Hostname: page-component-7c8c6479df-8mjnm Total loading time: 0 Render date: 2024-03-27T16:07:21.370Z Has data issue: false hasContentIssue false

The Mediterranean diet protects against waist circumference enlargement in 12Ala carriers for the PPARγ gene: 2 years' follow-up of 774 subjects at high cardiovascular risk

Published online by Cambridge University Press:  14 September 2009

Cristina Razquin
Affiliation:
Department of Nutrition and Food Sciences, Physiology and Toxicology, University of Navarra, Irunlarrea 1, 31080Pamplona, Navarra, Spain
J. Alfredo Martinez
Affiliation:
Department of Nutrition and Food Sciences, Physiology and Toxicology, University of Navarra, Irunlarrea 1, 31080Pamplona, Navarra, Spain
Miguel A. Martinez-Gonzalez
Affiliation:
Department of Preventive Medicine and Public Health, School of Medicine–Clinica Universitaria, University of Navarra, Spain
Dolores Corella
Affiliation:
Department of Preventive Medicine, School of Medicine, University of Valencia, Spain
José Manuel Santos
Affiliation:
Department of Family Medicine, San Pablo Health Centre and CIBER Fisiopatología Obesidad y Nutrición (CB06/03), Instituto de Salud Carlos III, Spain
Amelia Marti*
Affiliation:
Department of Nutrition and Food Sciences, Physiology and Toxicology, University of Navarra, Irunlarrea 1, 31080Pamplona, Navarra, Spain
*
*Corresponding author: Professor Amelia Marti, fax +34 948 42 56 49, email amarti@unav.es
Rights & Permissions [Opens in a new window]

Abstract

The PPARγ gene regulates insulin sensitivity and adipogenesis. The Pro12Ala polymorphism of this gene has been related to fat accumulation. Our aim was to analyse the effects of a 2-year nutritional intervention with Mediterranean-style diets on adiposity in high-cardiovascular risk patients depending on the Pro12Ala polymorphism of the PPARγ gene. The population consisted of a substudy (774 high-risk subjects aged 55–80 years) of the Prevención con Dieta Mediterránea (PREDIMED) randomised trial aimed at assessing the effect of the Mediterranean diet for CVD prevention. There were three nutritional intervention groups: two of them of a Mediterranean-style diet and the third was a control group advised to follow a conventional low-fat diet. All the participants were genotyped by PCR-restriction fragment length polymorphism (RFLP). The results showed that carriers of the 12Ala allele allocated to the control group had a statistically significant higher change in waist circumference (adjusted difference coefficient = 2·37 cm; P = 0·014) compared with wild-type subjects after 2 years of nutritional intervention. This adverse effect was not observed among 12Ala carriers allocated to both Mediterranean diet groups. In diabetic patients a statistically significant interaction between Mediterranean diet and the 12Ala allele regarding waist circumference change was observed ( − 5·85 cm; P = 0·003). In conclusion, the Mediterranean diet seems to be able to reduce waist circumference in a high-cardiovascular risk population, reversing the negative effect that the 12Ala allele carriers of the PPARγ gene appeared to have. The beneficial effect of this dietary pattern seems to be higher among type 2 diabetic subjects.

Type
Full Papers
Copyright
Copyright © The Authors 2009

CVD, the leading cause of death in developed countries(Reference Estruch, Martinez-Gonzalez and Corella1, Reference Robitaille, Perusse and Bouchard2), is a good example of complex and multifactorial disease caused by genetic, environmental factors and their interactions(Reference Ordovas3). Excess of adiposity is a major risk factor for CVD, especially when this excess is located in visceral depots(Reference Despres4). The prevention of CVD is currently directed to potential modifiable factors, such as diet and lifestyle factors. In this sense, adiposity indexes (waist circumference and body weight) could be modified by interventions aimed to obtain changes in dietary pattern.

Since the Mediterranean diet was first defined by Keys & Grande in 1957(Reference Keys and Grande5), it has been postulated as beneficial against CVD(Reference Martinez-Gonzalez and Sanchez-Villegas6, Reference Trichopoulou7) and type 2 diabetes(Reference Schroder8). Olive oil consumption, which was the main source of fat (MUFA) in traditional Mediterranean diets, has been widely related to this protective effect(Reference Estruch, Martinez-Gonzalez and Corella1, Reference Trichopoulou7, Reference Fernandez-Jarne, Martinez-Losa and Prado-Santamaria9, Reference Martinez-Gonzalez, Fernandez-Jarne and Serrano-Martinez10). Also, tree nuts (a good source of PUFA) are an integral part of the Mediterranean food pattern and previous studies have found a protective association between nut consumption and reduced risk of CVD(Reference Albert, Gaziano and Willett11Reference Sabate and Fraser13).

One of the most studied genes linked to interactions with dietary compounds is the PPARγ(Reference Luan, Browne and Harding14Reference Ochoa, Razquin and Martinez-Gonzalez17) that encodes a transcription factor (PPARγ2) that controls the expression of genes involved in adipocyte differentiation, lipid storage and insulin sensitisation. The effects of the 12Ala allele have been studied in functional analysis, revealing that the receptor carrying this allele displays reduced DNA-binding affinity and impaired transcriptional activity in target genes(Reference Deeb, Fajas and Nemoto18). Therefore the 12Ala in vivo would be expected to protect against increased adiposity due to the reduced activity of the receptor. However, previous studies in human subjects showed that the 12Ala allele was associated with increased adiposity(Reference Ek, Urhammer and Sorensen19Reference Tok, Ertunc and Bilgin21). These contradictory results in vitro and in vivo might be explained by a potential enhancing effect of the anti-lipolytic action of insulin, which leads to reduced release of NEFA(Reference Deeb, Fajas and Nemoto18).

The Pro12Ala substitution in this gene has also been widely associated with diabetes(Reference Tonjes and Stumvoll22). Moreover, a positive association between the 12Ala allele of the PPARγ and obesity-related traits was reported among diabetic patients(Reference Franks, Jablonski and Delahanty23).

Most of the associations between diet, risk factors for chronic diseases and genetics have come from observational epidemiological studies and, as with conventional nutrition research, such observations need to be verified by randomised dietary intervention trials to provide the strongest evidence of causality(Reference Joost, Gibney and Cashman24). As PPARγ is known to be activated by fatty acids(Reference Xu, Lambert and Montana25), we hypothesised that the effects of a high consumption of virgin olive oil or nuts could be partly explained by means of the effects mediated by PPARγ. Thus, our aim was to analyse the effects of a 2-year nutritional intervention with Mediterranean-style diets on adiposity in high-cardiovascular risk patients depending on the Pro12Ala polymorphism of the PPARγ gene.

Research methods and procedures

Study population

The present study has been conducted within the frame of the Prevención con Dieta Mediterránea (PREDIMED) trial. The design of the PREDIMED trial has been reported in detail elsewhere(Reference Estruch, Martinez-Gonzalez and Corella1, Reference Zazpe, Sanchez-Tainta and Estruch26). Briefly, the PREDIMED trial is a large, parallel-group, multicentre, randomised and controlled 4-year clinical trial that aims to assess the effects of a Mediterranean-type diet on CVD and in which participants are assigned to one of three different dietary patterns (low-fat diet; Mediterranean diet supplemented with nuts; Mediterranean diet supplemented with virgin olive oil)(Reference Estruch, Martinez-Gonzalez and Corella1). The study population is composed of men aged 55–80 years and women aged 60–80 years with no previously documented CVD but at high cardiovascular risk. Inclusion criteria were either diabetes mellitus type 2 or at least three of the following risk factors: current smoking, hypertension, hyperlipidaemia, HDL-cholesterol < 1·034 mm, overweight or obesity, or family history of premature CHD.

We included data from 774 participants enrolled in the AP-UNAV recruitment centre in Pamplona (total number of participants was 1055), in which the retention rate during the second year was greater than 80 %. Of the subjects, 141 were excluded for the present study since we were not able to obtain DNA samples or their adiposity measures at year 2 were not recorded. All participants provided informed consent and the protocol was approved by the institutional review boards of all participating centres according to the Declaration of Helsinki.

Dietary assessment

The dietary habits of the participants, both at baseline and after follow-up for 24 months, were assessed using a semi-quantitative 137-item FFQ previously validated in Spain(Reference Martin-Moreno, Boyle and Gorgojo27). More details about the dietary assessment are described elsewhere(Reference Estruch, Martinez-Gonzalez and Corella1). We focused the nutritional study on fat intake changes.

Genotyping

Overnight fasting venous blood samples were collected in tubes containing EDTA. DNA was extracted from the buffy coat fraction using a commercial kit (Master PureTM; Epicentre, Madison, WI, USA). All the subjects were genotyped for the Pro12Ala polymorphism of the PPARγ gene (rs1801282) using the PCR-restriction fragment length polymorphism (RFLP) method as described elsewhere(Reference Ochoa, Marti and Azcona28).

Statistical analysis

The χ2 test was used to evaluate the Hardy–Weinberg equilibrium. The Kolmogorov–Smirnov test was used to determine variable distribution. Descriptive analyses of variables between the three intervention groups were performed using parametric tests (Student's t tests; ANOVA followed by Bonferroni post hoc tests). Means of changes in body weight and waist circumference were compared among the three randomised groups and among genotypes using general linear models adjusting for age, baseline BMI (continuous), sex, diabetes and smoking habit.

The effects of the polymorphism on changes in anthropometric variables were evaluated with multivariate linear regression models adjusting for relevant variables: age, sex (males as the reference category), baseline BMI, diabetes and smoking habits. For each adiposity index (changes in waist circumference (2nd year waist circumference − baseline waist circumference), body weight (2nd year weight − baseline weight) and BMI (2nd year BMI − baseline BMI)) three regression models were fit: the first model analysed the effects of the nutritional intervention and the polymorphism separately on changes in adiposity indexes. The second model included the interaction product-term between the nutritional intervention (olive oil+nuts v. control) and the Pro12Ala polymorphism (12Ala carriers v. non-carriers) and the third model included dummy variables to consider each possible combination of the genetic variants and the nutritional intervention group simultaneously.

We repeated all these analyses including only diabetic patients.

Results

Baseline characteristics of the participants, according to the nutritional intervention and the polymorphism, are presented in Table 1. As expected, because of the randomisation, there were no statistically significant differences among the three nutritional groups in any of the biological variables. Most of the PREDIMED subjects, at baseline, were overweight or obese (91 %) and more than one-third had a diagnosis of type 2 diabetes.

Table 1 Baseline characteristics of the participants according to intervention group allocation

(Mean values and standard deviations)

* One subject was not included because of missing data for this analysis.

The analysis of macronutrient intake at the beginning of the study and after 2 years of nutritional intervention revealed that the intervention brought about an effective change in the dietary pattern of the subjects with significant differences in the expected direction according to the respective allocated group. These changes were remarkable regarding fat consumption (Table 2). A significantly higher consumption of total fat, monounsaturated and polyunsaturated fat (but not saturated fat) was observed in both Mediterranean diet groups compared with the control group (Table 2). Moreover, the control group had a significantly higher ratio of saturated fat:total fat than those subjects allocated to the Mediterranean diet groups (Table 2).

Table 2 Energy and fat intake after 2 years of nutritional intervention according to intervention group

(Mean values and standard deviations)

Mean value was significantly different from that of the control group: *P = 0·009, **P = 0·008, ***P = 0·004, ****P < 0·001.

Mean value was significantly different from that of the virgin olive oil group: † P = 0·019, †† P = 0·001, ††† P < 0·001.

To analyse whether the Mediterranean diet had a beneficial effect on adiposity indexes, we first investigated body weight and waist circumference changes depending on the intervention group. Mean body weight changes for the control, olive oil and nuts groups were similar (0·34 (sem 0·28), 0·33 (sem 0·22) and 0·87 (sem 0·21) kg, respectively). The comparison for mean body weight changes was not statistically significant, neither between control and olive oil (P = 0·992) nor control and nuts (P = 0·131). Mean waist circumference changes were also alike in the control, olive oil and nuts groups (0·35 (sem 0·35), 0·14 (sem 0·28) and 0·51 (sem 0·26) cm, respectively). No differences were found when control and olive oil (P = 0·632) or control and nuts (P = 0·711) were compared.

The distribution of genotypes for the Pro12Ala of PPARγ gene polymorphism was in Hardy–Weinberg equilibrium in the study groups, being the 12Ala allele frequency 0·07 in each group. Due to the low frequency of 12Ala of the PPARγ gene, only four out of the 774 subjects were homozygous for the polymorphism (Ala12Ala). Therefore we decided to merge in all the analysis the Ala carriers (both heterozygous and homozygous merged together as a single group, representing 14·5 % of the total population) and to compare them with the non-mutated subjects. Therefore this variable had always 1 df.

Adjusted mean body weight changes were 0·56 and 0·92 kg for ProPro and Ala carriers, respectively. Mean waist changes were 0·30 cm in non-mutated (ProPro) and 1·28 cm in mutated subjects (Ala carriers), this difference being statistically significant (P = 0·046). Moreover, when the three nutritional groups were also considered in the model, we observed that in the control group the 12Ala carriers had a significantly higher (P = 0·011) waist circumference increase compared with non-carriers after 2 years of nutritional intervention (Fig. 1). The same tendency was observed when weight changes were analysed, although the differences within control subjects carrying the 12Ala allele v. non-mutated subjects were not statistically significant (data not shown). The polymorphism was not associated with differences in waist circumference changes within both Mediterranean diet groups.

Fig. 1 Waist circumference changes according to the Pro12Ala polymorphism (, ProPro; ■, Ala) of the PPARγ gene after 2 years of nutritional intervention in the total population. Data are means, with standard deviation represented by vertical bars. ** Mean value was significantly different from that of the non-mutated subjects allocated to the control group (P = 0·011).

To go deeper in our analysis, several multiple linear regression models were fitted in order to find out the model that best predicted the effects of the Mediterranean diet on adiposity indexes depending on the Pro12Ala polymorphism of the PPARγ gene. Table 3 shows the model to predict the effects of the diet (considering the three intervention groups) and the two groups of genetic variants on waist circumference changes (2nd year − baseline), adjusting for sex, age, baseline BMI, diabetes and smoking habit. The obtained results supported the finding of the analysis of covariance (ANCOVA) showing that carriers of the 12Ala allele had a statistically significant increase in waist circumference compared with non-mutated subjects (ProPro) (P = 0·046). Moreover, we found that subjects with higher baseline BMI were those with the highest waist circumference reduction (adjusted changes: − 0·20 cm; P < 0·001) after 2 years of nutritional intervention regardless of nutritional intervention and genotype (Table 3).

Table 3 Multiple regression models for the analysis of the effects of the Pro12Ala polymorphism of the PPARγ2 gene on waist circumference change after 2 years of nutritional intervention

(B coefficients and 95 % confidence intervals)

* Dependent variable: waist circumference change (2nd year waist circumference − baseline waist circumference).

When the interaction was formally tested, the P value for interaction (2 df; three diet groups ×  two genotype groups) was P = 0·214.

Homozygous and heterozygous subjects merged together in a single group.

When a 1 df product-term to assess effect modification (interaction) between the nutritional intervention (control (0) v. Mediterranean diet groups (1)) and the two groups of genotypes (Ala carriers (1) v. non-carriers (0)) on waist circumference changes was introduced in the model (data not shown), the P value was not statistically significant (B = − 1·89; P = 0·095).

Thus, to better analyse the effects of the Mediterranean diet and the polymorphism together, we applied a multiple regression model with five dummy variables to include the combinations of the three nutritional groups and the two groups of genotypes as compared with non-mutated subjects (ProPro genotype) allocated to the control group, who were the single reference category (Table 3). This model was adjusted for sex, age, baseline BMI, diabetes and smoking habit and the dependent variable was 2-year waist circumference change (2nd year − baseline). We found that subjects allocated to the control group who also were carriers of the PPARγ polymorphism (12Ala) had the highest 2-year waist circumference increase compared with the reference group (adjusted difference = 2·37 cm; P = 0·014). In agreement with the ANCOVA results, there were no statistically significant differences within the subjects allocated to the Mediterranean diet groups, independently of the genotype, and the reference category.

Similar models were fit considering as the dependent variables body weight or BMI changes. The results were in line with findings regarding waist circumference, but neither of them reached significance for the product-terms of interaction (data not shown).

Of special interest was the analysis that we performed with the subset of diabetic patients (n 276) in order to investigate the effects of the Mediterranean diet on adiposity depending on the genotype for the PPARγ within this group. The general linear model showed that the detrimental effect of the 12Ala allele on waist changes was statistically significant (P < 0·001) within the subset of diabetic subjects allocated to the control group (Fig. 2), and the magnitude of this effect was higher than in the whole population (Figs. 1 and 2). In diabetic mutated subjects, the Mediterranean diet was able to reduce waist circumference, reversing the negative effect of the 12Ala allele (Fig. 2). As shown in the total population this effect was also observed when analysing body weight changes although it did not reach statistical significance (data not shown).

Fig. 2 Waist circumference changes according to the Pro12Ala polymorphism (, ProPro; ■, Ala) of the PPARγ gene after 2 years of nutritional intervention in diabetic patients. Data are means, with standard deviation represented by vertical bars. *** Mean value was significantly different from that of the non-mutated subjects allocated to the control group (P < 0·001).

A statistically significant interaction within diabetic subjects was observed between the nutritional intervention (control (0) v. Mediterranean diet groups (1)) and the polymorphism (Ala carriers (1) v. non-carriers (0)) in the multiple linear regression model. Ala carriers allocated to the Mediterranean diet groups were predicted to have a statistically significant lower increase in waist circumference compared with control subjects (B = − 5·85; P = 0·003), although the main effect of the polymorphism per se predicted a statistically higher waist circumference change after 2 years of nutritional intervention (B = 0·980; P = 0·046) (Table 4).

Table 4 Multiple regression models for the analysis of the effects of the Pro12Ala polymorphism of the PPARγ2 gene on waist circumference change in diabetic and non-diabetic patients after 2 years of nutritional intervention

(B coefficients and 95 % confidence intervals)

* Dependent variable: waist circumference change (2nd year waist circumference − baseline waist circumference).

Intervention refers to the comparison between olive oil and nuts groups v. control group (1 df). The two intervention groups (olive oil and nuts) were merged together.

12Ala allele: comparison between 12Ala carriers (homozygotes and heterozygotes merged together) and non-carriers (ProPro genotype) (1 df).

The same models fitted among diabetic patients were performed in non-diabetic patients, but they did not reach statistical significance (Table 4).

Regarding these results a three-way interaction assessing genotype (1 df; Ala carriers v. non-carriers) × intervention (1 df; control v. Mediterranean diet groups) × diabetes (1 df; yes v. no) was tested in our regression model for waist circumference change after 2 years of intervention. We obtained a non-statistically significant P value (B = − 0·66; P = 0·270). In this context the interaction product diabetes × genotype (1 df) resulted statistically significant within the control group (P = 0·012) when analysing separately the control and the Mediterranean diet groups (olive oil and nuts merged together).

Discussion

The excess of adiposity is considered as the main cause of the increased morbidity and mortality by CVD, especially when this excess is localised in the abdominal depots(Reference Despres4). Waist circumference is an excellent marker of abdominal obesity and consequently of CVD risk(Reference Despres4) and mortality(Reference Pischon, Boeing and Hoffmann29). Therefore, the reduction of adiposity indexes, such as waist circumference and body weight, would result in an effective prevention for CVD. We have studied the effect of the Pro12Ala polymorphism of the PPARγ gene on 2-year changes in adiposity indexes in a population assigned to a nutritional intervention trial with the Mediterranean diet (the PREDIMED trial), a randomised and controlled study for the prevention of CVD, which allowed us to control not only for established confounding factors such as sex, age and baseline parameters but also for unknown or unmeasured confounders.

First of all, the present results proved the effectiveness of the Mediterranean diet nutritional intervention since the fat intake distribution in the three nutritional groups changed in the expected direction(Reference Zazpe, Sanchez-Tainta and Estruch26). After 2 years, both Mediterranean dietary groups had a significantly higher consumption of total fat compared with the control group and the distribution of fatty acid intake was also very different. The Mediterranean diet groups had a significantly higher consumption of MUFA and PUFA and significantly lower saturated fat consumption, which are important characteristics of the Mediterranean dietary pattern(Reference Trichopoulou7).

When we analysed the effects of the nutritional intervention, no differences in changes in waist circumference and weight were found according to the nutritional group. This could be explained because the control subjects recruited for the trial already followed a healthier lifestyle before being included in the present study. This fact may justify the small differences in weight and waist changes between the Mediterranean groups and the control group after 2 years of intervention.

Our main objective was to analyse the effects of a Mediterranean dietary pattern depending on the Pro12Ala polymorphism of the PPARγ gene. We found a minor 12Ala allele frequency of 0·07, lower than values reported in Spanish subjects (0·13)(Reference Ochoa, Marti and Azcona28) but similar to those obtained in European populations(Reference Lagou, Scott and Manios30).

The analysis of the genotype variants on changes on adiposity indexes, adjusted for the nutritional intervention, revealed that the 12Ala allele was associated with statistically significant higher increase in waist circumferences (P = 0·046). It also showed higher, but not significant, body weight changes and BMI after 2 years of nutritional intervention. The majority of the studies analysing the effects of the polymorphism on adiposity indexes are directed to BMI, but there are few studies analysing the effects of the polymorphism on waist circumference. Similarly to our findings, Pihlajamaki et al. reported a significant association of waist circumference and the polymorphism in a healthy Finnish population(Reference Pihlajamaki, Vanhala and Vanhala31).

When considering the nutritional intervention we observed that the negative effect of the Ala allele on waist changes was only present in the control group (P = 0·011) and it was highly attenuated in the two Mediterranean diet groups (Fig. 1). It seems that the Mediterranean diet reversed the detrimental effect that the 12Ala minor allele appears to have on waist circumference. One plausible hypothesis that could partly explained this finding is that PPARγ gene is known to be activated by fatty acids(Reference Xu, Lambert and Montana25) and, as stated before, the Mediterranean diet consists of a high consumption of monounsaturated (olive oil) and polyunsaturated fat (nuts).

The interaction between the Mediterranean diet and the Pro12Ala polymorphism of the PPARγ gene on adiposity presented was not statistically significant (B = − 1·863; P = 0·095) although Robitaille et al. reported a Pro12Ala × diet interaction on waist circumference: ProPro subjects had significantly higher waist circumference when higher levels of total fat (P = 0·007) or saturated fat (P = 0·002) were consumed(Reference Robitaille, Perusse and Bouchard2).

Due to the potential role that the Pro12Ala polymorphism of the PPARγ gene seemed to have on type 2 diabetes(Reference Tonjes and Stumvoll22, Reference Ghoussaini, Meyre and Lobbens32), we performed a specific analysis on diabetic patients. This analysis revealed that the detrimental effect that the polymorphism appeared to have on waist circumference and weight changes was more relevant among these patients than in the total study population (Fig. 2). Moreover, the reversing effect of the nutritional intervention was also higher in diabetic 12Ala carriers allocated to the two Mediterranean diet groups. The interaction product-term, intervention × 12Ala, was statistically significant, confirming the protective effect of the Mediterranean diet against higher waist circumference, especially in 12Ala carriers. Some authors have reported the beneficial effects of the Mediterranean diet on diabetes(Reference Schroder8, Reference Biesalski33, Reference Martinez-Gonzalez, de la Fuente-Arrillaga and Nunez-Cordoba34) and glycaemic control(Reference Shai, Schwarzfuchs and Henkin35). Likewise, there are also studies analysing the effects of the polymorphism and dietary compounds on adiposity indexes in diabetic patients. In agreement with our findings, a higher visceral fat deposition on individuals carrying the 12Ala allele and with a low polyunsaturated fat intake was reported(Reference Franks, Jablonski and Delahanty23).

Following the present results, it seems that the negative effect of the 12Ala allele in the total population could be mostly attributed to diabetic patients. Moreover, we observed that this effect was only attenuated when these patients followed a Mediterranean dietary pattern. This attenuation may be partly explained by the fact that fatty acids (high intake of PUFA and MUFA in this dietary pattern) are activating the PPARγ protein and could potentially modify the action of the Pro12Ala substitution on the receptor activity, especially in the presence of diabetes(Reference Deeb, Fajas and Nemoto18, Reference Franks, Jablonski and Delahanty23, Reference Xu, Lambert and Montana25).

The present study has several strengths, such as its design, which is able to provide first-level scientific evidence(Reference Woolf, Battista and Anderson36), as we work in real-life conditions (home-prepared foods). But the study also has some limitations. The first was to ensure participants' compliance with the aid of continuous dietary instructions. In fact, we could observe that the dietary pattern (especially fat distribution) was significantly different between the Mediterranean diet groups and the control group after 2 years of intervention (Table 2). Another limitation of the statistical procedure is that multiple comparisons were performed in order to examine the contribution of the three nutritional groups and the different genotypes, but the differences remained statistically significant even after the Bonferroni correction.

In summary, the present study shows that Mediterranean-style diets seemed to be able to reduce 2-year waist circumference enlargement in a high-cardiovascular risk population, reversing the negative effect that carrying the 12Ala allele of the PPARγ gene appeared to have. The beneficial effect of this dietary pattern seemed to be more important among type 2 diabetic subjects.

Acknowledgements

C. R. has a predoctoral fellowship from IBERCAJA.

We would also like to thank all the participants and members of the PREDIMED project.

The authors acknowledge the following sources of support: Department of Health of the Navarra Government, Linea Especial sobre Nutrición, Obesidad y Salud (LE 97), RETICS Scheme funded by the Spanish Ministry of Health (PREDIMED Project, reference RD 06/0045/0000). CIBERobn is an initiative of ISCIII (CB06/03/1017).

C. R. carried out the experimental procedures, the analysis and interpretation of the data, and drafted the manuscript. A. M. participated in the design of the study, contributed to the analysis and interpretation of the data, and helped to draft the manuscript. M. A. M.-G. helped with the statistical analysis and helped to draft the manuscript. J. A. M. helped with the interpretation of the data and helped to draft the manuscript. M. A. M.-G., D. C. and J. M. S. participated in the initiation and design of the study and in the recruitment of the subjects. All authors read and approved the final version of the manuscript.

The authors have no conflicts of interest to declare.

References

1Estruch, R, Martinez-Gonzalez, MA, Corella, D, et al. (2006) Effects of a Mediterranean-style diet on cardiovascular risk factors: a randomized trial. Ann Intern Med 145, 111.CrossRefGoogle ScholarPubMed
2Robitaille, J, Perusse, L, Bouchard, C, et al. (2007) Genes, fat intake, and cardiovascular disease risk factors in the Quebec Family Study. Obesity (Silver Spring) 15, 23362347.CrossRefGoogle ScholarPubMed
3Ordovas, JM (2006) Genetic interactions with diet influence the risk of cardiovascular disease. Am J Clin Nutr 83, 443S446S.CrossRefGoogle ScholarPubMed
4Despres, JP (2006) Intra-abdominal obesity: an untreated risk factor for type 2 diabetes and cardiovascular disease. J Endocrinol Invest 29, 7782.Google ScholarPubMed
5Keys, A & Grande, F (1957) Role of dietary fat in human nutrition. III. Diet and the epidemiology of coronary heart disease. Am J Public Health Nations Health 47, 15201530.CrossRefGoogle ScholarPubMed
6Martinez-Gonzalez, MA & Sanchez-Villegas, A (2004) The emerging role of Mediterranean diets in cardiovascular epidemiology: monounsaturated fats, olive oil, red wine or the whole pattern? Eur J Epidemiol 19, 913.CrossRefGoogle ScholarPubMed
7Trichopoulou, A (2004) Traditional Mediterranean diet and longevity in the elderly: a review. Public Health Nutr 7, 943947.CrossRefGoogle ScholarPubMed
8Schroder, H (2007) Protective mechanisms of the Mediterranean diet in obesity and type 2 diabetes. J Nutr Biochem 18, 149160.CrossRefGoogle ScholarPubMed
9Fernandez-Jarne, E, Martinez-Losa, E, Prado-Santamaria, M, et al. (2002) Risk of first non-fatal myocardial infarction negatively associated with olive oil consumption: a case–control study in Spain. Int J Epidemiol 31, 474480.CrossRefGoogle ScholarPubMed
10Martinez-Gonzalez, MA, Fernandez-Jarne, E, Serrano-Martinez, M, et al. (2004) Development of a short dietary intake questionnaire for the quantitative estimation of adherence to a cardioprotective Mediterranean diet. Eur J Clin Nutr 58, 15501552.CrossRefGoogle Scholar
11Albert, CM, Gaziano, JM, Willett, WC, et al. (2002) Nut consumption and decreased risk of sudden cardiac death in the Physicians' Health Study. Arch Intern Med 162, 13821387.CrossRefGoogle ScholarPubMed
12Hu, FB, Stampfer, MJ, Manson, JE, et al. (1998) Frequent nut consumption and risk of coronary heart disease in women: prospective cohort study. BMJ 317, 13411345.CrossRefGoogle ScholarPubMed
13Sabate, J & Fraser, GE (1994) Nuts: a new protective food against coronary heart disease. Curr Opin Lipidol 5, 1116.CrossRefGoogle ScholarPubMed
14Luan, J, Browne, PO, Harding, AH, et al. (2001) Evidence for gene–nutrient interaction at the PPARγ locus. Diabetes 50, 686689.CrossRefGoogle ScholarPubMed
15Memisoglu, A, Hu, FB, Hankinson, SE, et al. (2003) Interaction between a peroxisome proliferator-activated receptor γ gene polymorphism and dietary fat intake in relation to body mass. Hum Mol Genet 12, 29232929.CrossRefGoogle ScholarPubMed
16Nieters, A, Becker, N & Linseisen, J (2002) Polymorphisms in candidate obesity genes and their interaction with dietary intake of n-6 polyunsaturated fatty acids affect obesity risk in a sub-sample of the EPIC-Heidelberg cohort. Eur J Nutr 41, 210221.CrossRefGoogle Scholar
17Ochoa, MC, Razquin, C, Martinez-Gonzalez, MA, et al. (2008) Role of PPARG2 polymorphims in body weight regulation. Future Lipidol 3, 3141.CrossRefGoogle Scholar
18Deeb, SS, Fajas, L, Nemoto, M, et al. (1998) A Pro12Ala substitution in PPARγ2 associated with decreased receptor activity, lower body mass index and improved insulin sensitivity. Nat Genet 20, 284287.CrossRefGoogle ScholarPubMed
19Ek, J, Urhammer, SA, Sorensen, TI, et al. (1999) Homozygosity of the Pro12Ala variant of the peroxisome proliferation-activated receptor-γ2 (PPAR-γ2): divergent modulating effects on body mass index in obese and lean Caucasian men. Diabetologia 42, 892895.CrossRefGoogle ScholarPubMed
20Masud, S & Ye, S (2003) Effect of the peroxisome proliferator activated receptor-γ gene Pro12Ala variant on body mass index: a meta-analysis. J Med Genet 40, 773780.CrossRefGoogle ScholarPubMed
21Tok, EC, Ertunc, D, Bilgin, O, et al. (2006) PPAR-γ2 Pro12Ala polymorphism is associated with weight gain in women with gestational diabetes mellitus. Eur J Obstet Gynecol Reprod Biol 129, 2530.CrossRefGoogle ScholarPubMed
22Tonjes, A & Stumvoll, M (2007) The role of the Pro12Ala polymorphism in peroxisome proliferator-activated receptor γ in diabetes risk. Curr Opin Clin Nutr Metab Care 10, 410414.CrossRefGoogle ScholarPubMed
23Franks, PW, Jablonski, KA, Delahanty, L, et al. (2007) The Pro12Ala variant at the peroxisome proliferator-activated receptor γ gene and change in obesity-related traits in the Diabetes Prevention Program. Diabetologia 50, 24512460.CrossRefGoogle ScholarPubMed
24Joost, HG, Gibney, MJ, Cashman, KD, et al. (2007) Personalised nutrition: status and perspectives. Br J Nutr 98, 2631.CrossRefGoogle Scholar
25Xu, HE, Lambert, MH, Montana, VG, et al. (1999) Molecular recognition of fatty acids by peroxisome proliferator-activated receptors. Mol Cell 3, 397403.CrossRefGoogle ScholarPubMed
26Zazpe, I, Sanchez-Tainta, A, Estruch, R, et al. (2008) A large randomized individual and group intervention conducted by registered dietitians increased adherence to Mediterranean-type diets: the PREDIMED study. J Am Diet Assoc 108, 11341145.CrossRefGoogle ScholarPubMed
27Martin-Moreno, JM, Boyle, P, Gorgojo, L, et al. (1993) Development and validation of a food frequency questionnaire in Spain. Int J Epidemiol 22, 512519.CrossRefGoogle ScholarPubMed
28Ochoa, MC, Marti, A, Azcona, C, et al. (2004) Gene–gene interaction between PPARγ2 and ADRβ3 increases obesity risk in children and adolescents. Int J Obes Relat Metab Disord 28, Suppl. 3, S37S41.CrossRefGoogle Scholar
29Pischon, T, Boeing, H, Hoffmann, K, et al. (2008) General and abdominal adiposity and risk of death in Europe. N Engl J Med 359, 21052120.CrossRefGoogle ScholarPubMed
30Lagou, V, Scott, RA, Manios, Y, et al. (2008) Impact of peroxisome proliferator-activated receptors γ and δ on adiposity in toddlers and preschoolers in the GENESIS Study. Obesity (Silver Spring) 16, 913918.CrossRefGoogle ScholarPubMed
31Pihlajamaki, J, Vanhala, M, Vanhala, P, et al. (2004) The Pro12Ala polymorphism of the PPARγ2 gene regulates weight from birth to adulthood. Obes Res 12, 187190.CrossRefGoogle ScholarPubMed
32Ghoussaini, M, Meyre, D, Lobbens, S, et al. (2005) Implication of the Pro12Ala polymorphism of the PPAR-γ2 gene in type 2 diabetes and obesity in the French population. BMC Med Genet 6, 11.CrossRefGoogle Scholar
33Biesalski, HK (2004) Diabetes preventive components in the Mediterranean diet. Eur J Nutr 43, Suppl. 1, I/26I/30.CrossRefGoogle ScholarPubMed
34Martinez-Gonzalez, MA, de la Fuente-Arrillaga, C, Nunez-Cordoba, JM, et al. (2008) Adherence to Mediterranean diet and risk of developing diabetes: prospective cohort study. BMJ 336, 13481351.CrossRefGoogle ScholarPubMed
35Shai, I, Schwarzfuchs, D, Henkin, Y, et al. (2008) Weight loss with a low-carbohydrate. Mediterranean, or low-fat diet. N Engl J Med 359, 229241.CrossRefGoogle ScholarPubMed
36Woolf, SH, Battista, RN, Anderson, GM, et al. (1990) Assessing the clinical effectiveness of preventive maneuvers: analytic principles and systematic methods in reviewing evidence and developing clinical practice recommendations. A report by the Canadian Task Force on the Periodic Health Examination. J Clin Epidemiol 43, 891905.CrossRefGoogle ScholarPubMed
Figure 0

Table 1 Baseline characteristics of the participants according to intervention group allocation(Mean values and standard deviations)

Figure 1

Table 2 Energy and fat intake after 2 years of nutritional intervention according to intervention group(Mean values and standard deviations)

Figure 2

Fig. 1 Waist circumference changes according to the Pro12Ala polymorphism (, ProPro; ■, Ala) of the PPARγ gene after 2 years of nutritional intervention in the total population. Data are means, with standard deviation represented by vertical bars. ** Mean value was significantly different from that of the non-mutated subjects allocated to the control group (P = 0·011).

Figure 3

Table 3 Multiple regression models for the analysis of the effects of the Pro12Ala polymorphism of the PPARγ2 gene on waist circumference change after 2 years of nutritional intervention(B coefficients and 95 % confidence intervals)

Figure 4

Fig. 2 Waist circumference changes according to the Pro12Ala polymorphism (, ProPro; ■, Ala) of the PPARγ gene after 2 years of nutritional intervention in diabetic patients. Data are means, with standard deviation represented by vertical bars. *** Mean value was significantly different from that of the non-mutated subjects allocated to the control group (P < 0·001).

Figure 5

Table 4 Multiple regression models for the analysis of the effects of the Pro12Ala polymorphism of the PPARγ2 gene on waist circumference change in diabetic and non-diabetic patients after 2 years of nutritional intervention(B coefficients and 95 % confidence intervals)