Dietary fat intake and polymorphisms at the PPARG locus ... - Nature

3 downloads 70 Views 316KB Size Report
May 3, 2011 - 2 Anghel SI, Wahli W. Fat poetry: a kingdom for PPAR gamma. Cell. Res 2007; 17: ... Diabetes 1998; 47: 1806–1808. 18 Masud S, Ye S. Effect ...
International Journal of Obesity (2012) 36, 218–224 & 2012 Macmillan Publishers Limited All rights reserved 0307-0565/12 www.nature.com/ijo

ORIGINAL ARTICLE Dietary fat intake and polymorphisms at the PPARG locus modulate BMI and type 2 diabetes risk in the D.E.S.I.R. prospective study A Lamri1,2, C Abi Khalil1,2,3, R Jaziri1,2, G Velho1, O Lantieri4, S Vol4, P Froguel5,6, B Balkau7,8, M Marre1,2,3, F Fumeron1,2 and the D.E.S.I.R. Study Group 1 INSERM, U695 (Genetic determinants of type 2 diabetes and its vascular complications), Paris, France; 2Univ Paris Diderot, Sorbonne Paris Cite´, UMRS 695, UFR de Me´decine Site Bichat, Paris, France; 3AP-HP, Department of Endocrinology, Diabetology, Nutrition, and Metabolic Diseases, Bichat Claude Bernard Hospital, Paris, France; 4Institut inter Re´gional Pour la Sante´ (IRSA), La Riche, France; 5CNRS, UMR8199, Institute of Biology, Pasteur Institute, Lille, France; 6Department of Genomics of Common Disease, School of Public Health, Imperial College London, Hammersmith Hospital, London, UK; 7 INSERM, CESP Centre for Research in Epidemiology and Population Health, U1018, Villejuif, France and 8University Paris Sud 11, UMRS 1018, Villejuif, France

Context: Fat-rich diets are involved in many disorders such as obesity and type 2 diabetes (T2D). The Pro12Ala variant of peroxisome proliferator-activated receptor-g (PPARg) is known to modulate body mass index (BMI) and T2D risk. Objective: Our aim was to study the interaction effect between PPARg gene (PPARG) polymorphisms Pro12Ala and 1431C4T and fat intake on incident T2D and BMI in a 9-year prospective cohort drawn from the French general population, the D.E.S.I.R. (Data from an Epidemiological Study on the Insulin Resistance Syndrome) study (n ¼ 4676). Methods: Nutritional intake was assessed by a food frequency self-questionnaire completed by each participant. Statistical analyses included logistic regression, analysis of covariance and haplotype analysis, with adjustment for confounding variables. Results: A high fat consumption (the third sex-specific tertile of fat intake, as a percentage of energy intake) was associated with an increased T2D risk among ProPro and CC homozygotes (Pinteraction ¼ 0.05, odds ratio (OR) (95% confidence interval (95% CI)) ¼ 1.73 (1.19–2.52) P ¼ 0.004 and OR ¼ 1.85 (1.27–2.71) P ¼ 0.001, respectively) but not in Ala and T carriers. There was a significant interaction effect between Pro12Ala and 1431C4T on BMI (Pinteraction ¼ 0.004); Ala was associated with lower BMI in CC homozygotes and with higher BMI in T carriers while the opposite was found for ProPro. There was also an interaction effect between Pro12Ala and dietary fat intake on BMI (Pinteraction ¼ 0.02); AlaAla individuals had a higher BMI than Pro carriers among high fat consumers (27.1±1.0 versus 24.9±0.1 for AlaAla and Pro þ , respectively). There was no interaction effect between the 1431C4T single-nucleotide polymorphism and fat intake on BMI. Conclusion: Our results indicate strong genetic and nutritional interaction effects on BMI and T2D risk at the PPARG locus in a general population. International Journal of Obesity (2012) 36, 218–224; doi:10.1038/ijo.2011.91; published online 3 May 2011 Keywords: genetic variation; PPARG; fat intake; gene–environment interactions; type 2 diabetes; body mass index

Introduction Fat-rich diets contribute to the onset of many disorders such as obesity and type 2 diabetes (T2D). The metabolic response to a high fat diet is, however, influenced by other factors and

Correspondence: Dr F Fumeron, Institut National de la Sante´ et de la Recherche Me´dicale (INSERM) Unite´ 695, Xavier Bichat Medical School, 16 rue Henri Huchard, Paris 75018, France. E-mail: [email protected] Received 29 October 2010; revised 15 March 2011; accepted 30 March 2011; published online 3 May 2011

remains highly dependent on the genetic background of individuals.1 Peroxisome proliferator-activated receptor-g (PPARg) is a transcription factor that directly regulates target genes, mediating lipid metabolism and adipocyte differentiation.2 The common Pro12Ala variant of PPARg is expressed only in the adipocyte-specific g2 isoform and results from a C to G transversion in the exon B of the PPARg encoding gene (PPARG).3–5 A number of studies have associated the Pro12Ala polymorphism with T2D and hyperglycemia with a protective effect of the minor allele Ala.6–12 To a lesser extent, this variant has also been associated

Dietary fat intake and PPARG polymorphisms modulate BMI and T2D risk A Lamri et al

219 with obesity and body weight changes, with conflicting results.12–18 Another PPARG polymorphism in linkage disequilibrium with Pro12Ala, the 1431C4T single-nucleotide polymorphism (SNP), has also been associated with T2D. In some studies, it has been shown to modulate the Pro12Ala effect on both T2D and body mass index (BMI).19–21 Interaction effects between fat intake and PPARG polymorphisms on BMI and other related parameters have been documented, but the results have not always been concordant.22–26 Our aim was to study potential interactions between Pro12Ala and 1431C4T polymorphisms and dietary fat intake on T2D and BMI, in a prospective cohort drawn from a French general population, the D.E.S.I.R. (Data from an Epidemiological Study on the Insulin Resistance Syndrome) study.

Materials and methods Population The study population, men and women aged 30–65 years, participated in the cohort D.E.S.I.R., a 9-year follow-up study that aimed to clarify the development of the insulin resistance syndrome.27 Participants were recruited from 10 health examination centers from the western part of France. These volunteers were insured by the French social security system, which offers periodic health examinations free of charge. All participants signed an informed consent. The protocol was approved by the consultative committee for the protection of participants for biomedical research of Biceˆtre Hospital. A total of 4676 men and women were genotyped for Pro12Ala and 1431C4T, and had data available at baseline. Among them, 3646 were non-diabetic at baseline and could be followed for incident diabetes during the 9-year period. In accordance with the 1997 American Diabetes Association criteria, T2D was defined as fasting plasma glucose X7.0 mmol l1 and/or treatment by drugs for diabetes.28 Incident cases (n ¼ 191) were defined for those free of disease at entry, who developed T2D at any time during the follow-up.

Measurements Weight, height and waist circumferences were measured by trained personnel, and BMI (kg m2) was calculated. Venous blood samples were collected in the morning after a 12-h fast. Fasting plasma glucose was assayed by the glucose oxidase method applied to fluoro-oxalated plasma, using a Technicon RA 1000 (Bayer, Puteaux, France) or a Specific or a Delta Automate (Konelab, Evry, France). A 23-item questionnaire was completed by each participant on the habitual frequency and quantity of consumption of different foods. This questionnaire has been validated by comparison with a dietary history method questionnaire, administrated by trained dieticians with an interview of at least 30 min (correlation coefficients were between 0.44 and 0.80).29 This validation study enabled the determination of

multiple linear regression equations to estimate the main nutrient intakes from the questions on the consumption of different foods.29 Fat intake was thus determined from the questions on the consumption of meat, fish, pork products, fried foods, butter, cheese and other dairy products. Low, medium and high fat consumption were defined by the sexspecific tertiles of fat intake (as a percentage of total energy). Total energy derived from fat varied from 22.5 to 32.9%, 32.9 to 36.4% and from 36.4 to 48.6% in the first, second and third tertiles, respectively, for men, and from 25.4 to 35.4%, 35.4 to 38.4% and 38.4 to 49.3% for women.

Genotyping The Pro12Ala SNP was genotyped using an assay on demand kit (Applied Biosystems, Foster City, CA, USA). The PCR was performed with a GeneAmp 9700 PCR system (Applied Biosystems). The conditions for the TaqMan reaction were 95 1C for 10 s and 40 cycles of 92 1C for 15 s, 60 1C for 1 min and 15 1C for 5 s. 1431C4T SNP genotyping was performed by Applied Biosystems SNPlex technology based on an oligonucleotide ligation assay combined with multiplex PCR target amplification (http://www.appliedbiosystems.com). Allelic discrimination was performed through capillary electrophoresis analysis, using an Applied Biosystems 3730xl DNA analyzer (Applied Biosystems) and GeneMapper3.7 software (Applied Biosystems). The genotypes were determined with an ABI PRISM 7900 HT sequence detection system (Applied Biosystems). Of those genotyped for Pro12Ala, 100 could not be genotyped for 1431C4T (genotyping success rates: 99.7 and 97.0% for Pro12Ala and 1431C4T, respectively). Statistical analysis Departure from Hardy–Weinberg equilibrium was tested using a w2 with one degree of freedom. The odds ratios (ORs) for T2D risk associated with genotypes were estimated by multivariable logistic regression adjusted for age, sex, BMI and alcohol consumption. Interactions between genotypes and fat intake on T2D risk were also assessed by multivariable logistic regression analysis, with the same adjustments. The effect of genotypes and the interaction effects of genotypes with fat intake on BMI were tested by analysis of covariance in the individuals non-diabetic at baseline, with adjustments for baseline age, sex, alcohol consumption and physical activity, using Systat for Windows software (version 13) (Systat Software, Inc., Chicago, IL, USA). P-values of multiple pairwise comparisons were corrected by the Bonferroni method. Haplotypic effects on BMI were estimated by the THESIAS software with the same adjustments.30

Results General characteristics of the study population at baseline and at the 9-year follow-up are shown in Table 1. The genotypic distributions of both the Pro12Ala and 1431C4T polymorphisms were in Hardy–Weinberg International Journal of Obesity

Dietary fat intake and PPARG polymorphisms modulate BMI and T2D risk A Lamri et al

International Journal of Obesity

P ¼ 0.21

0.58 (0.25–1.36) 1

P ¼ 0.004

1.73 (1.19–2.52) 1

P ¼ 0.36

0.67 (0.28–1.58) 1 OR for diabetes incidencea (95% CI) (low+medium versus high)

Abbreviations: BMI, body mass index; CI, confidence interval; D.E.S.I.R., Data from an Epidemiological Study on the Insulin Resistance Syndrome; ORs, odds ratios; SNPs, single-nucleotide polymorphisms. Pinteraction ¼ 0.05 for both SNPs. aBy multivariable logistic regression analysis (adjusted for baseline age, sex, BMI and alcohol consumption).

1

P ¼ 0.001

1.85 (1.27–2.71)

64 (6.6) 40 (4.3) 47 (5.1) 9 (3.4) 14 (5.4) 17 (5.8) 64 (6.3) 41 (4.2) 49 (5.0) 9 (3.6) 14 (5.5) 15 (5.7) With diabetes n (%)

910 (93.4) 254 (96.6) 882 (94.9) 891 (95.7)

High Medium Low High Medium

277 (94.2) 247 (94.6) 960 (93.7) 935 (95.0) 935 (95.8) 239 (96.4)

Influence of PPARG polymorphisms and fat intake on T2D No significant association was found for either of the PPARG SNPs with T2D incidence when tested either separately or in combination. The incidence of T2D was higher in the high fat intake group (third tertile) when compared with the low fat (first tertile, OR (95% confidence interval (95% CI)) ¼ 1.52 (1.02– 2.27), P ¼ 0.04) or to the medium fat intake group (second tertile, OR (95% CI) ¼ 1.46 (1.00–2.14), P ¼ 0.05). As there was no difference for T2D incidence between the first and the second tertile groups of fat intake, OR (95% CI) ¼ 0.96 (0.64–1.43), P ¼ 0.83, the interaction effects with the PPARG polymorphisms were investigated using only two categories: low and medium fat intake (first þ second tertile groups) versus high fat intake category (third tertile). There were significant interaction effects between the PPARG polymorphisms and total energy derived from fat consumption on T2D risk (Pinteraction ¼ 0.05 for both Pro12Ala and 1431C4T). An increase in fat intake was associated with an increased risk of T2D only among the ProPro and the CC homozygotes (Table 2).

248 (94.3) 239 (94.5)

equilibrium in the whole cohort at inclusion (P ¼ 0.48 for Pro12Ala and P ¼ 0.98 for 1431C4T polymorphism) with a minor allele frequency of 10.4 and 11.8% for Pro12Ala and 1431C4T SNP, respectively. Both polymorphisms were in linkage disequilibrium (D’ ¼ 0.669, r2 ¼ 0.41, Po0.001).

Without diabetes n (%)

Abbreviations: BMI, body mass index; D.E.S.I.R., Data from an Epidemiological Study on the Insulin Resistance Syndrome. Data are mean values±s.d., unless otherwise indicated. aNumbers with data available at follow-up: 3586. b According to the International Diabetes Federation definition.

Low

258 (7.1) 1212 (33.2) 729 (20.0) 796 (21.8)

High

124 (2.6) 838 (17.5) 373 (7.7) 531 (11.1)

Type 2 diabetes n (%) Metabolic syndromeb n (%) Lipid treatment n (%) Hypertension treatment n (%)

Medium

987 (27.5) 1708 (47.6) 891 (24.9)

Low

1256 (26.2) 2551 (53.1) 993 (20.7)

High

Physical activitya n (%) Low Medium High

Medium

2023±453 79.0±17.3 73.1±11.1 227.9±53.8 14.8±17.4

Low

2128±508 84.3±19.4 74.4±11.9 239.0±53.9 16.7±20.4

CC

Energy intake (kcal per day1) Fat (g) Proteins (g) Carbohydrates (g) Alcohol (g)

T+

1769/1877 56.1±9.9 25.7±4.0 86.4±12.2 5.43±0.92 5.71±0.93 135.2±19.0 398 (11.0)

ProPro

2369/2431 46.8±10.0 24.7±3.8 83.4±11.8 5.36±0.84 5.74±1,02 131.0±15 1057 (22.0)

Ala+

Men/women, n Age (years) BMI (kg m2) Waist (cm) Fasting glucose (mmol l1) Total cholesterol (mmol l1) Systolic blood pressure (mm Hg) Current smokers, n (%)

1431C4T

9-year follow-up

Pro12Ala

Baseline

Fat intake

Table 1 Baseline and 9-year follow-up characteristics of subjects from the D.E.S.I.R. cohort study

Table 2 Numbers with and without incident diabetes and odds ratios (95% CI) for cumulative 9-year incidence of type 2 diabetes associated with high fat intake at baseline as compared with medium/low fat intake, according to Pro12Ala, 1431C4T SNPs. The D.E.S.I.R. Study

220

Dietary fat intake and PPARG polymorphisms modulate BMI and T2D risk A Lamri et al

221

Figure 1 Mean (s.e.) BMI (kg m2) at baseline according to Pro12Ala and 1431C4T genotypes. The D.E.S.I.R. Study. Analysis of covariance adjusted for baseline age, sex, alcohol consumption and physical activity. Pinteraction ¼ 0.0006; Bonferroni-corrected P-values: aPc ¼ 0.02, bPc ¼ 0.36, c Pc ¼ 0.03, dPc ¼ 0.16.

Influence of PPARG polymorphisms and fat intake on BMI No significant effect of PPARG polymorphisms on BMI was detected, although the homozygotes for the minor alleles had the highest BMI (25.4±5.1, 24.6±3.7 and 24.7±3.8 kg m2 for AlaAla, ProAla and ProPro, respectively; P ¼ 0.38 and BMI ¼ 25.3±3.4, 24.6±3.8 and 24.8±3.8 kg m2 for TT, CT and CC genotype, respectively; P ¼ 0.14). Nevertheless, a very significant interaction effect between these SNPs on BMI was detected (Pinteraction ¼ 0.0006). Ala was associated with lower BMI at baseline in CC homozygotes and with higher BMI in T carriers. Conversely, T was associated with a higher BMI among Ala carriers and with a lower BMI in ProPro homozygotes (Figure 1). The results were similar when testing the interaction effect on BMI at 9 years or on repeated measures of BMI (data not shown). The genotypes did not influence BMI changes over the follow-up. These results were confirmed by a haplotype study where the genetic background modulated significantly the effect of each SNP at baseline (Figure 2). The interaction between fat intake and Pro12Ala on BMI was significant for the recessive model (Pinteraction ¼ 0.02). In each genotype, a higher fat intake was associated with a higher BMI at baseline; in the high fat consumers group only, the AlaAla homozygotes had a higher BMI than Pro carriers (Figure 3). No interaction between 1431C4T SNP and fat intake was detected. There were no significant interactions of PPARG SNPs genotype combinations or haplotypes with fat intake on BMI. All the results remained similar after additional adjustment for drug treatments (data not shown).

Figure 2 Estimated haplotype effects on BMI. The D.E.S.I.R. Study. Estimated means (95% CI) according to PPARG haplotypes, calculated by the THESIAS program. Adjusted for baseline age, sex, alcohol consumption and physical activity.

Figure 3 Mean (s.e.) BMI (kg m2) at baseline according to Pro12Ala polymorphism and fat intake. The D.E.S.I.R. Study. Analysis of covariance adjusted for baseline age, sex, alcohol consumption and physical activity. Pinteraction ¼ 0.02; Bonferroni-corrected Pc-values: aPc ¼ 0.13, bPc ¼ 0.03, c Pc ¼ 0.006.

Discussion In previous studies, we have shown a modest association of T2D and impaired fasting glycemia with PPARG polymorphisms,6 or only a trend.31 In the present study, we focus on the interaction issue. We report that a high fat intake is associated with the incidence of T2D over a 9-year period, in interaction with PPARG polymorphisms. There was a International Journal of Obesity

Dietary fat intake and PPARG polymorphisms modulate BMI and T2D risk A Lamri et al

222 gene–gene interaction effect between these two polymorphisms on BMI and there was also an interaction between fat intake and the Pro12Ala polymorphism. A high fat intake was associated with an increased T2D risk at 9 years only among the ProPro and the CC homozygotes, but not in the other genotypes. Therefore, the beneficial effect of the Ala allele on T2D risk usually reported6–12 might be due in part to the lack of response to a higher fat intake for Ala carriers. To our knowledge, this interaction with fat intake on T2D risk has not been described in the literature. Robitaille et al.26 described a similar effect on waist circumference, and showed that a high fat intake was correlated with an increased visceral adipose tissue accumulation among ProPro individuals only. However, this interaction was not detected in our cohort. The association between Pro12Ala and BMI has been the subject of several studies with controversial results. A metaanalysis with data from 30 independent studies, included cases with different pathologies related to metabolic disorders. In that study, the effect of the Ala allele on BMI was recessive, and AlaAla homozygotes had a higher BMI than ProPro in obese individuals.18 A second meta-analysis included data from 57 independent studies with healthy participants showed an effect of PPAR on BMI for Caucasians only. In this case, AlaAla individuals had a slightly higher BMI than ProPro.14 The higher BMI of Ala homozygotes in our study is in agreement with these analyses. Moreover, our results show an interaction between the Pro12Ala SNP and dietary fat intake on BMI. AlaAla homozygotes had significantly higher BMI than Pro carriers among the high fat consumers, but not among the low or medium fat consumers. Similar results have been described in a diabetic population.24 According to Luan et al.,23 a low polyunsaturated/saturated fatty acid ratio is associated with a higher BMI among Ala carriers. These results are in agreement with ours, since an increase in total fat intake is highly correlated with an increase in saturated fatty acid consumption. A study with Pro12Ala knock-in mice model is also in agreement with our observations: there was a trend for AlaAla mice to have a higher BMI, body weight and fat mass than ProPro homozygotes under a high fat diet.25 Another study shows results contradictory to ours: Ala carriers had higher BMI among the low fat consumers (first quintile) while an increase in fat intake was associated with an increased BMI in ProPro homozygotes but not in Ala carriers.22 However, this study included only healthy women. In our results, sex did not modulate the PPARG–fat intake interaction (data not shown). The reason for this difference is unclear. As nutritional habits vary between different populations, this Pro12Ala–fat intake interaction could partially explain the controversial results previously described for the effect of this SNP on BMI in the different ethnic groups. The second polymorphism 1431C4T did not show any direct effect on BMI. Nevertheless, there was an interaction between this SNP and Pro12Ala, affecting BMI. International Journal of Obesity

This modulation of 1431C4T on Pro12Ala effect could also explain the conflicting BMI data observed in populations with different Pro12Ala and 1431C4T allelic and haplotype frequencies. A similar interaction effect has previously been described by Doney et al.,19 but with a weaker significance, probably because of low statistical power. The main results of our study seem paradoxical. A high fat intake increases BMI in AlaAla individuals but increases T2D risk in ProPro homozygotes. This may be in line with the effects of the PPARg agonists, the glitazones which are drugs for treatment of diabetes that increase body weight. However, the results concerning Pro12Ala and fat intake interaction on BMI should be considered with caution as the difference in BMI between genotypes in high fat consumers was no longer significant after Bonferroni correction, probably because of the small number of AlaAla homozygotes in this study. Therefore, this observation needs replication. The second polymorphism 1431C4T acts similarly to Pro12Ala on T2D risk, but not on BMI, despite the strong linkage disequilibrium between the two SNPs. Nevertheless, the r2-value (0.41) indicates that they are not completely correlated. Therefore, the effects of Pro12Ala and fat intake on T2D and BMI could involve two distinct mechanisms. The interaction effect of fat intake with Pro12Ala (and the absence of interaction with 1431C4T) on BMI may be due to the fact that Pro12Ala is specific to PPARg2, the PPARg isoform expressed only in adipose tissue. Indirectly, this may also mean that the effect of Pro12Ala and 1431C4T on T2D could be due to the effect of the protein not only in the adipocytes, but also in other tissues where it is expressed and could imply other variants or mutations of the gene, in linkage disequilibrium with these SNPs. The Ala allele was associated with a decreased transactivation of the target genes, and a reduced adipogenic capacity due to a reduced DNA-binding capacity.7,12 In a study of Pro12Ala knock-in mice,25 under a chow diet, the pattern of expression of PPARg target genes of PPARG was similar between ProPro and AlaAla mice and only two genes were differentially expressed. Under a high fat diet, a total of 46 genes were differentially expressed between the two genotypes in white adipose tissue, with a majority being downregulated in the AlaAla mice, thus pointing to a different response of these genotype groups to an increase in fat intake that could have an impact on BMI and diabetes risk, in agreement with what we observed in our study. The prospective design of the D.E.S.I.R. study allowed us to avoid many biases of case–control studies. Nevertheless, at each follow-up examination, glycemia was measured only once, whereas T2D is diagnosed with at least two measures of glycemia. Our study has another important limitation: although it has been validated, the data from the selfquestionnaire remain semi-quantitative and do not allow the estimation of these interactions with different fat components (monounsaturated, polyunsaturated or saturated fatty

Dietary fat intake and PPARG polymorphisms modulate BMI and T2D risk A Lamri et al

223 acids). Unfortunately, plasma fatty acids could not be measured either. In conclusion, our results are indicative of strong genetic and nutritional interaction effects on BMI and T2D at the PPARG locus.

Conflict of interest The authors declare no conflict of interest.

Acknowledgements The D.E.S.I.R. study has been supported by INSERM contracts with CNAMTS, Lilly, Novartis Pharma and Sanofi-Aventis; by INSERM (Re´seaux en Sante´ Publique, Interactions entre les de´terminants de la sante´, Cohortes Sante´ TGIR 2008), the Association Diabe`te Risque Vasculaire, the Fe´de´ration Francaise de Cardiologie, La Fondation de France, ALFEDIAM, CNIEL, ONIVINS, Socie´te´ Francophone du Diabe`te, Ardix Medical, Bayer Diagnostics, Becton Dickinson, Cardionics, Merck Sante´, Novo Nordisk, Pierre Fabre, Roche, Topcon. The D.E.S.I.R. Study Group: INSERM CESP U1018: B Balkau, P Ducimetie`re and E Eschwe`ge; INSERM U367: F Alhenc-Gelas; CHU D’Angers: Y Gallois and A Girault; Bichat Hospital: F Fumeron and M Marre; CHU de Rennes: F Bonnet; CNRS UMR8090, Lille: P Froguel; Centres d’Examens de Sante´: Alenc- on, Angers, Blois, Caen, Chartres, Chateauroux, Cholet, Le Mans, Orle´ans and Tours; Institute de Recherche Me´decine Ge´ne´rale: J Cogneau; General practitioners of the region; Institute inter-Regional pour la Sante´: C Born, E Caces, M Cailleau, JG Moreau, O Lantieri, F Rakotozafy, J Tichet and S Vol.

References 1 Andreasen CH, Andersen G. Gene-environment interactions and obesityFfurther aspects of genomewide association studies. Nutrition 2009; 25: 998–1003. 2 Anghel SI, Wahli W. Fat poetry: a kingdom for PPAR gamma. Cell Res 2007; 17: 486–511. 3 Fajas L, Auboeuf D, Raspe E, Schoonjans K, Lefebvre AM, Saladin R et al. The organization, promoter analysis, and expression of the human PPARgamma gene. J Biol Chem 1997; 272: 18779–18789. 4 Mueller E, Drori S, Aiyer A, Yie J, Sarraf P, Chen H et al. Genetic analysis of adipogenesis through peroxisome proliferator-activated receptor gamma isoforms. J Biol Chem 2002; 277: 41925–41930. 5 Yen CJ, Beamer BA, Negri C, Silver K, Brown KA, Yarnall DP et al. Molecular scanning of the human peroxisome proliferator activated receptor gamma (hPPAR gamma) gene in diabetic Caucasians: identification of a Pro12Ala PPAR gamma 2 missense mutation. Biochem Biophys Res Commun 1997; 241: 270–274. 6 Jaziri R, Lobbens S, Aubert R, Pean F, Lahmidi S, Vaxillaire M et al. The PPARG Pro12Ala polymorphism is associated with a decreased risk of developing hyperglycemia over 6 years and combines with the effect of the APM1 G-11391A single nucleotide polymorphism: the Data from an Epidemiological Study on the Insulin Resistance Syndrome (DESIR) study. Diabetes 2006; 55: 1157–1162.

7 Masugi J, Tamori Y, Mori H, Koike T, Kasuga M. Inhibitory effect of a proline-to-alanine substitution at codon 12 of peroxisome proliferator-activated receptor-gamma 2 on thiazolidinedioneinduced adipogenesis. Biochem Biophys Res Commun 2000; 268: 178–182. 8 Saxena R, Voight BF, Lyssenko V, Burtt NP, de Bakker PI, Chen H et al. Genome-wide association analysis identifies loci for type 2 diabetes and triglyceride levels. Science 2007; 316: 1331–1336. 9 Ludovico O, Pellegrini F, Di Paola R, Minenna A, Mastroianno S, Cardellini M et al. Heterogeneous effect of peroxisome proliferator-activated receptor gamma2 Ala12 variant on type 2 diabetes risk. Obesity (Silver Spring) 2007; 15: 1076–1081. 10 Lohmueller KE, Pearce CL, Pike M, Lander ES, Hirschhorn JN. Meta-analysis of genetic association studies supports a contribution of common variants to susceptibility to common disease. Nat Genet 2003; 33: 177–182. 11 Memisoglu A, Hu FB, Hankinson SE, Liu S, Meigs JB, Altshuler DM et al. Prospective study of the association between the proline to alanine codon 12 polymorphism in the PPARgamma gene and type 2 diabetes. Diabetes Care 2003; 26: 2915–2917. 12 Deeb SS, Fajas L, Nemoto M, Pihlajamaki J, Mykkanen L, Kuusisto J et al. A Pro12Ala substitution in PPARgamma2 associated with decreased receptor activity, lower body mass index and improved insulin sensitivity. Nat Genet 1998; 20: 284–287. 13 Meirhaeghe A, Fajas L, Helbecque N, Cottel D, Auwerx J, Deeb SS et al. Impact of the peroxisome proliferator activated receptor gamma2 Pro12Ala polymorphism on adiposity, lipids and noninsulin-dependent diabetes mellitus. Int J Obes Relat Metab Disord 2000; 24: 195–199. 14 Tonjes A, Scholz M, Loeffler M, Stumvoll M. Association of Pro12Ala polymorphism in peroxisome proliferator-activated receptor gamma with pre-diabetic phenotypes: meta-analysis of 57 studies on nondiabetic individuals. Diabetes Care 2006; 29: 2489–2497. 15 Valve R, Sivenius K, Miettinen R, Pihlajamaki J, Rissanen A, Deeb SS et al. Two polymorphisms in the peroxisome proliferatoractivated receptor-gamma gene are associated with severe overweight among obese women. J Clin Endocrinol Metab 1999; 84: 3708–3712. 16 Ek J, Urhammer SA, Sorensen TI, Andersen T, Auwerx J, Pedersen O. Homozygosity of the Pro12Ala variant of the peroxisome proliferation-activated receptor-gamma2 (PPAR-gamma2): divergent modulating effects on body mass index in obese and lean Caucasian men. Diabetologia 1999; 42: 892–895. 17 Beamer BA, Yen CJ, Andersen RE, Muller D, Elahi D, Cheskin LJ et al. Association of the Pro12Ala variant in the peroxisome proliferator-activated receptor-gamma2 gene with obesity in two Caucasian populations. Diabetes 1998; 47: 1806–1808. 18 Masud S, Ye S. Effect of the peroxisome proliferator activated receptor-gamma gene Pro12Ala variant on body mass index: a meta-analysis. J Med Genet 2003; 40: 773–780. 19 Doney A, Fischer B, Frew D, Cumming A, Flavell DM, World M et al. Haplotype analysis of the PPARgamma Pro12Ala and C1431T variants reveals opposing associations with body weight. BMC Genet 2002; 3: 21. 20 Doney AS, Fischer B, Cecil JE, Boylan K, McGuigan FE, Ralston SH et al. Association of the Pro12Ala and C1431T variants of PPARG and their haplotypes with susceptibility to Type 2 diabetes. Diabetologia 2004; 47: 555–558. 21 Wei Q, Jacobs Jr DR, Schreiner PJ, Siscovick DS, Steffes MW, Fornage M. Patterns of association between PPARgamma genetic variation and indices of adiposity and insulin action in AfricanAmericans and whites: the CARDIA Study. J Mol Med 2006; 84: 955–965. 22 Memisoglu A, Hu FB, Hankinson SE, Manson JE, De Vivo I, Willett WC et al. Interaction between a peroxisome proliferatoractivated receptor gamma gene polymorphism and dietary fat intake in relation to body mass. Hum Mol Genet 2003; 12: 2923–2929.

International Journal of Obesity

Dietary fat intake and PPARG polymorphisms modulate BMI and T2D risk A Lamri et al

224 23 Luan J, Browne PO, Harding AH, Halsall DJ, O’Rahilly S, Chatterjee VK et al. Evidence for gene-nutrient interaction at the PPARgamma locus. Diabetes 2001; 50: 686–689. 24 Vaccaro O, Lapice E, Monticelli A, Giacchetti M, Castaldo I, Galasso R et al. Pro12Ala polymorphism of the PPARgamma2 locus modulates the relationship between energy intake and body weight in type 2 diabetic patients. Diabetes Care 2007; 30: 1156–1161. 25 Heikkinen S, Argmann C, Feige JN, Koutnikova H, Champy MF, Dali-Youcef N et al. The Pro12Ala PPARgamma2 variant determines metabolism at the gene-environment interface. Cell Metab 2009; 9: 88–98. 26 Robitaille J, Despres JP, Perusse L, Vohl MC. The PPAR-gamma P12A polymorphism modulates the relationship between dietary fat intake and components of the metabolic syndrome: results from the Quebec Family Study. Clin Genet 2003; 63: 109–116.

International Journal of Obesity

27 Balkau B. An epidemiologic survey from a network of French Health Examination Centres (D.E.S.I.R.): epidemiologic data on the insulin resistance syndrome. Rev Epidemiol Sante Publique 1996; 44: 373–375. 28 The Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care 1997; 20: 1183–1197. 29 Lasfargues G, Vol S, Le Clesiau H, Bedouet M, Hagel L, Constans T et al. Validity of a short self-administered dietary questionnaire compared with a dietetic interview. Presse Med 1990; 19: 953–957. 30 Tregouet DA, Garelle V. A new JAVA interface implementation of THESIAS: testing haplotype effects in association studies. Bioinformatics 2007; 23: 1038–1039. 31 Vaxillaire M, Veslot J, Dina C, Proenca C, Cauchi S, Charpentier G et al. Impact of common type 2 diabetes risk polymorphisms in the DESIR prospective study. Diabetes 2008; 57: 244–254.