Effects of Pro12Ala polymorphism in peroxisome proliferator-activated receptor-γ2 gene on metabolic syndrome risk: A meta-analysis

Gene 535 (2014) 79–87

Contents lists available at ScienceDirect

Gene journal homepage: www.elsevier.com/locate/gene

Effects of Pro12Ala polymorphism in peroxisome proliferator-activated receptor-γ2 gene on metabolic syndrome risk: A meta-analysis Ruyi Zhang, 1,2, Jiao Wang 1,3, Rui Yang 4, Jia Sun 5, Rongping Chen 6, Haizhao Luo 7, Duan Liu 8, Dehong Cai ⁎ Department of Endocrinology, Southern Medical University, Zhujiang Hospital, 253# Industry Road, 510282 Guangzhou, Guangdong, China

a r t i c l e

i n f o

Article history: Accepted 23 July 2013 Available online 5 September 2013 Keywords: PPARγ Polymorphism Metabolic syndrome Meta-analysis

a b s t r a c t Background: Associations between peroxisome proliferator-activated receptor γ2 (PPARγ2) gene polymorphism and metabolic syndrome risk remained controversial and ambiguous. Thus, we performed a meta-analysis to assess the association between Pro12Ala polymorphism in PPARγ2 gene and metabolic syndrome susceptibility. Methods: An electronic literature search was conducted on Medline, OVID, Cochrane Library database, and the China National Knowledge Internet up to March 2013. Odds ratios (ORs) with 95% confidence intervals (CIs) were used to calculate the strength of association in the fixed or random effects model. Results: Ten studies involving a total of 4456 cases and 10343 controls were included in this meta-analysis. No statistical evidence of association was found between Pro12Ala polymorphism and metabolic syndrome risk in all genetic models (homozygote model: OR = 0.83, 95% CI = 0.62–1.12; heterozygote model: OR = 1.04, 95% CI = 0.94–1.14; dominant model: OR = 1.02, 95% CI = 0.93–1.12; recessive model: OR = 0.83, 95% CI = 0.62–1.11). No statistical evidence of significant association was observed when stratified by ethnicity, definition of metabolic syndrome, source of control groups and quality score of the selected articles. All in all, the results did not support a major role of the Pro12Ala variant of the PPARγ2 gene in metabolic syndrome risk. Conclusions: This meta-analysis suggested that the effect of Pro12Ala polymorphism in PPARγ2 gene may not be related to metabolic syndrome as an entity. However, Pro12Ala may affect the single component of metabolic syndrome. A large, well designed study is required to more adequately assess the role for Pro12Ala polymorphism on metabolic syndrome. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Abbreviations: PPARγ2 (gamma2), peroxisome proliferator-activated receptor γ2 (gamma2); OR, odds ratio; CI, confidence interval; Pro12Ala, substitution of proline to alanine; Ala, alanine; TC, total cholesterol; LDL-C, low density lipoprotein cholesterol; HDL-C, high density lipoprotein cholesterol; TG, triglyceride; BMI, body mass index; HWE, Hardy–Weinberg equilibrium; PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-analysis; MOOSE, Meta-analysis of Observational Studies in Epidemiology; NCEP ATP III, National Cholesterol Education Program Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adult; EGIR, European Group for the study of Insulin Resistance; HOMA, Homeostasis Model Assessment; WHO, World Health Organization; IDF, International Diabetes Federation; PB, population-based; HB, hospital-based; NOS, Newcastle–Ottawa Scale; SNP, single nucleotide polymorphism. ⁎ Corresponding author. Tel.: +86 13808843648; fax: +86 20 61643036. E-mail addresses: [email protected] (R. Zhang,), [email protected] (J. Wang), [email protected] (R. Yang), [email protected] (J. Sun), [email protected] (R. Chen), [email protected] (H. Luo), [email protected] (D. Liu), [email protected] (D. Cai). 1 These authors contributed equally to this work. 2 Tel.:+86 015920941699; fax: +86 020 61643036. 3 Tel.: +86 015989271179; fax: +86 020 61643036. 4 Tel.: +86 013724004245; fax: +86 020 61643036. 5 Tel.: +86 13751822925; fax: +86 20 61643036. 6 Tel.: +86 13632101107; fax: +86 20 61643036. 7 Tel.: +86 18819419199; fax: +86 20 61643036. 8 Tel.: +86 13824403698; fax: +86 20 61643036. 0378-1119/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.gene.2013.07.087

Metabolic syndrome, characterized by clustering of chronic degenerative disorders such as abdominal obesity, glucose intolerance and atherogenic dyslipidemia and it confers an increased risk for diabetes and cardiovascular diseases (Robitaille et al., 2003). It is estimated that the prevalence of metabolic syndrome in adult was around 20%–25% all over the world (Sun et al., 2012). The current concept of the development of metabolic syndrome is that environmental factors add to an underlying genetic preponderance to insulin resistance (Hong et al., 1997; Mayer et al., 1996). Many genes are candidates for metabolic syndrome. Among the genes potentially involved in metabolic syndrome, peroxisome proliferator-activated receptor gamma (PPARγ) is a nuclear receptor that regulates adipocytes differentiation and lipid storage, and modulates the action of insulin, and therefore it has been intensively studied in relation to diabetes, insulin resistance and coronary artery diseases (Rhee et al., 2007; Sundvold and Lien, 2001). The PPARγ gene spans about 100 kb, located on chromosome 3p25 and is composed of 9 exons with the exons 1–6 being the common region (Fajas et al., 1997) and it produces four mRNA by alternative splicing and promoter usage, giving rise to two different proteins that are distributed selectively in adipose cells (Fajas et al., 1997, 1998; Ricote et al., 1998; Sundvold and

80

R. Zhang, et al. / Gene 535 (2014) 79–87

Lien, 2001; Tontonoz et al., 1994). The most frequently occurring PPARγ polymorphism is substitution of proline to alanine (Pro12Ala) at codon 12 of the exon B of the PPARγ gene, which encodes the amino-terminal domain defining the adipocytes-specific PPARγ2, with a minor allelic frequency ranging from 2% to 23% in different ethnic groups (Stumvoll and Haring, 2002). Some meta-analysis had reported the association between PPARγ2 polymorphism and the component of metabolic syndrome. Gouda et al. (2010) reported that Pro12Ala polymorphism was associated with a reduction in type 2 diabetes risk. Wang and Liu (2012) had demonstrated the significant association of Pro12Ala polymorphism with hypertension susceptibility among East Asians, but the association was not found in Caucasians. Huang et al. (2011) performed a meta-analysis to assess the association between Pro12Ala polymorphism and lipid profile in East Asians, Caucasians and other ethnic origins. In that study (Huang et al., 2011), the overall association of the Pro12Ala polymorphism with TC, LDL-C or HDL-C was not observed after excluding the studies not in HWE, but lower level of TG was only confirmed in subjects with Ala allele in Caucasian even after exclusion of the outlier studies. Masud and Ye (2003) detected that Ala allele carriers had a significantly higher BMI than non-carriers in a sample with a mean BMI value ≥27, but the difference was not detected in the samples with a mean BMI value b27. Although many studies have been performed on the association between this polymorphism and the components of metabolic syndrome, the significance of such associations remains an issue. On the other hand, metabolic syndrome comprises different components and some component could not represent the whole syndrome. Previously existing literature and primary analyses on the association of polymorphism Pro12Ala with the risk of metabolic syndrome as an entirety were inconsistent. Therefore, the aim of the present meta-analysis is to assess metabolic syndrome risk of the polymorphism, and quantitatively analyze the potential influencing factors, with an expectation to provide the most comprehensive evidence for the relationship between this variant and metabolic syndrome susceptibility.

2. Methods 2.1. Search strategy We conducted a meta-analysis of the published works with English or Chinese language restrictions and in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) criteria (Moher et al., 2009) and Meta-analysis of Observational Studies in Epidemiology (MOOSE) guidelines (Stroup et al., 2000). We searched Medline (1966–March 2013), OVID (1966–March 2013), the Cochrane Library database (1996–March 2013) and the China National Knowledge

Internet (1979–March 2013). The search used the following keyword strings: “PPAR gamma”, “peroxisome proliferator activated receptor gamma”, “metabolic syndrome”, “insulin resistance syndrome”, “metabolic syndrome X”, “syndrome X”, “polymorphism”, “mutation”, “variant”, “gene”, “genotype”, “SNP”, and “allele”. Reference list from published original articles and previous reviews were scanned for more relevant studies not identified in the database search. Detail search strategy is available at Appendix S1. 2.2. Inclusion and exclusion criteria Studies were included in the meta-analysis if they met the following criteria: (1). A case–control or cohort study published as an original study evaluating the association of the Pro12Ala polymorphism in PPARγ2 with the risk of metabolic syndrome; (2). metabolic syndrome was defined by standard criteria such as the National Cholesterol Education Program Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adult (NCEP ATP III), the European Group for the Study of Insulin Resistance, the International Diabetes Federation criteria and others; and (3). numbers in case and control groups or exposed and unexposed groups were reported for each genotype or data provided from which numbers could be calculated. The major exclusion criteria were: (1). Only case population and duplicate of previous publication; (2). family-based control; and (3). genotype distribution of the control population was deviated from Hardy–Weinberg equilibrium (HWE). 2.3. Data Extraction Data were independently extracted in duplicate by two investigators (Ruyi Zhang, Jiao Wang) using a data-collecting form according to the inclusion criteria. Original extraction data were checked by another investigator (Dehong Cai), and any disagreement was resolved by discussion among the three investigators. The following information was collected from each study: first author's name, year of publication, ethnicity, country of the selected subjects, source of the control groups, definition of metabolic syndrome, frequencies of genotypes in both groups and genotyping methods. Diverse ethnicity descents were categorized as Asian, Caucasian, and African. Definitions of metabolic syndrome were as follows: (1) EGIR: The European Group for the study of insulin resistance criteria was defined by the presence of insulin resistance (highest quartile of HOMA values) and two of the following: hyperglycemia (fasting plasma glucose ≥ 6.1 mmol/l, but nondiabetic); hypertension (systolic/ diastolic N 140/90 mm Hg or treated for hypertension); dyslipidemia (triglyceride N 2.0 mmol/l or HDL-C b 1.0 mmol/l or treated for dyslipidemia); and central obesity (waist circumference ≥ 94 cm in men and ≥80 cm in women). (2) NCEP ATP III: National Cholesterol

Table 1 Characteristics of studies included in the meta-analysis. Author

Year

Ethnicity

Country of the Control Definitiona subjects sourceb

Cases/controls Cases

Controls

Pro/Pro Pro/Ala Ala/Ala Pro/Pro Pro/Ala Ala/Ala

HWE Genotyping method

Frederiksen L Meirhaeghe A Rhee E J Vimaleswaran KS Liu Dongxia Montagnana M Ranjith N Haseeb A

2002 2005 2006 2007

Caucasian Caucasian Asian Asian

Denmark France Korea India

PB PB HB PB

EGIR NCEP ATP III Modified NCEP ATP III Modified NCEP ATP III

294/1951 267/838 44/209 613/887

222 207 41 503

70 60 3 104

2 7 0 6

1449 666 185 718

448 160 24 161

54 12 0 8

Yes Yes Yes Yes

PCR-RFLP PCR-RFLP TaqMan PCR-RFLP

2008 2008 2008 2009

Asian Caucasian African Asian

China Sweden South Africa India

HB PB HB HB

331/461 1165/4176 282 460/239

302 852 229 348

29 294 50 102

0 19 3 10

418 3035 150 190

43 1038 33 47

0 103 2 2

Yes Yes Yes Yes

TaqMan TaqMan TaqMan PCR-RFLP

Tellechea M L Shi Hui

2009 Caucasian Argentina 2012 Asian China

IDF NCEP ATP III/othersa NCEP ATP III/IDFa Modified NCEP ATP III/IDFa NCEP ATP III IDF

150/422 850/975

120 740

30 85

0 25

367 851

55 94

0 30

Yes Yes

PCR-RFLP TaqMan

a b

PB PB

These definitions were confirmed by NCEP ATP III or Modified NCEP ATP III. Control source: PB: population-based; HB: hospital-based.

R. Zhang, et al. / Gene 535 (2014) 79–87

81

Table 2 Assessment of study quality included in the meta-analysis. Source

L Frederiksen (2002) A Meirhaeghe (2005) E J Rhee (2006) K S Vimaleswaran (2007) Liu Dongxia (2008) M Montagnana (2008) N Ranjith (2008) A Haseeb (2009) M L Tellechea (2009) Shi Hui (2012)

Comparability

Exposure

1

Selection 2

3

4

1A

1B

1A

1B

2

3

Total scores

☆ ☆ ☆ ☆ ☆ ☆ ☆ ☆ ☆ ☆

☆ – – – – ☆ – ☆ – ☆

☆ ☆ – ☆ – ☆ – – ☆ ☆

☆ ☆ ☆ ☆ ☆ ☆ ☆ ☆ ☆ ☆

☆ ☆ ☆ ☆ ☆ ☆ ☆ ☆ ☆ ☆

☆ ☆ – – – – ☆ – – –

☆ ☆ ☆ ☆ ☆ ☆ ☆ ☆ ☆ ☆

– – – – – – – – – –

☆ ☆ ☆ ☆ ☆ ☆ ☆ ☆ ☆ ☆

☆ ☆ ☆ ☆ ☆ ☆ ☆ ☆ ☆ ☆

9 8 6 7 6 8 7 7 7 8

Selection: 1. Is the case definition adequate; 2. representativeness of the cases; 3. selection of controls and 4. definition of controls. Comparability: 1. Comparability of cases and controls on the basis of design or analysis. Exposure: 1. Ascertainment of exposure; 2. same method of ascertainment for cases and controls; and 3. non-response rate.

Education Program Expert Panel on Detection, Evaluation and Treatment of High Blood Cholesterol in Adults was defined by the presence of at least three or more of the following abnormalities: waist circumference N102 cm in men and N88 cm in women; triglycerides ≥ 1.50 g/l or HDL-C b 40 mg/dl in men or b50 mg/dl in women or treated for dyslipidemia; blood pressure ≥ 130/85 mm Hg or treatment with medication; and fasting glucose ≥ 1.10 g/l or treated for diabetes.(3) Modified NCEP ATP III: The waist circumference in NCEP ATP III was modified by the criteria of Asia Pacific WHO guidelines or IDF criteria.(4) IDF: The metabolic syndrome was defined by central obesity (waist circumference cutoff was different in different ethnicity) and the presence of any two or more of the following: triglycerides N 150 mg/dl or HDL-C b 40 mg/dl in men or b50 mg/dl in women or treated for dyslipidemia; blood

pressure ≥ 130/85 mm Hg or previously diagnosed with hypertension; and fasting glucose N 100 m g/dl or previously diagnosed with type 2 diabetes. Definition of metabolic syndrome was stratified to NCEP ATP III and other modified criteria including EGIR, modified NCEP ATP III, and IDF criteria. Study design was stratified to populationbased (PB) studies and hospital-based (HB) studies. If necessary data were not reported in the primary manuscripts, we contacted the corresponding authors by email to request the missing data. 2.4. Quality score assessment The qualities of included studies were assessed using the Newcastle– Ottawa Scale (NOS) (Wells et al., 2011). The NOS uses a “☆” rating

Fig. 1. Forest plot of metabolic syndrome risk associated with Pro12Ala polymorphism under homozygote comparison in overall study. Three studies had insufficient data about homozygote model. The squares and horizontal lines correspond to the study-specific OR and 95% CI. The area of the squares reflects the study specific weight. The diamond represents the pooled OR and 95% CI.

82

R. Zhang, et al. / Gene 535 (2014) 79–87

Fig. 2. Forest plot of metabolic syndrome risk associated with Pro12Ala polymorphism under heterozygote comparison in overall study. The squares and horizontal lines correspond to the study-specific OR and 95% CI. The area of the squares reflects the study specific weight. The diamond represents the pooled OR and 95% CI.

system to judge quality based on three aspects of the case–control study: selection, comparability and exposure. Scores were ranged from 0 to 9. We assessed the quality of the studies in a consensus meeting with all authors (the criteria of NOS were presented in Appendix S2). Studies with a score ≥7 were considered to be of high quality.

2.5. Statistical analysis The strength of the association between the Pro12Ala polymorphism and metabolic syndrome risk was measured by odds ratios (ORs) with 95% confidence intervals (CIs). The statistical significance of the pooled OR was determined using the Z-test. Pooled estimates of the OR were obtained by calculating a weighted average of OR from each study. The Pro12Ala genotypes included Pro/Pro, Pro/Ala, and Ala/Ala. The overall pooled analysis was performed for homozygote model (Ala/Ala versus Pro/Pro), heterozygote model (Pro/Ala versus Pro/Pro), dominant model (Pro/Ala and Ala/Ala versus Pro/Pro), and the recessive model (Ala/Ala versus Pro/Ala and Pro/Pro), respectively. Subgroup analyses were according to ethnicity (Asian/Caucasian/African), definition of metabolic syndrome (NCEP ATP III/other modified criteria), source of the control (PB/HB) and study quality (high, NOS score ≥ 7/ low, NOS scores b 7). In consideration of the possibility of heterogeneity across the studies, I2 was applied to assess heterogeneity between studies. If Phet b 0.05 or I2 N 50%, it was considered substantial heterogeneity among studies, and the random-effects model (the DerSimonian and Laird method) was applied as the preferred method for estimating the summary ORs and 95% CIs; the fixed-effects model (the Mantel– Haenszel method) was used when there was no substantive heterogeneity. Moreover, sensitivity analysis, by which a single study in the metaanalysis was deleted each time to determine the influence of the

individual data set for the overall pooled OR, was performed to assess the stability of the results. To assess the potential publication bias, visual inspection of Begg's funnel plot symmetry was performed. Egger's test was also conducted to analyze the publication bias statistically. The meta-analysis was conducted using Stata software (version 12.0; StataCorp LP, College Station, Texas), with all the P values were two-sided. 3. Results 3.1. Characteristics of studies The present study met the PRISMA statement (Appendix S3) and MOOSE guidelines (Appendix S4) and the study selection process is detailed in Appendix S5. Based on our preliminary search criteria, a total of ten publications were eligible (Dongxia et al., 2008; Frederiksen et al., 2002; Haseeb et al., 2009; Meirhaeghe et al., 2005; Montagnana et al., 2008; Ranjith et al., 2008; Rhee et al., 2006; Shi et al., 2012; Tellechea et al., 2009; Vimaleswaran et al., 2007). Among these articles, five focused on Asians, four on Caucasians, and one on Africans. The countries of these studies included India, China, Korea, France, Sweden, Argentina, and Denmark. All studies were case–control in design. Table 1 showed the studies identified and their main characteristics. Assessment of the study specific quality scores from NOS system was showed in Table 2. The median score of included studies was 7, with a range from 6 to 9, and 80.0% of the studies were identified as relatively high-quality. 3.2. Quantitative synthesis Fixed effect models were adopted as no heterogeneity was observed among studies for all the genetic models (I2 = 36.6%, Phet = 0.149;

R. Zhang, et al. / Gene 535 (2014) 79–87

I2 = 0.0%, Phet = 0.655; I2 = 0.0%, Phet = 0.467; I2 = 34.6%, Phet = 0.164; respectively). When all ten studies including 4456 cases and 10,343 controls were pooled into the meta-analysis, there was no statistical evidence of association between Pro12Ala polymorphism and overall risk of metabolic syndrome (homozygote model: OR = 0.83, 95% CI = 0.62–1.12; heterozygote model: OR = 1.04, 95% CI = 0.94– 1.14; dominant model: OR = 1.02, 95% CI = 0.93–1.12; recessive model: OR = 0.83, 95% CI = 0.62–1.11) (Figs. 1, 2, 3, 4). In a stratified analysis by specific ethnicity, definition of metabolic syndrome, source of control groups or quality score, no relationship between polymorphism and metabolic syndrome was observed in all genetic models (Table 3).

83

3.5. Publication bias Begg's funnel plot and Egger's test were performed to assess the publication bias of the selected articles. As shown in Figs. 5, 6, 7, 8, the shapes of the funnel plots did not reveal any evidence of an obvious asymmetry in all comparison models. Moreover, Egger's test was used to provide statistical evidence of funnel plot symmetry. The results still did not show any evidence of publication bias (P = 0.597 for homozygote model; P = 0.716 for heterozygote model; P = 0.543 for dominant model; P = 0.602 for recessive model, respectively).

4. Discussion 3.3. Test for heterogeneity For the overall comparisons, no significant heterogeneity among studies was observed, which suggested that there was no substantial heterogeneity between studies, except significant heterogeneity in the stratified analysis among Caucasian population (homozygote model: I2 = 69.4%, Phet = 0.038; dominant model: I2 = 50.5%, Phet = 0.108; recessive model: I2 = 68.1%, Phet = 0.043, respectively), for which the random-effects model was conducted in the pooled analysis. 3.4. Sensitivity analyses In the sensitivity analyses, the influence of each study on the pooled OR was checked by repeating the meta-analysis while omitting each study, one at a time. The corresponding pooled ORs were not materially altered, indicating that our results were statistically robust.

The metabolic syndrome is composed of different metabolic and hemodynamic disturbance presumed to be determined by a common pattern of insulin resistance that results in their clustering together (Cai et al., 2012; Shen et al., 2012). Metabolic syndrome is a multifactorial complex trait that is influenced by both environmental and genetic factors. Evidence from several studies supports the role of genetic factors in the development of metabolic syndrome (Bayoglu et al., 2013; Chiefari et al., 2013; Fernandez et al., 2012; Song et al., 2013; Wang et al., 2012). Among the genes potentially involved, peroxisome proliferator-activated receptor gamma (PPARγ) is a nuclear receptor that regulates adipocyte differentiation and possibly lipid metabolism and can therefore be a key regulator of fat storage. A variant in PPARγ2 and Pro12Ala, which results in the substitution of an alanine for a proline at amino acid 12, decreased PPARγ activity (Deeb et al., 1998; Yen et al., 1997). Recently, a variety of studies have focused on the association between Pro12Ala polymorphism in PPARγ2 gene and metabolic syndrome

Fig. 3. Forest plot of metabolic syndrome risk associated with Pro12Ala polymorphism under dominant comparison in overall study. The squares and horizontal lines correspond to the study-specific OR and 95% CI. The area of the squares reflects the study specific weight. The diamond represents the pooled OR and 95% CI.

84

R. Zhang, et al. / Gene 535 (2014) 79–87

Fig. 4. Forest plot of metabolic syndrome risk associated with Pro12Ala polymorphism under recessive comparison in overall study. Three studies had insufficient data about recessive model. The squares and horizontal lines correspond to the study-specific OR and 95% CI. The area of the squares reflects the study specific weight. The diamond represents the pooled OR and 95% CI.

showed that homozygosity of Ala allele confers a decreased risk of insulin resistance syndrome (Frederiksen et al., 2002). However, Tellechea et al. (2009) reported that the Pro12Ala polymorphism increased the risk of metabolic syndrome. For overall comparison of pooled ORs, our meta-analysis did not show significant association between all genetic models and metabolic syndrome susceptibility. Possible explanation may be that: (1) Only

or its components (Bego et al., 2011; Jurkowski et al., 2011; Passaro et al., 2011; Yang et al., 2009). However, the conclusions of these studies were inconsistent. Consequently, meta-analysis is needed to provide a quantitative approach for combining the different results. The present meta-analysis, which included 4456 cases and 10,343 controls from 10 case–control studies, explored the relationship between the Pro12Ala polymorphism and metabolic syndrome risk. Frederiksen

Table 3 Stratified analyses of the Pro12Ala polymorphism on metabolic syndrome risk. Variables

N

Ala/Ala versus Pro/Pro OR (95% CI)

Overall 10 0.83 (0.62–1.12) Ethnicity Caucasian 4 0.74 (0.29–1.89)a Asian 5 1.10 (0.70–1.73) African 1 0.98 (0.16–5.95) Definition of metabolic syndromeb NCEP ATP III 4 0.80 (0.53–1.22) Others 6 0.86 (0.57–1.30) Source of control groups Population-based 6 0.78 (0.57–1.06) Hospital-based 4 1.88 (0.61–5.85) Quality scorec High (≥7) 8 0.83 (0.62–1.12) Low (b7) 2 – a

Pro/Ala versus Pro/Pro 2

Pro/Ala + Ala/Ala versus Pro/Pro 2

Phet

I (%)

OR (95% CI)

Phet

I (%)

OR (95% CI)

0.149

36.6

1.04 (0.94–1.14)

0.655

0.0

1.02 (0.93–1.12) a

2

Ala/Ala versus Pro/Ala + Pro/Pro

Phet

I (%)

OR (95% CI)

0.467

0.0

0.83 (0.62–1.11) a

Phet

I2 (%)

0.164

34.6

0.038 0.446 –

69.4 0.0 –

1.06 (0.94–1.20) 0.99 (0.84–1.17) 0.99 (0.61–1.61)

0.227 0.732 –

30.8 0.0 –

1.09 (0.89–1.33) 1.00 (0.85–1.17) 0.99 (0.62–1.59)

0.108 0.643 –

50.5 0.0 –

0.73 (0.29–1.83) 1.10 (0.70–1.72) 0.98 (0.16–5.95)

0.043 0.467 –

68.1 0.0 –

0.151 0.134

47.0 46.2

1.07 (0.94–1.21) 1.00 (0.86–1.15)

0.228 0.843

30.7 0.0

1.05 (0.93–1.19) 0.99 (0.86–1.13)

0.135 0.749

46.1 0.0

0.80 (0.52–1.21) 0.86 (0.57–1.29)

0.170 0.138

43.6 45.5

0.121 0.394

45.2 0.0

1.04 (0.94–1.15) 1.02 (0.80–1.32)

0.386 0.666

4.8 0.0

1.01 (0.92–1.12) 1.05 (0.82–1.34)

0.253 0.568

24.1 0.0

0.78 (0.57–1.05) 1.84 (0.59–5.72)

0.130 0.411

43.7 0.0

0.149 –

36.6 –

1.05 (0.95–1.15) 0.86 (0.55–1.36)

0.575 0.461

0.0 0.0

1.03 (0.93–1.13) 0.86 (0.55–1.36)

0.364 0.461

8.6 0.0

0.83 (0.62–1.11) –

0.164 –

34.6 –

These results were obtained from random-effect model test. NCEP ATP III: National Cholesterol Education Program Expert Panel on Detection, Evaluation and Treatment of High Blood Cholesterol in Adults; others: EGIR, modified NCEP ATP III, and IDF criteria. c Quality score: the score assessed by NOS scale. b

R. Zhang, et al. / Gene 535 (2014) 79–87

85

Fig. 5. Funnel plot for homozygote model in overall study. Each point represents a separate study for the indicated association. Log OR, natural logarithm of OR. Horizontal line, mean effect size.

Fig. 7. Funnel plot for dominant model in overall study. Each point represents a separate study for the indicated association. Log OR, natural logarithm of OR. Horizontal line, mean effect size.

ten articles were included in this meta-analysis, thus the combined sample sizes were still inadequate to detect the association between Pro12Ala polymorphism and metabolic syndrome. The frequency of Ala allele was so small that several studies did not show the people with homozygote (Dongxia et al., 2008; Rhee et al., 2006; Tellechea et al., 2009). And the frequency of heterozygote may be also not large enough to explain the relationship. (2) It was difficult to get the full papers published in various languages; we only included the studies published in English and Chinese. Undoubtedly, the limitation may affect our final conclusions. (3) Other polymorphism in PPARγ gene such as C161T and C1431T may have an impact on the susceptibility of metabolic syndrome (Dongxia et al., 2008; Meirhaeghe et al., 2005). The present study only focused on the variant of Pro12Ala but omitted the other possible related polymorphism. Nevertheless, the Pro12Ala polymorphism on metabolic syndrome might be confounded by other SNPs, so further studies were required. (4) Metabolic syndrome is not well defined, being composed of different metabolic and hemodynamic disturbances presumed be determined by a common pattern of insulin resistance that results in their clustering together. However, the different contribution of the various components in diverse individuals makes this syndrome quite heterogeneous and its definition hazardous. Considerable research has shown that Pro12Ala polymorphism was associated with BMI (Passaro et al., 2011), adipose tissue distribution (Franks et al., 2007; Passaro et al., 2011), lipid profile, obesity (Ben Ali et al., 2009; Bhatt et al., 2012), insulin resistance (Estivalet et al., 2011),

hypertension (Gao et al., 2010), and type 2 diabetes (Lamri et al., 2012). (5) Environmental factors could influence the impact of the polymorphism. Luan et al. (2001) reported a strong interaction between the ratio of dietary polyunsaturated fat to saturated fat and the PPARγ Pro12Ala polymorphism on BMI and fasting insulin. Robitaille et al. (2003) showed that a greater response to fat or saturated fat intake in Pro/Pro homozygote and the variant may influence the magnitude of the association between fat intake and components of the insulin resistance syndrome. Lamri et al. (2012) showed that there was a significant genetic and nutritional interaction on BMI and type 2 diabetes at PPARγ gene in general population. Ruchat SM et al. (2010) demonstrated that Ala allele experienced greater improvements in glucose and insulin metabolism in response to regular endurance training. It seemed that nutrition and exercise may be related to genetic variant to alleviate the metabolic disturbance. At the same time, no statistical evidence of significant association was observed when stratified by ethnicity, definition of metabolic syndrome, source of control groups and quality score of the selected articles. There were some reasons for that: (1) There were only ten studies included in our meta-analysis. Five focused on Asians, four on Caucasians, and one on Africans. The small size limited the quality of stratified statistical results. Meanwhile, only one study was about the mutation on African population, and the frequencies of the genotype or alleles in different ethnicities were not the same (Stumvoll and Haring, 2002), which may limit the efficacy of the results. (2) It should be

Fig. 6. Funnel plot for heterozygote model in overall study. Each point represents a separate study for the indicated association. Log OR, natural logarithm of OR. Horizontal line, mean effect size.

Fig. 8. Funnel plot for recessive model in overall study. Each point represents a separate study for the indicated association. Log OR, natural logarithm of OR. Horizontal line, mean effect size.

86

R. Zhang, et al. / Gene 535 (2014) 79–87

noted that metabolic syndrome was not well defined and different criteria may lead to inconsistencies (Chackrewarthy et al., 2013; Sirdah et al., 2012). In our meta-analysis, metabolic syndrome was defined through several approaches, which limited the suitability. (3) The source of control groups in some articles (Dongxia et al., 2008; Haseeb et al., 2009; Ranjith et al., 2008; Rhee et al., 2006) was hospital-based population, which may have an impact on the results. (4) Population from different countries may have different dietary habit and lifestyle, which may be related to genetic variant to alleviate the metabolic syndrome. (5) The additional factors (e.g., age, gender) that may influence the study should have been analyzed; however, such variables are difficult to model with our metaanalysis. (6)Although publication bias was absent in our study, and several international databases were searched, we cannot rule out that the missing articles published in other languages existed and contributed to the final results. All in all, as PPARγ is specific to adipose tissue and regulates adipocytes differentiation, loss-of-function mutations in PPARγ had been associated with a phenotype resembling that of the metabolic syndrome, including insulin resistance, lipodystrophy, and hypertension (Barroso et al., 1999). However, no relationship was observed between PPARγ mutation and metabolic syndrome in present meta-analysis. For better interpreting the results, some limitations of this meta-analysis should be acknowledged. Firstly, the small number of studies and sample size limited the ability to draw more solid conclusions. Secondly, this meta-analysis was based on limited case–control studies, which reminded us that high risk of potential bias might exist, and cohort studies were required for further research. Thirdly, lacking of original data limited our further evaluation of potential interactions among gene–gene and gene–environment. Nonetheless, the present study is the first meta-analysis of the relationship between Pro12Ala polymorphism and metabolic syndrome risk. Our results indicated that no significant statistical difference was observed between the variant and metabolic syndrome, even if stratified by ethnicity, definition of metabolic syndrome, source of control groups, and quality score of selected studies. However, this conclusion should be interpreted with caution due to the small sample size and the inclusion of case–control studies. Larger sample-size studies with well-matched controls and follow-ups are required. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.gene.2013.07.087. Conflict of interest None declared. Acknowledgments The authors thank Zikan Sun M.D. at Cardiovascular Department, Zhujiang Hospital, Southern Medical University for assisting in the statistical analysis. References Barroso, I., et al., 1999. Dominant negative mutations in human PPARgamma associated with severe insulin resistance, diabetes mellitus and hypertension. Nature 402, 880–883. Bayoglu, B., et al., 2013. Chromosome 9p21 rs10757278 polymorphism is associated with the risk of metabolic syndrome. Mol Cell Biochem 379, 77–85. Bego, T., et al., 2011. Association of PPARG and LPIN1 gene polymorphisms with metabolic syndrome and type 2 diabetes. Med. Glas. (Zenica) 8, 76–83. Ben Ali, S., et al., 2009. Gender-specific effect of Pro12Ala polymorphism in peroxisome proliferator-activated receptor gamma-2 gene on obesity risk and leptin levels in a Tunisian population. Clin. Biochem. 42, 1642–1647. Bhatt, S.P., et al., 2012. Ala/Ala genotype of Pro12Ala polymorphism in the peroxisome proliferator-activated receptor-gamma2 gene is associated with obesity and insulin resistance in Asian Indians. Diabetes Technol. Ther. 14, 828–834. Cai, H., et al., 2012. Prevalence and determinants of metabolic syndrome among women in Chinese rural areas. PLoS One 7, e36936.

Chackrewarthy, S., et al., 2013. A comparison between revised NCEP ATP III and IDF definitions in diagnosing metabolic syndrome in an urban Sri Lankan population: the Ragama health study. ISRN Endocrinol. 2013, 320176. Chiefari, E., et al., 2013. A polymorphism of HMGA1 is associated with increased risk of metabolic syndrome and related components. Sci. Rep. 3, 1491. Deeb, S.S., et al., 1998. A Pro12Ala substitution in PPARgamma2 associated with decreased receptor activity, lower body mass index and improved insulin sensitivity. Nat. Genet. 20, 284–287. Dongxia, L., Qi, H., Lisong, L., Jincheng, G., 2008. Association of peroxisome proliferatoractivated receptorgamma gene Pro12Ala and C161T polymorphisms with metabolic syndrome. Circ. J. 72, 551–557. Estivalet, A.A., et al., 2011. D2 Thr92Ala and PPARgamma2 Pro12Ala polymorphisms interact in the modulation of insulin resistance in type 2 diabetic patients. Obesity (Silver Spring) 19, 825–832. Fajas, L., et al., 1997. The organization, promoter analysis, and expression of the human PPARgamma gene. J. Biol. Chem. 272, 18779–18789. Fajas, L., Fruchart, J.C., Auwerx, J., 1998. PPARgamma3 mRNA: a distinct PPARgamma mRNA subtype transcribed from an independent promoter. FEBS Lett. 438, 55–60. Fernandez, E., Carrizo, E., Connell, L., Baptista, T., 2012. Pro12Ala polymorphism of the PPAR-gamma2 gene, metabolic syndrome and response to metformin in clozapinetreated patients. Schizophr. Res. 137, 262–263. Franks, P.W., et al., 2007. The Pro12Ala variant at the peroxisome proliferator-activated receptor gamma gene and change in obesity-related traits in the Diabetes Prevention Program. Diabetologia 50, 2451–2460. Frederiksen, L., et al., 2002. Comment: studies of the Pro12Ala polymorphism of the PPAR-gamma gene in the Danish MONICA cohort: homozygosity of the Ala allele confers a decreased risk of the insulin resistance syndrome. J. Clin. Endocrinol. Metab. 87, 3989–3992. Gao, L., Wang, L., Yun, H., Su, L., Su, X., 2010. Association of the PPARgamma2 gene Pro12Ala variant with primary hypertension and metabolic lipid disorders in Han Chinese of Inner Mongolia. Genet. Mol. Res. 9, 1312–1320. Gouda, H.N., Sagoo, G.S., Harding, A.H., Yates, J., Sandhu, M.S., Higgins, J.P., 2010. The association between the peroxisome proliferator-activated receptor-gamma2 (PPARG2) Pro12Ala gene variant and type 2 diabetes mellitus: a HuGE review and metaanalysis. Am. J. Epidemiol. 171, 645–655. Haseeb, A., et al., 2009. Single-nucleotide polymorphisms in peroxisome proliferatoractivated receptor gamma and their association with plasma levels of resistin and the metabolic syndrome in a south Indian population. J. Biosci. 34, 405–414. Hong, Y., Pedersen, N.L., Brismar, K., de Faire, U., 1997. Genetic and environmental architecture of the features of the insulin-resistance syndrome. Am. J. Hum. Genet. 60, 143–152. Huang, X., Zhao, J., Zhao, T., 2011. Effects of peroxisome proliferator activated receptorgamma 2 gene Pro12Ala polymorphism on fasting blood lipids: a meta-analysis. Atherosclerosis 215, 136–144. Jurkowski, W., Roomp, K., Crespo, I., Schneider, J.G., Del Sol, A., 2011. PPARgamma population shift produces disease-related changes in molecular networks associated with metabolic syndrome. Cell Death Dis. 2, e192. Lamri, A., et al., 2012. 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. Int. J. Obes. (Lond) 36, 218–224. Luan, J., et al., 2001. Evidence for gene–nutrient interaction at the PPARgamma locus. Diabetes 50, 686–689. Masud, S., Ye, S., 2003. Effect of the peroxisome proliferator activated receptor-gamma gene Pro12Ala variant on body mass index: a meta-analysis. J. Med. Genet. 40, 773–780. Mayer, E.J., et al., 1996. Genetic and environmental influences on insulin levels and the insulin resistance syndrome: an analysis of women twins. Am. J. Epidemiol. 143, 323–332. Meirhaeghe, A., Cottel, D., Amouyel, P., Dallongeville, J., 2005. Association between peroxisome proliferator-activated receptor gamma haplotypes and the metabolic syndrome in French men and women. Diabetes 54, 3043–3048. Moher, D., Liberati, A., Tetzlaff, J., Altman, D.G., 2009. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ 339, b2535. Montagnana, M., et al., 2008. The Pro12Ala polymorphism of the PPARG gene is not associated with the metabolic syndrome in an urban population of middle-aged Swedish individuals. Diabet. Med. 25, 902–908. Passaro, A., et al., 2011. PPARgamma Pro12Ala and ACE ID polymorphisms are associated with BMI and fat distribution, but not metabolic syndrome. Cardiovasc. Diabetol. 10, 112. Ranjith, N., Pegoraro, R.J., Naidoo, D.P., Shanmugam, R., Rom, L., 2008. Genetic variants associated with insulin resistance and metabolic syndrome in young Asian Indians with myocardial infarction. Metab. Syndr. Relat. Disord. 6, 209–214. Rhee, E.J., et al., 2006. Effects of two common polymorphisms of peroxisome proliferatoractivated receptor-gamma gene on metabolic syndrome. Arch. Med. Res. 37, 86–94. Rhee, E.J., et al., 2007. No association of Pro12Ala polymorphism of PPAR-gamma gene with coronary artery disease in Korean subjects. Circ. J. 71, 338–342. Ricote, M., et al., 1998. Expression of the peroxisome proliferator-activated receptor gamma (PPARgamma) in human atherosclerosis and regulation in macrophages by colony stimulating factors and oxidized low density lipoprotein. Proc. Natl. Acad. Sci. U. S. A. 95, 7614–7619. Robitaille, J., Despres, J.P., Perusse, L., Vohl, M.C., 2003. 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. 63, 109–116. Ruchat, S.M., et al., 2010. Improvements in glucose homeostasis in response to regular exercise are influenced by the PPARG Pro12Ala variant: results from the HERITAGE Family Study. Diabetologia 53, 679–689.

R. Zhang, et al. / Gene 535 (2014) 79–87 Shen, J., Goyal, A., Sperling, L., 2012. The emerging epidemic of obesity, diabetes, and the metabolic syndrome in China. Cardiol. Res. Pract. 2012, 178675. Shi, H., et al., 2012. Association between PPAR-gamma and RXR-alpha gene polymorphism and metabolic syndrome risk: a case–control study of a Chinese Han population. Arch. Med. Res. 43, 233–242. Sirdah, M.M., Abu Ghali, A.S., Al Laham, N.A., 2012. The reliability of the National Cholesterol Education Program's Adult Treatment Panel III (NCEP/ATP III) and the International Diabetes Federation (IDF) definitions in diagnosing metabolic syndrome (MetS) among Gaza Strip Palestinians. Diabetes Metab. Syndr. 6, 4–8. Song, K.H., Cha, S., Yu, S.G., Yu, H., Oh, S.A., Kang, N.S., 2013. Association of apolipoprotein A5 Gene-1131TNC polymorphism with the risk of metabolic syndrome in Korean subjects. Biomed. Res. Int. 2013, 585134. Stroup, D.F., et al., 2000. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA 283, 2008–2012. Stumvoll, M., Haring, H., 2002. The peroxisome proliferator-activated receptor-gamma2 Pro12Ala polymorphism. Diabetes 51, 2341–2347. Sun, K., Liu, J., Ning, G., 2012. Active smoking and risk of metabolic syndrome: a metaanalysis of prospective studies. PLoS One 7, e47791. Sundvold, H., Lien, S., 2001. Identification of a novel peroxisome proliferator-activated receptor (PPAR) gamma promoter in man and transactivation by the nuclear receptor RORalpha1. Biochem. Biophys. Res. Commun. 287, 383–390.

87

Tellechea, M.L., Aranguren, F., Perez, M.S., Cerrone, G.E., Frechtel, G.D., Taverna, M.J., 2009. Pro12Ala polymorphism of the peroxisome proliferatoractivated receptor-gamma gene is associated with metabolic syndrome and surrogate measures of insulin resistance in healthy men: interaction with smoking status. Circ. J. 73, 2118–2124. Tontonoz, P., Hu, E., Graves, R.A., Budavari, A.I., Spiegelman, B.M., 1994. mPPAR gamma 2: tissue-specific regulator of an adipocyte enhancer. Genes Dev. 8, 1224–1234. Vimaleswaran, K.S., Radha, V., Deepa, R., Mohan, V., 2007. Absence of association of metabolic syndrome with PPARGC1A, PPARG and UCP1 gene polymorphisms in Asian Indians. Metab. Syndr. Relat. Disord. 5, 153–162. Wang, Y., Liu, C., 2012. Quantitative evaluation of common polymorphism (rs1801282) in the PPARgamma2 gene and hypertension susceptibility. Gene 502, 159–162. Wang, C., et al., 2012. Association between insulin receptor gene polymorphism and the metabolic syndrome in Han and Yi Chinese. Asia Pac. J. Clin. Nutr. 21, 457–463. Wells, G.A., Shea, B., O'Connell, D., Peterson, J., Welch, V., Losos, M., Tugwell, P., 2011. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. Health Research Institute, Ottawa. Yang, L.L., Hua, Q., Liu, R.K., Yang, Z., 2009. Association between two common polymorphisms of PPARgamma gene and metabolic syndrome families in a Chinese population. Arch. Med. Res. 40, 89–96. Yen, C.J., et al., 1997. 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. 241, 270–274.