SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Objective

To assess associations and genotype × treatment interactions for melanocortin 4 receptor (MC4R) locus variants and obesity-related traits.

Design and Methods

Diabetes prevention program (DPP) participants (N = 3,819, of whom 3,356 were genotyped for baseline and 3,234 for longitudinal analyses) were randomized into intensive lifestyle modification (diet, exercise, weight loss), metformin or placebo control. Adiposity was assessed in a subgroup (n = 909) using computed tomography. All analyses were adjusted for age, sex, ethnicity and treatment.

Results

The rs1943218 minor allele was nominally associated with short-term (6 month; P = 0.032) and long-term (2 year; P = 0.038) weight change. Eight SNPs modified response to treatment on short-term (rs17066856, rs9966412, rs17066859, rs8091237, rs17066866, rs7240064) or long-term (rs12970134, rs17066866) reduction in body weight, or diabetes incidence (rs17066829) (all Pinteraction < 0.05).

Conclusion

This is the first study to comprehensively assess the role of MC4R variants and weight regulation in a weight loss intervention trial. One MC4R variant was directly associated with obesity-related traits or diabetes; numerous other variants appear to influence body weight and diabetes risk by modifying the protective effects of the DPP interventions.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Obesity is a highly heritable [1], complex trait presumed to originate from interactions between multiple genetic and environment risk factors. Several well-defined genetic causes of monogenic obesity are known [2], as are multiple low-penetrance obesogenic variants that are prevalent in the general population [3].

One of the loci implicated in both monogenic and common forms of obesity is the melanocortin 4 receptor (encoded by MC4R). Mutations in MC4R are the most frequent known genetic causes of congenital obesity [4] and a variant mapping to a region upstream of MC4R is strongly associated with variations in BMI and adiposity in multiple populations [5]. MC4R variants have also been associated with energy balance behaviors [6, 7], blood pressure [8], dyslipidemia [9], stature [4], and type 2 diabetes [10].

Two important risk factors for obesity are physical inactivity and excessive caloric intake. Thus, studies focusing on the interactions between obesity-predisposing genes and therapies that influence body weight may help elucidate the molecular mechanisms underlying obesity and its consequences, such as cardiovascular disease, certain cancers, osteoarthritis, and type 2 diabetes. In this regard, lifestyle interventions focused on diet, exercise, and pharmacologic interventions that influence weight loss (e.g., metformin) are relevant.

The purpose of this study was to extend earlier studies that center on MC4R variants and BMI by examining whether MC4R variation per se is associated with obesity-related traits over time or type 2 diabetes incidence. We tagged the entire MC4R region and examined whether associations with obesity-related traits are mediated by dietary preference or physical activity or modified by metformin treatment or lifestyle intervention. All analyses were performed within the Diabetes Prevention Program (DPP), a randomized clinical trial of intensive lifestyle modification, metformin or troglitazone therapy, or placebo intervention.

Methods and Procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

The Diabetes Prevention Program (DPP)

The DPP was a multicenter randomized clinical trial. The study design and details of the preventive interventions have been described in detail previously [11-13]. Briefly, nondiabetic, overweight persons (n = 3,234) with raised fasting and postchallenge glucoses were randomized to placebo, metformin (850 mg twice daily), troglitazone (400 mg daily), or a program of intensive lifestyle modification (aimed at ∼7% weight loss and ∼150 min of physical activity per week). A further 585 participants were randomized to receive troglitazone treatment; owing to liver toxicity, this aspect of the DPP trial was terminated early and only the baseline results were used in these analyses [11]. Thus, a total of 3,819 participants were included in the baseline analyses reported here, whereas 3,234 are included in the prospective analyses. The principal endpoint in the DPP was the development of diabetes by confirmed oral glucose tolerance testing (OGTT). Other phenotypes, such as changes in weight, waist circumference, abdominal fat area, lipids, insulin and glucose, were also collected. The primary focus of this report is on the obesity-related traits; the secondary focus is on diabetes incidence. Each of the 27 DPP centers obtained approval from their respective institutional review boards prior to initiation of the study protocol.

Consent for genetic analyses was obtained in 87.9% (n = 3,356/3,819) of participants who were originally assigned to a treatment arm in the DPP. Of the participants studied in this report, by self-report 56.1% were white, 20.4% African American, 16.7% Hispanic, 4.4% Asian American and 2.5% American Indian. Concordant with the entire DPP, the participants' mean age at enrollment was 51 (SD = 11) years and mean body mass index (BMI) was 34 (SD = 7) kg m−2 (see 1).

Quantification of energy intake, physical activity, and body composition

In-person interviews were conducted to obtain data on calorie intake, percent of calories from fat and protein, absolute intake of fat and protein and food preference behaviors at baseline and 1 year later using the semiquantitative DPP food frequency questionnaire (FFQ) [14]. To evaluate food preference behaviors, we assessed the frequency of consumption of snacks, sweets and desserts, vegetables and fruits, expressed as servings per day. Occupational and leisure-time physical activity (LTPA) was assessed using the validated modifiable activity questionnaire (MAQ) at baseline and annually thereafter [15]. Weight and BMI were measured in all participants.

Computed tomography (CT) was performed at baseline in 909 randomly selected subjects DPP participants, among whom 718 underwent a repeated measure at 12-months postrandomization to quantify subcutaneous and visceral fat areas (at the level of the lumbar 2–3 disc space) [16] and gave consent for genetic analyses. Among the participants with both baseline and 1-year CT measurements, 241 were from the intensive lifestyle group, 246 from the metformin group and 231 from the placebo group.

Genotyping

SNPs were selected through multimarker tagging of the MC4R locus [as previously described [17]]. Genotyping was performed by allele-specific primer extension of multiplex amplified products and detection with matrix-assisted laser desorption ionization-time of flight mass spectroscopy on a Sequenom iPLEX platform. The genotyping success rates for the SNPs studied here were >99%. All 20 SNPs were in Hardy–Weinberg equilibrium in all five ethnic groups (P > 0.05). For each SNP, genomic location, alleles and their frequencies are reported in Supporting Information Table 1. Supporting Information Figure 1a–e shows the LD plots for these variants.

Table 1. Characteristics of DPP study participants included in these analyses (N = 3,891)
Baseline characteristicLifestyleMetforminPlaceboTroglitazoneP value
  1. Data are percentage or Mean (SD), stratified by treatment assignment. P values from ANOVA F-tests and Chi-Square test of independence. LTPA: leisure-time physical activity. The sample size (N = 3,819) is the maximum number used in baseline analyses and includes participants who were subsequently randomized to troglitazone treatment, in addition to those randomized to lifestyle, metformin, or placebo. Sample sizes for: short-term weight loss (n = 3,085); long-term weight loss (n = 3,015); diabetes risk (n = 3,234). We include individuals subsequently randomized to troglitazone in the baseline analyses to maximize statistical power. However, these individuals are not included in the prospective analyses as they represent a relatively small subgroup of the DPP population, which is likely underpowered for treatment stratified analyses.

  2. a

    The troglitazone arm was terminated early [11] and, by consequence, only baseline data are used in this article.

N107910731082585 
Male32%34%31%34%0.404
Age (years)51 (11)51 (10)50 (10)50 (10)0.469
Weight (kg)94 (21)94 (20)94 (20)93 (19)0.774
BMI (kg m−2)34 (7)34 (7)34 (7)34 (6)0.311
Waist circumference (cm)105 (15)105 (14)105 (14)104 (14)0.605
SAT L2-3 (cm2)308 (125)301 (129)314 (127) 0.449
SAT L4-5 (cm2)443 (152)433 (154)449 (151) 0.408
VAT L2-3 (cm2)200 (85)195 (83)199 (90) 0.701
VAT L4-5 (cm2)163 (65)158 (65)159 (65) 0.500
LTPA (MET-hrs/wk)16 (22)16 (26)17 (29)15 (16)0.390
Energy intake (kcal day−1)2137 (1071)2144 (986)2098 (1052)2118 (989)0.739
Short-term (6 m) weight loss (kg)6.7 (5.6)2.3 (4.0)0.3 (4.0) <0.001
Long-term (2 y) weight loss (kg)5.4 (7.6)2.1 (5.7)0.0 (5.5) <0.001
Diabetes incidence per 1,000 p-yr508010074<0.001

Statistical analysis

Analyses were performed using the SAS software (SAS Institute, Cary, NC). Weight loss and abdominal fat area follow-up data were normally distributed and linear regressions were employed for these outcomes. SNP associations with diabetes incidence were assessed with Cox proportional hazards models. Potential mediating effects of leisure time physical activity (LTPA) and energy intake were tested by comparing the SNP regression coefficients with and without adjustment for the potential mediators [17]. All models were adjusted for genotype, intervention, baseline value of the dependent variable, age at randomization, sex, and self-reported ethnicity. Further adjustment for study center made no material difference to the results (data not shown) and is thus not included in the current models. To determine whether it was appropriate to combine data from different treatment and ethnicity groups, we tested genotype × treatment and genotype × ethnicity interactions. If these tests were statistically significant (P < 0.05), we stratified by group; otherwise data were pooled and adjusted for treatment and/or ethnicity. For all variants, we analyzed additive genetic models. Two-sided P values without multiplicity adjustment are reported.

MC4R is a strong candidate gene for obesity-related traits [4]; thus, the prior probability that variants at this locus are associated with obesity traits is likely to be above the null. Moreover, although the DPP is the largest randomized controlled trial of intensive lifestyle modification, it is modestly powered for most of the tests performed here (see Ref. [18] for description of the statistical power characteristics of the DPP). Thus, correcting for multiple testing is likely to increase type 2 error (false-negative) rates. Nonetheless, we are conscious that many readers will be keen to compare the nominal test statistics with those corrected for multiple testing. After considering the linkage disequilibrium between the 20 SNPs, we determined that there are 17.46 independent SNP tests [19]. Therefore, instead of the Bonferroni cutoff of 0.05, a cutoff of 0.05/17.46 = 0.003 is used to determine a conservative estimate of statistical significance for the results reported in this article.

Exploratory analyses were performed to determine whether any of the statistically significant genetic effects detected in the main analyses were likely to be mediated by specific lifestyle factors. The following factors were considered potential intermediate factors (mediators) of the observed genetic effects: leisure time physical activity in MET-hrs per week (metabolic equivalents), total caloric intake (kcal/day), percent of calories from fat, percent of calories from protein, fat intake (g/day), protein intake (g/day), snacking (number of snacks/day), sweet and dessert (servings/day), vegetable consumption (servings/day) and fruit consumption (servings/day).

While testing each candidate mediator, two models were fit, one with the candidate mediator as a covariate and the other without. From each pair of models, the difference in the two SNP regression coefficients and the standard errors of these differences were calculated [18]. In these analyses, a statistically significant nonzero difference represents a discrepancy between the marginal SNP effect estimate and the SNP effect estimate that is conditional on the putative mediator, thus validating the intermediate function of the mediator from genotype to the outcome.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Participant characteristics

Participant characteristics for the DPP cohort used for genetic analyses are reported elsewhere [17]. Table 1 reports anthropometric traits, energy intake, and physical activity at baseline, as well as the weight loss outcomes.

Table 2 shows results from the analysis focused on the entire MC4R region; the table shows regression coefficients (representing the average difference in the dependent variable for each additional copy of the minor allele) and P values for all 20 MC4R SNPs and baseline weight and BMI in DPP participants. After adjustment for age, sex, and self-reported ethnicity, no SNP was statistically associated with baseline weight or BMI.

Table 2. Association between MC4R SNPs with baseline weight (kg) and BMI (kg m−2) (N = 3,346–3,356)
SNPMinor (major) alleleWeight (kg/allele)BMI (kg/m2/allele)
Coefficient (SE)P valueCoefficient (SE)P value
  1. Coefficients, standard errors (SE), and P values correspond to the additive allele effects adjusted for age at randomization, sex and ethnicity. The reported P values are not adjusted for multiple comparison.

rs1943226G(T)0.29 (0.94)0.759−0.19 (0.32)0.556
rs11875096C(T)0.80 (1.21)0.5090.12 (0.41)0.766
rs1943224G(A)−1.08 (2.36)0.6480.21 (0.80)0.790
rs7235242C(A)0.92 (2.06)0.654−0.10 (0.69)0.882
rs11872992A(G)−0.19 (0.73)0.791−0.00 (0.25)0.992
rs8093815T(C)0.10 (0.52)0.848−0.01 (0.17)0.955
rs17066856C(T)0.11 (0.67)0.8740.11 (0.23)0.611
rs17066836G(C)0.33 (1.41)0.8150.07 (0.47)0.876
rs1943227G(A)−0.57 (0.80)0.480−0.09 (0.27)0.747
rs1943218C(T)−0.18 (0.54)0.732−0.08 (0.18)0.651
rs17066829A(T)−0.33 (0.51)0.510−0.16 (0.17)0.340
rs9966412A(G)0.88 (0.64)0.1680.35 (0.21)0.102
rs17066859A(G)0.44 (0.76)0.5680.24 (0.26)0.351
rs9965495T(C)−0.01 (0.52)0.983−0.02 (0.18)0.890
rs12970134A(G)0.59 (0.59)0.3160.13 (0.20)0.509
rs17700633T(C)0.40(0.50)0.4250.00 (0.19)0.989
rs11873305G(T)−0.29 (0.93)0.752−0.08 (0.31)0.805
rs8091237G(C)0.12 (0.64)0.854−0.14 (0.22)0.510
rs17066866T(A)0.52 (1.09)0.6330.29 (0.37)0.429
rs7240064T(C)−0.14 (0.70)0.838−0.15 (0.24)0.526

SNP associations with change in weight or adiposity

Tables 3 and 4 list the overall and treatment specific genetic effects for the three major outcomes: short-term (baseline to 6 months) and long-term (baseline to 2 years) weight change, and diabetes incidence. The major “C” allele at SNP rs1943218 was associated with less short-term (P = 0.032) and long-term (P = 0.038) weight loss. Seven SNPs modified response to treatment on short-term (rs17066856, rs9966412, rs17066859, rs8091237, rs17066866, rs7240064) or long-term (rs12970134, rs17066866) reduction in body weight. Among them, nominally statistically significant associations with weight change for rs7240064 in the placebo arm and rs9966412 and rs17066859 in the metformin arm were observed, with each SNP's minor allele being associated with less short-term weight loss. The minor allele of rs17066866 was associated with less short-term (P = 0.006) and less long-term (P = 0.004) weight loss in the lifestyle intervention group, the latter of which approached the multiple-testing corrected significance threshold (P = 0.003).

Table 3. SNP effects on short-term, weight loss, long-term weight loss and diabetes risk
SNPMinor (major) allele6-month weight change (kg/allele)2-year weight change (kg/allele)Diabetes risk (hazard ratio)
N = 3,085N = 3,015N = 3,234
Coefficient (SE)P valueCoefficient (SE)P valueHR (95% CI)P value
  1. All analyses are adjusted for baseline age, sex, ethnicity and treatment assignment. Allele effects are estimated within the pooled sample of three treatment groups, unless the test of treatment*SNP interaction was statistically significant (P < 0.05), in which case the cells in the table are left blank and the relevant results are reported in Table 4.

rs1943226G(T)0.35 (0.24)0.1440.10 (0.33)0.7741.00 (0.91-1.11)0.980
rs11875096C(T)0.06 (0.31)0.8350.30 (0.42)0.4730.97 (0.77-1.23)0.805
rs1943224G(A)0.02 (0.59)0.9680.30 (0.82)0.7151.18 (0.78-1.79)0.431
rs7235242C(A)0.06 (0.53)0.908-0.07 (0.73)0.9201.28 (0.88-1.86)0.194
rs11872992A(G)-0.11 (0.18)0.566-0.18 (0.26)0.4741.04 (0.90-1.20)0.606
rs8093815T(C)0.13 (0.13)0.3040.32 (0.18)0.0810.99 (0.90-1.09)0.854
rs17066856C(T)--0.20 (0.24)0.3920.94 (0.83-1.07)0.375
rs17066836G(C)-0.12 (0.36)0.734-0.10 (0.49)0.8340.84 (0.63-1.11)0.216
rs1943227G(A)0.13 (0.20)0.5070.05 (0.28)0.8470.95 (0.81-1.12)0.540
rs1943218C(T)0.29 (0.14)0.0320.39 (0.19)0.0380.99 (0.89-1.10)0.869
rs17066829A(T)0.02 (0.13)0.8840.14 (0.18)0.420--
rs9966412A(G)--0.25 (0.23)0.2600.98 (0.86-1.11)0.737
rs17066859A(G)--0.44 (0.27)0.1030.94 (0.81-1.09)0.403
rs9965495T(C)0.17 (0.13)0.2070.20 (0.18)0.2651.00 (0.91-1.11)0.980
rs12970134A(G)-0.09 (0.15)0.548--1.02 (0.91-1.14)0.788
rs17700633T(C)0.15 (0.14)0.2730.10 (0.20)0.5960.97 (0.87-1.08)0.595
rs11873305G(T)-0.08 (0.24)0.738-0.18 (0.32)0.5770.99 (0.83-1.18)0.903
rs8091237G(C)--0.15 (0.23)0.5210.99 (0.87-1.12)0.815
rs17066866T(A)----0.98 (0.80-1.21)0.854
rs7240064T(C)--0.27 (0.25)0.2681.01 (0.88-1.16)0.875
Table 4. Treatment-specific SNP effects on short-term weight loss and diabetes risk
SNPMinor (major) alleleLifestyleMetforminPlacebo
6 m weight loss (kg/allele)Coefficient (SE)P valueCoefficient (SE)P valueCoefficient (SE)P value
  1. All analyses are adjusted for baseline age, sex, ethnicity, and treatment assignment.

N 1,0411,0241,020
rs17066856C(T)−0.01 (0.33)0.9800.50 (0.27)0.064−0.33 (0.27)0.214
rs9966412A(G)0.15 (0.32)0.6400.54 (0.26)0.034−0.28 (0.25)0.258
rs17066859A(G)0.35 (0.39)0.3580.71 (0.30)0.016−0.36 (0.31)0.245
rs8091237G(C)−0.36 (0.32)0.2590.40 (0.24)0.0940.32 (0.26)0.215
rs17066866T(A)1.48 (0.53)0.0060.54 (0.43)0.2070.05 (0.45)0.917
rs7240064T(C)−0.11 (0.34)0.7580.29 (0.26)0.2660.64 (0.28)0.024
2 yr weight loss (kg/allele)Coefficient (SE)P valueCoefficient (SE)P valueCoefficient (SE)P value
N 9991,0041,012
rs12970134A(G)−0.75 (0.43)0.0810.20 (0.31)0.5130.24 (0.32)0.450
rs17066866T(A)2.25 (0.77)0.0040.10 (0.60)0.8660.02 (0.60)0.979
Diabetes risk HR (95% CI)P valueHR (95% CI)P valueHR (95% CI)P value
N 1,0791,0731,082
rs17066829A(T)0.80 (0.64-1.00)0.0541.06 (0.89-1.27)0.5141.09 (0.92-1.29)0.309

Associations with diabetes incidence

Table 4 shows the SNP association results for type 2 diabetes incidence. SNP rs17066829 showed nominal evidence of interaction with treatment for diabetes risk (Pinteraction < 0.05); for example, each copy of the minor allele at SNP rs17066829 corresponded to a HR of 0.80 (95% CI: 0.64-1.00) in the lifestyle group, which contrasted the effect observed in the placebo group (HR = 1.09; 95% CI: 0.92-1.29).

Examining dietary factors and physical activity as putative mediators of genetic effects on obesity traits

Potential mediators were examined for all SNP associations and interactions that were of at least nominal statistical significance (as shown in Tables 3 and 4). No significant differences between the SNP effects before or after the adjustment of each potential mediator were detected.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

MC4R is a strong candidate gene for diabetes and obesity-related traits. The rs17782313 SNP, which is upstream of the MC4R locus, has for example been reliably associated with common obesity in a GWAS meta-analysis [5] and also in the DPP [17], but the MC4R region in general has been of continued interest in obesity research for many years [3, 5, 10, 20], largely because of its established role in monogenic obesity [4]. Hence, we examined common variation across MC4R to seek a more detailed understanding of the role the gene plays in obesity and diabetes, and to determine whether the apparent obesogenic effect of variation at this locus is modified by weight loss interventions in the DPP.

In analyses adjusted for multiple statistical comparisons, we found nominal evidence of association between the rs1943218 variant and short- (baseline to 6 months) and long- (baseline to 2 years) term weight loss. Interactions between metformin treatment and the rs9966412 and rs17066859 variants on short-term weight loss were also found. Moreover, we observed a nominally significant statistical interaction between the rs17066829 SNP and treatment on diabetes incidence, which is of potential interest, as a recent study also showed that this variant was associated with BMI [21].

MC4R is centrally expressed and binds the α, β, and γ isoforms of melanocortin-stimulating hormone (MSH) and adrenocoticotropic hormone (ACTH) with varying affinity, while Agouti and Agouti-related protein (AgRP) are endogenous MC4R antagonists [22]. Mc4r knockout mice are hyperphagic and are prone to maturity-onset obesity [23]. Although restricting food availability reduces body weight in female Mc4r−/− mice, these animals remain susceptible to obesity [24], indicating that MC4R ablation also influences energy expenditure; direct evidence for this comes from observations that Mc4r null mice have lower resting oxidative energy metabolism [24] and dark-phase locomotor activity [24, 25] compared with wild type animals. People heterozygous for specific MC4R mutations are MC4R deficient and are often hyperphagic, heavier and taller, with more fat and lean mass, and have higher bone mineral content than people who do not carry these mutations [26]. Genome-wide scans of White Europeans have identified two independent signals for BMI at the MC4R locus, which are also associated with adipose mass and type 2 diabetes [10]. One of the most thorough investigations of the MC4R locus in humans to date was undertaken by Thearle et al. in a study of 6,760 persons of predominantly Pima Indian ancestry [27]. The authors sequenced the MC4R exon to discover six low frequency mutations (∼2.4%), which were subsequently functionally interrogated in vitro. These mutations conveyed an effect on weight gain but this was only evident in childhood, and the effect on type 2 diabetes that the authors observed was mediated by adult BMI.

MC4R is a strong candidate gene for obesity-related traits; thus, the prior probability that variants at this locus are associated with obesity traits is likely to be above the null, in which case the nominal P values we report here may be somewhat conservative estimates of the underlying probability that an association exists. Moreover, although the DPP is the largest randomized controlled trial of intensive lifestyle modification, it is modestly powered for most of the tests performed here [17] for description of the statistical power characteristics of the DPP). Thus, correcting for multiple testing is likely to increase type 2 error (false-negative) rates. Nonetheless, we are conscious that many readers will be keen to compare the nominal test statistics with those corrected for multiple testing, which is why the multiple-test corrected significance threshold is given above.

Assessing epidemiological observations in an experimental setting is important step in the process of translating such observations into preventive action. Nevertheless, it is important to bear in mind that DPP participants were recruited on the basis of being overweight or obese and having impaired fasting and postchallenge glucose. Thus, the findings from this study may not generalize to other population subgroups. We were also unable to identify behavioral factors that might mediate the effects of MC4R on obesity-related traits, which would have aided the interpretation of the results described above.

In summary, we have conducted the first comprehensive analysis of common genomic variation at the MC4R locus in relation to changes in body composition and incidence of type 2 diabetes within a randomized controlled trial of weight loss interventions. Our findings, that several MC4R variants are nominally related to these traits, in some instances by modifying the effects of weight loss interventions on diabetes incidence and adipose accumulation, adds novel information on the role of the MC4R locus in the development of obesity and type 2 diabetes.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Bristol-Myers Squibb and Parke-Davis provided medication. LifeScan Inc., Health O Meter, Hoechst Marion Roussel, Inc., Merck-Medco Managed Care, Inc., Merck and Co., Nike Sports Marketing, Slim Fast Foods Co., and Quaker Oats Co. donated materials, equipment, or medicines for concomitant conditions. The opinions expressed are those of the investigators and do not necessarily reflect the views of the Indian Health Service or other funding agencies. A complete list of Centers, investigators, and staff can be found in the Appendix. The investigators gratefully acknowledge the commitment and dedication of all participants in the DPP, without whom this work would not have been possible.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information
  • 1
    Schousboe K, Willemsen G, Kyvik KO, et al. Sex differences in heritability of BMI: a comparative study of results from twin studies in eight countries. Twin Res 2003;6:409-421.
  • 2
    Ramachandrappa S, Farooqi IS. Genetic approaches to understanding human obesity. J Clin Invest 2011;121:2080-2086.
  • 3
    Speliotes EK, Willer CJ, Berndt SI, et al. Association analyses of 249,796 individuals reveal 18 new loci associated with body mass index. Nat Genet 2010;42:937-948.
  • 4
    Martinelli CE, Keogh JM, Greenfield JR, et al. Obesity due to melanocortin 4 receptor (MC4R) deficiency is associated with increased linear growth and final height, fasting hyperinsulinemia, and incompletely suppressed growth hormone secretion. J Clin Endocrinol Metab 2011;96:E181-E188.
  • 5
    Loos RJ, Lindgren CM, Li S, et al. Common variants near MC4R are associated with fat mass, weight and risk of obesity. Nat Genet 2008;40:768-775.
  • 6
    Qi L, Kraft P, Hunter DJ, et al. The common obesity variant near MC4R gene is associated with higher intakes of total energy and dietary fat, weight change and diabetes risk in women. Hum Mol Genet 2008;17:3502-3508.
  • 7
    Branson R, Potoczna N, Kral JG, et al. Binge eating as a major phenotype of melanocortin 4 receptor gene mutations. N Engl J Med 2003;348:1096-1103.
  • 8
    Greenfield JR, Miller JW, Keogh JM, et al. Modulation of blood pressure by central melanocortinergic pathways. N Engl J Med 2009;360:44-52.
  • 9
    Nogueiras R, Wiedmer P, Perez-Tilve D, et al. The central melanocortin system directly controls peripheral lipid metabolism. J Clin Invest 2007;117:3475-3488.
  • 10
    Renstrom F, Payne F, Nordstrom A, et al. Replication and extension of genome-wide association study results for obesity in 4923 adults from northern Sweden. Hum Mol Genet 2009;18:1489-1496.
  • 11
    Knowler WC, Hamman RF, Edelstein SL, et al. Prevention of type 2 diabetes with troglitazone in the Diabetes Prevention Program. Diabetes 2005;54:1150-1156.
  • 12
    The Diabetes Prevention Program Research Group. The diabetes prevention program: baseline characteristics of the randomized cohort. Diabetes Care 2000;23:1619-1629.
  • 13
    Knowler WC, Barrett-Connor E, Fowler SE, et al. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med 2002;346:393-403.
  • 14
    Mayer-Davis EJ, Sparks KC, Hirst K, et al. Dietary intake in the diabetes prevention program cohort: baseline and 1-year post randomization. Ann Epidemiol 2004;14:763-772.
  • 15
    Kriska AM, Knowler WC, LaPorte RE, et al. Development of questionnaire to examine relationship of physical activity and diabetes in Pima Indians. Diabetes Care 1990;13:401-411.
  • 16
    Bray GA, Jablonski KA, Fujimoto WY, et al. Relation of central adiposity and body mass index to the development of diabetes in the Diabetes Prevention Program. Am J Clin Nutr 2008;87:1212-1218.
  • 17
    Delahanty LM, Pan Q, Jablonski KA, et al. Genetic predictors of weight loss and weight regain after intensive lifestyle modification, metformin treatment, or standard care in the diabetes prevention program. Diabetes Care 2012;35:363-366.
  • 18
    Jablonski KA, McAteer JB, de Bakker PI, et al. 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.
  • 19
    Li J, Ji L. Adjusting multiple testing in multilocus analyses using the eigenvalues of a correlation matrix. Hereditry 2005;95:221-227.
  • 20
    Young EH, Wareham NJ, Farooqi S, et al. The V103I polymorphism of the MC4R gene and obesity: population based studies and meta-analysis of 29 563 individuals. Int J Obes (Lond) 2007;31:1437-1441.
  • 21
    Edwards TL, Velez Edwards DR, Villegas R, et al. HTR1B, ADIPOR1, PPARGC1A, and CYP19A1 and obesity in a cohort of Caucasians and African Americans: an evaluation of gene-environment interactions and candidate genes. Am J Epidemiol 2012;175:11-21.
  • 22
    Adan RA, Tiesjema B, Hillebrand JJ, et al. The MC4 receptor and control of appetite. Br J Pharmacol 2006;149:815-827.
  • 23
    Huszar D, Lynch CA, Fairchild-Huntress V, et al. Targeted disruption of the melanocortin-4 receptor results in obesity in mice. Cell 1997;88:131-141.
  • 24
    Ste Marie L, Miura GI, Marsh DJ, et al. A metabolic defect promotes obesity in mice lacking melanocortin-4 receptors. Proc Natl Acad Sci USA 2000;97:12339-12344.
  • 25
    Marsh DJ, Hollopeter G, Huszar D, et al. Response of melanocortin-4 receptor-deficient mice to anorectic and orexigenic peptides. Nat Genet 1999;21:119-122.
  • 26
    Farooqi IS, Keogh JM, Yeo GS, et al. Clinical spectrum of obesity and mutations in the melanocortin 4 receptor gene. N Engl J Med 2003;348:1085-1095.
  • 27
    Thearle MS, Muller YL, Hanson RL, et al. Greater impact of melanocortin-4 receptor deficiency on rates of growth and risk of type 2 diabetes during childhood compared with adulthood in Pima Indians. Diabetes 2012;61:250-257.
  • 28
    Chambers JC, Elliott P, Zabaneh D, et al. Common genetic variation near MC4R is associated with waist circumference and insulin resistance. Nat Genet 2008;40:716-718.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
oby20459-sup-0001-suppinfo.doc59KSupplementary Appendix: DPP Research Group Investigators
oby20459-sup-0002-suppinfo.doc60KOnline Supplementary Table 1. SNP identities and frequencies.
oby20459-sup-0003-suppinfo.tif25666KSupplementary Figure 1

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.