans and rodent models. POMC is a precursor protein that is cleaved into
several peptides, including alpha-melanocyte-stimulating hormone (alpha-MSH), which binds to
MC4R to promote satiety. Mutations in the POMC gene can lead to a deficiency in alpha-MSH
production, resulting in decreased activation of MC4R and reduced satiety signaling. Similarly,
mutations in the MC4R gene can impair receptor function, leading to an inability to respond to
satiety signals effectively. These genetic alterations disrupt the normal balance of hunger and
satiety, leading to hyperphagia and weight gain29-31. Other factors that impair glucose mediated
appetite control within the hypothalamus include leptin resistance, hyperinsulinemia and insulin
resistance, chronic inflammation of the hypothalamus, and alteration in gut-brain communication
through gut hormones such as ghrelin, peptide YY (PYY), glucagon-like peptide-1 (GLP-1).
Functional Neuroimaging Tools to Study the Hypothalamic Response to Glucose: a
Biomarker of Satiety and Weight Change
Insights from Lesion Studies in Animal Models
Lesion studies on the medial and lateral nuclei of the hypothalamus in rodent models have
significantly advanced our understanding of hypothalamic circuits involved in feeding behavior. In
the 1950s, pioneering studies by Anand and Brobeck revealed distinct and opposing feeding

7
behaviors associated with these hypothalamic regions16,32. Subsequent research by Anand,
Mayer, and Oomura in the 1960s and 1970s showed that neurons in the ventromedial and lateral
hypothalamic regions respond to glucose, with increased extracellular glucose levels exciting
glucose-sensing neurons in the ventromedial hypothalamus and inhibiting those in the lateral
hypothalamus 33-35. In the late 1990s, Borg and Sherwin demonstrated that glucose levels in the
ventromedial hypothalamus modulate the activation of the sympathetic nervous system and
counterregulatory hormone responses during hypoglycemia in rat models36. Although the initial
view of the hypothalamus as merely feeding and satiety centers is now considered oversimplistic,
the foundational divisions proposed by Anand, Brobeck, Mayer, and Oomura still inform
contemporary research. These studies have not only shaped current rodent research on gut
peptides and other regulatory signals but also laid the groundwork for translational animal and
human imaging studies of feeding behavior and weight change.
Functional Neuroimaging Techniques
Two in vivo functional MR Imaging (fMRI) methods that are of particular interest to measure the
brain response to glucose metabolism in animal and humans are Blood-Oxygenation Level
Dependent (BOLD) and Arterial Spin Labeling Perfusion MR Imaging (ASL). Magnetic Resonance
Imaging (MRI) utilizes the behavior of hydrogen nuclei, which align with an external magnetic field
(B0) and precess in a low-energy state. When a subject is placed in a scanner a radiofrequency
pulse is applied at a specific flip angle to B0 which tips the proton spins, causing the now excited
nuclei to gain energy. When the radiofrequency pulse is turned off, the protons return to a state
of equilibrium by realigning with B0, and emit energy as radio waves that are measured and
converted into images. Gradient coils alter the magnetic field strength to provide spatial
information, while T1 and T2 relaxation times refer to the realignment and dephasing of spins,
respectively, with local inhomogeneities affecting the apparent T2 (T2*). The blood oxygen level
dependent (BOLD) contrast is then utilized to visualize brain activity with high spatial and temporal

8
resolution. Specifically paramagnetic properties of endogenous deoxygenated hemoglobin is
used as a source of contrast. Increased neuronal activity increases local blood flow by reducing
deoxyhemoglobin concentration and altering local magnetic fields. This creates a high image
intensity on T2* weighted images.
Pulsed Arterial Spin Labeling (pASL) is another type of functional MRI technique used to measure
absolute cerebral blood flow, which involves magnetically labeling arterial blood water as an
endogenous tracer, providing a non-invasive means to assess perfusion. In pASL, a short
radiofrequency pulse is applied to invert the magnetization of arterial blood water proximal to the
imaging plane. This labeled blood flows into the imaging slice, and the difference in magnetization
between labeled and non-labeled images is used to calculate absolute cerebral blood flow. This
technique is beneficial for examining slower metabolic responses triggered by nutrient ingestion,
and the temporal stability of ASL aligns with the prolonged glucose response curves compared to
fMRI techniques. While stimulus-dependent BOLD signal has a better signal to noise ratio, the
changes are expressed as a percent signal change and lack direct quantification in physiological
units. pASL, however, enables quantitative measurement of absolute cerebral blood flow
measures in tissue-specific units of mL/100g/min37.
Insights from Rodent Neuroimaging Studies
Functional neuroimaging techniques in rodent models have significantly advanced our
understanding by aligning findings from rodent lesion studies with human neuroimaging research.
In the 1990s, Yokokawa et al. used functional fMRI to study the neural response to L-lysine (Lys)
deficiency in rats, finding that Lys deficiency led to decreased plasma and brain levels, resulting
in anorexia. Lys supplementation normalized food intake and growth, with MRI scans showing
increased signal intensity in the medial and lateral hypothalamus, indicating these regions'
involvement in nutrient recognition and homeostasis 38. Extending these findings Min and

9
colleagues examined the hypothalamic response after glucose ingestion in lean, overweight, and
obese rats using MRI39. They found that the BOLD signal intensity decreased in rats within 19.5
to 25.5 minutes of glucose consumption, with a greater decrease in lean rats compared to
overweight ones. No change was observed in control animals ingesting water. Another recent
study by Mohr et al. studied the hypothalamic response to glucose in mice fed a high-fat, highsugar diet or a control diet for seven days. Mice on the high-fat, high-sugar diet showed a blunted
response in the arcuate, lateral, and ventromedial nuclei of the hypothalamus40. Taken together
these studies further provide evidence that the hypothalamic signal is dynamically modulated by
glucose ingestion, with attenuation in obese versus lean rats using animal MR imaging
techniques.
History of Human Functional Imaging Studies of the Brain’s Response to Glucose
Four influential studies from Matsuda et al, Liu et al, Smeets et al, and Page and colleagues
provided evidence that the hypothalamus differentially responds to the ingestion of a glucose load,
and information regarding the temporal nature of the response. In the first experiment, Matsuda
and colleagues demonstrated that oral glucose ingestion resulted in a significant and transient
reduction in hypothalamic BOLD fMRI signal in lean subjects41. After administering a 75 g oral
glucose load to ten obese and ten lean adults following a 12-hour fast, they observed a 4-8%
deactivation in the lower posterior hypothalamus beginning 4 minutes post-ingestion and lasting
approximately 10 minutes. A slower and smaller response was reported in obese individuals.
These findings suggest delayed activation of satiety centers following glucose consumption may
contribute to excessive food intake in obese individuals. Liu et al. studied 21 healthy adults given
a 75 g glucose load after a 12-hour fast, showing a reduction in hypothalamic activity up to 4%,
initially at one to two minutes and then again at seven to twelve minutes post-ingestion42.
Temporal clustering analysis, a technique which provides information of peak responses of neural
activity without prior knowledge of timing, revealed peak neural activity within these time frames.

10
Smeets and colleagues extended these findings, showing a dose-dependent decrease in
hypothalamic MRI signal in response to varying oral glucose loads, lasting up to 30 minutes43.
Follow up experiment from Smeets and colleagues explored whether the hypothalamic signaling
could be reproduced by sweet taste alone (aspartame) or non-sweet caloric content
(maltodextrin), finding only glucose not aspartame (sweet taste alone) or maltodextrin
(carbohydrate alone), triggered the decrease in hypothalamic fMRI signal44. These studies
suggest that neural signals are triggered prior to and during nutrient absorption, playing a pivotal
role in adaptive food intake responses. However, these studies had limitations, primarily due to
artefacts caused from fMRI imaging, small sample sizes, and lack of ethnic diversity.
Page and colleagues investigated hypothalamic responses to glucose using pASL techniques.
The first study examined the effect of small decrements in circulating glucose on hypothalamic
blood flow and CBF measurements45. They found an increase in hypothalamic CBF after a small
reduction in glucose from 93.1 to 77 mg/dL compared to euglycemic controls. This response
preceded the significant elevation of counterregulatory hormones during hypoglycemia. A followup study compared fructose and glucose consumption, finding that glucose reduced hypothalamic
CBF in lean adults, suggesting it as a central biomarker of satiety46. In contrast, fructose did not
decrease the hypothalamic response but induced a small increase in hypothalamic activity. These
studies were the first to use pASL techniques to provide further support of the hypothalamus as
a purported biomarker of satiety signaling.
Prenatal and Childhood Metabolic Programming of Obesity and Type 2 Diabetes – Insights
from Multimodal Neuroimaging Studies
Classifying Childhood Obesity
Classifying obesity in children is complex due to the significant effects of age, gender, pubertal
status, and race/ethnicity on growth. The CDC utilize the BMI percentile method, considering age

11
and sex differences, defining overweight as a BMI at or above the 85th percentile but below the
95th percentile, and obesity as a BMI at or above the 95th percentile8
. While BMI is widely used,
it does not differentiate between fat and muscle mass or indicate where fat is distributed
throughout the body, making additional measurements such as body fat percentage and waist
circumference important for assessing where central adiposity is located in the child’s body.
Gestational Diabetes Mellitus Pathophysiology
While the primary factors contributing to obesity in children are similar to those of adults, recent
research indicates that maternal and fetal factors during intrauterine development, such as
exposure to Gestational Diabetes mellitus (GDM) are also linked to an increased risk of
developing obesity and type II diabetes in adulthood. Intrauterine exposure to GDM is one of the
most common pregnancy disorders of affecting 2-10% percent of all pregnancies in the US
annually47. GDM is characterized by varying degrees of glucose intolerance first identified during
pregnancy, which leads to maternal hyperglycemia and consequently fetal hyperglycemia and
hyperinsulinemia.
Glucose intake and disposal primarily occur in the skeletal muscle and adipose tissue. Pregnancy
naturally induces a state of insulin resistance due to placental hormones that promote fetal
development but inhibit maternal insulin action, potentially requiring the mother to produce up to
three times the normal amount of insulin, and decrease in glucose disposal by 50% in order to
maintain a euglycemic state 48. When the pregnant mother cannot produce sufficient insulin to
maintain glucose levels, the mother develops, maternal hyperglycemia, which causes an
elevation of circulating nutrients for the mother. Although insulin does not cross the placenta,
glucose and other nutrients do, leading to an excess supply of glucose to the fetus. This excess
glucose stimulates the fetal pancreas to produce additional insulin, which acts as a growth
hormone, promoting fetal growth and adiposity 49.

12
Understanding Gestational Diabetes Mellitus and links to Adiposity of the Offspring
While several studies have reported that the exposure to gestational diabetes mellitus is
associated with risk of obesity during childhood and adolescence, a few studies attribute this
effect either to genetic or shared environment factors between the mother child pair50-53. A study
led by Dabelea and colleagues examined sibling pairs concordant for exposure to maternal GDM
in a Pima Indian population in order to isolate the effects of intrauterine exposure from hereditary
effects. Results revealed that offspring who had mothers with diabetes mellitus, have a higher risk
for developing obesity than offspring of fathers with diabetes mellitus 54. Furthermore, among
sibling pairs where one was exposed to maternal GDM and the other was not, the sibling born
after the mother developed diabetes mellitus had a higher BMI and risk for developing diabetes
mellitus compared to the sibling born before the mother developed gestational diabetes (OR: 3.0,
P < 0.01). Another study by Silverman and colleagues revealed that children of mothers with type
I diabetes mellitus, who are generally not obese also have a higher BMI by age 14 to 17 years
and have impaired glucose tolerance compared to that of children of non-diabetic mothers55.
Taken together, these studies provide support for the independent effect of in utero exposure to
GDM on adiposity outcomes in children.
Evidence from cross sectional studies have shown that exposure to GDM is associated with
greater adiposity outcomes including larger levels of abdominal fat in children and
adolescents53,56,57. A few studies have started to map the adiposity trajectories of children
exposed to GDM compared to trajectories of control children longitudinally. One study from
Silverman and colleagues showed that the effects of GDM on increased BMI begin to appear after
the age of two in the affected children55. Another study from Crume et al. mapped the adiposity
trajectories until the age of 14 in GDM exposed and unexposed children56. Higher BMI trajectories

13
were found in children, exposed to GDM, compared to children who are unexposed. Differences
were mainly attributed to a significantly higher growth rate velocity in the exposed children
between 10 and 13 years. These two studies provide preliminary evidence that BMI trajectories
are differently affected by gestational diabetes mellitus, and that growth rates in children exposed
to GDM may change during pubertal maturation.
Maternal Diabetes associated with Altered Fetal Brain Programming- Evidence from Animal
Studies
While studies have linked GDM, or maternal obesity exposure to greater adiposity in offspring,
the biological mechanism of transgenerational influence mediating this adverse metabolic
programming are not well understood. Serum levels of neurotrophins, such as nerve growth factor
and brain-derived neurotrophic factor, which are essential for neuronal growth, development, and
differentiation, are lower in pregnant women with diabetes and in infants exposed to GDM in
utero58,59. Early research by Plagemann and colleagues showed that rodent offspring of diabetic
mothers (induced by streptozotocin injections) were overweight and hyperinsulinemic, with
decreased mean areas of nuclei in the PVN and ventromedial hypothalamic nucleus60. Follow-up
experiments found that adult offspring had altered levels of hypothalamic NPY neurons, impaired
glucose tolerance, and increased weight gain. Other rodent studies have confirmed that maternal
diabetes impairs hypothalamic and subcortical neuron development 61,62.
Altered Neural Programming Links Gestational Diabetes Mellitus Exposure, and Adiposity in
Offspring: Evidence from Multimodal Human Imaging Studies
Functional imaging studies show altered hypothalamic responses to glucose in adults with obesity
and type II diabetes, with recent research extending these findings to children and adolescents.
Studies by Jastreboff et al. and Ge et al. found increased hypothalamic responses to glucose in
obese adolescents and overweight/obese children compared to their lean peers63,64. Page et al.

14
linked GDM exposure to increased hypothalamic glucose response in children, predicting greater
BMI increases, with GDM-exposed children also showing higher waist-to-height ratios65. A
handful of cross-sectional neuroimaging studies have linked in utero exposure to GDM with
alterations in brain structure and function, including alterations in: cortical volume66, and white
matter integrity in sensorimotor regions67. A recent study by Lou and colleagues shows that
prenatal exposure to GDM is significantly associated with reduced cortical gray matter volume in
a sample of 9 and 10 year old children66. These studies are also linking GDM exposure to
subcortical structures including the hippocampal volume and surface area68, and amygdala69.
Another recent study by Cai and colleagues demonstrated that variation in fetal fractional
anisotropic values, a metric representing microstructural tissue properties based on Diffusion
Weighted MRI, within the neonatal amygdala of children mediated the associated between
maternal antenatal glycemia and offspring adiposity in mothers with gestational diabetes69. These
studies suggest that homeostatic and hedonic brain regions may be impacting adiposity
trajectories in children born to mothers with gestational diabetes.

15
Chapter 2: Non-Caloric Sweetener Effects on Brain Appetite Regulation in
Individuals Across Varying Body Weights
Sandhya P. Chakravartti, MS1,2,3, Kay Jann, PhD4
, Ralf Veit, PhD5
, Hanyang Liu, MS2,3
,
Alexandra G. Yunker, MPH2,3
, Brendan Angelo, MS2,3
, John R. Monterosso1,6, PhD, Anny H.
Xiang, PhD7
, Stephanie Kullmann, PhD5,8,9* and Kathleen A. Page, MD1,2,3*
1
Neuroscience Graduate Program, University of Southern California, Los Angeles, CA, USA.
2
Division of Endocrinology and Diabetes, Department of Medicine, Keck School of Medicine,
University of Southern California, Los Angeles, CA, United States.
3
Diabetes and Obesity Research Institute, Keck School of Medicine, University of Southern
California, Los Angeles, CA, United States.
4
Mark & Mary Stevens Neuroimaging & Informatics Institute, Keck School of Medicine,
University of Southern California, Los Angeles, CA, United States.
5
Institute for Diabetes Research and Metabolic Diseases of the Helmholtz Center Munich at the
University of Tübingen, Tübingen, Germany.
6
Department of Psychology, University Southern California, Los Angeles, CA, United States.
7
Department of Research and Evaluation, Kaiser Permanente Southern California, Pasadena,
CA 91101, United States
7
Department of Internal Medicine, Division of Endocrinology, Diabetology and Nephrology,
Eberhard Karls University Tübingen, Tübingen, Germany.
8
German Center for Diabetes Research (DZD), Tübingen, Germany.
*Contributed equally as senior authors.
Submitted to Nature Metabolism

16
Introduction
Obesity rates have risen dramatically over the last three decades, posing a significant
public health challenge70. A growing body of evidence links the rise in sugar-sweetened beverage
consumption to weight gain and obesity71-73. To address this, non-caloric sweeteners are
increasingly consumed as a calorie-free alternative to satisfy the craving for sweetness 2 . Despite
their widespread use, the health implications of non-caloric sweeteners remain uncertain and
subject to debate 3
. While epidemiological studies link non-caloric sweetener consumption to
weight gain4
, obesity5
, and type 2 diabetes6
, randomized controlled trials report that non-caloric
sweeteners have neutral or beneficial effects on body weight and glucose metabolism7
. Studies
conducted in rodents suggest that non-caloric sweeteners stimulate hunger by interfering with the
conventional neural responses to sweet taste and nutrient signaling that occur with caloric
sugar74. Human studies using functional magnetic resonance imaging (fMRI) also indicate that
the brain may respond differently to beverages containing non-caloric sweeteners compared to
caloric sugar 44,75,76 . However, previous fMRI studies have often been constrained by small
sample sizes, comprising healthy-weight individuals 44,75-79. Furthermore, prior studies have
shown a lack of diversity in sex and race/ethnicity, primarily focusing on male and white
participants, which limits their external validity10–12,15,16,18.
In this study, we included a demographically diverse cohort to investigate how sucralose,
a prevalent non-caloric sweetener80, influences hypothalamic activity and glucose signaling
across a range of body weights. Our investigation extends the findings of Yunker et al. (2021),
linking obesity with an increased brain response to food cues after sucralose consumption,
contrasting with the response to sucrose.
Sucralose, chemically altered from sucrose by replacing hydroxyl groups with chlorine,
offers sweetness without caloric absorption81. Drinks with sucralose were calibrated to match the
sweetness of sucrose-containing drinks, allowing us to assess the differential effects of a sweet
taste without nutrients (sucralose) compared to a sweet taste with nutrients (sucrose) on

17
hypothalamic activity, glucose concentrations, and hunger responses. We also compared
sucralose with water to examine the specific effects of sweetness on hypothalamic activity and
the corresponding physiological responses.
The hypothalamus plays a crucial role in appetite and homeostatic metabolic
regulation82,83, and functional connectivity between the hypothalamus and other brain areas
coordinates homeostatic energy balance82,84,85. Prior work shows that ingestion of the simple
sugar, glucose, exerts suppressive effects on hypothalamic activation (evidenced in MRI studies
as reduced blood oxygen level dependent (BOLD) signal or reduced cerebral blood flow (CBF))41-
43,46,65,83,86,87. The glucose-linked reductions in hypothalamic activity are associated with
reductions in hunger46,87, while an increase in hypothalamic activity is associated with heightened
hunger45,88. Based on prior findings, we hypothesized that sucralose, compared to sucrose and
water, would cause greater increases in hypothalamic blood flow and would alter functional
connectivity between the hypothalamus and other brain regions. Additionally, we expected
sucrose, but not sucralose, to raise blood glucose levels, inversely affecting hypothalamic blood
flow. We expected to observe differences between the lateral and medial subfields of the
hypothalamus, given their distinct functional roles. Finally, we projected that these hypothalamic
responses would differ according to participants' weight status.
Methods
Study Overview
Data are from the Brain Response to Sugar study, an investigation of neuroendocrine responses
to high-reward foods (NCT02945475)89. Data presented are the primary results of the arterial spin
labeling perfusion MRI analyses (pASL) from the sucralose, sucrose, and water conditions from
the randomized crossover trial (Figure 1). Glucose was included in the larger trial for the purposes
of testing differences in equicaloric sugars on outcomes89. Participants provided written informed
consent compliant with the University of Southern California Institutional Review Board (IRB #H-

18
09-00395). This study followed Consolidated Standards of Reporting Trials (CONSORT)
guidelines (trial protocol can be found in online digital repository90).
The study included an initial screening visit and MRI visits performed at the Dornsife
Cognitive Neuroimaging Center of the University of Southern California at approximately 8:00 AM
after a 12-hour overnight fast. During the screening visit, height was measured to the nearest 0.1
cm using a stadiometer and weight to the nearest 0.1 kg using a calibrated digital scale. Body
mass index (BMI) was calculated as weight in kg divided by height in meters squared.
The three MRI visits were performed in blinded, random order (using function randperm,
a computer-generated randomization procedure in Matlab) on separate days between 2 and 2
months apart with ingestion of 300 mL drinks containing either sucrose (75 g), sucralose
(individually sweetness matched to sucrose, as previously described89) or a water control to test
their effects on changes in hypothalamic blood flow, circulating glucose levels, and ratings of
hunger (Figure 2). Hypothalamic blood flow (measured by pulsed arterial spin labeling perfusion
MRI), hunger ratings, and glycemic responses were measured fasting, +10 and +35min after drink
ingestion. Hunger ratings and glycemic responses were additionally measured 120 min after drink
ingestion. Drinks were flavored with 0.25 tsp (1.07g) of non-sweetened cherry flavoring (Kraft
Foods Kool-Aid® Unsweetened Cherry Drink Mix) to improve palatability. Females underwent
MRI visits during the follicular phase of the menstrual cycle to reduce variability in hunger91,92.
Participants included in this analysis were 75 adults (43 female) ages 18 to 35 years with
healthy weight, overweight, or obesity (Table 1). They were right-handed, nonsmokers, weightstable for at least 3 months before the study visits, nondieters, not taking medication (except oral
contraceptives), and with no history of diabetes, eating disorders, illicit drug use, or other medical
diagnoses.
The prespecified primary outcome was relative changes in hypothalamic blood flow in
response to acute sucralose vs. sucrose and water among the whole cohort and stratified by
weight status (healthy-weight, overweight, obese). Secondary outcomes included (1) associations

19
between changes in circulating glucose levels, hypothalamic blood flow, and ratings of hunger in
response to sucralose and sucrose; and (2) functional connectivity analysis to investigate brain
regions with MR signal responses that were correlated with the hypothalamic response after
sucralose relative to sucrose and water.
MRI Acquisition Parameters
Participants were scanned at the USC Dana and David Dornsife Neuroimaging Center. Data were
collected using a 3T Siemens MAGNETOM Prismafit MRI System, with a 32-channel head coil.
A high-resolution 3D magnetization prepared rapid gradient echo (MPRAGE) sequence
(TR=1950ms; TE=2.26ms; bandwidth=200Hz/pixel; flip angle=9°; slice thickness=1mm;
FOV=224mm×256mm; matrix=224×256) was used to acquire structural images for multi-subject
registration.
Pulsed arterial spin labeling (pASL) was used to quantify cerebral blood flow (CBF) changes in
response to acute consumption of the different drinks. pASL provides a measure of CBF by
magnetically tagging arterial blood directly before it enters the brain and measuring the amount
of tagged blood to reach specific brain areas93,94. To acquire CBF maps, pulsed arterial spin
labelling images were obtained with a PICORE-Q2TIPS (proximal inversion with control for offresonance effects—quantitative imaging of perfusion by using a single subtraction) sequence by
using a frequency offset corrected inversion pulse and echo planar imaging readout for
acquisition95. The pASL acquisition parameters used in this study were: FOV= 192 mm; matrix
=64x64; bandwidth= 2232 Hz/Pixel; slice thickness=5 mm; in-plane resolution = 3 × 3 mm2
;
interslice spacing= 0 mm; TR= 4000 ms, and flip angle=90; bolus duration (TI1) = .7 seconds,
inversion time (TI2) =1.8 seconds; number of label/control pairs = 60. The first control volume of
the pASL sequence was used as calibration image for CBF quantification. The temporal stability
of pASL compliments the longer curve of glucose responses that we measured with the
accompanying blood draws during the study sessions. Blood oxygen level-dependent (BOLD)
fMRI was acquired with a multi-band interleaved gradient echo planar imaging sequence. Eighty-

20
eight 1.5-mm thick slices covering the whole brain were acquired using the following parameters:
TR=1,000ms; TE=43.20ms; bandwidth=2,055Hz/pixel; flip angle=52°; Multi Band factor = 8;
FOV=128mm×112mm, matrix=128×112. number of volumes = 376 volumes.
Arterial Spin Labeling Analysis
We used the Bayesian Inference for Arterial Spin Labeling (BASIL) toolbox, part of the Oxford
FMRIB Software Library (FSL), to determine mean CBF across the entire brain, as well as regional
CBF in the hypothalamus and hypothalamic subfields. Following motion correction of the whole
image sequence, CBF quantification was performed based on a single compartment model and
voxel-wise calibration. The hypothalamic response to drink was measured as change in regional
CBF (see section below) after drink (averaged across post drink time points) divided by whole
brain CBF to adjust for changes occurring across the whole brain. This post drink value was then
subtracted from pre drink value to correct for baseline CBF. To reduce confounding effects due
to limited spatial resolution, we adjusted for partial volume effects.
Hypothalamus Regions of Interest
We examined the hypothalamic response to different drink conditions by focusing on two distinct
hypothalamic subfields, the lateral hypothalamic (LH) and medial hypothalamic (MH). This
approach was driven by the well-recognized functional differences between these two areas of
the hypothalamus 82. The lateral and medial hypothalamic subfield ROI masks were anatomically
defined according to Baroncini et al.96 and previously used to examine hypothalamic appetite
regulation97. We included an additional exploratory subfield based on the high-resolution
Neudorfer anatomical atlas of hypothalamic nuclei related to energy regulation, in order to enable
direct comparison with future studies that may employ similar methodologies 98. The Neudorfer
hypothalamus ROI includes all hypothalamic nuclei related to energy regulation: lateral
hypothalamus, ventromedial nucleus, dorsomedial hypothalamic nucleus, arcuate nucleus, and
paraventricular nucleus. A single mask was created and normalized into MNI152 space. The

21
Neudorfer ROI was used in bilateral functional connectivity analyses employing the
comprehensive mask encompassing hypothalamic nuclei involved in energy regulation.
Functional Connectivity Analysis
BOLD-fMRI data were collected during a visual food-cue task, where food and non-food images
were presented to participants in random order as previously described89. To explore the primary
effects of different drinks on hypothalamic activity and its functional connections, we performed
an analysis across the whole fMRI time series including both food and non-food stimuli. This
approach allowed us to identify differences in hypothalamic functional connections after the
consumption of sucralose relative to sucrose and water. A similar methodology was previously
applied to investigate how increments in peripheral glucose affect brain connectivity during visual
food and non-food tasks99. Data were processed using Conn Toolbox v21.a
(https://www.nitrc.org/projects/conn)
100,101 and Statistical Parametric Mapping (SPM) v12.7
(https://www.fil.ion.ucl.ac.uk/spm/)
102. Preprocessing included motion alignment, regression of
physiological noise fluctuations and bandpass filtering between 0.008 Hz and 0.09 Hz. Noisecorrected fMRI images were then co-registered to anatomical T1 weighted images and
normalized into MNI standard space. We then performed seed-to-voxel analyses starting from a
seed in the hypothalamus as described above. Results were utilized in second-level group
analyses comparing functional connectivity between drink conditions using a full factorial model.
This model incorporated drink condition as a between-subjects factor and included 4 covariates
correcting for age, sex, BMI, and race/ethnicity. Group-level analyses were performed using a
weighted general linear model (GLM)103, which evaluated voxel-level hypotheses and accounted
for random effects across subjects and sample covariance estimation across multiple
measurements. Statistical inferences for clusters were based on Gaussian Random Field
theory103,104, with significance determined using a voxel-level threshold of p <.001 and an FDRcorrected cluster-size threshold of p < 0.05105.

22
Glucose Assay
Plasma glucose was measured enzymatically using glucose oxidase (YSI 2300 STAT PLUS
Enzymatic Electrode-YSI analyzer, Yellow Springs Instruments).
Hunger Ratings
Visual analogue scales were used to assess feelings of hunger on a scale from 1 to 10 where 1
was “not at all” and 10 was” very much”. Hunger was assessed at baseline, +10, +35 and +120
min after drink consumption. Prior studies have demonstrated good reproducibility and validity of
these VAS scores for assessing subjective sensations of hunger106 .
Statistical Analysis
Descriptive statistics were used to characterize frequency for categorical variables and mean (SD)
and median (interquartile range) for continuous variables and to check distributional properties.
Linear mixed-effects models were employed for comparisons between sucralose and sucrose,
aiming to assess the impact of sweet taste without calories (sucralose) versus sweet taste with
calories (sucrose), and sucralose versus water to investigate the effects of a sweetened drink
without calories (sucralose) versus a non-sweet drink without calories (water). A linear contrast
with a significance threshold of p < .05 was used to compare changes from before to after
ingestion between the sucralose vs sucrose and sucralose vs water conditions. Since there was
no significant interaction between drink condition and time on hypothalamic response, drink
condition was collapsed across both time points. Glycemic measures and hunger ratings were
collapsed across three time points. We also examined the main effect of BMI group on the
hypothalamic response across all drinks, tested for interactions between BMI group and drink
comparisons, and stratified drink comparisons by weight status group. Linear mixed effects
models assumed random intercept for each subject. General linear regression models were used
to assess the association of changes in circulating glucose (independent variable) with

23
hypothalamic response (dependent variable) after drink ingestion and the association between
hypothalamic responses (independent variable) with hunger ratings (dependent variable) after
drink ingestion. Models were adjusted for age, sex, BMI and race/ethnicity. Post hoc comparisons
of drink contrasts and time points were adjusted for multiple comparisons using a Bonferroni
correction when necessary, with significance levels set at .05. Model fits were examined using
the r2beta function from the r2glmm package. All statistical analyses were performed used
Rstudio (version 2023.06.1).
Results
Hypothalamic Blood Flow Response: Comparing Sucralose to Sucrose and Water
There were no baseline differences in hypothalamic blood flow among the three drink sessions
(p=0.40). There was a significant effect of drink condition on the lateral hypothalamic blood flow
response (F (2, 363.88) = 5.05, p < .007, R2 model = .059). Specifically, the response in the lateral
ROI was higher after consuming sucralose compared to sucrose (Mean difference = .079 ± 0.028;
p < .018) and sucralose compared to water (Mean difference = .078 ± 0.028; p < .019), as shown
in Figure 3 and Supplemental Table 1. While there was no interaction between drink and time
on the hypothalamic response (p = 0.26), the hypothalamic response to drinks over time is shown
in Supplemental Figure 2A and Supplemental Table 8 for completeness. No significant difference
was observed in the medial hypothalamic ROI after sucralose compared to sucrose, although the
response was greater after sucralose compared to water (Supplemental Table 1). Additionally,
differences in the Neudorfer ROI between sucralose and water were observed (Supplemental
Table 1).
Although the interaction between weight status and drink contrasts did not reach statistical
significance, we stratified the results by weight status to explore our hypothesis that weight status
influences hypothalamic responses to sucralose compared to sucrose and water. In the lateral
hypothalamic ROI, individuals with obesity had greater hypothalamic responses to sucralose vs

24
water (beta = .105 ± 0.052; p = .042) but not sucralose vs sucrose (beta = .046 ± 0.051; p = .37).
In contrast, individuals with healthy-weight had greater hypothalamic responses to sucralose vs
sucrose (beta = .106 ± 0.048; p = .027) and no differences between sucralose vs water (beta =
.051 ± 0.047; p = .28). There were no differences in individuals with overweight to either sucralose
vs sucrose (beta = .081 ± 0.049; p = .102) or sucralose vs water (beta = .081 ± 0.049; p = .103)
(Figure 4, Supplemental Table 2). The Neudorfer and medial hypothalamic ROIs showed
differential responses to sucralose relative to sucrose comparisons only among those with
healthy-weight (Supplemental Table 2).
We also examined the hypothalamic responses to sucrose compared with water. There
were no differences in hypothalamic blood flow when comparing sucrose to water conditions
across BMI groups (Supplemental Table 1). However, the data revealed a pattern where
individuals with healthy-weight showed a non-significant trend towards decreased lateral
hypothalamic blood flow in response to sucrose compared to water (mean difference = -0.06, ±
.047 p = .25), in contrast to individuals with obesity, who showed an increase, though this was
also not statistically significant (mean difference = 0.059, ± .051, p = .25) (Supplemental Table
2).
Hypothalamic Seed-to-Voxel Connectivity
Statistical analysis of the connectivity from the bilateral hypothalamus to all other voxels in the
brain resulted in different connectivity patterns after the ingestion of sucralose, relative to sucrose
and water. After ingestion of sucralose relative to sucrose, we observed increased connectivity
between the left hypothalamus and anterior cingulate cortex (Figure 5A). After ingestion of
sucralose relative to water, we observed increased connectivity between the right hypothalamus
and left superior parietal lobule (Figure 5B). After ingestion of sucrose relative to water, we
observed increased connectivity between the right hypothalamus and precuneus cortex and
decreased connection between right hypothalamus and occipital pole (Figure 5C).

25
Circulating Glucose Responses
Mean results for the effects of sucrose, sucralose, and water on changes in peripheral glucose
levels were previously reported89. There were no baseline differences in peripheral glucose levels
among the three drink sessions (p=0.596). There was a significant effect of drink on the peripheral
glucose response (F (2, 552.65) = 139.36, p = < 0.00001, R2 model = .34), adjusting for age, sex,
race/ethnicity, and BMI. Post hoc analysis showed a marked increase in peripheral glucose
following sucrose compared to sucralose intake (p < 0.0003), but no differences were observed
in peripheral glucose levels when comparing sucralose to water (p = .99) (Supplemental Table
3). There was an interaction between time and drink on peripheral glucose levels (P<0.002), and
drink comparison by time on peripheral glucose levels shown in Supplemental Figure 2B and
Supplemental Table 8. There were no differences in peripheral glucose levels by weight status
(p < .47).
Hunger responses
There were no differences in baseline hunger ratings among the drink sessions (p=0.678),
however, there was a significant effect of drink condition on changes in hunger (F (2,580.21) =
10.79, p < .00005, R2 model = .153). Post hoc analysis revealed a significant increase in hunger
after sucralose compared to sucrose (mean diff = .575 ± 0.16; p < .001) but no differences after
sucralose vs water (mean diff = -.090 ± 0.16; p = .99) (Supplemental Table 4, Supplemental
Figure 2). While there was no interaction between time and drink on hunger ratings (p=0.38),
drink comparisons by time on hunger are shown in Supplemental Figure 2C and Supplemental
Table 8 for informational purposes. A trending difference in hunger was observed by weight status
(p = .053), with greater hunger among individuals with obesity compared to healthy-weight (mean
diff = .975 ± 0.40; p = .056).

26
Associations between changes in peripheral glucose levels and hypothalamic blood flow
There was a significant relationship between changes in circulating glucose levels and blood flow
in the medial hypothalamus 30 minutes after consuming sucrose (beta = −0.005 ± 0.002, p <
0.007) (see Figure 6A and Supplemental Table 5). However, no significant association was
found between changes in circulating glucose and hypothalamic blood flow after consuming
sucralose (p = 0.19; Figure 6A). Associations were suggested between circulating glucose and
responses in other hypothalamic ROIs after sucrose ingestion, as noted in Supplemental Table 5.
In an exploratory analysis stratified by weight status, negative associations were observed
between increments in peripheral glucose and the medial hypothalamic response to sucrose in
individuals with healthy weight and those overweight, but not in individuals with obesity (see
Supplemental Table 6)
Associations between changes in hypothalamic blood flow and hunger
Decreases in medial hypothalamic blood flow observed within 10 minutes after consuming
sucrose were associated with reduced hunger post-ingestion (Supplemental Table 7, Figure
6B). Similar trends were observed in both the lateral hypothalamus and Neudorfer ROI, although
these did not reach statistical significance (Supplemental Table 7). No such associations were
found following sucralose ingestion (Supplemental Table 7, Figure 6B).
Discussion
In this randomized cross-over trial involving healthy young adults of varying weights, we
show that drinks sweetened with sucralose, a non-caloric sweetener, led to an increase in
hypothalamic blood flow—a purported MRI marker of hunger—when compared to caloric sugar
(sucrose) and water. Sucrose, compared to sucralose, had a hunger-dampening effect while also
raising peripheral glucose levels, which corresponded to reduced medial hypothalamic blood flow.
These results support the notion, initially observed in rodents74,107, that non-caloric sweeteners

27
may alter appetite by interfering with the conventional neural responses to sweet taste and
nutrient signaling observed with caloric sugar.
Stratified analyses based on weight categories revealed variations in the hypothalamic
responses to sucralose relative to sucrose and water. Specifically, healthy-weight individuals
displayed a marked increase in hypothalamic blood flow in response to sucralose compared to
sucrose. This suggests that for individuals of healthy weight, the hypothalamic response to drinks
containing the non-caloric sweetener, sucralose, differs significantly from that of the caloric
sweetener, sucrose. In contrast, individuals with obesity exhibited a stronger increase in the
hypothalamic response to sucralose than to water, indicating a potential relationship between
obesity and heightened hypothalamic sensitivity to sweetness. Understanding this relationship
may have important implications for the use of non-caloric sweeteners in weight management
strategies.
Our findings revealed that oral intake of sucrose led to increased peripheral glucose levels
and reduced hypothalamic activity, while sucralose had no such effect. This supports existing
evidence suggesting that metabolic signals like glucose are tightly connected to changes in
hypothalamic activity41,42,46,88,108. Importantly, the relationship between peripheral glucose
fluctuations and hypothalamic activity was less evident in individuals with obesity, further
reinforcing the idea that obesity may disrupt glucose signaling within the hypothalamus.
By including two distinct subfields of the hypothalamus—the lateral hypothalamus and the
medial hypothalamus, which comprises the ventromedial hypothalamic nucleus (VMH) and the
arcuate nucleus—we aimed to explore subfield-specific roles. Previous research in rats showed
that bilateral lesions in the VMH led to hyperphagia, while lesions in the lateral hypothalamic area
(LHA) resulted in hypophagia109. These foundational studies identified the VMH as the "satiety
center," which suppresses food intake, and the LHA as the "hunger center," which stimulates
eating. Our findings revealed that sucralose, compared to sucrose or water, induced the most
pronounced differences in the lateral hypothalamic subfield. However, the response patterns in

28
the medial hypothalamic subfield and the hypothalamus ROI defined by the Neudorfer highresolution atlas were similar. Importantly, associations between changes in peripheral glucose
levels post-sucrose ingestion were notable in the medial hypothalamic subfield but not in the
lateral subfield. Furthermore, reductions in blood flow in the medial hypothalamus correlated with
decreased hunger. These findings align with prior studies demonstrating that glucose ingestion
reduced the fMRI signal in the hypothalamic area corresponding to the VMH, correlating with
lower fasting plasma glucose and insulin levels, highlighting the significant role of the medial
hypothalamus in nutrient sensing41.
Finally, our functional connectivity analysis revealed that acute consumption of sucralose,
as opposed to sucrose, significantly increased coupling between the hypothalamus and the
anterior cingulate cortex (ACC)—an area of the brain that plays a crucial role in attention,
motivation, and reward processing 110. Additionally, compared to water, sucralose intake led to
greater connectivity between the right hypothalamus and the left superior parietal lobule, a region
integral to somatosensory integration111. These results suggest that sucralose may enhance the
functional connection between brain regions coordinating appetite with reward and motivation,
potentially influencing food-seeking behavior.
Limitations
There are several limitations to be considered. Our study investigated both the medial and
lateral hypothalamic subfields to determine if the varying effects of sucralose, compared to
sucrose and water, could be attributed to specific areas within the hypothalamus. However, due
to the limitations in spatial resolution, we were unable to precisely attribute these effects to
individual neurons within these hypothalamic subfields. While the primary focus of this study is to
investigate the isolated effects of the non-caloric sweetener, sucralose, considering that
beverages are frequently consumed as part of a meal containing carbohydrates112 and proteins,
further investigation into whether hypothalamic responses are differentially modulated by the

29
ingestion of sucralose within a mixed-meal context would be valuable. Given that the unique
chemical structure of each type of non-caloric sweetener may elicit varying physiological
responses113, future studies should examine whether altered hypothalamic and metabolic
responses are also observed in other types of non-caloric sweeteners. Finally, since the study
focused on examining the acute effects of sucralose consumption, we did not investigate the
impact of chronic consumption of non-caloric sweeteners on hypothalamic signaling and
appetitive behaviors. Given that rodent studies have shown chronic consumption of non-nutritive
sweeteners to alter central appetite signaling mechanisms114, more research addressing the longterm effects of non-caloric sweeteners is warranted.
Conclusion
Our findings indicate that non-caloric sweeteners could affect key mechanisms in the
hypothalamus responsible for appetite regulation, and that individuals with obesity might be
particularly susceptible to effects of non-caloric sweeteners on appetite regulation. Considering
the prevalent consumption of non-caloric sweeteners, it is vital to conduct comprehensive studies
to clarify their long-term health ramifications.

30
Chapter 2 Figures and Tables
Table 1. [Chapter 2] Participant Characteristics
N=75 Mean or Freq
(SD or Freq %)
Range
Age (years) 23.33 (3.97) 18.15-34.51
BMI (kg/m2) 27.16 (5.17) 19.18-40.28
BMI Group
Obese 23 (30.7%)
Overweight 24 (32%)
Healthy Weight 28 (37.3%)
Sex
F 43 (57.33%)
M 32 (42.66%)
Race & Ethnicity
Asian 23 (30.66%)
Hispanic 19 (25.33%)
Non-Hispanic Black 12 (16%)
Non-Hispanic White 21 (28%)
Figure 1: [Chapter 2] Participant Enrollment Flowchart for the Randomized Crossover
Brain Response to Sugar II Trial and Final Analysis
a Of the 76 participants enrolled, who received at least 1 drink allocation, 1 participant did not receive any of the
drinks (i.e., sucralose, sucrose, or water) included in this analysis because of dropout and was excluded from this
analysis (N=75).

31
Figure 2: Schematic of Study Design
BOLD indicates blood oxygen level-dependent.

32
Figure 3: Differential Hypothalamic Response to Drink Comparisons
Figure 3: Significant increases in hypothalamic blood flow were observed after sucralose vs sucrose (p = 0.018) and
sucralose vs water, (p = 0.019) comparisons, adjusted for multiple comparisons using a Bonferroni correction. Data
showing lateral hypothalamic ROI.
Figure 4: Difference in Hypothalamic Response to Drinks by Weight Status
Figure 4. Mean Change in lateral hypothalamic blood flow after water, sucrose or sucralose ingestion stratified by
weight status. Individuals with healthy weight (green circles) had greater hypothalamic blood flow after sucralose vs

33
sucrose (p=0.027) but not sucralose vs water (p=0.279). Individuals with obesity (red circles) had greater
hypothalamic activation after sucralose vs water (p=0.042) but not sucralose vs sucrose (p=0.370). Individuals with
overweight (yellow circles) had no differences in hypothalamic response to either drink comparison.
Figure 5: Differential functional connectivity from hypothalamus seed region after
sucralose ingestion relative to sucrose and water
Figure 5. Seed-to-voxel analysis comparing functional connectivity between the hypothalamus (seed region) and
other brain regions after sucralose ingestion relative to sucrose or water, adjusting for age, sex, BMI, and
race/ethnicity. (A) Sucralose compared to sucrose resulted in increased connectivity between the left hypothalamus
and anterior cingulate cortex, (B) Sucralose compared to water resulted in increased connectivity between the right
hypothalamus and left superior parietal lobule. Significance was set at p < 0.05 with correction for multiple
comparisons using the false discovery rate (FDR q < 0.05). Hot colors in red, orange and yellow indicate a more
positive z score, suggesting greater connectivity after sucralose relative to comparison drink. Neudorfer hypothalamic
ROI was used as the seed region. (C) Sucrose, compared to water, resulted in increased connectivity between the
right hypothalamus and precuneus cortex and decreased connectivity between the right hypothalamus and the
occipital pole. Five participants had excessive motion during the BOLD acquisition and were excluded from the
functional connectivity analysis (N=70).
A. Sucralose vs. Sucrose
B. Sucralose vs. Water
C. Sucrose vs. Water

34
Figure 6: Associations between Peripheral Glucose, Hunger, and Changes in Medial
Hypothalamic Blood Flow
Figures show scatterplots with Pearson correlations for visual purposes. Linear regression models were used in the
statistical analysis and data are reported in Supplemental Tables 5 and 7.

35
Supplemental Data
Table 2, Supplemental Table 1. Hypothalamic Response to Sucralose compared to
Sucrose and Water
Contrast Mean difference SE t-stat p-value p-adjusted†
Lateral Hypothalamus Sucralose vs Sucrose 0.079 0.028 2.764 0.006 0.018
Sucralose vs Water 0.078 0.028 2.746 0.006 0.019
Sucrose vs Water -0.001 0.028 -0.019 0.985 .999
Medial Hypothalamus Sucralose vs Sucrose 0.038 0.031 1.200 0.231 0.693
Sucralose vs Water 0.095 0.031 3.033 0.003 0.008
Sucrose vs Water 0.058 0.031 1.837 0.067 0.201
Hypothalamus (Neudorfer) Sucralose vs Sucrose 0.043 0.021 2.036 0.043 0.128
Sucralose vs Water 0.055 0.021 2.585 0.010 0.030
Sucrose vs Water 0.012 0.021 0.564 0.573 .999
Models covaried for Age, Sex, BMI, and Race/ethnicity. Change in Hypothalamic blood flow collapsed across 35 minutes; N = 75. †Correction for
multiple comparisons (3 drink contrasts) using a Bonferroni correction.
Table 3, Supplemental Table 2. Drink Comparisons of Hypothalamic Response Stratified
by Weight Status
Drink Comparisons Stratified by Weight Status
Lateral Hypothalamus Mean difference SE t-stat p-value
Sucralose vs Sucrose Obese 0.046 0.051 0.897 0.370
Overweight 0.081 0.049 1.637 0.102
Healthy weight 0.106 0.048 2.224 0.027
Sucralose vs Water Obese 0.105 0.052 2.040 0.042
Overweight 0.081 0.049 1.636 0.103
Healthy weight 0.051 0.047 1.085 0.279
Sucrose vs Water Obese 0.059 0.051 1.159 0.247
Overweight 0.000 0.049 -0.001 0.999
Healthy weight -0.055 0.047 -1.155 0.249
Medial Hypothalamus
Sucralose vs Sucrose Obese -0.009 0.057 -0.152 0.879
Overweight 0.020 0.055 0.372 0.710
Healthy weight 0.095 0.053 1.798 0.073
Sucralose vs Water Obese 0.113 0.057 1.978 0.049
Overweight 0.066 0.055 1.202 0.230
Healthy weight 0.108 0.052 2.054 0.041
Sucrose vs Water Obese 0.123 0.057 2.165 0.031
Overweight 0.045 0.055 0.830 0.407
Healthy weight 0.014 0.052 0.260 0.795
Hypothalamus (Neudorfer)
Sucralose vs Sucrose Obese -0.007 0.038 -0.188 0.851
Overweight 0.049 0.037 1.331 0.184
Healthy weight 0.082 0.036 2.315 0.021
Sucralose vs Water Obese 0.038 0.039 0.986 0.325
Overweight 0.049 0.037 1.331 0.184
Healthy weight 0.059 0.036 1.628 0.104
Sucrose vs Water Obese 0.045 0.038 1.182 0.238
Overweight 0.018 0.037 0.492 0.623
Healthy weight -0.024 0.036 -0.658 0.511
Exploratory analysis. Models covaried for Age, Sex, and Race/ethnicity. Change in Hypothalamic blood flow collapsed across 35 minutes; N = 75.

36
Table 4, Supplemental Table 3. Differential Effects of Drinks on Peripheral Glucose
Levels
Contrast Mean difference SE t-stat p-value p-adjusted†
By condition
Sucralose vs Sucrose -21.459 1.456 -14.743 <0.00001 <0.00003
Sucralose vs Water -0.522 1.491 -0.350 0.727 1.00
Sucrose vs Water 20.937 1.491 14.038 <0.00001 <0.00004
Models covaried for Age, Sex, BMI, and Race/ethnicity. Change in peripheral glucose collapsed across 120 minutes; N = 72 (3 subjects did not
complete blood sampling at all 3 visits). †Correction for multiple comparisons (3 drink contrasts) using a Bonferroni correction.
Table 5, Supplemental Table 4. Differential Effects of Drinks on Changes in Hunger
Hunger Contrast Mean
difference
SE t-stat p-value p-adjusted†
By condition1
Sucralose vs Sucrose 0.575 0.155 3.700 <0.0002 <0.001
Sucralose vs Water -0.090 0.155 -0.580 0.562 .999
Sucrose vs Water -0.665 0.155 -4.285 <.00002 <0.00006
1Models covaried for Age, Sex, BMI, and Race/ethnicity. N= 75. †Correction for multiple comparisons (3 drink contrasts) using a Bonferroni correction.
Table 6, Supplemental Table 5. Associations Between Peripheral Glucose Levels and
Hypothalamic Response to Sucrose Consumption
IV DV Beta SE t-stat p-value p-adjusted†
Glucose: Time 1 Lateral Hypothalamus Time 1 0.000 0.002 -0.146 0.884 .999
Glucose: Time 2 Lateral Hypothalamus Time 2 -0.003 0.001 -1.772 0.084 0.169
Glucose: Time 1 Medial Hypothalamus Time 1 -0.002 0.003 -0.700 0.487 0.973
Glucose: Time 2 Medial Hypothalamus Time 2 -0.005 0.002 -3.056 0.003 0.007
Glucose: Time 1 Hypothalamus (Neudorfer) Time 1 -0.002 0.002 -0.978 0.332 0.663
Glucose: Time 2 Hypothalamus (Neudorfer) Time 2 -0.002 0.001 -1.753 0.081 0.163
All models covaried for Age, Sex, BMI, and Race/ethnicity; Time 1 is 10 min after drink ingestion; Time 2 is 35 min after drink ingestion; N=69 (6
subjects did not receive sucrose drink allocation and complete blood glucose sampling). †Correction for multiple comparisons (2 time points) using a
Bonferroni correction.
Table 7, Supplemental Table 6. Associations Between Peripheral Glucose and Medial
Hypothalamic Blood Flow Response to Sucrose, Stratified by Weight Status
IV Contrast Beta SE t-stat p-value
By BMI group
Obesity -0.003 0.003 -0.973 0.335
Overweight -0.006 0.003 -2.284 0.026
Healthy weight -0.006 0.003 -2.260 0.027
Exploratory analysis. Models covaried for Age, Sex, and Race/ethnicity. Associations were observed 35 min after drink ingestion; N=69 (6 subjects did
not receive sucrose drink allocation and complete blood glucose sampling).

37
Table 8, Supplemental Table 7. Associations Between Hypothalamic Response and
Hunger
IV‡ DV Beta SE t-stat p-value p-adjusted†
By Sucralose Condition 1
Lateral Hypothalamus Time 1 Hunger Time 1 -0.61 1.018 -0.599 0.551 .999
Lateral Hypothalamus Time 1 Hunger Time 2 -1.188 1.09 -1.09 0.28 0.839
Lateral Hypothalamus Time 1 Hunger Time 3 0.676 1.198 0.565 0.574 .999
Medial Hypothalamus Time 1 Hunger Time 1 0.491 0.874 0.563 0.576 .999
Medial Hypothalamus Time 1 Hunger Time 2 -0.064 0.944 -0.068 0.946 .999
Medial Hypothalamus Time 1 Hunger Time 3 1.22 1.019 1.197 0.236 0.707
Hypothalamus (Neudorfer) Time 1 Hunger Time 1 0.09 1.2 0.075 0.941 .999
Hypothalamus (Neudorfer) Time 1 Hunger Time 2 -1.081 1.286 -0.84 0.404 .999
Hypothalamus (Neudorfer) Time 1 Hunger Time 3 1.464 1.4 1.045 0.3 0.899
By Sucrose Condition 2
Lateral Hypothalamus Time 1 Hunger Time 1 -1.041 0.988 -1.054 0.296 0.887
Lateral Hypothalamus Time 1 Hunger Time 2 -2.372 1.037 -2.288 0.025 0.076
Lateral Hypothalamus Time 1 Hunger Time 3 -1.993 1.080 -1.846 0.069 0.208
Medial Hypothalamus Time 1 Hunger Time 1 -1.643 0.753 -2.181 0.033 0.098
Medial Hypothalamus Time 1 Hunger Time 2 -1.795 0.813 -2.207 0.031 0.092
Medial Hypothalamus Time 1 Hunger Time 3 -2.096 0.833 -2.517 0.014 0.043
Hypothalamus (Neudorfer) Time 1 Hunger Time 1 -1.819 1.194 -1.523 0.133 0.398
Hypothalamus (Neudorfer) Time 1 Hunger Time 2 -2.679 1.272 -2.106 0.039 0.117
Hypothalamus (Neudorfer) Time 1 Hunger Time 3 -2.507 1.314 -1.908 0.061 0.182 ‡ Hypothalamic response to drink variables (independent variable) were transformed by multiplication with -1 to reflect their inverse relationship with the
dependent variable. Models covaried for Age, Sex, BMI, and Race/ethnicity. Time 1 is 10 min after drink ingestion; Time 2 is 35 min after drink
ingestion; Time 3 is 120 min after drink ingestion. 1 N=72 (3 subjects did not receive sucralose drink allocation); 2 N=73 (2 subjects did not receive
sucrose drink allocation). †Correction for multiple comparisons (3 time points) were conducted using a Bonferroni correction.
Figure 7, Supplemental Figure 1: Visual Display of Hypothalamic ROI
Hypothalamic Region of interest (ROIs) and corresponding coordinates. A) Lateral hypothalamus (green), B) Medial hypothalamus
(blue), C) Neudorfer (yellow). The images are displayed in neurological convention.

38
Figure 8, Supplemental Figure 2: Visual Display of Changes in Hypothalamic, Peripheral
Glucose, and Hunger Responses over Time
Supplemental Figure 2 shows change in: A) hypothalamic response (lateral hypothalamus), B) peripheral glucose (mg/dl), and C)
self-reported hunger ratings to each drink condition: water (green), sucrose (blue), and sucralose (red). Time 0 represents no
change in responses in measurement at baseline. Time 1 represents change in measurement at 10 minutes from baseline, Time 2
represents change in measurement at 35 minutes from baseline, and Time 3 represents change in measurement at 120 minutes
from baseline.

39
Table 9, Supplemental Table 8. Drinks Comparisons by Time for Lateral Hypothalamic
Response, Peripheral Glucose Levels, and Hunger Ratings
Contrast Time Mean difference SE t-stat p-value
Lateral Hypothalamus
Sucralose vs Sucrose 1 0.076 0.040 1.889 0.060
Sucralose vs Sucrose 2 0.081 0.040 2.018 0.044
Sucralose vs Water 1 0.097 0.040 2.415 0.016
Sucralose vs Water 2 0.059 0.040 1.467 0.143
Sucrose vs Water 1 0.021 0.040 0.527 0.598
Sucrose vs Water 2 -0.022 0.040 -0.553 0.580
Peripheral Glucose Levels
Sucralose vs Sucrose 1 -25.361 1.897 -13.368 < .000001
Sucralose vs Sucrose 2 -38.154 1.897 -20.112 < .000001
Sucralose vs Sucrose 3 -0.850 1.897 -0.448 0.654
Sucralose vs Water 1 -1.817 1.943 -0.935 0.350
Sucralose vs Water 2 -0.435 1.943 -0.224 0.823
Sucralose vs Water 3 0.293 1.943 0.151 0.880
Sucrose vs Water 1 23.544 1.945 12.106 < .000001
Sucrose vs Water 2 37.719 1.945 19.394 < .000001
Sucrose vs Water 3 1.142 1.945 0.587 0.557
Hunger Ratings
Sucralose vs Sucrose 1 0.465 0.256 1.815 0.070
Sucralose vs Sucrose 2 0.776 0.256 3.025 0.003
Sucralose vs Sucrose 3 0.475 0.257 1.843 0.066
Sucralose vs Water 1 -0.029 0.256 -0.113 0.910
Sucralose vs Water 2 -0.308 0.256 -1.200 0.231
Sucralose vs Water 3 0.069 0.256 0.268 0.789
Sucrose vs Water 1 -0.494 0.256 -1.933 0.054
Sucrose vs Water 2 -1.083 0.256 -4.236 < .00002
Sucrose vs Water 3 -0.406 0.257 -1.581 0.114

40
Chapter 3: Differential Effects of Timing and Weight Status on
Hypothalamic Response to Sucrose
Sandhya P. Chakravartti, MS1,2,3, Kay Jann, PhD4
, Ralf Veit, PhD5
, Alexandra G. Yunker, MPH2,3
,
Brendan Angelo, MS2,3
, John R. Monterosso1,6, PhD, Anny H. Xiang, PhD7
, Stephanie Kullmann,
PhD5,8,9* and Kathleen A. Page, MD1,2,3*
1
Neuroscience Graduate Program, University of Southern California, Los Angeles, CA, USA.
2
Division of Endocrinology and Diabetes, Department of Medicine, Keck School of Medicine,
University of Southern California, Los Angeles, CA, United States.
3
Diabetes and Obesity Research Institute, Keck School of Medicine, University of Southern
California, Los Angeles, CA, United States.
4
Mark & Mary Stevens Neuroimaging & Informatics Institute, Keck School of Medicine,
University of Southern California, Los Angeles, CA, United States.
5
Institute for Diabetes Research and Metabolic Diseases of the Helmholtz Center Munich at the
University of Tübingen, Tübingen, Germany.
6
Department of Psychology, University Southern California, Los Angeles, CA, United States.
7
Department of Research and Evaluation, Kaiser Permanente Southern California, Pasadena,
CA 91101, United States
7
Department of Internal Medicine, Division of Endocrinology, Diabetology and Nephrology,
Eberhard Karls University Tübingen, Tübingen, Germany.
8
German Center for Diabetes Research (DZD), Tübingen, Germany.
*Contributed equally as senior authors.

41
Introduction
The hypothalamus plays a central role in regulating energy homeostasis and glucose
metabolism by integrating signals related to hunger, satiety, and energy expenditure.
Dysregulation of these mechanisms can lead to impaired glucose regulation and contribute to the
development of obesity in individuals. Early studies mapping the hypothalamic response to
glucose in humans have highlighted the temporal nature of this response, and its variation
between lean and obese individuals. Using BOLD fMRI, a technique that measures changes in
blood oxygenation levels which reflect neural activity, Matsuda and colleagues demonstrated that
oral glucose ingestion led to a significant, transient reduction in hypothalamic signals in lean
subjects, beginning four minutes post-ingestion and lasting approximately ten minutes41. This
reduction was slower and smaller in obese individuals, with a delayed response observed in the
upper anterior hypothalamic region, suggesting that delayed activation of satiety centers may
contribute to excessive food intake in obesity. Similarly, Liu et al. found that healthy adults
exhibited a reduction in hypothalamic activity within one to two minutes and again at seven to
twelve minutes after glucose ingestion, correlating with fasting plasma insulin levels42. Extending
these findings, Smeets and colleagues showed a prolonged, dose-dependent decrease in
hypothalamic MRI signals for up to thirty minutes following glucose ingestion and identified that
only glucose, not sweet taste alone (aspartame) or carbohydrate alone (maltodextrin), elicited this
response44. In a follow up study, Vidarsdottir and colleagues examined the fMRI signal in
response to glucose ingestion in patients with type 2 diabetes compared to control participants
and found an impairment of the fMRI signal115. These studies indicate that the timing and intensity
of the hypothalamic signals are triggered before and during nutrient absorption, play a crucial role
in the adaptive response to food intake, and are shaped by weight status. However, there are
limitations to the fMRI methods, as it is an indirect measure of neural response, and is susceptible
to artifacts.

42
Pulsed Arterial Spin Labeling (ASL) is a functional magnetic resonance imaging technique
used to measure absolute cerebral blood flow, which involves magnetically labeling arterial blood
water as an endogenous tracer, providing a non-invasive means to assess perfusion. ASL is
particularly beneficial for examining slower metabolic responses triggered by pharmacological
interventions116 or nutrient ingestion, and the temporal stability of ASL aligns with the prolonged
glucose response curves compared to BOLD techniques. Studies from Page and colleagues
utilized pASL techniques to reveal that glucose significantly reduced hypothalamic response
within 15 minutes of ingestion, suggesting concordance in temporality between fMRI and pASL
methods46. Furthermore, experiments detailed in Chapter 2 revealed significant differences in
hypothalamic blood flow between sucrose and water across the 10 and 35-minute time frame, as
well attenuation of differences between sucrose and water in individuals with obesity. However,
the temporal specifics of the hypothalamic response to sucrose in adults with varying weight
statuses, including those who are overweight, remain unclear.
This study aims to investigate the temporal mechanisms of hypothalamic response to
sucrose ingestion in young adults of different weight statuses: obese, overweight, and healthy
weight using pASL techniques. Prior research indicates that adults with obesity and type II
diabetes exhibit glucose dysregulation, characterized by a slower and attenuated fMRI signal
following glucose ingestion. Based on this, we hypothesize a slower and attenuated ASL signal
in overweight and obese participants compared to healthy weight participants.
Methods
Methods presented below are also presented in Chapter 2. The analysis included 75
adults (43 females) aged 18 to 35 years, categorized as healthy weight, overweight, or obese
(Table 1). All participants were right-handed, nonsmokers, weight-stable for at least 3 months
prior to the study visits, and nondieters. They were not taking any medication, except for oral
contraceptives, and had no history of diabetes, eating disorders, illicit drug use, or other medical
conditions. The study involved an initial screening visit and subsequent MRI visits at the Dornsife

43
Cognitive Neuroimaging Center, University of Southern California, at approximately 8:00 AM after
a 12-hour overnight fast. The three MRI visits were conducted in a blinded, random order using a
computer-generated randomization procedure in Matlab (randperm function) on separate days,
spaced between 2 weeks and 2 months apart. During these visits, participants ingested 300 mL
drinks containing either sucrose (75 g), sucralose (matched for sweetness to sucrose89), or a
water control to test their effects on hypothalamic blood flow. Hypothalamic blood flow (measured
by pulsed arterial spin labeling perfusion MRI). were measured fasting, +10 and +35min after
drink ingestion. Drinks were flavored with 0.25 tsp (1.07g) of non-sweetened cherry flavoring
(Kraft Foods Kool-Aid® Unsweetened Cherry Drink Mix) to improve palatability. Females
underwent MRI visits during the follicular phase of the menstrual cycle to reduce variability in
hunger29,30.
The analysis included 75 adults (43 females) aged 18 to 35 years, categorized as healthy
weight, overweight, or obese (Table 1). All participants were right-handed, nonsmokers, weightstable for at least 3 months prior to the study visits, and nondieters. They were not taking any
medication, except for oral contraceptives, and had no history of diabetes, eating disorders, illicit
drug use, or other medical conditions.
MRI Acquisition Parameters
Participants were scanned at the USC Dana and David Dornsife Neuroimaging Center
using a 3T Siemens MAGNETOM Prismafit MRI System with a 32-channel head coil. Highresolution structural images were acquired using a 3D magnetization prepared rapid gradient
echo (MPRAGE) sequence (TR=1950ms; TE=2.26ms; bandwidth=200Hz/pixel; flip angle=9°;
slice thickness=1mm; FOV=224mm×256mm; matrix=224×256).
Arterial Spin Labeling Analysis and Hypothalamus Region of Interest

44
Pulsed arterial spin labeling (pASL) was employed to quantify cerebral blood flow (CBF)
changes following the acute consumption of different drinks. pASL tags arterial blood just before
it enters the brain, allowing measurement of tagged blood reaching specific brain areas. CBF
maps were obtained with a PICORE-Q2TIPS sequence using a frequency offset corrected
inversion pulse and echo planar imaging readout. Acquisition parameters were: FOV=192 mm;
matrix=64x64; bandwidth=2232 Hz/Pixel; slice thickness=5 mm; in-plane resolution=3x3 mm²;
interslice spacing=0 mm; TR=4000 ms; flip angle=90°; bolus duration (TI1)=0.7 seconds;
inversion time (TI2)=1.8 seconds; number of label/control pairs=60. The first control volume of the
pASL sequence served as a calibration image for CBF quantification.
Mean CBF across the entire brain and regional CBF in the hypothalamus and its subfields
were determined using the Bayesian Inference for Arterial Spin Labeling (BASIL) toolbox within
the Oxford FMRIB Software Library (FSL). Following motion correction, CBF quantification was
performed using a single compartment model and voxel-wise calibration. Hypothalamic response
to the drinks was measured as the change in regional CBF (averaged across post-drink time
points), divided by whole brain CBF to adjust for global changes. This post-drink value was
subtracted from the pre-drink value to correct for baseline CBF. Partial volume effects were
adjusted to reduce confounding due to limited spatial resolution. We investigated the
hypothalamic response to different drink conditions by focusing on the functionally distinct82 lateral
hypothalamic (LH) and medial hypothalamic (MH) subregions. The LH and MH subfield ROI
masks were anatomically defined based on the work of Baroncini et al. 96, which has been
previously used to study hypothalamic appetite regulation97.
Statistical Analysis
General linear regression models were used to assess differences in hypothalamic and
whole brain cerebral responses by BMI. Models were adjusted for age, sex, and race/ethnicity. All
statistical analyses were performed used Rstudio (version 2023.06.1).

45
Results
At 10 minutes mean change in lateral hypothalamic response to sucrose for healthy
weight individuals was -.114 (SD = .20), .001 (SD = .20) for overweight individuals, and .013 (SD
= .28) for individuals with obesity. At 35 minutes mean change in lateral hypothalamic response
was -.157 (SD = .26) for healthy weight individuals, -.089 (SD = .24) for individuals who are
overweight, and .024 (SD = .235). for individuals with obesity. In the medial hypothalamus at 10
minutes mean change in hypothalamic response for healthy weight individuals was -.063 (SD =
.30), -.02 (SD = .242) for overweight individuals, and -.089 (SD = .34) for individuals with obesity
in the lateral hypothalamus. At 35 minutes mean change in medial hypothalamic response was -
.070 (SD = .31) for healthy weight individuals, -.022 (SD = .23) for individuals who are overweight,
and .077 (SD = .30) for individuals with obesity.
At 10 minutes significant differences in lateral hypothalamic response were observed in
healthy vs obese (mean diff = -.148, p = .030) and healthy vs overweight individuals (mean diff =
-.157, p = .027) (Table 2). There were no significant differences observed in between overweight
and obese individuals (p = .90). At 35 minutes differences in the lateral hypothalamus were only
observed between the healthy and obese groups (mean diff = -.190; p = .01) (Table 3). We did
not observe any difference between the healthy and overweight groups (p = .240) or the
overweight and obese groups (p = .193). At 10 minutes within the medial hypothalamus,
differences were only observed in the healthy vs obese groups (diff = -.173, p = .049) with no
differences in the healthy vs overweight or overweight (p = .22) vs obese (p = .48) groups (Table
2). At 35 minutes there were no differences observed between any of the weight groups within
the medial hypothalamus (all p’s > .10) (Table 3).
Follow-up analyses examined the amount of hypothalamic and whole brain cerebral blood
flow at baseline. No differences in hypothalamic blood flow prior to the ingestion of the drink
(baseline) were found between different weight groups in both the lateral (all p’s > .09) and medial

46
(all p’s > .07) hypothalamus (Table 4). Within the lateral hypothalamus mean relative
hypothalamic blood flow values (divided by whole brain CBF) at baseline for healthy weight
individuals were: 1.09 mL/100g/min (SD = .25), 1.211 mL/100g/min (SD = .28) for overweight
individuals, and 1.24 mL/100g/min (SD = .23) for individuals with obesity. Similar values were
observed in the medial hypothalamus (Table 5). Whole brain gray matter cerebral blood flow
values at baseline for healthy weight individuals were: 36.9 mL/100g/min (SD = 4.9), 35.9
mL/100g/min (SD = 3.9) for overweight individuals, and 36.9 mL/100g/min (SD = 5.2) for obese
individuals (Table 5). There were no significant differences in whole brain cerebral blood flow prior
to drink consumption among the three weight groups (all p’s > .62).
As a control drink condition, we examined changes in hypothalamic blood flow at 15 and
30 minutes between varying weight groups after the consumption of water. There were no
significant differences in hypothalamic response to water ingestion at 10 and 35 minutes (all p’s
> .12) in the lateral and medial hypothalamus (all p’s > .48).
Discussion
In this follow-up experiment we show that the hypothalamic response to sucrose shows
the most pronounced differences in blood flow between healthy weight and obese/overweight
groups at 10 minutes in the lateral hypothalamus, followed by maximum reduction in blood flow
at 35 minutes in the healthy weight group. These results are in line with prior studies including
those from Mastuda, Lui, Smeets, and Page and colleagues which show a prolonged decrease
in wither fMRI or ASL signal within the hypothalamus after glucose ingestion. Our results are in
line with earlier studies from Matsuda and Lui indicate the greatest reduction in blood flow from
7-15 minutes using fMRI41,42, as well as later studies from Smeets et al which detail a reduction
that is sustained for up 30 minutes43. Using pASL recent findings from Page and colleagues also
showed a significant reduction from baseline occurring at 15 minutes, sustaining for 55 minutes46.
In obese individuals, the reduction in blood flow in response to sucrose ingestion was attenuated,
and we observed slight increase in blood flow over the course of 35 minutes. These results are in

47
line with initial findings from Matsuda which indicate a blunted and delayed response of fMRI
signal in obese compared to healthy individuals, and findings from Smeets which also observe an
attenuation of the signal in individuals with type 2 diabetes using fMRI. In contrast to the prior
studies which mention a delay and attenuation of the hypothalamic blood flow in obese
individuals, we observed a slight increase in blood flow over the course of 35 minutes using pASL.
Interestingly, in overweight individuals (BMI > 25 to < 30) we observed a reduction in hypothalamic
response compared to obese individuals, and a significant attenuation of the hypothalamic
response to sucrose compared to healthy weight individuals at 10 minutes. This could indicate
that overweight individuals may display a delay in the reduction of response (i.e taking longer to
peak), while obese individuals display a transient increase in response to sucrose.
In order to confirm that these findings are specific to the ingestion of glucose, we examined
1) hypothalamic blood flow prior to the ingestion of sucrose (baseline) as well as 2) hypothalamic
blood flow in response to water ingestion, and found no differences in blood flow between all three
weight groups. Prior studies from Smeets and colleagues have demonstrated that only glucose,
not aspartame (sweet taste alone) or maltodextrin (carbohydrate alone), caused a decrease in
the hypothalamic fMRI signal44. Results indicate that neural signals are activated before and
during the absorption of nutrients into the bloodstream, crucially influencing the adaptive response
to food intake. Furthermore, Chapter 2 results showed that ingestions sweet taste without a
calorie (sucralose) did not elicit a hypothalamic response. Results from the water condition
experiment are in line with previous work suggesting this signal to be specific to sweet taste
coupled to a nutrient. Finally in order to confirm that these results are specific to the hypothalamus,
we examined cerebral blood flow across the whole brain prior to the ingestion of the drink and did
not find differences in blood flow across weight groups, suggesting these effects are not a result
of whole-brain specific vascular effects.
There are several study limitations to consider. First, we have included a modest sample
size of 20 to 30 individuals per weight status group, larger sample sizes may be helpful to detect

48
differences in hypothalamic blood flow across genders within each weight status category.
Furthermore, we were only able to evaluate changes in hypothalamic blood flow at 35 minutes. It
would be interesting to examine the signal change at the 1-hour mark in each weight status
category to investigate whether the signal in the overweight groups is truly delayed or just
attenuated, as well as to examine if the hypothalamic signal in healthy weight groups continue to
decrease. Finally, this study was conducted in a sample of young adults, it would be interesting
to see if the hypothalamic response to sucrose follows similar patterns in each weight status group
in children and older adults. A study by Ge and colleagues utilized pASL to show that overweight
and obese children (ages 7 to 11) had higher hypothalamic responses to glucose compared to
healthy weight children and adults (ages 19 to 24) and young adults (19-24)64. These results
suggest that hypothalamic response to sucrose ingestion could be influences by a number of
factors, and more studies as needed to examine associations between brain response to glucose
and metabolic dysfunction across the lifespan.
Conclusion
This study reveals significant differences in the hypothalamic response to sucrose
ingestion across weight groups, with healthy weight individuals showing the most substantial
decrease in blood flow, particularly in the lateral hypothalamus, within 10 to 35 minutes postingestion. Obese individuals exhibited a blunted and delayed response, with a slight increase in
blood flow, while overweight individuals showed an intermediate response. These findings
highlight the critical role of the hypothalamus in regulating feeding behaviors and emphasize the
need for further research to explore the temporal dynamics of hypothalamic responses in varying
weight groups, genders, and age groups.
Chapter 3 Figures and Tables
Table 10, [Chapter 3] Table 1. Participant Characteristics

49
N=73 Mean or Freq
(SD or Freq %)
Range
Age (years) 23.45 (3.96) 18.15-34.51
BMI (kg/m2) 27.29 (5.18) 19.18-40.28
BMI Group
Obese 23 (31.51%)
Overweight 24 (32.88%)
Healthy Weight 26 (35.62%)
Sex
F 42 (57.53%)
M 31 (42.47%)
Race & Ethnicity
Asian 22 (30.14%)
Hispanic 19 (25.02%)
Non-Hispanic Black 12 (16.44%)
Non-Hispanic White 20 (27.40%)
Table 11, Table 2. Change Hypothalamic Blood Flow at 10 minutes Stratified by BMI
Status
Contrast Beta SE t-stat p-value
Lateral Hypothalamus Overweight vs Obese 0.009 0.070 0.125 0.901
Healthy vs Obese -0.148 0.067 -2.214 0.030
Healthy vs Overweight -0.157 0.069 -2.261 0.027
Medial Hypothalamus Overweight vs Obese -0.064 0.090 -0.712 0.479
Healthy vs Obese -0.173 0.086 -2.008 0.049
Healthy vs Overweight -0.109 0.090 -1.220 0.227
Models covaried for Age, Sex, and Race/ethnicity.
Table 12, Table 3. Change Hypothalamic Blood Flow at 35 minutes Stratified by BMI
Status
Contrast Beta SE t-stat p-value
Lateral Hypothalamus Overweight vs Obese -0.100 0.076 -1.314 0.193
Healthy vs Obese -0.190 0.073 -2.598 0.012
Healthy vs Overweight -0.090 0.076 -1.185 0.240
Medial Hypothalamus Overweight vs Obese -0.091 0.088 -1.040 0.302
Healthy vs Obese -0.140 0.084 -1.660 0.102
Healthy vs Overweight -0.048 0.087 -0.555 0.581
Models covaried for Age, Sex, and Race/ethnicity.
Table 13, Table 4. Pre-drink Hypothalamic Blood Flow Stratified by BMI Status
Contrast Beta SE t-stat p-value
Sucrose Condition
Lateral Hypothalamus Overweight vs Obese 0.106 0.078 1.362 0.178
Healthy vs Obese 0.127 0.074 1.709 0.092
Healthy vs Overweight -0.106 0.078 -1.362 0.178
Medial Hypothalamus Overweight vs Obese 0.137 0.075 1.829 0.072
Healthy vs Obese 0.100 0.072 1.386 0.170
Healthy vs Overweight -0.037 0.075 -0.501 0.618
Models covaried for Age, Sex, and Race/ethnicity.
Table 14, Table 5. Mean Values of Hypothalamic and Whole Brain Cerebral Blood Flow at
Baseline Stratified by Weight Status
Region of Interest Weight group Mean SD

50
Mean hypothalamic blood flow
values at baseline
Lateral Hypothalamus Obese 1.088 0.250
Overweight 1.211 0.281
Healthy weight 1.237 0.231
Medial Hypothalamus Obese 0.802 0.275
Overweight 0.922 0.238
Healthy weight 0.919 0.227
Whole brain cerebral blood flow
at baseline
Whole Brain Obese 36.907 5.216
Overweight 35.906 3.886
Healthy weight 36.864 4.855
Figure 9, [Chapter 3] Figure 1: Change in Hypothalamic Response to Sucrose by BMI
Group

51
Chapter 4: Impact of Gestational Diabetes Exposure on Developing Brain
and Adiposity Trajectories: A 6 Year Longitudinal Study
Sandhya P. Chakravartti, MS1,2,3, Kay Jann, PhD4
, Brendan Angelo, MS2,3
, John R.
Monterosso1,6, PhD, Sandrah P. Eckel7
, Anny H. Xiang, PhD8
* and Kathleen A. Page, MD1,2,3*
1
Neuroscience Graduate Program, University of Southern California, Los Angeles, CA, USA.
2
Division of Endocrinology and Diabetes, Department of Medicine, Keck School of Medicine,
University of Southern California, Los Angeles, CA, United States.
3
Diabetes and Obesity Research Institute, Keck School of Medicine, University of Southern
California, Los Angeles, CA, United States.
4
Mark & Mary Stevens Neuroimaging & Informatics Institute, Keck School of Medicine,
University of Southern California, Los Angeles, CA, United States.
6
Department of Psychology, University Southern California, Los Angeles, CA, United States.
7
Department of Preventive Medicine, Keck School of Medicine, University of Southern
California, Los Angeles, CA, USA.
8
Department of Research and Evaluation, Kaiser Permanente Southern California, Pasadena,
CA 91101, United States
*Contributed equally as senior authors.

52
Introduction
The obesity epidemic is rising among children at an alarming rate. Evidence suggests that
the risk for diabetes and obesity begins even before birth, with the prenatal environment playing
a crucial role in shaping lifelong disease risk. Understanding the mechanisms associated with the
regulation of appetite behaviors is essential to breaking the intergenerational cycle of diabetes.
Gestational Diabetes Mellitus (GDM) is the first recognition of glucose intolerance during
pregnancy, and is estimated to affect 2-10% of all pregnancies each year47. A growing body of
evidence suggests a neural basis for increased obesity risk in GDM-exposed children61,62,117.
Evidence from animal and human studies suggests that in utero exposure to GDM increases
metabolic and cognitive dysfunction in offspring118,119, thereby raising their risk for obesity and
related metabolic and neurodevelopmental consequences later in life.120,121 However, the
longitudinal pattern of the impact of in utero diabetes exposure on brain volume and adiposity
trajectories as children transition from childhood to adolescence remains unknown.
Several cross-sectional neuroimaging studies have linked in utero exposure to GDM with
alterations in brain structure and function. A recent study by Lou and colleagues shows that
prenatal exposure to maternal diabetes is significantly associated with reduced cortical gray
matter volume in a sample of 9 and 10 year old children66. Results indicate that these brain
alterations mediate the increased risk of obesity and metabolic disorders in exposed children,
however no study till date has examined brain volume growth longitudinally in this population.
Cross-sectional studies of adiposity indicate that children exposed to GDM have an
increased risk of pediatric obesity, higher BMI, waist-to-hip circumference, and waist-to-height
ratio,53,56,57,65,122,123. A recent longitudinal study from Crume and colleagues showed that GDM
exposure leads to higher BMI trajectories, particularly between ages 10 and 13, suggesting that
GDM could differentially influence BMI growth rates especially during puberty56.
This study aims to map longitudinal growth patterns of cortical and subcortical brain
volume, and adiposity measures over 6 years in a diverse sample of 204 7- and 8-year-old

53
children (110 GDM exposed; 94 unexposed; 61% female). Children underwent brain MRIs up to
4 times over 6 years. Adiposity measurements were obtained yearly over 6 years. Adiposity
measures and FreeSurfer-derived brain volumes were compared between GDM-exposed and
unexposed groups using Generalized Additive Mixed Effects Models. Quantifying developmental
trajectories is essential to gain a more comprehensive understanding of how and when in utero
exposures influence postnatal development during this critical period of child development, and
may provide critical insights into optimal time periods to conduct effect treatment and prevention
strategies.
Methods
Participants and Study Design
The BrainChild Study is a prospective cohort study recruiting healthy children between 7-
11 years of age born at Kaiser Permanente Southern California hospital a large healthcare
organization with an integrated electronic medical record (EMR) system with documented
exposure to maternal GDM or no GDM exposure during the index pregnancy.65 Children born to
mothers with pre-gestational diabetes, using medications altering metabolism, or without a brain
scan were excluded.124 Exclusions also applied to those using medications that alter metabolism
(e.g., glucocorticoids) or had contraindications to MRI (e.g., permanent metal implants,
claustrophobia) or left-handedness. During the study visits children underwent brain MRIs up to
4 times over 6 years. Adiposity measurements were obtained yearly over 6 years. The study was
approved by the Institutional Review Boards of the University of Southern California (USC # HS14-00034) and KPSC (# 10282). Parents provided written informed consent, and children gave
written informed assent.
Exposure to Gestational Diabetes and Maternal pre-pregnancy BMI

54
GDM diagnosis was confirmed from electronic medical records (EMR) based on 1)
plasma glucose ≥200mg/dL from 50g glucose challenge test or 2) ≥2 elevated values on the 100g
or 75g OGTT (fasting ≥95mg/dL, 1 hour ≥180 mg/dL, 2 hour ≥155 mg/dL, 3 hour ≥140 mg/dL).125
Gestational age at GDM diagnosis was calculated using the date of the glucose test result that
met GDM diagnosis criteria, date of delivery, and gestational age at delivery per EMR. Maternal
pre-pregnancy BMI was calculated from documented maternal height and weight measurements
in the EMR obtained closest to date of last menstrual period. Maternal pre-pregnancy BMI (kg/m2)
was calculated from maternal height (cm) and weight (kg) measurements closest to last menstrual
period within 180 days, using EMR.
Anthropometric measures
We examined the trajectories of 4 measures of adiposity: weight, BMI, body fat %, and
waist circumference across children ages 7 to 16. The Tanita Body Composition Analyzer SC
331S (Tanita Corporation of America, Arlington Heights, IL) was used to measure weight (kg) and
body fat %. Waist circumference (cm) was measured using a measuring tape positioned in a
horizontal plane at the mid-point between the iliac crest and the lower costal margin in the
midaxillary line. Tanner stage of puberty was assessed either by a physical exam or by a
previously validated questionnaire126. Anthropometric measures and body composition Using the
Center for Disease Control (CDC) standards, each child’s BMI z-score (BMIZ) and BMI percentile
were calculated from their height and weight. Children with BMI percentiles 5th to < 85th were
classified as healthy weight; BMI percentiles 85th to <95th percentile classified as over- weight,
and BMI percentiles ≥95th classified as obesity. Body fat % was measured using bioelectrical
impedance analysis (BIA) with the Tanita scale using the frequency current, 50 kHz, 90 μA.
Imaging Data Acquisition

55
Children came in for a study visit at the Dana & David Dornsife Cognitive Neuroimaging
Center at USC, where a brain MRI was collected. After a mock scanner training session, magnetic
resonance imaging (MRI) was performed using a Siemens MAGNETOM Prismafit 3 Tesla MRI
scanner (Siemens Medical Systems) with a 20-channel phased array coil. A high-resolution
magnetic resonance imaging scan was acquired using a T1-weighted three-dimensional
magnetization prepared rapid gradient echo (MP-RAGE) sequence with the parameters: 256 ×
256 × 176-matrix size with 1 × 1 × 1-mm3 resolution; inversion time = 900 ms; repetition time (TR)
= 1950 ms; echo time (TE) = 2.26 ms; flip angle = 90◦; Total scan duration was 4 min and 14 s.
MRI Structural Analyses Processing
Image processing was performed with the FreeSurfer image analysis suite (http://
surfer.nmr.mgh.harvard.edu/). Details of the procedure have been previously outlined127-130.
FreeSurfer employs a spherical atlas to register the cortical surface across subjects, leveraging
individual cortical folding patterns to match cortical geometry. The software calculates vertex-wise
gray matter volume (GMV), surface area (SA) at the gray matter-white matter boundary, mean
cortical thickness (CT), and curvature for each vertex, with higher curvature values indicating
deeper sulci. Surface-based analyses are smoothed with a Gaussian kernel of 15 mm full width
at half maximum. Cortical reconstruction and volumetric segmentation were optimized for
longitudinal processing131,132. In this process, unbiased within-subject template space and image
were created, using robust, inverse consistent registration131. The individual template was utilized
for several preprocessing steps including: skull stripping, Talairach transforms, atlas registration,
spherical surface maps and parcellations. This results in significantly improved statistical power
and reliability. All segmented regions were visually examined by a trained reviewer with particular
attention to the consistency of segmentations across time points for each participant using the
ENIGMA Consortium Cortical Quality Control Protocol 2.0 (http://enigma.usc.edu/). For the
primary analysis, the following brain regions were examined: total gray matter volume. And

56
subcortical gray matter volume based on prior literature linking GDM and obesity to alterations in
these regions66,133,134. Total gray matter includes both surface-based volume and voxel counts,
while subcortical gray matter is calculated using voxel based methods127,129.
Statistical Analysis
GDM exposure was modeled as yes or no exposure to GDM. Age was grand-mean
centered at 10.5 years. Descriptive analyses were conducted to characterize the study population
(Table 1). Because 90% of children were Tanner stage 1, Tanner stage was not adjusted in the
regression models at baseline. Conditional Intraclass correlations coefficients (ICC) were
calculated from linear mixed effects models with a random intercept for participant to quantify the
consistency of longitudinal adiposity and brain measurements across the 6-year study period.
General linear regression models were used to assess the associations between 1) GDM
exposure on adiposity and brain volumes, and 2) adiposity measures on brain volumes at baseline
with covariates of intracranial volume, sex, mean centered age, and tanner stage (since we were
including children of all ages in the analysis). Adiposity and brain volume trajectories were plotted
as a function of age and were modeled fitting generalized additive mixed models (GAMMs), with
a participant-level random intercept and a random slope for mean centered age using restricted
maximum likelihood 135,136. Random effects allow for each participant’s starting value to vary from
the population average (intercept) and the longitudinal trajectory to vary from the population
average longitudinal trajectory (slope). This method does not assume a specific parametric shape
(e.g., linear or quadratic) for the trajectory, allowing GAMM to find the best-fit trajectory by
considering both cross-sectional and longitudinal data aspects. This avoids the weaknesses of
global polynomials, where trajectory endpoints can be overly influenced by factors like the
sampled age range137. Smooth functions of age were fitted using a penalized regression spline
with k parameters specifying the stiffness of the model curves set to five.

57
We then evaluated whether age trajectories were influenced by GDM exposure, adjusting
for sex and pre-pregnancy BMI, by fitting a model with a common smooth surface for age and
covariates, and separate surfaces for the GDM exposure levels using the 'by' variable in the
gamm() function. This function which allowed to test whether the smooth for a given group was
different from that of the overall population smooth, and allows for potentially only one complex
smooth surface to be estimated136. A first order penalty was imposed on the “difference” smooths
for identifiability, encouraging shrinkage to the mean. Model fit was assessed using likelihood
ratio test (LRT) comparing 2 nested models: the full model (term of interested estimated) and a
null model (dropping the term of interest). In addition to the primary model fit metrics, secondary
assessments such as the standard deviation of the random intercept were examined to assess
whether the interaction term reduced the variability in the model's predictions. Predicted geometric
means and their 95% confidence intervals were calculated using the predict function from the
respective GAMM models. To quantify the relative change in predicted means over time, percent
differences between predicted means at different ages were computed by taking the absolute
difference between the two predicted means, dividing this difference by their average, and then
multiplying by 100. Similar methodologies have been previously applied to investigate trajectories
of brain volumes and markers of airway inflammation in cohorts 8 year old children followed
longitudinally for 7 years 138,139.
Finally, Linear mixed-effects models (LMER) with a random effect of subject were used to
assess associations between measures of adiposity and brain volumes across the whole cohort,
with covariates of mean centered age, sex, intracranial volume, and tanner stage as well as
random intercept of subject. Alpha levels of 0.05 were used for all statistical tests, with results
between 0.05 and 0.1 noted as a statistical trend. All statistical analyses were performed used
Rstudio (version 2023.06.1). GAMM models were run using “mgcv” package. Model fits were
examined using the r2beta function from the “r2glmm” package.

58
Results
A total of 204 (110 GDM exposed children ages 7-11 years and 94 age matched controls)
participated in the BrainChild Study. Clinical characteristics of the study participants are included
in (Table 1). Children underwent brain MRIs up to 4 times over 6 years and adiposity
measurements were obtained yearly over 6 years. The total longitudinal sample consisted of 204
scans from 154 participants, and 568 adiposity measurements from 204 participants. Of these
subjects 83 were re-recruited and returned for a second Brain assessment 2 years later, 35
returned for a 3rd Brain scan 4 years later, and 8 returned for a 4th brain scan 6 years later (Table
2). Adiposity measurements were assessed for 97 participants 2 years later, 47 participants 4
years later, and 9 participants 6 years later (Table 2).
Larger measures of adiposity, but not brain volumes associated with GDM exposure at
baseline
At the initial study visit mean age for all child participants was 8.62 (range: 7.19 -11.44). We found
significant associations between exposure to GDM in utero and adiposity measurements of BMI
(β = 1.21, p = .042) and waist circumference (β = 3.43, p = .023) after adjusting for mean centered
child age, sex, and maternal pre-pregnancy BMI (Table 3). Associations remained trending
significant for BMI (β = 1.20, p = .079) and significant for waist circumference (β = 3.57, p = .04)
after additionally covarying for tanner stage. Associations between GDM and weight and body
fat percentage showed a trend to significance (p = .069; p = .078) after covarying for age, sex,
and maternal pre-pregnancy BMI (Table 3). At baseline, all measurements of brain volume
including total cortex volume (p< .34), total gray matter volume (p< .63), did not vary by GDM
exposure (Table 4).
Brain volumes associated with measures of adiposity at baseline

59
At baseline higher measurements adiposity including weight (β = -885.1; p <.0004), BMI (β = -
2034.1; p < .0005), body fat (β = -1063.7; p < .0001), and waist circumference (β = -785.3; p <
.0005) were associated with lower total gray matter volume after adjusting for total intracranial
volume, child age, sex, and tanner stage (Table 5). Higher measurements adiposity including
weight (β = -66.64; p = .026), BMI (β = -153.2; p < .03), body fat (β = -90; p < .007), and waist
circumference (β = -64.7; p = .02) were also associated with lower total subcortical gray matter
volume after adjusting for total intracranial volume, child age, sex (Table 5).
Measures of adiposity and brain volume vary across time
For BMI ICC was .722, and for gray matter volume ICC was .87 suggesting moderately strong
between-participant correlation of both adiposity and brain measurements. All adiposity
measurements, including weight, BMI, body fat, and waist circumference, increased from baseline
through year 6, with significant non-linear trajectories observed for each measure (all p's <
.00002) (Figure 1; Table 6). Estimated geometric mean of BMI increased from 18.8 kg/m2 (CI:
18.1, 19.57) at age 4 to 23.42 (CI: 22.44, 24.41) at age 12, to 25.63 (CI: 22.44, 24.41) at age 16
(Table 7).
There was also a significant non-linear decrease of cortical volume (edf = 3.36; p<.00002)
from 8 to 16 (Figure 2; Table 8). Estimated geometric mean was 717959.2 mm3 (CI: 707,667.09,
728251.33) at age 8, with an average of 3% (CI: 5.96%, 0.07%) decrease in volume by age 12,
and a 10.2% decrease in volume by age 16 (CI: 13.93%, 6.31%) (Table 7). We observed a slight
non-linear increase in subcortical volume (edf = 1.52; p = .009) over time after adjusting for
gender, across all children (Figure 2). On average mean volume of subcortical structures was
59,952.7 mm3 (CI: 59042.87, 60862.58) at age 8 and increased by 2.3% by age 16 (CI: 2.43%,
7.3%) (Table 7).
Measures of Adiposity and Brain Volumes Vary by GDM exposure across time

60
There was a significant interaction of GDM and age on adiposity outcomes of BMI (interaction p
= .03), and body fat percentage (interaction p = .02). Interactions between GDM exposure and
age on weight and waist circumference were trending (interaction p = .09; p = .08), however there
was a significant effect of GDM on higher waist circumference across time (p =.042; Table 6) with
GDM participants having an average of 3.26 cm larger waist circumference compared to controls
across time. For example, GDM exposed subjects showing higher BMI measures compared to
control participants who had a non-linear and lower growth trajectory (edf = 3.17, p = .002) (e.g.,
1.81 kg/m2 higher at age 12;) (Table 6). By age 16 there was a 35.14% (CI: 20.25%, 52.18%)
increase in BMI in GDM exposed children compared to a 23.95% (CI: 7.24%, 43.6%) increase in
control children (Figure 3; Tables 9 and 10).
Interactions between GDM exposure and age on gray matter volume were not significant.
However, volume of subcortical trajectories varied by GDM exposure (interaction p = .007) after
adjusting for gender, maternal pre-pregnancy BMI. Control participants start out higher, with a
higher non-linear trajectory (edf = 1.87, p = .03; Table 8). Specifically subcortical volume in
controls increase by 4.31% (CI: 4.19%, 13.58%) from age 8 to 16, while GDM exposed children
increase subcortical volume by 1.38% (CI: 2.47%, 5.38%) from age 8 to 16 (Figure 4; Tables 9
and 10).
Higher Adiposity is associated with lower subcortical volumes across all ages
Across the entire sample higher weight showed a trending association with lower volumes of total
gray matter volumes after correcting for adjusting for child age, total intracranial volume, gender,
and tanner stage (β = -251.71, p < .06). BMI, body fat and waist were not associated with total
gray matter volume across all ages (all p’s > .19). Higher measures of adiposity including weight
(β = -42.07, p = .015), BMI (β = -105.2, p = .02), body fat percentage (β = -54.8, p = .016), and
waist circumference (β = -46.6, p = .009) were associated lower subcortical volumes after
adjusting for age, intracranial volume, gender, and tanner stage (Table 11).

61
Discussion
Emerging research indicates a neural basis of diabetes58. In light of recent evidence
showing alterations of brain structure and function in adults and children with type I and II
diabetes140-143, we sought to investigate whether in utero exposure, impacted brain and adiposity
trajectories of preadolescent children as they transitioned from childhood to adolescence.
Longitudinal structural MRI scans examining the transition from childhood to adolescence
in typically developing youth highlight protracted brain maturation marked by non-linear inverted
U-shaped trajectories of grey matter and subcortical volumes peaking in early childhood and
decreasing throughout adolescence144. The decreases in cortical gray matter volume from ages
8 – 16, are in line with these prior findings. A handful of cross-sectional neuroimaging studies
have linked in utero exposure to GDM with alterations in brain structure and function, including
alterations in: cortical66, hippocampal volume and surface area68, and white matter integrity in
sensorimotor regions67. Our results, extend prior findings to show that there is a sustained
reduction in growth in subcortical volumes over a 6 years period in GDM exposed children
compared to unexposed children. The sustained reduced subcortical brain volume growth in the
GDM group aligns with 2 recent longitudinal neuroimaging studies in children. The first study from
Jiang and colleagues showed that children with obesity have lower subcortical gray matter
volumes in regions such as the hippocampus, amygdala, caudate, and thalamus over 2 years
compared to normal-weight children145 . The second from Adise and colleagues, showed that
greater increases in BMI over 2 years were associated with lower subcortical volumes in
preadolescent girls146. Obesity triggers the release of cytokines and chemokines that cross the
blood-brain barrier, causing microglia to release proinflammatory cytokines and leading to
neuronal loss through apoptotic signaling147. In GDM-exposed children, excess adiposity may
disrupt synaptic pruning and gray matter maturation during puberty, delaying the development of
subcortical structures.

62
A few studies have begun to map adiposity trajectories in children exposed to gestational
diabetes mellitus (GDM) compared to those unexposed. A study by Silverman and colleagues
found that increased BMI in GDM-exposed children starts to appear after age two55. Crume et al.
extended these findings, showing that GDM-exposed children have higher BMI trajectories up to
age 14, primarily due to a significantly higher growth rate velocity between ages 10 and 1356.
These results support and extend findings to show that GDM exposed children display higher BMI
trajectories which continue to diverge throughout puberty.
There are of course several limitations that need to be taken into consideration. Structural
imaging at its current resolution cannot reveal details of the underlying cellular morphology.
Future imaging studies should employ a multimodal approach including diffusion weighted
imaging (DWI), in order to better elucidate macro and microstructural differences that may be
contributing to the volumetric differences observed. Recent studies suggest the subcortical
volumes to show region-specific regions undergoing distinct sex- specific developmental
trajectories144,148. It would be of interest to further investigate how GDM exposure affects
individual subcortical structures over time in boys and girls separately.
Sub-cortical brain structures such as the amygdala, hippocampus, caudate, and thalamus,
play crucial roles in food intake control and body weight regulation149-151. Along with reduced gray
matter volume, studies have shown and greater activation in these regions in response to food
cues150. Future research should explore how GDM affects and alters reward and motivation
behaviors longitudinally during the transition from childhood to adulthood to better understand the
neurobehavioral mechanisms associated with GDM during this critical period of development.
In conclusion, detectable longitudinal changes adiposity measures persist over time in
children exposed to gestational diabetes in utero as they transition from childhood to adolescence.
These results demonstrate that GDM has widespread effects on the growth subcortical gray
matter in exposed children, and presents a possible mechanism by which in utero exposure to
maternal metabolic disorders could contribute to an increased risk of obesity later in life.

63
Chapter 4 Figures and Tables
Table 15, [Chapter 4] Table 1. Participant Characteristics at Baseline
GDM (n = 110) Control (n= 94)
Age 8.53(1) 8.74(1)
Age range 7.2-11.4 7.3-11.3
Sex
Male 42 (38.2) 37 (39.4)
Female 68 (68.8) 57 (60.6)
Tanner Stage
1 99 (90) 84 (89.4)
2 8 (7.3) 7 (7.4)
3 2 (1.8) 2 (2.1)
4 1 (.9) 1 (1.1)
Race
Hispanic 68 (61.8) 48 (51.1)
Black 11 (10) 12 (12.8)
Non-Hispanic White 14 (12.7) 18 (19.1)
Other 17 (15.5) 16 (17)
Maternal Pre-pregnancy BMI
Normal 22 (20) 23 (24.5)
Overweight 33 (30) 29 (30.8)
Obesity 55 (50) 42 (44.7)
Table 16, Table 2. Participant Counts by Study Visit
Baseline Year 2 Year 4 Year 6
Adiposity 204 97 47 9
Brain volume 154 83 35 8
Table indicates how many unique participants have adiposity measurements (weight), or brain volume measurements (cortical
volume) separated by study visit. Year 2 indicates data point 2 years after the initial (Baseline) study visit. Year 4 indicates 4 years
after initial visit, and Year 6 indicates 6 years after initial study visit.
Table 17, Table 3. Associations between GDM exposure and Adiposity Measurements at
Baseline
Outcome Beta coefficient SE P-value
Weight (kg)
Model 0 2.05 1.70 0.228
Model 1 3.41 1.44 0.019
Model 2 2.62 1.43 0.068
BMI (kg/m2
)
Model 0 1.24 0.62 0.047
Model 1 1.53 0.60 0.011
Model 2 1.21 0.59 0.042
Body fat (%)
Model 0 2.60 1.28 0.043
Model 1 2.90 1.27 0.024
Model 2 2.23 1.26 0.077
Waist (cm)
Model 0 3.45 1.67 0.040
Model 1 4.46 1.54 0.004
Model 2 3.43 1.50 0.023
Model 0: no adjustment; Model 1: adjusted for child age and sex; Model 2: adjusted for mean centered child age, sex, and maternal
pre-pregnancy BMI

64
N = 204; GDM exposure reference group is unexposed; SE = Standard Error
Table 18, Table 4. Associations between GDM Exposure and Brain Volume Measurements
at Baseline
Outcome Beta coefficient SE P-value
Total Gray Matter
(mm3
)
Model 0 -4055.55 5321.72 0.447
Model 1 -5104.41 4997.58 0.309
Model 2 -4784.08 5014.14 0.342
Subcortical
Volume (mm3
)
Model 0 -397.35 595.17 0.505
Model 1 -364.19 599.00 0.544
Model 2 -284.99 595.50 0.633
Model 0: adjusted for total intracranial volume (ICV); Model 1: adjusted for child age, sex, and ICV; Model 2: adjusted for mean
centered child age, sex, maternal pre-pregnancy BMI and ICV.
N = 147; GDM exposure reference group is unexposed; SE = Standard Error
Table 19, Table 5. Associations between Adiposity and Brain Volume Measurements at
Baseline
Outcome Independent Variable Beta
coefficient
SE P-value
Total gray
matter volume
Weight (kg)
Model 0 -1176.23 209.90 < 0.00001
Model 1 -885.15 241.64 0.0004
BMI (kg/m2
)
Model 0 -2630.65 565.75 < 0.00001
Model 1 -2034.07 566.50 0.0005
Body fat (%)
Model 0 -1232.79 271.80 <0.00001
Model 1 -1063.68 265.14 0.0001
Waist (cm)
Model 0 -1054.18 213.39 <0.00001
Model 1 -785.28 221.77 0.0005
Subcortical
volume
Weight (kg)
Model 0 -58.75 25.42 0.0222
Model 1 -66.64 29.70 0.0264
BMI (kg/m2
)
Model 0 -160.76 66.47 0.0168
Model 1 -153.21 69.56 0.0292
Body fat (%)
Model 0 -95.34 31.49 0.0029
Model 1 -89.96 32.58 0.0065
Waist (cm)
Model 0 -64.46 25.22 0.0116
Model 1 -64.72 27.12 0.0183
Model 0: adjusted for total intracranial volume (ICV); Model 1: adjusted for mean centered child age, sex, and ICV
N = 147; SE = Standard Error

65
Table 20, Table 6. Summary of GAMM Models: Impact of GDM on Adiposity Trajectories
Adiposity
outcome
n-obs Model Model/parameter1 Regression
coefficient
EDF† p-value
Weight (kg) 574
Age only S(Age)
intercept 45.49 <0.00002
s(Age) 4.99 <0.00002
Age with
covariates
S(Age)+pre-pregnancy BMI + sex
intercept 39.36 <0.00002
Sex: Female -0.29 0.820
Pre-pregnancy BMI 0.21 0.019
s(Age) 5.01 <0.00002
Full model S(Age*gdm) +gdm +pre-pregnancy BMI+
sex‡
intercept 38.81 <.000002
Sex: Female -0.31 0.8087
Gdm: Exposed 2.21 0.1858
Pre-pregnancy BMI 0.19 0.0344
s(Age) 4.40 <.000002
s(age: Unexposed) 2.55 0.0377
s(age: Exposed) 0.00 0.6210
BMI
(kg/m2)
574
Age only S(Age)
intercept 21.26 <0.00002
s(Age) 3.39 <0.00002
Age with
covariates
S(Age)+pre-pregnancy BMI + sex
intercept 17.66 <0.00002
Sex: Female -0.77 0.189384
Pre-pregnancy BMI 0.13 0.000966
s(Age) 3.45 <0.00002
Full model S(Age*gdm) +pre-pregnancy BMI+ sex‡
intercept 17.51 <.000002
Sex: Female -0.76 0.1916
Gdm: Exposed 1.08 0.1080
Pre-pregnancy BMI 0.12 0.0032
s(Age) 2.15 <.000002
s(age: Unexposed) 3.18 0.002
s(age: Exposed) 0.00 0.445
Body fat
(%)
568
Age only S(Age)
intercept 27.55 <.000002
s(Age) 2.67 <.000002
Age with
covariates
S(Age)+pre- pregnancy BMI + sex
intercept 16.55 <.000001
Sex: Female 0.27 0.836

66
Pre-pregnancy BMI 0.35 <0.00006
s(Age) 2.76 <.000002
Full Model S(Age*gdm) +pre-pregnancy BMI + sex‡
intercept 16.08 <.000002
Sex: Female 0.26 0.8373
Gdm: Exposed 2.39 0.0729
Pre-pregnancy BMI 0.33 0.0002
s(Age) 3.08 0.0012
s(age: Unexposed) 0.23 0.1197
s(age: Exposed) 2.94 0.0032
Waist (cm) 572
Age only S(Age)
intercept 72.89 <.000002
s(Age) 3.01 <.000002
Age with
covariates
S(Age)+pre-pregnancy BMI + sex
intercept 58.72 <.000002
Sex: Female -0.59 0.703
Pre-pregnancy BMI 0.48 <.000009
s(Age) 3.08 <.000002
Full Model S(Age*gdm) +pre-pregnancy BMI+ sex‡
intercept 57.95 <.000002
Sex Female -0.66 0.667
Gdm: Exposed 3.26 0.042
Pre-pregnancy BMI 0.44 0.000033
s(age) 2.76 <.000002
s(age: Unexposed) 0.00 0.793
s(age: Exposed) 1.32 0.147
Summary of generalized additive mixed models that model longitudinal adiposity trajectories and smooth functions of time varying
age and time-constant covariates of sex and prepregnancy bmi, plus a participant-level random intercept and random slope on age. 1
Smooth terms definitions: s(): penalized regression spline. †
Effective degrees of freedom (EDF) for the smooth term; a value of 1
indicates a linear effect, higher values indicate a higher degree of non-linearity. ‡
K=5 specified in the s(age, by=sex, m=1, k=5) term
for improved model convergence. For covariate sex, reference variable is male. For covariate gdm, reference variable is
unexposed/control.
Table 21, Table 7. Predicted Geometric Means Measuring Effect of Age on Outcomes
using GAMM Models
Variable Age 8 CI Age 12 Age 12 CI
Percent
difference
8 to 12
Percent
difference
CI 8 to 12 Age 16 Age 16 CI
Percent
difference
8 to 12
Percent
difference CI
8 to 16
Weight (kg) 33.73 (32.01,
35.44) 58.99
(56.28,
61.69) 54.49
(42.91,
67.24) 73.54
(68.67,
78.41) 74.24 (58.39, 92.17)
BMI
(kg/m2)
18.84 (18.1,
19.57) 23.42
(22.44,
24.41) 21.71
(13.05,
31.1) 25.63
(23.92,
27.34) 30.54 (18.52, 43.96)
Body Fat
(%)
25.94 (24.36,
27.53) 29.40 (27.5, 31.3) 12.48 (0.11, 26.75) 30.35
(27.19,
33.52) 15.66 (1.1, 35.52)
Waist (cm) 65.17 (63.26,
67.09) 80.44
(77.98,
82.9) 20.97
(14.52,
27.82) 91.16 (86.82, 95.5) 33.24 (24.27, 42.97)
Cortical
Volume
(mm3)
717959.21
(707667.09,
728251.33) 696769.77
(685418.07,
708121.46) 3.00 (5.96, 0.07) 648318.46
(631184.87,
665452.06) 10.19 (13.93, 6.31)
Subcortical
volume
(mm3)
59952.72
(59042.87,
60862.58) 61181.27
(60144.76,
62217.78) 2.03 (1.17, 5.33) 61359.41
(59352.04,
63366.78) 2.32 (2.43, 7.3)

67
Predicted geometric mean values across all participants at ages 8, 12, and 15 as well as the percent differences from ages 8 to 12
and from ages 8 to 16. Predicted values are from “Age with Covariates” models from Table 5. For covariates sex, female variable
was held constant, and pre-pregnancy BMI was the mean value of cohort: 30.46961. CI = Confidence Interval
Calculation: To quantify the relative change in predicted means over time, percent differences between predicted means at different
ages were computed by taking the absolute difference between the two means, dividing by their average, and multiplying by 100.
Confidence intervals for these percent differences were calculated as follows: The lower bound was determined by the absolute
difference between the lower bound of the second CI and the upper bound of the first CI, divided by the average of the lower
bounds, and multiplied by 100. The upper bound was calculated similarly, using the upper bound of the second CI and the lower
bound of the first CI, divided by the average of the upper bounds, and multiplied by 100.
Table 22, Table 8. Summary of GAMM Models: Impact of GDM on Gray Matter Trajectories
Adiposity
outcome
n-obs Model Model/parameter1 Regression
coefficient
EDF† p-value
Gray Matter
Volume (mm3)
280
Age only S(Age)
intercept 736012.00 <.000002
s(Age) 3.56 <.000002
Age with
covariates
S(Age)+prepregnancybmi + sex
intercept 773447.36 <.000002
Sex: Female -63208.08 <.000002
Pre-pregnancy BMI -31.49 0.956
s(Age) 3.54 <.000002
Full model S(Age*gdm) +gdm
+prepregnancybmi+ sex‡
intercept 773418.62 <.000002
Sex: Female -63215.53 <.000002
Gdm: Exposed 80.97 0.992
Pre-pregnancy BMI -31.93 0.956
s(Age) 3.54 <.000002
s(age: Unexposed) 0.00 0.800
s(age: Exposed) 0.00 0.938
Subcortical gray
matter volume
(mm3)
280
Age only S(Age)
intercept 62430.30 <.000002
s(Age) 2.01 0.0032
Age with
covariates
S(Age)+prepregnancybmi + sex
intercept 67067.52 <.000002
Sex: Female -4668.96 <.000002
Pre-pregnancy BMI -64.37 .193
s(Age) 2.02 0.00239
Full model S(Age*gdm) +gdm
+prepregnancybmi+ sex‡
intercept 67146.34 <.000002
Sex: Female -4640.31 <.000002
Gdm: Exposed -396.41 0.564
Pre-pregnancy BMI -60.36 0.223
s(Age) 1.31 0.0844

68
s(age: Unexposed) 1.87 0.0322
s(age: Exposed) 0.00 0.9371
Summary of generalized additive mixed models that model longitudinal adiposity trajectories and smooth functions of time varying
age and time-constant covariates of sex and prepregnancy bmi, plus a participant-level random intercept and random slope on age. 1
Smooth terms definitions: s(): penalized regression spline. †
Effective degrees of freedom (EDF) for the smooth term; a value of 1
indicates a linear effect, higher values indicate a higher degree of non-linearity. ‡
K=5 specified in the s(age, by=sex, m=1, k=5) term
for improved model convergence. For covariate sex, reference variable is male. For covariate gdm, reference variable is
unexposed/control.
Table 23, Table 9. Predicted Geometric Mean of Age on Outcomes in GDM Exposed
Children using GAMM Models
Variable Age 8 CI Age 12 Age 12 CI
Percent
difference
8 to 12
Percent
difference
CI 8 to 12 Age 16 Age 16 CI
Percent
difference
8 to 12
Percent
difference CI
8 to 16
Weight (kg)
35.66
(33.11,
38.21) 60.25
(56.26,
64.23) 51.28
(35.25,
69.64) 81.03 (74.33, 87.73) 77.76 (57.36, 101.67)
BMI
(kg/m2) 19.65
(18.56,
20.74) 24.02
(22.59,
25.46) 20.01
(7.99,
33.51) 28.03 (25.9, 30.16) 35.14 (20.25, 52.18)
Body Fat
(%) 26.88
(24.62,
29.14) 30.10
(27.41,
32.78) 11.29 (5.6, 31.38) 34.29 (30.23, 38.35) 24.23 (3.22, 50.09)
Waist (cm)
66.81
(63.95,
69.66) 82.69
(79.03,
86.36) 21.25
(12.01,
31.34) 96.65 (90.81, 102.5) 36.52 (24.57, 49.81)
Cortical
Volume
(mm3) 722135.43
(708396.5,
735874.37) 700451.23
(685072.09,
715830.37) 3.05 (7, 1.07) 648381.75
(625455.64,
671307.86) 10.76 (15.69, 5.56)
Subcortical
volume
(mm3) 60040.53
(58997.96,
61083.09) 60761.29
(59686.34,
61836.24) 1.19 (2.27, 4.78) 60874.92
(59563.71,
62186.13 ) 1.38 (2.47 , 5.38 )
Predicted geometric mean values in GDM exposed participants at ages 8, 12, and 15 as well as the percent differences from ages 8
to 12 and from ages 8 to 16. Predicted values are from “Age with Covariates” models from Table 5. For covariates sex, female
variable was held constant, and pre-pregnancy BMI was the mean value of cohort: 30.46961. CI = Confidence Interval
Table 24, Table 10. Predicted Geometric Mean of Age on Outcomes in Unexposed
Children using GAMM Models
Variable Age 8 CI Age 12 Age 12 CI
Percent
difference
8 to 12
Percent
difference CI
8 to 12 Age 16 Age 16 CI
Percent
difference
8 to 12
Percent
difference CI
8 to 16
Weight (kg)
31.49
(29.46,
33.52) 56.77
(53.33,
60.22) 57.30
(42.26,
74.33) 68.25 (62.33, 74.17) 73.71 31.49
BMI
(kg/m2) 17.94
(17.01,
18.88) 22.81
(21.51,
24.12) 23.90
(12.24,
36.92) 22.82 (20.47, 25.18 23.95 17.94
Body Fat
(%) 24.40
(22.19,
26.61) 28.18
(25.44,
30.92) 14.38 (4.08, 36.67) 25.05 (20.33, 29.77) 2.62 24.40
Waist (cm)
63.04
(60.62,
65.47) 77.36 (74.2, 80.51) 20.39 (11.97, 29.51) 86.66 (81.27, 92.04) 31.55 63.04
Cortical
Volume
(mm3) 712101.55
(696621.7,
727581.41) 691642.33
(675165.69,
708118.96) 2.91 (7.3, 1.68) 652434.97
(628966.05,
675903.89) 8.75 712101.55
Subcortical
volume
(mm3) 59940.97
(58273.56,
61608.38) 61595.26
(59782.73,
63407.79) 2.72 (2.92, 8.7) 62583.69
(58933.16,
66234.23) 4.31 59940.97
Predicted geometric mean values across unexposed participants at ages 8, 12, and 15 as well as the percent differences from ages
8 to 12 and from ages 8 to 16. Predicted values are from “Age with Covariates” models from Table 5. For covariates sex, female
variable was held constant, and pre-pregnancy BMI was the mean value of cohort: 30.46961. CI = Confidence Interval
Table 25, Table 11. Associations between Adiposity Measurements and Brain Volume
Measurements across Time
Outcome Independent Variable Beta coefficient SE P-value
Total gray matter
volume (mm3
)
Weight (kg)
Model 0 -410.63 130.19 0.002
Model 1 -363.30 127.13 0.005
Model 2 -251.71 132.74 0.059
BMI (kg/m2
)
Model 0 -749.75 362.05 0.039
Model 1 -651.78 353.03 0.066
Model 2 -462.44 353.25 0.192

69
Body fat (%)
Model 0 -382.83 185.97 0.041
Model 1 -292.97 182.52 0.110
Model 2 -200.79 181.68 0.270
Waist (cm)
Model 0 -261.82 145.41 0.073
Model 1 -203.86 142.14 0.153
Model 2 -133.82 141.81 0.346
Subcortical volume
(mm3
)
Weight (kg)
Model 0 -38.60 16.00 0.017
Model 1 -37.03 16.13 0.022
Model 2 -42.07 17.16 0.015
BMI (kg/m2
)
Model 0 -100.24 42.92 0.020
Model 1 -96.82 43.12 0.026
Model 2 -105.18 44.57 0.019
Body fat (%)
Model 0 -52.87 21.70 0.016
Model 1 -50.46 21.96 0.022
Model 2 -54.78 22.47 0.016
Waist (cm)
Model 0 -44.01 17.07 0.011
Model 1 -42.45 17.19 0.014
Model 2 -46.55 17.67 0.009
Models are linear mixed effects models: Model 0: adjusted for age and ICV model 1: adjusted for age, ICV, and sex, Model2 are
adjusted for mean centered child age, sex, tanner stage, and ICV and a random intercept of subject.

70
Figure 10, [Chapter 4] Figure 1: Adiposity Trajectories by Age
Figure 1: Population average trajectories with a 95% confidence interval of the following adiposity trajectories: weight, BMI, body fat
%, and waist circumference (all p < .000001). Results are modeled as a smooth function of age. Figures were created using GAMM
models using a smooth term for child age, random intercept for subject and random slope for child age. N = 204; 568 observations

71
Figure 11, Figure 2: Brain Trajectories by Age

72
Figure 2: Population average trajectories with a 95% confidence interval of the following adiposity trajectories: weight, BMI, body fat
%, and waist circumference. Results are modeled as a smooth function of age. Figures were created using GAMM models using a
smooth term for child age, random intercept for subject and random slope for child age. 280 observations
Figure 12, Figure 3: Adiposity trajectories stratified by GDM exposure
Figure 3: Population averages trajectories stratified by GDM exposure (red) or control groups (blue) of the following adiposity
trajectories: weight (p = .037), BMI (p = .002), body fat % (p = .003), and waist circumference (p = .04 for overall effect of GDM
exposure on outcome across all ages). Results are modeled as a smooth function of age. Figures were created using GAMM
models using a smooth term for child age, random intercept for subject and random slope for child age. 280 observations.

73
Figure 13, Figure 4: Total Subcortical Brain Volume trajectories stratified by GDM
exposure
Figure 4: Population averages trajectories stratified by GDM exposure (red) or control groups (blue) for subcortical brain volume
trajectories (p =.03). Results are modeled as a smooth function of age. Figures were created using GAMM models using a smooth
term for child age, random intercept for subject and random slope for child age. 280 observations.
(mm3)

74
General Discussion
This dissertation aimed to characterize neurophenotypes related to appetite regulation and
obesity, with a focus on how prenatal factors contribute to the development of obesity in children.
Towards this goal, we conducted three studies to 1) examine the neurohumoral and 2) temporal
mechanisms of hypothalamic signaling mechanisms in adults, and 3) examine the influence of in
utero exposure to gestational diabetes, to predict adiposity and brain volume trajectories in a
group of 7–8-year-old children as they transition from childhood to adolescence.
The first study examined the effects of sucralose ingestion, a non-caloric sweetener, in
comparison to caloric sugar (sucrose) and water on hypothalamic, glycemic, and hunger
responses by utilizing a combination pulsed arterial spin labeling imaging and functional
connectivity, neuroimaging techniques, blood sampling and appetite assessments. Our results
demonstrated that sucralose when compare to both sucrose and water significantly altered the
hypothalamic responses and its functional connectivity to brain regions involved in somatosensory
processing. Additionally, sucrose but not sucrose led to increases in peripheral glucose that
corresponded with decreases in hypothalamic blood flow, and decreases in hunger. These
findings highlight the distinct differences in how non-caloric and caloric sweeteners affect the
hypothalamic regulation of appetite.
The second study, a follow up to the first examined the temporal mechanism of hypothalamic
response to sucrose ingestion in adults of varying status (healthy weight, overweight, and obesity
at 10- and 35-minutes post ingestion). Prior animal and human translational paradigms have
shown a significant reduction in blood flow to the hypothalamic response to glucose peaking at
15 minutes, with individuals with obesity showing a blunted response. Our study examined the

75
hypothalamic response to sucrose and water, and showed that sucrose ingestion elicited
sustained a reduction in blood flow at 10 minutes and 35 minutes in healthy-weight individuals.
Individuals with obesity displayed a slight increase in hypothalamic response to sucrose over 35
minutes, while overweight individuals displayed an attenuated reduction to sucrose in between
the healthy-weight and overweight signal. Pre-drink hypothalamic and whole brain blood flow did
not vary by weight status. These findings characterize the temporality of the hypothalamic
response to sucrose, and suggest that this signal is uniquely modified by weight status.
The third study examined how intrauterine exposure to GDM and timing of GDM exposure affects
brain and adiposity trajectories in a longitudinal sample of children 204 followed from ages 7
through 16, and whether hypothalamic response to glucose at age 7 predicts adiposity trajectories
in children as they transition from childhood to adolescence. Prior studies have linked GDM
exposure to diverging increases in BMI trajectories around puberty, and timing of GDM exposure
(< 26 weeks’ gestation) to an increased hypothalamic response to glucose compared to
unexposed children. Our study supports and extend these findings, as adiposity trajectories
continue to accelerate and diverge throughout adolescence in GDM exposed children. Our study
is also the first to provide evidence of differences in brain volume trajectories in children exposed
to GDM. These findings suggest a mechanistic link between GDM exposure and risk for adverse
metabolic and neural outcomes.
Strengths
There are several strengths to be noted from these experiments. Studies from Chapters 2 and 3
highlight the benefit of animal to human translational paradigms in human imaging research. We
provided further evidence that the hypothalamic response to glucose is a purported biomarker for
satiety, signaling, and weight status. We also show evidence for a tightly coupled link between
changes in peripheral glucose physiology, medial hypothalamic blood flow responses, and

76
hunger. Our findings align with previous imaging studies that show a prolonged decrease in fMRI
or pASL signals in the hypothalamus after glucose ingestion41,42,46,86. Our results are also
interestingly consistent with classic rodent lesion studies which have linked circulating glucose
with neurons in the ventromedial hypothalamus33, and later studies which demonstrated that the
ventromedial hypothalamus modulate the sympathetic nervous system and hormone responses
during hypoglycemia in rats36.
Second, we were able to use a multimodal approach of pASL and fMRI functional connectivity to
assess nutrient signaling in Chapter 2. The BOLD signal on its own is highly variable and transient
in nature, as connectivity patterns can vary significantly depending on the task or brain state of
the subject, making it challenging to generalize functional connectivity findings152-154. pASL on the
other hand provides an absolute measure of cerebral blood flow is particularly beneficial for
examining slower metabolic responses triggered by nutrient ingestion. Additionally, the wellstudied pASL response to glucose within the hypothalamus complements fMRI BOLD signal
changes by linking functional connectivity patterns to a reliable physiological signal. Since we only
assessed pASL responses to sucrose ingestions in adults of varying weight groups in Chapter 3,
it would be interesting to assess whether functional connectivity to sucrose ingestion varied by
BMI group.
Finally in Chapter 4 we leveraged a large unique longitudinal cohort of children to be the first study
to assess the impact of in utero exposure on child brain and adiposity development. The 6-year
study design allows us to better establish a temporal sequence between prenatal exposures and
later childhood, thus aiding in understanding causality of diabetes transmission from mother to
child. Furthermore, we assessed the natural progression of brain and weight changes in children
as they undergo puberty. This enables us to better understand how in utero exposure may
differentially impact brain development and adiposity trajectories during this sensitive period.

77
Limitations
There are of course several limitations that need to be taken into consideration. While the both of
the study samples had an even split of males and females, we did not investigate the effects of
how non-nutritive sweeteners may differentially impact hypothalamic blood flow in males and
females. Prior work from the Brain Response to Sugar study has shown sex related differences
related to increased reward-related functional activation during a food cue task after ingestion of
sucralose versus sucrose89. Mechanistic associations between hypothalamic blood flow, and
hunger signaling may also be significantly different in boys versus girls. Furthermore, cortical
volume growth trajectories are significantly different in boys versus girls during the time of preadolescence through adolescence155. While the mechanisms underlying these differences are
unclear, cross-sectional and longitudinal studies also show that maturational trajectories of
subcortical volume significantly vary by gender148. Furthermore, a recent investigation from Adise
and colleagues demonstrate associations between changes BMI and subcortical volumes may
be differentially impacted by sex146.
Another limitation is that we did not investigate whether brain and adiposity changes observed in
all three studies varied in minority ethnic groups. While we had a diverse cohort in each study
sample, minority ethnic groups have shown to have higher rates and obesity156, and may display
altered brain and adiposity trajectories from the population cohort. Future research should
specifically examine these variations to better understand the unique impacts on minority
populations and to develop targeted interventions.
Future Directions
Collectively, these studies show that maladaptive appetitive behaviors, excess adiposity, and
prenatal exposure to diabetes are characterized by structural and functional alterations in both

78
reward/hedonic and homeostatic brain regions in children and young adults. These findings align
with prior research indicating that prenatal exposure to gestational diabetes and maternal obesity
leads to functional alterations in brain reward circuitry of children150. Findings from these studies
highlight the importance of examining the interplay between top-down and bottom-up brain
regions in regulating appetite and metabolic outcomes. Evidence from rodent literature map
extensive connections between the hypothalamus and higher-order brain regions to regulate
feeding behavior. Future research should aim to elucidate the specific neural circuits and
mechanisms through which these top-down and bottom-up regions interact to control appetite.
This includes investigating how alterations in these circuits contribute to the development of
obesity and other metabolic disorders, particularly in children exposed to adverse prenatal
environments like gestational diabetes. A better understanding of how these regions are
connected and are related to feeding behavior will provide valuable insights into the mechanism
of childhood obesity and help develop targeted interventions in youth to promote healthy eating
behaviors and metabolic outcomes during this sensitive period of growth and development.
Another interesting future direction would be to explore which subcortical gray matter volumes
are altered longitudinally in GDM exposed children, and examine whether the structural volume
of subcortical regions is associated with the hypothalamic response to glucose, and eating
behaviors. Prior studies have linked GDM exposure, and higher BMI in children with a higher
hypothalamic response to glucose64,65. It may be that children with higher BMI and/or exposed to
GDM may have subcortical amygdala or hippocampal volumes68, which is in turn related to
behaviors of stressed or emotional eating. A recent study from Aviel-Shekler and colleagues
showed that in utero exposure to gestational diabetes caused dysregulation in genes related to
forebrain developed in the frontal cortex and lead to autistic behaviors in mice157. In light of recent
epidemiological evidence linking in utero exposure to GDM with neuropsychiatric disorders,
including autism, ADHD, and eating disorders118,119,158, it may be that the reduction of subcortical

79
volumes in GDM and obese children represents not just maladaptive eating behaviors but also
cognitive dysfunction and social impairments.
Conclusion
This thesis underscores the critical role of prenatal and early life factors in shaping
neurodevelopmental and metabolic outcomes. By identifying how in utero exposure and
sweetener types affect hypothalamic signaling and brain development, we highlight the potential
for early interventions to prevent obesity and related metabolic disorders. These findings
emphasize the importance of early-life prevention strategies to address the growing public health
challenge of obesity.

80
References
1 Popkin, B. M. & Gordon-Larsen, P. The nutrition transition: worldwide obesity
dynamics and their determinants. International journal of obesity 28, S2-S9
(2004).
2 Wang, Y. & Lobstein, T. Worldwide trends in childhood overweight and obesity.
International journal of pediatric obesity 1, 11-25 (2006).
3 Di Cesare, M. et al. The epidemiological burden of obesity in childhood: a
worldwide epidemic requiring urgent action. BMC medicine 17, 1-20 (2019).
4 Chung, S. T., Krenek, A. & Magge, S. N. Childhood obesity and cardiovascular
disease risk. Current atherosclerosis reports 25, 405-415 (2023).
5 Organization, W. H. Report of the commission on ending childhood obesity.
(World Health Organization, 2016).
6 Selman, A., Burns, S., Reddy, A. P., Culberson, J. & Reddy, P. H. The role of
obesity and diabetes in dementia. International journal of molecular sciences 23,
9267 (2022).
7 Slomski, A. Obesity is now the top modifiable dementia risk factor in the US.
Jama 328, 10-10 (2022).
8 Cole, T. J., Bellizzi, M. C., Flegal, K. M. & Dietz, W. H. Establishing a standard
definition for child overweight and obesity worldwide: international survey. Bmj
320, 1240 (2000).
9 Ward, Z. J. et al. Simulation of growth trajectories of childhood obesity into
adulthood. New England Journal of Medicine 377, 2145-2153 (2017).
10 Lister, N. B. et al. Child and adolescent obesity. Nature Reviews Disease Primers
9, 24 (2023).
11 Speakman, J. R. Thrifty genes for obesity, an attractive but flawed idea, and an
alternative perspective: the ‘drifty gene’hypothesis. International journal of
obesity 32, 1611-1617 (2008).

81
12 Nussey, S. S. & Whitehead, S. A. Endocrinology: an integrated approach.
(2001).
13 Elmquist, J. K., Elias, C. F. & Saper, C. B. From lesions to leptin: hypothalamic
control of food intake and body weight. Neuron 22, 221-232 (1999).
14 Crosby, E. C. & Woodburne, R. T. The comparative anatomy of the preoptic area
and the hypothalamus. RESEARCH PUBLICATIONS-ASSOCIATION FOR
RESEARCH IN NERVOUS AND MENTAL DISEASE 20, 52-169 (1939).
15 Hill, J. W. PVN pathways controlling energy homeostasis. Indian journal of
endocrinology and metabolism 16, S627-S636 (2012).
16 Anand, B. K. & Brobeck, J. R. Localization of a “feeding center” in the
hypothalamus of the rat. Proceedings of the society for experimental biology and
Medicine 77, 323-325 (1951).
17 Grossman, S. P. Role of the hypothalamus in the regulation of food and water
intake. Psychological review 82, 200 (1975).
18 Brobeck, J. R. Mechanism of the development of obesity in animals with
hypothalamic lesions. Physiological reviews 26, 541-559 (1946).
19 Hetherington, A. & Ranson, S. The relation of various hypothalamic lesions to
adiposity in the rat. Journal of Comparative Neurology 76, 475-499 (1942).
20 Jansen, G. & Hutchison, C. Production of hypothalamic obesity by microsurgery.
American Journal of Physiology-Legacy Content 217, 487-493 (1969).
21 Mayer, J. & Marshall, N. B. Specificity of gold thioglucose for ventromedial
hypothalamic lesions and hyperphagia. Nature 178, 1399-1400 (1956).
22 Oomura, Y. Central mechanism of feeding. Advances in biophysics 5, 65-142
(1973).
23 Morton, G., Cummings, D., Baskin, D., Barsh, G. & Schwartz, M. Central nervous
system control of food intake and body weight. Nature 443, 289-295 (2006).

82
24 Schwartz, M. W. Brain pathways controlling food intake and body weight.
Experimental Biology and Medicine 226, 978-981 (2001).
25 Schwartz, M. W., Woods, S. C., Porte Jr, D., Seeley, R. J. & Baskin, D. G. Central
nervous system control of food intake. Nature 404, 661-671 (2000).
26 Levin, B. E., Dunn-Meynell, A. A. & Routh, V. H. Brain glucose sensing and body
energy homeostasis: role in obesity and diabetes. American Journal of
Physiology-Regulatory, Integrative and Comparative Physiology 276, R1223-
R1231 (1999).
27 Cone, R. et al. The arcuate nucleus as a conduit for diverse signals relevant to
energy homeostasis. International Journal of Obesity 25, S63-S67 (2001).
28 Cone, R. D. Anatomy and regulation of the central melanocortin system. Nature
neuroscience 8, 571-578 (2005).
29 Huszar, D. et al. Targeted disruption of the melanocortin-4 receptor results in
obesity in mice. Cell 88, 131-141 (1997).
30 Yaswen, L., Diehl, N., Brennan, M. B. & Hochgeschwender, U. Obesity in the
mouse model of pro-opiomelanocortin deficiency responds to peripheral
melanocortin. Nature medicine 5, 1066-1070 (1999).
31 Farooqi, I. S. et al. Clinical spectrum of obesity and mutations in the melanocortin
4 receptor gene. New England Journal of Medicine 348, 1085-1095 (2003).
32 Anand, B. K. & Brobeck, J. R. Hypothalamic control of food intake in rats and
cats. The Yale journal of biology and medicine 24, 123 (1951).
33 Anand, B., Chhina, G. & Singh, B. Effect of Glucose on the Activity of
Hypothalamic" Feeding Centers". Science 138, 597-598 (1962).
34 Mayer, J. Glucostatic mechanism of regulation of food intake. New England
Journal of Medicine 249, 13-16 (1953).
35 Oomura, Y., Ono, T., Ooyama, H. & Wayner, M. Glucose and osmosensitive
neurones of the rat hypothalamus. Nature 222, 282-284 (1969).

83
36 Borg, M. A., Sherwin, R. S., Borg, W. P., Tamborlane, W. V. & Shulman, G. I.
Local ventromedial hypothalamus glucose perfusion blocks counterregulation
during systemic hypoglycemia in awake rats. The Journal of clinical investigation
99, 361-365 (1997).
37 Alsop, D. C. et al. Recommended implementation of arterial spin-labeled
perfusion MRI for clinical applications: a consensus of the ISMRM perfusion
study group and the European consortium for ASL in dementia. Magnetic
resonance in medicine 73, 102-116 (2015).
38 Yokawa, T., Tabuchi, E., Takezawa, M., Ono, T. & Torii, K. Recognition and neural
plasticity responding to deficient nutrient intake scanned by a functional MRI in
the brain of rats with L-lysine deficiency. Obesity research 3, 685S-688S (1995).
39 Min, C. et al. Functional magnetic resonance imaging and immunohistochemical
study of hypothalamic function following oral glucose ingestion in rats. Chinese
medical journal 120, 1232-1235 (2007).
40 Mohr, A. A. et al. A glucose-stimulated BOLD fMRI study of hypothalamic
dysfunction in mice fed a high-fat and high-sucrose diet. Journal of Cerebral
Blood Flow & Metabolism 41, 1734-1743 (2021).
41 Matsuda, M. et al. Altered hypothalamic function in response to glucose ingestion
in obese humans. Diabetes 48, 1801-1806 (1999).
42 Liu, Y., Gao, J.-H., Liu, H.-L. & Fox, P. T. The temporal response of the brain after
eating revealed by functional MRI. Nature 405, 1058-1062 (2000).
43 Smeets, P. A. et al. Oral glucose intake inhibits hypothalamic neuronal activity
more effectively than glucose infusion. American Journal of PhysiologyEndocrinology and Metabolism 293, E754-E758 (2007).
44 Smeets, P. A., de Graaf, C., Stafleu, A., van Osch, M. J. & van der Grond, J.
Functional magnetic resonance imaging of human hypothalamic responses to
sweet taste and calories. The American journal of clinical nutrition 82, 1011-1016
(2005).
45 Page, K. A. et al. Small decrements in systemic glucose provoke increases in
hypothalamic blood flow prior to the release of counterregulatory hormones.
Diabetes 58, 448-452 (2009).

84
46 Page, K. A. et al. Effects of fructose vs glucose on regional cerebral blood flow in
brain regions involved with appetite and reward pathways. Jama 309, 63-70
(2013).
47 DeSisto, C. L., Kim, S. Y. & Sharma, A. J. Peer reviewed: Prevalence estimates
of gestational diabetes mellitus in the United States, pregnancy risk assessment
monitoring system (prams), 2007–2010. Preventing chronic disease 11 (2014).
48 Barbour, L. A. et al. Cellular mechanisms for insulin resistance in normal
pregnancy and gestational diabetes. Diabetes care 30, S112-S119 (2007).
49 Pedersen, J. The pregnant diabetic and her newborn: problems and
management, by Farquhar J. The Williams and Wilkins Company Baltimore
(1968).
50 Silverman, B. L. et al. Long-term prospective evaluation of offspring of diabetic
mothers. Diabetes 40, 121-125 (1991).
51 Hillier, T. A. et al. Childhood obesity and metabolic imprinting: the ongoing effects
of maternal hyperglycemia. Diabetes care 30, 2287-2292 (2007).
52 Dabelea, D. et al. Association of intrauterine exposure to maternal diabetes and
obesity with type 2 diabetes in youth: the SEARCH Case-Control Study. Diabetes
care 31, 1422-1426 (2008).
53 Krishnaveni, G. V. et al. Intrauterine exposure to maternal diabetes is associated
with higher adiposity and insulin resistance and clustering of cardiovascular risk
markers in Indian children. Diabetes care 33, 402-404 (2010).
54 Dabelea, D. et al. Intrauterine exposure to diabetes conveys risks for type 2
diabetes and obesity: a study of discordant sibships. Diabetes 49, 2208-2211
(2000).
55 Silverman, B. L., Rizzo, T. A., Cho, N. H. & Metzger, B. E. Long-term effects of
the intrauterine environment: the Northwestern University Diabetes in Pregnancy
Center. Diabetes care 21, B142 (1998).

85
56 Crume, T. 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 54, 87-92 (2011).
57 Wright, C. S. et al. Intrauterine exposure to gestational diabetes, child adiposity,
and blood pressure. American journal of hypertension 22, 215-220 (2009).
58 Jadhav, A. et al. Placental neurotrophin levels in gestational diabetes mellitus.
International Journal of Developmental Neuroscience 81, 352-363 (2021).
59 Su, C.-H. et al. Correlations between serum BDNF levels and
neurodevelopmental outcomes in infants of mothers with gestational diabetes.
Pediatrics & Neonatology 62, 298-304 (2021).
60 Plagemann, A. et al. Malformations of hypothalamic nuclei in hyperinsulinemic
offspring of rats with gestational diabetes. Developmental neuroscience 21, 58-
67 (1999).
61 Franke, K. et al. ‘Programming’of orexigenic and anorexigenic hypothalamic
neurons in offspring of treated and untreated diabetic mother rats. Brain research
1031, 276-283 (2005).
62 Steculorum, S. M. & Bouret, S. G. Maternal diabetes compromises the
organization of hypothalamic feeding circuits and impairs leptin sensitivity in
offspring. Endocrinology 152, 4171-4179 (2011).
63 Jastreboff, A. M. et al. Altered brain response to drinking glucose and fructose in
obese adolescents. Diabetes 65, 1929-1939 (2016).
64 Ge, B. B. et al. Brain responses to glucose ingestion are greater in children than
adults and are associated with overweight and obesity. Obesity 29, 2081-2088
(2021).
65 Page, K. A. 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 42, 1473-1480 (2019).

86
66 Luo, S. et al. Associations among prenatal exposure to gestational diabetes
mellitus, brain structure, and child adiposity markers. Obesity 31, 2699-2708
(2023).
67 Xuan, D.-S. et al. Brain development in infants of mothers with gestational
diabetes mellitus: a diffusion tensor imaging study. Journal of Computer Assisted
Tomography 44, 947-952 (2020).
68 Lynch, K. M. et al. Selective morphological and volumetric alterations in the
hippocampus of children exposed in utero to gestational diabetes mellitus.
Human brain mapping 42, 2583-2592 (2021).
69 Cai, S. et al. Neonatal amygdala microstructure mediates the relationship
between gestational glycemia and offspring adiposity. BMJ Open Diabetes
Research and Care 9, e001396 (2021).
70 Hales, C. M., Carroll, M. D., Fryar, C. D. & Ogden, C. L. Prevalence of obesity
among adults and youth: United States, 2015–2016. (2017).
71 Hu, F. B. & Malik, V. S. Sugar-sweetened beverages and risk of obesity and type
2 diabetes: epidemiologic evidence. Physiology & behavior 100, 47-54 (2010).
72 Malik, V. S., Pan, A., Willett, W. C. & Hu, F. B. Sugar-sweetened beverages and
weight gain in children and adults: a systematic review and meta-analysis. The
American journal of clinical nutrition 98, 1084-1102 (2013).
73 Bray, G. A. & Popkin, B. M. Dietary sugar and body weight: have we reached a
crisis in the epidemic of obesity and diabetes? Health be damned! Pour on the
sugar. Diabetes care 37, 950-956 (2014).
74 Swithers, S. E., Sample, C. H. & Davidson, T. L. Adverse effects of high-intensity
sweeteners on energy intake and weight control in male and obesity-prone
female rats. Behavioral neuroscience 127, 262 (2013).
75 Van Opstal, A. et al. Dietary sugars and non-caloric sweeteners elicit different
homeostatic and hedonic responses in the brain. Nutrition 60, 80-86 (2019).

87
76 Van Opstal, A. M. et al. Brain activity and connectivity changes in response to
nutritive natural sugars, non-nutritive natural sugar replacements and artificial
sweeteners. Nutritional neuroscience 24, 395-405 (2021).
77 Frank, G. K. et al. Sucrose activates human taste pathways differently from
artificial sweetener. Neuroimage 39, 1559-1569 (2008).
78 Smeets, P. A., Weijzen, P., de Graaf, C. & Viergever, M. A. Consumption of
caloric and non-caloric versions of a soft drink differentially affects brain
activation during tasting. Neuroimage 54, 1367-1374 (2011).
79 Zhang, X. et al. Impacts of Acute Sucralose and Glucose on Brain Activity during
Food Decisions in Humans. Nutrients 12, 3283 (2020).
80 Sylvetsky, A. C. & Rother, K. I. Trends in the consumption of low-calorie
sweeteners. Physiology & behavior 164, 446-450 (2016).
81 Schiffman, S. S. & Rother, K. I. Sucralose, a synthetic organochlorine sweetener:
overview of biological issues. Journal of Toxicology and Environmental Health,
Part B 16, 399-451 (2013).
82 Roger, C. et al. The role of the human hypothalamus in food intake networks: An
MRI perspective. Frontiers in nutrition 8, 760914 (2022).
83 Osada, T. et al. Functional subdivisions of the hypothalamus using areal
parcellation and their signal changes related to glucose metabolism. Neuroimage
162, 1-12 (2017).
84 Wright, H. et al. Differential effects of hunger and satiety on insular cortex and
hypothalamic functional connectivity. European Journal of Neuroscience 43,
1181-1189 (2016).
85 Kullmann, S. et al. The effect of hunger state on hypothalamic functional
connectivity in response to food cues. Human Brain Mapping 44, 418-428 (2023).
86 Smeets, P. A., de Graaf, C., Stafleu, A., van Osch, M. J. & van der Grond, J.
Functional MRI of human hypothalamic responses following glucose ingestion.
Neuroimage 24, 363-368 (2005).

88
87 Luo, S. et al. Resting state hypothalamic response to glucose predicts glucoseinduced attenuation in the ventral striatal response to food cues. Appetite 116,
464-470 (2017).
88 Page, K. A. et al. Circulating glucose levels modulate neural control of desire for
high-calorie foods in humans. The Journal of clinical investigation 121, 4161-
4169 (2011).
89 Yunker, A. G. et al. Obesity and sex-related associations with differential effects
of sucralose vs sucrose on appetite and reward processing: a randomized
crossover trial. JAMA Network Open 4, e2126313-e2126313 (2021).
90 Page, K., Luo, S. & Dorton, H. Neural mechanisms for appetitive response for
high reward foods. Open Science Framework. doi 10 (2020).
91 Dye, L. & Blundell, J. Menstrual cycle and appetite control: implications for weight
regulation. Human reproduction (Oxford, England) 12, 1142-1151 (1997).
92 Krishnan, S., Tryon, R. R., Horn, W. F., Welch, L. & Keim, N. L. Estradiol, SHBG
and leptin interplay with food craving and intake across the menstrual cycle.
Physiology & behavior 165, 304-312 (2016).
93 Aguirre, G. K., Detre, J. A. & Wang, J. Perfusion fMRI for functional
neuroimaging. International review of neurobiology 66, 213-236 (2005).
94 Detre, J. A., Wang, J., Wang, Z. & Rao, H. Arterial spin-labeled perfusion MRI in
basic and clinical neuroscience. Current opinion in neurology 22, 348-355 (2009).
95 Wang, Y. et al. Regional reproducibility of pulsed arterial spin labeling perfusion
imaging at 3T. Neuroimage 54, 1188-1195 (2011).
96 Baroncini, M. et al. MRI atlas of the human hypothalamus. Neuroimage 59, 168-
180 (2012).
97 Kullmann, S. et al. Resting-state functional connectivity of the human
hypothalamus. Human brain mapping 35, 6088-6096 (2014).

89
98 Neudorfer, C. et al. A high-resolution in vivo magnetic resonance imaging atlas of
the human hypothalamic region. Scientific Data 7, 305 (2020).
99 Hoang, H. et al. Low-calorie diet-induced weight loss is associated with altered
brain connectivity and food desire in obesity. Obesity
100 Whitfield-Gabrieli, S. & Nieto-Castanon, A. Conn: a functional connectivity
toolbox for correlated and anticorrelated brain networks. Brain connectivity 2,
125-141 (2012).
101 Nieto-Castanon, A. & Whitfield-Gabrieli, S. CONN functional connectivity toolbox
(RRID: SCR_009550), Version 21. Series CONN functional connectivity toolbox
(RRID: SCR_009550), Version 21 (2021).
102 Penny, W. D., Friston, K. J., Ashburner, J. T., Kiebel, S. J. & Nichols, T. E.
Statistical parametric mapping: the analysis of functional brain images. (Elsevier,
2011).
103 Nieto-Castanon, A. Handbook of functional connectivity magnetic resonance
imaging methods in CONN. (Hilbert Press, 2020).
104 Worsley, K. J. et al. A unified statistical approach for determining significant
signals in images of cerebral activation. Human brain mapping 4, 58-73 (1996).
105 Chumbley, J., Worsley, K., Flandin, G. & Friston, K. Topological FDR for
neuroimaging. Neuroimage 49, 3057-3064 (2010).
106 Flint, A., Raben, A., Blundell, J. & Astrup, A. Reproducibility, power and validity of
visual analogue scales in assessment of appetite sensations in single test meal
studies. International journal of obesity 24, 38-48 (2000).
107 Swithers, S. E., Baker, C. R. & Davidson, T. General and persistent effects of
high-intensity sweeteners on body weight gain and caloric compensation in rats.
Behavioral neuroscience 123, 772 (2009).
108 Kullmann, S. et al. Selective insulin resistance in homeostatic and cognitive
control brain areas in overweight and obese adults. Diabetes Care 38, 1044-
1050 (2015).

90
109 Hetherington, A. & Ranson, S. Hypothalamic lesions and adiposity in the rat. The
Anatomical Record 78, 149-172 (1940).
110 Apps, M. A., Rushworth, M. F. & Chang, S. W. The anterior cingulate gyrus and
social cognition: tracking the motivation of others. Neuron 90, 692-707 (2016).
111 Wang, J. et al. Convergent functional architecture of the superior parietal lobule
unraveled with multimodal neuroimaging approaches. Human brain mapping 36,
238-257 (2015).
112 Dalenberg, J. R. et al. Short-term consumption of sucralose with, but not without,
carbohydrate impairs neural and metabolic sensitivity to sugar in humans. Cell
Metabolism 31, 493-502. e497 (2020).
113 Higgins, K. A. & Mattes, R. D. A randomized controlled trial contrasting the effects
of 4 low-calorie sweeteners and sucrose on body weight in adults with
overweight or obesity. The American journal of clinical nutrition 109, 1288-1301
(2019).
114 Contreras-Chavez, G. G., Estrada, J. A. & Contreras, I. Changes in appetite
regulation-related signaling pathways in the brain of mice supplemented with
non-nutritive sweeteners. Journal of Molecular Neuroscience 71, 1144-1155
(2021).
115 Vidarsdottir, S. et al. Glucose ingestion fails to inhibit hypothalamic neuronal
activity in patients with type 2 diabetes. Diabetes 56, 2547-2550 (2007).
116 Wang, D. J., Chen, Y., Fernández-Seara, M. A. & Detre, J. A. Potentials and
challenges for arterial spin labeling in pharmacological magnetic resonance
imaging. Journal of Pharmacology and Experimental Therapeutics 337, 359-366
(2011).
117 Plagemann, A. et al. Perinatal elevation of hypothalamic insulin, acquired
malformation of hypothalamic galaninergic neurons, and syndrome x-like
alterations in adulthood of neonatally overfed rats. Brain research 836, 146-155
(1999).
118 Xiang, A. H. et al. Maternal gestational diabetes mellitus, type 1 diabetes, and
type 2 diabetes during pregnancy and risk of ADHD in offspring. Diabetes Care
41, 2502-2508 (2018).

91
119 Xiang, A. H. et al. Association of maternal diabetes with autism in offspring. Jama
313, 1425-1434 (2015).
120 Sun, Y. et al. Effects of pre-pregnancy body mass index and gestational weight
gain on maternal and infant complications. BMC Pregnancy Childbirth 20, 390
(2020). https://doi.org:10.1186/s12884-020-03071-y
121 Tam, W. H. et al. Glucose intolerance and cardiometabolic risk in adolescents
exposed to maternal gestational diabetes: a 15-year follow-up study. Diabetes
Care 33, 1382-1384 (2010). https://doi.org:10.2337/dc09-2343
122 Kubo, A. et al. Maternal hyperglycemia during pregnancy predicts adiposity of the
offspring. Diabetes care 37, 2996-3002 (2014).
123 Nilsson, C., Carlsson, A. & Landin-Olsson, M. Increased risk for overweight
among Swedish children born to mothers with gestational diabetes mellitus.
Pediatric diabetes 15, 57-66 (2014).
124 Luo, S. et al. Associations Between Exposure to Gestational Diabetes Mellitus In
Utero and Daily Energy Intake, Brain Responses to Food Cues, and Adiposity in
Children. Diabetes Care (2021). https://doi.org:10.2337/dc20-3006
125 Goyal, A., Gupta, Y., Singla, R., Kalra, S. & Tandon, N. American Diabetes
Association "Standards of Medical Care-2020 for Gestational Diabetes Mellitus":
A Critical Appraisal. Diabetes Ther 11, 1639-1644 (2020).
https://doi.org:10.1007/s13300-020-00865-3
126 Rasmussen, A. R. et al. Validity of self-assessment of pubertal maturation.
Pediatrics 135, 86-93 (2015).
127 Dale, A. M., Fischl, B. & Sereno, M. I. Cortical surface-based analysis: I.
Segmentation and surface reconstruction. Neuroimage 9, 179-194 (1999).
128 Fischl, B., Sereno, M. I. & Dale, A. M. Cortical surface-based analysis: II:
inflation, flattening, and a surface-based coordinate system. Neuroimage 9, 195-
207 (1999).
129 Fischl, B. et al. Whole brain segmentation: automated labeling of
neuroanatomical structures in the human brain. Neuron 33, 341-355 (2002).

92
130 Angelo, B. C. et al. Relationships between physical activity, healthy eating and
cortical thickness in children and young adults. Brain Imaging and Behavior 16,
2690-2704 (2022).
131 Reuter, M., Rosas, H. D. & Fischl, B. Highly accurate inverse consistent
registration: a robust approach. Neuroimage 53, 1181-1196 (2010).
132 Greve, D. N. et al. A surface-based analysis of language lateralization and
cortical asymmetry. Journal of cognitive neuroscience 25, 1477-1492 (2013).
133 Alosco, M. L. et al. Body mass index and brain structure in healthy children and
adolescents. International Journal of Neuroscience 124, 49-55 (2014).
134 Laurent, J. S. et al. Associations among body mass index, cortical thickness, and
executive function in children. JAMA pediatrics 174, 170-177 (2020).
135 Wood, S. N. Stable and efficient multiple smoothing parameter estimation for
generalized additive models. Journal of the American Statistical Association 99,
673-686 (2004).
136 Wood, S. N. Generalized additive models: an introduction with R. (chapman and
hall/CRC, 2017).
137 Fjell, A. M. et al. When does brain aging accelerate? Dangers of quadratic fits in
cross-sectional studies. Neuroimage 50, 1376-1383 (2010).
138 Mills, K. L. et al. Inter-individual variability in structural brain development from
late childhood to young adulthood. NeuroImage 242, 118450 (2021).
139 Garcia, E. et al. Patterns and determinants of exhaled nitric oxide trajectories in
schoolchildren over a 7-year period. European Respiratory Journal 56 (2020).
140 Foland-Ross, L. C. et al. Brain function differences in children with type 1
diabetes: a functional MRI study of working memory. Diabetes 69, 1770-1778
(2020).
141 Mauras, N. et al. Impact of type 1 diabetes in the developing brain in children: a
longitudinal study. Diabetes Care 44, 983-992 (2021).

93
142 Mazaika, P. K. et al. Variations in brain volume and growth in young children with
type 1 diabetes. Diabetes 65, 476-485 (2016).
143 Bruehl, H., Sweat, V., Tirsi, A., Shah, B. & Convit, A. Obese adolescents with type
2 diabetes mellitus have hippocampal and frontal lobe volume reductions.
Neuroscience and medicine 2, 34 (2011).
144 Giedd, J. N. & Rapoport, J. L. Structural MRI of pediatric brain development:
what have we learned and where are we going? Neuron 67, 728-734 (2010).
145 Jiang, F. et al. Obesity is associated with decreased gray matter volume in
children: a longitudinal study. Cerebral Cortex 33, 3674-3682 (2023).
146 Adise, S. et al. Smaller subcortical volume relates to greater weight gain in girls
with initially healthy weight. Obesity (2024).
147 Gómez-Apo, E., Mondragón-Maya, A., Ferrari-Díaz, M. & Silva-Pereyra, J.
Structural brain changes associated with overweight and obesity. Journal of
obesity 2021, 6613385 (2021).
148 Herting, M. M. et al. Development of subcortical volumes across adolescence in
males and females: A multisample study of longitudinal changes. NeuroImage
172, 194-205 (2018).
149 Areias, M. F. C. & Prada, P. O. Mechanisms of insulin resistance in the amygdala:
influences on food intake. Behavioural brain research 282, 209-217 (2015).
150 Luo, S. et al. The role of maternal BMI on brain food cue reactivity in children: a
preliminary study. Brain imaging and behavior, 1-10 (2021).
151 Kim, M. S. et al. Prefrontal cortex and amygdala subregion morphology are
associated with obesity and dietary self-control in children and adolescents.
Frontiers in Human Neuroscience 14, 563415 (2020).
152 Geerligs, L., Rubinov, M. & Henson, R. N. State and trait components of
functional connectivity: individual differences vary with mental state. Journal of
Neuroscience 35, 13949-13961 (2015).

94
153 Gratton, C. et al. Functional brain networks are dominated by stable group and
individual factors, not cognitive or daily variation. Neuron 98, 439-452. e435
(2018).
154 Xie, H. et al. Whole-brain connectivity dynamics reflect both task-specific and
individual-specific modulation: A multitask study. Neuroimage 180, 495-504
(2018).
155 Lenroot, R. K. & Giedd, J. N. Sex differences in the adolescent brain. Brain and
cognition 72, 46-55 (2010).
156 Quader, Z. S., Gazmararian, J. A. & McCullough, L. E. Obesity and understudied
minority children: existing challenges and opportunities in epidemiology. BMC
pediatrics 19, 1-4 (2019).
157 Aviel-Shekler, K. et al. Gestational diabetes induces behavioral and brain gene
transcription dysregulation in adult offspring. Translational psychiatry 10, 412
(2020).
158 Wang, H. et al. Maternal diabetes and the risk of feeding and eating disorders in
offspring: a national population-based cohort study. BMJ Open Diabetes
Research and Care 8, e001738 (2020).
159 Henley, C. Foundations of neuroscience. (2021).

95
Appendix
Figure 14, Appendix Figure 1: Visual Display of Hypothalamic Nuclei
Figure shows lateral hypothalamic nucleus, labeled in green, ventromedial hypothalamic nucleus labeled in purple. The figure was
created by Valerie Hedges, and is licensed under CC BY-NC-SA 4.159
Table 26, Appendix Table 1: Anatomical Inputs, Outputs, and Functions of the Medial and Lateral
Hypothalamus
Inputs Outputs Function
Lateral Hypothalamus Prefrontal cortex
Nucleus accumbens
Amygdala
Hippocampus
Brainstem structures
Sensory inputs
(metabolic parameters)
Ventromedial
hypothalamus
Medial and lateral septal
nuclei
Central medial amygdala
Zona incerta
Periaqueductal gray
Widespread projections
via orexinergic neurons
Acts as the "satiety
center" - lesions induce
hyperphagia and obesity.
Involved in energy
homeostasis, glucose
sensing, and autonomic
regulation.
Medial Hypothalamus Lateral hypothalamus
Arcuate nucleus
Lateral Hypothalamus
Paraventricular nucleus
of hypothalamus
Periaqueductal gray
Brainstem
Acts as the "feeding
center" - lesions cause
anorexia. Involved in
hunger, feeding behavior,
motivation, reward, and
autonomic regulation.

Abstract (if available)

Abstract This dissertation aims to characterize the neurohumoral and temporal mechanisms of hypothalamic signaling across various nutrients in adults of varying weight status, and evaluate the link between in utero exposure to Gestational Diabetes Mellitus (GDM) and changes in adiposity and brain volume from childhood to adolescence.

The first study examined the effects of sucralose, a non-caloric sweetener, compared to caloric sugar (sucrose) and water on hypothalamic, glycemic, and hunger responses using imaging techniques. Results showed that sucralose significantly altered hypothalamic responses and connectivity to somatosensory processing regions, highlighting differences in how nonnutritive sweeteners affect appetite regulation. The second study examined the temporal mechanism of hypothalamic response to sucrose ingestion in adults of varying weight statuses (healthy weight, overweight, and obesity) at 10- and 35-minutes post-ingestion. Sucrose ingestion elicited a sustained reduction in hypothalamic blood flow in healthy-weight individuals, while individuals with obesity and overweight individuals displayed an attenuated reduction, suggesting that the hypothalamic response to sucrose is dynamically modulated by weight status.

The third study investigated how in utero exposure to GDM affects adiposity, and brain volume trajectories in children aged 7 to 16. Results showed that GDM exposure predicted higher adiposity trajectories, which continued to diverge throughout adolescence in GDM-exposed children. This study is also the first to reveal reduced growth of subcortical brain volume trajectories in GDM compared to control children, suggesting a link between GDM exposure, and adverse metabolic and neural outcomes.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Effects of sugar and non-nutritive sweetener consumption on neural processing, ingestive behavior, and appetite regulation

PDF

Prenatal and lifestyle predictors of metabolic health and neurocognition during childhood

PDF

Classically conditioned responses to food cues among obese and normal weight individuals: conditioning as an explanatory mechanism for excessive eating

PDF

The acute impact of glucose and sucralose on food decisions and brain responses to visual food cues

PDF

The association of cerebrovascular disease risk factors with brain structure and its modification by genetic variation

PDF

Relationships between lifetime chronic stress exposure, vascular risk, cognition, and brain structure in HIV

PDF

Effects of western dietary factors during early life on glucose metabolism, the gut microbiome, and neurocognition

PDF

Independent and interactive effects of depression genetic risk and household socioeconomic status on emotional behavior and brain development

PDF

Dedifferentiation of emotion regulation strategies in the aging brain: an MVPA investigation

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

Homeostatic imbalance and monetary delay discounting: effects of feeding on RT, choice, and brain response

PDF

Air pollution, mitochondrial function, and growth in children

PDF

Morphological and microstructural models of typical development in the hippocampus and white matter

PDF

The evolution of decision-making quality over the life cycle: evidence from behavioral and neuroeconomic experiments with different age groups

PDF

Pediatric magnetic resonance image processing: applications to posterior fossa cancer and normal development

Asset Metadata

Creator Prathap, Sandhya (author)

Core Title Prenatal and brain factors shape appetite regulation and weight from childhood onwards

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Neuroscience

Degree Conferral Date 2024-08

Publication Date 07/15/2024

Defense Date 06/19/2024

Publisher Los Angeles, California (original), University of Southern California (original), University of Southern California. Libraries (digital)

Tag appetite regulation,brain development,Child development,Hypothalamus,OAI-PMH Harvest,obesity,prenatal factors

Format theses (aat)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Mather, Mara (committee chair), Jann, Kay (committee member), Monterosso, John R. (committee member), Page, Kathleen A. (committee member), Xiang, Anny H. (committee member)

Creator Email prathap@usc.edu,sandhya.prathap@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC113997OH9

Unique identifier UC113997OH9

Identifier etd-PrathapSan-13230.pdf (filename)

Legacy Identifier etd-PrathapSan-13230

Document Type Dissertation

Format theses (aat)

Rights Prathap, Sandhya

Internet Media Type application/pdf

Type texts

Source 20240715-usctheses-batch-1182 (batch), 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 author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright.

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

Repository Email cisadmin@lib.usc.edu