Sex and Race Differences in Obesity-Related Genetic Susceptibility and Risk of Cardiometabolic Disease in Older US Adults

Hairui Yu; Nicole Armstrong; Greg Pavela; Kathryn Kaiser

doi:10.1001/jamanetworkopen.2023.47171

. 2023 Dec 8;6(12):e2347171. doi: 10.1001/jamanetworkopen.2023.47171

Sex and Race Differences in Obesity-Related Genetic Susceptibility and Risk of Cardiometabolic Disease in Older US Adults

Hairui Yu ^1,², Nicole Armstrong ³, Greg Pavela ¹, Kathryn Kaiser ^1,^✉

¹Department of Health Behavior, School of Public Health, University of Alabama at Birmingham

²Department of Family and Community Medicine, School of Medicine, University of Alabama at Birmingham

³Department of Epidemiology, School of Public Health, University of Alabama at Birmingham

Accepted for Publication: October 29, 2023.

Published: December 8, 2023. doi:10.1001/jamanetworkopen.2023.47171

^✉

Corresponding Author: Kathryn Kaiser, PhD, Department of Health Behavior, School of Public Health, The University of Alabama at Birmingham, 1720 Second Ave S, RPHB 227, Birmingham, AL 35294 (kakaiser@uab.edu).

Author Contributions: Drs Yu and Kaiser had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Yu, Pavela, Kaiser.

Acquisition, analysis, or interpretation of data: Yu, Armstrong, Kaiser.

Drafting of the manuscript: Yu, Kaiser.

Critical review of the manuscript for important intellectual content: All authors.

Statistical analysis: Yu, Armstrong, Pavela.

Administrative, technical, or material support: Yu, Armstrong, Kaiser.

Supervision: Kaiser.

Conflict of Interest Disclosures: Dr Yu reported receiving a HALE Graduate Fellowship for PhD Students in the Department of Health Behavior at University of Alabama at Birmingham. Drs Pavela and Kaiser reported being members of the University of Alabama at Birmingham NORC.

Funding/Support: The Reasons for Geographic and Racial Differences in Stroke (REGARDS) genetic study was supported by the National Institutes of Health National Heart, Lung, and Blood Institute grant R01HL136666 (MRI). The REGARDS study is supported by cooperative agreement U01 NS041588 from the National Institute of Neurological Disorders and Stroke, National Institutes of Health, US Department of Health and Human Services.

Role of the Funder/Sponsor: The funders of the REGARDS study had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of Neurological Disorders and Stroke, National Institute of Diabetes and Digestive and Kidney Diseases, or the National Institutes of Health.

Data Sharing Statement: See Supplement 2.

Additional Contributions: The authors thank the other investigators, the staff, and the participants of the REGARDS study for their valuable contributions. A full list of participating REGARDS investigators and institutions can be found at http://www.regardsstudy.org. We also thank Ryan Irvin, PhD, and Olivia Affuso, PhD (Department of Epidemiology, University of Alabama at Birmingham), for technical and editorial comments on earlier drafts of this paper. They were not compensated for their time.

^✉

Corresponding author.

PMCID: PMC10709778 PMID: 38064210

Key Points

Question

Are there race and sex differences in the association of FTO single-nucleotide variants (SNVs) with obesity phenotypes and cardiometabolic disease among self-identified Black and White US residents?

Findings

In this cross-sectional study of 10 447 participants, none of the tested FTO SNVs were significantly associated with the risk of obesity in Black participants, while 11 FTO SNVs were associated with the risk of obesity in White participants. White males had a higher risk of obesity than White females. The SNV rs1121980 was associated with increased risk of heart disease in Black participants, independent of obesity.

Meaning

The findings of this study suggest that sex-specific mechanisms for obesity risk and cardiometabolic diseases may require further investigation to inform genetic risk, and these may differ by race.

This cross-sectional study investigates sex differences in the association between FTO single-nucleotide variants and obesity traits among self-identified non-Hispanic Black and non-Hispanic White US adults.

Abstract

Importance

The fat mass and obesity–associated gene (FTO) is associated with obesity phenotypes, but the association is inconsistent across populations. Within-population differences may explain some of the variability observed.

Objective

To investigate sex differences in the association between FTO single-nucleotide variants (SNVs) and obesity traits among self-identified non-Hispanic Black and non-Hispanic White US adults, to examine whether the SNVs were associated with cardiometabolic diseases, and to evaluate whether obesity mediated the association between FTO SNVs and cardiometabolic diseases.

Design, Setting, and Participants

This cross-sectional study used data from the Reasons for Geographic and Racial Differences in Stroke (REGARDS) study, a US population-based cohort study with available genetic data (assayed in 2018) and phenotypic data at baseline (enrolled 2003-2007). Participants were aged 45 to 98 years at baseline. Data were analyzed from October 2021 to October 2022.

Exposures

Eleven SNVs in the FTO gene present among both Black and White participants.

Main Outcomes and Measures

Objectively measured obesity indicators (body mass index and waist-to-height ratio), objectively measured and/or self-reported cardiometabolic diseases (hypertension, stroke history, heart disease, and diabetes), and self-reported social-economic and psychosocial status.

Results

A total of 10 447 participants (mean [SD] age, 64.4 [9.7] years; 5276 [55.8%] women; 8743 [83.7%] Black and 1704 [16.3%] White) were included. In the White group, 11 FTO SNVs were significantly associated with obesity, hypertension, and diabetes using linear models (eg, body mass index: β = 0.536; 95% CI, 0.197-0.875), but none of the FTO SNVs were associated with obesity traits in the Black group. White males had a higher risk of obesity while White females had a higher risk of hypertension and diabetes. However, 1 FTO SNV (rs1121980) was associated with a direct increase in the risk of heart disease in Black participants not mediated by obesity (c′ = 0.145 [SE, 0.0517]; P = .01).

Conclusions and Relevance

In this cross-sectional study of obesity phenotypes and their association with cardiometabolic diseases, the tested FTO SNVs reflected sex differences in White participants. Different patterns of associations were observed among self-identified Black participants. Therefore, these results could inform future work discovering risk alleles or risk scores unique to Black individuals or further investigating genetic risk in all US residents.

Introduction

The fat mass and obesity–associated (FTO [OMIM 610966]) gene is a well-established obesity-associated gene identified in multiple genomewide association studies (GWAS) since 2007.¹ Its strong associations with body mass index, body composition, waist circumference, and metabolism have drawn wide attention from researchers.^2,3 FTO is extensively expressed in human fat tissues and skeletal muscles as well as the arcuate nucleus of the hypothalamus that controls energy balance and energy metabolism.^1,4 The critical role of the FTO variants in energy intake, energy consumption, and obesity has been demonstrated in previous studies.^5,6

The mechanistic pathways relative to the specific risk leading to obesity remain poorly understood.⁷ Many candidate genes are associated with obesity-related phenotypes, and sexual dimorphism has been demonstrated in human and animal models, not simply due to the presence of gonadal hormones but also to the sex chromosome complement.^8,9,10 Further complicating potential risk factors for obesity beyond obesity genes are the inconsistently observed ranks of prevalence between men and women within the same racial category in the United States as seen over time in the National Health and Nutrition Examination Survey cohorts.^11,12 Since the intervals when race data were reported separately, there are more than 10–percentage point gaps between men and women for obesity rates, but with Black women being higher than their male counterparts and White women being 10 or more percentage points lower than their male counterparts.¹² Besides obesity phenotypes, other sex differences in risk factors for cardiovascular disease, such as blood lipid levels and glucose regulation, have been reported.^13,14

While the number of possibly influential genes has continued to increase in the literature, FTO continues to be observed with some inconsistencies.^15,16 For example, research on the association between a genetic risk score and BMI across 5 birth cohorts using the Health and Retirement Study revealed that an additional copy of the A allele or the FTO allele (rs1558902: A) was significantly associated with a 0.37 (95% CI, 0.21-0.54)–unit higher BMI among 7482 people with primarily European identity and a 1.03 (95% CI, 0.27-1.79)–unit higher BMI among 1306 people with primarily African identity.¹⁷ Another meta-analysis of GWASs among 20 cohorts of individuals with European identity reported that FTO (rs9936385) was significantly associated with BMI and whole-body lean mass.¹⁸

Socioeconomic, racial, and sex differences have been reported in gene-environment interactions with obesity, along with an increased risk of cardiometabolic diseases.^13,19 Prior research concluded that socioeconomic status (income and/or education) was negatively associated with the obesity risk of adult women in developed countries.^19,20,21 Stronger obesity-related genetic effects have been associated with lower socioeconomic groups, such as those with less income or educational attainment.²² Previous analysis from the Reasons for Geographic and Racial Differences in Stroke (REGARDS) study shows that excess biological and social risk factors associated with coronary heart disease events in Black individuals remain a public health problem, along with a higher risk for premature death.^23,24,25,26 The psychological factors, such as depressive symptoms, also account for increased mortality across racial and income groups.^27,28

FTO is one of the most commonly identified obesity-related genes (eTable 1 in Supplement 1), but prior studies have not included a wide range of racial groups or social backgrounds across the lifespan. We aimed to extend this literature using FTO single-nucleotide variants (SNVs) to evaluate sex differences in obesity phenotypes in participants who self-identify as non-Hispanic Black and non-Hispanic White individuals to bridge the research gap.

To extend these previous studies reporting racial disparities in the United States, 4 associations were examined. First, the association between FTO SNVs (11 effect alleles: rs1558902: A; rs1121980: A; rs17817449: G; rs8050136: A; rs9935401: A; rs3751812: T; rs9936385: C; rs9939609: A; rs9941349: T; rs9930506: G; and rs9922708: T) and obesity indicators (body mass index [BMI, calculated as weight in kilograms divided by height in meters squared] and waist-to-height ratio [WHtR]) was investigated, controlling for sex and race. Second, we examined whether the association between FTO SNVs and obesity indicators varies by sex, controlling for age, socioeconomic status (education level, income level), and psychosocial status (perceived stress, depressive symptoms) among Black and White individuals separately. Third, based on any observed significant associations in the previous examination, we tested whether there is an association between FTO SNVs and cardiometabolic diseases (hypertension, history of stroke, heart disease, and diabetes) and whether this association varies by sex. Finally, we examined whether the association between FTO SNVs and cardiometabolic disease is mediated by obesity indicators and whether the mediating effect varies by sex.

Methods

Participants

The Reasons for Geographic and Racial Differences in Stroke (REGARDS) study is a large, longitudinal, population-based cohort focusing on risk for stroke in older men and women, heavily sampling from the southeastern United States. A detailed description of the study and its measurements has been published.^29,30 The study protocol for this research was approved by the institutional review board at University of Alabama at Birmingham, and all participants provided written informed consent. This report follows the Strengthening the Reporting of Genetic Association Studies (STREGA) reporting guidelines for cross-sectional studies of genetic association studies.³¹ REGARDS staff enrolled 30 239 unrelated non-Hispanic Black and non-Hispanic White participants, aged 45 years and older at baseline, with oversampling of Black individuals and residents of the southeastern US. Sociodemographic, psychological, and medical history data were collected using a computer-assisted telephone interview at baseline. During an in-home examination, researchers measured anthropometrics, conducted a medication inventory, and collected blood and urine specimens.

The subsample that made up the genotyped cohort (10 543 with complete baseline data) was reviewed for obesity indicators, ie, BMI and WHtR, for the present analysis. Participants with extreme values in obesity indicators (BMI ≤18.5 or ≥40 [10.29%]; waist circumference [WC] ≤60 or ≥180 cm [0.91%]) were excluded.³² The final analysis cohort was identified, with 9458 with BMI and 10 447 with WHtR. eFigure 1 in Supplement 1 illustrates the cohort selection process.

Genetic Assessment

The genetic exposures were the 11 available FTO SNVs located in the FTO gene that met the 5% minor allele frequency (MAF) and imputation quality (0.3) threshold and were present in Black and White strata (rs1558902: A; rs1121980: A; rs17817449: G; rs8050136: A; rs9935401: A; rs3751812: T; rs9936385: C; rs9939609: A; rs9941349: T; rs9930506: G; and rs9922708: T) (eTable 1 in Supplement 1). All SNVs in the study population did not deviate from the Hardy-Weinberg equilibrium. LDmatrix was used to determine linkage disequilibrium (LD) among the 11 SNVs by self-identified race group.³³ The eAppendix in Supplement 1 provides details of genotyping methods and quality control. LD blocks by race group were created based on an LD threshold of 0.8. For significant LD block proxy SNVs in the models for obesity traits, we used those for mediation analysis of cardiovascular disease status at baseline in an exploratory analysis, as defined later.

Study Outcomes

The primary outcomes of this study were obesity indicators and cardiometabolic disease. The obesity indicators based on researcher measured height, weight, and waist circumference were BMI and WHtR.^34,35 Aligned with REGARDS protocol, the variables of cardiometabolic diseases were defined as current hypertension (systolic blood pressure ≥130 mm Hg or diastolic blood pressure ≥80 mm Hg or self-reported use of medication to control blood pressure), stroke (baseline self-reported stroke history), heart disease (self-reported history of myocardial infarction, coronary artery bypass graft, angioplasty, or stenting or evidence of myocardial infarction via electrocardiogram), and type 2 diabetes defined as fasting glucose level of 126 mg/dL or greater; nonfasting glucose, 200 mg/dL or greater, or self-reported use of pills or insulin to manage blood glucose (to convert glucose to millimoles per liter, multiply by 0.0555).

Covariates

Participants self-reported sex (female and male), race (self-identified Black or White), and age at baseline. Black and White were the only categories available to participants in the REGARDS study. Self-reported income (annual household categories: <$20 000, $21 000-$34,000, $35 000-$74,000, or ≥$75 000), education levels (highest grade or year of school completed: <high school, high school, some college, or ≥college), perceived stress (Perceived Stress Scale),³⁶ and depression (Center for Epidemiologic Studies Depression Scale)³⁷ were included as covariates in the tested models.

Statistical Analysis

Descriptive statistics were analyzed for each race and sex group. All continuous variables were evaluated for normality of distribution and multicollinearity. Baseline obesity phenotypes, age, and psychosocial status variables were compared between race and sex groups using analysis of variance and χ² tests. The significance level was assessed by using the Bonferroni correction (α = .05 / No. of independent tests).³⁸ All analyses were conducted using SAS version 9.4 (SAS Institute) and SPSS version 28.0 (IBM Corp).

Multiple linear regression analysis was performed to examine the associations between the explanatory variables and the obesity indicators (BMI and WHtR) in each race group. For significant SNVs identified from the main models, an interaction term (the product of sex and number of copies of FTO SNVs) was calculated and tested separately in each race group with multiple linear regressions. Based on significant associations identified in regression models, per methods of Hayes,³⁹ mediation and moderation analyses with bootstrapping (5000 bootstrap samples) were performed to test whether, for each race group, an obesity indicator mediated the association between FTO SNVs and cardiometabolic diseases, which was potentially moderated by sex. Any with significant associations were evaluated in mediation models to investigate whether obesity played a role in indirect effects between sex groups. To augment power to detect genuine associations,^40,41,42 the top 10 genetic principal components (PC 1 to PC 10) were included in each model to account for genetic ancestry.^43,44 Based on LD patterns of the 11 FTO SNVs in each race group, statistical significance assessment was adjusted using a Bonferroni correction for 6 LD blocks, while in the White stratum, all 11 FTO variants were strongly linked (r² > 0.8). The α level for Black participants was calculated as .05 / 6 independent tests, or .00833, while the α for White participants was .05 to assess statistical significance.

Results

At enrollment, the overall mean (SD) age was 64.4 (9.7) years, there were 5276 (55.8%) women, 8743 (83.7%) Black participants, and 1704 (16.3%) White participants. See Table 1 for the mean differences in obesity indicators between race and sex groups. eTable 2 in Supplement 1 shows the frequency of the 11 FTO SNVs between race groups. There was a significant difference (>20%) in SNV frequencies for rs1558902: A, rs3751812: T, rs9941349: G, rs9930506: G, and rs9922708: T between the Black and White groups. However, all were greater than 10% MAF, so these differences would not affect statistical power substantially. Black participants had significantly higher BMI (difference, 1.6 [95% CI, 1.4-1.8]; P < .001) and WHtR (difference, 0.72 [95% CI, 0.68-0.76]; P < .001) compared with White participants.

Table 1. Comparisons of Obesity Indicators Between Sex and Self-Identified Race Groups by Analysis of Variance, Excluding Cases Analysis by Analysis.

Phenotype	Race		F	P value^a	Sex		F	P value^a
Phenotype	Black	White	F	P value^a	Female	Male	F	P value^a
BMI
No.	7382	1626	NA	NA	5276	4182	NA	NA
Mean (SD) [range]	29.50 (4.86) [18.51-39.99]	27.90 (4.54) [18.52-39.87]	149.09	<.001	29.90 (5.02) [18.51-39.99]	28.40 (4.48) [18.52-39.99]	231.56	<.001
WHtR
No.	8743	1704	NA	NA	6043	4404	NA	NA
Mean (SD) [range]	0.58 (0.09) [0.35-1.13]	0.57 (0.08) [0.36-0.97]	51.86	<.001	0.59 (0.96) [0.36-1.13]	0.56 (0.78) [0.35-0.97]	274.99	<.001

Open in a new tab

Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); NA, not applicable; WHtR, waist to height ratio.

^{^a}

Welch tests performed for groups with unequal variances based on results from Levene tests.

Females had significantly higher BMI (difference, 1.5 [95% CI, 1.3-1.7]; P < .001) and WHtR (difference, 0.028 [95% CI, 0.025-0.032]; P < .001) compared with males. Compared with White participants, Black participants were significantly younger (difference, 4.78 [95% CI, 4.25-5.30] years; P < .001), reported higher stress and depression scores, and had lower education and income levels (eTables 3 and 4 in Supplement 1). Between sex groups, women were younger (difference, 1.09 [95% CI, 0.72-1.46] years; P < .001); reported higher stress and depression scores; and had lower-income levels (eTables 3 and 4 in Supplement 1).

eTables 5 and 6 in Supplement 1 provide results for LD for White participants and Black participants, for LD values and LD blocks, respectively, formed from these associations. Due to significant LD values observed among the 11 SNVs, we selected a proxy SNV to represent others, where the LD values were greater than 0.80. Thus, 6 proxy SNVs from the LD blocks were used in the adjusted regression models for Black participants (rs1558902, rs1121980, rs17817449, rs8050136, rs9941349, and rs9930506) and 1 for White participants (rs1558902). All proxy SNVs were used for the mediation analysis for Black (6 SNVs) and White (1 SNV) participants.

Results of Regression Models

For White participants, all 11 FTO SNVs were significantly and positively associated with BMI and WHtR (eTable 7 in Supplement 1). In contrast, results of the unadjusted models show that only 2 of 6 FTO LD block SNVs were significantly and positively associated with BMI in Black participants.

In adjusted models using age, education, income, perceived stress, and depressive symptoms as covariates (Table 2), none of the LD block SNVs were associated with an obesity indicator in Black participants. All 11 FTO SNVs remained associated with BMI in White participants (β = 0.536 [95% CI, 0.197-0.875]) and WHtR (β = 0.007 [95% CI, 0.001-0.013]). Table 3 presents the adjusted models summary. Older age was significantly associated with lower BMI but significantly associated with higher WHtR in each race group. Education levels were significantly negatively associated with BMI and WHtR in each race group. In Black participants, depression was significantly positively associated with WHtR. Only in White participants were significant interactions between FTO and sex as well as FTO and age identified for BMI and WHtR (eTables 8 and 9 in Supplement 1). For example, with rs1558902: A, the coefficient for BMI was higher for men (β = 0.781; SE, 0.192; P < .001).

Table 2. Associations of BMI and WHtR With FTO SNVs, Controlling for Sex and Psychosocial Status in Self-Identified Black Participants^a.

FTO SNV LD block proxy	FTO		Sex		Age		Education		Income		Stress		Depression
FTO SNV LD block proxy	β (SE)	P value	β (SE)	P value	β (SE)	P value	β (SE)	P value	β (SE)	P value	β (SE)	P value	β (SE)	P value
BMI^b
rs1558902	0.272 (0.132)	.04	−1.807 (0.121)	<.001	−0.061 (0.007)	<.001	−0.287 (0.064)	<.001	0.199 (0.072)	.006	−0.028 (0.024)	.25	0.042 (0.03)	.16
rs1121980	0.093 (0.084)	.27	−1.805 (0.122)	<.001	−0.061 (0.007)	<.001	−0.287 (0.064)	<.001	0.199 (0.072)	.006	−0.028 (0.024)	.25	0.042 (0.03)	.16
rs17817449	0.041 (0.085)	.63	−1.807 (0.122)	<.001	−0.061 (0.007)	<.001	−0.287 (0.064)	<.001	0.199 (0.072)	.006	−0.028 (0.024)	.25	0.042 (0.03)	.17
rs8050136	0.054 (0.084)	.52	−1.807 (0.122)	<.001	−0.061 (0.007)	<.001	−0.287 (0.064)	<.001	0.199 (0.072)	.006	−0.028 (0.024)	.25	0.042 (0.03)	.16
rs9941349	0.222 (0.106)	.04	−1.806 (0.122)	<.001	−0.061 (0.007)	<.001	−0.287 (0.064)	<.001	0.199 (0.072)	.006	−0.028 (0.024)	.25	0.042 (0.03)	.16
rs9930506	0.184 (0.101)	.07	−1.807 (0.122)	<.001	−0.061 (0.007)	<.001	−0.287 (0.064)	<.001	0.199 (0.072)	.006	−0.028 (0.024)	.25	0.042 (0.03)	.16
WHtR^c
rs1558902	0.003 (0.002)	.21	−0.021 (0.002)	<.001	0.001 (<0.001)	<.001	−0.006 (0.001)	<.001	−0.002 (0.001)	.05	−0.001 (<0.001)	.23	0.002 (<0.001)	<.001
rs1121980	0.001 (0.002)	.38	−0.021 (0.002)	<.001	0.001 (<0.001)	<.001	−0.006 (0.001)	<.001	−0.002 (0.001)	.05	−0.001 (<0.001)	.23	0.002 (<0.001)	<.001
rs17817449	0.001 (0.002)	.94	−0.021 (0.002)	<.001	0.001 (<0.001)	<.001	−0.006 (0.001)	<.001	−0.002 (0.001)	.05	−0.001 (<0.001)	.23	0.002 (<0.001)	<.001
rs8050136	0.001 (0.002)	.60	−0.021 (0.002)	<.001	0.001 (<0.001)	<.001	−0.006 (0.001)	<.001	−0.002 (0.001)	.05	−0.001 (<0.001)	.23	0.002 (<0.001)	<.001
rs9941349	0.002 (0.002)	.19	−0.021 (0.002)	<.001	0.001 (<0.001)	<.001	−0.006 (0.001)	<.001	−0.002 (0.001)	.05	−0.001 (<0.001)	.24	0.002 (<0.001)	<.001
rs9930506	0.003 (0.002)	.06	−0.021 (0.002)	<.001	0.001 (<0.001)	<.001	−0.006 (0.001)	<.001	−0.002 (0.001)	.05	−0.001 (<0.001)	.23	0.002 (<0.001)	<.001

Open in a new tab

Abbreviation: BMI, body mass index; LD, linkage disequilibrium; SNV, single-nucleotide variant; WHtR, waist-to-height ratio.

^{^a}

Female was a reference group for sex, and lower education and income level was the reference group from psychosocial status. The significance level was assessed at .05.

^{^b}

General linear model: BMI = intercept + SNV (eg, LD Block 1: rs1558902: A) + sex + age + income + education + stress + depression + PC 1-10 + error.

^{^c}

General linear model: WHtR = intercept + SNV (eg, LD Block 1: rs1558902: A) + sex + age + income + education + stress + depression + PC 1-10 + error.

Table 3. Associations of BMI and WHtR With FTO SNV rs1558902, Controlling for Sex and Psychosocial Status in Self-Identified White Participants.

FTO SNV LD block proxy	FTO		Sex		Age		Education		Income		Stress		Depression
FTO SNV LD block proxy	β (SE)	P value	β (SE)	P value	β (SE)	P value	β (SE)	P value	β (SE)	P value	β (SE)	P value	β (SE)	P value
BMI^b
rs1558902	0.536 (0.173)	.002	0.525 (0.256)	.04	−0.082 (0.012)	<.001	−0.400 (0.136)	.003	−0.042 (0.149)	.78	0.007 (0.056)	.91	0.010 (0.067)	.88
WHtR^c
rs1558902	0.007 (0.003)	.02	0.013 (0.005)	.004	0.001 (<0.001)	.04	−0.009 (0.002)	<.001	−0.006 (0.003)	.03	0.001 (0.001)	.61	0.002 (0.001)	.07

Open in a new tab

Abbreviation: BMI, body mass index; LD, linkage disequilibrium; SNV, single-nucleotide variant; WHtR, waist-to-height ratio.

^{^a}

Female was a reference group for sex, and lower education and income level was the reference group from psychosocial status. The Bonferroni-correct significance level was .05.

^{^b}

General linear model: BMI = intercept + SNV (proxy LD block SNV: rs1558902) + sex + age + income + education + stress + depression + PC 1-10 + error.

^{^c}

General linear model: WHtR = intercept + SNV (proxy LD block SNV: rs1558902) + sex + age + income + education + stress + depression + PC 1-10 + error.

Results of Mediation Analysis for Cardiovascular Disease Associations by Race and Sex

Given the initial model results and commonly observed associations of BMI and WHtR with cardiometabolic disease, we explored whether these associations might differ between race and sex groups. In the mediation and moderation analyses (Tables 4 and 5; eFigure 2 in Supplement 1), only FTO LD block 2 had significant direct effects on heart disease among Black participants (c′ = 0.145 [SE, 0.0517]; P = .01), not mediated through BMI or WHtR. The direct effect did not vary by sex (P = .76) (eTable 10 in Supplement 1). There were positive (but not significant) indirect effects of FTO through obesity indicators and direct effects of FTO on other cardiometabolic diseases in Black participants (Table 4). In White participants, FTO had significant indirect effects via obesity indicators on hypertension (Table 4; eFigure 3A in Supplement 1) (eg, via WHtR: a = 0.0071 [SE, 0.0028]; P = .01; b = 7.9672 [SE, 0.9951]; P < .001) and diabetes (eTable 11 and eFigure 3B in Supplement 1) (eg, via WHtR: a, 0.0062 [SE, 0.0028]; P = .03; b, 14.60 [SE, 3.79]; P < .001), which differed by sex (eg, via WHtR on hypertension, index = −0.026; via WHtR on diabetes, index = −0.016) (eTable 11 in Supplement 1).

Table 4. Results of Mediation and Moderation Analyses of BMI, WHtR, FTO (rs1121980), and Heart Disease in Self-Identified Black Participants and of BMI, WHtR, FTO (rs1558902), and Hypertension Among Self-Identified White Partcipants^a.

Model	Consequents
	Path M (obesity indicator)		Path Y (outcome)
	Coefficient (SE)	P value	Coefficient (SE)	P value
Model BMI-heart disease among Black participants
Antecedent
X, rs1121980	a, 0.0852 (0.0841)	.31	c′, 0.1450 (0.0517)	.01
M, BMI	NA	NA	b, 0.0212 (0.0076)	.01
Constant	i_m, 38.1841 (0.5852)	<.001	i_y, −4.7185 (0.4622)	<.001
R ²	0.0529^b	<.001	0.0451	<.001
Model WHtR-heart disease among Black participants
Antecedent
X, rs1121980	a, 0.0008 (0.0015)	.59	c′, 0.1149 (0.0493)	.02
M, BMI	NA	NA	b, 1.2860 (0.3871)	.001
Constant	i_m, 0.6638 (0.0103)	<.001	i_y, −4.7370 (0.4335)	<.001
R ²	0.0485^c	<.001	0.0451	<.001
Model BMI-hypertension among White participants
Antecedent
X, proxy LD block SNV rs1558902	a, 0.5407 (0.1734)	.002	c′, −0.0276 (0.0940)	.77
M, BMI	NA	NA	b, 0.1168 (0.0157)	<.001
Constant	i_m, 32.8935 (1.5525)	<.001	i_y, −3.4320 (1.2979	.01
R ²	0.0588^d	<.001	0.1519	<.001
Model WHtR-hypertension among White participants
Antecedent
X, proxy LD block SNV rs1558902	a, 0.0071 (0.0028)	.01	c′, −0.0175 (0.0944)	.85
M, WHtR	NA	NA	b, 7.9672 (0.9951)	<.001
Constant	i_m, 0.5585 (0.0250)	<.001	i_y, −4.1055 (1.2936)	.002
R ²	0.0548^e	<.001	0.1602	<.001

Open in a new tab

Abbreviations: a, effect coefficient between genotype and obesity indicator; b, effect coefficient between mediator and outcome; BMI, body mass index, c′, direct coefficient between genotype and outcome; i_m, intercept of mediator equation; i_y, intercept of direct outcome equation; M, mediator tested; NA, not applicable; SNV, single-nucleotide variant; WHtR, waist-to-height ratio; X, genotype adjusted for ancestry principal components, age and psychosocial variables; Y, outcome.

^{^a}

Indirect effect is calculated as a × b; direct effect, c′; total effect, a × b + c′.

^{^b}

F = 21.3216.

^{^c}

F = 21.5925.

^{^d}

F = 4.9571.

^{^e}

F = 4.5871.

Table 5. Results of Moderation Analyses of BMI, WHtR, Hypertension, and Sex in Self-identified White Participants.

Subgroup	Consequent
	M (obesity indicator)		Y (hypertension)
	Coefficient (SE)	P value	Coefficient (SE) [95% CI]	Index of moderated mediation
Model BMI-hypertension
Female	0.1440 (0.0227)	<.001	0.0779 (0.0291) [0.0282-0.1422]	−0.0289
Male	0.0906 (0.0215)	<.001	0.0490 (0.0215) [0.0151-0.0037]	−0.0289
Model WHtR-hypertension
Female	9.4080 (1.2549)	<.001	0.0719 (0.0719) [0.0126-0.1398]	−0.0255
Male	6.0693 (1.3042)	<.001	0.0464 (0.0464) [0.0081-0.1022]	−0.0255

Open in a new tab

Abbreviations: BMI, body mass index; M, mediator; WHtR, waist-to-height ratio; Y, outcome.

Discussion

These results show important differences in genotype-phenotype association by sex among White participants, which have different patterns for Black participants. The often-reported FTO obesity risk alleles did not show the same strong associations for the Black group, and associations between obesity and cardiometabolic disease were different by sex for White individuals.

Many of the essential socioeconomic variables are consistent with prior investigations, which concluded that socioeconomic status (education level) was negatively associated with obesity risk for adults in developed countries.^20,21 Depression was positively associated with obesity risk in this sample of older US residents. Notably, there was significantly higher risk of abdominal obesity and elevated visceral fat for Black women. Through physiological mechanisms, the hypothalamic-pituitary-adrenal (HPA) axis, the regulation of the metabolic pathways, or inflammation, the negative psychosocial status may interrupt energy metabolism and increase risk of abdominal obesity, an immuno-metabolic trait.^45,46 Because of an age-related increase in visceral fat rather than subcutaneous or total body fat and a decrease in skeletal muscle indicating osteosarcopenic obesity,^47,48 negative psychosocial statuses combined with a considerable decline in body composition may exacerbate health risks in vulnerable populations.^49,50,51

For risks of obesity and cardiometabolic diseases, this study found that the degree of variance using FTO SNVs was different between self-identified race groups. Prior studies showed that Black individuals disproportionately experience heart disease mortality.^23,24 A recent study highlighted the importance of exploring the circulating metabolome as a potential intermediary explaining differences between Black and White individuals regarding the associations between circulating metabolites and incident coronary heart disease.⁵² Our study indicated that independent of obesity phenotypes, FTO rs1121980: A had a significant direct association with heart disease in Black participants. Similar to prior studies, the significant association between rs1121980 and obesity was identified in a small West African sample but not in the Black sample⁵³ and suggested a stronger association with obesity in White participants.⁵⁴ In a GWAS study (GOLDN study on 707 US adults, not race specific), rs1121980 appeared to have important pleiotropic effects, with a significant direct association with high-density lipoprotein levels and an indirect effect on triglyceride levels.⁵⁵ During early adipocyte differentiation and through a tissue-autonomous manner, the FTO allele associated with obesity could promote a reduction in mitochondrial thermogenesis, excessive accumulation of triglycerides, and an increase in lipid storage.⁵⁶ Our results may identify more complete gene ontology pathways to understand the genes and variants that balance blood triglycerides as an important cardiovascular disease risk factor^57,58 but showing disparities between White and Black adults.^59,60

In contrast to the finding that no FTO variants were associated with obesity risk among Black participants, the 11 FTO SNVs examined here were significantly associated with obesity phenotypes (higher in White males) and risks of hypertension and diabetes (higher in White females) in the White group. Similar to previous findings in Europeans (490 Finnish participants), the AA genotype of rs9939609 was associated with an adjusted 2-fold increased risk (hazard ratio, 2.09; 95% CI, 1.17-3.73) of cardiovascular disease (morbidity and mortality) among male participants,⁶¹ and a higher incidence of coronary heart disease and cardiovascular disease events or death was associated with rs9939609 (AA genotype, OR = 1.90; P = .002 and P = .004, respectively) among 1200 Finnish people at baseline.⁶² Further, a meta-analysis of 1497 Black and West African adults concluded that FTO, along with IRS1, KLF14, and PPARG, played active roles via insulin resistance in the development of type 2 diabetes and other metabolic disorders.⁶³ However, previous genetic risk score modeling (including rs9939609 as 1 of 46 gene variants) of 990 Black and 1644 White individuals aged 44 to 69 years found that cumulative allele load was associated with the risk of type 2 diabetes in White individuals, but only marginally in Black individuals.⁶⁴

Limitations

This study has limitations. The present analysis shows substantial differences between sex and self-identified race groups regarding obesity and cardiometabolic disease risk in this large national sample of 8743 Black and 1704 White individuals. However, we used only SNVs from 1 obesity risk gene (FTO) rather than some of the proposed genetic risk scores. For obesity and associated cardiometabolic diseases, the FTO SNVs may explain more of the variance in White individuals, or there may be other yet-to-be-identified risk or protective factors that were not included here. Participants in the REGARDS study were only allowed to self-identify as Black or White. These self-identifications do not preclude admixture in an individual’s ancestry. Additionally, we did not control for any potential familial relationships between participants, but these were screened out on enrollment in REGARDS.

Conclusions

In this cross-sectional study of a large sample of US adults of self-identified African and European race groups, we observed that the often-reported FTO obesity risk alleles were not associated with obesity for the Black group while controlling for race and psychosocial variables and that association between obesity and cardiometabolic disease varied by sex for White participants with rs1558902. More studies should focus on discovering risk alleles or risk scores unique to Black individuals and exploring the influence of genotypes, including sex chromosome complement, to inform mechanisms in the prevention and management of obesity and cardiometabolic diseases. These data support future work refining genetic risk scores and personalized medicine for all US residents.

Supplement 1.

eTable 1. Selected Obesity-Related Human Studies That Investigated 11 FTO SNVs for Further Analyses

eAppendix. Genotyping and Quality Control Details for REGARDS Participants

eFigure 1. Study Participant Selection

eTable 2. Comparisons of Effect Allele Frequencies of FTO SNVs Between Self-Identified Race Groups

eTable 3. Comparisons of Baseline Age, Income, Education Levels, Depression Score, and Stress Score Between Sex and Self-Identified Race Groups for BMI Analysis

eTable 4. Comparisons of Baseline Age, Income, Education Levels, Depression Score, and Stress Score Between Sex and Self-identified Race Groups for Waist-to-Height Ratio Analysis

eTable 5. Results of Linkage Disequilibrium Among the 11 FTO SNVs in Self-Identified White Participants

eTable 6. Results of Linkage Disequilibrium Among the 11 FTO SNVs in Self-Identified Black Participants

eTable 7. Associations Between BMI and WHtR by FTO Proxy SNVs, Controlling for Sex (Females as the Reference Group) in Self-Identified Black Participants or White Participants

eTable 8. Interaction Terms Between FTO SNVs, Sex, and Age in Black Participants, Controlling for Psychosocial Status (Lower Education Levels and Lower Income Levels as the Reference Group)

eTable 9. Significant Interaction Terms Between FTO SNV rs1558902:A, Sex, and Age in White Participants, Controlling for Psychosocial Status (Lower Education Levels and Lower Income Levels as the Reference Group)

eFigure 2. Mediation Model of Obesity Indicators (BMI) in Relation With 6 FTO LD Block Proxy SNVs and Heart Disease in Black Participants

eTable 10. Interaction Terms Between Genetic Susceptibility by FTO SNV rs1121980: A and Sex in Black Participants

eFigure 3. Mediation Model of Obesity Indicators (BMI and WHtR) in Relation With FTO SNVs, and Hypertension/Diabetes in White Participants

eTable 11. Results of Mediation and Moderation analyses of BMI/WHtR, LD Block 1 (rs1558902:A), and Diabetes (No Diabetes as the Reference Group) in White Participants

Click here for additional data file.^{(831.4KB, pdf)}

Supplement 2.

Data Sharing Statement

Click here for additional data file.^{(239.5KB, pdf)}

References

1.Loos RJ, Yeo GS. The bigger picture of FTO: the first GWAS-identified obesity gene. Nat Rev Endocrinol. 2014;10(1):51-61. doi: 10.1038/nrendo.2013.227 [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Frayling TM, Timpson NJ, Weedon MN, et al. A common variant in the FTO gene is associated with body mass index and predisposes to childhood and adult obesity. Science. 2007;316(5826):889-894. doi: 10.1126/science.1141634 [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Scuteri A, Sanna S, Chen WM, et al. Genome-wide association scan shows genetic variants in the FTO gene are associated with obesity-related traits. PLoS Genet. 2007;3(7):e115. doi: 10.1371/journal.pgen.0030115 [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Fawcett KA, Barroso I. The genetics of obesity: FTO leads the way. Trends Genet. 2010;26(6):266-274. doi: 10.1016/j.tig.2010.02.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Lan N, Lu Y, Zhang Y, et al. FTO—a common genetic basis for obesity and cancer. Front Genet. 2020;11:559138-559138. doi: 10.3389/fgene.2020.559138 [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Speakman JR. The ‘fat mass and obesity related’ (FTO) gene: mechanisms of impact on obesity and energy balance. Curr Obes Rep. 2015;4(1):73-91. doi: 10.1007/s13679-015-0143-1 [DOI] [PubMed] [Google Scholar]
7.Rohde K, Keller M, la Cour Poulsen L, Blüher M, Kovacs P, Böttcher Y. Genetics and epigenetics in obesity. Metabolism. 2019;92:37-50. doi: 10.1016/j.metabol.2018.10.007 [DOI] [PubMed] [Google Scholar]
8.Link JC, Reue K. Genetic basis for sex differences in obesity and lipid metabolism. Annu Rev Nutr. 2017;37:225-245. doi: 10.1146/annurev-nutr-071816-064827 [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Pulit SL, Stoneman C, Morris AP, et al. ; GIANT Consortium . Meta-analysis of genome-wide association studies for body fat distribution in 694 649 individuals of European ancestry. Hum Mol Genet. 2019;28(1):166-174. doi: 10.1093/hmg/ddy327 [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Heid IM, Jackson AU, Randall JC, et al. ; MAGIC . Meta-analysis identifies 13 new loci associated with waist-hip ratio and reveals sexual dimorphism in the genetic basis of fat distribution. Nat Genet. 2010;42(11):949-960. doi: 10.1038/ng.685 [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Cameron NA, Petito LC, McCabe M, et al. Quantifying the sex-race/ethnicity-specific burden of obesity on incident diabetes mellitus in the United States, 2001 to 2016: MESA and NHANES. J Am Heart Assoc. 2021;10(4):e018799. doi: 10.1161/JAHA.120.018799 [DOI] [PMC free article] [PubMed] [Google Scholar]
12.National Center for Health Statistics . Health, United States—data finder; table 026. Accessed November 8, 2023. https://www.cdc.gov/nchs/hus/data-finder.htm?&subject=Obesity%20and%20overweight
13.Wang X, Magkos F, Mittendorfer B. Sex differences in lipid and lipoprotein metabolism: it’s not just about sex hormones. J Clin Endocrinol Metab. 2011;96(4):885-893. doi: 10.1210/jc.2010-2061 [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Tramunt B, Smati S, Grandgeorge N, et al. Sex differences in metabolic regulation and diabetes susceptibility. Diabetologia. 2020;63(3):453-461. doi: 10.1007/s00125-019-05040-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Rankinen T, Zuberi A, Chagnon YC, et al. The human obesity gene map: the 2005 update. Obesity (Silver Spring). 2006;14(4):529-644. doi: 10.1038/oby.2006.71 [DOI] [PubMed] [Google Scholar]
16.Bouchard C. Genetics of obesity: what we have learned over decades of research. Obesity (Silver Spring). 2021;29(5):802-820. doi: 10.1002/oby.23116 [DOI] [PubMed] [Google Scholar]
17.Walter S, Mejía-Guevara I, Estrada K, Liu SY, Glymour MM. Association of a genetic risk score with body mass index across different birth cohorts. JAMA. 2016;316(1):63-69. doi: 10.1001/jama.2016.8729 [DOI] [PubMed] [Google Scholar]
18.Zillikens MC, Demissie S, Hsu YH, et al. Large meta-analysis of genome-wide association studies identifies five loci for lean body mass. Nat Commun. 2017;8(1):80. doi: 10.1038/s41467-017-00031-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Sobal J, Stunkard AJ. Socioeconomic status and obesity: a review of the literature. Psychol Bull. 1989;105(2):260-275. doi: 10.1037/0033-2909.105.2.260 [DOI] [PubMed] [Google Scholar]
20.Cohen AK, Rai M, Rehkopf DH, Abrams B. Educational attainment and obesity: a systematic review. Obes Rev. 2013;14(12):989-1005. doi: 10.1111/obr.12062 [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Kaiser KA, Smith DL Jr, Allison DB. Conjectures on some curious connections among social status, calorie restriction, hunger, fatness, and longevity. Ann N Y Acad Sci. 2012;1264(1):1-12. doi: 10.1111/j.1749-6632.2012.06672.x [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Frank M, Dragano N, Arendt M, et al. A genetic sum score of risk alleles associated with body mass index interacts with socioeconomic position in the Heinz Nixdorf Recall Study. PLoS One. 2019;14(8):e0221252. doi: 10.1371/journal.pone.0221252 [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Safford MM, Brown TM, Muntner PM, et al. ; REGARDS Investigators . Association of race and sex with risk of incident acute coronary heart disease events. JAMA. 2012;308(17):1768-1774. doi: 10.1001/jama.2012.14306 [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Colantonio LD, Gamboa CM, Richman JS, et al. Black-White differences in incident fatal, nonfatal, and total coronary heart disease. Circulation. 2017;136(2):152-166. doi: 10.1161/CIRCULATIONAHA.116.025848 [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Safford MM, Reshetnyak E, Sterling MR, et al. Number of social determinants of health and fatal and nonfatal incident coronary heart disease in the REGARDS study. Circulation. 2021;143(3):244-253. doi: 10.1161/CIRCULATIONAHA.120.048026 [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Gutiérrez OM, Khodneva YA, Muntner P, et al. ; REGARDS Investigators . Association between urinary albumin excretion and coronary heart disease in Black vs White adults. JAMA. 2013;310(7):706-714. doi: 10.1001/jama.2013.8777 [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Jannat-Khah DP, Khodneva Y, Bryant K, et al. Depressive symptoms do not discriminate: racial and economic influences between time-varying depressive symptoms and mortality among REGARDS participants. Ann Epidemiol. 2020;46:31-40.e2. doi: 10.1016/j.annepidem.2020.04.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Gowey MA, Khodneva Y, Tison SE, et al. Depressive symptoms, perceived stress, and metabolic health: The REGARDS study. Int J Obes (Lond). 2019;43(3):615-632. doi: 10.1038/s41366-018-0270-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Howard VJ, Cushman M, Pulley L, et al. The reasons for geographic and racial differences in stroke study: objectives and design. Neuroepidemiology. 2005;25(3):135-143. doi: 10.1159/000086678 [DOI] [PubMed] [Google Scholar]
30.Howard VJ, Kleindorfer DO, Judd SE, et al. Disparities in stroke incidence contributing to disparities in stroke mortality. Ann Neurol. 2011;69(4):619-627. doi: 10.1002/ana.22385 [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Little J, Higgins JPT, Ioannidis JPA, et al. STrengthening the REporting of Genetic Association Studies (STREGA)–an extension of the STROBE statement. Genet Epidemiol. 2009;33(7):581-598. doi: 10.1002/gepi.20410 [DOI] [PubMed] [Google Scholar]
32.Azzalini A. The multivariate skew-normal distribution. Biometrika. 1996;83(4):715-726. doi: 10.1093/biomet/83.4.715 [DOI] [Google Scholar]
33.Machiela MJ, Chanock SJ. LDlink: a web-based application for exploring population-specific haplotype structure and linking correlated alleles of possible functional variants. Bioinformatics. 2015;31(21):3555-3557. doi: 10.1093/bioinformatics/btv402 [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Browning LM, Hsieh SD, Ashwell M. A systematic review of waist-to-height ratio as a screening tool for the prediction of cardiovascular disease and diabetes: 0·5 could be a suitable global boundary value. Nutr Res Rev. 2010;23(2):247-269. doi: 10.1017/S0954422410000144 [DOI] [PubMed] [Google Scholar]
35.Kazlauskaite R, Avery-Mamer EF, Li H, et al. Race/ethnic comparisons of waist-to-height ratio for cardiometabolic screening: The study of women’s health across the nation. Am J Hum Biol. 2017;29(1). doi: 10.1002/ajhb.22909 [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Cohen S. Perceived stress in a probability sample of the United States. The social psychology of health. In: Spacapan S, Oskamp S, eds. The Social Psychology of Health. Sage Publications, Inc. 1988:31-67. [Google Scholar]
37.Radloff LS. The CES-D Scale. Appl Psychol Meas. 2016;1(3):385-401. doi: 10.1177/014662167700100306 [DOI] [Google Scholar]
38.Gao X, Starmer J, Martin ER. A multiple testing correction method for genetic association studies using correlated single nucleotide polymorphisms. Genet Epidemiol. 2008;32(4):361-369. doi: 10.1002/gepi.20310 [DOI] [PubMed] [Google Scholar]
39.Preacher KJ, Rucker DD, Hayes AF. Addressing moderated mediation hypotheses: theory, methods, and prescriptions. Multivariate Behav Res. 2007;42(1):185-227. doi: 10.1080/00273170701341316 [DOI] [PubMed] [Google Scholar]
40.Hsu CY, Yang W, Parikh RV, et al. ; CRIC Study Investigators . Race, genetic ancestry, and estimating kidney function in CKD. N Engl J Med. 2021;385(19):1750-1760. doi: 10.1056/NEJMoa2103753 [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Hauser MA, Allingham RR, Aung T, et al. ; Genetics of Glaucoma in People of African Descent (GGLAD) Consortium . Association of genetic variants with primary open-angle glaucoma among individuals with African Ancestry. JAMA. 2019;322(17):1682-1691. doi: 10.1001/jama.2019.16161 [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Parcha V, Heindl B, Kalra R, et al. Genetic european ancestry and incident diabetes in Black individuals: insights from the SPRINT Trial. Circ Genom Precis Med. 2022;15(1):e003468. doi: 10.1161/CIRCGEN.121.003468 [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Price AL, Patterson NJ, Plenge RM, Weinblatt ME, Shadick NA, Reich D. Principal components analysis corrects for stratification in genome-wide association studies. Nat Genet. 2006;38(8):904-909. doi: 10.1038/ng1847 [DOI] [PubMed] [Google Scholar]
44.Tian C, Gregersen PK, Seldin MF. Accounting for ancestry: population substructure and genome-wide association studies. Hum Mol Genet. 2008;17(R2):R143-R150. doi: 10.1093/hmg/ddn268 [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Everson-Rose SA, Lewis TT, Karavolos K, Dugan SA, Wesley D, Powell LH. Depressive symptoms and increased visceral fat in middle-aged women. Psychosom Med. 2009;71(4):410-416. doi: 10.1097/PSY.0b013e3181a20c9c [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Alshehri T, Boone S, de Mutsert R, et al. The association between overall and abdominal adiposity and depressive mood: a cross-sectional analysis in 6459 participants. Psychoneuroendocrinology. 2019;110:104429. doi: 10.1016/j.psyneuen.2019.104429 [DOI] [PubMed] [Google Scholar]
47.Beaufrère B, Morio B. Fat and protein redistribution with aging: metabolic considerations. Eur J Clin Nutr. 2000;54(suppl 3):S48-S53. doi: 10.1038/sj.ejcn.1601025 [DOI] [PubMed] [Google Scholar]
48.JafariNasabian P, Inglis JE, Reilly W, Kelly OJ, Ilich JZ. Aging human body: changes in bone, muscle and body fat with consequent changes in nutrient intake. J Endocrinol. 2017;234(1):R37-R51. doi: 10.1530/JOE-16-0603 [DOI] [PubMed] [Google Scholar]
49.Bray GA, Heisel WE, Afshin A, et al. The science of obesity management: an endocrine society scientific statement. Endocr Rev. 2018;39(2):79-132. doi: 10.1210/er.2017-00253 [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Chen X, Anderson CS. Revisiting the adiposity paradox for the risk of stroke subtypes. JAMA Netw Open. 2022;5(12):e2246621. doi: 10.1001/jamanetworkopen.2022.46621 [DOI] [PubMed] [Google Scholar]
51.Cardel MI, Guo Y, Sims M, et al. Objective and subjective socioeconomic status associated with metabolic syndrome severity among African American adults in Jackson Heart Study. Psychoneuroendocrinology. 2020;117:104686. doi: 10.1016/j.psyneuen.2020.104686 [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Cruz DE, Tahir UA, Hu J, et al. Metabolomic analysis of coronary heart disease in an African American cohort from the Jackson Heart Study. JAMA Cardiol. 2022;7(2):184-194. doi: 10.1001/jamacardio.2021.4925 [DOI] [PMC free article] [PubMed] [Google Scholar]
53.Adeyemo A, Chen G, Zhou J, et al. FTO genetic variation and association with obesity in West Africans and African Americans. Diabetes. 2010;59(6):1549-1554. doi: 10.2337/db09-1252 [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Fesinmeyer MD, North KE, Ritchie MD, et al. Genetic risk factors for BMI and obesity in an ethnically diverse population: results from the Population Architecture Using Genomics and Epidemiology (PAGE) study. Obesity (Silver Spring). 2013;21(4):835-846. doi: 10.1002/oby.20268 [DOI] [PMC free article] [PubMed] [Google Scholar]
55.Justice AE, Howard AG, Fernández-Rhodes L, Graff M, Tao R, North KE. Direct and indirect genetic effects on triglycerides through omics and correlated phenotypes. BMC Proc. 2018;12(suppl 9):22. doi: 10.1186/s12919-018-0118-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
56.Claussnitzer M, Dankel SN, Kim KH, et al. FTO obesity variant circuitry and adipocyte browning in humans. N Engl J Med. 2015;373(10):895-907. doi: 10.1056/NEJMoa1502214 [DOI] [PMC free article] [PubMed] [Google Scholar]
57.Johansen CT, Kathiresan S, Hegele RA. Genetic determinants of plasma triglycerides. J Lipid Res. 2011;52(2):189-206. doi: 10.1194/jlr.R009720 [DOI] [PMC free article] [PubMed] [Google Scholar]
58.Laufs U, Parhofer KG, Ginsberg HN, Hegele RA. Clinical review on triglycerides. Eur Heart J. 2020;41(1):99-109c. doi: 10.1093/eurheartj/ehz785 [DOI] [PMC free article] [PubMed] [Google Scholar]
59.Chung ST, Cravalho CKL, Meyers AG, et al. Triglyceride paradox is related to lipoprotein size, visceral adiposity and stearoyl-coa desaturase activity in Black versus White women. Circ Res. 2020;126(1):94-108. doi: 10.1161/CIRCRESAHA.119.315701 [DOI] [PMC free article] [PubMed] [Google Scholar]
60.Yu SS, Castillo DC, Courville AB, Sumner AE. The triglyceride paradox in people of African descent. Metab Syndr Relat Disord. 2012;10(2):77-82. doi: 10.1089/met.2011.0108 [DOI] [PMC free article] [PubMed] [Google Scholar]
61.Lappalainen T, Kolehmainen M, Schwab US, et al. ; Finnish Diabetes Prevention Study Group . Association of the FTO gene variant (rs9939609) with cardiovascular disease in men with abnormal glucose metabolism—the Finnish Diabetes Prevention Study. Nutr Metab Cardiovasc Dis. 2011;21(9):691-698. doi: 10.1016/j.numecd.2010.01.006 [DOI] [PubMed] [Google Scholar]
62.Äijälä M, Ronkainen J, Huusko T, et al. The fat mass and obesity-associated (FTO) gene variant rs9939609 predicts long-term incidence of cardiovascular disease and related death independent of the traditional risk factors. Ann Med. 2015;47(8):655-663. doi: 10.3109/07853890.2015.1091088 [DOI] [PubMed] [Google Scholar]
63.Chen G, Bentley A, Adeyemo A, et al. Genome-wide association study identifies novel loci association with fasting insulin and insulin resistance in African Americans. Hum Mol Genet. 2012;21(20):4530-4536. doi: 10.1093/hmg/dds282 [DOI] [PMC free article] [PubMed] [Google Scholar]
64.Keaton JM, Cooke Bailey JN, Palmer ND, et al. A comparison of type 2 diabetes risk allele load between African Americans and European Americans. Hum Genet. 2014;133(12):1487-1495. doi: 10.1007/s00439-014-1486-5 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials