Skip to main content
NCBI home page
Medicine logoLink to Medicine
. 2021 Jul 2;100(26):e26104. doi: 10.1097/MD.0000000000026104

Individual and combined effects of GSTM1, GSTT1, and GSTP1 polymorphisms on lung cancer risk

A meta-analysis and re-analysis of systematic meta-analyses

Wen-Ping Zhang a, Chen Yang b, Ling-Jun Xu c, Wei Wang d, Liang Song e, Xiao-Feng He f,∗
Editor: Meixia Lu
PMCID: PMC8257913  PMID: 34190143

Abstract

Thirty-five previous meta-analyses have been reported on the individual glutathione S-transferase M1 (GSTM1) present/null, glutathione S-transferase T1 (GSTT1) present/null, and glutathione S-transferase P1 (GSTP1) IIe105Val polymorphisms with lung cancer (LC) risk. However, they did not appraise the credibility and explore the combined effects between the 3 genes and LC risk.

We performed a meta-analysis and re-analysis of systematic previous meta-analyses to solve the above problems.

Meta-analyses of Observational Studies in Epidemiology guidelines were used. Moreover, we employed false-positive report probability (FPRP), Bayesian false discovery probability (BFDP), and the Venice criteria to verify the credibility of current and previous meta-analyses.

Significantly increased LC risk was considered as “highly credible” or “positive” for GSTM1 null genotype in Japanese (odds ratio (OR) = 1.30, 95% confidence interval (CI) = 1.17–1.44, I2 = 0.0%, statistical power = 0.997, FPRP = 0.008, BFDP = 0.037, and Venice criteria: AAB), for GSTT1 null genotype in Asians (OR = 1.23, 95% CI = 1.12–1.36, I2 = 49.1%, statistical power = 1.000, FPRP = 0.051, BFDP = 0.771, and Venice criteria: ABB), especially Chinese populations (OR = 1.31, 95% CI = 1.16–1.49, I2 = 48.9%, Statistical power = 0.980, FPRP = 0.039, BFDP = 0.673, and Venice criteria: ABB), and for GSTP1 IIe105Val polymorphism in Asians (Val vs IIe: OR = 1.28, 95% CI = 1.17–1.42, I2 = 30.3%, statistical power = 0.999, FPRP = 0.003, BFDP = 0.183, and Venice criteria: ABB). Significantly increased lung adenocarcinoma (AC) risk was also considered as “highly credible” or “positive” in Asians for the GSTM1 (OR = 1.35, 95% CI = 1.22–1.48, I2 = 25.5%, statistical power = 0.988, FPRP < 0.001, BFDP < 0.001, and Venice criteria: ABB) and GSTT1 (OR = 1.36, 95% CI = 1.17–1.58, I2 = 30.2%, statistical power = 0.900, FPRP = 0.061, BFDP = 0.727, and Venice criteria: ABB) null genotype.

This study indicates that GSTM1 null genotype is associated with increased LC risk in Japanese and lung AC risk in Asians; GSTT1 null genotype is associated with increased LC risk in Chinese, and GSTP1 IIe105Val polymorphism is associated with increased LC risk in Asians.

Keywords: BFDP, FPRP, GSTM1, GSTP1, GSTT1, lung cancer

1. Introduction

Lung cancer (LC) is the most common malignancy worldwide, accounting for more deaths than any other cancer in India.[1,2] There were about 228,190 new LC cases and 159,480 deaths in America in 2013.[3] It is calculated that over one million Chinese may be diagnosed with LC by 2025 in China.[4] Up to now, it is still not clear on the mechanism of LC. Studies have indicated that smoking was one of the most important risk factors,[5,6] however, only a small fraction of people, who are exposed to such risk factors, will develop LC. This indicates that host factors including genetic polymorphism may be an important role in LC development.

The glutathione S-transferases (GSTs) are a supergene family of phase II detoxifying enzymes, which play important role in the detoxification of toxic, potentially carcinogenic compounds, and a series of basic physiological processes of the human body.[7–9] In human, GSTs enzymes have been observed to be five classes (α, μ, π, σ, and θ)[10] and each class is encoded as an independent gene or family gene (such as GSTA, GSTM, GSTP, GSTO, and GSTT genes). Glutathione S-transferase M1 (GSTM1), glutathione S-transferase P1 (GSTP1), and glutathione S-transferase T1 (GSTT1) polymorphisms have been identified resulting in possible impaired activity for the elimination of carcinogenic compounds and raised risk of cancer.[11] The GSTM1 and GSTT1 show deletion (null genotype), which causes enzyme activity loss.[11] They are located on chromosome 1 (1p13.3) and chromosome 22 (22q11.2), respectively.[12] A codon 105 A to G mutation at exon 5 in GSTP1 polymorphism leads to change in isoleucine (IIe) to valine (Val), which also brings about decreased enzymatic activity.[13–14]

To date, 291 publications[supplementalreference1–291] have been reported on the individual and combined effects of GSTM1 present/null, GSTT1 present/null, and GSTP1 IIe105Val polymorphisms with LC risk. However, these results were still contradictory. In addition, 35 meta-analyses[15–29,31–50] have been reported on the individual and the combined effects of GSTM1 present/null, GSTT1 present/null, and GSTP1 IIe105Val polymorphisms with LC risk. However, a lot of studies have been published to investigate these associations recently. Hence, an updated meta-analysis should be performed to explore these problems. For all we know, this is the first meta-analysis to investigate the combined effects of GSTM1 and GSTP1, GSTT1 and GSTP1, and GSTM1, GSTT1, and GSTP1 with LC risk in the overall population. Moreover, there has been no study investigating whether the previous meta-analyses are “credible” on these associations. Therefore, 2 Bayesian methods (false-positive report probability (FPRP) and Bayesian false discovery probability (BFDP)) and the Venice criteria were applied to evaluate the credibility of these findings. We aimed to provide true associations on these problems and discuss the identified positive findings in terms of biological mechanisms involved in LC.

2. Materials and methods

2.1. Search strategy

Meta-analyses of Observational Studies in Epidemiology guidelines were used.[51] PubMed and China National Knowledge Infrastructure (CNKI) databases were applied to search literature in this meta-analysis (update to April 22, 2019). The following search strategy (it was designed to be sensitive and broad) was applied: (glutathione S-transferase T1 OR GSTT1 OR glutathione S-transferase P1 OR GSTP1 OR glutathione S-transferase M1 OR GSTM1) AND lung AND (polymorphism OR genotype OR allele OR variant OR mutation). In addition, the reference lists of identified articles and reviews (including published meta-analyses) were examined as appropriate. Moreover, Finally, the corresponding authors were contacted via e-mail if necessary. There was no limit or restriction on language in this study.

2.2. Inclusion and exclusion criteria

Inclusion criteria were as listed below: (1) case–control or cohort studies; (2) publications on GSTM1 present/null, GSTT1 present/null, GSTP1 IIe105Val, and their combined effects with LC risk; and (3) complete genotype data between LC cases and controls. Exclusion criteria were as listed below: (1) duplicate genotype data; (2) no case–control studies; (3) meta-analyses, reviews, or letters; and (4) other SNP.

2.3. Data extraction and quality score assessment

Two authors independently collected data of all eligible studies applying Excel. If necessary, any disagreement was resolved by discussion. The following data were extracted: (1) first author's surname, (2) year of publication, (3) country, (4) ethnicity, (5) sample size, (6) cases source, (7) controls source, (8) type of controls, (9) matching, (10) material used for assessment of genotype, and (11) genotype distribution of GSTM1 present/null, GSTT1 present/null, GSTP1 IIe105Val, and their combined effects in cases and controls. Races were considered as “Caucasians,” “Asians,” “Indians,” and “Africans.” “Mixed populations” was defined if race was not stated or the sample size cannot be separated. The scale of quality assessment criteria are listed by 2 previous meta-analyses[52,53] in Table 12, Supplemental Digital Content. Tables 2 and 3, Supplemental Digital Content list the quality assessment by included studies. Studies scoring >12 will be considered as high quality.

2.4. Statistical analysis

Crude odds ratios (ORs) and their 95% confidence intervals (CIs) were used to assess the associations between the individual and combined effects of GSTM1, GSTT1, and GSTP1 IIe105Val polymorphisms with LC risk. Either a fixed-effect model (Mantel–Haenszel method)[54] or a random-effect model (DerSimonian–Laird model)[55] was applied in this meta-analysis. Between-study heterogeneity was evaluated by calculating the Q statistic and I2 value (a random-effect model was applied if P < .10 and/or I2 > 50%). Subgroups were also calculated if heterogeneity was significant. In addition, we applied a meta-regression analysis to assess the source of heterogeneity. Sensitivity analysis was performed by removing a single study each time. Begg funnel plot[56] and Egger regression asymmetry test[57] were used to identify publication bias. A nonparametric “trim and fill” method[58] was considered to add missing studies if publication bias was observed in this meta-analysis. Moreover, Chi-square goodness-of-fit test was applied to check Hardy–Weinberg equilibrium (HWE), and significant deviation was considered in control groups if P < .05. All statistical analyses were calculated using STATA version 12.0 (STATA Corporation, College Station, TX).

2.5. Credibility of genetic association

We employed FPRP,[59] BFDP,[60] and the Venice criteria[61] to verify the credibility of current and previous meta-analyses. FPRP and BFDP can clarify the probability of no true association between genetic association and disease. The FPRP and BFDP were calculated by applying the Excel spreadsheet. A cutoff value of FPRP and BRDP was set up to be a level of 0.2 and 0.8 to assess whether the significant associations were noteworthy or not, respectively. Concerning the Venice criteria, we evaluated the criteria of the amount of evidence by statistical power[62]: A: 80% or more; B: 50% to 79%; and C: <50%. For replication, we applied the I2 recommended by Ioannidis et al[61]: A: less than 25%, B: 25% to 50%, and C: more than 50%. For protection from bias, we considered using the criteria proposed by Ioannidis et al[61] The following criteria were applied to assess the credibility of genetic association by FPRP, BFDP, and the Venice criteria. Firstly, associations were considered as positive results if they met the following criteria[62]: (1) statistically significant associations were observed in at least 2 of the genetic model (individual GSTM1 and GSTT1 polymorphisms with LC risk did not need to meet the criteria); (2) FPRP < 0.2 and BFDP < 0.8; (3) I2 < 50%; and (4) statistical power >80%. All other significant results were considered as less-credible positives. Previous meta-analyses were selected to assess the credibility by the following criteria: (1) more recent meta-analysis with the larger number of participants was selected and (2) studies supplying complete information involving OR and 95% CI.

3. Results

3.1. Study characteristics

A flowchart of study selection is listed in Figure 1. Overall, 756 publications were identified by PubMed and CNKI databases. Among these publications, 291 were selected by carefully screening titles, abstracts, and full text. In addition, 66 studies[supplemental references 5, 11, 12, 16, 18, 23, 24, 33, 38, 48, 54, 62, 73, 86, 89, 96, 101, 103, 112, 116, 166, 173, 177, 180, 183, 187, 189, 190, 193, 195, 196, 198, 199, 202, 206, 208, 212, 217, 218, 221, 222, 223, 224, 225, 228, 233, 235, 237, 238, 239, 241, 243, 244, 245, 251, 253, 254, 264, 267, 269, 270, 272, 280, 283, 284, 285] were excluded because another 47 publications included their cases and controls. Finally, 225 publications met the inclusion criteria. The general characteristics of publications are listed in Table 1, Supplemental Digital Content. There were 205 case–control studies from 197 articles (involving 45,726 cases and 58,788 controls, as shown in Table 4, Supplemental Digital Content) on GSTM1 present/null polymorphism, 103 case–control studies from 98 articles (involving 29,476 cases and 35,305 controls, as shown in Table 4, Supplemental Digital Content) on GSTT1 present/null polymorphism, 69 case–control studies from 66 publications regarding GSTP1 IIe105Val polymorphism (including 18,852 cases and 21,941 controls, as shown in Table 4, Supplemental Digital Content), 45 case–control studies from 42 publications on the combined effects of GSTM1 and GSTT1 present/null polymorphisms (involving 15,560 cases and 15,914 controls, as shown in Table 8, Supplemental Digital Content), 21 case–control studies from 19 publications on the combined effects of GSTM1 present/null and GSTP1 IIe105Val polymorphisms (involving 4538 cases and 5604 controls, as shown in Table 9, Supplemental Digital Content), 17 case–control studies from 15 publications regarding the combined effects of GSTT1 present/null and GSTP1 IIe105Val polymorphisms (involving 3507 cases and 4151 controls, as shown in Table 10, Supplemental Digital Content), and 7 case–control studies from 6 publications concerning the combined effects of the GSTM1 present/null, GSTT1 present/null, and GSTP1 IIe105Val polymorphisms (including 436 cases and 672 controls, as shown in Table 11, Supplemental Digital Content) with LC risk. In addition, we also collected the genotype frequencies of the GSTM1, GSTT1, and GSTP1 polymorphisms by histological type, smoking status, and gender, as shown in Tables 5 to 7, Supplemental Digital Content, respectively. In the end, Tables 2 and 3, Supplemental Digital Content show the quality assessment in this meta-analysis by scale for quality assessment of molecular association studies of lung cancer (as shown in Table 12, Supplemental Digital Content).

Figure 1.

Figure 1

Open in a new tab

Flow diagram for identifying and including studies in the current meta-analysis.

3.2. Quantitative synthesis

The GSTM1 null genotype was associated with an increased LC risk (OR = 1.24, 95% CI: 1.19–1.30) in the overall analysis and some subgroups, such as Asians, Caucasians, Chinese populations, Japanese populations, and so on, as shown in Table 13, Supplemental Digital Content.

The GSTT1 null genotype was also associated with an increased LC risk (OR = 1.16, 95% CI: 1.08–1.24) in the overall analysis and several subgroups, such as Indians, Asians, Chinese populations, Japanese populations, high-quality studies, large-sized studies, smokers, and so on, as shown in Table 14, Supplemental Digital Content.

The pooled data from all eligible studies yielded a significant association between the GSTP1 IIe105Val polymorphism and LC risk (Val/Val + IIe/Val vs IIe/IIe: OR = 1.06, 95% CI = 1.00–2.11 and Val vs IIe: OR = 1.40, 95% CI = 1.34–1.46, Table 15, Supplemental Digital Content). In addition, a significantly increased LC risk was also found in several subgroups, such as Africans, Asians, Caucasians, and so on, as shown in Table 15, Supplemental Digital Content.

A significant association was observed (model 1: OR = 1.34, 95% CI = 1.11–1.61; model 2: OR = 1.27, 95% CI = 1.11–1.46; model 3: OR = 1.53, 95% CI = 1.30–1.80; model 4: OR = 1.20, 95% CI = 1.08–1.33; model 5: OR = 1.28, 95% CI = 1.15–1.42; and model 6: OR = 1.30, 95% CI = 1.17–1.45, Table 16, Supplemental Digital Content) between the combined effects of GSTM1 and GSTT1 null genotypes in the overall analysis and several subgroups, such as Caucasians, Asians, Indians, population-based studies, high-quality studies, and so on, as shown in Table 16, Supplemental Digital Content.

A significantly increased LC risk was found (model 1: OR = 1.15, 95% CI = 1.01–1.31; model 4: OR = 1.31, 95% CI = 1.09–1.56; model 5: OR = 1.18, 95% CI = 1.03–1.36; and model 6: OR = 1.20, 95% CI = 1.06–1.35, Table 17, Supplemental Digital Content) between the combined effects of GSTM1 present/null and GSTP1 IIe105Val polymorphisms in the overall analysis and several subgroups, such as Caucasians, Asians, Indians, Africans, and so on (Table 17, Supplemental Digital Content).

A significantly increased LC risk was observed (model 1: OR = 1.32, 95% CI = 1.10–1.58; model 4: OR = 1.55, 95% CI = 1.18–2.02; and model 6: OR = 1.47, 95% CI = 1.15–1.88) between the combined effects of GSTT1 present/null and GSTP1 IIe105Val polymorphism in the overall analysis and several subgroups, such as Caucasians, Asians, Indians, and so on, as shown in Table 18, Supplemental Digital Content.

Last, a significant association was observed (model 7: OR = 2.81, 95% CI = 1.02–7.79; model 8: OR = 1.44, 95% CI = 1.13–1.85; model 9: OR = 2.09, 95% CI = 1.42–3.08; and model 10: OR = 1.73, 95% CI = 1.17–2.56) between the combined effects of GSTM1 present/null, GSTT1 present/null and GSTP1 IIe105Val polymorphisms when all eligible studies were pooled, as shown in Table 19, Supplemental Digital Content.

3.3. Heterogeneity and sensitivity analyses

Between-studies heterogeneity was observed, as shown in Tables 13 to 19, Supplemental Digital Content. A meta-regression analysis indicates that ethnicity (P = .006) and type of controls (P = .019) are sources of heterogeneity between the GSTM1 null genotype and LC risk. For the GSTT1 null genotype, a meta-regression analysis suggests that ethnicity (P = .017), source of controls (P < .001), and type of controls (P < .001) are sources of heterogeneity. We found that HWE (model 1: P = .046) and quality score (model 6: P = .043) were the sources of heterogeneity by meta-regression analysis for the combined effects of GSTM1 present/null and GSTP1 IIe105Val polymorphisms. Moreover, we have not observed any change when 1 single study was excluded each time in the overall analysis.

3.4. Evaluation of publication bias

There was obvious evidence of publication bias for GSTM1 null genotype (P < .001), GSTT1 null genotype (P = .044), GSTP1 IIe105Val (Val/Val vs IIe/IIe: P = .010; IIe/Val vs IIe/IIe: P < .001; Val/Val + IIe/Val vs IIe/IIe: P < .001), the combined effects of GSTM1 and GSTT1 (model 1: P = .022; model 2: P = .013; model 3: P = .032; model 5: P = .037; and model 6: P = .004), the combined effects of GSTM1 present/null and GSTP1 IIe105Val (model 4: P = .001 and model 6: P = .002) by the Begg funnel plot shape and Egger test in the current meta-analysis. Figures 1 to 12, Supplemental Digital Content lists the funnel plots of the nonparametric “trim and fill” method. No significant association was observed (Val/Val + IIe/Val vs IIe/IIe: OR = 0.97, 95% CI = 0.92–1.03) for the GSTP1 IIe105Val when we applied the nonparametric “trim and fill” method in the overall analysis. The results of a pooled analysis from all studies changed in the following genetic models (model 1: OR = 1.08, 95% CI = 0.89–1.31; model 2: OR = 1.03, 95% CI = 0.89–1.19; model 5: OR = 1.10, 95% CI = 0.98–1.24; and model 6: OR = 1.10, 95% CI = 0.98–1.24) for the combined effects of GSTM1 and GSTT1 null genotypes when we applied the nonparametric “trim and fill” method. The results of a pooled analysis from all studies changed in model 4 (OR = 1.01, 95% CI = 0.83–1.24) and model 6 (OR = 1.00, 95% CI = 0.87–1.14) when we applied the nonparametric “trim and fill” method.

3.5. Credibility of the previous meta-analyses

To evaluate the credibility of the previously published meta-analyses with the largest number of cases and controls on the associations between the GSTM1 present/null, GSTT1 present/null, and/or GSTP1 IIe105Val polymorphisms with LC risk, we applied the FPRP, BFDP, and the Venice criteria. Table 1 shows the results of the credibility on these issues. Gao et al[18] on the combined effects of GSTM1 present/null and GSTT1 present/null polymorphisms with LC risk will be considered as “positive” result in the overall population, Ye et al[15] on the GSTM1 null genotype with LC risk in all races, Liu et al[41] on the GSTM1 null genotype with LC risk in Chinese populations, and Xu et al[33] on the GSTP1 IIe105Val polymorphism with LC risk will be considered as “positive” results because their studies represent the most credible findings. Li et al,[28] Sengupta et al,[50] Yang et al,[19] Yang et al,[34] Wang et al,[40] and Feng et al[21] will be classified as less-credible results (higher heterogeneity, lower statistical power, FPRP > 0.2 and BFDP > 0.8).

Table 1.

Credibility of previously published meta-analysis with the largest number of participants.

Credibility
Prior probability of 0.001
Author Gene Model n Case/control Variable OR (95% CI) I2 (%) Statistical power FPRP BFDP Venice criteria
Gao et al[18] 2017 GSTM1-GSTT1 Model 1 34 5886/5224 Overall 1.58 (1.34–1.87) 57.8 0.273 <0.001 0.006 CCB
Gao et al[18] 2017 GSTM1-GSTT1 Model 2 23 3309/2063 Overall 1.26 (1.13–1.42) 4.7 0.998 0.131 0.885 AAB
Gao et al[18] 2017 GSTM1-GSTT1 Model 3 23 4447/3198 Overall 1.26 (1.08–1.48) 31.5 0.983 0.832 0.993 ABB
Gao et al[18] 2017 GSTM1-GSTT1 Model 4 34 8177/6586 Overall 1.27 (1.13–1.42) 28.2 0.998 0.026 0.619 ABB
Gao et al[18] 2017 GSTM1-GSTT1 Model 5 44 13,706/13,093 Overall 1.33 (1.19–1.48) 45.9 0.986 <0.001 0.013 ABB
Gao et al[18] 2017 GSTM1-GSTT1 Model 1 16 2608/2893 Caucasian 1.23 (1.07–1.41) 12 0.998 0.748 0.991 AAB
Gao et al[18] 2017 GSTM1-GSTT1 Model 1 3 348/391 Indian 2.53 (1.61–3.98) 0.0 0.012 0.833 0.727 CAB
Gao et al[18] 2017 GSTM1-GSTT1 Model 4 3 348/273 Indian 1.69 (1.07–2.67) 2.0 0.305 0.988 0.997 CAB
Gao et al[18] 2017 GSTM1-GSTT1 Model 5 3 632/632 Indian 2.11 (1.36–3.28) 1.2 0.065 0.933 0.959 CAB
Gao et al[18] 2017 GSTM1-GSTT1 Model 3 10 2948/1592 Asian 1.24 (1.10–1.41) 33.2 0.998 0.508 0.977 ABB
Gao et al[18] 2017 GSTM1-GSTT1 Model 4 11 4159/2403 Asian 1.45 (1.19–1.77) 39.8 0.631 0.292 0.898 BBB
Gao et al[18] 2017 GSTM1-GSTT1 Model 5 14 5766/4337 Asian 1.53 (1.24–1.90) 68.1 0.429 0.217 0.806 CCB
Li et al[28] 2015 GSTM1-GSTP1 Model a 2 209/316 Chinese 1.68 (1.08–2.60) NA 0.306 0.985 0.996 C-B
Li et al[28] 2015 GSTM1-GSTP1 Model b 2 209/316 Chinese 2.13 (1.27–3.56) NA 0.090 0.977 0.987 C-B
Li et al[28] 2015 GSTM1-GSTP1 Model c 2 209/316 Chinese 2.29 (1.33–3.93) NA 0.062 0.977 0.983 C-B
Ye et al[15] 2006 GSTM1 null vs present 119 19,729/25,931 Overall 1.22 (1.14–1.23) 44 1.000 <0.001 <0.001 ABB
Liu et al[41] 2014 GSTM1 null vs present 68 8649/10,380 Chinese 1.20 (1.16–1.25) 45.1 1.000 <0.001 <0.001 ABB
Sengupta et al[50] 2017 GSTM1 null vs present 13 NA Indian 1.30 (1.01–1.68) 11.9 0.863 0.981 0.998 AAB
Yang et al[19] 2014 GSTT1 null vs present 55 15,140/16,662 Overall 1.14 (1.03–1.25) 62.8 1.000 0.841 0.996 ACB
Yang et al[34] 2013 GSTT1 null vs present 23 4065/5390 Asian 1.28 (1.10–1.49) 62.0 0.980 0.596 0.981 ACB
Wang et al[40] 2015 GSTT1 null vs present 20 3351/4683 Chinese 1.31 (1.12–1.52) 59 0.963 0.278 0.937 ACB
Sengupta et al[50] 2017 GSTT1 null vs present 12 NA Indian 1.39 (1.04–1.87) 34.7 0.693 0.977 0.998 BBB
Li et al[28] 2015 GSTP1 Val/Val vs IIe/IIe 13 2026/2451 Chinese 1.36 (1.01–1.84) 31.7 0.737 0.984 0.998 BBB
Xu et al[33] 2014 GSTP1 Val/Val vs IIe/IIe 18 3175/5516 Asian 1.22 (1.16–1.59) NA 0.937 0.993 0.999 A-B
Xu et al[33] 2014 GSTP1 Val/Val + IIe/Val vs IIe/IIe 18 3175/5516 Asian 1.24 (1.12–1.37) 18.4 1.000 0.023 0.609 AAB
Feng et al[21] 2012 GSTP1 Val vs IIe 44 12,363/13,948 Overall 1.08 (1.02–1.15) 44 1.000 0.942 0.999 ABB
Open in a new tab

CI = confidence interval, OR = odds ratio, SC = squamous carcinoma, Model a = M1 null/P1 IIe/IIe vs M1 present/P1 IIe/IIe, Model b = M1 null/P1 Val∗ vs M1 present/P1 IIe/IIe, Model c = T1 null/P1 Val∗ vs T1 present/P1 IIe/IIe, Model 1 = M1 null/T1 null vs M1 present/T1 present, Model 2 = M1 null/T1 null vs M1 present/T1 null, Model 3 = M1 null/T1 null vs M1 null/T1 present, Model 4 = M1 null/T1 null vs All 1 risk genotypes, Model 5 = M1 null/T1 null vs (M1 present/T1 present + M1 present/T1 null + M1 null/T1 present).

The significance of bold values indicated that these positive results were credible.

3.6. Credibility of the current meta-analysis

To evaluate the credibility of the present meta-analysis, we also applied the FPRP, BFDP, and the Venice criteria. Table 2 lists the credibility of the current meta-analysis on the individual and combined effects of GSTM1 present/null, GSTT1 present/null, and GSTP1 IIe105Val polymorphisms with LC risk. They will be considered as “positive” results on the GSTM1 null genotype with LC risk in Japanese population (OR = 1.30, 95% CI = 1.17–1.44, I2 = 0.0%, statistical power = 0.997, FPRP = 0.008, BFDP = 0.037, and Venice criteria: AAB), GSTM1 null genotype with lung AC risk in Asians (OR = 1.35, 95% CI = 1.22–1.48, I2 = 25.5%, statistical power = 0.988, FPRP < 0.001, BFDP < 0.001, and Venice criteria: ABB), GSTT1 null genotype with LC risk in Asians (OR = 1.23, 95% CI = 1.12–1.36, I2 = 49.1%, statistical power = 1.000, FPRP = 0.051, BFDP = 0.771, and Venice criteria: ABB), especially in Chinese population (OR = 1.31, 95% CI = 1.16–1.49, I2 = 48.9%, statistical power = 0.980, FPRP = 0.039, BFDP = 0.673, and Venice criteria: ABB), GSTT1 null genotype with lung AC risk in Asians (OR = 1.36, 95% CI = 1.17–1.58, I2 = 30.2%, statistical power = 0.900, FPRP = 0.061, BFDP = 0.727, and Venice criteria: ABB), and GSTP1 IIe105Val polymorphism with LC risk in overall population, especially in Asians (Val vs IIe: OR = 1.28, 95% CI = 1.17–1.42, I2 = 30.3%, statistical power = 0.999, FPRP = 0.003, BFDP = 0.183, and Venice criteria: ABB). All other significant associations will be considered as less-credible results, as also shown in Table 2.

Table 2.

Credibility of the current meta-analysis.

Credibility
Prior probability of 0.001
Variables Model OR (95% CI) I2 (%) Statistical power FPRP BFDP Venice criteria
GSTM1
Overall Null vs present 1.24 (1.19–1.30) 58.5 1.000 <0.001 <0.001 ACB
Asian Null vs present 1.43 (1.33–1.53) 54.8 0.988 <0.001 <0.001 ACB
Caucasian Null vs present 1.07 (1.01–1.13) 39.4 1.000 0.938 0.999 ABB
China Null vs present 1.52 (1.40–1.65) 53.3 1.000 <0.001 <0.001 ACB
Japan Null vs present 1.30 (1.17–1.44) 0.0 0.997 0.008 0.037 AAB
HB Null vs present 1.30 (1.21–1.39) 64.0 1.000 <0.001 <0.001 ACB
PB Null vs present 1.14 (1.05–1.24) 55.6 1.000 0.718 0.992 ACB
Matching Null vs present 1.18 (1.10–1.25) 55.1 1.000 <0.001 0.003 ACB
Non-matching Null vs present 1.30 (1.23–1.39) 60.8 1.000 <0.001 <0.001 ACB
Quality score >12 Null vs present 1.14 (1.07–1.21) 57.8 1.000 0.017 0.637 ACB
Quality score ≤12 Null vs present 1.31 (1.24–1.39) 56.9 1.000 <0.001 <0.001 ACB
Sample size >200 Null vs present 1.21 (1.16–1.27) 63.2 1.000 <0.001 <0.001 ACB
Sample size ≤200 Null vs present 1.42 (1.29–1.57) 18.8 0.858 <0.001 <0.001 AAB
SCLC Null vs present 1.38 (1.16–1.63) 50.2 0.837 0.152 0.855 ACB
SCLC/Asian Null vs present 1.43 (1.04–1.97) 43.3 0.615 0.979 0.997 BBB
SCLC/Caucasian Null vs present 1.33 (1.01–1.76) 65.7 0.800 0.983 0.998 ACB
SCLC/Indian Null vs present 1.66 (1.21–2.28) 0.0 0.266 0.868 0.975 CAB
SC Null vs present 1.33 (1.22–1.45) 55.2 0.997 <0.001 <0.001 ACB
SC/Asian Null vs present 1.52 (1.38–1.66) 10.9 0.403 <0.001 <0.001 CAB
SC/Indian Null vs present 1.37 (1.13–1.67) 0.0 0.815 0.692 0.981 AAB
AC Null vs present 1.24 (1.13–1.36) 52.0 1.000 0.005 0.277 ACB
AC/Asian Null vs present 1.35 (1.22–1.48) 25.5 0.988 <0.001 <0.001 ABB
AC/Indian Null vs present 1.49 (1.17–1.90) 19.2 0.522 0.714 0.971 BAB
Smoking Null vs present 1.27 (1.17–1.39) 61.7 1.000 <0.001 0.019 ACB
Non-smoking Null vs present 1.36 (1.21–1.53) 50.4 0.948 <0.001 0.022 ACB
Male Null vs present 1.16 (1.06–1.26) 47.1 1.000 0.303 0.966 ABB
GSTT1
Overall Null vs present 1.16 (1.08–1.24) 59.2 1.000 0.013 0.558 ACB
Indian Null vs present 1.54 (1.13–2.11) 78.5 0.435 0.943 0.992 CCB
Asian Null vs present 1.23 (1.12–1.36) 49.1 1.000 0.051 0.771 ABB
China Null vs present 1.31 (1.16–1.49) 48.9 0.980 0.039 0.673 ABB
Japan Null vs present 1.22 (1.01–1.47) 8.2 0.985 0.974 0.999 AAB
North India Null vs present 2.99 (1.88–4.78) 51.8 0.002 0.706 0.267 CCB
HB Null vs present 1.17 (1.06–1.29) 63.1 1.000 0.619 0.988 ACB
Matching Null vs present 1.12 (1.02–1.24) 56.3 1.000 0.967 0.999 ACB
Non-matching Null vs present 1.19 (1.08–1.30) 61.9 1.000 0.103 0.885 ACB
Quality score >12 Null vs present 1.11 (1.02–1.21) 54.8 1.000 0.947 0.999 ACB
Quality score ≤12 Null vs present 1.20 (1.08–1.33) 62.0 1.000 0.339 0.964 ACB
Sample size >200 Null vs present 1.15 (1.08–1.23) 60.7 1.000 0.044 0.808 ACB
LCLC Null vs present 0.39 (0.17–0.94) 36.3 0.114 0.997 0.998 CBB
SC/Asian Null vs present 1.38 (1.02–1.87) 63.5 0.705 0.982 0.998 BCB
AC/Asian Null vs present 1.36 (1.17–1.58) 30.2 0.900 0.061 0.727 ABB
AC/Indian Null vs present 2.02 (1.51–2.70) 0.0 0.022 0.084 0.094 CAB
Smoking Null vs present 1.23 (1.08–1.40) 56.1 0.999 0.633 0.985 ACB
GSTP1
Overall Val/Val + IIe/Val vs IIe/IIe 1.06 (1.00–1.11) 29.0 1.000 0.930 0.999 ABB
Val vs IIe 1.40 (1.34–1.46) 23.3 1.000 <0.001 <0.001 AAB
African Val vs IIe 1.65 (1.27–2.15) 0.0 0.240 0.465 0.865 CAB
Asian Val/Val vs IIe/IIe 1.45 (1.16–1.80) 7.6 0.621 0.549 0.957 BAB
IIe/Val vs IIe/IIe 1.13 (1.02–1.24) 12.0 1.000 0.908 0.998 AAB
Val/Val vs IIe/IIe + IIe/Val 1.39 (1.12–1.72) 0.0 0.758 0.763 0.984 BAB
Val/Val + IIe/Val vs IIe/IIe 1.16 (1.06–1.26) 23.0 1.000 0.303 0.966 AAB
Val vs IIe 1.28 (1.17–1.42) 30.3 0.999 0.003 0.183 ABB
Caucasian Val vs IIe 1.44 (1.38–1.50) 0.0 1.000 <0.001 <0.001 AAB
HB Val/Val vs IIe/IIe 1.12 (1.01–1.25) 11.5 1.000 0.977 0.999 AAB
Val/Val vs IIe/IIe + IIe/Val 1.12 (1.01–1.24) 6.0 1.000 0.967 0.999 AAB
Val/Val + IIe/Val vs IIe/IIe 1.08 (1.01–1.16) 35.5 1.000 0.972 0.999 ABB
Val vs IIe 1.38 (1.31–1.47) 26.4 1.000 <0.001 <0.001 ABB
PB Val vs IIe 1.42 (1.34–1.51) 2.5 1.000 <0.001 <0.001 AAB
Matching Val vs IIe 1.38 (1.32–1.45) 18.4 1.000 <0.001 <0.001 AAB
Non-matching Val vs IIe 1.42 (1.33–1.51) 29.0 1.000 <0.001 <0.001 ABB
Quality score >12 Val vs IIe 1.40 (1.34–1.46) 0.0 1.000 <0.001 <0.001 AAB
Quality score ≤12 Val/Val vs IIe/IIe 1.23 (1.06–1.42) 9.8 0.997 0.826 0.993 AAB
IIe/Val vs IIe/IIe 1.13 (1.05–1.23) 24.6 1.000 0.825 0.996 AAB
Val/Val vs IIe/IIe + IIe/Val 1.16 (1.01–1.34) 1.8 1.000 0.978 0.999 AAB
Val/Val + IIe/Val vs IIe/IIe 1.16 (1.07–1.25) 26.5 1.000 0.090 0.886 ABB
Val vs IIe 1.39 (1.27–1.51) 44.1 0.964 <0.001 <0.001 ABB
Sample size >200 Val/Val + IIe/Val vs IIe/IIe 1.06 (1.00–1.12) 32.9 1.000 0.974 1.000 ABB
Val vs IIe 1.41 (1.35–1.47) 23.2 1.000 <0.001 <0.001 ABB
HWE (yes) Val/Val vs IIe/IIe 1.08 (1.00–1.17) 17.6 1.000 0.983 1.000 AAB
Val vs IIe 1.41 (1.36–1.46) 9.6 1.000 <0.001 <0.001 AAB
HWE (no) Val/Val vs IIe/IIe 0.73 (0.54–0.99) 0.0 0.709 0.984 0.998 BAB
Val/Val vs IIe/IIe + IIe/Val 0.71 (0.53–0.95) 0.0 0.652 0.970 0.997 BAB
SCLC Val/Val vs IIe/IIe 1.34 (1.01–1.77) 0.0 0.787 0.980 0.998 BAB
Val/Val vs IIe/IIe + IIe/Val 1.32 (1.01–1.72) 21.8 0.828 0.980 0.998 AAB
SCLC/Caucasian Val/Val vs IIe/IIe 1.42 (1.05–1.92) 0.0 0.639 0.973 0.997 BAB
Val/Val vs IIe/IIe + IIe/Val 1.41 (1.07–1.87) 20.6 0.666 0.962 0.996 BAB
Smoking Val/Val vs IIe/IIe 1.33 (1.08–1.64) 0.0 0.870 0.898 0.994 AAB
Val/Val vs IIe/IIe + IIe/Val 1.29 (1.01–1.57) 0.0 0.934 0.922 0.996 AAB
Val vs IIe 1.10 (1.01–1.21) 0.0 1.000 0.980 0.999 AAB
The combined effects of GSTM1 and GSTT1 polymorphisms
Overall Model 1 1.34 (1.11–1.61) 54.7 0.886 0.667 0.981 ACB
Model 2 1.27 (1.11–1.46) 57.0 0.990 0.440 0.968 ACB
Model 3 1.53 (1.30–1.80) 61.6 0.406 0.001 0.017 CCB
Model 4 1.20 (1.08–1.33) 51.5 1.000 0.339 0.964 ACB
Model 5 1.28 (1.15–1.42) 61.3 0.999 0.003 0.183 ACB
Model 6 1.30 (1.17–1.45) 50.6 0.995 0.002 0.147 ACB
Caucasian Model 3 1.14 (1.02–1.28) 20.5 1.000 0.964 0.999 AAB
Model 5 1.14 (1.02–1.27) 43.3 1.000 0.946 0.998 ABB
Asian Model 1 1.40 (1.06–1.84) 47.1 0.690 0.958 0.996 BBB
Model 2 1.52 (1.17–1.98) 48.2 0.461 0.805 0.978 CBB
Model 3 1.99 (1.40–2.85) 75.6 0.061 0.738 0.846 CCB
Model 4 1.40 (1.10–1.79) 56.8 0.709 0.911 0.993 BCB
Model 5 1.61 (1.22–2.11) 69.9 0.304 0.647 0.938 CCB
Model 6 1.51 (1.22–1.86) 68.1 0.475 0.183 0.793 CCB
Indian Model 2 1.53 (1.13–2.07) 0.0 0.449 0.928 0.991 CAB
Model 3 2.53 (1.61–3.98) 0.0 0.012 0.833 0.727 CAB
Model 4 1.49 (1.18–1.88) 0.0 0.522 0.597 0.956 BAB
Model 5 1.62 (1.29–2.02) 0.0 0.247 0.069 0.427 CAB
Model 6 2.11 (1.36–3.28) 1.2 0.065 0.933 0.959 CAB
HB Model 1 1.30 (1.01–1.68) 42.2 0.863 0.981 0.998 ABB
Model 2 1.36 (1.12–1.66) 48.2 0.832 0.750 0.985 ABB
Model 3 1.57 (1.27–1.94) 45.1 0.336 0.080 0.539 CBB
Model 4 1.24 (1.06–1.45) 50.3 0.991 0.876 0.995 ACB
Model 5 1.32 (1.13–1.55) 55.8 0.941 0.428 0.962 ACB
Model 6 1.36 (1.18–1.57) 39.2 0.909 0.029 0.572 ABB
PB Model 1 1.73 (1.13–2.65) 75.1 0.256 0.979 0.994 CCB
Model 3 1.54 (1.12–2.13) 76.1 0.437 0.954 0.994 CCB
Model 5 1.25 (1.02–1.53) 70.3 0.961 0.969 0.998 ACB
Model 6 1.27 (1.05–1.53) 64.1 0.960 0.925 0.996 ACB
Matching Model 3 1.43 (1.04–1.97) 51.7 0.615 0.979 0.997 BCB
Model 5 1.34 (1.01–1.78) 70.9 0.782 0.982 0.998 BCB
Non-matching Model 1 1.34 (1.10–1.64) 49.0 0.863 0.840 0.991 ABB
Model 2 1.21 (1.06–1.38) 40.1 0.999 0.818 0.994 ABB
Model 3 1.57 (1.29–1.91) 66.0 0.324 0.020 0.230 CCB
Model 4 1.16 (1.05–1.28) 38.9 1.000 0.757 0.993 ABB
Model 5 1.25 (1.11–1.40) 56.8 0.999 0.102 0.860 ACB
Model 6 1.39 (1.20–1.61) 60.1 0.845 0.013 0.366 ACB
Quality score >12 Model 2 1.24 (1.01–1.55) 78.4 0.953 0.984 0.999 ACB
Model 3 1.50 (1.17–1.92) 71.8 0.500 0.720 0.970 BCB
Model 4 1.19 (1.00–1.41) 69.0 0.996 0.978 0.999 ACB
Model 5 1.27 (1.06–1.52) 74.8 0.965 0.904 0.996 ACB
Model 6 1.25 (1.08–1.45) 55.4 0.992 0.764 0.991 ACB
Quality score ≤12 Model 1 1.39 (1.06–1.84) 37.4 0.703 0.968 0.997 BBB
Model 2 1.28 (1.10–1.49) 0.0 0.980 0.596 0.981 AAB
Model 3 1.56 (1.24–1.96) 47.5 0.368 0.267 0.819 CBB
Model 4 1.17 (1.07–1.28) 10.0 1.000 0.381 0.973 AAB
Model 5 1.28 (1.13–1.46) 32.5 0.991 0.192 0.915 ABB
Model 6 1.36 (1.15–1.61) 47.1 0.872 0.289 0.928 ABB
Sample size >200 Model 1 1.37 (1.11–1.69) 64.1 0.801 0.804 0.988 ACB
Model 2 1.25 (1.08–1.46) 65.7 0.989 0.831 0.993 ACB
Model 3 1.53 (1.29–1.83) 66.5 0.414 0.008 0.139 CCB
Model 4 1.19 (1.07–1.33) 58.6 1.000 0.685 0.990 ACB
Model 5 1.27 (1.13–1.42) 66.8 0.998 0.020 0.619 ACB
Model 6 1.27 (1.14–1.42) 53.0 0.998 0.026 0.619 ACB
Sample size ≤200 Model 2 1.49 (1.02–2.20) 0.0 0.513 0.989 0.998 BAB
Model 3 1.48 (1.01–2.17) 18.6 0.527 0.988 0.998 BAB
Model 5 1.39 (1.01–1.92) 0.0 0.678 0.985 0.998 BAB
Model 6 1.52 (1.16–1.99) 63.0 0.462 0.834 0.981 CCB
The combined effects of GSTM1 and GSTP1 polymorphisms
Overall Model a 1.15 (1.01–1.31) 24.9 1.000 0.973 0.999 AAB
Model d 1.31 (1.09–1.56) 46.8 0.936 0.723 0.986 ABB
Model e 1.18 (1.03–1.36) 51.7 1.000 0.957 1.000 ACB
Model f 1.20 (1.06–1.35) 30.0 1.000 0.707 0.990 ABB
Caucasian Model d 1.21 (1.00–1.47) 41.2 0.985 0.982 0.999 ABB
Model f 1.16 (1.01–1.35) 39.6 1.000 0.982 0.999 ABB
Asian Model a 1.68 (1.08–2.60) 0.0 0.306 0.985 0.996 CAB
Model c 1.56 (1.03–2.35) 0.0 0.426 0.987 0.997 CAB
Model d 2.54 (1.50–4.33) 0.0 0.026 0.959 0.952 CAB
Model e 1.76 (1.19–2.60) 0.0 0.211 0.955 0.988 CAB
Model f 1.90 (1.20–3.03) 0.0 0.160 0.978 0.992 CAB
Indian Model a 1.44 (1.09–1.90) 48.8 0.614 0.942 0.994 BBB
African Model c 1.99 (1.00–3.94) 46.8 0.209 0.996 0.998 CBB
Model e 1.98 (1.02–3.86) 43.4 0.207 0.995 0.998 CBB
PB Model c 1.43 (1.05–1.94) 12.5 0.621 0.972 0.997 BAB
Model d 1.46 (1.04–2.05) 0.0 0.562 0.981 0.997 BAB
Model e 1.44 (1.08–1.93) 18.0 0.608 0.960 0.996 BAB
Matching Model a 1.34 (1.12–1.61) 37.9 0.886 0.667 0.981 ABB
Model c 1.32 (1.09–1.61) 46.7 0.896 0.873 0.993 ABB
Model d 1.55 (1.17–2.06) 56.5 0.411 0.860 0.982 CCB
Model e 1.39 (1.14–1.71) 55.2 0.764 0.706 0.980 BCB
Model f 1.28 (1.05–1.57) 45.6 0.936 0.950 0.997 ABB
Quality score >12 Model a 1.32 (1.01–1.71) 53.6 0.833 0.977 0.998 ACB
Model c 1.26 (1.05–1.52) 47.6 0.966 0.942 0.997 ABB
Model d 1.31 (1.02–1.68) 52.5 0.857 0.975 0.998 ACB
Model e 1.29 (1.06–1.57) 56.9 0.934 0.922 0.996 ACB
Quality score ≤12 Model d 1.30 (1.07–1.58) 39.1 0.925 0.901 0.995 ABB
Model f 1.34 (1.14–1.57) 0.0 0.919 0.242 0.919 AAB
HWE (yes) Model d 1.34 (1.10–1.62) 51.0 0.878 0.740 0.986 ACB
Model e 1.17 (1.02–1.34) 42.5 1.000 0.959 0.998 ABB
Model f 1.22 (1.07–1.39) 36.7 0.999 0.738 0.990 ABB
The combined effects of GSTT1 and GSTP1 polymorphisms
Overall Model g 1.32 (1.10–1.58) 0.0 0.918 0.729 0.986 AAB
Model h 1.55 (1.18–2.02) 53.7 0.404 0.745 0.967 CCB
Model k 1.47 (1.15–1.88) 52.7 0.564 0.792 0.981 BCB
Caucasian Model h 1.42 (1.03–1.95) 50.2 0.633 0.979 0.997 BCB
Model k 1.41 (1.03–1.93) 55.8 0.650 0.980 0.998 BCB
Asian Model h 2.29 (1.33–3.93) 0.0 0.062 0.977 0.983 CAB
Model j 1.47 (1.01–2.14) 0.0 0.542 0.988 0.998 BAB
Indian Model g 1.75 (1.21–2.55) 20.5 0.211 0.944 0.986 CAB
HB Model g 1.32 (1.06–1.64) 14.5 0.876 0.933 0.996 AAB
Model h 1.54 (1.01–2.37) 67.9 0.452 0.991 0.998 CCB
Model k 1.50 (1.03–2.18) 62.8 0.500 0.985 0.998 BCB
PB Model h 1.70 (1.16–2.49) 26.7 0.260 0.961 0.991 CBB
Matching Model h 1.41 (1.02–1.95) 40.5 0.646 0.983 0.998 BBB
Non-matching Model g 1.50 (1.11–2.01) 0.0 0.500 0.930 0.992 BAB
Model h 1.71 (1.09–2.67) 63.4 0.282 0.985 0.996 CCB
Model k 1.76 (1.18–2.61) 57.4 0.213 0.958 0.989 CCB
Quality score >12 Model h 1.52 (1.09–2.12) 50.1 0.469 0.967 0.995 CCB
Model k 1.43 (1.04–1.99) 55.6 0.612 0.982 0.998 BCB
Quality score ≤12 Model g 1.47 (1.11–1.94) 0.5 0.557 0.921 0.992 BAB
Model k 1.53 (1.03–2.26) 50.4 0.460 0.986 0.997 CCB
HWE (yes) Model g 1.29 (1.06–1.58) 0.0 0.928 0.937 0.997 AAB
Model h 1.58 (1.18–2.10) 56.5 0.360 0.819 0.974 CCB
Model k 1.48 (1.13–1.93) 56.7 0.539 0.876 0.988 BCB
The combined effects of GSTT1 and GSTP1 polymorphisms
Overall Model 7 2.81 (1.02–7.79) – 0.114 0.998 0.998 C–B
Model 8 1.44 (1.13–1.85) 0.0 0.625 0.874 0.989 BAB
Model 9 2.09 (1.42–3.08) 0.0 0.047 0.806 0.862 CAB
Model 10 1.73 (1.17–2.56) 0.0 0.238 0.963 0.991 CAB
HWE (yes) Model 9 2.10 (1.41–3.14) 0.0 0.051 0.856 0.901 CAB
Open in a new tab

Model 1 = M1 present/T1 null vs M1 present/T1 present, Model 2 = M1 null/T1 present vs M1 present/T1 present, Model 3 = M1 null/T1 null vs M1 present/T1 present, Model 4 = all 1 risk genotypes vs M1 present/T1 present, Model 5 = all risk genotypes vs M1 present/T1 present, Model 6 = M1 null/T1 null vs M1 present/T1 present + M1 present/T1 null + M1 null/T1 present, Model a = M1 null/P1 IIe/IIe vs M1 present/P1 IIe/IIe, Model c = (M1 null/P1 IIe/IIe + M1 present/P1 Val∗) vs M1 present/P1 IIe/IIe, Model d = M1 null/P1 Val∗ vs M1 present/P1 IIe/IIe, Model e = all risk genotypes vs M1 present/P1 IIe/IIe, Model f = M1 null/P1 Val∗ vs (M1 present/P1 IIe/IIe + M1 null/P1 IIe/IIe + M1 present/P1 Val∗), Model g = T1 null/P1 IIe/IIe vs T1 present/P1 IIe/IIe, Model h = T1 null/P1 Val∗ vs T1 present/P1 IIe/IIe, Model j = all risk genotypes vs T1 present/P1 IIe/IIe, Model k = T1 null/P1 Val∗ vs (T1 present/P1 IIe/IIe + T1 null/P1 IIe/IIe + T1 present/P1 Val∗), Model 7 = M1 present/T1 null/P1 Val 1 vs M1 present/T1 present/P1 IIe/IIe, Model 8 = all 2 high-risk genotype vs M1 present/T1 present/P1 IIe/IIe, Model 9 = M1 null/T1 null/P1 Val1 vs M1 present/T1 present/P1 IIe/IIe, Model 10 = M1 null/T1 null/P1 Val1 vs (M1 present/T1 present/P1 IIe/IIe + all 1 high-risk genotype + all 2 high-risk genotypes).

CI = confidence interval, HB = hospital-based studies, HWE = Hardy–Weinberg equilibrium, LC = lung cancer, LCLC = large cell lung carcinoma, ORs = odds ratios, PB = population-based studies, SC = squamous carcinoma, SCLC = small-cell lung cancer.

The significance of bold values indicated that these positive results were credible.

4. Discussion

To the best of our knowledge, we reported the first meta-analysis to investigate the combined effects of GSTM1 and GSTP1, GSTT1 and GSTP1, and GSTM1, GSTT1, and GSTP1 IIe105Val polymorphisms with LC risk in the overall population. We also firstly reported the credibility of these genetic polymorphisms with LC risk using the FPRP, BFDP, and the Venice criteria.

Overall, a statistically significantly increased LC risk was observed in both individual and combined effects of the GSTM1, GSTT1, and GSTP1 polymorphisms in the current meta-analysis. However, the pooled P value must be adjusted because the present meta-analysis applied several subgroup analyses and genetic models at the expense of multiple comparisons.[63] In addition, random error and bias were common in the studies with small sample sizes so that the results were unreliable, especially in molecular epidemiological studies. Furthermore, small sample studies were easier to accept if there were positive reports as they tend to yield false-positive results because they may be not rigorous and are often of low quality. Figures 1 to 12, Supplemental Digital Content indicated that the asymmetry of the funnel plot was caused by a study of low-quality small samples. FPRP was reported to be an appropriate approach for assessing the probability of a positive result, “noteworthiness,” on the multiple hypothesis testing of molecular epidemiology studies.[59] Wakefield[60] in 2007 proposed a more precise Bayesian measure of false discovery in genetic epidemiology studies, for determining the “noteworthiness” of the positive association.[60] Hence, we considered FPRP and BFDP test to assess the false discovery in the current meta-analysis. Lack of replication or higher between-study heterogeneity (I2 > 50%) may be potential errors and biases, including genotype error, phenotype misclassification, population stratification, and selective reporting biases.[64–67] In addition, statistical power was also an important influence factor. A large amount of evidence (statistical power >80%) can reach a more stringent level of statistical significance or decreased lower false-discovery rate.[9] Therefore, we also applied for the Venice criteria to assess the credibility of the current meta-analysis.

Based on biochemical properties described for GSTM1, GSTT1, and GSTP1 polymorphisms, we expected that the individual and the combined effects of these genes were associated with the risk of LC risk in all races. However, the significant associations were considered in the Japanese population on the GSTM1 null genotype with LC risk, Asians on GSTM1 null genotype with lung AC risk, Chinese population on GSTT1 null genotype with LC risk, GSTT1 null genotype with lung AC risk in Asians, and Asians on GSTP1 IIe105Val polymorphism with LC risk as “highly credible” or “positive” results when we applied the FPRP, BFDP, and the Venice criteria to assess the credibility. These results indicated that the same genes may play different roles in cancer susceptibility in different races and countries, because cancer is a complicated multi genetic disease, and different genetic backgrounds and environmental factors (smoking or lifestyle) may contribute to the discrepancy.[30] It was a pity that all other significant associations were considered as “less-credible” (higher heterogeneity, lower statistic power, FPRP > 0.2 and BRDP > 0.8), such as the combined effects of GSTM1 and GSTT1 polymorphisms, GSTM1 and GSTP1 polymorphisms, GSTT1 and GSTP1 polymorphisms, GSTM1, GSTT1, and GSTP1 polymorphisms with lung cancer risk, and so on. These results indicated that potential gene–gene interactions are still required to investigate by a very much larger sample size. In addition, GSTM1 present/null (Table 13, Supplemental Digital Content) and GSTP1 IIe105Val (Table 15, Supplemental Digital Content) polymorphisms were not associated with LCLC risk, however, GSTT1 present/null was associated with LCLC risk (OR = 0.39, 95% CI = 0.17–0.94, Table 14, Supplemental Digital Content) in this meta-analysis.

We found that 8 studies only included 108 LCLC cases on GSTM1 present/null polymorphism, 3 studies only included 51 LCLC cases on GSTT1 present/null polymorphism, and 4 studies only included 193 LCLC cases on GSTP1 IIe105Val polymorphism. The results might be unreliable because random error and bias were common in the pooled meta-analysis with small sample sizes. Therefore, the results should be interpreted with caution and it is necessary that a well-designed large sample study to explore the true association on the 3 genetic polymorphisms with LCLC risk.

A total of 35 published meta-analyses[15–29,31–50] from 1995 to 2017 had been reported to investigate the individual and combined effects of GSTM1 present/null, GSTT1 present/null, and/or GSTP1 IIe105Val polymorphisms with LC risk. Several previous meta-analyses[15,18,19,21,28,33,34,40,41,50] indicated that the GSTM1 null genotype, GSTT1 null genotype, GSTP1 IIe105Val, the combined effects of GSTM1 present/null and GSTT1 present/null polymorphisms, and the combined effects of GSTM1 and GSTP1 were associated with significantly increased LC risk. However, when we applied the FPRP, BFDP, and the Venice criteria to evaluate the credibility of these meta-analyses, only 3 studies[18,33,41] were considered as “positive” results. In addition, a lot of studies did not be involved in the previously published meta-analysis, therefore their meta-analyses[18,33,41] are still not credible.

The present study has several limitations. First, only published studies were included in the current meta-analysis while positive results are known to be published more readily than negative ones. If negative results were included, an underestimation of the GSTM1 null effect may be observed. Second, we did not consider whether the genotype distribution in the controls was in HWE for GSTM1 and GSTT1 polymorphism because we cannot calculate the HWE on both genes. The current study also has several advantages over previously published meta-analyses.[15–29,31–50] First, the sample size was larger. There were 205 studies (45,726 LC cases and 58,788 controls for the GSTM1 null genotype, 103 studies (29,476 LC cases and 35,305 controls) for the GSTT1 null genotype, 69 studies (18,852 LC cases and 21,941 controls) for the GSTP1 IIe105Val polymorphism, and so on. Second, this is the first meta-analysis to investigate the combined effects of the 3 gene polymorphisms with LC risk in the overall population. Third, we collected more detailed data. Fourth, we evaluated the quality of the eligible studies. Fifth, we assess the credibility of the significant association in the current and previous meta-analyses.

In summary, this meta-analysis strongly indicated that the GSTM1 null genotype significantly increased LC risk in Japanese, GSTM1 null genotype was significantly increased lung AC risk in Asians, GSTT1 null genotype significantly increased LC risk in the Chinese population, and GSTP1 IIe105Val polymorphisms have an association with increased LC risk. Another significant association should be interpreted with caution and it is essential that future analyses be based on sample sizes well-powered to identify these variants having modest effects on LC risk, especially the combined effects of gene-gene.

Author contributions

The study was designed by Xiao-Feng He and Wei Wang. Chen Yang, Ling-Jun Xu, and Liang Song did the literature search, study quality assessment, and data extraction. Xiao-Feng He and Ling-Jun Xu performed the statistical analysis and drafted the tables and figures. Wen-Ping Zhang wrote the first draft of this analysis, and Xiao-Feng He and Wei Wang helped to finish the final version. All authors approved the conclusions of our study.

Conceptualization: Xiao-Feng He.

Data curation: Chen Yang, Ling-Jun Xu, Liang Song, Wen-Ping Zhang.

Formal analysis: Wei Wang, Liang Song, Xiao-Feng He, Wen-Ping Zhang.

Funding acquisition: No.

Investigation: Wei Wang, Ling-Jun Xu, Liang Song, Wen-Ping Zhang.

Methodology: Wei Wang, Liang Song, Wen-Ping Zhang, Xiao-Feng He.

Resources: Wei Wang, Chen Yang, Liang Song.

Software: Xiao-Feng He.

Supervision: Xiao-Feng He.

Validation: Xiao-Feng He.

Visualization: Xiao-Feng He.

Writing – original draft: Wen-Ping Zhang.

Writing – review & editing: Xiao-Feng He, Wei Wang.

Supplementary Material

Supplemental Digital Content
medi-100-e26104-s001.pdf (847.7KB, pdf)

Supplementary Material

Supplemental Digital Content
medi-100-e26104-s002.pdf (380.8KB, pdf)

Footnotes

Abbreviations: AC = adenocarcinoma, BFDP = Bayesian false discovery probability, CIs = confidence intervals, CNKI = China National Knowledge Infrastructure, FPRP = false-positive report probability, GSTM1 = glutathione S-transferase M1, GSTP1 = glutathione S-transferase P1, GSTs = glutathione S-transferases, GSTT1 = glutathione S-transferase T1, HB = hospital-based studies, HWE = Hardy–Weinberg equilibrium, LC = lung cancer, LCLC = large cell lung carcinoma, ORs = odds ratios, PB = population-based studies, SC = squamous carcinoma, SCLC = small-cell lung cancer.

How to cite this article: Zhang WP, Yang C, Xu LJ, Wang W, Song L, He XF. Individual and combined effects of GSTM1, GSTT1, and GSTP1 polymorphisms on lung cancer risk: a meta-analysis and re-analysis of systematic meta-analyses. Medicine. 2021;100:26(e26104).

WPZ, CY, and LJX contributed equally to this work.

This is a meta-analysis, hence, ethical approval was waived or not necessary.

The authors have no funding and conflicts of interest to disclose.

The datasets generated during and/or analyzed during the present study are publicly available.

Supplemental digital content is available for this article.

References

  • [1].Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. Global cancer statistics. CA Cancer J Clin 2011;61:69–90. [DOI] [PubMed] [Google Scholar]
  • [2].Krishnamurthy A, Vijayalakshmi R, Gadigi V, Ranganathan R, Sagar TG. The relevance of “Nonsmoking-associated lung cancer” in India: a single-centre experience. Indian J Cancer 2012;49:82–8. [DOI] [PubMed] [Google Scholar]
  • [3].Siegel R, Naishadham D, Jemal A. Cancer statistics, 2013. CA Cancer J Clin 2013;63:11–30. [DOI] [PubMed] [Google Scholar]
  • [4].Zhao P, Dai M, Chen W, Li N. Cancer trends in China. Jpn J Clin Oncol 2010;40:281–5. [DOI] [PubMed] [Google Scholar]
  • [5].Toh CK, Gao F, Lim WT, et al. Never smokers with lung cancer: epidemiologic evidence of a distinct disease entity. J Clin Oncol 2006;24:2245–51. [DOI] [PubMed] [Google Scholar]
  • [6].Luqman M, Javed MM, Daud S, Raheem N, Ahmad J, Khan AU. Risk factors for lung cancer in the Pakistani population. Asian Pac J Cancer Prev 2014;15:3035–9. [DOI] [PubMed] [Google Scholar]
  • [7].Hayes JD, Flanagan JU, Jowsey IR. Glutathione transferases. Annu Rev Pharmacol Toxicol 2005;45:51–88. [DOI] [PubMed] [Google Scholar]
  • [8].Udomsinprasert R, Pongjaroenkit S, Wongsantichon J, et al. Identification, characterization and structure of a new Delta class glutathione transferase isoenzyme. Biochem J 2005;388:763–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [9].Benjamini Y, Hochberg Y. Controlling the false discovery rate-a practical and powerful approach to multiple testing. J Royal Stat Soc B 1995;57:289–300. [Google Scholar]
  • [10].Board PG, Baker RT, Chelvanayagam G, Jermiin S. Zeta, a novel class of glutathione transferases in a range of species from plants to humans. Biochem J 1997;328:929–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [11].Hayes J, Strange R. Glutathione S-transferase polymorphisms and their biological consequences. Pharmacology 2000;61:154–66. [DOI] [PubMed] [Google Scholar]
  • [12].Mitrunen K, Jourenkova N, Kataja V, et al. Glutathione S-transferase M1, M3, P1, and T1 genetic polymorphisms and susceptibility to breast cancer. Cancer Epidemiol Biomarkers Prev 2001;10:229–36. [PubMed] [Google Scholar]
  • [13].Zimniak P, Nanduri B, Pikuła S, et al. Naturally occurring human glutathione S-transferase GSTP1-1 isoforms with isoleucine and valine in position 104 differ in enzymic properties. Eur J Biochem 1994;224:893–9. [DOI] [PubMed] [Google Scholar]
  • [14].Ali-Osman F, Akande O, Antoun G, Mao JX, Buolamwini J. Molecular cloning, characterization, and expression in Escherichia coli of full-length cDNAs of three human glutathione S-transferase Pi gene variants. Evidence for differential catalytic activity of the encoded proteins. J Biol Chem 1997;272:10004–12. [DOI] [PubMed] [Google Scholar]
  • [15].Ye Z, Song H, Higgins JP, Pharoah P, Danesh J. Five glutathione s-transferase gene variants in 23,452 cases of lung cancer and 30,397 controls: meta-analysis of 130 studies. PLoS Med 2006;3:e91. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [16].Zhong H, Feng Y, Zheng GX, et al. A meta-analysis of the association between glutathione S-transferase P1 gene polymorphism and the risk of adenocarcinomas of lung cancer. Cancer Biomark 2013;13:29–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [17].Zhou HF, Feng X, Zheng BS, Qian J, He W. A meta-analysis of the relationship between glutathione S-transferase T1 null/presence gene polymorphism and the risk of lung cancer including 31802 subjects. Mol Biol Rep 2013;40:5713–21. [DOI] [PubMed] [Google Scholar]
  • [18].Gao Y, Gao F, Hu TT, Li G, Sui YX. Combined effects of glutathione S-transferase M1 and T1 polymorphisms on risk of lung cancer: evidence from a meta-analysis. Oncotarget 2017;8:28135–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [19].Yang H, Shen X, Li B, Ma R. Association between glutathione S-transferase T1 null genotype and risk of lung cancer: a meta-analysis of 55 studies. Tumour Biol 2014;35:2359–66. [DOI] [PubMed] [Google Scholar]
  • [20].Liu H, Ma HF, Chen YK. Association between GSTM1 polymorphisms and lung cancer: an updated meta-analysis. Genet Mol Res 2015;14:1385–92. [DOI] [PubMed] [Google Scholar]
  • [21].Feng X, Zheng BS, Shi JJ, Qian J, He W, Zhou HF. Association of glutathione S-transferase P1 gene polymorphism with the susceptibility of lung cancer. Mol Biol Rep 2012;39:10313–23. [DOI] [PubMed] [Google Scholar]
  • [22].Feng X, Zhou HF, Zheng BS, Shi JJ, Luo C, Qin JJ. Association of glutathione S-transferase P1 gene polymorphism with the histological types of lung cancer: a meta-analysis. Mol Biol Rep 2013;40:2439–47. [DOI] [PubMed] [Google Scholar]
  • [23].Ou C, Zhu B, Zhao HL, et al. Association of glutathione S-transferase P1 gene polymorphism with the risk of small-cell carcinoma of lung cancer. J Recept Signal Transduct Res 2015;35:165–9. [DOI] [PubMed] [Google Scholar]
  • [24].Carlsten C, Sagoo GS, Frodsham AJ, Burke W, Higgins JP. Glutathione S-transferase M1 (GSTM1) polymorphisms and lung cancer: a literature-based systematic HuGE review and meta-analysis. Am J Epidemiol 2008;167:759–74. [DOI] [PubMed] [Google Scholar]
  • [25].McWilliams JE, Sanderson BJ, Harris EL, Richert-Boe KE, Henner WD. Glutathione S-transferase M1 (GSTM1) deficiency and lung cancer risk. Cancer Epidemiol Biomarkers Prev 1995;4:589–94. [PubMed] [Google Scholar]
  • [26].Houlston RS. Glutathione S-transferase M1 status and lung cancer risk: a meta-analysis. Cancer Epidemiol Biomarkers Prev 1999;8:675–82. [PubMed] [Google Scholar]
  • [27].Chen X, Liang L, Hu X, Chen Y. Glutathione S-transferase P1 gene Ile105Val polymorphism might be associated with lung cancer risk in the Chinese Han population. Tumour Biol 2012;33:1973–81. [DOI] [PubMed] [Google Scholar]
  • [28].Li XM, Yu XW, Yuan Y, et al. Glutathione S-transferase P1, gene-gene interaction, and lung cancer susceptibility in the Chinese population: an updated meta-analysis and review. J Cancer Res Ther 2015;11:565–70. [DOI] [PubMed] [Google Scholar]
  • [29].Wang Y, Yang H, Li L, Wang H. Glutathione S-transferase T1 gene deletion polymorphism and lung cancer risk in Chinese population: a meta-analysis. Cancer Epidemiol 2010;34:593–7. [DOI] [PubMed] [Google Scholar]
  • [30].Hirschhorn JN, Lohmueller K, Byrne E. A comprehensive review of genetic association studies. Genet Med 2002;4:45–61. [DOI] [PubMed] [Google Scholar]
  • [31].Zhao Y, Zeng J, Zhang Y, et al. GSTM1 polymorphism and lung cancer risk among East Asian populations: a meta-analysis. Tumour Biol 2014;35:6493–500. [DOI] [PubMed] [Google Scholar]
  • [32].Chen XP, Xu WH, Xu DF, Xie XH, Yao J, Fu SM. GSTM1 polymorphisms and lung cancer risk in the Chinese population: a meta-analysis based on 47 studies. Asian Pac J Cancer Prev 2014;15:7741–6. [DOI] [PubMed] [Google Scholar]
  • [33].Xu CH, Wang Q, Zhan P, Qian Q, Yu LK. GSTP1 Ile105Val polymorphism is associated with lung cancer risk among Asian population and smokers: an updated meta-analysis. Mol Biol Rep 2014;41:4199–212. [DOI] [PubMed] [Google Scholar]
  • [34].Yang X, Qiu MT, Hu JW, et al. GSTT1 null genotype contributes to lung cancer risk in asian populations: a meta-analysis of 23 studies. PLoS One 2013;8:e62181. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [35].Liu HZ, Peng J, Zheng F, Wang CH, Han MJ. Lack of association of glutathione S-transferase T1 gene null and susceptibility to lung cancer in china: a meta-analysis. Asian Pac J Cancer Prev 2013;14:7215–9. [DOI] [PubMed] [Google Scholar]
  • [36].Benhamou S, Lee WJ, Alexandrie AK, et al. Meta- and pooled analyses of the effects of glutathione S-transferase M1 polymorphisms and smoking on lung cancer risk. Carcinogenesis 2002;23:1343–50. [DOI] [PubMed] [Google Scholar]
  • [37].Cote ML, Chen W, Smith DW, et al. Meta- and pooled analysis of GSTP1 polymorphism and lung cancer: a HuGE-GSEC review. Am J Epidemiol 2009;169:802–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [38].Raimondi S, Paracchini V, Autrup H, et al. Meta- and pooled analysis of GSTT1 and lung cancer: a HuGE-GSEC review. Am J Epidemiol 2006;164:1027–42. [DOI] [PubMed] [Google Scholar]
  • [39].Liu X, Li Z, Zhang Z, et al. Meta-analysis of GSTM1 null genotype and lung cancer risk in Asians. Med Sci Monit 2014;20:1239–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [40].Wang Y, Yang H, Wang H. The association of GSTT1 deletion polymorphism with lung cancer risk among Chinese population: evidence based on a cumulative meta-analysis. Onco Targets Ther 2015;8:2875–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [41].Liu K, Lin X, Zhou Q, et al. The associations between two vital GSTs genetic polymorphisms and lung cancer risk in the Chinese population: evidence from 71 studies. PLoS One 2014;9:e102372. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [42].Zhang Y, Liu C. The interaction between smoking and GSTM1 variant on lung cancer in the Chinese population. Tumour Biol 2013;34:395–401. [DOI] [PubMed] [Google Scholar]
  • [43].Gui Q, Xiong J, Zhu C, Lu X. The present/null polymorphism in the GSTT1 gene and the risk of lung cancer in Chinese population. Tumour Biol 2013;34:3465–9. [DOI] [PubMed] [Google Scholar]
  • [44].Zhao Y, Wang B, Hu K, et al. Glutathione S-transferase θ1 polymorphism contributes to lung cancer susceptibility: a meta-analysis of 26 case–control studies. Oncol Lett 2015;9:1947–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [45].Yang H, Yang S, Liu J, Shao F, Wang H, Wang Y. The association of GSTM1 deletion polymorphism with lung cancer risk in Chinese population: evidence from an updated meta-analysis. Sci Rep 2015;5:9392. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [46].Ma W, Zhang BJ, Hua F, et al. A meta-analysis on relationship between GSTP1 IIe105Val genetic polymorphisms and lung cancer risk. Cancer Res Prev Treat 2011;38:1311–5. [Google Scholar]
  • [47].An WJ, Feng XY, Jiu T, et al. Association between GSTT1 polymorphism and lung cancer susceptibility in Chinese population: a meta-analysis. Chin J Gerontol 2017;37:3244–6. [Google Scholar]
  • [48].Bao XL, Yang L. Correlation between glutathione S-transferase M1 polymorphism and risk of lung cancer in non-smoking Chinese: a meta-analysis. J Chin PLA Postgrad Med Sch 2012;33:849–52. 92. [Google Scholar]
  • [49].Huang ZM, Li ZZ, Feng Y, et al. A meta-analysis of GSTM1 gene polymorphism and lung cancer risk in Chinese population. China Cancer 2014;23:883–90. [Google Scholar]
  • [50].Sengupta D, Guha U, Bhattacharjee S, Sengupta M. Association of 12 polymorphic variants conferring genetic risk to lung cancer in Indian population: an extensive meta-analysis. Environ Mol Mutagen 2017;58:688–700. [DOI] [PubMed] [Google Scholar]
  • [51].Stroup DF, Berlin JA, Morton SC, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA 2000;283:2008–12. [DOI] [PubMed] [Google Scholar]
  • [52].Thakkinstian A, McKay GJ, McEvoy M, et al. Systematic review and meta-analysis of the association between complement component 3 and age-related macular degeneration: a HuGE review and meta-analysis. Am J Epidemiol 2011;173:1365–79. [DOI] [PubMed] [Google Scholar]
  • [53].Xue WQ, He YQ, Zhu JH, Ma JQ, He JH, Jia WH. Association of BRCA2 N372H polymorphism with cancer susceptibility: a comprehensive review and meta-analysis. Sci Rep 2014;4:6791. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [54].Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. Natl Cancer Inst 1959;22:719–48. [PubMed] [Google Scholar]
  • [55].DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7:177–88. [DOI] [PubMed] [Google Scholar]
  • [56].Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics 1994;50:1088–101. [PubMed] [Google Scholar]
  • [57].Egger M, Smith DG, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. Br Med J 1997;315:629–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [58].Dual S, Tweedie R. A nonparametric “trim and fill” method of accounting for publication bias in meta-analysis. J Am Stat Assoc 2000;95:89–98. [Google Scholar]
  • [59].Wacholder S, Chanock S, Garcia-Closas M, El Ghormli L, Rothman N. Assessing the probability that a positive report is false: an approach for molecular epidemiology studies. J Natl Cancer Inst 2004;96:434–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [60].Wakefield J. A Bayesian measure of the probability of false discovery in genetic epidemiology studies. Am J Hum Genet 2007;81:208–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [61].Ioannidis JP, Boffetta P, Little J, et al. Assessment of cumulative evidence on genetic associations: interim guidelines. Int J Epidemiol 2008;37:120–32. [DOI] [PubMed] [Google Scholar]
  • [62].Theodoratou E, Montazeri Z, Hawken S, et al. Systematic meta-analyses and field synopsis of genetic association studies in colorectal cancer. J Natl Cancer Inst 2012;104:1433–57. [DOI] [PubMed] [Google Scholar]
  • [63].Attia J, Thakkinstian A, D’Este C. Meta-analyses of molecular association studies: methodologic lessons for genetic epidemiology. J Clin Epidemiol 2003;56:297–303. [DOI] [PubMed] [Google Scholar]
  • [64].Balding DJ. A tutorial on statistical methods for population association studies. Nat Rev Genet 2006;7:781–91. [DOI] [PubMed] [Google Scholar]
  • [65].Pompanon F, Bonin A, Bellemain E, Taberlet P. Genotyping errors: causes, consequences and solutions. Nat Rev Genet 2005;6:847–59. [DOI] [PubMed] [Google Scholar]
  • [66].Clayton DG, Walker NM, Smyth DJ, et al. Population structure, differential bias and genomic control in a largescale, case–control association study. Nat Genet 2005;37:1243–6. [DOI] [PubMed] [Google Scholar]
  • [67].Page GP, George V, Go RC, et al. ‘Are we there yet?’: deciding when one has demonstrated specific genetic causation in complex diseases and quantitative traits. Am J Hum Genet 2003;73:711–9. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Digital Content
medi-100-e26104-s001.pdf (847.7KB, pdf)
Supplemental Digital Content
medi-100-e26104-s002.pdf (380.8KB, pdf)

Articles from Medicine are provided here courtesy of Wolters Kluwer Health

RESOURCES