University of Southern California Dissertations and Theses

Ectopic fat and adipose tissue inflammation in overweight and obese African Americans and Hispanics

(USC Thesis Other)

Ectopic fat and adipose tissue inflammation in overweight and obese African Americans and Hispanics

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content ECTOPIC FAT AND ADIPOSE TISSUE INFLAMMATION IN OVERWEIGHT AND
OBESE AFRICAN AMERICANS AND HISPANICS

By
Tanya Lynn Alderete

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL OF THE
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(INTEGRATIVE BIOLOGY OF DISEASE)

August 2014

Copyright 2014 Tanya L. Alderete
! 2!
Table of Contents
Page 3. List of Tables
Page 4. List of Figures
Page 5. Acknowledgements
Page 6-7. Abstract
Page 8. Chapter 1: Obesity, Insulin Resistance, and Risk for Type 2 Diabetes
Page 8. Introduction
Page 8-9. Obesity and Cardiometabolic Disease Risk
Page 9-10. Ectopic Fat Accumulation: Visceral and Liver Fat
Page 10-11. Adipose Tissue: Endocrine Organ
Page 11-13. Adipose Tissue Inflammation: Link Between Obesity and Disease Risk
Page 13-14. Metabolically Benign Obesity
Page 14. Decreasing Systemic and Adipose Tissue Inflammation
Page 15-16. Anti-Inflammatory Medications and Adipose Tissue Inflammation
Page 16. Conclusion
Page 18-25. Chapter 1 References
Page 26. Chapter 2: Liver Fat Has a Stronger Association with Risk Factors for
Type 2 Diabetes in African Americans Compared with Hispanic Adolescents
Page 26. Abstract
Page 27. Introduction
Page 28-30. Methods
Page 30-31. Results
Page 31-35. Discussion
Page 40-42. Chapter 2 References
Page 43. Chapter 3: A Novel Biopsy Method to Increase Yield of Subcutaneous
Abdominal Adipose Tissue
Page 43. Abstract
Page 44. Introduction
Page 44-46. Methods
Page 46. Results
Page 47-48. Discussion
Page 52-54. Chapter 3 References
Page 55. Chapter 4: Salsalate Treatment Improves Glycemia but Does Not Alter
Adipose Tissue Inflammation in Non-Diabetic Obese Hispanic Young Adults
Page 55. Abstract
Page 56. Introduction
Page 56-60. Methods
Page 60-62. Results
Page 62-65. Discussion
Page 73-78. Chapter 4 References
Page 79. Chapter 5: Summary of Findings, Future Directions, and Conclusions
Page 92-100. Chapter 5 References

! 3!
List of Tables

Page 17. Table 1-1. Key Background Point
Page 36. Table 2-1: Baseline Descriptive Statistics
Page 67-68. Table 4-1: Baseline Group Comparisons
Page 69-70. Table 4-2: Variables at Baseline and After the 4-Week Intervention
Page 79. Table 5-1: The AA vs. Hispanic Paradox
Page 86. Table 5-2: Improvements Not Mediated by Adipose Tissue

! 4!
List of Figures

Page 12. Figure 1-1. NF-kβ Signaling Pathways
Page 37. Figure 2-1: High/low LF and high/low VAT Analysis by Ethnicity
Page 38. Figure 2-2: LF and VAT Regression Analysis by Ethnicity
Page 39. Figure 2-3: High/low LF and High/low VAT by Prediabetes Status
Page 49. Figure 3-1: Biopsy Method With the Bergström Biopsy Needle
Page 50. Figure 3-2: Subcutaneous Adipose Tissue Biopsies: Weight and Location
Page 51. Figure 3-3: Histology and Flow Cytometry From Adipose Tissue Biopsy
Page 66. Figure 4-1: Study Flow Diagram
Page 71. Figure 4-2: Fasting Measures From the 2-hr OGTT
Page 72. Figure 4-3: Glucose, Insulin, and C-Peptide During the 2-hr OGTT
Page 80. Figure 5-1: Disposition Index
Page 81. Figure 5-2: Four Groups Based on LF and VAT
Page 81. Figure 5-3: β-cell Compensation in Response to LF
Page 82. Figure 5-4: Fat Inflammation and Metabolic Dysfunction
Page 85. Figure 5-5: Adipose Tissue Viability
Page 88. Figure 5-6: Diet and Fat Inflammation
Page 89. Figure 5-7: Obesity and the Gut Microbiota
Page 89. Figure 5-8: Gut-Liver and Adipose Tissue
Page 90. Figure 5-9: Gut Microbiome and Inflammation

! 5!
Acknowledgements
It is with the upmost gratitude that I would like to acknowledge my mentors, Dr. Michael
Goran and Dr. Fred Sattler, for offering the essential guidance and support needed to
compete this body of work. I would like to especially thank Dr. Goran for offering me the
support and intellectual freedom needed to grow as an independent thinker. Dr. Sattler’s
generosity and genuine excitement for science has been essential to our work and my
professional growth. Additionally, I would like to thank each of my committee members,
including Drs. Hooman Allayee, Steven Mittelman, and Joyce Richey, for teaching me
the true meaning behind transdisciplinary and collaborative research.

Special thanks to my husband, Andrew P. Alderete, who has taught me the importance of
working and living passionately through example. His unwavering support has allowed
us to share in our successes and affords me the ability to fully devote myself to my
scientific research. Thank you to my mother and sister, who are never hesitant to listen
and offer encouragement. My mother’s fortitude, dedication, and ingrained compassion
inspire me to strive for the incredible; for this I am most grateful. I would also like to
thank my mother- and father-in-law for their support, welcoming me into their family,
and teaching me the true meaning of success. Lastly, I would like to express my gratitude
to Dr. Claudia Toledo-Corral for being one of my biggest supporters and advocates. Our
collaborations have shown me the importance of teamwork and I am continually grateful
for her friendship.

In addition to the individuals mentioned above, I would like to express my appreciation
to our research participants and their families. Their commitment to our studies serves as
a constant reminder that research does not take place in a vacuum. Collectively, none of
this work would have been possible without them or our dedicated research teams at the
Childhood Obesity Research Center and USC Clinical Trials Unit.

! 6!
Abstract
Studies have shown that obesity is linked to insulin resistance (IR) and type 2 diabetes,
yet the exact mechanism(s) linking excess adiposity to increased disease risk is unknown.
Studies suggest that visceral adipose tissue (VAT), and liver fat (LF) contribute to the
development of IR and risk for T2D. However, this theory is challenged by the fact that
overweight and obese African Americans (AAs) have less VAT and LF than Hispanics,
yet share a similar risk for IR and T2D. Recent data suggest that excess adipose tissue
contributes to chronic inflammation, thereby promoting the development of IR and other
metabolic complications. Supporting this idea, recent studies have found that anti-
inflammatory drugs, such as salsalate, decrease markers of inflammation and improve
glucose control in Caucasian adults with T2D and/or impaired glucose control.

Overall, the purpose of this dissertation was multi-faceted where we focused on both the
location and inflammatory state of adipose tissue. In order to examine adipose tissue
inflammation, we developed a novel biopsy method to sample large quantities of intact
subcutaneous adipose tissue for a variety of research purposes (e.g., cell sorting,
histology, gene expression). This allowed us to perform a randomized, placebo-controlled
trial in obese Hispanic young adults in order to determine whether any salsalate-induced
metabolic improvements were mediated by changes in adipose tissue inflammation.

From these studies, we show that high LF was more strongly associated with decreased
insulin sensitivity and β-cell function among African Americans (AAs) compared to
Hispanics. Our results suggest that Hispanics exhibited appropriate β-cell compensation
for liver fat induced insulin resistance where AAs did not. These findings may partially
explain why AAs are a great risker for T2D, despite having lower levels of LF and VAT
when compared to Hispanics. Our work focusing on adipose tissue inflammation led us to
develop a novel biopsy method where we were able to sample large quantities of intact
adipose tissue (~1.5 g). This method allowed us to systematically examine the effects of
salsalate on various markers of adipose tissue inflammation. Overall, we confirmed that
salsalate improved metabolic outcomes related to T2D risk in otherwise healthy obese
Hispanics. Our finding suggests that these improvements occurred in the absence of any
! 7!
significant alterations to adipose tissue inflammation. Taken together, our results suggest
that although specific patters of fat distribution and inflammation contribute to T2D risk,
salsalate does not target adipose tissue inflammation. Future studies should examine other
treatment strategies and/or therapeutic targets in an effort to decrease obesity-associated
inflammation as a means to decrease disease risk.

! 8!
Chapter 1: Obesity, Insulin Resistance, and Risk for Type 2 Diabetes

Introduction
In the United States, approximately 35% of 12- to 19-year olds and 69% of adults aged 20 years
or older were classified as overweight or obese (1). Obesity increases risk for a number of
metabolic diseases including insulin resistance (IR), type 2 diabetes (T2D), cardiovascular
disease (CVD), non-alcoholic fatty liver disease (NAFLD) and some forms of cancer (2).
Despite the well-known link between obesity and increased morbidity, the exact mechanism(s)
linking excess adiposity to increased disease risk is unknown. Currently, several putative
explanations exist for why fat affects metabolic health. One such theory is based on the
anatomical location of fat deposition and ectopic fat accumulation, namely liver fat (3).
Specifically, current literature suggests that visceral and liver fat accumulation affect organ
function and contribute to the development of insulin resistance (IR), fatty liver, and the
metabolic syndrome (4-8). However, even in individuals matched for body fat, significant
differences can exist in metabolic outcomes, and the phenomenon of metabolically healthy obese
has been well described (9). Therefore, recent data suggests an alternative hypothesis relating
excess adiposity to increased disease risk, which is based on the metabolic and morphological
properties of adipose tissue. In this scenario, excess adipose tissue contributes to a state of
chronic inflammation which promotes development of systemic IR as well as other metabolic
complications by stimulating nuclear factor κB (NF-κB) and Jun N-terminal kinase (JNK)
pathways in adipocytes (10). Overall, this chapter will serve to summarize the metabolic
complications associated with obesity, as well as how the anatomical location and inflammatory
state of adipose tissue can contribute to increased disease risk. Finally, considerable attention
will be given to anti-inflammatory therapies that have the potential to target adipose tissue
inflammation and improve glucose and lipid homeostasis.

Obesity and Cardiometabolic Disease Risk
Numerous studies have shown that obesity is linked to IR, T2D, CVD, and NAFLD. These
disease outcomes are observed in men and women of all ethnicities; however, despite similar
rates of obesity, African Americans (AAs) and Hispanics suffer a higher prevalence of these
comorbid conditions. In fact, death rates due to T2D and heart disease are greatest among AAs
! 9!
and Hispanics compared to other ethnic groups. Between 2005-2006, the total crude prevalence
of prediabetes and T2D in AAs, Hispanics, and Non-Hispanic Whites ≥12 years of age were
36.5%, 41.1%, and 38.6%, respectively (11). Interestingly, ethnic differences are also observed
in liver fat accumulation where 45% of Hispanics, 33% of Non-Hispanic Whites, and 24% of
AAs have been shown to be affected by NAFLD (12). For these reasons, obesity and related
metabolic diseases, especially in minority populations, are becoming an increasing health burden
in the Unites States. Therefore, studies aimed at understanding the physiological mechanisms
behind ethnic differences in obesity related disease risk are urgently needed.

Ectopic Fat Accumulation: Visceral and Liver Fat
Recent studies show that fat distribution and ectopic fat accumulation (5, 13-15), not overall
adiposity per se, is the causal link between obesity and disease risk. Theories describing the
manner in which ectopic fat is related to IR and diabetes include the Randle/portal hypothesis
(16, 17) and the concept of adipose tissue dysfunction (15). The former suggests that excess free
fatty acids (FFAs) released form the highly insulin-resistant visceral adipose tissue (VAT) spill
into the liver and/or pancreas via the portal vein, leading to excess ectopic fat and IR. The
alternate theory suggests an inability of adipose tissue to expand and/or proliferate to store
excess energy, resulting in triglyceride (TAG) accumulation in organs.

As previously mentioned, AAs and Hispanics suffer from higher rates of obesity than non-
Hispanics whites (18), placing them at increased risk for IR and T2D (19). Interestingly,
although AAs and Hispanics share similar rates of obesity, AAs have less VAT and liver fat (LF)
than Hispanics (20, 21). These ethnic differences are even more intriguing given that previous
studies have shown that increased VAT (22-25) and LF (15, 26) are associated with increased
risk for diabetes. Therefore, having lower VAT and LF implies that AAs should be partially
metabolically protected; however, studies have shown that AAs share a comparable risk for
insulin resistance and T2D when compared to Hispanics (20, 22, 23). This AA versus Hispanic
paradox suggests that the absolute amount of ectopic fat may not be as important as its ethnic-
specific quantity and location.

! 10!
A major limitation of studies examining ectopic fat and its relationship with T2D risk is the high
correlation between VAT and LF (27, 28), making it difficult to determine which fat depot(s)
drives metabolic risk. Elegant studies, mostly in Non-Hispanic Whites adults, provide evidence
that LF is the predominant compartment associated with metabolic disease risk in obese adults
(14, 29). Therefore, due to the various ethnic differences in ectopic fat accumulation, yet similar
risk for IR and T2D, one of the primary objectives of this dissertation was to examine the
separate and combined effects of VAT and LF on risk factors for T2D in overweight and
minority youth and adolescents. An additional focus of this work was aimed at determining
whether these observed relationships differed in AAs and Hispanics, thereby potentially
explaining the paradox of similar diabetes risk in AAs at lower levels of VAT and LF.

Adipose Tissue: Endocrine Organ
Recent studies have found that the metabolic and inflammatory state of adipose tissue is just as
important as its anatomical location. This is especially true since adipose tissue is not only
composed of adipocytes, but also includes immune cells, such as macrophages, that can
contribute to inflammation. It was once believed that adipose tissue was only involved in the
storage of TAG, but recent studies have demonstrated that it also acts as an endocrine organ.
Hotamisligil et al. and Karasik et al. first showed that pro-inflammatory cytokine TNF-α was
produced by adipocytes, induced IR, and increased with expanding fat volume (30). The concept
of adipose tissue as a site for the production of cytokines and other substances has expanded to
other factors including leptin, interleukin (IL)-6, resistin, monocyte chemoattractant protein-1
(MCP-1), plasminogen activator inhibitor-1 (PAI-1), angiotensinogen, visfatin, retinol-binding
protein-4, serum amyloid A, dipeptidyl peptidate (DPP) 4, and adipsin.

Adiponectin, DPP4, and adipsin are produced by adipocytes, with the exception of DPP4, levels
of these adipokines decrease with increasing adiposity. Leptin, adiponectin, and adipsin are
adipokines; while TNF-α, IL-6, MCP-1, visfatin, and PAI-1 are also expressed at high levels in
activated macrophages. Adiponectin has been shown to be insulin sensitizing while DPP4 is
suggested to impair insulin sensitivity by inhibiting incretins GLP-1 and GIP (31). Recent
unpublished data by Spiegelman et al., has shown that adipsin improves β-cell function by
stimulating insulin secretion. TNF-α, IL-6, resistin, and other pro-inflammatory cytokines
! 11!
participate in the induction and maintenance of the acute inflammatory response associated with
obesity. Additionally, MCP-1 and other chemokines recruit macrophages to adipose tissue.
These cytokines and chemokines activate intracellular pathways that promote the development of
IR, T2D and other metabolic complications associated with obesity (10). Therefore, adipose
tissue and macrophages within adipose tissue have been shown to play important roles in the
regulation of metabolic pathways through the excretion of adipokines and cytokines.

Adipose Tissue Inflammation: Link Between Obesity and Disease Risk
Obesity and high-fat western diets activate inflammatory processes in adipose tissue, which
promote development of IR as well as other metabolic complications by activating I kappa B
kinase β (IKKβ) / NF-κB and JNK pathways in adipocytes and the liver (32). Stimuli that
activate these pathways during metabolic dysfunction include ligands for: TNF-α, IL-1, Toll
(TLR2/4), or advanced glycation endproducts (AGE) receptors. Obesity-induced IKKβ
activation leads to NF-κB translocation and increased expression of mediators of inflammation
that can cause IR, T2D, and CVD (32). Other studies have shown that enhanced inducible nitric
oxide sythase (iNOS) production is associated with obesity and inflammatory responses.
Increases in iNOS negatively modulate insulin signaling through S-nitrosylation of proteins
involved in the early steps of insulin signaling, such as (IR)-β, IRS-1, and protein kinase B (Akt)
(33).

Clinical evidence suggests that the risk for obesity-related conditions such as T2D and CVD are
related to systemic inflammation arising from inflamed adipose tissue (32, 34-36). Mice fed a
high fat diet have enlarged adipocytes where adipose tissue immune cells increase in number and
switch from alternative anti-inflammatory macrophages (M2) to pro-inflammatory macrophages
(M1). These pro-inflammatory macrophages secrete high levels of pro-inflammatory cytokines
(TNF-α, IL-6), reactive oxygen species (ROS) / iNOS, and MCP-1 (36), which may spill over
into the blood stream (32, 37, 38). Once in the circulation, pro-inflammatory and pro-atherogenic
mediators may promote systemic IR, endothelial dysfunction and CVD (30, 39).

Supporting the role of adipose tissue inflammation in contributing to obesity-associated disease
risk, we have reported that obese AA and Hispanic young adults with crown-like structures
! 12!
(CLS) in their adipose tissue, had significantly higher levels of fasting plasma glucose, insulin,
HOMA-IR, and TNF-α compared to those without CLS. Furthermore, adipose tissue from
subjects with CLS showed increased expression of NF-κB stress pathway activity including the
TNF receptor superfamily 11b, IL-1ra, MCP-1, lipopolysaccharide binding protein (LBP), and
MMP-9 along with monocyte CD14. These results suggest that FFA/intestinal translocation of
microbial products are possible driving forces for increased adipose tissue inflammation (40).
These findings support the concept that adipose tissue inflammation significantly affects
metabolic health and are consistent with a model that suggests that products of the NF-κB stress
pathway mediate inflammation and secretion of proatherogenic mediators.

In this model, pro-inflammatory cytokines (e.g., TNF-α and IL-1β) secreted by inflamed adipose
tissue, along with lipopolysaccharide (LPS) and FFAs,
bind their respective cell receptors or intracellular
binding proteins and trigger TNFR/IL-1R, TLR, and
FFA/A-FABP signaling (Figure 1-1) (40, 41). These
signaling pathways result in nuclear translocation of
NF-κB and increased nuclear transcription of an entire
cascade of pro-inflammatory, pro-atherogenic
mediators, and MCP-1 which recruits
monocytes/macrophages from the systemic circulation
into the adipose tissue (42). We postulate that these
inflammatory mediators spill-over into the systemic
circulation and provide the substrates for systemic insulin resistance and vascular damage.

In addition to inflamed adipose tissue, recent reports indicate that sCD163 is a clinically useful
indicator of systemic inflammation/immune activation of monocytes and macrophages. In fact,
sCD163 has been shown to be closely linked to abdominal obesity, systemic TNF-α levels, IR,
endothelial dysfunction and vascular damage (43-45). In one study, levels of sCD163, but not
highly sensitive C-Reactive Protein (hsCRP), were associated with arterial inflammation of
atheromatous plaque (46). Furthermore, vascular adhesion molecules, such as sICAM and
sVCAM which are secreted by inflamed adipose tissue, are increased with obesity (47), systemic
Figure 1-1. NF-κB Signaling Pathways
!
! 13!
inflammation (48, 49) and IR (50, 51). Therefore, sCD163, vascular adhesion molecules (e.g.,
sICAM; also a marker of endothelial dysfunction) have shown utility in assessing the presence of
inflammation, vascular dysfunction, and early atherosclerosis during obesity.

Metabolically Benign Obesity
Despite the observed link between adiposity, NAFLD, and T2D, some individuals exhibit
“metabolically benign obesity” and are protected from the metabolic consequences of excess
adiposity, possibly due to differences in adipocyte tissue metabolism and macrophage infiltration.
In a study aimed at identifying insulin resistant individuals, 17% of the overweight and obese
participants were found to be insulin sensitive (52). Additionally, a review by Karelis et al.
determined that approximately 20% of the general population can be categorized as obese but
metabolically healthy. In contrast to this, 18% of the population were found to have a normal
body weight but suffered from severe metabolic abnormalities (9). In other studies among obese
adults, the degree of adipose tissue inflammation was closely associated with increased
metabolic risk for T2D, CVD and NAFLD, whereas obese adults without adipose tissue
inflammation have metabolic risk factors in the healthy range (40, 53).

Although most human research has focused on the links between fat distribution and metabolic
disease risk, other studies suggest a role for adipose tissue inflammation. As previously
mentioned, among obese young minority adults, we found that approximately 40% of subjects
had adipose tissue inflammation, whereas approximately 60% of subjects had no signs of
inflammation. Even though the two groups were identical for overall obesity and abdominal
subcutaneous adipose tissue (SAT) volume, those with inflamed adipose tissue had
approximately 30% greater VAT and 41% greater LF; 53% greater fasting insulin and 23%
lower β-cell function; and 22% higher TNF-α. (40). Another recent study by Bremer et al.
demonstrated that, among overweight and obese adults, those with metabolic syndrome had
significantly higher levels of infiltrating macrophages / CLS in their adipose tissue compared to
those without the metabolic syndrome (53). Given these observations, the disparities in
metabolic diseases among obese individuals, including AAs and Hispanics, may be explained by
the degree of chronic low-grade inflammation of adipose tissue. Therefore, targeting adipose
tissue inflammation has become an important new strategy in treating the metabolic conditions
! 14!
typically associated with obesity.

Decreasing Systemic and Adipose Tissue Inflammation
One of the most effective treatments for metabolic diseases associated with obesity is weight loss
(54). Studies in diet-induced obese mice have shown that reductions in adiposity result in
decreases in macrophage infiltration of adipose tissue as well as gene expression of pro-
inflammatory markers (55, 56). Among obese humans, caloric restriction to achieve weight loss
also decreased markers of inflammation. For example, a 28-day severe calorie restrictive diet
(800 kcal/day) among obese females reduced body weight by an average of 13 pounds and
decreased markers of inflammation (e.g., IL-12a, MMP-9) in adipose tissue (57). Studies
utilizing surgical methods to achieve significant weight loss also improve inflammation. For
example, one study observed a switch in the activation state of adipose tissue macrophages from
mostly pro-inflammatory to anti-inflammatory following ~44 pounds of weight loss resulting
from gastric bypass surgery (54). However, since weight loss is difficult to achieve and maintain,
identifying additional treatment strategies aimed at decreasing adipose tissue inflammation
presents a unique target in which to modify metabolic health.

On such approach has focused on dietary alterations to modulate systemic inflammation. This is
especially true since diets high in sugar (fructose, glucose) and fat have been shown to increase
systemic markers of inflammation such as hsCRP (58, 59). In addition to sugar intake, trans fat
consumption has been shown to be related to increased CRP and IL-6 levels (59). Conversely,
diets high in fiber have been shown to offer some therapeutic potential where 30 g/day decreased
CRP and IL-6 levels (60). Finally, recent studies in rodents have shown that omega-3
polyunsaturated fatty acids improves IR and decreases adipose tissue inflammation (61-63).
Despite these promising results, studies have largely shown that omega-3 supplementation does
not improve IR or adipose tissue inflammation in humans (64-66). Collectively, these studies
demonstrate that dietary changes have the ability to alter inflammation associated with obesity;
however, the direct mechanism and critical dietary components are known. Additionally, the
dietary changes necessary to elicit weight loss and decrease inflammation are likely to be too
drastic to sustain over time.

! 15!
Anti-Inflammatory Medications and Adipose Tissue Inflammation
Given the difficulty in decreasing obesity-associated inflammation through weight loss or dietary
interventions, studies have begun to look at alternative strategies for targeting inflammation. On
such approach involves the use of the non-steroidal anti-inflammatory drugs such as salsalate.
Salsalate, unlike aspirin, lacks an acetyl group and does not covalently modify COX enzymes
(32) but has been shown to inhibit NF-κB activity (67). For this reason, a handful of studies have
begun to investigate the utility of salsalate in regulating glycemia in predominately Caucasian
obese adults with impaired fasting glucose (IFG), impaired glucose tolerance (IGT), and/or T2D
(68-71).

Of these double-bind, placebo-controlled trials, numerous studies have shown that salsalate
treatment has the ability to decrease markers of inflammation and improve glucose control (68,
69, 71-73). In this regard, one study found that 4 g/day of salsalate treatment reduced fasting
glucose by 13%, glycemic response after oral glucose challenge by 20%, glycated albumin by
17%, and CRP by 34% (72). Notably, work by Goldfine et al. examined the effects of moderate
to high-dose salsalate treatment on glucose and lipid homeostasis (68, 69, 71). For example,
among those with T2D, 4 weeks of 4.5 g/day of salsalate had more pronounced beneficial effects
on fasting glucose, lipids, and systemic markers of inflammation compared to those receiving 3
g/day (68). Work by the same group found significant decreases in hemoglobin A1C levels with
3, 3.5, and 4 g/day of salsalate treatment over 14 weeks (69).

Potential side effects of high-dose salsalate treatment include hypoglycemia and tinnitus (68, 69,
71), which limit its utility in the prevention and/or treatment of diabetes. For this reason,
understanding the mechanism(s) by which salsalate exerts its beneficial metabolic effects may
lead to safer and more robust treatment strategies for the prevention and treatment of T2D and
CVD. Therefore, studies have begun to take a closer look at salsalate-induced changes in
inflammation, where reductions in NF-κB binding activity have been observed in peripheral
blood mononuclear cells (PBMC) (68) and adipose tissue (71). Taken together, these findings
suggest that salsalate exerts its protective cardiometabolic effects through decreases in the NF-
κB stress pathway. Findings from these studies are particularly important since they offer a
putatively safe and inexpensive treatment strategy for the prevention or reversal of
! 16!
cardiometabolic abnormalities observed during obesity. Despite this, only three studies have
examined salsalate treatment in those without T2D (72-74), and only one has examined adipose
tissue inflammation (71). Furthermore, no studies have examined the efficacy of this treatment
strategy in high-risk minority population (1).

Conclusion
Studies suggest that adipose tissue accumulation, specifically visceral and liver fat, contribute to
increased risk for T2D. Additionally, current work has identified adipose tissue inflammation as
largely contributing to the increased cardiometabolic disease risk seen during obesity. Therefore,
the first project of this dissertation focused on the location of adipose tissue and how it related to
diabetes risk among minority adolescents. Given that overweight and obese AAs have less VAT
and LF than Hispanics, yet similar risk for T2D, we were able to leverage data across five studies
in our laboratory in order to examine the independent contributions of each fat depot to T2D risk.
Project two focused on an anti-inflammatory drug, salsalate, in order to determine whether
treatment could decrease risk for T2D among non-diabetic obese Hispanics. Another central
component of this project was determining whether any metabolic improvements were mediated
by reductions in adipose tissue inflammation. In order to accomplish this, we developed a novel
subcutaneous abdominal adipose tissue biopsy technique that allowed us to consistently sample
large quantities of intact adipose tissue. Results from these studies contribute to the existing
literature by helping to elucidate the exact mechanisms linking excess adiposity to increased
disease risk, which is particularly important among high-risk AAs and Hispanics.

! 17!
Table 1-1. Key Background Points
KEY POINTS
• Ectopic fat accumulation, specifically visceral and liver fat, may link obesity with
risk for T2D.
• The AA vs. Hispanics paradox and the phenomenon of metabolically healthy
obesity suggest that other factors relating to adiposity may contribute to increased
disease risk.
• Excess adiposity is associated with inflammation in adipose tissue, which has been
hypothesized to mediate the link between obesity and risk for T2D. Therefore,
adipose tissue inflammation may serve as a therapeutic target for decreasing disease
risk.
• Salsalate, a non-steroidal anti-inflammatory drug, shows promise for preventing and
treating T2D; however, the exact mechanism(s) leading to improved glucose and
lipid homeostasis are unknown.
• Previous studies indicate that salsalate decreases inflammation through inhibition of
the NF-κB stress pathway, yet only one clinical trial has examined its effects on
adipose tissue inflammation.

! 18!
Chapter 1 References:

1. Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of Childhood and Adult Obesity
in the United States, 2011-2012. JAMA 2014;311:806.
2. Wang YC, McPherson K, Marsh T, Gortmaker SL, Brown M. Health and economic
burden of the projected obesity trends in the USA and the UK. Lancet 2011;378:815–825.
3. Mittendorfer B. Origins of metabolic complications in obesity: adipose tissue and free
fatty acid trafficking. Curr Opin Clin Nutr Metab Care 2011;14:535–541.
4. Byrne CD. Ectopic fat, insulin resistance and non-alcoholic fatty liver disease. Proc Nutr
Soc 2013;72:412–419.
5. Cruz ML, Bergman RN, Goran MI. Unique effect of visceral fat on insulin sensitivity in
obese Hispanic children with a family history of type 2 diabetes. Diabetes Care
2002;25:1631–1636.
6. Item F, Konrad D. Visceral fat and metabolic inflammation: the portal theory revisited.
Obes Rev 2012;13 Suppl 2:30–39.
7. Kantartzis K, Machann J, Schick F, Fritsche A, Häring H-U, Stefan N. The impact of liver
fat vs visceral fat in determining categories of prediabetes. Diabetologia 2010;53:882–889.
8. Tchernof A, Després JP. Pathophysiology of Human Visceral Obesity: An Update.
Physiological Reviews 2013;93:359–404.
9. Karelis AD, St-Pierre DH, Conus F, Rabasa-Lhoret R, Poehlman ET. Metabolic and body
composition factors in subgroups of obesity: what do we know? J Clin Endocrinol Metab
2004;89:2569–2575.
10. Shoelson SE, Herrero L, Naaz A. Obesity, Inflammation, and Insulin Resistance.
Gastroenterology 2007;132:1–12.
11. Cowie CC, Rust KF, Ford ES, et al. Full accounting of diabetes and pre-diabetes in the
! 19!
U.S. population in 1988-1994 and 2005-2006. Diabetes Care 2009;32:287–294.
12. Centers for Disease Control and Prevention (CDC). Differences in prevalence of obesity
among black, white, and Hispanic adults - United States, 2006-2008. MMWR Morb
Mortal Wkly Rep 2009;58:740–744.
13. Nazare J-A, Smith JD, Borel A-L, et al. Ethnic influences on the relations between
abdominal subcutaneous and visceral adiposity, liver fat, and cardiometabolic risk profile:
the International Study of Prediction of Intra-Abdominal Adiposity and Its Relationship
With Cardiometabolic Risk/Intra-Abdominal Adiposity. Am J Clin Nut 2012;96:714–726.
14. Fabbrini E, Magkos F, Mohammed BS, et al. Intrahepatic fat, not visceral fat, is linked
with metabolic complications of obesity. Proc Natl Acad Sci USA 2009;106:15430–15435.
15. Ravussin E, Smith SR. Increased fat intake, impaired fat oxidation, and failure of fat cell
proliferation result in ectopic fat storage, insulin resistance, and type 2 diabetes mellitus.
Ann NY Acad Sci 2002;967:363–378.
16. Bergman R. Free Fatty Acids and Pathogenesis of Type 2 Diabetes Mellitus. Trends
Endocrinol Metab 2000;11:351–356.
17. Randle PJ, Garland PB, Hales C, Newsholme E. The glucose fatty-acid cycle. Its role in
insulin sensitivity and the metabolic disturbances of diabetes mellitus. Lancet 1963;1:785–
789.
18. Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of obesity and trends in body
mass index among US children and adolescents, 1999-2010. JAMA 2012;307:483–490.
19. Dabelea D, Bell RA, D'Agostino RB, et al. Incidence of diabetes in youth in the United
States. JAMA 2007;297:2716–2724.
20. Hasson RE, Adam TC, Davis JN, et al. Ethnic differences in insulin action in obese
African-American and Latino adolescents. J Clin Endocrinol Metab 2010;95:4048–4051.
21. Toledo-Corral CM, Alderete TL, Hu HH, et al. Ectopic fat deposition in prediabetic
! 20!
overweight and obese minority adolescents. J Clin Endocrinol Metab 2013;98:1115–1121.
22. Goran MI. Ethnic-specific pathways to obesity-related disease: the Hispanic vs. African-
American paradox. Obesity (Silver Spring) 2008;16:2561–2565.
23. Goran MI, Bergman RN, Cruz ML, Watanabe R. Insulin resistance and associated
compensatory responses in African-American and Hispanic children. Diabetes Care
2002;25:2184–2190.
24. Goran MI, Gower BA. Relation between visceral fat and disease risk in children and
adolescents. Am J Clin Nut 1999;70:149S–56S.
25. Vague J, Vague P, Jubelin J. Fat distribution, obesities and health: evolution of concepts.
Curr Top Nut Dis 1988;17:9–41.
26. Bennett B, Larson-Meyer DE, Ravussin E, et al. Impaired Insulin Sensitivity and Elevated
Ectopic Fat in Healthy Obese vs. Nonobese Prepubertal Children. Obesity (Silver Spring)
2011;20:371–375.
27. Jakobsen M, Berentzen T, Sorensen T, Overvad K. Abdominal Obesity and Fatty Liver.
Epidemiologic Reviews 2007;29:77–87.
28. Korenblat KM, Fabbrini E, Mohammed BS, Klein S. Liver, muscle, and adipose tissue
insulin action is directly related to intrahepatic triglyceride content in obese subjects.
Gastroenterology 2008;134:1369–1375.
29. Fabbrini E, Tamboli RA, Magkos F, et al. Surgical Removal of Omental Fat Does Not
Improve Insulin Sensitivity and Cardiovascular Risk Factors in Obese Adults.
Gastroenterology 2010;139:448–455.
30. Hotamisligil GS, Shargill NS, Spiegelman BM. Adipose expression of tumor necrosis
factor-alpha: direct role in obesity-linked insulin resistance. Science 1993;259:87–91.
31. Lamers D, Famulla S, Wronkowitz N, et al. Dipeptidyl peptidase 4 is a novel adipokine
potentially linking obesity to the metabolic syndrome. Diabetes 2011;60:1917–1925.
! 21!
32. Shoelson SE, Lee J, Goldfine AB. Inflammation and insulin resistance. J Clin Invest
2006;116:1793–1801.
33. Carvalho-Filho MA, Ropelle ER, Pauli RJ, et al. Aspirin attenuates insulin resistance in
muscle of diet-induced obese rats by inhibiting inducible nitric oxide synthase production
and S-nitrosylation of IRbeta/IRS-1 and Akt. Diabetologia 2009;52:2425–2434.
34. Pou KM, Massaro JM, Hoffmann U, et al. Visceral and Subcutaneous Adipose Tissue
Volumes Are Cross-Sectionally Related to Markers of Inflammation and Oxidative Stress:
The Framingham Heart Study. Circulation 2007;116:1234–1241.
35. Hotamisligil GS. Inflammation and metabolic disorders. Nature 2006;444:860–867.
36. Lumeng CN, Saltiel AR. Inflammatory links between obesity and metabolic disease. J
Clin Invest 2011;121:2111–2117.
37. Fried SK, Bunkin DA, Greenberg AS. Omental and subcutaneous adipose tissues of obese
subjects release interleukin-6: depot difference and regulation by glucocorticoid. J Clin
Endocrinol Metab 1998;83:847–850.
38. Keaney JF, Larson MG, Vasan RS, et al. Obesity and systemic oxidative stress: clinical
correlates of oxidative stress in the Framingham Study. Arterioscler Thromb Vasc Biol
2003;23:434–439.
39. Feinstein R, Kanety H, Papa MZ, Lunenfeld B, Karasik A. Tumor necrosis factor-alpha
suppresses insulin-induced tyrosine phosphorylation of insulin receptor and its substrates.
J Biol Chem 1993;268:26055–26058.
40. Lê K-A, Mahurkar S, Alderete TL, et al. Subcutaneous adipose tissue macrophage
infiltration is associated with hepatic and visceral fat deposition, hyperinsulinemia, and
stimulation of NF-κB stress pathway. Diabetes 2011;60:2802–2809.
41. Furuhashi M, Fucho R, Görgün CZ, Tuncman G, Cao H, Hotamisligil GS.
Adipocyte/macrophage fatty acid–binding proteins contribute to metabolic deterioration
through actions in both macrophages and adipocytes in mice. J Clin Invest
! 22!
2008;118:2640–2650.
42. Lumeng CN, DelProposto JB, Westcott DJ, Saltiel AR. Phenotypic Switching of Adipose
Tissue Macrophages With Obesity Is Generated by Spatiotemporal Differences in
Macrophage Subtypes. Diabetes 2008;57:3239–3246.
43. Zanni MV, Burdo TH, Makimura H, Williams KC, Grinspoon SK. Relationship between
monocyte/macrophage activation marker soluble CD163 and insulin resistance in obese
and normal-weight subjects. Clin Endocrinol (Oxf) 2012;77:385–390.
44. Aristoteli LP, Møller HJ, Bailey B, Moestrup SK, Kritharides L. The monocytic lineage
specific soluble CD163 is a plasma marker of coronary atherosclerosis. Atherosclerosis
2006;184:342–347.
45. Moreno JA, Muñoz-García B, Martín-Ventura JL, et al. The CD163-expressing
macrophages recognize and internalize TWEAK. Atherosclerosis 2009;207:103–110.
46. Subramanian S, Tawakol A, Burdo TH, et al. Arterial inflammation in patients with HIV.
JAMA 2012;308:379–386.
47. Ito H, Ohshima A, Inoue M, et al. Weight reduction decreases soluble cellular adhesion
molecules in obese women. Clin Exp Pharmacol Physiol 2002;29:399–404.
48. Haught WH, Mansour M, Rothlein R, et al. Alterations in circulating intercellular
adhesion molecule-1 and L-selectin: further evidence for chronic inflammation in
ischemic heart disease. Am Heart J 1996;132:1–8.
49. Ross R. Atherosclerosis--an inflammatory disease. N Engl J Med 1999;340:115–126.
50. Leinonen E, Hurt-Camejo E, Wiklund O, Hultén LM, Hiukka A, Taskinen M-R. Insulin
resistance and adiposity correlate with acute-phase reaction and soluble cell adhesion
molecules in type 2 diabetes. Atherosclerosis 2003;166:387–394.
51. Matsumoto K, Sera Y, Nakamura H, Ueki Y, Miyake S. Serum concentrations of soluble
adhesion molecules are related to degree of hyperglycemia and insulin resistance in
! 23!
patients with type 2 diabetes mellitus. Diabetes Res Clin Pract 2002;55:131–138.
52. McLaughlin T, Abbasi F, Cheal K, Chu J, Lamendola C, Reaven G. Use of metabolic
markers to identify overweight individuals who are insulin resistant. Ann Intern Med
2003;139:802–809.
53. Bremer AA, Devaraj S, Afify A, Jialal I. Adipose tissue dysregulation in patients with
metabolic syndrome. J Clin Endocrinol Metab 2011;96:E1782–E1788.
54. Aron-Wisnewsky J, Tordjman J, Poitou C, et al. Human adipose tissue macrophages: m1
and m2 cell surface markers in subcutaneous and omental depots and after weight loss. J
Clin Endocrinol Metab 2009;94:4619–4623.
55. Kosteli A, Sugaru E, Haemmerle G, et al. Weight loss and lipolysis promote a dynamic
immune response in murine adipose tissue. J Clin Invest 2010;120:3466–3479.
56. Vieira VJ, Valentine RJ, Wilund KR, Antao N, Baynard T, Woods JA. Effects of exercise
and low-fat diet on adipose tissue inflammation and metabolic complications in obese
mice. Am J Physiol Endocrinol Metab 2009;296:E1164–1171.
57. Clément K, Viguerie N, Poitou C, et al. Weight loss regulates inflammation-related genes
in white adipose tissue of obese subjects. FASEB J 2004;18:1657–1669.
58. Aeberli I, Gerber PA, Hochuli M, et al. Low to moderate sugar-sweetened beverage
consumption impairs glucose and lipid metabolism and promotes inflammation in healthy
young men: a randomized controlled trial. Am J Clin Nutr 2011;94:479–485.
59. Lopez-Garcia E, Schulze MB, Meigs JB, et al. Consumption of trans fatty acids is related
to plasma biomarkers of inflammation and endothelial dysfunction. J Nutr 2005;135:562–
566.
60. King DE, Egan BM, Woolson RF, Mainous AGI, Al-Solaiman Y, Jesri A. Effect of a
High-Fiber Diet vs a Fiber-Supplemented Diet on C-Reactive Protein Level. Arch Intern
Med 2007;167:502.
! 24!
61. Todoric J, Löffler M, Huber J, et al. Adipose tissue inflammation induced by high-fat diet
in obese diabetic mice is prevented by n-3 polyunsaturated fatty acids. Diabetologia
2006;49:2109–2119.
62. Kalupahana NS, Claycombe K, Newman SJ. Eicosapentaenoic acid prevents and reverses
insulin resistance in high-fat diet-induced obese mice via modulation of adipose tissue
inflammation. J Nutr 2010;140:1915–1922.
63. Saraswathi V, Gao L, Morrow JD, Chait A, Niswender KD, Hasty AH. Fish oil increases
cholesterol storage in white adipose tissue with concomitant decreases in inflammation,
hepatic steatosis, and atherosclerosis in mice. J Nutr 2007;137:1776–1782.
64. Itariu BK, Zeyda M, Hochbrugger EE, et al. Long-chain n-3 PUFAs reduce adipose tissue
and systemic inflammation in severely obese nondiabetic patients: a randomized
controlled trial. Am J Clin Nut 2012;96:1137–1149.
65. Kabir M, Skurnik G, Naour N, et al. Treatment for 2 mo with n 3 polyunsaturated fatty
acids reduces adiposity and some atherogenic factors but does not improve insulin
sensitivity in women with type 2 diabetes: a randomized controlled study. Am J Clin Nut
2007;86:1670–1679.
66. Kratz M, Kuzma JN, Hagman DK, et al. n3 PUFAs do not affect adipose tissue
inflammation in overweight to moderately obese men and women. J Nutr 2013;143:1340–
1347.
67. Kopp E, Ghosh S. Inhibition of NF-kappa B by sodium salicylate and aspirin. Science
1994;265:956–959.
68. Goldfine AB, Silver R, Aldhahi W, et al. Use of salsalate to target inflammation in the
treatment of insulin resistance and type 2 diabetes. Clin Transl Sci 2008;1:36–43.
69. Goldfine AB, Fonseca V, Jablonski KA, et al. The effects of salsalate on glycemic control
in patients with type 2 diabetes: a randomized trial. Ann Intern Med 2010;152:346–357.
70. Faghihimani E, Aminorroaya A, Rezvanian H, Adibi P, Ismail-Beigi F, Amini M.
! 25!
Salsalate improves glycemic control in patients with newly diagnosed type 2 diabetes.
Acta Diabetol 2013;50:537–543.
71. Goldfine AB, Conlin PR, Halperin F, et al. A randomised trial of salsalate for insulin
resistance and cardiovascular risk factors in persons with abnormal glucose tolerance.
Diabetologia 2013;56:714–723.
72. Fleischman A, Shoelson SE, Bernier R, Goldfine AB. Salsalate improves glycemia and
inflammatory parameters in obese young adults. Diabetes Care 2008;31:289–294.
73. Koska J, Ortega E, Bunt JC, et al. The effect of salsalate on insulin action and glucose
tolerance in obese non-diabetic patients: results of a randomised double-blind placebo-
controlled study. Diabetologia 2009;52:385–393.
74. Faghihimani E, Aminorroaya A, Rezvanian H, Adibi P, Ismail-Beigi F, Amini M.
Reduction of insulin resistance and plasma glucose level by salsalate treatment in persons
with prediabetes. Endocr Pract 2012;18:826–833.

! 26
Chapter 2: Liver Fat Has a Stronger Association with Risk Factors for Type 2 Diabetes in
African Americans Compared with Hispanic Adolescents
Alderete TL, Toledo-Corral CM, Desai PD, Weigensberg MJ, Goran MI. Liver Fat Has a
Stronger Association with Risk Factors for Type 2 Diabetes in African-American Compared with
Hispanic Adolescents. J Clin Endocrinol Metab 2014;98:3748-3754.

Abstract
Context: Although overweight and obese African-Americans (AAs) have less visceral adipose
tissue (VAT) and liver fat (LF) than Hispanics, they have a similar risk for type 2 diabetes.
Objective: We examined ethnic differences in the association between VAT and LF with risk
factors for type 2 diabetes to help explain this paradox.
Design: We conducted a cross-sectional study in an academic pediatric care facility.
Subjects: Subjects were overweight and obese AA (n = 131; 15.5 ± 3.3 years old) and Hispanic
adolescents (n = 227; 14.7 ± 3.0 years old).
Main Outcome Measures: Outcome measures included insulin sensitivity (SI), acute insulin
response (AIR), and disposition index (DI) by frequently sampled iv glucose tolerance test and
minimal modeling.
Results: LF, not VAT, was inversely associated with SI, and the effect of high LF compared
with low was more pronounced in AAs (P
interaction
< .05). In Hispanics, high LF was
associated with a 24% lower SI (P < .01) and a 31% increase in AIR (P < .01) and was not
associated with DI (P = .35). In AAs, high LF was associated with a 49% lower SI (P < .001),
was not associated with an increase in AIR (P = .25), and was associated with a 42% lower DI (P
< .01), indicating failure of compensatory insulin secretion/clearance in response to insulin
resistance. Prediabetes changed the relationship between high/low LF and DI in Hispanics
(P
interaction
= .002) but not AAs such that prediabetic Hispanics with high LF had a 43% lower
DI (P = .03) with no difference in those without prediabetes (P = .06).
Conclusions: LF has a stronger effect on SI compared with VAT. Our results suggest that the
impact of high LF on poor β-cell compensation is more pronounced in AAs. In Hispanics, the
combination of high LF and prediabetes contributes to poor β-cell compensation.
! 27
Introduction
Studies show that for the same degree of overall adiposity, African-Americans (AA) have higher
levels of subcutaneous abdominal adipose tissue (SAAT) and lower levels of visceral adipose
tissue (VAT) and liver fat (LF) compared to Hispanics (1, 2). However, despite having lower
levels of VAT and LF, AA are more insulin resistant than Hispanics (1) and have similar or
increased risk for type 2 diabetes (T2D) (1-3). These observations conflict with the prevailing
hypothesis that higher VAT (4-6) and/or LF (7, 8) contribute to obesity-associated metabolic
disease risk. Given this, it is possible that VAT and LF contribute differently to metabolic risk in
AA and Hispanics. Differences in the impact of either fat depot across ethnicity may explain why
AA exhibit similar risk factors for T2D despite having less VAT and LF compared to Hispanics.

Although AA have a more protective fat profile when compared to Hispanics, they have a
comparable risk for T2D, establishing what we have termed the AA versus Hispanic paradox (2).
These ethnic differences are even more intriguing given that previous studies have shown that
increased VAT (2, 3, 5, 9) and LF (4, 10) are associated with increased T2D risk. Additionally,
recent work by our group has show that prediabetic minority adolescents have higher levels of
VAT and LF than those who are normal glucose tolerant, suggesting prediabetes may modify the
association between these fat depots and risk for T2D (11). A major limitation of these studies is
the high correlation between VAT and LF (12, 13), making it difficult to determine which
compartment of abdominal fat drives metabolic risk. Elegant studies, mostly in Caucasian adults,
provide evidence that LF is the predominant compartment associated with metabolic disease in
obese adults (7, 8). Therefore, the two primary connected goals of this study were to examine 1)
the separate and combined effects of VAT and LF on risk factors for T2D in overweight and
obese minorities and 2) to determine whether the association between these fat depots and risk
factors for T2D differed in AA and Hispanics, thereby potentially explaining the paradox of
similar diabetes risk in AA at lower levels of VAT and LF. As a secondary aim, we explored
whether prediabetes status altered the associations between these fat depots and risk factors for
T2D.

! 28
Methods:
Participants
We combined participants from 5 studies in our laboratory that used a common protocol for
assessment of body fat distribution and risk factors for T2D. Participants included 358
overweight and obese (BMI ≥85
th
percentile for ages <18 years or BMI >30 kg/m
2
for those ≥18
years) AA and Hispanics (131 AA / 227 Hispanics) aged 8-25 years who had complete measures
of subcutaneous abdominal adipose tissue (SAAT), VAT and LF. Ethnicity was defined as self,
parents, and grandparents being all of AA or Hispanic decent (by parental / self-report). Data
from some of these studies have been reported (1, 11, 14-16); however, this is the first combined
analysis to examine the independent effect of LF and VAT on metabolic outcomes in AA and
Hispanics who differ in their LF and VAT profiles. Prior to any testing, informed written consent
/ assent was obtained from the participant or parents. All studies were approved by the
University of Southern California Institutional Review Board.

Procedures
Procedures used in these studies have been previously reported (1, 11, 14-16). Total body fat
mass was assessed by dual-energy X-ray absorptiometry (Hologic QDR 4500W; Hologic,
Bedord, MA). Depending on machine availability, a 30-slice whole abdominal magnetic
resonance imaging (MRI) scan was performed on a General Electric 1.5-Tesla or 3-Tesla (T)
magnet in order to measure abdominal fat distribution and ectopic fat accumulation (17, 18) as
previously described (19). SAAT and VAT images were segmented using SliceOmatic
(TomoVision, Montreal, Canada). LF was measured during the same MRI test using a modified
Dixon three-point technique. For a subset of participants (N=37), we performed consecutive MRI
scans using both 1.5 and 3-T magnets and found that these two methods were highly correlated
for measures of SAAT, VAT, and LF (r=0.90, r=0.95, and r=0.96, respectively). Therefore, we
used this data to developed conversion equations to standardize data relative to the 1.5-T magnet
for each participant.

Metabolic Parameters
All but 34 participants had an overnight stay at the Clinical Trials Unit (CTU) and were fed a
standardized dinner prior to the frequently sampled intravenous glucose tolerance test
! 29
(FSIVGTT) the following morning. The remaining 34 participants arrived at the CTU for their
FSIVGTT following an overnight fast. After the FSIVGTT, all participants were fed a
standardized lunch (1, 11, 14-16). Plasma was analyzed for glucose and insulin and values
entered into Minmod Millenium 2003 (versions 5.16 and 6.02, Richard N. Bergman, USC) to
generate values for insulin sensitivity (SI), acute insulin response to glucose (AIR), and
disposition index (DI; the product of SI and AIR). Previous studies have shown that decreases in
DI is predictive of the development of T2D in Pima Indians as well as women with gestational
diabetes (20, 21). In addition to the FSIVGTT, a subset of our participants (n=299) underwent an
OGTT as previously described (11, 14, 22).
Statistical Analysis
Definition of high/low liver fat (LF), high/low visceral adipose tissue (VAT), and Normal
Glucose Tolerant/Prediabetic. Participants were classified into one of four groups based on:
high/low LF and high/low VAT. High LF was defined as an LF fraction >5.5% while low LF
was defined as an LF fraction <5.5%. We utilized this cut-point since it has been shown to be
likely indicative of fatty liver disease (23). High or low VAT, relative to SAAT, was defined by
positive or negative residuals from the regression between VAT and SAAT in each ethnic group.
This approach is supported by previous findings which found that in obese adolescents high
VAT, in the context of low SAAT, was associated with a greater disease risk (24). Participants
who underwent an OGTT were classified as normal glucose tolerant (NGT) (fasting glucose
<100-125 mg/dL and 2-hour glucose <140 mg/L or 2) or prediabetic (fasting glucose ≥100
mg/dL and/or 2-hour glucose of 140-199 mg/dL).

The standardized measures of metabolic indices, SAAT, VAT, and LF values for 358
participants are shown in Table 2-1. Analyses were done in SPSS Statistics version 18.0 and a
priori significance level was set at P<0.05. A full-factorial three-way analysis of covariance
(ANCOVA) was used to examine interactions between our main effect variables (high/low LF,
high/low VAT, ethnicity, and prediabetes status where appropriate) and metabolic outcomes (SI,
AIR, and DI). All three-way interactions were non-significant; therefore, we examined two-way
interactions between high/low LF, high/low VAT, ethnicity, and prediabetes status (where
appropriate). If an interaction term, which included ethnicity, was found to be significant, we
stratified our sample and reported the separate adjusted means for AA and Hispanics. A priori
! 30
covariates include age, sex, and total fat mass (TFM) as well as ethnicity, high/low LF, and
high/low VAT. Controlling for total percent fat mass instead of TFM, or including Tanner stage
as a covariate did not change our results (data not shown). Repeating these analyses in the subset
of participants who underwent an OGTT (n=299) yielded nearly identical results (data not
shown). We also examined the relationships between LF and VAT with SI, AIR, and DI using
multivariate linear regression analysis. Each model controlled for age, sex, TFM, and SAAT.
Ethnicity was included in each model and we examined interactions between LF, VAT, and
ethnicity. For all of the analyses, SI, AIR, and DI were log transformed in order to meet
assumptions. In some cases, results were back-transformed for presentation and reported as mean
± standard error or adjusted mean [95% confidence interval (CI)].

Results:
Table 2-1 displays the mean physical characteristics and metabolic parameters of the 131 AA
and 227 Hispanic males and females. Consistent with our previous studies, AA had lower LF and
VAT compared to Hispanics (P<0.01). Although the unadjusted values of SI were not different
by ethnicity, after controlling for age, sex, and TFM, AA had a lower SI (1.82 ± 0.11 vs. 2.01 ±
0.09 x10
4
min
-1
/µU/ml; P=0.01). LF was correlated with VAT in AA and Hispanics after
controlling for TFM (AA: r=0.27; P<0.01 and Hispanics: r=0.44; P<0.001).

High/Low LF and High/Low VAT Analysis. In regards to SI, AIR, and DI, the interaction
between high/low LF and ethnicity was significant in each model (P
LF*Ethnicity
<0.05) while the
interaction between high/low LF and high/low VAT was not (P
LF*VAT
≥0.18). Figure 2-1 shows
these results stratified by ethnicity. Here we hereafter report the adjusted means for AA and
Hispanics wherein the models do not include the interaction between high/low LF and high/low
VAT. High LF was associated with a 49% lower SI in AA (high LF: 0.77 [(x 10
-4
min
-
1
)/(µU/mL)] [0.58, 1.03] vs. low LF: 1.52 [1.35, 1.71]; P<0.001) and a 24% lower SI in
Hispanics (high LF: 1.39 [(x 10
-4
min
-1
)/(µU/mL)] [1.23, 1.57] vs. low LF: 1.83 [1.63, 2.05];
P<0.01). In Hispanics, high LF was associated with a 31% higher AIR (high LF: 1,264.74
µU/mL x 10 min [1,106.62, 1,445.44] vs. low LF: 963.83 [849.18, 1,091.44]; P<0.01), indicating
appropriate compensation as also demonstrated by no association of high LF with DI (high LF:
1,815.52 [1,595.88, 2,065.38] vs. low LF: 1,659.59 [1,472.32, 1,874.99]; P=0.35). In AA
! 31
however, there was no association of high LF with AIR, suggesting poor compensation as also
demonstrated by the 42% lower DI in those with high LF (high LF: 1,352.07; [1,000.00,
1828.10] vs. low LF: 2,328.09 [2,051.16, 2,648.50]; P<0.01).

Liver Fat and Visceral Fat Regression Analysis. As shown in Figure 2-2, we observed that
there was a stronger negative association between LF and SI in AA compared to Hispanics
(P
LF*Ethnicity
<0.01; β
AA
=-0.53; P=0.001 vs. β
Hisp
=-0.21; P=0.003). In AA and Hispanics there was
no relationship between LF and AIR while we did observe an ethnic difference in the
relationship between VAT and AIR (P
VAT*Ethnicity
<0.001). Specifically, in AA there was no
relationship between VAT and AIR, while in Hispanics there was a significant positive
relationship between VAT and AIR (β
AA
=0.03; P=0.55 vs. β
Hisp
=0.06; P=0.01). In AA, with
increasing LF, there was a significantly lower DI while this relationship was not observed in
Hispanics (P
VAT*Ethnicity
=0.04;

β
AA
=-0.38; P=0.007 vs. β
Hisp
=-0.02; P=0.80). There was no
relationship between VAT and SI or VAT and DI in AA or Hispanics.

High/Low LF, High/Low VAT, and Prediabetes Status Analysis. Eighty-four percent of the
participants had an OGTT, which we used to identify 15 AA and 48 Hispanics as being
prediabetic. Controlling for prediabetes status did not change our findings regarding SI and AIR
(data not shown). As shown in Figure 2-3, when examining DI we found that prediabetes status
significantly changed the relationship between high/low LF and DI in Hispanics
(P
LF*Prediabetes
=0.002) but not AA (P
LF*Prediabetes
=0.24). NGT Hispanics with high LF showed a
trend for a 26% higher DI (high LF: 2,084.49; [1,778.28, 2,454.71] vs. low LF: 1,655.77
[1,412.54, 1,949.84]; P=0.06) while prediabetic Hispanics with high LF had a 43% lower DI
(high LF: 1,088.93
1
; [790.68, 1,496.24] vs. low LF: 1,909.85 [1,361.44, 2,673.00]; P=0.03). In
NGT and prediabetic AA, high LF was associated with a 35% lower DI (high LF: 1,244.51;
[895.36, 1,729.82] vs. low LF: 1,905.46 [1,548.82, 2,338.84]; P=0.02).

Discussion
Elevated levels of VAT and LF have been shown to be associated with insulin resistance (1, 4, 7,
8, 25) while some previous studies have demonstrated the importance of LF over VAT in regards
to increased risk for T2D (7, 8). Given this, overweight and obese AA with lower LF and VAT,
! 32
should be protected from T2D risk when compared to Hispanics. Despite this, previous work by
our group has shown AA have higher or similar risk factors for T2D (1-3). Therefore, the
objective of this analysis was to determine the separate effects of LF and VAT on risk factors for
T2D and to see whether these relationships differed between AA and Hispanics. In order to
accomplish this, we examined LF and VAT by identifying sub-groups that were contrasted for
high versus low LF and high versus low VAT. Using this method, we demonstrated that high LF
was associated with increased insulin resistance while high VAT was not. This association was
more prominent in AA compared to Hispanics, suggesting that even though LF tends to be lower
in AA, its relationship with insulin resistance is more pronounced.

By examining those with high and low LF and VAT, we found that there was a stronger
association between high LF and low SI in AA compared to Hispanics. In AA we observed that
high LF was associated with a 49% lower SI when compared to those with low LF. Therefore,
based on the hyperbolic relationship between SI and AIR, we anticipated that AA with high LF
would have to compensate with an AIR that was double that of AA with low LF. Despite this, we
found that AA with high LF only had a 22% higher AIR. Conversely, in Hispanics those with
high compared to low LF had a 31% higher AIR, which was in line with the 24% lower SI that
we observed among those with high LF. These findings suggest that in response to a lower SI,
Hispanics but not AA are able to compensate for LF induced insulin resistance by adequately
increasing their AIR. Taken together, we found that in AA high LF was associated with a 49%
lower SI and a small and non-significant increase in their AIR, which translated into high LF
being associated with a 42% lower DI. In Hispanics, high LF was associated with a 24% lower
SI and an appropriate increase in AIR, which resulted in no relationship between high/low LF,
high/low VAT and DI.

In order to further explore the relationship between LF and VAT with risk factors for T2D, we
also used regression analyses to complement the approach of classifying participants based on
high/low LF and VAT. Overall, our results were similar using the two approaches and we found
that in AA there was a stronger inverse relationship between LF and SI when compared to
Hispanics. We also found that VAT, and not LF, was positively related to AIR in Hispanics but
not AA. In AA, LF was inversely associated with a lower DI while this relationship was not
! 33
observed in Hispanics. When examining our four groups, we found that high LF and high VAT
were associated with a higher AIR in Hispanics whereas when we examined LF and VAT
continuously, we observed that only VAT was related to a higher AIR. This difference is likely
due to the fact that these two analyses differed in the manner in which they examined LF, VAT,
and SAAT. A previous study has shown that VAT relative to SAAT is important when
considering metabolic disease risk (24). Therefore, it is possible that our high/low classification
method for VAT relative to SAAT is better suited to examine the relative importance of VAT
compared to LF. Furthermore, it is important to note that our regression analysis largely supports
our findings regarding the differential effects of high LF on risk factors for T2D in AA and
Hispanics.

Recent work by our group has show that minority adolescents with prediabetes differ in their
ectopic fat distribution and risk for T2D when compared to those without prediabetes (11).
Additionally, another study in overweight adults demonstrated that LF, more than VAT,
increases when glucose tolerance moves from normal to impaired fasting glucose/impaired
glucose tolerance (26). For this reason, we explored prediabetes status in the current study and
found that it changed the results for DI but not SI or AIR. Specifically, NGT Hispanics with high
LF showed a trend for a 26% higher DI while prediabetic Hispanics with high LF had a 43%
lower DI. In AA, including prediabetes status in the model resulted in high LF being associated
with a 35% lower DI. These findings suggest that the combination of high LF with prediabetes is
particularly problematic for beta-cell function. Specifically, high LF was associated with poor
beta-cell function only in Hispanics with prediabetes. It is important to note that this secondary
analysis is greatly limited by the small number of participants with prediabetes in this study (15
AA and 48 Hispanics).

Although it is impossible to determine causality from this study, we would postulate that AA and
Hispanics differ in their beta-cell compensation in response to LF induced insulin resistance. We
would hypothesize that Hispanics with high LF are able to compensate for a lower SI by
increasing their AIR; thereby showing appropriate beta-cell compensation to LF induced insulin
resistance. Conversely, AA would fail to have an adequate increase in their AIR, thus showing
poor beta-cell compensation to LF induced insulin resistance. In light of these findings, we were
! 34
left with the question of why LF might have a more prominent association with risk factors for
T2D in AA compared to Hispanics. One explanation is that AA already have a higher AIR when
compared to Hispanics. Due to this, their beta-cells could have a harder time compensating for
the added stress of LF induced insulin resistance because they are already on the steeper part of
the DI curve. It is also conceivable that AA and Hispanics not only differ in their ectopic fat
distribution, but also in the amount of adipokines that these fat depots secrete. This is particularly
important given that leptin and adiponectin have been shown to affect insulin secretion and beta-
cell apoptosis (27). Finally, another possible explanation is a high correlation between LF and
pancreatic fat (PF) (28), which may impair beta-cell compensation to insulin resistance. One of
our previous studies found that obese pre-diabetic AA have higher levels of PF than those with
normal glucose tolerance while this relationship was not observed in Hispanics (11). Given these
findings, it is possible that PF interferes with insulin secretion from beta-cells and may be why
AA show poor beta-cell compensation.

One limitation of this study is that we enrolled only overweight and obese AA and Hispanics and
did not include lean participants or Caucasians. Although we observed that AA had a lower
fasting insulin and higher fasting glucose compared to Hispanics, a previous study by our group
has demonstrated that fasting indicators of insulin sensitivity are not always reflective of SI from
an FSIVGTT with minimal modeling (29). Also, in the current study, we found that after
adjusting for age, sex, and TFM, AA had a lower SI than Hispanics. Another limitation of this
study is that, due to the low prevalence of high LF and VAT in AA compared to Hispanics (2),
we observed a small number of AA with high LF (n=19). For this reason, we completed a power
analysis and found that we were powered to detect up to a 39% higher AIR in AA with high
compared to low LF. As previously mentioned, given the 49% lower SI in AA with high LF, we
expected an AIR that was double that of those with low LF, which we were more than powered
to detect. Additionally, despite the small number of AA with high LF, we were able to confirm
these results using regression analyses.

This study is also limited in that we did not have data regarding hepatic insulin sensitivity,
hepatic insulin clearance, or C-peptide measures, making it possible that our DI findings reflect
changes in hepatic insulin clearance and/or beta-cell secretion of insulin in response to decreased
! 35
SI. We also can not determine whether our observed SI, primarily based on muscle, is a
consequence of chronic hyperinsulinemia induced by hepatic insulin resistance and/or reduced
insulin clearance due to high LF. For example, since Hispanics had higher insulin levels and a
smaller AIR compared to AA, it is possible that impaired hepatic insulin clearance could
partially explain our findings. Lastly, since race/ethnicity was defined based on self-report,
future studies should verify these finding by using a more definitive measure of ethnicity.

In summary, we found that LF, more than VAT, was associated with insulin resistance in AA
and Hispanics and that these effects are more pronounced in AA. Additionally, we found that
there was a greater association of LF with risk factors for T2D in AA as seen by compromised
beta-cell compensation to the LF induced insulin resistance in this group. Additionally, we found
that the combination of high LF and prediabetes was associated with decreased beta-cell function
in Hispanics but not AA. This is the first study to demonstrate that high LF has a more profound
effect on SI, AIR, and DI among AA compared to Hispanics. This finding could explain why AA
are at a similar risk for T2D despite having lower levels of LF and VAT (2) and suggest that
even though LF tends to be lower in AA, its association with metabolic risk is more pronounced.

Acknowledgements
This work was supported by National Institutes of Health (NIH) grants R01-HD033064-10
(M.I.G), R01-DK059211 (M.I.G), National Center on Minority Health and Health Disparities
P60-MD002254 (M.I.G.); General Clinical Research Center (GCRC) for Health Resources grant
(M01-RR 00043); and the Robert C. and Veronica Atkins Foundation Grant (M.I.G). Mrs.
Alderete was involved in data collection and was responsible for data analysis and manuscript
preparation. Dr. Toledo-Corral assisted with data collection, evaluation of statistical analysis,
and manuscript preparation. Ms. Desai assisted with data collection and manuscript preparation.
Dr. Weigensberg and Dr. Goran (principal investigators) supervised all aspects of the study. No
conflicts of interest existed for any of the authors. We would like to thank the nursing staff at the
Clinical Trials Unit as well as our participants and their families for their involvement.

! 36
Table 2-1. Baseline Descriptive Statistics
a
Unless indicated otherwise, result are shown as mean ± SD.
b
Nonparametric test.
c
Chi-Square test.
d
AA, n=128 and Hispanics, n=217.
e
Variables were not normally distributes so statistical tests were run on log-transformed data.
f
AAs, n=125 and Hispanics, n=169.
g
AAs, n=128 and Hispanics, n=190.
Variable AAs (n=131) Hispanics (n=227) P Value
General Characteristics
Age, y
b
15.5 ± 3.3 14.7 ± 3.0 .07
Pubertal stage, n
c

1-3 23 69 <.001
4 14 44
5 94 114
Sex (males/females), n
c
61/70 131/96 .04
Height, cm 164.2 ± 12.2 161.4 ± 11.6 .04
Weight, kg 89.7 ± 26.7 84.6 ± 22.8 .07
Body Composition
BMI, kg/m
2b
32.7 ± 7.7 32.1 ± 6.5 .74
Total fat mass, kg
b,d
31.6 ± 13.2 31.0 ± 11.1 .55
Total Lean Tissue Mass, kg
d
52.8 ± 14.4 49.0 ± 12.3 .01
Subcutaneous Fat, L
b
14.6 ± 6.3 13.7 ± 6.2 .29
Visceral Fat, L
b
1.5 ± 1.0 2.0 ± 1.3 <.01
Liver Fat Fraction, %
b
4.2 ± 3.6 9.0 ± 8.5 <.001
Metabolic Parameters
Fasting Insulin (µU/mL)
e,f
15.7 ± 10.0 26.9 ± 24.0 <.001
Fasting Glucose (mg/dL)
b,f
87.3 ± 6.7 83.0 ± 24.1 <.01
SI, [(x10
-4
min
-1
/(µU/mL)]
e,g
1.8 ± 1.4 1.9 ± 1.4 .14
AIR, (µU/mL x 10 min)
e,g
1,944.2 ± 1,326.7 1,326.7 ± 823.2 <.001
DI
b,g
2,626.9 ± 1,716.1 1,972.9 ± 1,044.4 .03
Prediabetes Status
c

NGT, n 90 146 .03
Prediabetic, n 15 48
! 37
Figure 2-1. High/low LF and high/low VAT Analysis by Ethnicity
Figure 2-1. A-C, Adjusted mean ± SE. Model included high/low LF, high/low VAT, age, sex, TFM, and high/low LF*high/low VAT. Although
the interaction between high/low LF and high/low VAT was nonsignificant in all models, we included this term to show the adjusted means for
each of the 4 groups in AAs and Hispanics: LL, low liver fat; HL, high liver fat; LV, low visceral fat; HV, high visceral fat. ***, P < .001; **, P
< .01; *, P < .05. Variables were log transformed to meet analysis of covariance assumptions; adjusted mean values were back transformed for
ease of interpretation.

African-Americans Hispanics
0
500
1000
1500
2000
2500
n=68 n=38 n=6 n=13 n=64 n=33 n=34 n=49
**
LL
LV
LL
HV
HL
LV
HL
HV
LL
LV
LL
HV
HL
LV
HL
HV
*
Adjusted Acute Insulin Response
(µU/mL x 10 min)
African-Americans Hispanics
0.0
0.4
0.8
1.2
1.6
2.0
2.4
n=68 n=38 n=6 n=13 n=64 n=33 n=34 n=49
***
LL
LV
LL
HV
HL
LV
HL
HV
LL
LV
LL
HV
HL
LV
HL
HV
**
LL: low liver fat; HL: high liver fat; LV: low visceral fat; HV: high visceral fat
Adjusted Insulin Sensitivity
[(x 10
-4
min
-1
)/(µU/mL)]
African-Americans Hispanics
500
1000
1500
2000
2500
3000
n=68 n=38 n=6 n=13 n=64 n=33 n=34 n=49
**
LL
LV
LL
HV
HL
LV
HL
HV
LL
LV
LL
HV
HL
LV
HL
HV
Adjusted Disposition Index
A B C
! 38
Figure 2-2. LF and VAT Regression Analysis by Ethnicity

Figure 2-2. Results from the multivariate linear regression analysis. Models included LF, VAT, age, sex, TFM, and SAAT. Results are shown as
predicated values of log SI, log AIR, and log DI versus log LF (A–C) or VAT (D–F). AAs are represented by solid red circles and red regression
line. Hispanics are represented by open black circles and dashed black regression line. Variables were log transformed to meet assumptions.
0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8
-0.4
-0.2
-0.0
0.2
0.4
0.6
0.8
African-Americans Hispanics
β
Hisp
= -0.21; P= 0.003
β
AA
= -0.53; P= 0.001
P
LF*Ethnicity
= 0.01
Log Liver Fat (%) Predicted Log Insulin Sensitivity [(x 10
-4
min
-1
)/(µU/mL)]
0 1 2 3 4 5 6
2.8
3.0
3.2
3.4
3.6
β
Hisp
= 0.06; P= 0.01
β
AA
= 0.03; P= 0.55
P
VAT*Ethnicity
< 0.001
Visceral Aipose Tissue (L)
Predicted Log Acute Insulin Response (µU/mL x 10 min)
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
2.8
3.0
3.2
3.4
3.6
3.8
β
AA
= -0.38; P= 0.007
β
Hisp
= -0.02; P= 0.80
P
LF*Ethnicity
= 0.15
Log Liver Fat (%)
Predicted Log Disposition Index
A B C
D
E
F
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8
2.8
3.0
3.2
3.4
3.6
β
Hisp
= 0.11; P= 0.12
β
AA
= 0.17; P= 0.23 P
LF*Ethnicity
= 0.14
Log Liver Fat (%)
Predicted Log Acute Insulin Response (µU/mL x 10 min)
1 2 3 4 5 6
-0.4
-0.2
-0.0
0.2
0.4
0.6
0.8
β
Hisp
= 0.004; P= 0.86
β
AA
= -0.01; P= 0.73
P
VAT*Ethnicity
= 0.43
Visceral Aipose Tissue (L)
Predicted Log Insulin Sensitivity [(x 10
-4
min
-1
)/(µU/mL)]
0 1 2 3 4 5 6
2.8
3.0
3.2
3.4
3.6
3.8
β
AA
= 0.02; P= 0.69
β
Hisp
= 0.02; P= 0.41
P
VAT*Ethnicity
= 0.04
Visceral Aipose Tissue (L)
Predicted Log Disposition Index
! 39
Figure 2-3. High/low LF and High/low VAT by Prediabetes Status

Figure 3. A and B, Adjusted mean ± SE. Model included high/low LF, high/low VAT, age, sex, TFM,
prediabetes status, and high/low LF*high/low VAT. Although the interaction between high/low LF and
high/low VAT was nonsignificant in all models, we included this term to show the adjusted means for
each of the 4 groups in AAs and Hispanics: LL, low liver fat; HL, high liver fat; LV, low visceral fat; HV,
high visceral fat. *, P < .05; τ = 0.06. Variables were log transformed to meet analysis of covariance
assumptions; adjusted mean values were back transformed for ease of interpretation.

Normal Glucose Tolerant and Prediabetic
0
500
1000
1500
2000
2500
n=52 n=32 n=6 n=9
LL
LV
LL
HV
HL
LV
HL
HV
LL: low liver fat; HL: high liver fat; LV: low visceral fat; HV: high visceral fat
African-Americans
*
A
Adjusted Disposition Index
Normal Glucose Tolerant Prediabetic
0
500
1000
1500
2000
2500
n=51 n=15 n=21 n=32 n=6 n=9 n=11 n=6
τ
LL
LV
LL
HV
HL
LV
HL
HV
LL
LV
LL
HV
HL
LV
HL
HV
*
Hispanics
B
Adjusted Disposition Index
! 40
Chapter 2 References:
1. Hasson RE, Adam TC, Davis JN, et al. Ethnic differences in insulin action in obese
African-American and Latino adolescents. J Clin Endocrinol Metab 2010;95:4048–4051.
2. Goran MI. Ethnic-specific pathways to obesity-related disease: the Hispanic vs. African-
American paradox. Obesity (Silver Spring) 2008;16:2561–2565.
3. Goran MI, Bergman RN, Cruz ML, Watanabe R. Insulin resistance and associated
compensatory responses in African-American and Hispanic children. Diabetes Care
2002;25:2184–2190.
4. Bennett B, Larson-Meyer DE, Ravussin E, et al. Impaired Insulin Sensitivity and Elevated
Ectopic Fat in Healthy Obese vs. Nonobese Prepubertal Children. Obesity (Silver Spring)
2011;20:371–375.
5. Goran MI, Gower BA. Relation between visceral fat and disease risk in children and
adolescents. Am J Clin Nut 1999;70:149S–56S.
6. Cruz ML, Bergman RN, Goran MI. Unique effect of visceral fat on insulin sensitivity in
obese Hispanic children with a family history of type 2 diabetes. Diabetes Care
2002;25:1631–1636.
7. Fabbrini E, Magkos F, Mohammed BS, et al. Intrahepatic fat, not visceral fat, is linked
with metabolic complications of obesity. Proc Natl Acad Sci USA 2009;106:15430–15435.
8. Fabbrini E, Tamboli RA, Magkos F, et al. Surgical Removal of Omental Fat Does Not
Improve Insulin Sensitivity and Cardiovascular Risk Factors in Obese Adults.
Gastroenterology 2010;139:448–455.
9. Vague J, Vague P, Jubelin J. Fat distribution, obesities and health: evolution of concepts.
Curr Top Nut Dis 1988;17:9–41.
10. Ravussin E, Smith SR. Increased fat intake, impaired fat oxidation, and failure of fat cell
proliferation result in ectopic fat storage, insulin resistance, and type 2 diabetes mellitus.
Ann NY Acad Sci 2002;967:363–378.
! 41
11. Toledo-Corral CM, Alderete TL, Hu HH, et al. Ectopic fat deposition in prediabetic
overweight and obese minority adolescents. J Clin Endocrinol Metab 2013;98:1115–1121.
12. Jakobsen M, Berentzen T, Sorensen T, Overvad K. Abdominal Obesity and Fatty Liver.
Epidemiologic Reviews 2007;29:77–87.
13. Korenblat KM, Fabbrini E, Mohammed BS, Klein S. Liver, muscle, and adipose tissue
insulin action is directly related to intrahepatic triglyceride content in obese subjects.
Gastroenterology 2008;134:1369–1375.
14. Toledo-Corral CM, Vargas LG, Goran MI, Weigensberg MJ. Hemoglobin A1c above
threshold level is associated with decreased β-cell function in overweight Latino youth. J
Pediatr 2012;160:751–756.
15. Davis JN, Kelly LA, Lane CJ, et al. Randomized control trial to improve adiposity and
insulin resistance in overweight Latino adolescents. Obesity (Silver Spring)
2009;17:1542–1548.
16. Lê K-A, Mahurkar S, Alderete TL, et al. Subcutaneous adipose tissue macrophage
infiltration is associated with hepatic and visceral fat deposition, hyperinsulinemia, and
stimulation of NF-κB stress pathway. Diabetes 2011;60:2802–2809.
17. Glover G, Schneider E. Three‐point dixon technique for true water/fat decomposition with
B0 inhomogeneity correction. Magn Reson Med 1991;18:371–383.
18. Hu HH, Kim H-W, Nayak KS, Goran MI. Comparison of fat-water MRI and single-voxel
MRS in the assessment of hepatic and pancreatic fat fractions in humans. Obesity (Silver
Spring) 2010;18:841–847.
19. Goran MI, Walker R, Lê K-A, et al. Effects of PNPLA3 on liver fat and metabolic profile
in Hispanic children and adolescents. Diabetes 2010;59:3127–3130.
20. Bergman RN, Ader M, Huecking K, Van Citters G. Accurate assessment of beta-cell
function: the hyperbolic correction. Diabetes 2002;51 Suppl 1:S212–S220.
! 42
21. Lillioja S, Mott DM, Spraul M, et al. Insulin resistance and insulin secretory dysfunction
as precursors of non-insulin-dependent diabetes mellitus. Prospective studies of Pima
Indians. N Engl J Med 1993;329:1988–1992.
22. Hasson RE, Adam TC, Davis JN, et al. Randomized Controlled Trial to Improve
Adiposity, Inflammation, and Insulin Resistance in Obese African-American and Latino
Youth. Obesity 2012;20:811–818.
23. Szczepaniak LS, Nurenberg P, Leonard D, et al. Magnetic resonance spectroscopy to
measure hepatic triglyceride content: prevalence of hepatic steatosis in the general
population. Am J Phys 2005;288:E462–E468.
24. Taksali SE, Caprio S, Dziura J, et al. High visceral and low abdominal subcutaneous fat
stores in the obese adolescent: a determinant of an adverse metabolic phenotype. Diabetes
2008;57:367–371.
25. Gastaldelli A, Miyazaki Y, Pettiti M, et al. Metabolic effects of visceral fat accumulation
in type 2 diabetes. J Clin Endo & Metab 2002;87:5098–5103.
26. Kantartzis K, Machann J, Schick F, Fritsche A, Häring H-U, Stefan N. The impact of liver
fat vs visceral fat in determining categories of prediabetes. Diabetologia 2010;53:882–889.
27. Lee Y-H, Magkos F, Mantzoros CS, Kang ES. Effects of leptin and adiponectin on
pancreatic β-cell function. Metab Clin Exp 2011;60:1664–1672.
28. Lê K-A, Ventura EE, Fisher JQ, et al. Ethnic differences in pancreatic fat accumulation
and its relationship with other fat depots and inflammatory markers. Diabetes Care
2011;34:485–490.
29. Adam TC, Hasson RE, Lane CJ, et al. Fasting indicators of insulin sensitivity: effects of
ethnicity and pubertal status. Diabetes Care 2011;34:994–999.

! 43
Chapter 3: A Novel Biopsy Method to Increase Yield of Subcutaneous Abdominal Adipose
Tissue
Alderete TL, Sattler FR, Sheng X, Tucci J, Mittelman SD, Grant EG, Goran MI. A Novel Biopsy
Method to Increase Yield of Subcutaneous Abdominal Adipose Tissue. Int J Obes (Lond) 2014;
epub ahead of print.

Abstract
Collection of abdominal subcutaneous adipose tissue (SAT) for research testing is traditionally
performed using punch biopsy or needle-aspiration techniques, yielding small amounts of very
superficial SAT (100-500 milligrams). Although liposuction techniques can be used to obtain
large amounts of SAT, these approaches can compromise the integrity of the adipose tissue.
Therefore, we investigated a novel method using a 6-mm Bergström side-cutting biopsy needle
to acquire suitable amounts of intact abdominal SAT for multiple complex studies such as flow
cytometry, RNA extraction, ex vivo expression of molecular and post-translational protein
mediators, and histology. Fifty biopsies were obtained from 29 participants using a Bergström
biopsy needle, applying transient manual suction, and shearing large pieces of fat within the
inner cutting trochar. Eighteen of the biopsies were performed under ultrasound guidance,
whereby we successfully sampled deep SAT (dSAT) from below Scarpa’s fascia. The average
weight of SAT sampled was 1.5±0.4 grams. There was no clinically important bleeding or
ecchymosis on the abdominal wall and no infection occurred with this procedure. The 6-mm
Bergström biopsy needle yielded substantially more SAT than what has been obtained from
superficial procedures and for the first time allowed sampling of dSAT by a percutaneous
approach.

! 44
Introduction
In obese persons, specific patters of adipose tissue distribution, particularly upper body fat, are
closely linked to increased risk for cardiovascular disease (CVD) and insulin resistance (1-4).
Upper body fat accumulation includes visceral adipose tissue (VAT), or intra-abdominal fat, and
subcutaneous adipose tissue (SAT). Within the abdominal SAT depot, there are two distinct
regions divided by the Scarpa’s fascia into superficial SAT (sSAT) and deep SAT (dSAT) (4, 5).
VAT shows the greatest linkage with risk for inflammation, CVD, and metabolic syndrome (6-8).
However, compared to sSAT, VAT and dSAT share more similar inflammatory characteristics
and associations with disease risk (4, 5, 9-11).

Traditionally, invasive surgical procedures are required to obtain VAT and sufficient amounts of
fat for a portfolio of laboratory tests (9, 12). Since sSAT is easily accessible and can reflect some
biological aspects of VAT (9, 13), non-invasive aspiration and punch biopsy procedures are used
to sample this depot (2, 13). However, these methods typically acquire only 100-500 milligrams
of sSAT (14-16), far less than what is needed for research studies that may include flow
cytometry, RNA extraction, ex vivo expression of molecular and post-translational protein
mediators, and histology. Further, some of these techniques, including liposuction type
approaches, require vigorous suctioning, which may disrupt the integrity of adipose tissue for
multiple, and complex research studies. Although two studies reported using the Bergström
biopsy needle to sample SAT, neither provided detailed information regarding their methodology,
the quantity of adipose tissue obtained, or the feasibility of sampling dSAT (17, 18). Therefore,
the aims of this study were to determine: 1. whether a Bergström side-cutting biopsy needle
could be used to safely obtain large amounts of intact SAT (≥1 gram) and 2. whether this method
could be used in conjunction with ultrasound guidance to specifically sample dSAT because of
its greater similarities to VAT.

Methods
SAT biopsies were obtained at the bedside from study participants who were enrolled in a study
at the University of Southern California (USC). The USC Institutional Review Board approved
the protocol and all participants signed informed consent before testing. To be eligible,
participants had to be 18-35 years-of-age, obese (BMI ≥30 kg/m
2
), non-diabetic, and free of any
! 45
active medical problems. Eligible participants were scheduled for two biopsies and 3-Tesla
whole abdominal MRI scans one month apart.

The biopsy site was prepared with three betadine scrubs and covered with a fenestrated sterile
drape. The dermis at the biopsy site (at the right anterior axillary line at the level of the
umbilicus) was infiltrated with 0.5 cc of 1% lidocaine followed by injection of 4 cc in the very
superficial layers of adipose tissue immediately below the skin (no more than ½-inch deep). As
shown in Figure 3-1, a 6-7 mm incision was made in the skin with a number 11 Bard parker
scalpel at the biopsy site. A 6-mm Bergström side-cutting needle (Micrins Surgical, Inc; Lake
Forest, IL) was introduced approximately 1-1.5 inches through the incision into the deeper SAT.
After the needle was angled obliquely, an assistant applied brief suction from a 60 cc irrigation
syringe attached to the Bergström needle with gastrointestinal irrigation tubing (Kendall; no. 1;
16 Fr/Ch x 48 inches; Mansfield, MA).

Four cuts were made with the cutting trochar as the needle was further advanced and rotated 90
degrees. The procedure was repeated with a second pass to generate eight total cuts. The SAT
samples were removed from the needle with ice saline in a separate irrigating syringe. Two of
the samples were stored in Dulbecco’s phosphate buffered saline (Invitrogen, Gibco 1X) in order
to: 1. quantify macrophage sub-types by flow cytometry and 2. perform ex vivo
adipokine/cytokine secretion for future studies. A third portion of SAT was immediately fixed in
paraformaldehyde for immunohistochemistry and the last was snap frozen in liquid nitrogen and
stored at -80°C for future gene expression analyses. We quantified macrophage subtypes using
flow cytometry where we gated for the live cell population (P1) and the total cell population (P2).
We gated for nucleated leukocytes using CD45+ (P3) and monocytes and macrophages using
CD45+/CD14+ (P4). The P4 population was further distinguished with CD40 (M1 marker) and
CD206 (M2 marker) antibodies to quantify macrophages subpopulations of different activation
states (Figure 3-3).

In 18 of the biopsies performed, we utilized ultrasound (US) guidance (Toshiba Aplio 500 with
iStyle, Nasu, Japan) to ensure sampling from dSAT below the Scarpa’s fascia, which is often 2-3
inches below the skin in obese participants. We also utilized color Doppler imaging to verify that
! 46
there were no large blood vessels present in our sampling region in dSAT. Some participants
experienced minor discomfort when the needle was advanced through Scarpa’s fascia. To rectify
the discomfort we now routinely anesthetize the superficial layer of Scarpa’s fascia with 1%
lidocaine administered through a spinal needle guided by US.

After the biopsy procedure, manual compression was applied for 10 minutes to prevent bleeding.
Following compression, a single 2.0 Ethicon nylon suture (LSL Industries, Chicago, IL, USA)
was used to close the wound. Bacitracin ointment was applied to the wound, which was then
covered with Tegaderm (3M Health Care, St. Paul, MN) dressing. After this, the participant
rolled onto their right side with a cold pack (Cardinal Health, McGaw Park, IL, USA) over the
wound for an additional 15-20 minutes. Participants were given written and verbal instructions
on how to care for their wound including no water immersion (swimming, bath, spa) and
avoidance of heavy lifting (more than 5 pounds) for 72 hours. The suture was removed 5-7 days
following the biopsy. Data are presented as mean ± standard deviation.

Results
This study included 17 male and 12 female obese (BMI 40.5±6.9 kg/m
2
) Hispanic young adults
(22.4±4.8 years) who underwent 50 abdominal SAT biopsies. The time period from prepping the
biopsy site to tissue processing averaged about 20 minutes. Other than minor discomfort from
the intradermal lidocaine injection, there was no evidence of substantial participant pain during
the procedure and the mild discomfort penetrating the Scarpa’s fascia was alleviated by first
anesthetizing the fascia.

We obtained an average of 1.5 grams (0.9-2.6 grams) of intact SA T, which ranged from 5-10 times
greater than we previously obtained using a punch biopsy method (Figure 3-2 A-B) (14). As
illustrated in the MRI scan (Figure 3-2 C), SAT is divided into sSAT and dSAT by the Scarpa’s
fascia. For 18 biopsies we used US guidance to sample intact dSAT (Figure 3-2 C-D). We
performed histology and cell sorting on SAT samples, which confirmed that this procedure did
not damage the adipose tissue (Figure 3-3). There was no clinically important bleeding (<1cc at
the site of incision), abdominal wall ecchymosis, or infections. Occasionally, biopsies would
result in several mm of bruising at the biopsy site.
! 47
Discussion
This novel abdominal fat biopsy procedure used a Bergström side-cutting biopsy needle and
yielded an average of 1.5 grams of intact SAT, which was several fold greater than what has
been previously reported using punch biopsy and needle aspiration techniques (14-16).
Additionally, to our knowledge, we are the first to report the ability to safely and consistently
sample adipose tissue from the dSAT compartment by a percutaneous bedside approach. This is
particularly important since dSAT better reflects VAT than sSAT in regards to inflammation,
macrophage accumulation, cardiometabolic risk factors and insulin resistance (4, 5, 9-11).

Although the current study was limited in that we did not directly compare our biopsy method to
aspiration, punch, liposuction techniques; the Bergström biopsy needle required only mild and
brief suction to aspirate adipose tissue into the side window of the large bore Bergström needle.
This did not damage the integrity of the tissue for histology because of the size of the pieces cut.
This is in contrast to other procedures whereby SAT samples are obtained by forceful extraction
with tissue forceps, continued manual suction using smaller gauge cutting needles, or wall
suction as utilized during liposuction. For example, although a tumescent mini liposuction
method may yield 3-15 grams of adipose tissue (19), it utilizes forceful continued suction,
increasing the potential for disruption of adipose tissue and extensive abdominal wall
ecchymosis as commonly occurs with liposuction. In contrast, we found that using the Bergström
method did not damage our SAT and only occasionally resulted in minor bruising. Although two
prior studies used the Bergström biopsy needle to sample SAT (17, 18), this is the first to report a
detailed description of this approach, including the quantity of adipose tissue obtained and the
ability to sample dSAT.

Other studies have shown that open surgical procedures are able to sample large quantities of
intact adipose tissue from sSAT, dSAT, and VAT (20, 21); however, these methods are invasive
and are not always feasible for research testing. Additionally, ascertainment of surgical adipose
tissue during bariatric or lap band procedures is restricted to persons with morbid obesity who
may have metabolic dysregulation different than patients with lesser amounts of upper body
obesity. Surgical biopsies are also limited by an inability to sample SAT over time, for example,
during research treatment interventions.
! 48

In summary, the 6-mm Bergström biopsy procedure with brief manual suction allowed us to
obtain an average of 1.5 grams of SAT before and after treatment interventions in obese healthy
adults. The procedure was associated with minimal discomfort and also afforded us the ability to
obtain substantial amounts of intact adipose tissue from the dSAT compartment that has been
shown to have metabolic and histologic properties more similar to VAT (4, 9).

Acknowledgements
This work was supported by the Robert C. and Veronica Atkins Foundation Grant (M.I.G) and
the SC CTSI (NIH/NCRR/NCATS) Grant # UL1TR000130 (T.L.A.). Mrs. Alderete (principal
investigator) supervised all aspects of the study, assisted with the biopsies, and was responsible
for data analysis and manuscript preparation. Dr. Sattler (co-investigator) was the study
physician, performed the biopsies, and contributed to manuscript preparation. Dr. Goran was co-
investigator and acted as sponsor and mentor to Mrs. Alderete. Dr. Mittelman, Ms. Sheng, and
Mr. Tucci developed the adipose tissue processing protocols, analyzed results, and assisted with
the manuscript. Dr. Edward Grant helped develop the US guidance protocol used in this this. No
conflicts of interest existed for any of the authors. We would like to thank the nursing staff at the
Clinical Trials Unit as well as our participants for their involvement. The authors declared no
conflict of interest.

! 49
Figure 3-1. Biopsy Method With the Bergström Biopsy Needle
A. B.

C. D.
!
!
Figure 3-1. A. Tubing was cut at a 45-degree angle in order to fit more securely into the top of the
Bergström biopsy needle for generating suction. B. Tubing inserted into the cutting trochar of the
Bergström biopsy needle. C. A 6-7 mm incision was made through the skin up to the hub of a number 11
Bard Parker blade. D. When US guidance was not used, the Bergström biopsy needle was inserted
approximately 1.5 inches in the incision site.

! 50
Figure 3-2. Subcutaneous Adipose Tissue Biopsies: Weight and Location

A. B.

C. D.

Figure 3-2. A. The average SAT biopsy weight taken was 1.5±0.4 grams. B. Example of one of our larger SAT biopsies that weighted approximately
2.6 grams. C. Single slice from 3-Tesla MRI scan shows V AT as well as sSAT and dSAT separated by the Scarpa’s fascia. Adipose tissue was obtained
from dSAT (denoted by circle) below the Scarpa’s fascia (arrowhead) D. Representative US images are from 1 of the 18 biopsies performed under US
guidance. Left: Bergström biopsy needle above the Scarpa's fascia (arrowhead). Right: Bergström biopsy needle below the Scarpa's fascia,
approximately 2-3 inches below the skin (arrowhead), where we sampled dSAT.
!
2.6 grams
!
!
! 51
Figure 3-3. Histology and Flow Cytometry From Adipose Tissue Biopsy
A. B.

C.

Figure 3-3. A-B. Representative histology from human adipose tissue biopsies. We obtained a total of 6
slices (5 micrometers thick) of adipose tissue. A. Hematoxylin and eosin staining of nuclei. B.
Macrophages were stained with CD68. Images of the adipose tissue were captured using 10x
magnification and examined for the presence of macrophages and crown-like structures (accumulation of
macrophages around dead/dying adipocytes detonated by black arrow). C. Representative dot plots from
flow cytometry analysis of stromal vascular fraction of human fat samples. P1 gates for the live cell
population, while P2 (the rectangular gate including P1) gates for the total cell population. The P4
population was further distinguished with CD40 (M1 marker) and CD206 (M2 marker) antibodies to
quantify macrophages subpopulations of different activation states.

! 52
Chapter 3 References:
1. Fox CS, Massaro JM, Hoffmann U, et al. Abdominal visceral and subcutaneous adipose
tissue compartments: association with metabolic risk factors in the Framingham Heart
Study. Circulation 2007;116:39–48.
2. Goodpaster B, Thaete F, Simoneau J, Kelley D. Subcutaneous abdominal fat and thigh
muscle composition predict insulin sensitivity independently of visceral fat. Diabetes
1997;46:1579–1585.
3. Patel P, Abate N. Body fat distribution and insulin resistance. Nutrients 2013;5:2019–
2027.
4. Kelley DE, Thaete FL, Troost F, Huwe T, Goodpaster BH. Subdivisions of subcutaneous
abdominal adipose tissue and insulin resistance. Am J Physiol Endocrinol Metab
2011;278:E941–948.
5. Walker GE, Verti B, Marzullo P, et al. Deep subcutaneous adipose tissue: a distinct
abdominal adipose depot. Obesity (Silver Spring) 2007;15:1933–1943.
6. Item F, Konrad D. Visceral fat and metabolic inflammation: the portal theory revisited.
Obes Rev 2012;13 Suppl 2:30–39.
7. Tchernof A, Després JP. Pathophysiology of Human Visceral Obesity: An Update.
Physiological Reviews 2013;93:359–404.
8. Alvehus M, Burén J, Sjöström M, Goedecke J, Olsson T. The human visceral fat depot has
a unique inflammatory profile. Obesity 2010;18:879–883.
9. Tordjman J, Divoux A, Prifti E, et al. Structural and inflammatory heterogeneity in
subcutaneous adipose tissue: relation with liver histopathology in morbid obesity. J
Hepatol 2012;56:1152–1158.
10. Marinou K, Hodson L, Vasan SK, et al. Structural and functional properties of deep
abdominal subcutaneous adipose tissue explain its association with insulin resistance and
cardiovascular risk in men. Diabetes Care 2014;37:821–829.
! 53
11. Smith SR, Lovejoy JC, Greenway F, et al. Contributions of total body fat, abdominal
subcutaneous adipose tissue compartments, and visceral adipose tissue to the metabolic
complications of obesity. Metab Clin Exp 2001;50:425–435.
12. Kolak M, Westerbacka J, Velagapudi VR, et al. Adipose tissue inflammation and
increased ceramide content characterize subjects with high liver fat content independent of
obesity. Diabetes 2007;56:1960–1968.
13. Bigornia SJ, Farb MG, Mott MM, et al. Relation of depot-specific adipose inflammation
to insulin resistance in human obesity. Nutr Diabetes 2012;2:e30.
14. Lê K-A, Mahurkar S, Alderete TL, et al. Subcutaneous adipose tissue macrophage
infiltration is associated with hepatic and visceral fat deposition, hyperinsulinemia, and
stimulation of NF-κB stress pathway. Diabetes 2011;60:2802–2809.
15. Campbell KL, Makar KW, Kratz M, Foster-Schubert KE, McTiernan A, Ulrich CM. A
pilot study of sampling subcutaneous adipose tissue to examine biomarkers of cancer risk.
Cancer Prev Res (Phila) 2009;2:37–42.
16. Daum SM, Knittle J, Roseman K, Rom WN, Holstein EC. A simple technique for fat
biopsy of PBB-exposed individuals. Environ Health Perspect 1978;23:183–185.
17. Pasarica M, Gowronska-Kozak B, Burk D, et al. Adipose tissue collagen VI in obesity. J
Clin Endocrinol Metab 2009;94:5155–5162.
18. Tam CS, Covington JD, Bajpeyi S, et al. Weight gain reveals dramatic increases in
skeletal muscle extracellular matrix remodeling. J Clin Endocrinol Metab 2014; epub
ahead of print.
19. Bastard J-P, Cuevas J, Cohen S, Jardel C, Hainque B. Percutaneous adipose tissue biopsy
by mini-liposuction for metabolic studies. Journal of Parenteral and Enteral Nutrition
1994;18:466–468.
20. van Beek L, Lips MA, Visser A, et al. Increased systemic and adipose tissue inflammation
differentiates obese women with T2DM from obese women with normal glucose tolerance.
! 54
Metab Clin Exp 2014;63:492–501.
21. Hagman DK, Kuzma JN, Larson I, et al. Characterizing and quantifying leukocyte
populations in human adipose tissue: impact of enzymatic tissue processing. J Immunol
Methods 2012;386:50–59.

! 55
Chapter 4: Salsalate Treatment Improves Glycemia but Does Not Alter Adipose Tissue
Inflammation in Non-Diabetic Obese Hispanic Young Adults

Abstract
The anti-inflammatory salicylate drugs improve glucose and lipid homeostasis in patients with
diabetes and/or impaired glucose control, but the mechanism of their effect are unknown. This
randomized double-blind and placebo-controlled trial examined the effects of 4 g/day of salsalate
(n=11) versus placebo (n=13) for 4 weeks on glycemia and adipose tissue inflammation in non-
diabetic obese Hispanics (18-35 years). Salsalate decreased fasting glucose by 3.4% and free
fatty acids by 42.5%. Salsalate increased insulin AUC by 38% but glucose and C-peptide AUC
were unchanged. Estimates of insulin sensitivity/resistance were unaffected while HOMA-B
increased by 47.2% with salsalate. Metabolic improvements occurred without changes in total,
abdominal, or liver fat. Adiponectin increased by 27.7% but high sensitivity IL-6, sTNRrII,
sCD14, and sCD163 were unchanged following salsalate. For adipose tissue, salsalate did not
change average adipocyte size, presence of crown-like structures, or components of the immune
cell repertoire. Further, salsalate did not alter adipose gene expression of adipokines, immune
cell markers, or cytokines downstream of NF-κB with the exception of IL-1β. Our findings show
that salsalate treatment resulted in clinical improvements in risk for diabetes that occurred
without significant alterations to adiposity or adipose tissue inflammation.
Trial registration ClinicalTrials.gov NCT02130804

! 56
Introduction
Obesity is associated with chronic low-grade inflammation, which increases the risk for
insulin resistance, metabolic complications, and type 2 diabetes (T2D) (1-5). Evidence suggests
that adipose tissue is a significant contributor to this inflammatory state (6-11). Treatment
strategies have included anti-inflammatory therapies to improve metabolic health. In rodents,
salicylates inhibit obesity-induced inflammation and improve insulin resistance (12-16). This is
important since high doses of salicylates appear to inhibit nuclear factor kappa B (NF-κB)
activity, a transcription factor that contributes to increased pro-inflammatory and pro-atherogenic
mediators occurring during obesity (17-19).
Recent clinical investigations have shown that salsalate (a prodrug of salicyate) favorably
affects glycemia in predominantly obese Caucasian adults with impaired fasting glucose (IFG),
impaired glucose tolerance (IGT), and/or T2D (20-24). In general, results from these studies
show that salsalate improved glucose and lipid homeostasis (21-24). Further, salsalate has been
shown to inhibit systemic inflammation and NF-κB activity in peripheral blood mononuclear
cells (21) and adipose tissue (24). Collectively, these findings suggest that salsalate-induced
metabolic improvements may be mediated by decreases in adipose tissue inflammation.
These clinical outcomes are particularly important since salsalate is a putatively safe and
inexpensive treatment that could be used for prevention or reversal of cardiometabolic
abnormalities occurring during obesity. Yet, there is limited data about the utility of salsalate to
improve metabolic health in persons without T2D (25-27) and its effects on adipose tissue
inflammation are uncertain. Furthermore, treatment with salsalate has not be evaluated
exclusively in Hispanics, who suffer from a greater prevalence of obesity (28) and
cardiometabolic disease risk than non-Hispanic whites (29, 30). We therefore conducted a
randomized, double-blind and placebo-controlled trial of salsalate in obese Hispanic young
adults without T2D to determine 1) whether salsalate improves glycemic control and 2) whether
metabolic improvements are mediated by attenuation of adipose tissue inflammation.

Methods
Study Design. The study was a 4-week, randomized double-blind and placebo-controlled
investigation, which compared 4 g/day of salsalate (2 g twice daily) with matching placebo
(Merical, Anaheim, CA, USC). Primary outcomes included effects on glycemia, insulin
! 57
resistance, and markers of systemic and adipose tissue inflammation. The protocol specified
stepped reductions of 500 mg/day for symptoms related to salicylate (e.g., tinnitus). We
evaluated adherence by pill counts and salicylate levels at baseline and 4 weeks (trough levels
~12 hours after previous dose and then again 2 hours after final dose). Participants were called
weekly to encourage adherence and inquire about potential adverse events. Participants were
instructed to maintain their current activity and dietary patterns during the study.

Participants and Screening. Participants signed an informed consent approved by the
University of Southern California’s (USC) Institutional Review Board approved prior to
undergoing any study measurements or interventions. Participants were recruited from Los
Angeles, CA and inclusion criteria required that participants be otherwise healthy obese (body
mass index [BMI] ≥30 kg/m
2
) Hispanic adults 18-35 years of age. Hispanic ethnicity required
that both parents and grandparents be of Hispanic descent (by self-report). Participants were
excluded if they had diabetes, peptic ulcer disease, history of gastrointestinal bleeding, blood
clotting disorder, liver or kidney function abnormalities, asthma, allergy to non-steroidal anti-
inflammatory drugs (NSAIDs), or were pregnant or lactating. Participants were also excluded if
they were taking any medications that could affect body composition, metabolism or
inflammation (e.g., thyroid replacement, β-blockers, NSAIDs, statins). Half of those enrolled had
to have at least two of the latter; HOMA-IR ≥3.5, elevated HbA1C (5.7-6.4%), or elevated
erythrocyte sedimentation rate.

2-Hour OGTT. A 2-hour OGTT was performed after an overnight fast at baseline and
immediately following the last treatment dose. Blood samples were collected at baseline and 15,
30, 45, 60, and 120 minutes following ingestion of glucose (75 g). HOMA-IR, an index measure
of insulin resistance, was calculated as fasting glucose (mg/dL) x insulin (µU/mL) / 205.
HOMA-B, an indirect measure of β-cell function, was calculated as fasting insulin (µU/mL) x
360 / [fasting glucose (mg/dL) – 63]. QUICKI, an index for insulin sensitivity, was calculated as
1 / (log (fasting insulin (µU/mL)) + log (fasting glucose (mg/dL))) (31). Finally, the Matsuda
Index, an indirect but more robust estimate of insulin sensitivity, was calculated as 10,000 / sqrt
[(fasting glucose (mg/dL) x fasting insulin (µU/mL)) x (mean glucose x mean insulin during the
OGTT)]. Glucose, insulin, and C-peptide area under the curve (AUC) were calculated from the
! 58
OGTT data. Dietary intake was collected using 24-hour recall multiple pass method at baseline
and after 4 weeks of treatment (32). Nutrition data were analyzed using the Nutrition Data
System for Research (NDS-R version 2012) developed at the University of Minnesota. Godin-
Shephard Leisure-Time Physical Activity questionnaires were also performed before and after
the 4 week study intervention in order to calculate weekly leisure-time activity (33).

Assays. Assays were performed in duplicate (except glucose and C-peptide) at the USC
Metabolic Assay Core and had coefficients of variation of less than fifteen percent. Glucose was
assayed on a Yellow Springs 2710 Analyzer (Yellow Springs, OH) using the glucose oxidase
method. Insulin was assayed using a specific human ELISA kit from EMD Millipore (St. Charles,
MO) and fasting FFA were quantified using a colorimetric kit (NEFA-HR(2)) from Wako
Diagnostics (Richmond, VA). C-peptide was assayed by an automated enzyme immunoassay
(Tosho Bioscience Inc., AIA 600 II analyzer, South San Francisco, CA). Circulating cytokines
and markers of immune activation, including high sensitivity interleukin-6 (hsIL-2) (34), soluble
tumor necrosis receptor-II (sTNFrII) (35, 36), sCD14 (1), and sCD163 (6) were measured by
human ELISA from R&D Systems Inc. (Minneapolis, MN). Adipokines, including adipsin and
adiponectin, were assayed using human ELISA from R&D Systems Inc. (Minneapolis, MN) and
EMD Millipore (St. Charles, MO). Complete blood counts and comprehensive chemistry panels
were performed by the USC Clinical Laboratory and plasma salicylate levels by Quest
Diagnostics (Chantilly, VA).

Adipose Tissue Biopsies. Baseline and post-test subcutaneous abdominal adipose tissue biopsies
were obtained at the right anterior axillary line at the level of the umbilicus using a 6-mm
Bergström side-cutting needle (Micrins Surgical, Inc; Lake Forest, IL). Adipose tissue was
rapidly irrigated with iced saline and immediately transported for flow cytometry, fixed for
immunohistochemistry, or flash frozen in liquid nitrogen for subsequent whole-tissue Fluidigm
gene expression. Briefly, flow cytometry gated for live cells, leukocytes (CD45
+
), and then
monocytes and macrophages (CD45
+
/CD14
+
). Percent monocytes and macrophages were
calculated as the percent difference between CD45
+
/CD14
+
and CD45
+
/CD14
-
isotype control
cells. For immunohistochemistry, adipose tissue samples were formalin-fixed and paraffin-
embedded. Four consecutive, 5-micron sections of adipose tissue were stained with hematoxylin-
! 59
eosin and CD68 antibody (Leica Biosystems, Newcastle, UK). Four independent fields at 20x
magnification were captured from the first mounted section on each slide for cell sizing. Adipose
cell size (µ
2
) was obtained from each field captured using Fiji quantitative microscopy software
(12). For samples at each time point, the mean value for adipose cell size was obtained. All
sections were examined by a single blinded technician for the presence or absence of
macrophage crown-like structures (CLS), an indicator of adipose tissue inflammation (17, 19).
The Genoseq Core at the University of California Los Angeles performed adipose tissue gene
expression assays using Fludigm DELTAgene assays. Gene expression from the post-test biopsy
was compared to baseline in order to calculate fold change (FC). Assays included genes involved
in adipose tissue inflammation (inflammatory cytokines and immune cell markers) as well as
insulin, FFA, and lipopolysaccharide (LPS) / Toll-like receptor (TLR) signaling.

Adiposity. A DEXA scan was used to control for possible changes in body composition (e.g.,
percent body fat and lean tissue mass) before and after the 4-week intervention. At baseline and
within one week of completing the intervention, a whole abdominal (top of liver to iliac crest) 3-
Tesla magnetic resonance imaging (MRI) scan (Excite HD; GE Healthcare, Waukesha,
Wisconsin) was used to measure subcutaneous abdominal adipose tissue (SAAT), visceral
adipose tissue (VAT), liver fat fraction, and pancreatic fat fraction (PF) as previously reported
(20). Briefly, the MRI utilized the IDEAL method (GE Healthcare) which separates water and fat
datasets of the anatomy on a voxel-wise basis (20, 37). This 3-D whole abdominal MRI scan
provides SAAT and VAT in liters as well as liver and pancreatic fat as the percentage of fat
found within each organ tissue. A single analyst performed all MRI image post-processing,
tissue segmentation, and analysis using SliceOmatic software (Tomovision, Inc., Montreal,
Québec, Canada).

Statistical Analysis. Study data were managed using Research Electronic Data Capture
(REDCap) tools hosted at USC (38). Statistical analyses were performed with IBM SPSS
Statistics (version 21). Data are reported as median ± interquartile range (IQR), median percent
change, or mean ± standard deviation. Paired comparisons (post-test versus baseline) and
unpaired group comparisons were performed by Wilcoxon signed rank tests, Mann-Whitney U
Tests, or the McNemar’s test and Fisher’s Exact test for CLS data. The effect of salsalate was
! 60
examined using analysis of covariance (ANCOVA) where the dependent variable was the change
from baseline in each variable of interest (e.g., plasma markers of metabolic health and adipose
tissue biopsy outcomes). These models controlled for the baseline measure of each dependent
variable and baseline percent of body fat. Change in CLS (presence or absence) was examined
using Fisher’s Exact test as well as logistic regression including the same covariates. Partial
correlations between variables of interest also controlled for baseline body fat percent. When
necessary, log transformations were performed to meet assumptions of normality. Results were
considered significant when two-sided P-value was less than 0.05.

Results
Baseline Assessments. As shown by the consort diagram in Figure 4-1, 30 participants were
randomized to the study and 25 completed the intervention. Table 4-1 lists the baseline
characteristics of the 24 participants with complete paired study data. On average, participants
were approximately 20 years old and 58% were male. Those in the placebo and salsalate group
did not significantly differ in any parameter except fasting insulin and estimates of insulin
sensitivity, resistance and secretion. The salsalate group had a higher fasting insulin (11.0±4.1 vs.
7.2±5.2 µU/mL, P=0.02), fasting insulin to C-peptide ratio (3.3±0.8 vs. 2.8±0.9, P=0.03),
HOMA-IR (2.2±0.8 vs. 1.5±1.3, P=0.03), and lower QUICKI (0.34±0.02 vs. 0.36±0.05, P=0.03)
and Matsuda Index (1.9±0.7 vs. 2.7±1.8, P=0.01).

Participation and Compliance. Two participants in the placebo and one in the salsalate arm
dropped out of the study. Reasons for dropping out included lack of time, loss of interest, or
illness / injury that was unrelated to study participation. There were four instances of tinnitus
(placebo n=1 and salsalate n=3) that were graded mild to moderate, resolved within 1-2 weeks
during continued treatment, and did not require a dose reduction. Changes in weight, blood
pressure (BP), liver function tests, blood urea nitrogen, serum creatinine, and estimated
glomerular filtration rate did not vary by treatment groups (data not shown). At the end of the
study, there were no differences in the number of pills remaining in the placebo group compared
to salsalate treatment (19.0±53.0 vs. 22.5±32.0 pills, P=0.73). All participants had non-detectable
levels of plasma salicylate at baseline. At post-test, the median salicylate trough levels remained
non-detectable in the placebo group and increased to 26.0±128.0 mg/L in the salsalate group.
! 61
Two hours after the final dose, salicylate levels remained non-detectable among those receiving
placebo and increased to 105.0±117.0 mg/L in the salsalate group. As shown in Table 4-2, BMI,
total body fat percent, SAAT, VAT, liver, and pancreatic fat did not change within or between
treatment groups. Finally, upper and lower body fat percent, lean tissue mass, total energy intake,
macronutrients (protein, fat, fiber, and carbohydrates), and leisure activity did not differ between
the groups at baseline or change within or between treatment groups (data not shown).

Fasting Measures and OGTT Results. Table 4-2 shows outcome variables before and after the
4-week treatment intervention. There were treatment related differences in plasma fasting
glucose levels (P
treatment
<0.01) as well as fasting FFA (P
treatment
=0.02), where the median blood
glucose and FFA levels decreased in the salsalate (-3.4%, P=0.003 and -42.5%, P=0.06,
respectively) but not placebo group (Figure 4-2A and D). Salsalate resulted in higher insulin
AUC (P
treatment
<0.01), where the median value increased in the salsalate (+38.0%, P=0.01) but
not placebo group (Figure 4-3C-D). Although there was a trend for an increased fasting C-
peptide in the placebo group (P
treatment
=0.05), salsalate treatment did not significantly alter levels
of fasting insulin and fasting C-peptide (Figure 4-2B-C) or glucose AUC and C-peptide AUC
(Figure 4-3A-B and E-F). Fasting insulin to C-peptide ratios were increased in the salsalate but
not placebo group (+57.6%, P<0.01; P
treatment
<0.01). Treatment with salsalate did not alter
estimates of insulin sensitivity or resistance (QUICKI, HOMA-IR or Matsuda Index); however, a
proxy for β-cell function (HOMA-B) increased in the salsalate (+47.2%, P<0.01; P
treatment
<0.01)
but not placebo group.

Adipokines and Markers of Inflammation. Salsalate treatment resulted in higher plasma
adiponectin levels (27.7%, P<0.01; P
treatment
<0.001) with no change in the placebo group (Table
4-2). However, circulating adipokines, cytokines, and markers of immune activation, including
adipsin, hsIL-6, sCD14, and sCD163, were unchanged within each group and there was no effect
of salsalate treatment. Although there was a trend towards decreased sTNFrII levels in the
salsalate group (-5.2%, p=0.09), there were no treatment related differences. From the fat
biopsies, treatment did not alter the average fat cell size (P
treatment
=0.90) or presence of CLS
(P
treatment
=0.40); however, there was a modest decrease in the average fat cell size among those in
the placebo group (P=0.046). To examine adipose tissue inflammation in a more quantitative
! 62
manner, the percent leukocytes and monocytes/macrophages in each fat sample we assessed
using flow cytometry. Overall, there were no within group or treatment related differences in
these immune cell populations. Despite systemic metabolic improvements in glycemia, FFA, and
adiponectin levels, there were no within group or treatment related differences in adipose tissue
gene expression for adiponectin, dipeptidyl peptidase-4 (DPP4), leptin, adipsin or any signaling
pathways of interest (i.e., insulin, FFA, LPS/TLR). Further, there were no significant changes in
gene expression of immune cell markers, including ACP5, CD68, CD163, MCP1, FOXP3, and
LIPA within or between treatment groups (data not shown). Of several inflammatory markers
affected by NF-κB (e.g., IL-6/8, IL1R, TNFα, NOS, ICAM1, VCAM1), salsalate treatment was
only associated with a 47% decreased IL1-β expression (P=0.002) while there was no change in
the placebo group (P
treatment
=0.29). Finally, although post-test plasma salicylate levels (2 hours
after the final dose of salsalate) were inversely related to IL1-β mRNA levels (r = -0.46), this
relationship did not reach statistical significance (P=0.25).

Discussion
In this study of obese non-diabetic Hispanics, who are at increased risk for
cardiometabolic disease (29, 30), we show for the first time that salsalate improved glucose,
insulin, and fat metabolism without overt changes in adiposity, liver fat, or adipose tissue
inflammation. In particular, blood glucose levels decreased significantly by 3.4% following
treatment with salsalate, even though ~88% of participants had baseline blood glucose levels
within the normal range (≤100 mg/dL). Treatment also reduced fasting FFA by 42.5% and
increased insulin concentrations during the OGTT (insulin AUC: +38.0%). These metabolic
improvements were accompanied by increased insulin secretion (HOMA-B: +47.2%), similar to
other reports (27, 39). Although there were no improvements in insulin sensitivity or resistance
(QUICKI, HOMA-IR, Matsuda Indices) following salsalate treatment, changes in insulin
sensitivity may not have been detectable since participants had relatively normal fasting insulin
levels and metabolic indices at baseline (40). Increased HOMA-B appeared to be related to
decreased insulin clearance as reported previously (21, 24, 26), since salsalate did not alter C-
peptide levels but increased insulin AUC and insulin to C-peptide ratios.
Decreased insulin clearance could have resulted in hyperinsulinemia, thereby indirectly
suppressing hepatic glucose production and/or stimulating glucose and FFA uptake by peripheral
! 63
tissues. Alternatively, since high-dose aspirin (a primary metabolite of salsalate) therapy
improves hepatic insulin sensitivity, glucose production, and inflammation (41), it is possible
that salsalate decreased fasting glucose and fasting FFA through these hepatic alterations or
decreased rates of insulin clearance. Supporting this, in Sprague-Dawley rats, salsalate
ameliorated hepatic steatosis through decreased fetuin-A expression via the AMPK-NF-κB
dependent pathway (42). In the current study, we found that metabolic improvements occurred
without reductions in liver fat, suggesting that changes in liver inflammation, prior to declines in
hepatic fat, may be contributing to metabolic improvements seen in humans treated with salsalate.
However, it is possible that this methodology was not sensitive enough to capture small, yet
important, alterations in hepatocyte lipid content, which could affect metabolism.
The fact that adiponectin levels increased by 27.7% following salsalate treatment,
suggests that metabolic improvements could also relate to this adiopokine since it stimulates
insulin sensitivity and modulates hepatic glucose production, fatty acid oxidation, and lipid
synthesis (42-46). The marginal reductions in plasma sTNFrII following salsalate (P=0.09)
suggest that increased adiponectin might have contributed to reductions in glucose and fasting
FFA by decreasing systemic inflammation (47). However, we did not detect changes in other
plasma markers of inflammation (i.e., hsIL-6, sCD14, sCD163, and adipsin). Finally, although
adiponectin is exclusively produced by adipocytes (47, 48), there were no changes in adipose
tissue gene expression of adiponectin following treatment. Thus, results from our study are
consistent with the possibility that increased circulating adiponectin levels did not result from
increased adipocyte gene expression, but were due to either enhanced protein translation and
secretion or decreased clearance by the liver or kidneys (47).
To our knowledge, this is the first study to extensively examine adipose tissue as a
potential target to explain the beneficial effects of salsalate treatment on glycemia in obese, non-
diabetic Hispanics. Using histology, flow cytometry, and gene expression, we found that
salsalate did not affect markers of adipose tissue inflammation, including fat cell size, presence
of CLS, or the percent of immune cells (monocytes/macrophages and leukocytes) in biopsy
specimens. Furthermore, we did not detect changes in expression of immune cell markers or IL-
6/8, IL1R, TNFα, NOS, ICAM1, and VCAM1, which is consistent with another report (24).
However, we did observe a significant decrease in IL-1β gene expression after salsalate.
Macrophage derived IL-1β stimulates IL-17 and IL-22 producing T cells (11), which are found
! 64
in adipose tissue of insulin resistance individuals (9). Therefore, it is possible that salsalate
exerted some of its metabolic effects through inhibition of macrophage derived IL-1β. However,
given that other markers of adipose tissue inflammation were unchanged, it is unlikely that
decreased IL-1β gene expression accounted for the improvements observed in glucose and
fasting FFA.
Although participants were randomly assigned to placebo or salsalate, those in the
salsalate group had somewhat higher baseline fasting insulin levels, were less insulin sensitive,
and showed lower rates of insulin clearance than those in the placebo group. These differences
likely resulted from the relatively small number of study participants. Notwithstanding, all
findings were corrected for these baseline differences. Despite these minor differences at
baseline, participants appeared to be adherent to all aspects of the protocol, which strengthens the
treatment results. Pill counts indicated good adherence over the 4-week treatment period.
Average salicylate levels two hours after the final dose were 105.0±117.0 mg/L, which are levels
known to be anti-inflammatory. Further, the fact that there were no changes in diet, exercise,
body fat percent, or ectopic fat provide further evidence that our findings can not be attributed to
changes in lifestyle or adiposity. Thus, the metabolic improvements in our Hispanic participants
were likely due solely to salsalate.
In conclusion, 4 weeks of salsalate treatment lowered fasting glucose and FFA levels
among non-diabetic obese Hispanic young adults. To our knowledge, this is the first study to
systematically examine the metabolic effects of salsalate as they related to adiposity, ectopic fat,
and various markers of adipose tissue inflammation. We demonstrated that salsalate treatment
improved metabolic outcomes without altering percent body fat, volume of abdominal adipose
tissue depots, liver fat, adipose tissue inflammation, or adipose tissue gene expression of
important adipokines (adiponectin, adipsin, leptin). By contrast, studies of salsalate in adult
diabetics suggest that improvements in glycemic control are due to salsalate based suppression of
inflammatory plasma biomarkers (CRP, leukocytes, sCD40L), presumed to originate from
abdominal fat (21, 49, 50). However, our study in non-insulin resistant obese Hispanics
indicates that salsalate has metabolic effects independent of its anti-inflammatory properties that
improve aspects of glucose, insulin, and FFA metabolism. Based on our findings, future studies
should examine how salsalate affects the liver, particularly hepatic insulin sensitivity and glucose
production, as well as its effects on insulin and adiponectin clearance.
! 65
Acknowledgments
We would like to thank the CTU and our dedicated project assistants; Preeya Desai, Karen Elano,
and Kimberly Chen, MPH. Our gratitude is especially extended to our participants for their
involvement.
Grant Support:
This work was supported by the SC CTSI (NIH/NCRR/NCATS) Grant # UL1TR000130
(T.L.A.) and the Robert C. and Veronica Atkins Foundation (M.I.G).
Attribution (Contributions to Project)
T.L.A. served as principal investigator and collected, analyzed, and interpreted data as well as
prepared the manuscript. M.I.G. (co-investigator, sponsor, and mentor), F.R.S. (co-investigator),
and E.G.G. assisted with data analysis, data interpretation, and manuscript preparation. F.R.S.
was the study physician and performed the biopsies; E.G.G. provided ultrasonic guidance for fat
biopsies. L.E.G., J.M.R., H.A., S.D.M., X.S., and J.T. assisted with data collection, evaluation of
statistical analysis, and manuscript preparation. S.D.M., X.S., and J.T. also developed adipose
tissue processing protocols for flow cytometry.
Conflict of Interest Disclosures:
The authors have nothing to disclose. No conflicts of interest existed for any of the authors.

! 66
Figure 4-1. Study Flow Diagram

Figure 4-1. Of those who completed the screening visit (n = 54), 18 were deemed ineligible, 5 lost
contact, and 1 lost interest.
*
Reasons for ineligibility after screening included smoking (n = 1), pregnancy
(n = 1), abnormal routine blood chemistry results (n = 1), low BMI (n = 3), or exclusion based on
HOMA-IR, HbA1C, and ESR (n = 12). Of the 30 randomized (15 placebo / 15 salsalate), only 5 dropped.
Reasons for dropout included lack of time (n = 2), loss of interest (n = 1), unrelated study illness (n = 1),
and unrelated study injury (n = 1). Injury involved a fractured rib that was unrelated to study
involvement.
#
Cold symptoms, including nausea, were unrelated to the intervention.
**
One participant did
not have complete fat biopsy data and was not included in the analysis.

! 67

Placebo
N=13
Salsalate
N=11

P-Value
General Characteristics
Age (years) 19.0(8) 22.0(8) 0.46
Sex (M/F) 8/5 6/5 0.53
Height (m) 1.7(0.18) 1.7(0.15) 0.82
Weight (kg)

101.8(22.9) 109.7(40.4) 0.17

Adiposity: DEXA and 3-T MRI
BMI (kg/m
2
) 35.5(10.5) 41.3(9.3) 0.13
Total Body Fat (%) 40.6(10.0) 44.8(10.3) 0.30
SAAT (L) 9.1(3.2) 11.0(3.6)
a
0.28
VAT (L) 3.3(2.0) 3.4(2.9)
a
0.52
Liver Fat (%) 8.6(9.2) 8.7(7.5)
a
0.69
Pancreatic Fat (%) 5.3(2.5) 6.9(8.8)
b
0.51

2-hr OGTT
Systolic BP (mm/Hg) 114.0(19.0) 122.0(35.0) 0.78
Diastolic BP (mm/Hg) 71.0(10.0) 67.0(10.0) 0.23
Fasting Glucose (mg/dL) 83.2(8.2) 86.2(12.2) 0.09
Fasting Insulin (µU/mL) 7.2(5.2) 11.0(4.1) 0.02
Fasting C-Peptide (ng/mL) 2.9(1.1) 3.6(1.3) 0.10
Fasting Insulin / C-peptide Ratio 2.8(0.9) 3.3(0.8) 0.03
Fasting FFA (µM/L) 459.2(126.7) 510.7(123.6) 0.46
QUICKI 0.36(0.05) 0.34(0.02) 0.03
HOMA-IR 1.5(1.3) 2.2(0.8) 0.03
HOMA-%B 142.2(101.7) 157.9(69.2) 0.33
Matsuda Index 2.7(1.8) 1.9(0.7) 0.01
Glucose AUC (mg/dL x min) 268.9(51.1) 280.6(60.1) 0.25
Insulin AUC (µU/mL x min) 88.0(38.2) 114.8(55.6) 0.06
C-Peptide AUC (ng/mL x min) 18.9(5.2) 22.4(6.9) 0.19

Table 4-1. Baseline Group Comparisons!
! 68
Table 4-1. Data are reported as median (IQR). P-value corresponds to Mann-Whitney U Test on
Medians.
a
N=10;
b
N=9;
c
N=12.

Plasma Markers
hsIL-6 (pg/mL) 2.4(2.0) 2.6(1.3) 0.96
sTNFrII (pg/mL) 2,283.5(863.5) 2,380.1(806.3) 0.42
sCD14 (ng/mL) 1,309.7(290.5) 1,267.9(153.8) 0.33
SCD163 (ng/mL) 41.7(34.8) 47.5(24.9) 0.82
Adiponectin (ng/mL) 6,446.0(3,973.9) 6,363.5(4,015.5) 0.87
Adipsin (ng/mL) 2,710.6(927.2) 3,173.8(867.8) 0.25

Adipose Tissue Biopsies
Average Fat Cell Size (µ
2
) 5,533.0(1,098) 5,947.0(2,771) 0.33
Crown-Like Structures (Y/N) 3/10(23%) 7/4(64%) 0.06
Leukocytes (%) 17.5(15.4)
c
20.9(26.2) 0.61
Monocytes/Macrophages (%) 34.0(16.7)
c
29.0(44.1) 0.70
! 69
Table 4-2. Variables at Baseline and After the 4-Week Intervention
Placebo (4 g/day) (N=13) Salsalate (4 g/day) (N=11) Treatment
Baseline Post-test Baseline Post-test P-Value
2-hr OGTT
Systolic BP (mm/Hg) 114.0(19.0) 117.0(16.0) 122.0 (35.0) 123.0(27.0) 0.69
Diastolic BP (mm/Hg) 71.0(10.0) 71.0(8.0)
‡
67.0(10.0) 69.0(12.0) 0.14
Fasting Glucose (mg/dL) 83.2(8.2) 83.8(4.5) 86.2(12.2) 83.3(10.9)
**
<0.01
Fasting Insulin (µU/mL) 7.2(5.2) 10.0(6.3)
‡
11.0(4.1) 14.8(6.3) 0.24
Fasting C-Peptide (ng/mL) 2.9(1.1) 3.2(1.1)
*
3.6(1.3) 3.3(1.7) 0.05
Fasting Insulin / C-peptide Ratio 2.8(0.9) 3.2(1.3) 3.3(0.8) 5.2(2.0)
**
<0.01
Fasting FFA (µM/L) 459.2(126.7) 491.9(111.7) 510.7(123.6) 293.8(201.8)
‡
0.02
QUICKI 0.36(0.05) 0.34(0.04) 0.34(0.02) 0.32(0.02) 0.56
HOMA-IR 1.5(1.3) 2.1(1.4)
‡
2.2(0.8) 3.2(1.4) 0.43
HOMA-B (%) 142.2(101.7) 165.1(130.9) 157.9(69.2) 232.5(243.2)
**
<0.01
Matsuda Index 2.7(1.8) 2.2(1.2) 1.9(0.7) 1.4(0.9) 0.40
Glucose AUC (mg/dL x min) 268.9(51.1) 261.5(47.0) 280.6(60.1) 260.0(21.7) 0.44
Insulin AUC (µU/mL x min) 88.0(38.2) 93.2(40.9) 114.8(55.6) 159.6(73.8)
**
<0.01
C-Peptide AUC (ng/mL x min) 18.9(5.2) 21.9(5.8)
*
22.4(6.9) 24.1(9.6) 0.16

Plasma Markers
hsIL-6 (pg/mL) 2.4(2.0) 2.3(1.4) 2.6(1.3) 1.7(2.8) 0.54
sTNFrII (pg/mL) 2,283.5(863.5) 2,389.1(1,029.4) 2,380.1(806.3) 2,256.1(544.6)
#
0.14
sCD14 (ng/mL) 1,309.7(290.5) 1,236.2(375.7) 1,267.9(153.8) 1,247.8(334.1) 0.30
SCD163 (ng/mL) 41.7(34.8) 47.7(39.9) 47.5(24.9) 49.8(12.6) 0.37
! 70
Table 4-2. Data are reported as median (IQR).
a
N=10;
b
N=9;
c
N=12. Paired P-values correspond to Related-Samples Wilcoxon Signed Rank Test
on the differences of the medians or
¶
McNemar test on proportions.
‡
P=0.06;
#
P=0.09; *P<0.05; **P≤0.01. Treatment P-values are from
ANCOVA, which controlled for the baseline dependent variable and body fat percent. A Fisher’s Exact test was performed on the change in
crown-like structures (CLS) to determine whether treatment altered the proportion of those with CLS at post-test compared to baseline.
Adiponectin (ng/mL) 6,446.0(3,973.9) 5,925.5(3,628.1) 6,363.5(4,015.5) 8,123.5(3,164.5)
**
<0.001
Adipsin (ng/mL) 2,710.6(927.2) 2,803.8(751.2) 3,173.8(867.8) 3,264.2(695.8) 0.99

Adiposity: DEXA and 3-T MRI
BMI (kg/m
2
) 35.5(10.5) 36.8(10.2) 41.3(9.3) 41.8(7.6) 0.90
Total Body Fat (%) 40.6(10.0) 39.7(9.6) 44.8(10.3) 45.2(12.9) 0.42
SAAT (L) 9.1(3.2) 8.4(3.1) 11.0(3.6)
a
9.9(4.5)
a
0.96
VAT (L) 3.3(2.0) 3.2(2.9) 3.4(2.9)
a
3.6(2.5)
a
0.53
Liver Fat (%) 8.6(9.2) 8.2(9.9) 8.7(7.5)
a
9.1(10.0)
a
0.19
Pancreatic Fat (%) 5.3(2.5) 6.2(3.7) 6.9(8.8)
b
7.8(3.3)
b
0.94

Adipose Tissue Biopsies

Average Fat Cell Size (µ
2
) 5,533.0(1,098) 4,478.0(1,528.0)
*
5,947.0(2,771) 5,274.3(2,479.0) 0.90
Crown-Like Structures (Y/N)
¶
3/10(23%) 0/13(0%) 7/4(64%) 3/8(27%) 0.40
Leukocytes (%) 17.5(15.4)
c
23.2(15.6)
c
20.9(26.2) 30.7(25.5) 0.14
Monocytes/Macrophages (%) 34.0(16.7)
c
33.3(25.7)
c
29.0(44.1) 27.3(60.4) 0.73
! 71
Figure 4-2. Fasting Measures From the 2-hr OGTT

Figure 4-1. Fasting glucose (A), insulin (B), C-peptide (C), and free fatty acids (D) levels from the 2-hr
OGTT. Individual participant data is shown for those in the placebo (grey squares and lines) and salsalate
(black circles and lines) groups. Treatment P-values are from ANCOVA, which controlled for the
baseline dependent variable and body fat percent. Within group P-values correspond to Related-Samples
Wilcoxon Signed Rank Test.

Baseline Post-Test Baseline Post-Test
70
75
80
85
90
95
100
105
110
P<0.01
Ptreatment<0.01
Placebo Salsalate
ΔMedian: +0.6 ΔMedian: -2.9
Fasting Glucose (mg/dL)
B. A.
Baseline Post-Test Baseline Post-Test
0
5
10
15
20
25
30
35
40
Placebo Salsalate
ΔMedian: +2.8 ΔMedian: +3.8
Fasting Insulin (µU/mL)
C. D.
Baseline Post-Test Baseline Post-Test
200
300
400
500
600
700
800
Placebo Salsalate
ΔMedian: +32.7 ΔMedian: -216.9
P=0.06
Ptreatment=0.02
Fasting Free Fatty Acids (µM/L)
Baseline Post-Test Baseline Post-Test
1
2
3
4
5
6
7
Placebo Salsalate
ΔMedian: +0.3 ΔMedian: -0.3
P=0.04
Ptreatment=0.05
Fasting C-Peptide (ng/mL)
! 72
Figure 4-3. Glucose, Insulin, and C-Peptide During the 2-hr OGTT

Figure 4-3. Results from the 2-hour OGTT. Plasma glucose (A, B), insulin (C, D), and C-Peptide (E, F)
responses are shown for placebo (left column) and salsalate (right column).
*
P<0.05,

τP<0.01. P-values were
obtained using ANCOVA analysis (change in placebo vs. change salsalate) with adjustment for the dependent
variable at baseline as well as percent body fat. A significant treatment effect was only observed for insulin
AUC (P
treatment
<0.01).
15 30 45 60 120 0 90
60
80
100
120
140
160
180
200
Post-Test Baseline
*
ΔMedian Glucose AUC: -7.4 mg/dL x min
Time (min)
Glucose (mg/dL)
15 30 45 60 120 0 90
60
80
100
120
140
160
180
200
Post-Test Baseline
τ
*
*
ΔMedian Glucose AUC: -20.6 mg/dL x min
Time (min)
Glucose (mg/dL)
15 30 45 60 120 0 90
0
20
40
60
80
100
120
Post-Test Baseline
τ
ΔMedian Insulin AUC: +5.2 µU/mL x min
Time (min)
Insulin (µU/mL)
15 30 45 60 120 0 90
0
20
40
60
80
100
120
Post-Test Basline
τ
τ
τ
ΔMedian Insulin AUC: +44.8 µU/mL x min
Time (minutes)
Insulin (µU/mL)
Placebo Salsalate
A. B.
C. D.
E. F.
15 30 45 60 120 0 90
0
2
4
6
8
10
12
14
16
18
Post-Test Baseline
ΔMedian C-peptide AUC: +3.0 ng/mL x min
*
τ
*
Time (min)
C-Peptide (ng/mL)
15 30 45 60 120 0 90
0
2
4
6
8
10
12
14
16
18
Post-Test Baseline
ΔMedian C-peptide AUC: +1.7 ng/mL x min
Time (min)
C-Peptide (ng/mL)
! 73
Chapter 4 References:
1. Fernández-Real JM, Pérez del Pulgar S, Luche E, et al. CD14 modulates inflammation-
driven insulin resistance. Diabetes 2011;60:2179–2186.
2. Lumeng CN, Saltiel AR. Inflammatory links between obesity and metabolic disease. J
Clin Invest 2011;121:2111–2117.
3. Reyes M, Quintanilla C, Burrows R, Blanco E, Cifuentes M, Gahagan S. Obesity is
associated with acute inflammation in a sample of adolescents. Pediatric Diabetes 2014;
epub ahead of print.
4. Zanni MV, Burdo TH, Makimura H, Williams KC, Grinspoon SK. Relationship between
monocyte/macrophage activation marker soluble CD163 and insulin resistance in obese
and normal-weight subjects. Clin Endocrinol (Oxf) 2012;77:385–390.
5. Gokulakrishnan K, Deepa R, Mohan V, Gross MD. Soluble P-selectin and CD40L levels
in subjects with prediabetes, diabetes mellitus, and metabolic syndrome--the Chennai
Urban Rural Epidemiology Study. Metab Clin Exp 2006;55:237–242.
6. Parkner T, Sørensen LP, Nielsen AR, et al. Soluble CD163: a biomarker linking
macrophages and insulin resistance. Diabetologia 2012;55:1856–1862.
7. Ortega Martinez de Victoria E, Xu X, Koska J, et al. Macrophage Content in
Subcutaneous Adipose Tissue: Associations With Adiposity, Age, Inflammatory Markers,
and Whole-Body Insulin Action in Healthy Pima Indians. Diabetes 2008;58:385–393.
8. Lê K-A, Mahurkar S, Alderete TL, et al. Subcutaneous adipose tissue macrophage
infiltration is associated with hepatic and visceral fat deposition, hyperinsulinemia, and
stimulation of NF-κB stress pathway. Diabetes 2011;60:2802–2809.
9. Fabbrini E, Cella M, McCartney SA, et al. Association between specific adipose tissue
CD4+ T-cell populations and insulin resistance in obese individuals. Gastroenterology
2013;145:366–74.e1–e3.
10. Ahlin S, Sjöholm K, Jacobson P, et al. Macrophage Gene Expression in Adipose Tissue is
! 74
Associated with Insulin Sensitivity and Serum Lipid Levels Independent of Obesity.
Obesity 2013;21:E571–576.
11. Dalmas E, Venteclef N, Caer C, et al. T cell-derived IL-22 amplifies IL-1β-driven
inflammation in human adipose tissue: relevance to obesity and type 2 diabetes. Diabetes
2014; epub ahead of print.
12. Schindelin J, Arganda-Carreras I, Frise E, et al. Fiji: an open-source platform for
biological-image analysis. Nat Meth 2012;9:676–682.
13. Yuan M. Reversal of Obesity- and Diet-Induced Insulin Resistance with Salicylates or
Targeted Disruption of Ikkbeta. Science 2001;293:1673–1677.
14. Arkan MC, Hevener AL, Greten FR, et al. IKK-beta links inflammation to obesity-
induced insulin resistance. Nat Med 2005;11:191–198.
15. Kim JK, Kim YJ, Fillmore JJ, et al. Prevention of fat-induced insulin resistance by
salicylate. J Clin Invest 2001;108:437–446.
16. Shoelson SE, Lee J, Yuan M. Inflammation and the IKK beta/I kappa B/NF-kappa B axis
in obesity- and diet-induced insulin resistance. Int J Obes Relat Metab Disord 2003;27
Suppl 3:S49–S52.
17. Cinti S, Mitchell G, Barbatelli G, et al. Adipocyte death defines macrophage localization
and function in adipose tissue of obese mice and humans. The Journal of Lipid Research
2005;46:2347–2355.
18. Shoelson SE, Lee J, Goldfine AB. Inflammation and insulin resistance. J Clin Invest
2006;116:1793–1801.
19. Wentworth JM, Naselli G, Brown WA, et al. Pro-inflammatory CD11c+CD206+ adipose
tissue macrophages are associated with insulin resistance in human obesity. Diabetes
2010;59:1648–1656.
20. Toledo-Corral CM, Alderete TL, Hu HH, et al. Ectopic fat deposition in prediabetic
! 75
overweight and obese minority adolescents. J Clin Endocrinol Metab 2013;98:1115–1121.
21. Goldfine AB, Silver R, Aldhahi W, et al. Use of salsalate to target inflammation in the
treatment of insulin resistance and type 2 diabetes. Clin Transl Sci 2008;1:36–43.
22. Goldfine AB, Fonseca V, Jablonski KA, et al. The effects of salsalate on glycemic control
in patients with type 2 diabetes: a randomized trial. Ann Intern Med 2010;152:346–357.
23. Faghihimani E, Aminorroaya A, Rezvanian H, Adibi P, Ismail-Beigi F, Amini M.
Salsalate improves glycemic control in patients with newly diagnosed type 2 diabetes.
Acta Diabetol 2013;50:537–543.
24. Goldfine AB, Conlin PR, Halperin F, et al. A randomised trial of salsalate for insulin
resistance and cardiovascular risk factors in persons with abnormal glucose tolerance.
Diabetologia 2013;56:714–723.
25. Koska J, Ortega E, Bunt JC, et al. The effect of salsalate on insulin action and glucose
tolerance in obese non-diabetic patients: results of a randomised double-blind placebo-
controlled study. Diabetologia 2009;52:385–393.
26. Fleischman A, Shoelson SE, Bernier R, Goldfine AB. Salsalate improves glycemia and
inflammatory parameters in obese young adults. Diabetes Care 2008;31:289–294.
27. Faghihimani E, Aminorroaya A, Rezvanian H, Adibi P, Ismail-Beigi F, Amini M.
Reduction of insulin resistance and plasma glucose level by salsalate treatment in persons
with prediabetes. Endocr Pract 2012;18:826–833.
28. Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of Childhood and Adult Obesity
in the United States, 2011-2012. JAMA 2014;311:806.
29. Centers for Disease Control and Prevention. National diabetes fact sheet: national
estimates and general information on diabetes and prediabetes in the United States, 2011.
US Department of Health and Human Services, Centers for Disease Control and
Prevention 2011.
! 76
30. National Center for Health Statistics. Summary Health Statistics for U.S. Adults: National
Health Interview Survey, 2010. US Department of Health and Human Services, Centers
for Disease Control and Prevention 2012;10:1–217.
31. Katz A, Nambi SS, Mather K, et al. Quantitative insulin sensitivity check index: a simple,
accurate method for assessing insulin sensitivity in humans. Journal of Clinical
Endocrinology & Metabolism 2000;85:2402–2410.
32. Burrows TL, Martin RJ, Collins CE. A systematic review of the validity of dietary
assessment methods in children when compared with the method of doubly labeled water.
J Am Diet Assoc 2010;110:1501–1510.
33. Godin G. The Godin-Shephard Leisure-Time Physical Activity Questionnaire. Health &
Fitness Journalof Canada 2011;4:18–22.
34. Kern PA, Ranganathan S, Li C, Wood L, Ranganathan G. Adipose tissue tumor necrosis
factor and interleukin-6 expression in human obesity and insulin resistance. Am J Physiol
Endocrinol Metab 2001;280:E745–E751.
35. Brinkley TE, Hsu F-C, Beavers KM, et al. Total and abdominal adiposity are associated
with inflammation in older adults using a factor analysis approach. J Gerontol A Biol Sci
Med Sci 2012;67:1099–1106.
36. Dzienis-Straczkowska S, Straczkowski M, Szelachowska M, Stepien A, Kowalska I,
Kinalska I. Soluble tumor necrosis factor-alpha receptors in young obese subjects with
normal and impaired glucose tolerance. Diabetes Care 2003;26:875–880.
37. Hu HH, Kim H-W, Nayak KS, Goran MI. Comparison of fat-water MRI and single-voxel
MRS in the assessment of hepatic and pancreatic fat fractions in humans. Obesity (Silver
Spring) 2010;18:841–847.
38. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data
capture (REDCap)--a metadata-driven methodology and workflow process for providing
translational research informatics support. J Biomed Inform 2009;42:377–381.
! 77
39. Fernández-Real JM, López-Bermejo A, Ropero A-B, et al. Salicylates increase insulin
secretion in healthy obese subjects. Journal of Clinical Endocrinology & Metabolism
2008;93:2523–2530.
40. Qu H-Q, Li Q, Rentfro AR, Fisher-Hoch SP, McCormick JB. The definition of insulin
resistance using HOMA-IR for Americans of Mexican descent using machine learning.
PLoS ONE 2011;6:e21041.
41. Hundal RS, Petersen KF, Mayerson AB, et al. Mechanism by which high-dose aspirin
improves glucose metabolism in type 2 diabetes. J Clin Invest 2002;109:1321–1326.
42. Jung TW, Youn B-S, Choi HY, et al. Salsalate and adiponectin ameliorate hepatic
steatosis by inhibition of the hepatokine fetuin-A. Biochem Pharmacol 2013;86:960–969.
43. Shehzad A, Iqbal W, Shehzad O, Lee YS. Adiponectin: Regulation of its production and
role in human diseases. Hormones 2012;11:6–18.
44. Kadowaki T, Yamauchi T, Okada-Iwabu M, Iwabu M. Adiponectin and its receptors:
implications for obesity-associated diseases and longevity. Lancet Diabetes Endocrinol
2014;2:8–9.
45. Kharroubi I, Rasschaert J, Eizirik DL, Cnop M. Expression of adiponectin receptors in
pancreatic beta cells. Biochem Biophys Res Commun 2003;312:1118–1122.
46. Ouchi N, Walsh K. A novel role for adiponectin in the regulation of inflammation.
Arteriosclerosis, Thrombosis, and Vascular Biology 2008;28:1219–1221.
47. Robinson K, Prins J, Venkatesh B. Clinical review: adiponectin biology and its role in
inflammation and critical illness. Crit Care 2011;15:221.
48. Scherer PE, Williams S, Fogliano M, Baldini G, Lodish HF. A novel serum protein
similar to C1q, produced exclusively in adipocytes. J Biol Chem 1995;270:26746–26749.
49. Goldfine AB, Fonseca V, Jablonski KA, et al. Salicylate (salsalate) in patients with type 2
diabetes: a randomized trial. Ann Intern Med 2013;159:1–12.
! 78
50. Goldfine AB, Buck JS, Desouza C, et al. Targeting inflammation using salsalate in
patients with type 2 diabetes: effects on flow-mediated dilation (TINSAL-FMD). Diabetes
Care 2013;36:4132–4139.

! 79
Chapter 5: Summary of Findings, Future Directions, and Conclusions

Summary
In the U.S., overweight and obesity is prevalent in 69% of adults (1), drastically increasing the
risk for IR, metabolic syndrome, and T2D. Despite this, the exact mechanism(s) linking excess
adiposity to these comorbidities is unknown. Therefore, the purpose of this dissertation was to
examine whether the location and/or inflammatory state of adipose tissue contributed to
metabolic dysfunction and diabetes risk. In order to accomplish this, we examined overweight
and obese AAs and Hispanic youth, a population that suffers from higher rates of obesity than
non-Hispanic whites (1). Studying these two minority groups is of particular importance since
they are also at increased risk for IR and T2D (2, 3). Overall, each of our studies provided us
with the unique opportunity to examine adipose tissue and how it relates to metabolic disease
risk. Using whole abdominal 3-Tesla MRI scans; we were able to uniquely identify LF as the
ectopic fat depot most strongly linked with increased risk for T2D. Furthermore, our novel
abdominal SAT biopsy method allowed us to demonstrate that anti-inflammatory salsalate
treatment improved metabolic outcomes without altering percent body fat, LF, adipose tissue
inflammation, or adipose tissue gene expression of important adipokines. Collectively, as
discussed at the end of this chapter, results from these studies lay the foundation for future trials
aimed at the prevention or reversal of cardiometabolic abnormalities seen during obesity.

LF, not VAT, Was Associated with Increased Risk for T2D
Interestingly, although AAs and Hispanics share
similar rates of obesity, AAs have higher levels of
SAAT and lower levels of VAT and LF than
Hispanics (Table 5-1) (4, 5). Lower levels of VAT
and LF imply that AAs should be metabolically
protected; however, we have shown that AAs share a
comparable risk for insulin resistance and T2D when
compared to Hispanics (4, 6, 7). This observation
conflicts with the prevailing hypothesis that elevated VAT (8, 9) and/or LF (10, 11) contribute to
obesity-associated metabolic disease risk. Therefore, we hypothesized that differences in the
AAs Hispanics
Total Fat Mass = =
SAAT ! "
VAT " !
LF " !
Risk for IR/T2D = =
Table 5-1. The AA vs. Hispanic Paradox
!
! 80
impact of either fat depot across ethnicity may explain why AAs exhibit similar risk factors for
T2D despite having less VAT and LF than Hispanics.

Although recent studies suggest that VAT and/or LF represent the causal link between obesity
and disease risk (9, 10, 12, 13), these studies are unable to determine which abdominal fat depot
drives metabolic risk since VAT and LF are highly correlated (14, 15). Despite this, studies in
Caucasian adults have provided evidence that LF is primarily associated with metabolic disease
in obese adults (10, 16). Therefore, the two aims of our first study were to examine 1) the
separate and combined effects of VAT and LF on risk factors for T2D in overweight and obese
minorities and 2) determine whether the association between these fat depots and risk factors for
T2D differed in AAs and Hispanics, thereby potentially explaining the paradox of similar
diabetes risk in AA at lowers level of VAT and LF.

In order to address these aims, we examined 358 overweight and obese AA and Hispanic
adolescents (8-25 years of age) with complete measures of SAAT, VAT, and LF. Our primary
outcomes of interest were derived from a FSIVGTT with Minimal Model where we explored SI,
AIR, and DI (17). DI is a constant derived from the product of AIR and SI. By examining DI, we
were able to assess β-cell function in the context of different insulin sensitivities. Plotting the
relationship between AIR and SI reveals a hyperbolic
sensitivity / secretion curve which characterizes DI. As
shown in Figure 5-1, healthy β-cells respond to a
decrease in SI by increasing their AIR (yellow circles
arrow). β-cells will compensate up to a point, after
which they become exhausted and do not adequate
respond, resulting in a “falling off” on the normal
curve and exhibiting a decreased DI (red circle and
arrow). Studies have shown that falling off of the DI curve is predictive of the development of
T2D in Pima Indians and women with gestational diabetes (17, 18). Therefore, we were able to
utilize these robust measures in our first study to examine how high VAT and LF related to
increased risk for diabetes.

!
!
Figure 5-1. Disposition Index
!
! 81
Since VAT and LF are inter-correlated, we classified participants into 1 of 4 groups based on
high/low LF and high/low VAT (19, 20). Figure 5-2 shows representative MRI images from
individuals who fell into the each of the 4 categories based on LF and VAT. Following this, we
used an ANCOVA to examine the main effects of our 4 groups on SI, AIR, and DI. Overall, as
outlined in chapter 2, our unique
categorical and statistical approach
allowed us to disassociate the inter-
relationships between VAT and LF. As
hypothesized, we found that ethnicity
modified the effect of high/low LF on
each of our outcomes of interest. To
further explore the relationship between
LF and VAT with risk factors for T2D,
we also used regression analysis to
complement the approach of classifying participants based on high/low VAT and LF.

In summary, results from this first study are the first to demonstrate that high LF has a more
pronounced association with
decreased SI and β-cell function
among AAs than Hispanics.
Specifically, AAs with high LF had
a 49% lower SI when compared to
those with low LF. Based on the
hyperbolic relationship between SI
and AIR, we anticipated that AAs
with high LF would have to
compensate with an AIR that would be double that of AAs with low LF. Despite this, we found
that AAs with high LF had only a 22% higher AIR. Conversely, Hispanics with high LF had a
24% lower SI and a 31% higher AIR compared to those with low LF. These results suggest that
Hispanics exhibit appropriate β-cell compensation for LF induced insulin resistance where AAs
do not (Figure 5-3). Supporting this conclusion, AAs with high LF had a 42% lower DI than
!
Figure 5-2. Four Groups Based on LF and VAT
!
!
Figure 5-3. β-cell Compensation in Response to LF
!
! 82
those with low LF while this relationship was not observed in Hispanics. These findings may
explain why AAs are at a similar risk for T2D, despite having lower levels of VAT and LF when
compared to Hispanics.

Overall, this study was limited in that we were not able to assess liver insulin sensitivity or
clearance, making it possible that our DI findings reflected changes in the liver and/or β-cell
section of insulin in response to decreased SI. We also could not determine whether our SI was
primarily based on muscle, chronic hyperinsulinemia induced by liver insulin resistance, and/or
reduced insulin clearance due to high LF. Therefore, it is feasible that ethnic differences were
partially driven by disparities in insulin clearance rates. To address this issue, future studies
should utilize similar measures of VAT and LF, but also incorporate a hyperinsulinemic-
euglycemic clamp in order to measure peripheral and liver insulin sensitivity. This would allow
more definitive conclusions to be made regarding the possible effects of elevated LF on β-cell
function. At the same time, recent studies have suggested that pancreatic fat may affect risk for
T2D (21-23). For this reason, quantifying pancreatic fat could also help to elucidate any
additional ethnic differences that could account for our findings.

Adipose Tissue Inflammation Leads to Ectopic Fat and Metabolic Dysfunction
Since our first study found that LF was strongly associated
with risk factors for T2D, we next explored the role of
adipose tissue inflammation in contributing to increased
ectopic fat and metabolic dysfunction. With excess energy
intake, insulin stimulates adipose tissue to increase storage
of FFA as triglycerides and decrease lipolysis (24). This
process results in increased total body fat and expansion of
the SAT and VAT depots (12, 25). Adipose tissue will
continue to store excess energy until demands outweigh
the ability of the adipose tissue to expand and/or become
adequately vascularized (25). Once this occurs, there is a
marked increase in adipose tissue inflammation and IR
(Figure 5-4). Adipose tissue inflammation and IR is
!
Figure 5-4. Fat Inflammation and
Metabolic Dysfunction
!
! 83
characterized by a decreased ability to take-up FFA and inhibit lipolysis. Therefore, studies
suggest that adipose tissue inflammation and dysfunction are the catalyst for ectopic fat
accumulation through increased plasma levels of FFA. In this regard, bot the VAT and SAT fat
depots likely contribute to this inflammatory state. Namely, VAT releases FFA and pro-
inflammatory cytokines into the liver and pancreas via the portal vein while large volumes of
SAT, which greatly contribute to total adiposity, are thought to equally contribute this state of
metabolic dysfunction (26).

Given the connections between adipose tissue inflammation and ectopic fat accumulation, the
second objective of this dissertation focused on decreasing adipose tissue inflammation. This
undertaking was largely driven by a previous study from our lab, which found that those with
CLS in their adipose tissue had higher LF, VAT, and fasting insulin levels compared to those
without CLS (27). Collectively, these studies lead us to explore treatment strategies that could be
used to decrease inflammation and improve risk factors for T2D. As discussed in the chapter 1,
anti-inflammatory salsalate therapy has been shown to improve glucose and lipid homeostasis in
patents with IFG, IGT and/or T2D (28-31). Reports from one group suggests that salsalate
inhibits systemic inflammation and NF-κB activity in peripheral blood mononuclear cells (28)
and adipose tissue (31). We therefore conducted a randomized, double-blind and placebo-
controlled trial of salsalate in obese Hispanic young adults without T2D to determine 1) whether
salsalate improves glycemic control and 2) whether metabolic improvements are mediated by
attenuation of adipose tissue inflammation. However, in order to conduct this study, we needed
to develop an adipose tissue biopsy method that would allow us to sample enough adipose tissue
to perform an array of studies surrounding adipose tissue inflammation, including flow cytomery,
histology, and gene expression

Sampling Abdominal Subcutaneous Adipose Tissue With a Bergström Biopsy Needle
As previously discussed, upper body fat accumulation includes VAT and SAT. Within the
abdominal SAT depot, there are two distinct regions divided by the Scarpa’s fascia into sSAT
and dSAT (32, 33). Past studies have shown that VAT has the greatest linkage with risk for
inflammation, CVD, and metabolic syndrome (34-36). However, compared to sSAT, VAT and
dSAT share more similar inflammatory characteristics and associations with disease risk (32, 37,
! 84
38). Given the importance of adipose tissue inflammation in contributing to ectopic fat
accumulation and risk for T2D, we focused our efforts on exploring whether metabolic
improvements observed with salsalate were mediated by improvements in adipose tissue
inflammation. In order to achieve this goal, it became necessary to safely and consistently
sample large quantities of SAT before and after salsalate treatment. Traditionally, invasive
surgical procedures are required to obtain sufficient amounts of fat for a portfolio of laboratory
tests (37, 39). Since sSAT is easily accessible, and thought to reflect biological aspects of VAT
(37, 40), non-invasive aspiration and punch biopsy procedures haven been used to sample this
depot (40, 41). However, these methods typically acquire only 100-500 milligrams of sSAT (27,
42, 43), far less than what is needed for research studies that may include flow cytometry, RNA
extraction, ex vivo expression of molecular and post-translational protein mediators, and
histology. Thus, our second study aimed at determining: 1) whether a 6-mm Bergström side-
cutting biopsy needle could be used to safely obtain large amounts of intact SAT (≥1 gram) and
2) whether this method could be used in conjunction with ultrasound guidance to specifically
sample dSAT, which has greater similarities to VAT.

Overall, we performed 50 biopsies that were obtained from 29 participants using the Bergström
biopsy needle. Eighteen of these biopsies were performed under ultrasound guidance, whereby
we were able to successfully sample dSAT from below the Scarpa’s fascia. Importantly, the
average weight of SAT sampled was 1.5 grams. Overall, the 6-mm Bergström needle yielded
substantially more than what has been obtained from superficial procedures and for the first time
allowed us to sample dSAT by a percutaneous approach. Using this methodology, we were
therefore able to consistently sample large quantities of intact SAT before and after a 4-week
clinical trial aimed at examining adipose tissue inflammation.

Interestingly, our biopsy results were limited in that we observed somewhat low levels of cell
viability (Figure 5-5) for cell sorting (44), perhaps contributing to our inability to adequately
characterize pro-inflammatory and anti-inflammatory macrophages. In order to explore immune
cells present in adipose tissue, specimens must be digested to yield the stromal-vascular fraction
(SVF). The SVF contains a number of leukocytes, including T cells, B cells, NK cells, mast cells,
neutrophils, and macrophages. To date, most studies have used rodent models of obesity and
! 85
aging to characterize various cell populations in adipose tissue (25, 45, 46). Dr. Carey Lumeng,
an expert in rodent adipose tissue cell sorting,
has reported that cell yield is significantly
impacted by the procedure used to digest
adipose tissue. In human adipose tissue,
collagenase alone has been shown to produce
better cell yields and viability than when
combined with additional proteases, and
liberase digestion has been shown to degrade
cell surface expression of immune cell markers
(44). Perhaps due to the difficulty in analyzing cell subsets (47), it is only recently that studies
have examined immune cells in human adipose tissue (11, 48-50) In this regard, one study
reported a limited ability to isolate T cells (11) while the remaining studies did not provide
information regarding cell viability and number of immune cells isolated (11, 48, 49). Therefore,
future studies should adequately develop and characterize human adipose tissue cell sorting
methodologies before utilizing this technique to quantify specialized immune cells.

Salsalate Improved Glycemia but Did Not Alter Adipose Tissue Inflammation
Armed with the ability to sample large quantities of intact abdominal SAT, we successfully
completed a randomized, double-blind and placebo-controlled trial of salsalate in obese Hispanic
young adults without T2D to determine 1) whether salsalate improves glycemic control and 2)
whether metabolic improvements are mediated by attenuation of adipose tissue inflammation.
Based on previous studies (28, 31, 51), we hypothesized that salsalate treatment would result in
metabolic improvements that would be mediated by decreases in adipose tissue inflammation.

As previously shown in Table 4-1, 24 participants (salsalate, n=11 and placebo, n=13) and were
included in our analysis. This study was unique in that, in addition to assessing metabolic
changes that resulted from salsalate treatment, we also sampled large quantities of adipose tissue
and measured body composition and ectopic fat accumulation with DEXA and MRI. These
robust measures allowed us to observe improvements in glycemia, insulin metabolism, and
!
Figure 5-5. Adipose Tissue Viability!
! 86
plasma FFA after salsalate treatment that could not be related to changes in adipose tissue
inflammation, total adiposity, LF, or changes in diet or physical activity (Table 5-2).

Briefly, although there were no improvements in estimates of insulin sensitivity or resistance
following salsalate treatment, we observed increases in HOMA-B, insulin AUC, and insulin to
C-peptide ratios. These findings
suggest that metabolic
improvements resulted from
decreased insulin clearance.
Contrary to our hypothesis, anti-
inflammatory salsalate treatment
did not significantly alter plasma
markers of inflammation (sTNFrII,
hsIL-6, sCD14, and sCD163), fat
cell size, CLS in adipose tissue, or
adipose gene expression of
adipokines, immune cell markers,
or cytokines downstream of NF-
κB. Collectively, these findings
suggest that metabolic improvements resulting form salsalate treatment occurred in the absence
of significant alterations to adiposity or systemic or adipose tissue inflammation (Table 5-2).

Salsalate Increased Adiponectin and Decreased Insulin Clearance
Consistent with previous reports (28, 29, 31, 51, 52), we saw large increases in adiponectin
levels after salsalate treatment. Although, anti-inflammatory salsalate did not exert metabolic
improvements through decreases in systemic or adipose tissue inflammation, increased plasma
adiponectin levels could explain observed decreases in glucose and fasting FFA. This is
particularly likely since adiponectin, an adipose-derived hormone, has been shown to stimulate
insulin sensitivity and modulate hepatic glucose production, fatty acid oxidation, and lipid
synthesis (53-57). Prior to this study, we speculated that decreased adipose tissue inflammation
was the primary cause of increased adiponectin levels. However, due to the fact that adipose
Salsalate Placebo
Fasting Glucose -3.4% −
Fasting FFA -42.5% −
HOMA-B +47.2% −
Insulin (INS) AUC +38% −
Fasting INS:C-Peptide Ratio +57.6% −
Adiponectin +27.7% −
Total Fat (%) / LF (%) − −
Crown-like Structures − −
Monocytes/Macrophages (%) − −
Adipose Tissue Gene Expression
§
− −
Table 5-2. Improvements Not Mediated by Adipose Tissue
!
§
Adipokines, immune cell markers, and cytokines (except IL-1β).!
! 87
tissue inflammation was largely unaffected by salsalate, our data suggest that salsalate directly
increased adipose tissue secretion of adiponectin and/or altered its clearance. Surprisingly,
adipose tissue gene expression of adiponectin was unchanged following salsalate treatment. Thus,
results from our study agree with the possibility that increased circulating adiponectin levels did
not result from increased adipocyte gene expression, but were due to either enhanced protein
translation and secretion or decreased clearance by the liver and/or kidneys (58). Alternatively,
decreased insulin clearance could have caused transient states of hyperinsulinemia, thereby
indirectly suppressing hepatic glucose production and/or stimulating glucose and FFA uptake by
peripheral tissues. This theory is supported by our findings that salsalate increased fasting insulin
to C-peptide ratios and are also reinforced by a recent study that reported that salsalate did not
affect on whole-body insulin sensitivity or systemic glucose disposal (31).

All Roads Lead to the Liver
The central thread connecting each of these potential mechanisms is undoubtedly the liver.
Namely, adiponectin has been shown to have profound affects on liver metabolism and
inflammation. Additionally, as the primary site of insulin and adiponectin clearance, this organ is
positioned to be a likely target of salsalate treatment. In fact, a recent study found that salsalate
treatment ameliorated liver steatosis in high fat fed rats through decreases in fetuin-A expression
via the AMPK-NF-κB dependent pathway (57). In our clinical trial, we used advanced MRI
imaging to show that metabolic improvements occurred without changes in LF percent; however,
we can not be certain that this methodology was sensitive enough to capture small, yet important,
alterations in hepatocyte lipid content. Furthermore, we did not assess liver inflammation.
Likewise, while we included measures of glucose, insulin and C-peptide from an OGTT, we
were unable to assess possible treatment effects on hepatic insulin sensitivity or glucose
production. Despite this, a 2-week long salsalate study found that lower fasting glucose levels
did not result from decreased endogenous glucose production (28). Given the complex nature of
these studies, future trials should include estimates of liver insulin sensitivity while also
incorporating liver biopsies. Although invasive, liver biopsies would provide the most
meaningful information regarding any potential changes in liver inflammation and insulin
signaling at the cellular level. Perhaps, studies among patients with NAFLD would present an
ideal study population to treat with anti-inflammatory agents and to explore such measures.
! 88
Future Directions: Additional Approaches to Decreasing Inflammation
Although we found that salsalate improved metabolic indices without reducing inflammation, a
vast amount of data continue to support a causal link between adipose tissue inflammation, LF,
and increased risk for cardiometabolic disease. Therefore, treatment strategies should continue to
focus on approaches that have the ability to decrease obesity-associated inflammation. In this
regard, preliminary cross-sectional data from our salsalate study suggests that specific dietary
alterations have the potential to dampen systemic and adipose tissue inflammation. Further,
current research indicates that alterations to the gut microbiome also have the ability to reduce
systemic, liver, and fat inflammation. In the reminder of this discussion, we will examine these
findings as well as future treatment strategies utilizing these approaches as means to decrease the
chronic low-grade inflammation observed during obesity.

Dietary Sugar is Associated with Adipose Tissue Inflammation
As previously described, our clinical trial among obese Hispanic young adults included adipose
tissue biopsies and 24-hour dietary recalls, allowing us
to examine cross-sectional relationships between sugar
intake and adipose tissue inflammation. Preliminary
data indicated that those with CLS in their adipose
tissue consumed more servings of sugary sweetened
beverages (SSB) and fructose than those without CLS
(Figure 5-6). This is particularly important since SSB
are a major source of added sugars, which contain 50-
60% more fructose than glucose (59). In fact, we also
found that, independent of total sugar intake and
percent body fat, each gram of added fructose
consumption increased the odds of having CLS in
adipose tissue by 6% (P=0.04). Furthermore, consumption of SSB, a primary source of dietary
fructose, was positively associated with plasma sTNFrII (r = 0.43; P=0.04). Given that excessive
dietary sugar consumption, including SSB, increases liver fat and cardiometabolic disease risk
(60-62), these preliminary results suggest that dietary reductions in SSB/fructose have the
potential to decrease adipose tissue inflammation.
!
Figure 5-6. Diet and Fat Inflammation
! 89
The Gut Microbiome is Altered by Diet and Obesity
Under normal circumstance, the gut microbiota play a vital role in human health by breaking
down complex dietary carbohydrates and synthesizing vitamins (63, 64). Within a healthy gut,
the intestinal mucosa acts as a barrier between the circulation and microbes and microbial
products in the lumen (63). Although the composition of the gut microbiota is well established
by one year of age (65), adult gut microbiota can
vary considerably due to diet (63). In this regard,
studies have shown that the composition of the
gut microbiota is altered during obesity where
the typical Western diet (high sugar and fat)
decreases microbial diversity, increases
Firmicutes, and decreases Bacteroidetes (66-68)
(Figure 5-7). These microbial alterations result
in intestinal bacterial overgrowth, breakdown of the intestinal barrier, bacterial translocation, and
chronic endotoxemia (e.g., LPS). These observations suggest that increased sugar consumption
has metabolic and inflammatory consequences that are mediated by the gut as discussed below.

Gut-Liver and Adipose Tissue Axis Contributes to Inflammation
The gut and liver are intrinsically connected, where blood drains from the gut via the portal vein
and is delivered to the liver. This crosstalk makes it
likely that gut microbiota play an important role in liver
insulin sensitivity, inflammation, and fat accumulation
(69, 70). Specifically, microbes and microbial products
carried in the blood from the gut are filtered by the liver
before reaching the systemic circulation (63). Therefore,
with chronic endotoxemia, unfavorable gut microbiota
can result in increased liver inflammation, insulin
resistance, and ectopic fat accumulation (Figure 5-8 and
5-9). As shown in Figure 5-9, the liver is not the only site affected by the gut microbiome.
Chronic bacterial translocation, due to increased intestinal permeability, can also contribute to
adipose tissue inflammation where animal studies have shown that LPS can induce adipose
!
Figure 5-7. Obesity and the Gut Microbiota !
!

Kemp et al, 2013 and Munukka et al, 2014

Figure 5-8. Gut-Liver and Adipose Tissue
! 90
tissue inflammation and increase serum lipids through TLR4 and NF-κB pathways in the liver
and adipose tissue (71-73). These findings appear relevant to humans since we found that those
with inflamed adipose tissue had increased expression of LPS binding protein and CD14 (27),
which interact with TLR4 and modulate adipose tissue inflammation and insulin resistance (74).
Finally, preliminary cross-sectional data
from our salsalate study suggest that
increased SSB consumption is related to
elevated levels of plasma sCD14 (r=0.35;
P=0.10). Since LPS binds to CD14
+

monocytes and macrophages, producing
sCD14, these findings indicate important
interactions between sugar consumption,
the gut microbiome, bacterial translocation,
and inflammation.

Decreasing Sugar Intake and Altering the Gut Microbiome to Decrease Inflammation
Not only have several studies found that dietary fructose is a major culprit in liver fat deposition
(62, 75, 76), increased sugar intake is also associated with increased systemic, liver, and adipose
tissue inflammation as well as negative alterations to the gut microbiome. These findings suggest
that interventions aimed at decreasing sugar, specifically SSB/fructose, offer another therapeutic
strategy that can be used to decrease ectopic fat and obesity-associated inflammation.
Furthermore, given the link between diet, the gut microbiome, and metabolic health, treatments
geared toward altering the composition of the flora (e.g., dietary alterations, probiotics,
transplants) (67, 77, 78) in conjunction with dietary alterations may prove the most beneficial.
For example, probiotic treatments and microbial transplants have been shown to decrease
bacterial translocation and improve metabolic health (Figure 5-8) (68, 79-82) where daily
consumption of probiotic supplements regulate blood sugar, decrease inflammation, and improve
LF severity among children with NAFLD (80, 82). Collectively, the interplay between diet and
the gut microbiome suggests that combination therapies may prove the most useful in treating
obese Hispanics who are at an increased risk for IR, T2D, and NAFLD (83-85).

Figure 5-9. Gut Microbiome and Inflammation
!
! 91
Conclusions
In summary, the central goal of this dissertation has been to understand the link between
increased adiposity and metabolic disease risk. In this regard, we began our investigation into the
importance of ectopic fat accumulation. From this study, we found that LF was more strongly
related to increased risk for T2D than VAT. Given that increased adipose tissue and insulin
resistance are believed to result in ectopic fat accumulation and metabolic dysfunction, we
decided to explore treatment strategies that had the potential to decrease adipose tissue
inflammation. In this regard, we utilized anti-inflammatory salsalate due to its safety, tolerability,
and ability to improve glucose and lipid homeostasis and decrease NF-κB activity (28, 31, 86).
Although contrary to our hypothesis, we found that salsalate lowered glycemia and fasting FFA
in the absence of any significant alterations to systemic or adipose tissue inflammation. Instead,
results suggested that salsalate exerted its metabolic effects through increases in adiponectin
and/or decrease in insulin clearance. Finally, preliminary cross-sectional data from our salsalate
study indicated that dietary factors, such as SSB/fructose consumption, contribute to increased
systemic and adipose tissue inflammation, possibly through increased bacterial translocation. At
the same time, emerging work in animals and humans suggests that gut microbiota play an
important role in adipose tissue dysfunction, liver inflammation, and metabolic disease risk (70).
Therefore, as outlined above, future studies should leverage results from this dissertation to
explore dietary interventions aimed at reducing SSB/fructose intake as well as various ways in
which favorable alterations can be made to the gut flora (e.g., fecal transplants, dietary
manipulations, and/or the use of pre/probiotics) in an effort to decrease inflammation and
improve metabolic health.

! 92
Chapter 5 References:
1. Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of Childhood and Adult Obesity
in the United States, 2011-2012. JAMA 2014;311:806.
2. Dabelea D, Bell RA, D'Agostino RB, et al. Incidence of diabetes in youth in the United
States. JAMA 2007;297:2716–2724.
3. National Center for Health Statistics. Summary Health Statistics for U.S. Adults: National
Health Interview Survey, 2010. US Department of Health and Human Services, Centers
for Disease Control and Prevention 2012;10:1–217.
4. Hasson RE, Adam TC, Davis JN, et al. Ethnic differences in insulin action in obese
African-American and Latino adolescents. J Clin Endocrinol Metab 2010;95:4048–4051.
5. Toledo-Corral CM, Alderete TL, Hu HH, et al. Ectopic fat deposition in prediabetic
overweight and obese minority adolescents. J Clin Endocrinol Metab 2013;98:1115–1121.
6. Goran MI. Ethnic-specific pathways to obesity-related disease: the Hispanic vs. African-
American paradox. Obesity (Silver Spring) 2008;16:2561–2565.
7. Goran MI, Bergman RN, Cruz ML, Watanabe R. Insulin resistance and associated
compensatory responses in African-American and Hispanic children. Diabetes Care
2002;25:2184–2190.
8. Bennett B, Larson-Meyer DE, Ravussin E, et al. Impaired Insulin Sensitivity and Elevated
Ectopic Fat in Healthy Obese vs. Nonobese Prepubertal Children. Obesity (Silver Spring)
2011;20:371–375.
9. Cruz ML, Bergman RN, Goran MI. Unique effect of visceral fat on insulin sensitivity in
obese Hispanic children with a family history of type 2 diabetes. Diabetes Care
2002;25:1631–1636.
10. Fabbrini E, Magkos F, Mohammed BS, et al. Intrahepatic fat, not visceral fat, is linked
with metabolic complications of obesity. Proc Natl Acad Sci USA 2009;106:15430–15435.
! 93
11. Fabbrini E, Cella M, McCartney SA, et al. Association between specific adipose tissue
CD4+ T-cell populations and insulin resistance in obese individuals. Gastroenterology
2013;145:366–74.e1–3.
12. Nazare J-A, Smith JD, Borel A-L, et al. Ethnic influences on the relations between
abdominal subcutaneous and visceral adiposity, liver fat, and cardiometabolic risk profile:
the International Study of Prediction of Intra-Abdominal Adiposity and Its Relationship
With Cardiometabolic Risk/Intra-Abdominal Adiposity. Am J Clin Nut 2012;96:714–726.
13. Ravussin E, Smith SR. Increased fat intake, impaired fat oxidation, and failure of fat cell
proliferation result in ectopic fat storage, insulin resistance, and type 2 diabetes mellitus.
Ann NY Acad Sci 2002;967:363–378.
14. Jakobsen M, Berentzen T, Sorensen T, Overvad K. Abdominal Obesity and Fatty Liver.
Epidemiologic Reviews 2007;29:77–87.
15. Korenblat KM, Fabbrini E, Mohammed BS, Klein S. Liver, muscle, and adipose tissue
insulin action is directly related to intrahepatic triglyceride content in obese subjects.
Gastroenterology 2008;134:1369–1375.
16. Fabbrini E, Tamboli RA, Magkos F, et al. Surgical Removal of Omental Fat Does Not
Improve Insulin Sensitivity and Cardiovascular Risk Factors in Obese Adults.
Gastroenterology 2010;139:448–455.
17. Bergman RN, Ader M, Huecking K, Van Citters G. Accurate assessment of beta-cell
function: the hyperbolic correction. Diabetes 2002;51 Suppl 1:S212–S220.
18. Lillioja S, Mott DM, Spraul M, et al. Insulin resistance and insulin secretory dysfunction
as precursors of non-insulin-dependent diabetes mellitus. Prospective studies of Pima
Indians. N Engl J Med 1993;329:1988–1992.
19. Szczepaniak LS, Nurenberg P, Leonard D, et al. Magnetic resonance spectroscopy to
measure hepatic triglyceride content: prevalence of hepatic steatosis in the general
population. Am J Phys 2005;288:E462–E468.
! 94
20. Taksali SE, Caprio S, Dziura J, et al. High visceral and low abdominal subcutaneous fat
stores in the obese adolescent: a determinant of an adverse metabolic phenotype. Diabetes
2008;57:367–371.
21. Szczepaniak LS, Victor RG, Mathur R, et al. Pancreatic steatosis and its relationship to β-
cell dysfunction in humans: racial and ethnic variations. Diabetes Care 2012;35:2377–
2383.
22. Targher G, Rossi AP, Zamboni GA, et al. Pancreatic Fat Accumulation and Its
Relationship with Liver Fat Content and Other Fat Depots in Obese Individuals. J
Endocrinol Invest 2012;35:748–753.
23. Heni M, Machann J, Staiger H, et al. Pancreatic fat is negatively associated with insulin
secretion in individuals with impaired fasting glucose and/or impaired glucose tolerance: a
nuclear magnetic resonance study. Diabetes Metab Res Rev 2010;26:200–205.
24. Maligie M, Crume T, Scherzinger A, Stamm E, Dabelea D. Adiposity, fat patterning, and
the metabolic syndrome among diverse youth: the EPOCH study. J Pediatr
2012;161:875–880.
25. Lumeng CN, DeYoung SM, Bodzin JL, Saltiel AR. Increased inflammatory properties of
adipose tissue macrophages recruited during diet-induced obesity. Diabetes 2007;56:16–
23.
26. Klein S. The case of visceral fat: argument for the defense. J Clin Invest 2004;113:1530–
1532.
27. Lê K-A, Mahurkar S, Alderete TL, et al. Subcutaneous adipose tissue macrophage
infiltration is associated with hepatic and visceral fat deposition, hyperinsulinemia, and
stimulation of NF-κB stress pathway. Diabetes 2011;60:2802–2809.
28. Goldfine AB, Silver R, Aldhahi W, et al. Use of salsalate to target inflammation in the
treatment of insulin resistance and type 2 diabetes. Clin Transl Sci 2008;1:36–43.
29. Goldfine AB, Fonseca V, Jablonski KA, et al. The effects of salsalate on glycemic control
! 95
in patients with type 2 diabetes: a randomized trial. Ann Intern Med 2010;152:346–357.
30. Faghihimani E, Aminorroaya A, Rezvanian H, Adibi P, Ismail-Beigi F, Amini M.
Salsalate improves glycemic control in patients with newly diagnosed type 2 diabetes.
Acta Diabetol 2013;50:537–543.
31. Goldfine AB, Conlin PR, Halperin F, et al. A randomised trial of salsalate for insulin
resistance and cardiovascular risk factors in persons with abnormal glucose tolerance.
Diabetologia 2013;56:714–723.
32. Kelley DE, Thaete FL, Troost F, Huwe T, Goodpaster BH. Subdivisions of subcutaneous
abdominal adipose tissue and insulin resistance. Am J Physiol Endocrinol Metab
2011;278:E941–E948.
33. Walker GE, Verti B, Marzullo P, et al. Deep subcutaneous adipose tissue: a distinct
abdominal adipose depot. Obesity (Silver Spring) 2007;15:1933–1943.
34. Item F, Konrad D. Visceral fat and metabolic inflammation: the portal theory revisited.
Obes Rev 2012;13 Suppl 2:30–39.
35. Tchernof A, Després JP. Pathophysiology of Human Visceral Obesity: An Update.
Physiological Reviews 2013;93:359–404.
36. Alvehus M, Burén J, Sjöström M, Goedecke J, Olsson T. The human visceral fat depot has
a unique inflammatory profile. Obesity 2010;18:879–883.
37. Tordjman J, Divoux A, Prifti E, et al. Structural and inflammatory heterogeneity in
subcutaneous adipose tissue: relation with liver histopathology in morbid obesity. J
Hepatol 2012;56:1152–1158.
38. Marinou K, Hodson L, Vasan SK, et al. Structural and functional properties of deep
abdominal subcutaneous adipose tissue explain its association with insulin resistance and
cardiovascular risk in men. Diabetes Care 2014;37:821–829.
39. Kolak M, Westerbacka J, Velagapudi VR, et al. Adipose tissue inflammation and
! 96
increased ceramide content characterize subjects with high liver fat content independent of
obesity. Diabetes 2007;56:1960–1968.
40. Bigornia SJ, Farb MG, Mott MM, et al. Relation of depot-specific adipose inflammation
to insulin resistance in human obesity. Nutr Diabetes 2012;2:e30.
41. Goodpaster B, Thaete F, Simoneau J, Kelley D. Subcutaneous abdominal fat and thigh
muscle composition predict insulin sensitivity independently of visceral fat. Diabetes
1997;46:1579–1585.
42. Campbell KL, Makar KW, Kratz M, Foster-Schubert KE, McTiernan A, Ulrich CM. A
pilot study of sampling subcutaneous adipose tissue to examine biomarkers of cancer risk.
Cancer Prev Res (Phila) 2009;2:37–42.
43. Daum SM, Knittle J, Roseman K, Rom WN, Holstein EC. A simple technique for fat
biopsy of PBB-exposed individuals. Environ Health Perspect 1978;23:183–185.
44. Hagman DK, Kuzma JN, Larson I, et al. Characterizing and quantifying leukocyte
populations in human adipose tissue: impact of enzymatic tissue processing. J Immunol
Methods 2012;386:50–59.
45. Lumeng CN, Liu J, Geletka L, et al. Aging is associated with an increase in T cells and
inflammatory macrophages in visceral adipose tissue. J Immunol 2011;187:6208–6216.
46. Lumeng CN, DelProposto JB, Westcott DJ, Saltiel AR. Phenotypic Switching of Adipose
Tissue Macrophages With Obesity Is Generated by Spatiotemporal Differences in
Macrophage Subtypes. Diabetes 2008;57:3239–3246.
47. Grant R, Youm Y-H, Ravussin A, Dixit VD. Quantification of adipose tissue leukocytosis
in obesity. Methods Mol Biol 2013;1040:195–209.
48. Klimcakova E, Roussel B, Kovacova Z, et al. Macrophage gene expression is related to
obesity and the metabolic syndrome in human subcutaneous fat as well as in visceral fat.
Diabetologia 2011;54:876–887.
! 97
49. Wentworth JM, Naselli G, Brown WA, et al. Pro-inflammatory CD11c+CD206+ adipose
tissue macrophages are associated with insulin resistance in human obesity. Diabetes
2010;59:1648–1656.
50. Nov O, Shapiro H, Ovadia H, et al. Interleukin-1β Regulates Fat-Liver Crosstalk in
Obesity by Auto-Paracrine Modulation of Adipose Tissue Inflammation and
Expandability. PLoS ONE 2013;8:e53626.
51. Koska J, Ortega E, Bunt JC, et al. The effect of salsalate on insulin action and glucose
tolerance in obese non-diabetic patients: results of a randomised double-blind placebo-
controlled study. Diabetologia 2009;52:385–393.
52. Fleischman A, Shoelson SE, Bernier R, Goldfine AB. Salsalate improves glycemia and
inflammatory parameters in obese young adults. Diabetes Care 2008;31:289–294.
53. Shehzad A, Iqbal W, Shehzad O, Lee YS. Adiponectin: Regulation of its production and
role in human diseases. Hormones 2012;11:6–18.
54. Kadowaki T, Yamauchi T, Okada-Iwabu M, Iwabu M. Adiponectin and its receptors:
implications for obesity-associated diseases and longevity. Lancet Diabetes Endocrinol
2014;2:8–9.
55. Kharroubi I, Rasschaert J, Eizirik DL, Cnop M. Expression of adiponectin receptors in
pancreatic beta cells. Biochem Biophys Res Commun 2003;312:1118–1122.
56. Ouchi N, Walsh K. A novel role for adiponectin in the regulation of inflammation.
Arteriosclerosis, Thrombosis, and Vascular Biology 2008;28:1219–1221.
57. Jung TW, Youn B-S, Choi HY, et al. Salsalate and adiponectin ameliorate hepatic
steatosis by inhibition of the hepatokine fetuin-A. Biochem Pharmacol 2013;86:960–969.
58. Robinson K, Prins J, Venkatesh B. Clinical review: adiponectin biology and its role in
inflammation and critical illness. Crit Care 2011;15:221.
59. Walker R, Dumke KA, Goran MI. Fructose content in popular beverages made with and
! 98
without high fructose syrup. Nutrition 2014; epub ahead of print.
60. Goran MI, Ulijaszek SJ, Ventura EE. High fructose corn syrup and diabetes prevalence: a
global perspective. Glob Public Health 2013;8:55–64.
61. Khitan Z, Kim DH. Fructose: A Key Factor in the Development of Metabolic Syndrome
and Hypertension. J Nutr Metab 2013; 682673.
62. Basaranoglu M, Basaranoglu G, Sabuncu T, Sentürk H. Fructose as a key player in the
development of fatty liver disease. World J Gastroenterol 2013;19:1166–1172.
63. Goel A, Gupta M, Aggarwal R. Gut microbiota and liver disease. J Gastroenterol Hepatol
2014; 29:1139-1148.
64. Neish AS. Microbes in gastrointestinal health and disease. Gastroenterology
2009;136:65–80.
65. Faith JJ, Guruge JL, Charbonneau M, et al. The long-term stability of the human gut
microbiota. Science 2013;341:1237439.
66. Ley RE, Turnbaugh PJ, Klein S, Gordon JI. Microbial ecology: human gut microbes
associated with obesity. Nature 2006;444:1022–1023.
67. Van den Abbeele P, Verstraete W, Aidy El S, Geirnaert A, Van de Wiele T. Prebiotics,
faecal transplants and microbial network units to stimulate biodiversity of the human gut
microbiome. Microb Biotechnol 2013;6:335–340.
68. Vrieze A, Van Nood E, Holleman F, et al. Transfer of intestinal microbiota from lean
donors increases insulin sensitivity in individuals with metabolic syndrome.
Gastroenterology 2012;143:913–6.e7.
69. Kemp DM. Does chronic low-grade endotoxemia define susceptibility of obese humans to
insulin resistance via dietary effects on gut microbiota? Adipocyte 2013;2:188–190.
70. Munukka E, Pekkala S, Wiklund P, et al. Gut-adipose tissue axis in hepatic fat
accumulation in humans. J Hepatol 2014; epub ahead of print.
! 99
71. Cani PD, Bibiloni R, Knauf C, et al. Changes in gut microbiota control metabolic
endotoxemia-induced inflammation in high-fat diet-induced obesity and diabetes in mice.
Diabetes 2008;57:1470–1481.
72. Cullberg KB, Larsen JØ, Pedersen SB, Richelsen B. Effects of LPS and dietary free fatty
acids on MCP-1 in 3T3-L1 adipocytes and macrophages in vitro. Nutr Diabetes
2014;4:e113.
73. Duca FA, Sakar Y, Lepage P, et al. Replication of obesity and associated signaling
pathways through transfer of microbiota from obese-prone rats. Diabetes 2014;63:1624–
1636.
74. Fernández-Real JM, Pérez del Pulgar S, Luche E, et al. CD14 modulates inflammation-
driven insulin resistance. Diabetes 2011;60:2179–2186.
75. Lê K-A, Ith M, Kreis R, et al. Fructose overconsumption causes dyslipidemia and ectopic
lipid deposition in healthy subjects with and without a family history of type 2 diabetes.
The American Journal of Clinical Nutrition 2009;89:1760–1765.
76. Faeh D, Minehira K, Schwarz J-M, et al. Effect of fructose overfeeding and fish oil
administration on hepatic de novo lipogenesis and insulin sensitivity in healthy men.
Diabetes 2005;54:1907–1913.
77. Delzenne NM, Neyrinck AM, Bäckhed F, Cani PD. Targeting gut microbiota in obesity:
effects of prebiotics and probiotics. Nat Rev Endocrinol 2011;7:639–646.
78. Kootte RS, Vrieze A, Holleman F, et al. The therapeutic potential of manipulating gut
microbiota in obesity and type 2 diabetes mellitus. Diabetes, Obesity, & Metabolism
2012;14:112–120.
79. Amar J, Chabo C, Waget A, et al. Intestinal mucosal adherence and translocation of
commensal bacteria at the early onset of type 2 diabetes: molecular mechanisms and
probiotic treatment. EMBO Mol Med 2011;3:559–572.
80. Asemi Z, Zare Z, Shakeri H, Sabihi S-S, Esmaillzadeh A. Effect of multispecies probiotic
! 100
supplements on metabolic profiles, hs-CRP, and oxidative stress in patients with type 2
diabetes. Ann Nutr Metab 2013;63:1–9.
81. Chang B, Sang L, Wang Y, Tong J, Zhang D, Wang B. The protective effect of VSL#3 on
intestinal permeability in a rat model of alcoholic intestinal injury. BMC Gastroenterol
2013;13:151.
82. Alisi A, Bedogni G, Baviera G, et al. Randomised clinical trial: the beneficial effects of
VSL#3 in obese children with non-alcoholic steatohepatitis. Aliment Pharmacol Ther
2014;39:1276–1285.
83. Ogden CL, Carroll MD, Curtin LR, Lamb MM, Flegal KM. Prevalence of high body mass
index in US children and adolescents, 2007-2008. JAMA 2010;303:242–249.
84. Centers for Disease Control and Prevention. National diabetes fact sheet: national
estimates and general information on diabetes and prediabetes in the United States, 2011.
US Department of Health and Human Services, Centers for Disease Control and
Prevention 2011.
85. Centers for Disease Control and Prevention (CDC). Differences in prevalence of obesity
among black, white, and Hispanic adults - United States, 2006-2008. MMWR Morb
Mortal Wkly Rep 2009;58:740–744.
86. Goldfine AB, Fonseca V, Jablonski KA, et al. Salicylate (salsalate) in patients with type 2
diabetes: a randomized trial. Ann Intern Med 2013;159:1–12.

Abstract (if available)

Abstract Studies have shown that obesity is linked to insulin resistance (IR) and type 2 diabetes, yet the exact mechanism(s) linking excess adiposity to increased disease risk is unknown. Studies suggest that visceral adipose tissue (VAT), and liver fat (LF) contribute to the development of IR and risk for T2D. However, this theory is challenged by the fact that overweight and obese African Americans (AAs) have less VAT and LF than Hispanics, yet share a similar risk for IR and T2D. Recent data suggest that excess adipose tissue contributes to chronic inflammation, thereby promoting the development of IR and other metabolic complications. Supporting this idea, recent studies have found that anti‐inflammatory drugs, such as salsalate, decrease markers of inflammation and improve glucose control in Caucasian adults with T2D and/or impaired glucose control. ❧ Overall, the purpose of this dissertation was multi‐faceted where we focused on both the location and inflammatory state of adipose tissue. In order to examine adipose tissue inflammation, we developed a novel biopsy method to sample large quantities of intact subcutaneous adipose tissue for a variety of research purposes (e.g., cell sorting, histology, gene expression). This allowed us to perform a randomized, placebo‐controlled trial in obese Hispanic young adults in order to determine whether any salsalate‐induced metabolic improvements were mediated by changes in adipose tissue inflammation. ❧ From these studies, we show that high LF was more strongly associated with decreased insulin sensitivity and β‐cell function among African Americans (AAs) compared to Hispanics. Our results suggest that Hispanics exhibited appropriate β‐cell compensation for liver fat induced insulin resistance where AAs did not. These findings may partially explain why AAs are a great risker for T2D, despite having lower levels of LF and VAT when compared to Hispanics. Our work focusing on adipose tissue inflammation led us to develop a novel biopsy method where we were able to sample large quantities of intact adipose tissue (~1.5 g). This method allowed us to systematically examine the effects of salsalate on various markers of adipose tissue inflammation. Overall, we confirmed that salsalate improved metabolic outcomes related to T2D risk in otherwise healthy obese Hispanics. Our finding suggests that these improvements occurred in the absence of any significant alterations to adipose tissue inflammation. Taken together, our results suggest that although specific patters of fat distribution and inflammation contribute to T2D risk, salsalate does not target adipose tissue inflammation. Future studies should examine other treatment strategies and/or therapeutic targets in an effort to decrease obesity‐associated inflammation as a means to decrease disease risk.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Quantity versus quality: how adipose tissue accumulation and immune cell profile associate with risk for type 2 diabetes in minority children and adults

PDF

Genetic and dietary determinants of nonalcoholic fatty liver disease in Hispanic children

PDF

Metabolic consequences of obesity-associated inflammation during puberty and perinatal development

PDF

Variation in insulin-like growth factor-2 binding protein 2 interacts with adiposity to alter insulin sensitivity in Mexican Americans

PDF

Role of sympathetic nervous system in metabolic function

PDF

Impact of cannabinoid receptor 2 stimulation on adipose tissue-resident ILC2s and its potential in alleviating obesity-induced type 2 diabetes mellitus

PDF

Omentectomy in the canine: a unique model to investigate the physiologic relationship between visceral fat and insulin action

PDF

Molecular mechanisms underlying the effects of CB1 antagonism in canine model of obesity

PDF

Mechanisms relating obesity to insulin resistance and the metabolic syndrome

PDF

Molecular mechanisms of young-onset type 2 diabetes: integration of diet and multi-omics biomarkers

PDF

Increased abdominal adiposity in adolescents with classical congenital adrenal hyperplasia due to 21-hydroxylase deficiency

PDF

Objectively measured physical activity and body fat distribution in overweight Hispanic and African American adolescents

PDF

The associations between ultra-processed food consumption and type 2 diabetes and obesity among young adults

PDF

The role of adipocytes in acute lymphoblastic leukemia cell migration and survival against daunorubicin

PDF

Targeted metabolic signatures and diet in associations with obesity and insulin resistance in young adults

PDF

The physiologic and pathophysiologic role of sympathetic nervous system induced pulsatile lipolysis in metabolism

PDF

The ADRB3 TRP64ARG variant and obesity in African American breast cancer cases

PDF

Renin-angiotensin system modulation for the prevention and treatment of metabolic dysfunction

PDF

Adaptive patterning during skin formation: assembly of vasculature and adipose tissue

PDF

Investigating a physiological pathway for the effect of guided imagery on insulin resistance

Asset Metadata

Creator Alderete, Tanya Lynn (author)

Core Title Ectopic fat and adipose tissue inflammation in overweight and obese African Americans and Hispanics

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Integrative Biology of Disease

Publication Date 07/08/2014

Defense Date 06/06/2014

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

Tag adipose tissue,adipose tissue biopsy,African Americans,Hispanics,Inflammation,insulin resistance,liver fat,OAI-PMH Harvest,obesity,salsalate,type 2 diabetes,visceral fat

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Goran, Michael (committee chair), Allayee, Hooman (committee member), Mittelman, Steven D. (committee member), Richey, Joyce (committee member), Sattler, Fred (committee member)

Creator Email tanya24lynn@gmail.com,tanyasmi@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-433182

Unique identifier UC11286525

Identifier etd-AldereteTa-2623.pdf (filename),usctheses-c3-433182 (legacy record id)

Legacy Identifier etd-AldereteTa-2623.pdf

Dmrecord 433182

Document Type Dissertation

Format application/pdf (imt)

Rights Alderete, Tanya Lynn

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA