Abstract
OBJECTIVE
We tested genetic associations with weight loss and weight regain in the Diabetes Prevention Program, a randomized controlled trial of weight loss–inducing interventions (lifestyle and metformin) versus placebo.
RESEARCH DESIGN AND METHODS
Sixteen obesity-predisposing single nucleotide polymorphisms (SNPs) were tested for association with short-term (baseline to 6 months) and long-term (baseline to 2 years) weight loss and weight regain (6 months to study end).
RESULTS
Irrespective of treatment, the Ala12 allele at PPARG associated with short- and long-term weight loss (−0.63 and −0.93 kg/allele, P ≤ 0.005, respectively). Gene–treatment interactions were observed for short-term (LYPLAL1 rs2605100, Plifestyle*SNP = 0.032; GNPDA2 rs10938397, Plifestyle*SNP = 0.016; MTCH2 rs10838738, Plifestyle*SNP = 0.022) and long-term (NEGR1 rs2815752, Pmetformin*SNP = 0.028; FTO rs9939609, Plifestyle*SNP = 0.044) weight loss. Three of 16 SNPs were associated with weight regain (NEGR1 rs2815752, BDNF rs6265, PPARG rs1801282), irrespective of treatment. TMEM18 rs6548238 and KTCD15 rs29941 showed treatment-specific effects (Plifestyle*SNP < 0.05).
CONCLUSIONS
Genetic information may help identify people who require additional support to maintain reduced weight after clinical intervention.
Multiple obesity-predisposing gene variants are known (1,2), which may interact with lifestyle to modify obesity risk (3). It is unknown whether these variants influence weight regain (WR) after intentional weight loss (WL). We therefore tested associations of 16 obesity-predisposing variants with weight change in Diabetes Prevention Program (DPP) participants.
RESEARCH DESIGN AND METHODS
The DPP is described elsewhere (4,5). In brief, 3,234 overweight/obese adults with impaired glucose tolerance were randomly assigned to placebo, 850 mg metformin twice daily, or intensive lifestyle modification aimed at ∼150 min of physical activity per week and ∼7% WL, to compare effects on diabetes incidence. Participants provided written informed consent, and institutional review boards of 27 DPP study centers approved the study.
Participants
Of 3,597 participants with baseline and 1-year data available, 93.3% consented to genetic analyses. Of these, 56.1% were non-Hispanic white (NHW), 20.4% were African American, 16.7% were Hispanic, 4.4% were Asian American, and 2.5% were American Indian; on average participants were middle-aged and obese (see Supplementary Table 1 for participant characteristics).
Genotyping
Sixteen obesity-predisposing variants reported elsewhere (1,2) or in the DPP (6) were genotyped as described previously (6); genotyping success rates exceeded 99% (Supplementary Table 2).
Statistical analysis
Analyses were performed using SAS v9.2 (Cary, NC). Predictor variables were single nucleotide polymorphisms (SNPs), with effect alleles coded consistent with the association of each SNP with BMI or waist circumference in published meta-analyses (Supplementary Table 3) (2). Genetic risk scores were constructed by summing effect alleles (see Supplementary Data) (7). Models are annotated in Supplementary Data.
Primary end points are 1) short-term WL (baseline to 6 months), 2) long-term WL (baseline to 2 years), and 3) average rate of WR (6 months to study end (range 2–4.5 years). WL analyses included all participants, whereas WR analyses included 1,411 participants who had achieved ≥3% WL at 6 months. Analyses were conducted in the pooled sample adjusting for self-reported ethnicity; sensitivity analyses were repeated in NHW only to rule out population stratification. Unless there was statistical evidence of gene x treatment interactions, data were pooled from the three study arms and models were adjusted for age, sex, ethnicity, treatment, and baseline value for the dependent variable. Where such interactions were observed, treatment-specific genetic effects were estimated. For general linear models assuming additive allele effects (except Pro12Ala, which was coded with Pro12Pro vs. Ala12×), nominal two-sided P values are reported. All P values for the same outcome are adjusted for multiple comparisons, and significant SNP effects are reported (8): for short-term WL, there are three significant SNP*treatment interactions (13 + (3*3) = 22 tests are corrected for); for long-term WL, there are two significant interactions (14 + (3*2) = 20 tests are corrected for); and for WR, there are six significant interactions (10 + (6*3) = 28 tests are adjusted for).
RESULTS
Baseline data are reported in Supplementary Tables 1 and 4. P values in Table 1 are obtained from the regressions; however, only SNPs that remain statistically significant after adjusting for multiple comparisons are reported in this section.
Table 1.
Summary of association data for each of 16 known obesity loci and short-term (6 months) change in weight (kilograms per year), long-term (2 years) change in weight (kilograms per year), and rate of WR from 6 months through trial end (kilograms per year) in the overall DPP cohort and each treatment arm if significant SNP treatment interactions were detected
WL
Short- and long-term WL were greatest in the lifestyle intervention group, and both lifestyle and metformin groups had significantly greater WL than the placebo (control) group (4,5). Irrespective of treatment, the minor Ala12 allele at PPARG was associated with short- and long-term WL (Table 1). Statistically significant gene-lifestyle interactions were observed for short-term (LYPLAL1 rs2605100; GNPDA2 rs10938397; MTCH2 rs10838738) and long-term (NEGR1 rs2815752; FTO rs9939609) WL (Pinteraction< 0.05).
WR
The rate of WR (in kilograms per year) from 6 months to study end was greatest in the lifestyle group and least in the placebo group (Supplementary Table 1). Those who lost ≥3% body weight from baseline to 6 months had a mean (SD) WR of 0.94 (±4.68) kg/year. Three of 16 SNPs were associated with WR (NEGR1 rs2815752, BDNF rs6265, PPARG rs1801282), irrespective of treatment. TMEM18 rs6548238 and KTCD15 rs29941 showed treatment-specific effects. In aggregate, the risk alleles associated with WR associated with faster WR (0.274 kg/year/allele [SE = 0.097]; P = 0.005), whereas these alleles had no detectable impact on WR in the control group (Supplementary Fig. 1).
Sensitivity analyses performed in NHW participants, who are essentially free of admixture (9), yielded effect estimates of comparable magnitude, indicating that population stratification does not confound our findings (Supplementary Table 5).
Mediator analyses
Analyses were also performed assessing putative mediating roles of specific lifestyle factors (details in Supplementary Data). However, none explained a statistically significant amount of variance in the SNP-phenotype relationships.
CONCLUSIONS
This is, to our knowledge, the first report of effects of validated obesity-predisposing genotypes on long-term WR after successful intentional WL. We found that three SNPs (NEGR1 rs2815752, BDNF rs6265, PPARG rs1801282) predicted WR, irrespective of type of WL therapy, two of which (BDNF rs6265, PPARG Pro12Ala) were robust to correction for multiple hypothesis testing. Two other variants (TMEM18 rs6548238, KTCD15 rs29941) interacted with treatment modality to influence WR. We also replicated several associations reported previously with baseline obesity metrics (Supplementary Table 4).
Our WR findings are perhaps most clinically relevant, since these might help target susceptible individuals and thus improve long-term effects of WL interventions. One of few published genetic association studies on WR found that the minor allele at the Pro12Ala locus associated with greater regain 1 year after a 6-month hypocaloric diet intervention ended, also noting reductions in the rate of fat oxidation in Ala12 allele carriers but not in Pro12 homozygotes (10), findings supporting those reported here. A second small study (11) found a similar association between the Pro12Ala genotype and WR, whereas a third small study (12) reported no effect.
Despite plausible mechanisms by which the genetic effects reported here are expressed, we were unable to detect statistically significant mediators. Although the genetic variants we studied may act independently of the selected mediators, it is also possible some of our findings are false-negative, owing to the relatively small sample and indirect measures. Our study is also limited by the absence of information on lifestyle behaviors after 12 months of intervention, which may have hampered the detection of mediators. Finally, the hyperglycemic nature of the DPP cohort may limit generalizability of our findings.
In summary, our findings offer novel insights into the mechanisms influencing the propensity for WR after intentional WL. This information may help target individuals who require additional support to maintain reduced weight in intervention settings.
Acknowledgments
This work was funded by R01 DK-072041-02 to J.C.F., K.A.J., and A.S. (P.W.F. and W.C.K. are coinvestigators). P.W.F. was supported by grants from Novo Nordisk, the Swedish Research Council, the Swedish Heart-Lung Foundation, and the Swedish Diabetes Association. S.E.K. is supported in part by the Department of Veterans Affairs. J.C.F. is supported by a Doris Duke Charitable Foundation Clinical Scientist Development Award. The National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) of the National Institutes of Health funded the clinical centers and the Coordinating Center for the design and conduct of the study and the collection, management, analysis, and interpretation of the data. The Southwestern American Indian Centers were supported directly by the NIDDK and the Indian Health Service. The General Clinical Research Center Program, National Center for Research Resources, supported data collection at many of the clinical centers. The Office of Research on Minority Health, the National Institute of Child Health and Human Development, the National Institute on Aging, the Centers for Disease Control and Prevention, Office of Research on Women's Health, the Department of Veterans Affairs, and the American Diabetes Association funded data collection and provided participant support. This research was also supported, in part, by the intramural research program of the NIDDK. The Henry M. Jackson Foundation provided support services under subcontract with the Coordinating Center.
LifeScan, Health O Meter, Hoechst Marion Roussel, Merck-Medco Managed Care, Merck and Co., Nike Sports Marketing, Slim Fast Foods Co., and Quaker Oats Co. donated materials, equipment, or medicines for concomitant conditions. Bristol-Myers Squibb and Parke-Davis provided medication. McKesson BioServices Corporation and Matthews Media Group provided support services under subcontract with the Coordinating Center. No other potential conflicts of interest relevant to this article were reported.
L.M.D. conceived the analysis, designed the analysis plan, interpreted the results, wrote the manuscript, and provided critical input on the manuscript revisions. Q.P. designed the analysis plan, conducted the statistical analyses, interpreted the results, wrote the manuscript, and provided critical input on the manuscript revisions. K.A.J. designed the analysis plan, conducted the statistical analyses, and provided critical input on the manuscript revisions. K.E.W., J.M.M., and A.S. provided critical input on the manuscript revisions. S.E.K. and W.C.K. conducted the clinical trial, provided the phenotypic data, and provided critical input on the manuscript revisions. J.C.F. designed the analysis plan, coordinated the genotyping, and provided critical input on the manuscript revisions. P.W.F. conceived the analysis, designed the analysis plan, interpreted the results, wrote the manuscript, and provided critical input on the manuscript revisions and is the guarantor of this article.
The investigators acknowledge the commitment and dedication of all participants in the Diabetes Prevention Program Research Group, without whom this work would not have been possible.
Footnotes
Clinical trial reg. no. NCT00004992, clinicaltrials.gov.
This article contains Supplementary Data online at http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc11-1328/-/DC1.
*A complete list of centers, investigators, staff, additional methods, results, and list of Diabetes Prevention Program Research Group investigators (Genetics version) can be found in the Supplementary Data online.
The opinions expressed are those of the investigators and do not necessarily reflect the views of the Indian Health Service or other funding agencies.
References
- 1.Lindgren CM, Heid IM, Randall JC, et al. ; Wellcome Trust Case Control Consortium; Procardis Consortia; Giant Consortium Genome-wide association scan meta-analysis identifies three Loci influencing adiposity and fat distribution. PLoS Genet 2009;5:e1000508. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Willer CJ, Speliotes EK, Loos RJ, et al. ; Wellcome Trust Case Control Consortium; Genetic Investigation of ANthropometric Traits Consortium Six new loci associated with body mass index highlight a neuronal influence on body weight regulation. Nat Genet 2009;41:25–34 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Li S, Zhao JH, Luan J, et al. Physical activity attenuates the genetic predisposition to obesity in 20,000 men and women from EPIC-Norfolk prospective population study. PLoS Med 2010;7:e1000332. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Knowler WC, Barrett-Connor E, Fowler SE, et al. ; Diabetes Prevention Program Research Group Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med 2002;346:393–403 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Rubin RR, Fujimoto WY, Marrero DG, et al. ; DPP Research Group The Diabetes Prevention Program: recruitment methods and results. Control Clin Trials 2002;23:157–171 [DOI] [PubMed] [Google Scholar]
- 6.Franks PW, Jablonski KA, Delahanty L, et al. ; Diabetes Prevention Program Research Group The Pro12Ala variant at the peroxisome proliferator-activated receptor gamma gene and change in obesity-related traits in the Diabetes Prevention Program. Diabetologia 2007;50:2451–2460 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Hivert MF, Jablonski KA, Perreault L, et al. ; DIAGRAM Consortium; Diabetes Prevention Program Research Group Updated genetic score based on 34 confirmed type 2 diabetes Loci is associated with diabetes incidence and regression to normoglycemia in the diabetes prevention program. Diabetes 2011;60:1340–1348 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Holm S. A simple sequentially rejective multiple test procedure. Scand J Statist 1979;6:65–70 [Google Scholar]
- 9.Jablonski KA, McAteer JB, de Bakker PI, et al. ; Diabetes Prevention Program Research Group Common variants in 40 genes assessed for diabetes incidence and response to metformin and lifestyle intervention in the diabetes prevention program. Diabetes 2010;59:2672–2681 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Nicklas BJ, van Rossum EF, Berman DM, Ryan AS, Dennis KE, Shuldiner AR. Genetic variation in the peroxisome proliferator-activated receptor-γ2 gene (Pro12Ala) affects metabolic responses to weight loss and subsequent weight regain. Diabetes 2001;50:2172–2176 [DOI] [PubMed] [Google Scholar]
- 11.Goyenechea E, Dolores Parra M, Alfredo Martínez J. Weight regain after slimming induced by an energy-restricted diet depends on interleukin-6 and peroxisome-proliferator-activated-receptor-gamma2 gene polymorphisms. Br J Nutr 2006;96:965–972 [DOI] [PubMed] [Google Scholar]
- 12.Vogels N, Mariman EC, Bouwman FG, Kester AD, Diepvens K, Westerterp-Plantenga MS. Relation of weight maintenance and dietary restraint to peroxisome proliferator-activated receptor gamma2, glucocorticoid receptor, and ciliary neurotrophic factor polymorphisms. Am J Clin Nutr 2005;82:740–746 [DOI] [PubMed] [Google Scholar]