Genetic underpinnings of regional adiposity distribution in African Americans: Assessments from the Jackson Heart Study

Mohammad Y. Anwar; Laura M. Raffield; Leslie A. Lange; Adolfo Correa; Kira C. Taylor

doi:10.1371/journal.pone.0255609

Abstract

Background

African ancestry individuals with comparable overall anthropometric measures to Europeans have lower abdominal adiposity. To explore the genetic underpinning of different adiposity patterns, we investigated whether genetic risk scores for well-studied adiposity phenotypes like body mass index (BMI) and waist circumference (WC) also predict other, less commonly measured adiposity measures in 2420 African American individuals from the Jackson Heart Study.

Methods

Polygenic risk scores (PRS) were calculated using GWAS-significant variants extracted from published studies mostly representing European ancestry populations for BMI, waist-hip ratio (WHR) adjusted for BMI (WHR_BMIadj), waist circumference adjusted for BMI (WC_BMIadj), and body fat percentage (BF%). Associations between each PRS and adiposity measures including BF%, subcutaneous adiposity tissue (SAT), visceral adiposity tissue (VAT) and VAT:SAT ratio (VSR) were examined using multivariable linear regression, with or without BMI adjustment.

Results

In non-BMI adjusted models, all phenotype-PRS were found to be positive predictors of BF%, SAT and VAT. WHR-PRS was a positive predictor of VSR, but BF% and BMI-PRS were negative predictors of VSR. After adjusting for BMI, WHR-PRS remained a positive predictor of BF%, VAT and VSR but not SAT. WC-PRS was a positive predictor of SAT and VAT; BF%-PRS was a positive predictor of BF% and SAT only.

Conclusion

These analyses suggest that genetically driven increases in BF% strongly associate with subcutaneous rather than visceral adiposity and BF% is strongly associated with BMI but not central adiposity-associated genetic variants. How common genetic variants may contribute to observed differences in adiposity patterns between African and European ancestry individuals requires further study.

Citation: Anwar MY, Raffield LM, Lange LA, Correa A, Taylor KC (2021) Genetic underpinnings of regional adiposity distribution in African Americans: Assessments from the Jackson Heart Study. PLoS ONE 16(8): e0255609. https://doi.org/10.1371/journal.pone.0255609

Editor: Heming Wang, Brigham and Women’s Hospital and Harvard Medical School, UNITED STATES

Received: February 1, 2021; Accepted: July 19, 2021; Published: August 4, 2021

Copyright: © 2021 Anwar et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: The data underlying the findings include potentially identifying participant information and cannot be made publicly available due to ethical/legal restrictions. However, data, including statistical code, from this manuscript are available to researchers who meet the criteria for access to confidential data. Access to JHS data is available through (https://www.jacksonheartstudy.org/Research/Study-Data/Data-Access). All JHS genetic data is available through dbGaP at phs000286.v6.p2. Additionally, much of the JHS data is available through BioLINCC (https://biolincc.nhlbi.nih.gov/studies/jhs/).

Funding: Acknowledgements The Jackson Heart Study (JHS) is supported and conducted in collaboration with Jackson State University (HHSN268201800013I), Tougaloo College (HHSN268201800014I), the Mississippi State Department of Health (HHSN268201800015I) and the University of Mississippi Medical Center (HHSN268201800010I, HHSN268201800011I and HHSN268201800012I) contracts from the National Heart, Lung, and Blood Institute (NHLBI) and the National Institute on Minority Health and Health Disparities (NIMHD). The authors also wish to thank the staffs and participants of the JHS. LMR was supported by T32 HL129982. The funders for LMR had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Despite the wide adoption of body mass index (BMI), a measure that correlates well with numerous health risk factors [1], there are limitations to this metric [2]; notably, it does not differentiate between variation in fat and lean mass. This can lead to an imprecise categorization of obesity [3], and hence misleading inferences for cardiometabolic outcomes [4]. This is a major drawback given that obesity-health outcome associations are differentially mediated by bodily composition, with adipose tissues more prominently linked with adverse outcomes [5]. Total adiposity [6], regional fat distribution [7], and especially visceral adiposity [8] are significant risk factors for cardiovascular diseases.

Since body fat is linearly associated with BMI in sedentary populations [9], a proportion of genetic variants associated with overall body mass expectedly overlap with loci linked to body fat percentage [10]. However, this BMI-body fat mass link is known to exhibit phenotypic variability across ethnicities [11], which extends to regional distribution of fat tissue as well: in African ancestry (AA) individuals with comparable BMI metrics to those with European ancestry (EA), the proportion of visceral adiposity is lower [12, 13]. Paradoxically, the prevalence of cardiometabolic diseases in AA, compared to EA with similar BMI, is higher [14]; genetics may also play a significant role in this differential adiposity distribution and merits a thorough examination.

Despite some evidence for AA-specific variants associated with BMI [15], most reported BMI-associated SNPs in AAs are variants first reported in EA individuals [16], and European-derived genetic risk scores have been found to be predictive of BMI variation in AA populations as well [17]. But evidence for generalization of variants associated with other adiposity traits to AAs, including measures of central obesity, is limited [18], and no study has examined replicability of body fat percentage-associated variants identified in EA populations to AA. Since genomic loci associated with body fat percentage are suggested to be more closely aligned with multiple cardiometabolic disease risks than BMI-associated variants [19], assessing if genetic variants associated with adiposity patterns in EA—particularly body fat percentage (BF%)—can be extrapolated to AA individuals is an important question.

The objective of this this study was to assess the utility of known variants associated with anthropometric and adiposity measures discovered in predominantly EA populations to predict BF%, SAT, VAT, and VSR among AA individuals from Jackson Heart Study (JHS). We also estimated the association of variants with evidence of directional replicability and nominal significance in JHS for their originally reported adiposity measure to other adiposity measures, allowing us to examine the relationships between adiposity measures.

Methods

Study population

The JHS recruited 5306 African American residents living in the Jackson, Mississippi, metropolitan area of Hinds, Madison, and Rankin Counties. Participants were recruited to participate in the study from four pools: random sampling (17% of participants), volunteers (30%), participants in the Atherosclerosis Risk in Communities (ARIC) study (31%), and secondary family members (22%). The age at enrollment for the unrelated cohort was 35–84 years; the family cohort included related individuals >21 years old. Extensive phenotypic data was collected during a baseline examination (September 2000 –March 2004), and two follow-up examinations (October 2005 –December 2008, and February 2009 –January 2013). A third follow-up examination is in progress. Annual follow-up interviews and cohort surveillance for cardiovascular events and mortality are also ongoing.

For BMI and WC genome-wide association tests, we used phenotypic observations from visit 1 to maximize sample size (N = 3020); for other traits, we used phenotypic measures from visit 2 for the same purpose (N = 2554). However, for polygenic risk score regression analyses, we exclusively used phenotypic observations from visit 2. In our analyses, 10 participants were excluded due to pregnancy during the visit 2 examination, and 124 were excluded for missing or biologically implausible values. The total sample size used was n = 2420 individuals.

The study protocol was approved by the participating JHS institutions including the Tougaloo College, the Jackson State University, and the University of Mississippi Medical Center IRB, under IRB number- UMMC IRB File#1998–6004. All JHS participants gave written informed consent. The IRB registration for UMMC is #00000061, DHHS FWA #00003630.

Genotyping

Genotyping for single nucleotide polymophisms (SNPs) was performed with the Affymetrix 6.0 SNP Array (Affymetrix, Santa Clara, Calif). Outliers based on principal components, sample swaps, duplicates, and one of each pair of monozygotic twins were excluded. Samples with a mismatch between pedigree vs genetic sex were also removed. Variants with a minor allele frequency ≥ 1%, a call rate ≥ 90%, and a Hardy Weinberg equilibrium (HWE) p-value >10–6 (n = 832,508 variants) were used for imputation to 1000 Genomes Project population SNP reference panel (Phase 3, Version 5), using Minimac3 on the Michigan Imputation Server. Only SNPs with an imputation quality r²>0.9 were selected for polygenic risk score analyses.

Polygenic risk scores

The use of polygenic risk scores (PRS) to predict complex traits [20], closely related phenotypes [21], and the same phenotype across different populations [22] has been validated. Both weighted and unweighted methods have been employed for estimation of PRS [23]. Weighted risk scores account for each variant’s effect size on the phenotype. Unweighted PRS assume that all variants have equal effect on the trait; this assumption is almost always violated, as in genome-wide association studies (GWAS), some variants have much larger effect sizes than others. In contrast, use of a weighted PRS method in a non-EA population may also introduce bias because nearly all known variants were identified in predominantly EA populations, and studies suggest dissimilar effect sizes for the same SNPs across ancestries, likely due to differential LD with true causal variants [24]. Although the weighted method can lead to reduced mean square error for prediction in some cases [25], the main applications for polygenic scores, namely association testing and prediction, do not appear to differ substantially between two methods. In addition, an unweighted score is more robust against error in estimating the effect sizes arising from limited samples, “winner’s curse bias” [26], and confounding by demographic structure [27]. Therefore, we used an unweighted PRS approach in the AA population studied here.

At each locus (SNP), participants were assigned a dosage value between 0 and 2 inclusive, based on the estimated number and frequency of phenotype-increasing alleles under an additive genetic effect model. The PRS value for each individual reflects the summation of risk alleles across all selected loci.

SNP risk sets

To construct SNP sets used for PRS calculation, we utilized both the European Bioinformatic Institute GWAS repository (ebi.ac.uk/gwas, accessed December 2020), and PubMed for extraction of SNPs linked to anthropometric measures including body mass index (BMI), waist to hip ratio adjusted for BMI (WHR_BMIadj), waist circumference adjusted for BMI (WC_BMIadj), and body fat percentage (BF%). Only SNPs reported at GWAS significance level (p≤5.00 ×10⁻⁸) were selected. Most variants were discovered in EA-only studies, but some are from large multi-ethnic GWAS meta-analyses or AA specific analyses. PRS derived from only AA-specific variants for target phenotypes were judged to be underpowered due to the low number of available variants, as well as PRS calculated from SNPs associated with VAT, SAT and VSR.

We conducted linkage disequilibrium (LD) analysis using LDlink (ldlink.nci.nih.gov) with multi-ethnic populations. If a pair or a group of SNPs were in LD with one another (R²≥0.1), we prioritized sentinel variants from larger and more recent studies with lower p-values and selected a single SNP to ensure independence of variants and avoid double counting the same functional locus (site). This set of LD-pruned variants was then used to calculate the PRS.

Anthropometry measures

WC was measured at the umbilicus level using a non-elastic tape measurer and rounded to the nearest centimeter; hip circumference (HC) was measured at the level of the widest circumference over the greater trochanter. WHR was obtained by dividing WC over HC.

Adiposity measures

A variety of different techniques for assessment of body composition exist [28]. For overall BF%, bioimpedance is a widely adopted method [29]. Under this technique, BF% is calculated based on the measured resistance of the adipose tissue as the person lays supine with electrodes placed on the arm and/or leg; bare foot-to-foot bioimpedance was conducted using the Tanita Body Fat Monitor (Tanita Corp, Tokyo). BF% was estimated using a programmed algorithm that incorporates bioimpedance readings with a height, weight, age and sex-specific equation and additional adjustment for physical activity levels.

To estimate visceral and subcutaneous adipose volumes, the study employed computed tomography (CT) technique at visit 2, where the heart and lower abdomen regions were scanned with 16-channel mutidetector CT machine (Lightspeed 16 Pro, GE Healthcare, Milwaukee, WI). Abdominal imaging slices covering the lower abdomen from L3 to S1 were used to quantify both VAT and SAT [29], such that 24 adjacent 2-mm thick slices centered on the lumbar disk space at L4 to L5 were used for quantification of both types of adiposity; 12 images before the center of L4 to L5 disk space and 12 images after that space [30].

Statistical analysis

To facilitate comparison across different phenotypes, we performed inverse-normal transformations prior to analysis. Genome-wide association analyses were completed for target traits. We used EPACTS 3.2.6 [31] to perform GWAS analyses, adjusting for age, sex, and a genetic relationship matrix using the EMMAX test; additionally, BMI was incorporated as covariate in WHR and WC analyses.

We calculated PRS under three separate but complementary scenarios using the following configurations: (a) set of all known loci (LD-pruned) reported at genome-wide significance (5×10⁻⁸) in multi-ethnic or European studies regardless of replicability in JHS (principle approach) (S1 Table); (b) the subset of risk loci with evidence of directional replication in the JHS-GWAS results (approach 2); and (c) a more restricted subset of risk loci with evidence for both directional replication as well as nominal statistical significance (p<5×10⁻²) in JHS results (approach 3, S2 Table). PRS for BMI, WHR_BMIadj,WC_BMIadj, and BF% were first tested against their respective phenotypes in JHS to ensure the validity of constructed predictors (S3 Table).

Phenotype-specific PRS obtained under various approaches were then tested for cross-sectional associations with phenotypic measures. Results for PRS obtained under the principle approach (e.g. all known loci) were reported in the manuscript, whereas models that were obtained with PRS under complementary approaches were provided as supplementary material. Both multivariable linear regression and mixed models [32] were employed to investigate the associations between PRS and adiposity measures, with age, sex, and the top 10 ancestry principal components as covariates; additionally, family ID was utilized as random component in mixed models. Both offered similar results; linear results were chosen for simplicity.

Some loci associated with obesity traits and/or fat distributions [33] are known to be sex-specific. Although we performed gender-stratified analysis, given the sex imbalance in this sample (36.9% males), results for the females largely mirrored the combined findings, while the male-only analysis lacked precision; therefore we chose to report the sex-combined analysis.

Finally, to characterize the association between EA-established variants for anthropometric traits with evidence of transferrability to AA (i.e. statistically significant in JHS-GWAS), and type of adiposity (BF%, SAT, VAT, VSR), we constructed heatmap plots of SNPs’ effect sizes (betas*100) to investigate if genetic variations underpinning obesity traits are closely correlated with overall body fat change or aligned to specific adiposity patterns. For phenotype and SNP clustering, Ward’s minimum variance method was used which aims at finding compact, spherical clusters [34].

Statistical analyses including PRS generation and regression models were performed using RStudio (V 1.1.463), and heatmap plots were obtained using the R package “pheatmap” [35].

Results

Population characteristics and adiposity measures

Table 1 provides a descriptive distribution of demographic, anthropometric and adiposity traits in the JHS population. Using the standard BMI cutpoints of ≥25 and ≥30 for overweight and obesity respectively, the majority of participants were either overweight (31.5%) or obese (54.2%). The mean WC and BF% also indicate a high prevalence of excess adiposity.

Download:

Table 1. Baseline characteristics of the study participants.

https://doi.org/10.1371/journal.pone.0255609.t001

Spearman correlation coefficients for anthropometric and adiposity measures (S4 Table) illustrated a high degree of correlation of BMI with WC (r = 0.82), but much weaker correlation with WHR (r = 0.16). BMI was also highly correlated with SAT (r = 0.83) and BF% (r = 0.70), but not as strongly with VAT (r = 0.49).

Association with polygenicc risk scores

In non-BMI adjusted models, all phenotype-specific PRS were found to be positive predictors of BF%; β = 2.4 (p = 2.1 ×10⁻¹⁰) for BF%-PRS; β = 1.2 (p = 3.3 ×10⁻⁵¹) for BMI-PRS; β = 1.1 (p = 2.2 ×10⁻¹⁹) for WC-PRS; and β = 0.7 (p = 1.4×10⁻⁸) for WHR-PRS respectively, where β represents % change in phenotype z-score per increase of 1 trait-increasing allele (Table 2). For the SAT phenotype, like BF%, all phenotype-specific PRS were positive predictors: β = 1.9 (p = 1.9 ×10⁻⁴) for BF%-PRS; β = 1.4 (p = 1.4 ×10⁻³⁷) for BMI-PRS; β = 1.3 (p = 9.8 ×10⁻¹⁵) for WC-PRS; and β = 0.6 (p = 5.8×10⁻⁵) for WHR-PRS respectively. For the VAT phenotype, BMI-PRS and WC-PRS were positively associated with similarly close coefficients (e.g. slope) as the SAT phenotype (Table 2), but for WHR-PRS, the coefficient was notably larger (β = 1.2 (p = 3.5×10⁻⁹)); BF%-PRS was positive but not strong predictor of VAT (β = 0.8 (p = 1.4×10⁻¹)). For the VSR, BF%-PRS was a significant negative predictor (β = -0.9 (p = 5.2×10⁻²)) and WHR-PRS was a positive predictor (β = 0.4 (p = 3.8×10⁻³)).

Download:

Table 2. Associations between phenotype-PRS (columns), and measures of adiposity (rows).

Betas are reported for standardized inverse normalized values, followed by their respective p-values. Nominally statistically significant results (p<5.00×10–2) are in bold font.

https://doi.org/10.1371/journal.pone.0255609.t002

In contrast, after adjusting for BMI in the models, only BF% and WHR were found to be positive predictors of BF%, with attenuated effect sizes (β = 1.3 (p = 2.1 ×10⁻⁷) for BF%-PRS and β = 0.2 (p = 2.6 ×10⁻³) for WHR-PRS respectively) (Table 2). For the SAT phenotype, BF% and WC-PRS remained positive predictors, but not WHR-PRS. For the VAT phenotype, both WHR and WC-PRS were positively associated, but not BF%-PRS. For the VSR, BF%-PRS was consistent negative predictor and WHR-PRS was positive predictor. Generally, the PRS of BMI and BF% phenotypes explained higher proportions of variance for BF% and SAT than VAT (S4 Table).

Results for PRS obtained under approach 2 and 3 are provided in S5 Table, which were largely reflective of estimates under the principle approach but less precise due to reduced number of SNPs used for the PRS calculation. Associations between WHR and WC-PRS were less consistent in both BMI and non-BMI adjusted models. BF%-PRS, in contrast, was consistently associated with BF% and SAT under both approach 2 and 3 and with/without BMI adjustment.

Association with individual SNPs

Since PRS were summary statistics of risk alleles and do not illustrate the scale of association of each risk allele, which may vary when compared to other variants, we performed separate multivariate regression tests with individual variants. We used 33, 8, 12, and 4 risk allele SNPs which were nominally significant in BMI, WC_BMIadj, WHR_BMIadj, and BF% GWAS results in JHS, to characterize individual variants’ association with BF%, SAT, VAT, and VSR ratio, respectively; each SNP represents an independent locus within the genome (based on linkage disequilibrium).

On the heatmap (S1 Fig), the primary clustering of adiposity pattern variables on the x-axis separate different types of fat. The primary clustering of the SNPs on the y-axis separates group of SNPs that are associated with increase in BF%, BMI, WC_BMIadj, and WHR_BMIadj. A clear majority of risk alleles associated with increased BMI and BF% are positively associated with SAT and overall BF%. Among WHR_BMIadj and WC_BMIadj-increasing alleles, no discernible or consistent pattern can be observed (S1 Fig).

Discussion

In this study we characterized the associations between genetic risk scores for BMI, WC_BMIadj, WHR_BMIadj, and BF% with body fat composition phenotypes, including overall BF%, VAT, SAT, and VSR among AA individuals in the JHS cohort.

Phenotypically, pairwise correlations among BMI, WC, BF% and SAT were strong, but considerably weaker correlations with VAT were observed (S4 Table). Although the pairwise correlation between BMI and WHR was somewhat weaker (ρ = 0.16) than observed in most studies (which often report a correlation of ~0.4 or more) [36–38], similarly weak correlations have been previously reported [39, 40], including in studies of African Americans individuals [41]. The BMI-WHR correlation is known to vary across age, sex, and ethnic groups [38].

All PRS were significant and positive predictors of BF% in AA; this illustrates that similar to most BMI-linked SNPs [16], most known BF%-associated loci, initially observed in EA individuals [19],, are likely transferrable to AA populations (e.g. associatons may be replicable across populations). Furthermore, BMI-associated loci have cross-phenotypic effects on body fat [19]; the significant association between BMI-PRS and BF% measures in our analysis may also be indicative of transferrability of similar biological mechanisms for these GWAS identified variants from EA to AA populations, though the causal SNPs at most of these loci are unknown.

Both WC_BMIadj and WHR_BMIadj-PRS were also found to be positive predictors of BF% when associations were not adjusted for BMI; only WHR_BMIadj-PRS remained positively associated with BF% after adjusting for BMI, albeit at an attenuated scale (Table 2). Similarly, both WC_BMIadj and WHR_BMIadj-PRS were positive predictors of SAT, but only WC_BMIadj-PRS remained associated after adjustment for BMI. Similarly inconsistent patterns were also observed at the individual SNP level.

Examination of individual SNPs that exhibited nominally significant associations to BF%, BMI, WC_BMIadj, and WHR_BMIadj in this cohort shows that larger proportions of BMI- and BF%- associated SNPs cluster with BF% and subcutaneous adiposity measures. WC_BMIadj- associated SNPs, in contrast, do not appreciably cluster with any of the adiposity traits.

Collectively, results from both PRS models and supplementary cluster analyses suggest that BF% and BMI SNPs seem to better predict SAT, whereas the more deleterious VAT (and VSR) are more likely represented by WHR-associated variants, of the more commonly measured adiposity indices. Difference in patterns of visceral adiposity in AA [42], with significantly lower levels of VAT in BMI-adjusted analysis noted previously [13], that suggests a favorable visceral adiposity profile compared to EA [43]. Results highlighted the importance of population-specific GWAS studies on central adiposity measures in addition to overall adiposity phenotypes like BMI. Due to the limited availability of GWAS on VAT or VSR phenotypes [44], with none in a well-powered AA cohort, we are unable to asssess whether these observed differences are primarily genetically driven. Larger studies in AA populations are needed to assess the genetics of these traits across ancestries.

This is the first study to examine the relationship between adiposity PRS, including BF%, WC_BMIadj and WHR_BMIadj, and adiposity traits in AA. Utilization of several PRS with different risk set configurations made it possible to examine associations under comparative scenarios, minimizing potential sources of bias in estimates.

Relatively few African American cohorts with comprehensive assessment of adiposity metrics (including fat mass and CT based adiposity metrics) and genetic data exist, limiting our statistical power for these analyses. However, JHS is to our knowledge the largest such cohort and consistency of results under multiple approaches suggested a measure of robustness for estimates.

Results of this assessment pertain only to adults. Many genetic variants associated with obesity traits are age-dependent [45]. For example, in one prior study, there was no association between BMI-PRS and BF% in <5 years olds [46]; thus, extrapolation of results from this study to younger age categories in AA individuals may be unwarranted. This limitation is likely to be extended to WC_BMIadj- and WHR_BMIadj- associated SNPs since they also exhibit interaction with age [47]. We could not account for differential associations of dimorphic or sex-specific variants because of smaller numbers of male participants. Bioimpedance methods also have limitations [48], where the method slightly underestimates BF% in highly obese males [49]. However, we do not expect this limitation to substantially bias our results since only 6% of males had BMI measures >40 in this dataset.

Due to limited sample size for deriving adequate training and testing datasets, we did not perform extensive comparisons of different PRS derivation methods in JHS (for example comparisons of LDpred and pruning [22] and thresholding methods). Our unweighted risk score method, including an LD-pruned list of previously reported genome-wide significant variants in PRS derivation for each trait, may not provide ideal predictive power. However, this simple unweighted method is also likely to be less strongly influenced by differential LD and variant effect sizes across ancestral populations. Additionally, this approach allowed us to utilize SNPs both from large European focused GWAS for these traits and smaller studies in African American populations. PRS derivation and weighting by previously reported effect sizes is complex in admixed African American populations [50, 51], especially given the lack of representation of participants of diverse genetic ancestry in prior GWAS [52, 53]. Comparisons of different methods and use of ancestry-specific effect size estimates for PRS should be explored in future work for adiposity traits. Finally, the cross-sectional nature of the assessment restricts causal inferences.

In conclusion, these analyses illustrate that anthropometric phenotype-associated loci, initially explored in predominantly EA populations, are generally transferrable to AAs. Our results suggest that total gain in fat mass in AA, at least for gains in fat mass mediated by genetic factors, may be mostly through subcutaneous rather than visceral adiposity, but a comparable assessment in other populations, including Europeans, would be required to make firm conclusions about any population differences. Absence of association between anthropometric PRS, particularly WC_BMIadj and WHR_BMIadj, and adiposity traits like BF% may imply that the latter phenotypic measure are likely driven by genetic variants which influence overall adiposity versus central obesity.

Supporting information

S1 Table. SNPs used for configuration of polygenic risk scores.

https://doi.org/10.1371/journal.pone.0255609.s001

(XLSX)

S2 Table. Number of SNPs configurations used for calculation of polygenic risk scores under complementary approaches.

https://doi.org/10.1371/journal.pone.0255609.s002

(DOCX)

S3 Table. Polygenic risk score validation.

Estimates for all phenotypes other than percentage body fat (%BF).

https://doi.org/10.1371/journal.pone.0255609.s003

(DOCX)

S4 Table. Matrix of correlation between observed phenotypic measures for Jackson Heart Study.

Spearman correlation coefficients were calculated.

https://doi.org/10.1371/journal.pone.0255609.s004

(DOCX)

S5 Table. Coefficients of determination (R²) for PRS.

They exhibit the proportion of adiposity traits’(BF%, SAT, VAT, VSR) variances were predicted by each individual PRS.

https://doi.org/10.1371/journal.pone.0255609.s005

(DOCX)

S6 Table. Associations between phenotype-PRS (columns), and measures of adiposity (rows).

Betas are reported for standardized inverse normalized values, followed by their respective p-values. Nominally statistically significant results (p<5.00×10⁻²) are in bold font.

https://doi.org/10.1371/journal.pone.0255609.s006

(DOCX)

S1 Fig. Heatmap plot of association between top phenotype-linked SNPs and adiposity traits.

These SNPs were used for PRS calculation under approach 3 and represent polymorphisms with evidence of directionally consistent and statistically significant associations with their respective traits in JHS-genome wide assessment. Ward’s method used for SNP and trait clustering.

https://doi.org/10.1371/journal.pone.0255609.s007

(TIF)

Acknowledgments

Disclaimer: The views expressed in this manuscript are those of the authors and do not necessarily represent the views of the National Heart, Lung, and Blood Institute; the National Institutes of Health; or the U.S. Department of Health and Human Services.

References

1. Bell J. A. et al., "Associations of body mass and fat indexes with cardiometabolic traits," Journal of the American College of Cardiology, vol. 72, no. 24, pp. 3142–3154, 2018. pmid:30545453
- View Article
- PubMed/NCBI
- Google Scholar
2. Nevill A. M., Stewart A. D., Olds T., and Holder R., "Relationship between adiposity and body size reveals limitations of BMI," American Journal of Physical Anthropology: The Official Publication of the American Association of Physical Anthropologists, vol. 129, no. 1, pp. 151–156, 2006. pmid:16270304
- View Article
- PubMed/NCBI
- Google Scholar
3. Romero-Corral A. et al., "Accuracy of body mass index in diagnosing obesity in the adult general population," International journal of obesity, vol. 32, no. 6, pp. 959–966, 2008. pmid:18283284
- View Article
- PubMed/NCBI
- Google Scholar
4. Gómez-Ambrosi J. et al., "Body mass index classification misses subjects with increased cardiometabolic risk factors related to elevated adiposity," International journal of obesity, vol. 36, no. 2, pp. 286–294, 2012. pmid:21587201
- View Article
- PubMed/NCBI
- Google Scholar
5. De Lorenzo A. et al., "Adiposity rather than BMI determines metabolic risk," International journal of cardiology, vol. 166, no. 1, pp. 111–117, 2013. pmid:22088224
- View Article
- PubMed/NCBI
- Google Scholar
6. Zeng Q., Dong S.-Y., Sun X.-N., Xie J., and Cui Y., "Percent body fat is a better predictor of cardiovascular risk factors than body mass index," Brazilian Journal of Medical and Biological Research, vol. 45, no. 7, pp. 591–600, 2012. pmid:22510779
- View Article
- PubMed/NCBI
- Google Scholar
7. Van Pelt R., Evans E., Schechtman K., Ehsani A., and Kohrt W., "Contributions of total and regional fat mass to risk for cardiovascular disease in older women," American Journal of Physiology-Endocrinology and Metabolism, 2002. pmid:11934666
- View Article
- PubMed/NCBI
- Google Scholar
8. Janssen I., Katzmarzyk P. T., and Ross R., "Waist circumference and not body mass index explains obesity-related health risk," The American journal of clinical nutrition, vol. 79, no. 3, pp. 379–384, 2004. pmid:14985210
- View Article
- PubMed/NCBI
- Google Scholar
9. Deurenberg P., Weststrate J. A., and Seidell J. C., "Body mass index as a measure of body fatness: age-and sex-specific prediction formulas," British journal of nutrition, vol. 65, no. 2, pp. 105–114, 1991. pmid:2043597
- View Article
- PubMed/NCBI
- Google Scholar
10. Clifton E. A. et al., "Associations between body mass index-related genetic variants and adult body composition: The Fenland cohort study," International journal of obesity, vol. 41, no. 4, pp. 613–619, 2017. pmid:28096530
- View Article
- PubMed/NCBI
- Google Scholar
11. Lear S. A., Humphries K. H., Kohli S., and Birmingham C. L., "The use of BMI and waist circumference as surrogates of body fat differs by ethnicity," Obesity, vol. 15, no. 11, pp. 2817–2824, 2007. pmid:18070773
- View Article
- PubMed/NCBI
- Google Scholar
12. Hill J. O., Sidney S., Lewis C. E., Tolan K., Scherzinger A. L., and Stamm E. R., "Racial differences in amounts of visceral adipose tissue in young adults: the CARDIA (Coronary Artery Risk Development in Young Adults) study," The American journal of clinical nutrition, vol. 69, no. 3, pp. 381–387, 1999. pmid:10075320
- View Article
- PubMed/NCBI
- Google Scholar
13. Carroll J. F. et al., "Visceral fat, waist circumference, and BMI: impact of race/ethnicity," Obesity, vol. 16, no. 3, pp. 600–607, 2008. pmid:18239557
- View Article
- PubMed/NCBI
- Google Scholar
14. Taylor H. A. Jr et al., "Relationships of BMI to cardiovascular risk factors differ by ethnicity," Obesity, vol. 18, no. 8, pp. 1638–1645, 2010. pmid:19927137
- View Article
- PubMed/NCBI
- Google Scholar
15. Chen G. et al., "Genome‐wide analysis identifies an african‐specific variant in SEMA4D associated with body mass index," Obesity, vol. 25, no. 4, pp. 794–800, 2017. pmid:28296344
- View Article
- PubMed/NCBI
- Google Scholar
16. Monda K. L. et al., "A meta-analysis identifies new loci associated with body mass index in individuals of African ancestry," Nature genetics, vol. 45, no. 6, pp. 690–696, 2013. pmid:23583978
- View Article
- PubMed/NCBI
- Google Scholar
17. Munthali R. J. et al., "Genetic risk score for adult body mass index associations with childhood and adolescent weight gain in an African population," Genes & nutrition, vol. 13, no. 1, pp. 1–9, 2018.
- View Article
- Google Scholar
18. Ng M. C. et al., "Discovery and fine-mapping of adiposity loci using high density imputation of genome-wide association studies in individuals of African ancestry: African Ancestry Anthropometry Genetics Consortium," PLoS genetics, vol. 13, no. 4, p. e1006719, 2017. pmid:28430825
- View Article
- PubMed/NCBI
- Google Scholar
19. Lu Y. et al., "New loci for body fat percentage reveal link between adiposity and cardiometabolic disease risk," Nature communications, vol. 7, no. 1, pp. 1–15, 2016. pmid:26833246
- View Article
- PubMed/NCBI
- Google Scholar
20. Peterson R. E. et al., "Genetic risk sum score comprised of common polygenic variation is associated with body mass index," Human genetics, vol. 129, no. 2, pp. 221–230, 2011. pmid:21104096
- View Article
- PubMed/NCBI
- Google Scholar
21. HLA-C H., "Common polygenic variation contributes to risk of schizophrenia and bipolar disorder," 2009. pmid:19571811
- View Article
- PubMed/NCBI
- Google Scholar
22. Vilhjálmsson B. J. et al., "Modeling linkage disequilibrium increases accuracy of polygenic risk scores," The american journal of human genetics, vol. 97, no. 4, pp. 576–592, 2015. pmid:26430803
- View Article
- PubMed/NCBI
- Google Scholar
23. Dudbridge F., "Power and predictive accuracy of polygenic risk scores," PLoS Genet, vol. 9, no. 3, p. e1003348, 2013. pmid:23555274
- View Article
- PubMed/NCBI
- Google Scholar
24. Hu Y. et al., "Minority-centric meta-analyses of blood lipid levels identify novel loci in the Population Architecture using Genomics and Epidemiology (PAGE) study," PLoS genetics, vol. 16, no. 3, p. e1008684, 2020. pmid:32226016
- View Article
- PubMed/NCBI
- Google Scholar
25. Goddard M. E., Wray N. R., Verbyla K., and Visscher P. M., "Estimating effects and making predictions from genome-wide marker data," Statistical science, vol. 24, no. 4, pp. 517–529, 2009.
- View Article
- Google Scholar
26. Zöllner S. and Pritchard J. K., "Overcoming the winner’s curse: estimating penetrance parameters from case-control data," The American Journal of Human Genetics, vol. 80, no. 4, pp. 605–615, 2007. pmid:17357068
- View Article
- PubMed/NCBI
- Google Scholar
27. Martin A. R. et al., "Human demographic history impacts genetic risk prediction across diverse populations," The American Journal of Human Genetics, vol. 100, no. 4, pp. 635–649, 2017. pmid:28366442
- View Article
- PubMed/NCBI
- Google Scholar
28. Lee S. Y. and Gallagher D., "Assessment methods in human body composition," Current opinion in clinical nutrition and metabolic care, vol. 11, no. 5, p. 566, 2008. pmid:18685451
- View Article
- PubMed/NCBI
- Google Scholar
29. Gutin B., Litaker M., Islam S., Manos T., Smith C., and Treiber F., "Body-composition measurement in 9–11-y-old children by dual-energy X-ray absorptiometry, skinfold-thickness measurements, and bioimpedance analysis," The American journal of clinical nutrition, vol. 63, no. 3, pp. 287–292, 1996. pmid:8602582
- View Article
- PubMed/NCBI
- Google Scholar
30. Liu J. et al., "Impact of abdominal visceral and subcutaneous adipose tissue on cardiometabolic risk factors: the Jackson Heart Study," The Journal of Clinical Endocrinology & Metabolism, vol. 95, no. 12, pp. 5419–5426, 2010. pmid:20843952
- View Article
- PubMed/NCBI
- Google Scholar
31. Kang H. M., "EPACTS: efficient and parallelizable association container toolbox," EPACTS: efficient and parallelizable association container toolbox, 2014.
- View Article
- Google Scholar
32. Tosto G. et al., "Polygenic risk scores in familial Alzheimer disease," Neurology, vol. 88, no. 12, pp. 1180–1186, 2017. pmid:28213371
- View Article
- PubMed/NCBI
- Google Scholar
33. Pulit S. L., Karaderi T., and Lindgren C. M., "Sexual dimorphisms in genetic loci linked to body fat distribution," Bioscience reports, vol. 37, no. 1, 2017. pmid:28073971
- View Article
- PubMed/NCBI
- Google Scholar
34. Ward J. H. Jr, "Hierarchical grouping to optimize an objective function," Journal of the American statistical association, vol. 58, no. 301, pp. 236–244, 1963.
- View Article
- Google Scholar
35. Kolde R. and Kolde M. R., "Package ‘pheatmap’," R package, vol. 1, no. 7, p. 790, 2015.
- View Article
- Google Scholar
36. Elsayed E. F. et al., "Waist-to-hip ratio, body mass index, and subsequent kidney disease and death," American Journal of Kidney Diseases, vol. 52, no. 1, pp. 29–38, 2008. pmid:18511168
- View Article
- PubMed/NCBI
- Google Scholar
37. Harris M. M., Stevens J., Thomas N., Schreiner P., and Folsom A. R., "Associations of fat distribution and obesity with hypertension in a bi‐ethnic population: the ARIC study," Obesity research, vol. 8, no. 7, pp. 516–524, 2000. pmid:11068957
- View Article
- PubMed/NCBI
- Google Scholar
38. Vazquez G., Duval S., Jacobs D. R. Jr, and Silventoinen K., "Comparison of body mass index, waist circumference, and waist/hip ratio in predicting incident diabetes: a meta-analysis," Epidemiologic reviews, vol. 29, no. 1, pp. 115–128, 2007.
- View Article
- Google Scholar
39. Ahmad N., Adam S. I. M., Nawi A. M., Hassan M. R., and Ghazi H. F., "Abdominal obesity indicators: waist circumference or waist-to-hip ratio in Malaysian adults population," International journal of preventive medicine, vol. 7, 2016. pmid:27330688
- View Article
- PubMed/NCBI
- Google Scholar
40. Fredriks A. M., van Buuren S., Fekkes M., Verloove-Vanhorick S. P., and Wit J. M., "Are age references for waist circumference, hip circumference and waist-hip ratio in Dutch children useful in clinical practice?," European journal of pediatrics, vol. 164, no. 4, pp. 216–222, 2005. pmid:15662504
- View Article
- PubMed/NCBI
- Google Scholar
41. Nassir R. et al., "Relationship between adiposity and admixture in African-American and Hispanic-American women," International journal of obesity, vol. 36, no. 2, pp. 304–313, 2012. pmid:21487399
- View Article
- PubMed/NCBI
- Google Scholar
42. Katzmarzyk P. T. et al., "Racial differences in abdominal depot–specific adiposity in white and African American adults," The American journal of clinical nutrition, vol. 91, no. 1, pp. 7–15, 2010. pmid:19828714
- View Article
- PubMed/NCBI
- Google Scholar
43. Després J.-P. et al., "Race, visceral adipose tissue, plasma lipids, and lipoprotein lipase activity in men and women: the Health, Risk Factors, Exercise Training, and Genetics (HERITAGE) family study," Arteriosclerosis, thrombosis, and vascular biology, vol. 20, no. 8, pp. 1932–1938, 2000. pmid:10938014
- View Article
- PubMed/NCBI
- Google Scholar
44. Fox C. S. et al., "Genome-wide association for abdominal subcutaneous and visceral adipose reveals a novel locus for visceral fat in women," PLoS Genet, vol. 8, no. 5, p. e1002695, 2012. pmid:22589738
- View Article
- PubMed/NCBI
- Google Scholar
45. Winkler T. W. et al., "The influence of age and sex on genetic associations with adult body size and shape: a large-scale genome-wide interaction study," PLoS Genet, vol. 11, no. 10, p. e1005378, 2015. pmid:26426971
- View Article
- PubMed/NCBI
- Google Scholar
46. Elks C. E. et al., "Associations between genetic obesity susceptibility and early postnatal fat and lean mass: an individual participant meta-analysis," JAMA pediatrics, vol. 168, no. 12, pp. 1122–1130, 2014. pmid:25329327
- View Article
- PubMed/NCBI
- Google Scholar
47. Perissinotto E., Pisent C., Sergi G., Grigoletto F., Enzi G., and Group I. W., "Anthropometric measurements in the elderly: age and gender differences," British Journal of nutrition, vol. 87, no. 2, pp. 177–186, 2002. pmid:11895170
- View Article
- PubMed/NCBI
- Google Scholar
48. Deurenberg P., "Limitations of the bioelectrical impedance method for the assessment of body fat in severe obesity," The American journal of clinical nutrition, vol. 64, no. 3, pp. 449S–452S, 1996. pmid:8780361
- View Article
- PubMed/NCBI
- Google Scholar
49. Leal A. A. et al., "Bioimpedance analysis: should it be used in morbid obesity?," American Journal of Human Biology, vol. 23, no. 3, pp. 420–422, 2011. pmid:21448907
- View Article
- PubMed/NCBI
- Google Scholar
50. Martin A. R., Kanai M., Kamatani Y., Okada Y., Neale B. M., and Daly M. J., "Clinical use of current polygenic risk scores may exacerbate health disparities," Nature genetics, vol. 51, no. 4, pp. 584–591, 2019. pmid:30926966
- View Article
- PubMed/NCBI
- Google Scholar
51. Duncan L. et al., "Analysis of polygenic risk score usage and performance in diverse human populations," Nature communications, vol. 10, no. 1, pp. 1–9, 2019. pmid:30602773
- View Article
- PubMed/NCBI
- Google Scholar
52. Mills M. C. and Rahal C., "The GWAS Diversity Monitor tracks diversity by disease in real time," Nature genetics, vol. 52, no. 3, pp. 242–243, 2020. pmid:32139905
- View Article
- PubMed/NCBI
- Google Scholar
53. Bentley A. R., Callier S. L., and Rotimi C. N., "Evaluating the promise of inclusion of African ancestry populations in genomics," NPJ genomic medicine, vol. 5, no. 1, pp. 1–9, 2020. pmid:32140257
- View Article
- PubMed/NCBI
- Google Scholar

[ref1] 1. Bell J. A. et al., "Associations of body mass and fat indexes with cardiometabolic traits," Journal of the American College of Cardiology, vol. 72, no. 24, pp. 3142–3154, 2018. pmid:30545453
View Article
PubMed/NCBI
Google Scholar

[2] View Article

[3] PubMed/NCBI

[4] Google Scholar

[ref2] 2. Nevill A. M., Stewart A. D., Olds T., and Holder R., "Relationship between adiposity and body size reveals limitations of BMI," American Journal of Physical Anthropology: The Official Publication of the American Association of Physical Anthropologists, vol. 129, no. 1, pp. 151–156, 2006. pmid:16270304
View Article
PubMed/NCBI
Google Scholar

[6] View Article

[7] PubMed/NCBI

[8] Google Scholar

[ref3] 3. Romero-Corral A. et al., "Accuracy of body mass index in diagnosing obesity in the adult general population," International journal of obesity, vol. 32, no. 6, pp. 959–966, 2008. pmid:18283284
View Article
PubMed/NCBI
Google Scholar

[10] View Article

[11] PubMed/NCBI

[12] Google Scholar

[ref4] 4. Gómez-Ambrosi J. et al., "Body mass index classification misses subjects with increased cardiometabolic risk factors related to elevated adiposity," International journal of obesity, vol. 36, no. 2, pp. 286–294, 2012. pmid:21587201
View Article
PubMed/NCBI
Google Scholar

[14] View Article

[15] PubMed/NCBI

[16] Google Scholar

[ref5] 5. De Lorenzo A. et al., "Adiposity rather than BMI determines metabolic risk," International journal of cardiology, vol. 166, no. 1, pp. 111–117, 2013. pmid:22088224
View Article
PubMed/NCBI
Google Scholar

[18] View Article

[19] PubMed/NCBI

[20] Google Scholar

[ref6] 6. Zeng Q., Dong S.-Y., Sun X.-N., Xie J., and Cui Y., "Percent body fat is a better predictor of cardiovascular risk factors than body mass index," Brazilian Journal of Medical and Biological Research, vol. 45, no. 7, pp. 591–600, 2012. pmid:22510779
View Article
PubMed/NCBI
Google Scholar

[22] View Article

[23] PubMed/NCBI

[24] Google Scholar

[ref7] 7. Van Pelt R., Evans E., Schechtman K., Ehsani A., and Kohrt W., "Contributions of total and regional fat mass to risk for cardiovascular disease in older women," American Journal of Physiology-Endocrinology and Metabolism, 2002. pmid:11934666
View Article
PubMed/NCBI
Google Scholar

[26] View Article

[27] PubMed/NCBI

[28] Google Scholar

[ref8] 8. Janssen I., Katzmarzyk P. T., and Ross R., "Waist circumference and not body mass index explains obesity-related health risk," The American journal of clinical nutrition, vol. 79, no. 3, pp. 379–384, 2004. pmid:14985210
View Article
PubMed/NCBI
Google Scholar

[30] View Article

[31] PubMed/NCBI

[32] Google Scholar

[ref9] 9. Deurenberg P., Weststrate J. A., and Seidell J. C., "Body mass index as a measure of body fatness: age-and sex-specific prediction formulas," British journal of nutrition, vol. 65, no. 2, pp. 105–114, 1991. pmid:2043597
View Article
PubMed/NCBI
Google Scholar

[34] View Article

[35] PubMed/NCBI

[36] Google Scholar

[ref10] 10. Clifton E. A. et al., "Associations between body mass index-related genetic variants and adult body composition: The Fenland cohort study," International journal of obesity, vol. 41, no. 4, pp. 613–619, 2017. pmid:28096530
View Article
PubMed/NCBI
Google Scholar

[38] View Article

[39] PubMed/NCBI

[40] Google Scholar

[ref11] 11. Lear S. A., Humphries K. H., Kohli S., and Birmingham C. L., "The use of BMI and waist circumference as surrogates of body fat differs by ethnicity," Obesity, vol. 15, no. 11, pp. 2817–2824, 2007. pmid:18070773
View Article
PubMed/NCBI
Google Scholar

[42] View Article

[43] PubMed/NCBI

[44] Google Scholar

[ref12] 12. Hill J. O., Sidney S., Lewis C. E., Tolan K., Scherzinger A. L., and Stamm E. R., "Racial differences in amounts of visceral adipose tissue in young adults: the CARDIA (Coronary Artery Risk Development in Young Adults) study," The American journal of clinical nutrition, vol. 69, no. 3, pp. 381–387, 1999. pmid:10075320
View Article
PubMed/NCBI
Google Scholar

[46] View Article

[47] PubMed/NCBI

[48] Google Scholar

[ref13] 13. Carroll J. F. et al., "Visceral fat, waist circumference, and BMI: impact of race/ethnicity," Obesity, vol. 16, no. 3, pp. 600–607, 2008. pmid:18239557
View Article
PubMed/NCBI
Google Scholar

[50] View Article

[51] PubMed/NCBI

[52] Google Scholar

[ref14] 14. Taylor H. A. Jr et al., "Relationships of BMI to cardiovascular risk factors differ by ethnicity," Obesity, vol. 18, no. 8, pp. 1638–1645, 2010. pmid:19927137
View Article
PubMed/NCBI
Google Scholar

[54] View Article

[55] PubMed/NCBI

[56] Google Scholar

[ref15] 15. Chen G. et al., "Genome‐wide analysis identifies an african‐specific variant in SEMA4D associated with body mass index," Obesity, vol. 25, no. 4, pp. 794–800, 2017. pmid:28296344
View Article
PubMed/NCBI
Google Scholar

[58] View Article

[59] PubMed/NCBI

[60] Google Scholar

[ref16] 16. Monda K. L. et al., "A meta-analysis identifies new loci associated with body mass index in individuals of African ancestry," Nature genetics, vol. 45, no. 6, pp. 690–696, 2013. pmid:23583978
View Article
PubMed/NCBI
Google Scholar

[62] View Article

[63] PubMed/NCBI

[64] Google Scholar

[ref17] 17. Munthali R. J. et al., "Genetic risk score for adult body mass index associations with childhood and adolescent weight gain in an African population," Genes & nutrition, vol. 13, no. 1, pp. 1–9, 2018.
View Article
Google Scholar

[66] View Article

[67] Google Scholar

[ref18] 18. Ng M. C. et al., "Discovery and fine-mapping of adiposity loci using high density imputation of genome-wide association studies in individuals of African ancestry: African Ancestry Anthropometry Genetics Consortium," PLoS genetics, vol. 13, no. 4, p. e1006719, 2017. pmid:28430825
View Article
PubMed/NCBI
Google Scholar

[69] View Article

[70] PubMed/NCBI

[71] Google Scholar

[ref19] 19. Lu Y. et al., "New loci for body fat percentage reveal link between adiposity and cardiometabolic disease risk," Nature communications, vol. 7, no. 1, pp. 1–15, 2016. pmid:26833246
View Article
PubMed/NCBI
Google Scholar

[73] View Article

[74] PubMed/NCBI

[75] Google Scholar

[ref20] 20. Peterson R. E. et al., "Genetic risk sum score comprised of common polygenic variation is associated with body mass index," Human genetics, vol. 129, no. 2, pp. 221–230, 2011. pmid:21104096
View Article
PubMed/NCBI
Google Scholar

[77] View Article

[78] PubMed/NCBI

[79] Google Scholar

[ref21] 21. HLA-C H., "Common polygenic variation contributes to risk of schizophrenia and bipolar disorder," 2009. pmid:19571811
View Article
PubMed/NCBI
Google Scholar

[81] View Article

[82] PubMed/NCBI

[83] Google Scholar

[ref22] 22. Vilhjálmsson B. J. et al., "Modeling linkage disequilibrium increases accuracy of polygenic risk scores," The american journal of human genetics, vol. 97, no. 4, pp. 576–592, 2015. pmid:26430803
View Article
PubMed/NCBI
Google Scholar

[85] View Article

[86] PubMed/NCBI

[87] Google Scholar

[ref23] 23. Dudbridge F., "Power and predictive accuracy of polygenic risk scores," PLoS Genet, vol. 9, no. 3, p. e1003348, 2013. pmid:23555274
View Article
PubMed/NCBI
Google Scholar

[89] View Article

[90] PubMed/NCBI

[91] Google Scholar

[ref24] 24. Hu Y. et al., "Minority-centric meta-analyses of blood lipid levels identify novel loci in the Population Architecture using Genomics and Epidemiology (PAGE) study," PLoS genetics, vol. 16, no. 3, p. e1008684, 2020. pmid:32226016
View Article
PubMed/NCBI
Google Scholar

[93] View Article

[94] PubMed/NCBI

[95] Google Scholar

[ref25] 25. Goddard M. E., Wray N. R., Verbyla K., and Visscher P. M., "Estimating effects and making predictions from genome-wide marker data," Statistical science, vol. 24, no. 4, pp. 517–529, 2009.
View Article
Google Scholar

[97] View Article

[98] Google Scholar

[ref26] 26. Zöllner S. and Pritchard J. K., "Overcoming the winner’s curse: estimating penetrance parameters from case-control data," The American Journal of Human Genetics, vol. 80, no. 4, pp. 605–615, 2007. pmid:17357068
View Article
PubMed/NCBI
Google Scholar

[100] View Article

[101] PubMed/NCBI

[102] Google Scholar

[ref27] 27. Martin A. R. et al., "Human demographic history impacts genetic risk prediction across diverse populations," The American Journal of Human Genetics, vol. 100, no. 4, pp. 635–649, 2017. pmid:28366442
View Article
PubMed/NCBI
Google Scholar

[104] View Article

[105] PubMed/NCBI

[106] Google Scholar

[ref28] 28. Lee S. Y. and Gallagher D., "Assessment methods in human body composition," Current opinion in clinical nutrition and metabolic care, vol. 11, no. 5, p. 566, 2008. pmid:18685451
View Article
PubMed/NCBI
Google Scholar

[108] View Article

[109] PubMed/NCBI

[110] Google Scholar

[ref29] 29. Gutin B., Litaker M., Islam S., Manos T., Smith C., and Treiber F., "Body-composition measurement in 9–11-y-old children by dual-energy X-ray absorptiometry, skinfold-thickness measurements, and bioimpedance analysis," The American journal of clinical nutrition, vol. 63, no. 3, pp. 287–292, 1996. pmid:8602582
View Article
PubMed/NCBI
Google Scholar

[112] View Article

[113] PubMed/NCBI

[114] Google Scholar

[ref30] 30. Liu J. et al., "Impact of abdominal visceral and subcutaneous adipose tissue on cardiometabolic risk factors: the Jackson Heart Study," The Journal of Clinical Endocrinology & Metabolism, vol. 95, no. 12, pp. 5419–5426, 2010. pmid:20843952
View Article
PubMed/NCBI
Google Scholar

[116] View Article

[117] PubMed/NCBI

[118] Google Scholar

[ref31] 31. Kang H. M., "EPACTS: efficient and parallelizable association container toolbox," EPACTS: efficient and parallelizable association container toolbox, 2014.
View Article
Google Scholar

[120] View Article

[121] Google Scholar

[ref32] 32. Tosto G. et al., "Polygenic risk scores in familial Alzheimer disease," Neurology, vol. 88, no. 12, pp. 1180–1186, 2017. pmid:28213371
View Article
PubMed/NCBI
Google Scholar

[123] View Article

[124] PubMed/NCBI

[125] Google Scholar

[ref33] 33. Pulit S. L., Karaderi T., and Lindgren C. M., "Sexual dimorphisms in genetic loci linked to body fat distribution," Bioscience reports, vol. 37, no. 1, 2017. pmid:28073971
View Article
PubMed/NCBI
Google Scholar

[127] View Article

[128] PubMed/NCBI

[129] Google Scholar

[ref34] 34. Ward J. H. Jr, "Hierarchical grouping to optimize an objective function," Journal of the American statistical association, vol. 58, no. 301, pp. 236–244, 1963.
View Article
Google Scholar

[131] View Article

[132] Google Scholar

[ref35] 35. Kolde R. and Kolde M. R., "Package ‘pheatmap’," R package, vol. 1, no. 7, p. 790, 2015.
View Article
Google Scholar

[134] View Article

[135] Google Scholar

[ref36] 36. Elsayed E. F. et al., "Waist-to-hip ratio, body mass index, and subsequent kidney disease and death," American Journal of Kidney Diseases, vol. 52, no. 1, pp. 29–38, 2008. pmid:18511168
View Article
PubMed/NCBI
Google Scholar

[137] View Article

[138] PubMed/NCBI

[139] Google Scholar

[ref37] 37. Harris M. M., Stevens J., Thomas N., Schreiner P., and Folsom A. R., "Associations of fat distribution and obesity with hypertension in a bi‐ethnic population: the ARIC study," Obesity research, vol. 8, no. 7, pp. 516–524, 2000. pmid:11068957
View Article
PubMed/NCBI
Google Scholar

[141] View Article

[142] PubMed/NCBI

[143] Google Scholar

[ref38] 38. Vazquez G., Duval S., Jacobs D. R. Jr, and Silventoinen K., "Comparison of body mass index, waist circumference, and waist/hip ratio in predicting incident diabetes: a meta-analysis," Epidemiologic reviews, vol. 29, no. 1, pp. 115–128, 2007.
View Article
Google Scholar

[145] View Article

[146] Google Scholar

[ref39] 39. Ahmad N., Adam S. I. M., Nawi A. M., Hassan M. R., and Ghazi H. F., "Abdominal obesity indicators: waist circumference or waist-to-hip ratio in Malaysian adults population," International journal of preventive medicine, vol. 7, 2016. pmid:27330688
View Article
PubMed/NCBI
Google Scholar

[148] View Article

[149] PubMed/NCBI

[150] Google Scholar

[ref40] 40. Fredriks A. M., van Buuren S., Fekkes M., Verloove-Vanhorick S. P., and Wit J. M., "Are age references for waist circumference, hip circumference and waist-hip ratio in Dutch children useful in clinical practice?," European journal of pediatrics, vol. 164, no. 4, pp. 216–222, 2005. pmid:15662504
View Article
PubMed/NCBI
Google Scholar

[152] View Article

[153] PubMed/NCBI

[154] Google Scholar

[ref41] 41. Nassir R. et al., "Relationship between adiposity and admixture in African-American and Hispanic-American women," International journal of obesity, vol. 36, no. 2, pp. 304–313, 2012. pmid:21487399
View Article
PubMed/NCBI
Google Scholar

[156] View Article

[157] PubMed/NCBI

[158] Google Scholar

[ref42] 42. Katzmarzyk P. T. et al., "Racial differences in abdominal depot–specific adiposity in white and African American adults," The American journal of clinical nutrition, vol. 91, no. 1, pp. 7–15, 2010. pmid:19828714
View Article
PubMed/NCBI
Google Scholar

[160] View Article

[161] PubMed/NCBI

[162] Google Scholar

[ref43] 43. Després J.-P. et al., "Race, visceral adipose tissue, plasma lipids, and lipoprotein lipase activity in men and women: the Health, Risk Factors, Exercise Training, and Genetics (HERITAGE) family study," Arteriosclerosis, thrombosis, and vascular biology, vol. 20, no. 8, pp. 1932–1938, 2000. pmid:10938014
View Article
PubMed/NCBI
Google Scholar

[164] View Article

[165] PubMed/NCBI

[166] Google Scholar

[ref44] 44. Fox C. S. et al., "Genome-wide association for abdominal subcutaneous and visceral adipose reveals a novel locus for visceral fat in women," PLoS Genet, vol. 8, no. 5, p. e1002695, 2012. pmid:22589738
View Article
PubMed/NCBI
Google Scholar

[168] View Article

[169] PubMed/NCBI

[170] Google Scholar

[ref45] 45. Winkler T. W. et al., "The influence of age and sex on genetic associations with adult body size and shape: a large-scale genome-wide interaction study," PLoS Genet, vol. 11, no. 10, p. e1005378, 2015. pmid:26426971
View Article
PubMed/NCBI
Google Scholar

[172] View Article

[173] PubMed/NCBI

[174] Google Scholar

[ref46] 46. Elks C. E. et al., "Associations between genetic obesity susceptibility and early postnatal fat and lean mass: an individual participant meta-analysis," JAMA pediatrics, vol. 168, no. 12, pp. 1122–1130, 2014. pmid:25329327
View Article
PubMed/NCBI
Google Scholar

[176] View Article

[177] PubMed/NCBI

[178] Google Scholar

[ref47] 47. Perissinotto E., Pisent C., Sergi G., Grigoletto F., Enzi G., and Group I. W., "Anthropometric measurements in the elderly: age and gender differences," British Journal of nutrition, vol. 87, no. 2, pp. 177–186, 2002. pmid:11895170
View Article
PubMed/NCBI
Google Scholar

[180] View Article

[181] PubMed/NCBI

[182] Google Scholar

[ref48] 48. Deurenberg P., "Limitations of the bioelectrical impedance method for the assessment of body fat in severe obesity," The American journal of clinical nutrition, vol. 64, no. 3, pp. 449S–452S, 1996. pmid:8780361
View Article
PubMed/NCBI
Google Scholar

[184] View Article

[185] PubMed/NCBI

[186] Google Scholar

[ref49] 49. Leal A. A. et al., "Bioimpedance analysis: should it be used in morbid obesity?," American Journal of Human Biology, vol. 23, no. 3, pp. 420–422, 2011. pmid:21448907
View Article
PubMed/NCBI
Google Scholar

[188] View Article

[189] PubMed/NCBI

[190] Google Scholar

[ref50] 50. Martin A. R., Kanai M., Kamatani Y., Okada Y., Neale B. M., and Daly M. J., "Clinical use of current polygenic risk scores may exacerbate health disparities," Nature genetics, vol. 51, no. 4, pp. 584–591, 2019. pmid:30926966
View Article
PubMed/NCBI
Google Scholar

[192] View Article

[193] PubMed/NCBI

[194] Google Scholar

[ref51] 51. Duncan L. et al., "Analysis of polygenic risk score usage and performance in diverse human populations," Nature communications, vol. 10, no. 1, pp. 1–9, 2019. pmid:30602773
View Article
PubMed/NCBI
Google Scholar

[196] View Article

[197] PubMed/NCBI

[198] Google Scholar

[ref52] 52. Mills M. C. and Rahal C., "The GWAS Diversity Monitor tracks diversity by disease in real time," Nature genetics, vol. 52, no. 3, pp. 242–243, 2020. pmid:32139905
View Article
PubMed/NCBI
Google Scholar

[200] View Article

[201] PubMed/NCBI

[202] Google Scholar

[ref53] 53. Bentley A. R., Callier S. L., and Rotimi C. N., "Evaluating the promise of inclusion of African ancestry populations in genomics," NPJ genomic medicine, vol. 5, no. 1, pp. 1–9, 2020. pmid:32140257
View Article
PubMed/NCBI
Google Scholar

[204] View Article

[205] PubMed/NCBI

[206] Google Scholar

Figures

Abstract

Background

Methods

Results

Conclusion

Introduction

Methods

Study population

Genotyping

Polygenic risk scores

SNP risk sets

Anthropometry measures

Adiposity measures

Statistical analysis

Results

Population characteristics and adiposity measures

Association with polygenicc risk scores

Association with individual SNPs

Discussion

Supporting information

S1 Table. SNPs used for configuration of polygenic risk scores.

S2 Table. Number of SNPs configurations used for calculation of polygenic risk scores under complementary approaches.

S3 Table. Polygenic risk score validation.

S4 Table. Matrix of correlation between observed phenotypic measures for Jackson Heart Study.

S5 Table. Coefficients of determination (R2) for PRS.

S6 Table. Associations between phenotype-PRS (columns), and measures of adiposity (rows).

S1 Fig. Heatmap plot of association between top phenotype-linked SNPs and adiposity traits.

Acknowledgments

References

S5 Table. Coefficients of determination (R²) for PRS.