The Single Nucleotide Polymorphism PPARG2 Pro12Ala Affects Body Mass Index, Fat Mass, and Blood Pressure in Severely Obese Patients

Background The PPARG2 Pro12Ala (rs1801282) and IL6 -174G >C (rs1800795) have important function in body weight regulation and a potential role in obesity risk. We aimed to investigate the association between PPARG2 Pro12Ala and IL6 -174G >C variants and the genotypes interaction with body composition, metabolic markers, food consumption, and physical activity in severely obese patients. Methods 150 severely obese patients (body mass index (BMI) ≥ 35 kg/m2) from Central Brazil were recruited. Body composition, metabolic parameters, physical activity, and dietary intake were measured. The genotype was determined by the qPCR TaqMan Assays System. Multiple linear regression and multiple logistic regression models were fitted adjusting for confounders. Results Ala carriers of the Pro12Ala polymorphism had higher adiposity measures (BMI: p=0.031, and fat mass: p=0.049) and systolic blood pressure (p=0.026) compared to Pro homozygotes. We found no important associations between the -174G >C polymorphism and obesity phenotypes. When genotypes were combined, individuals with genotypes ProAla + AlaAla and GC + CC presented higher BMI (p=0.029) and higher polyunsaturated fatty acids (PUFAs) consumption (p=0.045) compared to the ones with genotypes ProPro and GG, and individuals carriers of the PPARG2 Ala allele only (genotype ProAla + AlaAla and GG) had higher fat mass and systolic and diastolic blood pressure compared to the ones with genotypes ProPro and GG. Conclusions Severely obese individuals carrying the Ala allele of the PPARG2 Pro12Ala polymorphism had higher measures of adiposity and blood pressure, while no important associations were found for the IL6 -174G >C polymorphism.


Introduction
Studies about the genetic obesity susceptibility have investigated polymorphisms related to genes encoding factors involved in food and energy intake regulation, energy expenditure, and adipogenesis control [1]. In this context, the polymorphisms PPARG2 Pro12Ala (rs1801282) and IL6 -174G >C (rs1800795) have received attention for their possible influence in body weight regulation [1][2][3].
e PPARG2 has an important role in modulating the expression of genes involved in the regulation of adipose tissue differentiation and lipid metabolism [4,5]. e mostly studied polymorphism of PPARG2, the Pro12Ala, may promote lower affinity of the PPAR-c for the response element and about 50% lower transcriptional capacity [6]. e Ala allele has been associated with lower body mass index (BMI) [6,7]; nevertheless, some meta-analyses have found contradictory results showing higher BMI in Ala carriers, especially in severely obese individuals [8][9][10][11]. Considering this potential association of the Ala allele and increased BMI, it is important to investigate whether this variant influences other parameters in severely obese individuals.
us, the role of Pro12Ala and -174G >C in modulating BMI is still to be confirmed and also the information regarding other body composition parameters such as fat mass, fat-free mass, percentage body fat, and lean mass.
Interaction of genes with environmental factors, such as diet and physical activity, may be involved in the discrepancies of associations between studies [1,20,21]. Gene-diet interaction studies with Pro12Ala and -174G >C polymorphisms have found that the energy content and composition of the diet may affect obesity phenotypes, showing the importance to assess the interactions among genotypes and potentially modifiable lifestyle factors [20,22,23]. Considering that the PPARG2 Pro12Ala and IL6 -174G >C polymorphisms have important functions in body weight regulation with a potential role in obesity risk and the alarming increase in severe obesity worldwide [24], they are promising single nucleotide polymorphisms (SNPs) for association studies of obesity phenotypes. us, we aimed to investigate the association between PPARG2 Pro12Ala and IL6 -174G >C variants and the genotypes interaction with body composition, metabolic markers, food consumption, and physical activity in severely obese patients.

Subjects.
is study is an analysis of baseline data from participants of the clinical trial "Effect of Nutritional

Anthropometric and Body Composition Measurements.
Body weight and height were measured using standardized procedures [25]. BMI (kg/m 2 ) was calculated dividing the body mass (kg) by the squared height (m 2 ). Severe obesity was defined as BMI ≥ 35 kg/m 2 [24].
For BIA assessment, patients were instructed to fast for 12 h and avoid strenuous physical activity and alcohol, as well as food and drinks containing caffeine on the previous day [26]. BIA assessment was conducted according to the manufacturer's guidelines.

Physical Activity Assessment.
Physical activity was assessed using a triaxial accelerometer ActiGraph wGT3X (ActiGraph, Pensacola, FL, USA) for movement registration. Patients were instructed to wear the accelerometer 24 h a day for six consecutive days over the nondominant wrist, even during shower and water activities, as the device was waterproof. e sampling frequency of the accelerometer was set at 30 Hz, and the data collection interval was set at one min. Accelerometers were set up and downloaded at ActiLife 6 software. Output data were processed using the R-package GGIR (http://cran.r-project.org). e outcome measures used in the present study were moderate-to-vigorous physical activity (MVPA) (>100 mg) defined as estimated time spent in ≥ 10 min per bout during a week and the sedentary time (<50 mg, without bouts) in min per day. Journal of Obesity 742INT (Omron Healthcare Inc., Kyoto, Japan) automatic blood pressure monitor with an appropriately sized cuff, and the mean was calculated. e presence of comorbidities was analyzed as a dichotomous variable (presence/absence). Subjects with systolic/diastolic blood pressure higher than 140/90 mmHg or under antihypertensive therapy were considered hypertensive [28]. Subjects with fasting glucose ≥ 126 mg/dL or under glucose-lowering therapy were considered diabetic [29]. Individuals with low-density lipoprotein (LDL) cholesterol ≥ 160 mg/dL and/or triglycerides ≥ 150 mg/dL or high-density lipoprotein (HDL) cholesterol < 40 mg/dL for men and < 50 mg/dL for women were classified as dyslipidaemic [30].

Laboratory
Tests. Blood samples were collected for metabolic markers and genomic DNA extraction after 12 h overnight fasting. Serum glucose, total cholesterol, HDL cholesterol, LDL cholesterol, and triglycerides were measured by enzyme-colorimetric methods. Serum insulin was measured by chemiluminescence, and hemoglobin A1c (HbA1c) was measured by liquid chromatography. e homeostasis model assessment of insulin resistance (HOMA-IR) was calculated following the formula derived by Matthews et al. [31].

Statistical Analysis.
e dataset was structured using EpiData 3.1, and double entry typing with validation was performed.
e data were presented as mean ± SD for continuous variables and frequencies for categorical variables.
e chi-squared test was used to analyze the agreement of genotype frequencies with the Hardy-Weinberg equilibrium expectation. Allele frequency was determined by manual counting. Normal distribution was tested for all measured variables using the Kolmogorov-Smirnov test and histograms; skewed variables were normalized by log transformation and then backtransformed for results presentation.
Individual genotype analysis and the combination of the two genotypes analysis (PPARG2 + IL6) were performed. For the combination of genotypes, patients were grouped as follows: no variants (genotypes ProPro and GG, n � 78), IL6 only (genotypes ProPro and GC + CC, n � 48), PPARG2 only (genotypes ProAla + AlaAla and GG, n � 15), and both variants (genotypes ProAla + AlaAla and GC + CC, n � 4). Comparisons were performed using Student's ttest or ANOVA and chi-squared test or Fisher's exact test. Due to the low frequency of the variant allele, we compared carriers versus noncarriers of the minor allele. We fitted multiple linear regression models to adjust the analysis for potential confounders (age, sex, BMI, sedentary time, and diabetes). For binary variables, odds ratios (ORs) and 95% confidence intervals (95% CIs) were calculated and multiple logistic regression models were fitted adjusting for the same confounders. Statistical analyses were performed in Stata 12.

Results
All the study participants (n � 150) had DNA extracted; however, genetic samples were viable in 146 individuals for the PPARG2 Pro12Ala polymorphism and in 148 individuals for the IL6 -174G >C polymorphism. e genotype distribution for PPARG2 Pro12Ala was 86.9%, 12.4%, and 0.7% for ProPro, ProAla, and AlaAla, respectively. e minor allele frequency of the Ala allele was 0.065. For IL6 -174G >C, the frequencies of GG, GC, and CC genotypes were 65.3%, 31.3%, and 3.4%, respectively. e minor allele frequency of the C allele was 0.193. Observed genotype frequencies were in agreement with the Hardy-Weinberg equilibrium (p � 0.689 for PPARG2 Pro12Ala and p � 0.863 for IL6 -174G >C). e characteristics of the study participants according to the PPARG2 Pro12Ala polymorphism are displayed in Table 1. Analysis of the Pro12Ala polymorphism showed higher BMI (p � 0.031) and fat mass (p � 0.049) for Ala carriers, even after adjustment for age, sex, sedentary time, and diabetes. Ala carriers presented significantly higher SBP and DBP, but after adjustments, only SBP (p � 0.026) remained associated (Table 1). e characteristics of the study participants according to the IL6 -174G >C polymorphism are displayed in Table 2. Sex and MVPA were associated with the -174G >C polymorphism after adjustments (p � 0.043 and p � 0.024, respectively). Males had triple probability (OR: 3.60; 95% CI: 1.04-12.48) to be C carriers, and the C carriers spent lower amount of time in MVPA (Table 2).
For combined genotypes, participants were grouped as follows: no variants (n � 78), IL6 variant only (n � 48), PPARG2 variant only (n � 15), and both variants (n � 4). e same variables of the individual genotype analysis were tested for the combined genotypes, but the results were presented in figures only for the variables with a Journal of Obesity 3 significant statistical difference. Analysis of the combined effects of the two genotypes showed association with BMI, fat mass, SBP, DBP, and polyunsaturated fat consumption after adjustments. Individuals with both variants had higher BMI (p � 0.023) compared to the ones with no variants. Fat mass, SBP, and DBP were higher for participants with the PPARG2 variant only compared to those with no variants (p � 0.045, p � 0.018, and p � 0.030, respectively). Individuals with both variants presented higher consumption of PUFA compared to the ones with no variants (p � 0.045) (Figure 1). e analysis of BMI in categories did not show association with the polymorphisms or the genotype combination (Table 3).

Discussion
e Ala allele of PPARG2 Pro12Ala (rs1801282) is potentially associated with higher BMI; however, few studies have investigated other key factors that may interact with the variant leading to severe obesity susceptibility such as metabolic, food consumption, and clinical parameters. To our knowledge, this is the first study to find association of the PPARG2 Pro12Ala (rs1801282) polymorphism with an obesity phenotype other than BMI and blood pressure in severely obese patients. We found higher measures of adiposity (BMI and fat mass) and higher SBP in Ala carriers of the Pro12Ala polymorphism compared to the ProPro individuals.
ese results may contribute to a better understanding of the pathophysiology of severe obesity and translate into more effective preventive and treatment measures to halt its increasing prevalence.
Despite controversial results regarding the PPARG2 Pro12Ala polymorphism, meta-analyses have shown association of the Ala allele with higher BMI and fat mass, especially in severely obese individuals, corroborating our results [6-8, 10, 11, 40]. e controversies observed in other studies may be explained by a suggested interaction between dietary fat intake and the Pro12Ala polymorphism, showing that when the dietary P : S ratio is similar or lower than 0.66, Ala carriers present higher BMI than Pro homozygotes, while when the P : S ratio is higher than 0.66, the opposite occurs [41]. Our patients had the mean P : S ratio ≤ 0.66, Journal of Obesity 5 possibly explaining the higher BMI among Ala carriers and demonstrating the effect of dietary fat intake on the phenotype determination. PPAR-c2 has a crucial role in modulating lipid metabolism and adipose tissue accumulation [4,5]. e PPARG2 Pro12Ala polymorphism has been suggested to have a role in the variance of fat mass among analyses of candidate genes for adiposity changes [42]. Few studies have investigated the influence of the Pro12Ala polymorphism on body composition parameters other than body mass, BMI, and waist circumference. Association between the Ala allele and higher fat mass was found in Caucasians from the Québec Family   Journal of Obesity Study [43] and in Italian women [44], corroborating the results of the current study. us, more detailed measurements of body composition are important to be addressed in studies of the Pro12Ala polymorphism.
We found an association between the Ala allele of the Pro12Ala polymorphism and higher blood pressure in severely obese individuals. Although the association between the Pro12Ala variants and blood pressure is controversial, a meta-analysis of eight studies with 3281 individuals (1865 cases and 1416 controls) suggested that the Ala allele might be protective for hypertension among East Asians, but not among Caucasians [45]. e mechanisms behind this association are unclear. Studies have suggested that the activation of the PPAR-c2 inhibits processes related to inflammation and hypertension.
e Ala allele has been associated with impaired function of the PPAR-c2, leading to increased blood pressure [46][47][48].
We found higher probability of severely obese males to be C carriers of the IL6 -174G >C polymorphism after adjustments. Studies comparing elderly (octa/nonagenarians and centenarians) with younger individuals have found a tendency in the reduction of GG genotypes in males [49][50][51][52]. Regarding MVPA, we did not find any evidence that justifies this association. More investigation is needed to clarify the association of the higher frequency of C alleles in severely obese males and the -174G >C polymorphism and also the association of the C allele with lower time spent in MVPA.
When the Pro12Ala and the -174G >C genotypes were combined, association was found between the PPARG2 variant only and fat mass, SBP, and DBP compared to the ones with no variants, as demonstrated in the independent analysis of genotypes. We also observed higher BMI and higher PUFA consumption in the presence of both variants compared to the ones with no variants. Nevertheless, this association of the combined genotypes seems to be determined by the PPARG2 Pro12Ala polymorphism as the IL6 -174G >C polymorphism was not associated with adiposity indexes in the independent analysis of genotypes and also due to the low number of individuals with both variants.
Our study has limitations such as the small sample size, especially in the analysis of combined genotypes, and the impossibility to demonstrate causality due to the study design. However, we may address some strengths, such as the analysis of the association between SNPs and also the combination of genotypes, with a variety of parameters, such as anthropometrical, clinical, biochemical, physical activity, and dietary intake, in severely obese patients; the adjusted analysis for potential confounders, such as sociodemographic, anthropometrical, and physical activity measurements; and the high-quality methods used to assess physical activity (triaxial accelerometer) and body composition (BIA).

Conclusions
In summary, we found that severely obese individuals carrying the Ala allele of the PPARG2 Pro12Ala polymorphism have higher measures of adiposity (BMI and fat mass) and blood pressure, while no important associations were found for the IL6 -174G >C polymorphism. Further studies on gene-diet/gene-environment interactions are necessary to clarify the associations and underlying mechanisms between SNPs and severe obesity. is information may eventually be used to develop tailored interventions specific to the individual's genotype and enable more effective prevention and treatment measures to address this increasing public health problem.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request. Disclosure e funder had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Conflicts of Interest
e authors declare that they have no conflicts of interest. Journal of Obesity 7