Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Feb 1.
Published in final edited form as: Obesity (Silver Spring). 2021 Feb;29(2):446–453. doi: 10.1002/oby.23070

A Genome-Wide Association Study of Childhood Body Fatness

Erica T Warner 1, Lai Jiang 2, David Nana Adjei 3, Constance Turman 2, William Gordon 4, Lu Wang 5, Rulla Tamimi 2,6, Peter Kraft 2,3, Sara Lindström 4,7
PMCID: PMC7842657  NIHMSID: NIHMS1640815  PMID: 33491310

Abstract

Objective:

To uncover genetic contributors to adiposity in early-life.

Methods:

We conducted a genome-wide association study (GWAS) of childhood body fatness in 34,401 individuals within the Nurses’ Health Studies and the Health Professionals Follow-up Study. We imputed using the 1000 Genomes Phase 3 v5 reference panel.

Results:

We identified 1,354 SNPs (p<10−4) from a previously published GWAS of childhood body mass index (BMI). We observed 19 genome-wide significant (p<5x 10−8) regions, of which 14 were previously associated with childhood obesity, while five were novel: BNDF (p= 7.58 x 10−13), PRKD1 (p= 1.43 x10−10), 20p13 (p= 2.05 x10−10), FHIT (p= 1.77 x10−08), and LOC101927575 (p=3.22 x 10−08). The BNDF, FHIT and PRKD1 regions were previously associated with adult BMI. LOC101927575 and 20p13 regions have not previously been associated with adiposity phenotypes. In transcriptome wide analyses (TWAS), we observed associations for POMC at 2p23.3 (p=3.36x10−6) and with TMEM18 at 2p25.3 (p=3.53x10−7). Childhood body fatness was genetically correlated with hip (rg=0.42, p=4.44 x 10−16) and waist circumference (rg=0.39, p= 5.56 x 10−16), and age at menarche (rg=−0.37, p= 7.96 x 10−19).

Conclusions:

We identified additional loci that contribute to childhood adiposity, further explicating its genetic architecture.

Keywords: obesity, childhood, adolescence, genome wide association study

Introduction

While recent growth in the prevalence of childhood obesity is likely due to environmental exposures and behaviors (1), heritability estimates for body fatness in childhood and adolescence are high (40-70%) (2). Obesity in childhood and adolescence is associated with obesity in adulthood, but the correlation between childhood and adult body mass index (BMI) is modest (r=0.24-0.30) (3, 4) and BMI heritability estimates are higher in children than in adults (5). These findings suggest that body size in childhood and adulthood represent two different phenotypes that likely have different etiologies. Yet, much of the literature on genetic loci influencing adiposity related traits, has focused on adulthood, implicating key genes such as INSIG2, FTO, MC4R, TMEM18, and GNPDA2 (6) and several previous reports demonstrate that the impact of FTO on adiposity begins in childhood (2, 7, 8). Others have examined whether loci associated with adult BMI are also associated with childhood BMI (9, 10). Among those focused on childhood BMI, several have examined early onset-extreme obesity phenotypes while others have identified copy-number variants (11, 12, 13). One recent paper examined eight loci identified in nine independent datasets (2,818 cases and 4,083 controls), two of which reached genome-wide significance when meta-analyzed (OLFM4 at 13q14; HOXB5 at 17q21) (14).

To identify additional genetic loci associated with body size in childhood, we conducted a GWAS among 34,401participants of the Nurses’ Health Studies I and II (NHS and NHSII) and the Health Professionals Follow-up Study (HPFS). We replicated promising findings (p< 10−4) in the latest childhood BMI GWAS from the EGG consortium (2013) including 35,688 children from 20 studies in the discovery phase and 11,873 children from 13 studies in the replication phase (15). We further performed a multi-tissue transcriptome-wide association study (TWAS) (16) to identify expression-trait associations based on three different gene expression reference panels, two in blood and one in adipose tissue. Lastly, we assessed cross-trait genetic correlations to estimate the shared heritability between childhood body fatness, adult BMI, and other traits.

Methods

Study Population

NHS is a prospective cohort study of 121,700 female registered nurses in 11 states in the United States who were 30-55 years of age at enrollment in 1976. NHSII was established in 1989 with 116,429 female registered nurses between the ages of 30 and 55. The HPFS is an ongoing cohort study established in 1986 when 51,529 men who were aged 40–75. In each cohort, participants complete self-administered questionnaires on disease risk factors every two years. The study protocol was approved by the Institutional Review Boards of the Brigham and Women’s Hospital and the Harvard T.H. Chan School of Public Health

Childhood Body Fatness

In 1988 (NHS and HPFS) and 1989 (NHSII), men and women were asked to select the figure from a validated nine-figure drawing (Figure 1) which best corresponded to their body fatness at age 5, 10 and 20 (17, 18). The correlations between recalled body fatness and BMI assessed at similar ages were moderate (r=0.53 to r=0.75), however correlations were lowest (r=0.36) among males at age 5. Childhood body fatness is calculated as the average of figures selected for ages 5 and 10 and the distribution is provided in Suppl Table 1.

Figure 1:

Figure 1:

Open in a new tab

Childhood and Adolescent Body Size Pictogram

GWAS data and imputation

The pipeline for generating, cleaning, harmonizing and imputing GWAS data in NHS, NHSII and HPFS has been described previously (19). Briefly, subjects were genotyped as part of (12+HumanCore+OncoArray) separate outcomes including cancer, diabetes, coronary heart disease, mammographic density among others. GWAS data was merged by platform to maximize SNP content and minimize errors due to platform differences. We created five different GWAS datasets: Illumina HumanHap, Illumina OmniExpress, Illumina HumanCore, Illumina OncoArray and Affymetrix 6.0. We applied standard quality control filters. We excluded variants or samples using the following criteria: 1) variants with all rate <95%, 2) variants with HWE p-value < 1x10-6; 3) Samples with call rate < 90%; 4) Samples with gender discordance; 5) Samples with extreme heterozygosity rate; 6) European ancestry outliers (identified by principal components analysis). The number of subjects included in our analyses by platform can be found in Suppl Table 2. Each dataset was separately imputed using the phase 3 v5 1000 Genome reference panel (20).

GWAS analysis

Sex- and platform-stratified linear regression was applied for the GWAS analysis with package RVtest (21), adjusting for the top four principal components to account for the population structure. We selected four principal components because in a GWAS of hair color in NHS, adjusting for the top four PCs eliminated most of the apparent residual confounding due to population stratification; further control for up to 50 PCs did not alter the genomic inflation factor (22). This is also confirmed by visual inspection of the PC plots, which indicates no structure beyond the fourth PC in our data and Tracy-Widom tests (23). Fixed-effect meta-analysis across the five platforms and sex was conducted to explore the overall genetic effect on child body fatness. METAL (24) was used for the meta-analysis. The heterogeneity across sex was tested by Cochran’s Q statistics. LD score regression (25) was applied to test the bias vs. polygenic signals in observed inflation in test statistics.

Replication analysis

5,348 SNPs was associated with childhood body fatness with p<10−4. We set out to replicate these suggestive SNPs in the EGG consortium which has made their genome-wide summary statistics from a GWAS of childhood BMI publicly available. Data on the childhood BMI trait has been contributed by the EGG Consortium and has been downloaded from www.egg-consortium.org. Since the EGG GWAS was imputed to HapMap 2 and all coordinates were in hg18, we first lifted over their GWAS to hg19 using liftOver from Bioconductor R package rtracklayer. Out of the 5,348 SNPs with suggestive findings, 1,682 (31%) were also found in EGG data. We then conducted a meta-analysis using sample size based approach in METAL, whereby a weighted sum of Z-scores for each allele is generated across studies. The weights are proportional to the square-root of the sample size for each study (24, 26). 709 SNPs reached genome-wide significance in the meta-analysis. These SNPs correspond to 19 independent regions at least 500kb apart.

TWAS

We conducted a TWAS using SNP weights from three reference panels as described previously.(16) Gene-specific SNP weights were downloaded from http://gusevlab.org/projects/fusion/. Briefly, the adipose tissue reference panel was generated from the Metabolic Syndrome in Men (METSIM; n = 563) and the two blood reference panels were generated from the Young Finns Study (YFS; n = 1,264) and the Netherlands Twins Registry (NTR; n = 1,247). We used a significance threshold of p<4.24x10−6 which corresponds to p=0.05/11,787 tests (genes) across the three target tissues.

Genetic Correlation analysis

We used LDHub (27, 28) to estimate the genetic correlation between childhood body fatness, adult BMI and other traits. For adult BMI, we used the GIANT BMI statistics from Locke et al. (29) and downloaded from https://portals.broadinstitute.org/collaboration/giant/index.php/GIANT_consortium_data_files

Results

Genome-Wide Association Study and Meta-Analysis

In total, we assessed the association between childhood body fatness and 10,031,233 SNPs and indels using data on 34,401 participants of NHS, NHSII and HPFS. We observed an inflation in association statistics λ=1.10 but further analyses using LD score regression analyses (25) suggest that this inflation is mainly due to a polygenic signal (intercept: 1.02). In total, we discovered eleven regions that reached genome-wide significance (Figure 2), of which, one region at 20p13, had not been associated with childhood obesity before. One region (3p12.1) was genome-wide significant in the discovery GWAS but was not replicated in EGG (p>0.5) and thus, not further pursued. We conducted in silico look ups for 1,682 SNPs with p-values less than 10−4 and that were also present in the publicly available GWAS summary statistics of childhood BMI from the EGG consortium. We meta-analyzed results from our study and from those of the EGG consortium. In this meta-analysis, 19 regions reached genome-wide significance (Table 1), five of which were not previously associated with body fatness or BMI in childhood (rs17637930 in FHIT, rs2378662 in LOC101927575, rs962369 in BDNF, rs10151686 in PRKD1 and rs650759 at 20p13). Regional plots for these loci are displayed in Figure 3.

Figure 2:

Figure 2:

Open in a new tab

Manhattan plot of the childhood body fatness GWAS

Table 1.

Genome wide significant (combined P value 5<10−8) loci associated with childhood body fatness in joint analysis, showing the most significant SNP from each locus

Top SNP Position1 A1 A2 A1 freq Imputation
r2 		Z-score P-value Discovery2 Replication3 Previous4
SNP		 LD Closest Gene Other GWAS associations
rs1421085 16:53800954 T C 0.59 0.9875 −12.975 1.70E-38 6.28E-21 3.20E-19 rs1421085 1 FTO (intronic) Childhood/adult BMI, age at menarche, hip circumference, breast cancer
rs11676272 2:25141538 A G 0.53 0.98472 −12.279 1.17E-34 5.45E-14 8.55E-23 rs11676272 1 ADCY3 (missense variant) Adult/childhood BMI, adult weight circumference, CD/IBD, breast cancer etc.
rs6728726 2:623976 T C 0.18 0.99759 −11.527 9.63E-31 2.07E-12 1.96E-20 rs4854349 0.96 44 kb upstream of TMEM18B Adult/childhood BMI, Type II Diabetes,
rs7132908 12:50263148 A G 0.39 0.98342 11.3 1.31E-29 9.31E-13 4.99E-19 rs7132908 1 FAIM2 (3' UTR) Childhood/adult BMI, age at menarche, hip circumference
rs543874 1:177889480 A G 0.81 0.99884 −10.209 1.81E-24 2.82E-09 2.38E-17 rs543874 1 8 kb upstream of SEC16B Childhood/adult BMI, age at menarche, hip circumference, body mass percentage
rs12041852 1:75003500 A G 0.57 0.99224 −9.774 1.45E-22 1.16E-13 1.77E-10 rs12041852 1 TNNI3K (intronic) Childhood/adult BMI
rs11662368 18:57757978 A G 0.75 0.99619 −8.911 5.04E-19 1.37E-09 6.14E-11 rs6567160 0.78 Intergenic (MC4R) Childhood/adult BMI, waist/hip circumference
rs987237 6:50803050 A G 0.82 0.99165 −8.637 5.77E-18 8.14E-07 3.81E-13 rs987237 1 TFAP2B (intronic) Childhood/adult BMI, waist/hip circumference
rs13130484 4:45175691 T C 0.43 0.99637 8.589 8.81E-18 1.56E-08 8.94E-11 rs13130484 1 Intergenic Childhood/adult BMI, age at menarche, hip circumference
rs8046312 16:19979334 A C 0.81 0.93322 7.686 1.52E-14 2.92E-06 4.06E-10 rs8046312 1 Intergenic Adult BMI, age at menarche. Waist/hip circumference
rs7550711 1:110082886 T C 0.03 0.98089 7.591 3.17E-14 3.80E-07 1.50E-08 rs7550711 1 GPR61 (intronic) Childhood/adult BMI
rs3101336 1:72751185 T C 0.37 0.99329 −7.48 7.46E-14 3.58E-07 4.14E-08 rs3101336 1 AL513166.1 (Intronic) Childhood extreme obesity, Adult BMI
rs962369 11:27734420 T C 0.72 0.97219 −7.169 7.58E-13 1.76E-07 8.75E-07 None   BDNF (intronic) Adult BMI, waist circumference, diastolic blood pressure
rs13107325 4:103188709 T C 0.07 0.96301 7.14 9.32E-13 1.14E-05 1.19E-08 rs13107325 1 SLC39A8 (missense) Childhood/adult BMI, waist/hip ratio
rs13253111 8:28061974 A G 0.55 0.99505 6.935 4.05E-12 6.41E-05 4.13E-09 rs13253111 1 Intergenic Childhood BMI
rs10151686 14:30466466 A G 0.04 0.93411 6.413 1.43E-10 2.07E-05 1.50E-06 None   PRKD1 (intronic) r2=1 with age at menarche SNP (rs10136330), r2=1.0 with adult BMI SNP (rs61980008)
rs650759 20:791957 A C 0.02 0.58798 6.358 2.05E-10 5.60E-10 0.05691 None   22 kb upstream of FAM110A N/A
rs17637930 3:61142441 A G 0.88 0.96916 5.633 1.77E-08 3.16E-05 0.0001466 None   FHIT (intronic) Adult BMI (r2=0.44 rs17668356)
rs2378662 9:86707289 A G 0.54 0.93911 5.529 3.22E-08 7.98E-05 9.93E-05 None   LOC101927575 (intronic) Age at menarche (same SNP)
Open in a new tab
1.

Position is based on NCBI build 37 (UCSC hg19)

2.

Discovery p-values are based on analyses we conducted in 34,401 individuals enrolled in NHS, NHSII, or HPFS cohorts.

3.

Replication p-values are from EGG Consortium GWAS

4.

Previous SNP: SNP in this region that has been previously associated with childhood BMI

Figure 3:

Figure 3:

Open in a new tab

Regional plots of the newly identified genome-wide significant loci. In order to show data on as many SNPs as possible, P-values are based on the discovery GWAS in the Harvard cohorts. a) rs17637930 (FHIT), b) rs2378662 (BX537783), c) rs962369 (BDNF), d) rs10151686 (PRKD1), e) rs650759 (20p13).

Transcriptome-Wide Association Study (TWAS)

We conducted a TWAS (16) to assess if predicted gene expression was associated with childhood body fatness. Results are presented for three gene expression reference panels including two in blood (YFS and NTR) and one in adipose tissue (METSIM). We found a heritable component (p<0.01) for 2,447 (NTR), 4,686 (YFS) and 4,654 (METSIM) genes. Positive TWAS associations suggest that higher expression of the gene is associated with a higher level of the phenotype of interest (16), in this case greater childhood body fatness. The top three genes for each reference tissue are displayed in Table 2. Expression of two genes, POMC at 2p23.3 (p=3.36 x 10−6) on the YFS blood reference panel and TMEM18 at 2p25.3 (p=3.53x10−7) on the METSIM adiposity reference panel crossed our significance threshold of p<4.24x10−6. Both POMC and TMEM18 have been previously linked to childhood BMI (30, 31, 32).

Table 2.

Top TWAS hits according to reference tissue, METSIM (adipose), NTR (blood) and YFS (blood)

Panel GENE Location HSQ TWAS.Z TWAS.P
METSIM          
  TMEM18 2p25.3 0.20 5.09 3.53x10−7
  MAP2K5 15q23 0.07 4.34 1.43x10−5
  EFR3B 2p23.3 0.04 −3.77 1.61x10−4
NTR          
  BANF1 11q13.1 0.03 −4.26 2.02x10−5
  PROK2 3p13 0.06 4.02 5.79x10−5
  CXCR6 3p21.31 0.02 3.92 8.85x10−5
YFS          
  POMC 2p23.3 0.35 −4.65 3.36x10−6
  DDX23 12q.13.12 0.24 −3.89 1.02x10−4
  MAP2K5 15q23 0.05 3.88 1.03x10−4
Open in a new tab

Genetic Correlation Analysis

The observed SNP heritability hSNP2 for childhood body fatness was 0.156 (se: 0.019). We identified several expected genetic correlations between childhood body fatness and adult adiposity traits (Suppl Table 1) including BMI (rg=0.58, p=3.03 x 10−23), class 3 obesity (rg=0.66, p=1.81 x 10−11), hip circumference (rg=0.42, p=4.44 x 10−16) and waist circumference (rg=0.39, p= 5.56 x 10−16). We also observed several genetic correlations with BMI-related biomarkers such as leptin (not adjusted for BMI; rg=0.40, p=1.1 x 10−5) and fasting insulin (rg=0.20, p=0.01), as well as with smoking initation (rg=0.33, p=0.002).We found an inverse genetic correlation between childhood body fatness and age at menarche (rg=−0.37, p= 7.96 x 10−19), and neuroticism (rg=−0.25, p=0.02). Genetic correlations with biomarkers/metabolites tended to be stronger for adult BMI than for childhood body fatness and genetic correlations for adult BMI and childhood body fatness were modest for other disease traits (Suppl Table 3).

Discussion

This study’s goal was to strengthen existing evidence and identify novel loci that may specifically impact adiposity in childhood and adolescence. Our underlying hypothesis was that different loci may impact obesity risk at different times in life, and that the impact of specific loci may vary over the life course. In this GWAS of nearly 37,000 individuals, with joint analysis based on over 72,000 individuals, we identified 19 genome-wide significant loci, five of which had not previously been associated with childhood body fatness or BMI (rs17637930 in FHIT, rs2378662 in LOC101927575, rs962369 in BDNF, rs10151686 in PRKD1 and rs650759 at 20p13). The 3p14.2 lead SNP rs17637930 is an intronic SNP in FHIT and in LD (r2=0.41) with rs1916799, a previously reported adult BMI SNP (33). SNP rs2378662 at 9q21.32 has previously been associated with age at menarche (34). The 11p14.1 lead SNP rs962369 is an intronic SNP in BDNF, in strong LD (r2=0.93) with SNPs associated with adult BMI and waist circumference (35, 36). A SNP in moderate LD (r2=0.5) with rs962369 showed suggestive association with childhood BMI in Felix et al. but did not reach genome-wide significance (p=1.4x10−7) (15). SNP rs10151686 at 14q12 is in perfect LD (r2=1) with rs10136330 (37) and rs61980008 (33) which have been associated with age at menarche, and adult BMI respectively. The 20p13 lead SNP rs650759 is 22kb located upstream of FAM110A. We could not identify any previous genome-wide associations in this region. Top hits in our joint analysis included many of the 15 loci identified in Felix et al., including ADCY3, SEC16B, FAIM2, TMEM18, TNNI3K, and BDNF (15). Our TWAS confirmed previous associations with the POMC and TMEM18 genes, suggesting that these associations might be mediated through gene expression. However, we did not identify any novel regions with our TWAS. In our genetic correlation analyses, most traits were more strongly correlated with adult BMI than with childhood body fatness, and none were correlated with childhood body fatness only.

Except for our completely new finding on 20p13, our other novel SNPs in FHIT, PRKD1 and BDNF were previously associated with either adult BMI or adult waist circumference. Similarly, we identified significant correlations with adult adiposity and related traits suggesting significant overlap in the genetic architecture of adiposity across the life course. While we identified a novel variant in BDNF at rs962369, other SNPs in this gene have been previously associated with childhood (rs11030104) (9) and/or adult adiposity (rs10767664, rs11030104, rs12291063) in European and East-Asian populations (38, 39, 40). Additionally, a different SNP in FHIT (rs2365389) was associated with adult BMI among individuals of European ancestry (29). To date, GWAS meta-analyses have identified over 450 independent variants associated with BMI in adulthood (41). Significantly fewer studies have specifically evaluated adiposity in childhood, yielding 20 independent loci. Of these, just five: rs650759 on 20p13 and rs2378662 on LOC101927575 from the current study, and rs13253111 near ELP3, rs8092503 near RAB27B and rs13387838 near ADAM23 from Felix et al., are not also associated with adult BMI. However, rs8092503 near RAB27B was identified in a recent meta-analysis of adult BMI GWAS, but did not reach genome-wide significance (p= 5 x 10−7) (41). Similarly, Felix et al. reports that rs13253111 and rs13387838 are near two SNPs (rs4319045 and rs972540) that just missed genome-wide significance in an adult BMI GWAS meta-analysis, however due to low linkage disequilibrium they hypothesize these may be independent signals (15). Our genetic correlation analyses demonstrated that our childhood adiposity phenotype was associated with other adiposity-related traits including hip and waist circumference, and waist to hip ratio. Consistent with our genetic correlation findings, there is recent evidence of genetic pleiotropy between childhood BMI and 13 adult traits (BMI, waist-to-hip ratio, waist circumference, type 2 diabetes, fasting plasma insulin, glycated hemoglobin, insulin sensitivity, coronary artery disease, myocardial infarction, low-density lipoprotein (LDL) cholesterol, high-density lipoprotein (HDL) cholesterol, total cholesterol, and triglycerides) (42). This report also showed that 97.1% of SNPs associated with childhood BMI were shared with adult BMI, while 47.9% of SNPs associated with adult BMI overlapped with childhood BMI (42). Taken together, these results demonstrate the shared genetic architecture of adiposity across the life course.

Our study’s strengths include the large sample size, and integration of multiple analyses including GWAS, TWAS and genetic correlation to comprehensively assess the genetic underpinnings of childhood adiposity. However, there are several important limitations to note. Our phenotype is obtained in adulthood using a somatotype pictogram. While this measure has been previously validated (43), it is based on participant recall, and there is opportunity for misclassification. In addition to errors of recall, the images in the figure are more representative of adults than children, and changes in adiposity are also accompanied by changes in body shape (44). However, despite these opportunities for missclassification, our study replicated findings where measured BMI in childhood was used (15). Lastly, our findings are based on a European ancestry population, and may not apply to other continental ancestry populations.

Conclusion

In conclusion, this study identified five novel loci associated with childhood body fatness. It adds to the small, but growing body of work delineating the genetic contributors to obesity in childhood and adolescence. Consistent with previous studies, our results demonstrate the significant overlap between genetic variants associated with childhood and adult and adiposity.

Supplementary Material

TableS1-S3

  • The correlation between childhood and adult body mass index (BMI) is modest (r=0.24-0.30) and it appears the genetic influence on body fatness is higher in childhood.

  • Most previous studies of the genetics of obesity have focused on adiposity in adulthood, with few studies on childhood. To date, 15 independent loci associated with childhood adiposity have been identified compared to over 450 for adult adiposity.

  • Our GWAS of 34,401 European ancestry participants in the Nurses’ Health Studies and Health Professional Follow-Up Study identified five novel loci associated with childhood body fatness rs17637930 in FHIT, rs2378662 in LOC101927575, rs962369 in BDNF, rs10151686 in PRKD1, and rs650759 at 20p13.

  • The BNDF, FHIT and PRKD1 regions were previously associated with adult BMI. LOC101927575 and 20p13 regions have not previously been associated with adiposity phenotypes.

  • We found the correspondence between genetic loci are similar between childhood and adult adiposity.

Acknowledgements

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Funding: The study is supported by funds from the National Cancer Institute (R03CA165131). E.T. Warner is supported by the National Cancer Institute (K01CA188075). The Nurses’ Health Study is supported by NCI grants UM1CA186107 and R01CA49449. The Nurses’ Health Study II is supported by NCI grants UM1CA176726 and R01CA67262. The Health Professionals Follow-up Study is supported by UM1CA167552.

Footnotes

Disclosure: The authors declared no conflict of interest

References

  • 1.Anderson PM, Butcher KF, Schanzenbach DW. Understanding recent trends in childhood obesity in the United States. Econ Hum Biol 2019;34: 16–25. [DOI] [PubMed] [Google Scholar]
  • 2.Haworth CM, Carnell S, Meaburn EL, Davis OS, Plomin R, Wardle J. Increasing heritability of BMI and stronger associations with the FTO gene over childhood. Obesity (Silver Spring) 2008;16: 2663–2668. [DOI] [PubMed] [Google Scholar]
  • 3.Power C, Lake JK, Cole TJ. Body mass index and height from childhood to adulthood in the 1958 British born cohort. Am J Clin Nutr 1997;66: 1094–1101. [DOI] [PubMed] [Google Scholar]
  • 4.Ylihärsilä H, Kajantie E, Osmond C, Forsén T, Barker DJ, Eriksson JG. Body mass index during childhood and adult body composition in men and women aged 56-70 y. Am J Clin Nutr 2008;87: 1769–1775. [DOI] [PubMed] [Google Scholar]
  • 5.Elks CE, den Hoed M, Zhao JH, Sharp SJ, Wareham NJ, Loos RJ, et al. Variability in the heritability of body mass index: a systematic review and meta-regression. Front Endocrinol (Lausanne) 2012;3: 29. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Speliotes EK, Willer CJ, Berndt SI, Monda KL, Thorleifsson G, Jackson AU, et al. Association analyses of 249,796 individuals reveal 18 new loci associated with body mass index. Nat Genet 2010;42: 937–948. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Dina C, Meyre D, Gallina S, Durand E, Körner A, Jacobson P, et al. Variation in FTO contributes to childhood obesity and severe adult obesity. Nat Genet 2007;39: 724–726. [DOI] [PubMed] [Google Scholar]
  • 8.Frayling TM, Timpson NJ, Weedon MN, Zeggini E, Freathy RM, Lindgren CM, 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: 889–894. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Zhao J, Bradfield JP, Zhang H, Sleiman PM, Kim CE, Glessner JT, et al. Role of BMI-associated loci identified in GWAS meta-analyses in the context of common childhood obesity in European Americans. Obesity (Silver Spring) 2011;19: 2436–2439. [DOI] [PubMed] [Google Scholar]
  • 10.Meng XR, Song JY, Ma J, Liu FH, Shang XR, Guo XJ, et al. Association study of childhood obesity with eight genetic variants recently identified by genome-wide association studies. Pediatr Res 2014;76: 310–315. [DOI] [PubMed] [Google Scholar]
  • 11.Jarick I, Vogel CI, Scherag S, Schäfer H, Hebebrand J, Hinney A, et al. Novel common copy number variation for early onset extreme obesity on chromosome 11q11 identified by a genome-wide analysis. Hum Mol Genet 2011;20: 840–852. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Scherag A, Jarick I, Grothe J, Biebermann H, Scherag S, Volckmar AL, et al. Investigation of a genome wide association signal for obesity: synthetic association and haplotype analyses at the melanocortin 4 receptor gene locus. PLoS One 2010;5: e13967. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Mitchell JA, Hakonarson H, Rebbeck TR, Grant SF. Obesity-susceptibility loci and the tails of the pediatric BMI distribution. Obesity (Silver Spring) 2013;21: 1256–1260. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Bradfield JP, Taal HR, Timpson NJ, Scherag A, Lecoeur C, Warrington NM, et al. A genome-wide association meta-analysis identifies new childhood obesity loci. Nat Genet 2012;44: 526–531. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Felix JF, Bradfield JP, Monnereau C, van der Valk RJ, Stergiakouli E, Chesi A, et al. Genome-wide association analysis identifies three new susceptibility loci for childhood body mass index. Hum Mol Genet 2016;25: 389–403. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Gusev A, Ko A, Shi H, Bhatia G, Chung W, Penninx BW, et al. Integrative approaches for large-scale transcriptome-wide association studies. Nat Genet 2016;48: 245–252. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Stunkard AJ, Sorensen T, Schulsinger F. Use of the Danish Adoption Register for the study of obesity and thinness. Research publications - Association for Research in Nervous and Mental Disease 1983;60: 115. [PubMed] [Google Scholar]
  • 18.Koprowski C, Coates RJ, Bernstein L. Ability of young women to recall past body size and age at menarche. Obesity research 2001;9: 478. [DOI] [PubMed] [Google Scholar]
  • 19.Lindström S, Loomis S, Turman C, Huang H, Huang J, Aschard H, et al. A comprehensive survey of genetic variation in 20,691 subjects from four large cohorts. PLoS One 2017;12: e0173997. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Sudmant PH, Rausch T, Gardner EJ, Handsaker RE, Abyzov A, Huddleston J, et al. An integrated map of structural variation in 2,504 human genomes. Nature 2015;526: 75–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Zhan X, Hu Y, Li B, Abecasis GR, Liu DJ. RVTESTS: an efficient and comprehensive tool for rare variant association analysis using sequence data. Bioinformatics 2016;32: 1423–1426. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Han J, Kraft P, Nan H, Guo Q, Chen C, Qureshi A, et al. A genome-wide association study identifies novel alleles associated with hair color and skin pigmentation. PLoS Genet 2008;4: e1000074. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Patterson N, Price AL, Reich D. Population structure and eigenanalysis. PLoS genetics 2006;2: e190. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Willer CJ, Li Y, Abecasis GR. METAL: fast and efficient meta-analysis of genomewide association scans. Bioinformatics 2010;26: 2190–2191. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Bulik-Sullivan BK, Loh PR, Finucane HK, Ripke S, Yang J, Patterson N, et al. LD Score regression distinguishes confounding from polygenicity in genome-wide association studies. Nat Genet 2015;47: 291–295. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Stouffer SA, Suchman EA, De Vinney LC, Star SA, Williams RM. Studies in Social Psychology in World War II. Vol. I: The American Soldier: Adjustment during Army Life; 1951. [Google Scholar]
  • 27.Zheng J, Erzurumluoglu AM, Elsworth BL, Kemp JP, Howe L, Haycock PC, et al. LD Hub: a centralized database and web interface to perform LD score regression that maximizes the potential of summary level GWAS data for SNP heritability and genetic correlation analysis. Bioinformatics 2017;33: 272–279. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Bulik-Sullivan B, Finucane HK, Anttila V, Gusev A, Day FR, Loh PR, et al. An atlas of genetic correlations across human diseases and traits. Nat Genet 2015;47: 1236–1241. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Locke AE, Kahali B, Berndt SI, Justice AE, Pers TH, Day FR, et al. Genetic studies of body mass index yield new insights for obesity biology. Nature 2015;518: 197–206. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Kuehnen P, Mischke M, Wiegand S, Sers C, Horsthemke B, Lau S, et al. An Alu element-associated hypermethylation variant of the POMC gene is associated with childhood obesity. PLoS Genet 2012;8: e1002543. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Mendiratta MS, Yang Y, Balazs AE, Willis AS, Eng CM, Karaviti LP, et al. Early onset obesity and adrenal insufficiency associated with a homozygous POMC mutation. Int J Pediatr Endocrinol 2011;2011: 5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Larder R, Sim MFM, Gulati P, Antrobus R, Tung YCL, Rimmington D, et al. Obesity-associated gene. Proc Natl Acad Sci U S A 2017;114: 9421–9426. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Tachmazidou I, Süveges D, Min JL, Ritchie GRS, Steinberg J, Walter K, et al. Whole-Genome Sequencing Coupled to Imputation Discovers Genetic Signals for Anthropometric Traits. Am J Hum Genet 2017;100: 865–884. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Pickrell JK, Berisa T, Liu JZ, Ségurel L, Tung JY, Hinds DA. Detection and interpretation of shared genetic influences on 42 human traits. Nat Genet 2016;48: 709–717. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Thorleifsson G, Walters GB, Gudbjartsson DF, Steinthorsdottir V, Sulem P, Helgadottir A, et al. Genome-wide association yields new sequence variants at seven loci that associate with measures of obesity. Nat Genet 2009;41: 18–24. [DOI] [PubMed] [Google Scholar]
  • 36.Shungin D, Winkler TW, Croteau-Chonka DC, Ferreira T, Locke AE, Mägi R, et al. New genetic loci link adipose and insulin biology to body fat distribution. Nature 2015;518: 187–196. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Kichaev G, Bhatia G, Loh PR, Gazal S, Burch K, Freund MK, et al. Leveraging Polygenic Functional Enrichment to Improve GWAS Power. Am J Hum Genet 2019;104: 65–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Okada Y, Kubo M, Ohmiya H, Takahashi A, Kumasaka N, Hosono N, et al. Common variants at CDKAL1 and KLF9 are associated with body mass index in east Asian populations. Nat Genet 2012;44: 302–306. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Wen W, Zheng W, Okada Y, Takeuchi F, Tabara Y, Hwang JY, et al. Meta-analysis of genome-wide association studies in East Asian-ancestry populations identifies four new loci for body mass index. Hum Mol Genet 2014;23: 5492–5504. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Mou Z, Hyde TM, Lipska BK, Martinowich K, Wei P, Ong C-J, et al. Human obesity associated with an intronic SNP in the brain-derived neurotrophic factor locus. Cell reports 2015;13: 1073–1080. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Hoffmann TJ, Choquet H, Yin J, Banda Y, Kvale MN, Glymour M, et al. A Large Multiethnic Genome-Wide Association Study of Adult Body Mass Index Identifies Novel Loci. Genetics 2018;210: 499–515. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Tekola-Ayele F, Lee A, Workalemahu T, Sánchez-Pozos K. Shared genetic underpinnings of childhood obesity and adult cardiometabolic diseases. Hum Genomics 2019;13: 17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Must A, Willett WC, Dietz WH. Remote recall of childhood height, weight, and body build by elderly subjects. American Journal of Epidemiology 1993;138: 56. [DOI] [PubMed] [Google Scholar]
  • 44.Magnusson C, Roddam AW. Breast cancer and childhood anthropometry: emerging hypotheses. Breast Cancer Res 2005;7: 83. [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

TableS1-S3
NIHMS1640815-supplement-TableS1-S3.xlsx (34.5KB, xlsx)

RESOURCES