Abstract

The melatonin receptor 1B (MTNR1B) polymorphism rs10830963 C>G has been reported to be associated with the risk of gestational diabetes mellitus (GDM) with inconsistent results. To clarify the effect of the polymorphism on the risk of GDM, a meta-analysis therefore was performed. Pooled OR with its corresponding 95%CI was used to estimate the strength of the association. Totally 14 eligible studies with a number of 5033 GDM patients and 5614 controls were included in this meta-analysis. Results indicated that the variant G allele was significantly associated with an increased GDM risk (CG vs. CC: OR = 1.25, 95% CI = 1.11−1.40, P < 0.001; GG vs. CC: OR = 1.78, 95% CI = 1.45−2.19, P < 0.001; G vs. C: OR = 1.33, 95% CI = 1.21−1.47, P < 0.001). In the stratified analysis by ethnicity, similar results were found in Asians (CG vs. CC: OR = 1.15, 95%CI = 1.02−1.28, P = 0.020; GG vs. CC: OR = 1.52, 95% CI = 1.23−1.89, P < 0.001; G vs. C: OR = 1.23, 95% CI = 1.10−1.37, P < 0.001) and in Caucasians (CG vs. CC: OR = 1.40, 95% CI = 1.16−1.70, P < 0.001; GG vs. CC: OR = 2.21, 95% CI = 1.54−3.17, P < 0.001; G vs. C: OR = 1.47, 95% CI = 1.24−1.73, P < 0.001). FPRP and TSA analyses confirmed findings support that the rs10830963 G allele increases the risk of GDM, and further functional experimental studies are warranted to explore and clarify the potential mechanism.

Introduction

Gestational diabetes mellitus (GDM) is defined as abnormal glucose tolerance with onset or first recognition during pregnancy [1]. Worldwide, it affects approximately 2–20% of all pregnancies [2]. GDM has been shown to be associated with poor pregnancy outcome and substantial long-term adverse consequences for mothers and their offspring [3–7]. So far, the major risk factors related to GDM are older age at pregnancy, obesity, family history of T2DM and past history of GDM, previous poor obstetric history and genetics [8–12]. Insulin secretory defect accompanied by peripheral insulin resistance is an important characteristic of GDM [13]. Melatonin receptor 1B (MTNR1B) is an integral membrane protein that coupled to an inhibitory G protein and is expressed in pancreatic islets and pancreatic β cells [14–16]. It has been found that the increased expression of MTNR1B on β cells diminished intracellular cyclic cAMP levels, thereby inhibited the insulin secretion [17–19]. These findings suggested that MTNR1B may be involved in the development of GDM.

MTNR1B is located on human chromosome 11q21-q22, spanning about 22 kb and consisting of 3 exons and 1 intron. So far, 64 single-nucleotide polymorphisms (SNPs) have been validated in the MTNR1B gene (http://www.ncbi.nlm.nih.gov/SNP), and some of which were reported to be associated with GDM risk. The SNP rs10830963 is located in the unique intron between exon 1 (+5.6 kb) and exon 2 (−5.9 kb) of MTNR1B gene. Genotype–phenotype study of the SNP rs10830963 C>G showed that compared with the wild-type C allele of rs10830963, the variant G allele was associated with increased MTNR1B transcript levels in human islets [20]. The study of Li et al. indicated that G allele carrying genotype means a higher MTNR1B protein level, fasting blood glucose, fasting insulin and homeostasis model assessment for insulin resistance [21].

Because of the functional consequence of rs10830963 C>G, many association studies have examined its effect on the risk of GDM [22–35]. However, these studies presented inconsistent results. To clarify the effect of the MTNR1B rs10830963 C>G on the risk of GDM, we therefore performed a meta-analysis with a total of 5033 GDM patients and 5614 controls from 14 published case–control studies.

Materials and methods

Literature search strategy

We searched NCBI PubMed, Google Scholar and the Chinese National Knowledge Infrastructure (CNKI) databases for the association studies of MTNR1B rs10830963 polymorphism with the risk of GDM. The following key words ‘MTNR1B’ or ‘Melatonin receptor 1B’, ‘gestational diabetes mellitus’ and ‘variation’ or ‘polymorphism’ were used. The corresponding Chinese terms were used in the Chinese library. All studies were published up to March 1, 2019. In addition, we manually searched for additional published studies on this topic in the references cited in the retrieved studies.

The included studies in this meta-analysis had to meet the following criteria: evaluation of the MTNR1B rs10830963 C>G polymorphism and GDM risk; case–control study; genotype or allele distribution information in cases and controls for calculating odds ratio (OR) with corresponding 95% confidence interval (CI); the study was written in English or Chinese. Accordingly, family based studies, abstract, case reports, comments and reviews were excluded. If studies had overlapped subjects, only the largest study was included in the final analysis.

Data extraction and quality score assessment

Two professional investigators (Huang B and Wang Y) independently reviewed the articles and extracted the data from eligible publications. The following information was extracted from each included study: first author, year of publication, country, diagnostic criteria, source of controls, number of cases and controls, genotype or allele distribution data of cases and controls, mean age, mean body mass index (BMI) and P-value of the Chi-square goodness of fit test for Hardy–Weinberg equilibrium (HWE) in controls.

In the present study, the Newcastle–Ottawa scale (NOS) scale was used to assess the quality of the eligible studies by Yu, X.Y. and Wang, Y.K. independently [36]. A score range of the scale was from 0 (the lowest) to 9 (the highest), and study with a score <5 were considered to be of low quality and those ≥5 to be of high quality. If there is disagreement in quality assessment, it can be resolved through discussion.

Statistical analysis

Deviation of genotype frequencies of the MTNR1B rs10830963 C>G polymorphism in controls from HWE was tested by using the Chi-square goodness of fit test, and a P-value less than 0.05 was considered a departure from HWE. The odds ratio (OR) and 95% confidence interval (CI) were used to assess the strength of association between the MTNR1B rs10830963 C>G and GDM risk. The heterogeneity across the studies was assessed by the Q test, and was considered significant when a P-value less than 0.1 [37]. A fixed-effect model was used to calculate the pooled OR if the heterogeneity was not significant, otherwise, the random-effect model was adopted [38]. The potential source of heterogeneity across the included studies was explored with meta-regression analyses by ethnicity (Asian and Caucasian), diagnostic criteria(ADA and others) and assessed literature quality (high and low). Both Begg’s and Egger’s tests were used to test for publication bias [39,40]. A P-value < 0.05 was considered as an indication for the potential presence of publication bias. Sensitivity analyses were done to assess the influence of individual study on the pooled ORs. All analyses were performed by using Stata software, version 12.0 (Stata Corp LP, College Station, TX, U.S.A.). In addition, false positive report probability (FPRP) was estimated to assess the robustness of foundings statistically significant association by using the method described by Wacholder et al. [41]. The FPRP threshold was set to 0.2, and the prior probability was set to 0.1 to detect the noteworthiness for OR of 1.5 or 0.67, with an alpha level equal to the observed P-value. An FPRP less than 0.2 was considered as a noteworthy association [41,42].

Trial sequential analysis (TSA)

Meta-analysis might be affected by type I error due to the increased risk of random error and repeated significance testing [43]. Trial sequential analysis (TSA) was used to reduce random errors and increase the robustness of the conclusions by estimating the amount of required information size (RIS) and the threshold for statistical significance [44]. A 5% significance level for type I errors, 20% significance level for type II errors (80% power) and 20% relative risk reduction (RRR) were defined and a TSA monitoring boundary were determined. TSA was conducted in allelic model and positive results of meta-analysis were tested. When the cumulative Z-curve crosses the TSA boundary or enters the insignificance area, a sufficient level of evidence has been reached, and no further studies are necessary. However, when the Z-curve does not exceed any of the boundaries and the required sample size has not been reached, evidence to reach a conclusion is insufficient [45]. Review Manager (RevMan) version 5.2 and TSA version 0.9.5.10 beta softwares were used in data processing.

Results

Characteristics of included studies

The flowchart of study selection for this meta-analysis is presented in Figure 1. A total of 69 studies were found using our literature search strategy, of which 43 studies were excluded because of duplicates or not on the topic of polymorphisms and GDM risk. After full-text reviews of the remaining 26 articles, 12 studies were excluded for the following reasons: 2 studies were case only studies, 3 studies were review or meta-analysis articles, 7 studies didn’t focus on the topic of the MTNR1B rs10830963 C>G and GDM risk. Finally, 14 studies with 5033 GDM patients and 5614 controls were selected in present study. According to the evaluation of NOS scale, 12 of the included literatures were considered to be of high quality (score ≥ 5) and 2 of them were of low quality (score < 5). Main characteristics of the included studies are shown in Table 1. The genotype frequency distributions of the rs10830963 C>G in controls were in agreement with HWE in all included studies except for the two by Vlassi et al. [25] and Liu et al. [32].

Flowchart of the process of identification of eligible studies

Figure 1
Flowchart of the process of identification of eligible studies
Figure 1
Flowchart of the process of identification of eligible studies
Table 1
Characteristics of the studies included in the meta-analysis
Author, year Country Diagnostic criteria Genotyping methods Controls No. of case/ control MAF case/ control Mean age of cases/controls Mean BMI of cases/controls pHWE for controls NOS score 
Deng, Z., 2011 China ADA Sequencing NGT 87/91 0.52/0.41 31.8 ± 4.6/29.7 ± 3.5 23.6 ± 3.0/21.5 ± 2.4 0.84 
Kim, J.Y., 2011 Korea ADA TaqMan NGT 908/966 0.52/0.45 33.1/32.2 23.3 ± 4.0/21.4 ± 2.9 0.53 
Wang, Y., 2011 China ADA TaqMan NGT 700/1029 0.46/0.43 30.0/32.0 21.5/21.7 0.81 
VlassiM, 2012 Greece ADA PCR-RFLP NGT 77/98 0.41/0.28 35.4 ± 4.4/31.3 ± 5.2 25.8 ± 5.1/26.7 ± 6.2 0.02 
HuopioH, 2013 Finland ADA Sequenom Assay/TaqMan NGT 533/407 0.47/0.35 32.6/29.9 26.3 ± 4.7/24.1 ± 3.8 0.98 
Li. C., 2013 China IADPSG PCR-RFLP NGT 350/480 0.45/0.40 32.4 ± 4.8/31.9 ± 5.2 25.3 ± 5.2/24.6 ± 4.6 0.79 
Qi, J., 2013 China IADPSG Sequencing NGT 110/110 0.54/0.44 28.7 ± 3.1/28.1 ± 2.4 NA/NA 0.43 
Vejrazkova, D., 2014 Czech WHO TaqMan NGT 458/422 0.38/0.29 34.1 ± 6.1/34.8 ± 15.1 24.3 ± 4.9/23.7 ± 4.2 0.48 
Wang, X., 2014 China ADA PCR-RFLP NGT 184/235 0.42/0.45 28.2 ± 3.8/27.9 ± 4.1 21.2 ± 1.8/20.7 ± 1.4 0.53 
Junior, J.P., 2015 Brazil ADA Real-time PCR Healthy pregnant 183/183 0.28/0.20 32/29 32.0/25.4 0.11 
Liu, Q., 2015 China ADA TaqMan NGT 674/674 0.51/0.44 31.6/32.1 24.4/25.2 0.02 
Tarnowski, M., 2017 Poland IADPSG TaqMan NGT 204/207 0.39/0.31 31.7 ± 4.5/29.2 ± 5.0 25.1 ± 5.5/23.0 ± 4.0 0.112 
Popova, P.V., 2017 Russia ADA RT-PCR Healthy pregnant 278/179 0.35/0.31 31.8 ± 4.8/29.4 ± 4.8 25.7 ± 5.9/22.9 ± 4.5 0.426 
Rosta, K., 2017 Hungary and Austria IADPSG KASP assay — 287/533 0.33/0.30 — — 0.975 
Author, year Country Diagnostic criteria Genotyping methods Controls No. of case/ control MAF case/ control Mean age of cases/controls Mean BMI of cases/controls pHWE for controls NOS score 
Deng, Z., 2011 China ADA Sequencing NGT 87/91 0.52/0.41 31.8 ± 4.6/29.7 ± 3.5 23.6 ± 3.0/21.5 ± 2.4 0.84 
Kim, J.Y., 2011 Korea ADA TaqMan NGT 908/966 0.52/0.45 33.1/32.2 23.3 ± 4.0/21.4 ± 2.9 0.53 
Wang, Y., 2011 China ADA TaqMan NGT 700/1029 0.46/0.43 30.0/32.0 21.5/21.7 0.81 
VlassiM, 2012 Greece ADA PCR-RFLP NGT 77/98 0.41/0.28 35.4 ± 4.4/31.3 ± 5.2 25.8 ± 5.1/26.7 ± 6.2 0.02 
HuopioH, 2013 Finland ADA Sequenom Assay/TaqMan NGT 533/407 0.47/0.35 32.6/29.9 26.3 ± 4.7/24.1 ± 3.8 0.98 
Li. C., 2013 China IADPSG PCR-RFLP NGT 350/480 0.45/0.40 32.4 ± 4.8/31.9 ± 5.2 25.3 ± 5.2/24.6 ± 4.6 0.79 
Qi, J., 2013 China IADPSG Sequencing NGT 110/110 0.54/0.44 28.7 ± 3.1/28.1 ± 2.4 NA/NA 0.43 
Vejrazkova, D., 2014 Czech WHO TaqMan NGT 458/422 0.38/0.29 34.1 ± 6.1/34.8 ± 15.1 24.3 ± 4.9/23.7 ± 4.2 0.48 
Wang, X., 2014 China ADA PCR-RFLP NGT 184/235 0.42/0.45 28.2 ± 3.8/27.9 ± 4.1 21.2 ± 1.8/20.7 ± 1.4 0.53 
Junior, J.P., 2015 Brazil ADA Real-time PCR Healthy pregnant 183/183 0.28/0.20 32/29 32.0/25.4 0.11 
Liu, Q., 2015 China ADA TaqMan NGT 674/674 0.51/0.44 31.6/32.1 24.4/25.2 0.02 
Tarnowski, M., 2017 Poland IADPSG TaqMan NGT 204/207 0.39/0.31 31.7 ± 4.5/29.2 ± 5.0 25.1 ± 5.5/23.0 ± 4.0 0.112 
Popova, P.V., 2017 Russia ADA RT-PCR Healthy pregnant 278/179 0.35/0.31 31.8 ± 4.8/29.4 ± 4.8 25.7 ± 5.9/22.9 ± 4.5 0.426 
Rosta, K., 2017 Hungary and Austria IADPSG KASP assay — 287/533 0.33/0.30 — — 0.975 

Note: ADA, American Diabetes Association; IADPSG, International Association of the Diabetes and Pregnancy Study Groups; NGT, Normal Glucose Tolerance; MAF, Minor Allele Frequency; BMI, Body Mass Index; HWE, Hardy–Weinberg Equilibrium; NOS, Newcastle–Ottawa Scale.

Association between the MTNR1B rs10830963 C>G and GDM risk

As shown in Table 2 and Figures 24, we found that the variant G allele of rs10830963 polymorphism was significantly associated with an increased risk of GDM (CG vs. CC: OR = 1.25, 95% CI = 1.11−1.40, P < 0.001, Pheterogeneity = 0.090, I2 = 35.6%; GG vs. CC: OR = 1.78, 95% CI = 1.45−2.19, P < 0.001, Pheterogeneity = 0.001, I2 = 61.9%. G vs. C: OR = 1.33, 95% CI = 1.21−1.47, P < 0.001, Pheterogeneity = 0.001, I2 = 64.1%), respectively.

Forest plot on the risk of GDM associated with rs10830963 (CG vs. CC)

Figure 2
Forest plot on the risk of GDM associated with rs10830963 (CG vs. CC)
Figure 2
Forest plot on the risk of GDM associated with rs10830963 (CG vs. CC)

Forest plot on the risk of GDM associated with rs10830963 (GG vs. CC)

Figure 3
Forest plot on the risk of GDM associated with rs10830963 (GG vs. CC)
Figure 3
Forest plot on the risk of GDM associated with rs10830963 (GG vs. CC)

Forest plot on the risk of GDM associated with rs10830963 (G vs. C)

Figure 4
Forest plot on the risk of GDM associated with rs10830963 (G vs. C)
Figure 4
Forest plot on the risk of GDM associated with rs10830963 (G vs. C)
Table 2
Meta-analysis of the MTNR1B rs10830963 polymorphism on GDM risk
Subgroup Heterozygous (CG vs. CC) Homozygous (GG vs. CC) Allele mogel (G vs. C) 
 No. of studies Case/ Control OR (95% CI) PEffect No. of studies Case/ Control OR (95% CI) PEffect No. of studies Case/Control OR (95% CI) PEffect 
Overall 14 3952/4736 1.25 (1.11–1.40) <0.001 14 2628/2966 1.78 (1.45–2.19) <0.001 14 10066/11228 1.33 (1.21–1.47) <0.001 
Ethnicity             
Asian 2271/2916 1.15 (1.02–1.28) 0.020 1543/1796 1.52 (1.23–1.89) <0.001 6026/7170 1.23 (1.10–1.37) <0.001 
Caucasian 1681/1820 1.40 (1.16–1.70) <0.001 1085/1170 2.21 (1.54–3.17) <0.001 4040/4058 1.47 (1.24–1.73) <0.001 
Subgroup Heterozygous (CG vs. CC) Homozygous (GG vs. CC) Allele mogel (G vs. C) 
 No. of studies Case/ Control OR (95% CI) PEffect No. of studies Case/ Control OR (95% CI) PEffect No. of studies Case/Control OR (95% CI) PEffect 
Overall 14 3952/4736 1.25 (1.11–1.40) <0.001 14 2628/2966 1.78 (1.45–2.19) <0.001 14 10066/11228 1.33 (1.21–1.47) <0.001 
Ethnicity             
Asian 2271/2916 1.15 (1.02–1.28) 0.020 1543/1796 1.52 (1.23–1.89) <0.001 6026/7170 1.23 (1.10–1.37) <0.001 
Caucasian 1681/1820 1.40 (1.16–1.70) <0.001 1085/1170 2.21 (1.54–3.17) <0.001 4040/4058 1.47 (1.24–1.73) <0.001 

In the stratified analysis by ethnicity, as shown in Table 2 and Figures 24, we found that rs10830963 polymorphism was significantly associated with a relatively higher GDM risk in the above three models in Asians (CG vs. CC: OR = 1.15, 95% CI = 1.02−1.28, P = 0.020, Pheterogeneity = 0.633, I2 = 0.0%; GG vs. CC: OR = 1.52, 95% CI = 1.23−1.89, P < 0.001, Pheterogeneity = 0.069, I2 = 48.7%; G vs. C: OR = 1.23, 95% CI = 1.10−1.37, P < 0.001, Pheterogeneity = 0.064, I2 = 49.7%) and in Caucasians (CG vs. CC: OR = 1.40, 95% CI = 1.16−1.70, P < 0.001, Pheterogeneity = 0.088, I2 = 45.5%; GG vs. CC: OR = 2.21, 95% CI = 1.54−3.17, P < 0.001, Pheterogeneity = 0.013, I2 = 63.0%; G vs. C: OR = 1.47, 95% CI = 1.24−1.73, P < 0.001, Pheterogeneity = 0.010, I2 = 64.5%).

Evaluation of heterogeneity

In this present study, the Q test was used to evaluate the heterogeneity across the included studies and the heterogeneity across studies was found in most of comparisons. As shown in Table 2, the Q test suggested that a low to high heterogeneity across studies presented in most of comparisons. We then used the meta-regression analysis to explore the source of heterogeneity by ethnicity (Asian and Caucasian), diagnostic criteria (ADA and others) and assessed literature quality (score ≥ 5 and <5), and found that they didn’t contribute to the main observed heterogeneity across the studies, effect of ethnicity (CG vs. CC: t = −2.05, P = 0.063; GG vs. CC: t = −1.78, P = 0.101; G vs. C: t = 1.91, P = 0.080), diagnostic criteria (CG vs. CC: t = 1.03, P = 0.325; GG vs. CC: t = 1.41, P = 0.183; G vs. C: t = 1.26, P = 0.230) and literature quality (CG vs. CC: t = −0.98, P = 0.344; GG vs. CC: t = −0.22, P = 0.826; G vs. C: t = −0.10, P = 0.923).

Sensitivity analyses

Sensitivity analyses of the association between the MTNR1B rs10830963 C>G polymorphism and GDM risk were performed to assess the stability of the pooled ORs under the CG vs. CC and GG vs. CC comparisons. The leave-one-out analysis showed that no single study dramatically influenced the pooled ORs (Figures 5 and 6).

Sensitivity analyses of the association between rs10830963 C>G and GDM risk under the CG vs. CC comparison

Figure 5
Sensitivity analyses of the association between rs10830963 C>G and GDM risk under the CG vs. CC comparison
Figure 5
Sensitivity analyses of the association between rs10830963 C>G and GDM risk under the CG vs. CC comparison

Sensitivity analyses of the association between rs10830963 C>G and GDM risk under the GG vs. CC comparison

Figure 6
Sensitivity analyses of the association between rs10830963 C>G and GDM risk under the GG vs. CC comparison
Figure 6
Sensitivity analyses of the association between rs10830963 C>G and GDM risk under the GG vs. CC comparison

Publication bias

Both Begg’s and Egger’s tests were performed to evaluate the publication bias of the included studies. The shape of the funnel plots did not reveal any evidence of obvious asymmetry for all genetic models (Figures 7 and 8), and the Begg’s and Egger’s tests did not present any significantly statistical evidence of publication bias for any of the genetic models in the overall meta-analysis (CG vs. CC: PBegg’s = 0.913 and PEgger’s = 0.655, GG vs. CC: PBegg’s = 0.063 and PEgger’s = 0.186).

Begg’s funnel plot for publication bias test (CG vs. CC)

Figure 7
Begg’s funnel plot for publication bias test (CG vs. CC)
Figure 7
Begg’s funnel plot for publication bias test (CG vs. CC)

Begg’s funnel plot for publication bias test (GG vs. CC)

Figure 8
Begg’s funnel plot for publication bias test (GG vs. CC)
Figure 8
Begg’s funnel plot for publication bias test (GG vs. CC)

FPRP analysis results

FPRP was adopted to assess the noteworthiness of the significant associations between the MTNR1B rs10830963 C>G polymorphism and GDM risk. At the prior probability of 0.1 and FPRP cut-off value of 0.2, the FPRP values for the significant findings in the heterozygous genotype comparison (CG vs. CC), homozygote model (GG vs. CC) and allele model (G vs. C) of overall were 0.005, 0.008 and 0.004, respectively. Moreover, the FPRP values for the significant findings in the studied three models were 0.153, 0.009 and 0.010 in Asians, 0.007, 0.047 and 0.006 in Caucasians, respectively. As shown in Table 3.

Table 3
FPRP analysis for the significant associations of the MTNR1B rs10830963 C>G polymorphism and GDM risk
 OR (95% CI) Prior probability 
  0.25 0.1 0.01 0.001 0.0001 0.00001 
Overall        
CG vs. CC 1.25 (1.11–1.40) 0.002 0.005 0.056 0.375 0.857 0.984 
GG vs. CC 1.78 (1.45–2.19) 0.003 0.008 0.083 0.477 0.901 0.989 
G vs. C 1.33 (1.21–1.47) 0.001 0.004 0.038 0.286 0.800 0.976 
Asian        
CG vs. CC 1.15 (1.02–1.28) 0.057 0.153 0.664 0.952 0.995 1.000 
GG vs. CC 1.52 (1.23–1.89) 0.003 0.009 0.092 0.506 0.911 0.990 
G vs. C 1.23 (1.10–1.37) 0.003 0.010 0.097 0.519 0.915 0.991 
Caucasian        
CG vs. CC 1.40 (1.16–1.70) 0.002 0.007 0.074 0.446 0.889 0.988 
GG vs. CC 2.21 (1.54–3.17) 0.016 0.047 0.351 0.845 0.982 0.998 
G vs. C 1.47 (1.24–1.73) 0.002 0.006 0.060 0.393 0.866 0.985 
 OR (95% CI) Prior probability 
  0.25 0.1 0.01 0.001 0.0001 0.00001 
Overall        
CG vs. CC 1.25 (1.11–1.40) 0.002 0.005 0.056 0.375 0.857 0.984 
GG vs. CC 1.78 (1.45–2.19) 0.003 0.008 0.083 0.477 0.901 0.989 
G vs. C 1.33 (1.21–1.47) 0.001 0.004 0.038 0.286 0.800 0.976 
Asian        
CG vs. CC 1.15 (1.02–1.28) 0.057 0.153 0.664 0.952 0.995 1.000 
GG vs. CC 1.52 (1.23–1.89) 0.003 0.009 0.092 0.506 0.911 0.990 
G vs. C 1.23 (1.10–1.37) 0.003 0.010 0.097 0.519 0.915 0.991 
Caucasian        
CG vs. CC 1.40 (1.16–1.70) 0.002 0.007 0.074 0.446 0.889 0.988 
GG vs. CC 2.21 (1.54–3.17) 0.016 0.047 0.351 0.845 0.982 0.998 
G vs. C 1.47 (1.24–1.73) 0.002 0.006 0.060 0.393 0.866 0.985 

Trial sequential analysis (TSA)

TSA was performed to reduce the random errors and increase the robustness of the conclusions. The TSA of the heterozygote (CG vs. CC) and homozygote models (GG vs. CC) for rs10830963 C>G among Asians and Caucasians (Figures 912), indicated that the cumulative Z-curve crossed both the conventional cut-off value and the TSA boundaries, and confirmed that the SNP was mainly associated with susceptibility to GDM. Even if the cumulative amount information didn’t reach the required information size (RIS), the results of TSA suggested that no further study evidence was needed to verify the conclusion.

TSA for rs10830963 under the heterozygote model among Asians (CG vs. CC)

Figure 9
TSA for rs10830963 under the heterozygote model among Asians (CG vs. CC)
Figure 9
TSA for rs10830963 under the heterozygote model among Asians (CG vs. CC)

TSA for rs10830963 under the heterozygote model among Caucasians (CG vs. CC)

Figure 10
TSA for rs10830963 under the heterozygote model among Caucasians (CG vs. CC)
Figure 10
TSA for rs10830963 under the heterozygote model among Caucasians (CG vs. CC)

TSA for rs10830963 under the homozygote model among Caucasians (GG vs.CC)

Figure 11
TSA for rs10830963 under the homozygote model among Caucasians (GG vs.CC)
Figure 11
TSA for rs10830963 under the homozygote model among Caucasians (GG vs.CC)

TSA for rs10830963 under the homozygote model among Caucasians (GG vs. CC)

Figure 12
TSA for rs10830963 under the homozygote model among Caucasians (GG vs. CC)
Figure 12
TSA for rs10830963 under the homozygote model among Caucasians (GG vs. CC)

Discussion

Gestational diabetes mellitus (GDM) is considered as an early form of diabetes, which might increase the risk of adverse pregnancy outcomes and substantial long-term adverse health among mothers and their offspring. Pregnant women with GDM are at a relatively high risk of developing Type 2 diabetes in the future [46]. Moreover, GDM also can be increase the hypertension, obesity, dyslipidemia and cardiovascular disease [4–6,47]. Therefore, it is urgent to clarify the etiology and pathogenesis of GDM.

A number of candidate gene-based association studies and genome-wide association studies (GWAS) have shown that some GDM relation genetic variants could provide insight into pathogenetic mechanisms underlying the disease [48–50]. Functional studies have shown that these diabetogenic genes took part in the process of developing GDM by impairing β-cell function, insulin resistance or abnormal utilization of glucose etc [48,49]. Our present study showed that the frequency of MTNR1B rs10830963 G allele was relatively higher in GDM cases than that in controls, suggesting a significant association with an increased GDM risk.

GDM could occur when the pancreatic islet β cells were impaired by an increased insulin resistance during pregnancy [51]. MTNR1B is a G-protein coupled 7-transmembrane receptor and could influence pancreatic β-cell function and fasting plasma glucose (FPG) level [14,15]. Studies have shown that genetic variations of MTNR1B gene are associated with insulin secretion and impaired β-cell function. A study of genetics and quantitative traits analysis of Palmer et al. revealed a significant association between MTNR1B and the glucose disposition index and acute insulin response among T2DM [52]. Meanwhile, experimental studies have confirmed that comparing with the wild-type C allele of rs10830963, the variant G allele caused an increased expression of MTNR1B and to be related to the risk of T2DM or GDM [52–54]. Our study showed that MTNR1B rs10830963 C>G might possibly increase the incidence of GDM. This may due to the observation that the increased expression of MTNR1B on β cells diminished intracellular cyclic cAMP level, thereby inhibited the glucose-stimulated insulin secretion.

A previous meta-analysis by Zhang et al. observed a statistically significant association between the MTNR1B variant CG/GG genotype and GDM risk (OR = 1.24, 95% CI = 1.14−1.35) [55]. However, this meta-analysis only included 5 studies (4 studies for Asians and 1 for Caucasians) with 2122 GDM patients and 2664 control subjects, and was unable to do the subgroup analysis by ethnicity to reveal the effect of MTNR1B rs10830963 C>G polymorphism on GDM risk in different ethnic populations.We summarized the evidence to date with 5033 GDM patients and 5614 controls by a meta-analysis with TSA, and found that compared with the wild CC genotype, the variant CG and GG genotype were significantly associated with an increased risk of GDM, respectively. Furthermore, the subgroup analysis by ethnicity revealed that rs10830963 C>G polymorphism was significantly associated with the risk of GDM both in Asians and in Caucasians. Obvious heterogeneity across studies was observed in data processing in this meta-analysis. We then used the meta-regression analysis to explore the potential source of heterogeneity across the studies by ethnicity, diagnostic criteria and assessed literature quality, and found that they didn’t contribute to the main observed heterogeneity.

FPRP analysis is an efective approach to verify the noteworthiness of signifcant association fndings. In the present study, we performed a relatively stringent FPRP threshold of 0.2. We found that the FPRP values of the observed signifcant associations between MTNR1B rs10830963 C>G and GDM risk was much lower than the preset threshold. It suggests that the positive findings both in overall analysis and racial-related subgroup analysis of heterozygous and homozygous models are probability authentic and reliable. Hence, we believe the association of rs10830963 C>G and GDM risk are credible to some extent. Further, a TSA indicated that the cumulative Z-curve crossed the conventional cut-off value and the TSA boundaries both in Aisan and Caucasian subgroup analyzes and confirmed that the rs10830963 SNP was mainly associated with susceptibility to GDM. The results of TSA suggesting that no additional researches are required to further evaluate the findings.

The current meta-analysis more comprehensively makes the relationship clear between MTNR1B rs10830963 and GDM risk. However, some limitations should be point out in the current meta-analysis. First, although the Begg’s and Egger’s tests did not detect any significantly statistical evidence of publication bias, selection bias could exist because only published case–control studies were included. Second, the small sample size might limit the statistical power of the study. Third, the findings of present study were based on the unadjusted results. Due to lack of individual-level data prevented us from making further analysis to identify any genotype–environment interaction between rs10830963 C>G and metabolic traits, such as FPG, pancreatic β-cell function, acute insulin response or indices for insulin sensitivity.

Conclusion

In summary, the current meta-analysis with TSA indicates that the variant G allele of MTNR1B rs10830963 C>G polymorphism significantly increases the risk of GDM. Further functional experimental studies are warranted to help explore and clarify the potential mechanism.

Author Contribution

X.Y., Y.W., and B.H. wrote the manuscripts; L.Q.,Q.W., N.L. and J.M. did the document retrieval, data collection and processing; X.Y., B.H. and H.Y. conceived and designed the study; Y.W., B.H. and J.M. created all tables and figures. All authors read and approved the final manuscript.

Funding

This study was supported by the National Natural Science Foundation of China [grant number 81760614]; the Key R&D Program [grant number AB18050023]; the Scientific Research Project [grant numbers 14124004-1-15 and KY2015YB223]; and College Students’ Innovation and Entrepreneurship Project [grant numbers 201710601103 and 201810601031] of Guangxi, China.

Competing Interests

The authors declare that there are no competing interests associated with the manuscript.

Abbreviations

     
  • CI

    confidence interval

  •  
  • FPG

    fasting plasma glucose

  •  
  • FPRP

    false positive report probability

  •  
  • GDM

    gestational diabetes mellitus

  •  
  • HWE

    Hardy–Weinberg equilibrium

  •  
  • MTNR1B

    melatonin receptor 1B

  •  
  • OR

    odds ratio

  •  
  • RIS

    required information size

  •  
  • TSA

    trial sequential analysis

References

References
1.
American Diabetes Association
. (
2004
)
Gestational diabetes mellitus
.
Diabetes Care
27
,
S88
S90
[PubMed]
2.
Ben-Haroush
A.
,
Yogev
Y.
and
Hod
M.
(
2004
)
Epidemiology of gestational diabetes mellitus and its association with Type 2 diabetes
.
Diabetes Med.
21
,
103
113
[PubMed]
3.
Galtier
F.
(
2010
)
Definition, epidemiology, risk factors
.
Diabetes Metab.
36
,
628
651
[PubMed]
4.
Bellamy
L.
,
Casas
J.P.
and
Hingorani
A.D.
(
2009
)
Type 2 diabetes mellitus after gestational diabetes: a systematic review and meta-analysis
.
Lancet
373
,
1773
1779
[PubMed]
5.
Nicotra
F.
,
Molinari
C.
and
Dozio
N.
(
2015
)
Screening for gestational diabetes in the Lombardy region: A population-based study
.
Diabetes Metab.
41
,
319
325
[PubMed]
6.
Martin
K.E.
,
Grivell
R.M.
and
Yelland
L.N.
(
2015
)
The influence of maternal BMI and gestational diabetes on pregnancy outcome
.
Diabetes Res. Clin. Pract.
108
,
508
513
[PubMed]
7.
Damm
P.
,
Houshmand-Oeregaard
A.
and
Kelstrup
L.
(
2016
)
Gestational diabetes mellitus and long-term consequences for mother and offspring: a view from Denmark
.
Diabetologia
59
,
1396
1399
[PubMed]
8.
Leng
J.
,
Shao
P.
and
Zhang
C.
(
2015
)
Prevalence of gestational diabetes mellitus and its risk factors in Chinese pregnant women: a prospective population-based study in Tianjin, China
.
PLoS One
10
,
e0121029
[PubMed]
9.
Chu
S.Y.
,
Callaghan
W.M.
and
Kim
S.Y.
(
2007
)
Maternal obesity and risk of gestational diabetes mellitus
.
Diabetes Care
30
,
2070
2076
[PubMed]
10.
Rhee
S.Y.
,
Kim
J.Y.
and
Woo
J.T.
(
2010
)
Familial clustering of type 2 diabetes in Korean women with gestational diabetes mellitus
.
Korean. J. Intern. Med.
25
,
269
272
[PubMed]
11.
Zhang
C.
,
Bao
W.
and
Rong
Y.
(
2013
)
Genetic variants and the risk of gestational diabetes mellitus: a systematic review
.
Hum. Reprod. Update
19
,
376
390
[PubMed]
12.
Anzaku
A.S.
and
Musa
J.
(
2013
)
Prevalence and associated risk factors for gestational diabetes in Jos, North-central, Nigeria
.
Arch. Gynecol. Obstet.
287
,
859
863
[PubMed]
13.
Buchanan
T.A.
(
2001
)
Pancreatic B-cell defects in gestational diabetes: implications for the pathogenesis and prevention of type 2 diabetes
.
J. Clin. Endocrinol. Metab.
86
,
989
993
[PubMed]
14.
Pandi-Perumal
S.R.
,
Trakht
I.
and
Srinivasan
V.
(
2008
)
Physiological effects of melatonin: role of melatonin receptors and signal transduction pathways
.
Prog. Neurobiol.
85
,
335
353
[PubMed]
15.
Ramracheya
R.D.
,
Muller
D.S.
and
Squires
P.E.
(
2008
)
Function and expression of melatonin receptors on human pancreatic islets
.
J. Pineal Res.
44
,
273
279
[PubMed]
16.
Li
Y.
,
Wu
H.
and
Liu
N.
(
2018
)
Melatonin exerts an inhibitory effect on insulin gene transcription via MTNR1B and the downstream Raf1/ERK signaling pathway
.
Int. J. Mol. Med.
41
,
955
961
[PubMed]
17.
Van Cauter
E.
(
1998
)
Putative roles of melatonin in glucose regulation
.
Therapie
53
,
467
472
[PubMed]
18.
Ren
J.
,
Xiang
A.H.
and
Trigo
E.
(
2014
)
Genetic variation in MTNR1B is associated with gestational diabetes mellitus and contributes only to the absolute level of beta cell compensation in Mexican Americans
.
Diabetologia
57
,
1391
1399
[PubMed]
19.
Karamitri
A.
and
Jockers
R.
(
2019
)
Melatonin in type 2 diabetes mellitus and obesity
.
Nat. Rev. Endocrinol.
15
,
105
125
[PubMed]
20.
Lyssenko
V.
,
Nagorny
C.L.
and
Erdos
M.R.
(
2009
)
Common variant in MTNR1B associated with increased risk of type 2 diabetes and impaired early insulin secretion
.
Nat. Genet.
41
,
82
88
[PubMed]
21.
Li
C.
,
Zhou
Y.
and
Qiao
B.
(
2019
)
Association Between a Melatonin Receptor 1B Genetic Polymorphism and Its Protein Expression in Gestational Diabetes Mellitus
.
Reprod. Sci.
26
,
1382
1388
1933719118765983
22.
Deng
Z
,
Shu
Q
,
Chen
Y
et al.
(
2011
)
Association of genetic variant rs10830963 of melatonin receptor 1B gene in women with gestational diabetes mellitus
.
Chin. J. Perinat. Med.
14
,
666
669
23.
Kim
J.Y.
,
Cheong
H.S.
and
Park
B.L.
(
2011
)
Melatonin receptor 1 B polymorphisms associated with the risk of gestational diabetes mellitus
.
BMC Med. Genet.
12
,
82
[PubMed]
24.
Wang
Y.
,
Nie
M.
and
Li
W.
(
2011
)
Association of six single nucleotide polymorphisms with gestational diabetes mellitus in a Chinese population
.
PLoS One
6
,
e26953
[PubMed]
25.
Vlassi
M.
,
Gazouli
M.
and
Paltoglou
G.
(
2012
)
The rs10830963 variant of melatonin receptor MTNR1B is associated with increased risk for gestational diabetes mellitus in a Greek population
.
Hormones (Athens)
11
,
70
76
[PubMed]
26.
Huopio
H.
,
Cederberg
H.
and
Vangipurapu
J.
(
2013
)
Association of risk variants for type 2 diabetes and hyperglycemia with gestational diabetes
.
Eur. J. Endocrinol.
169
,
291
297
[PubMed]
27.
Li
C.
,
Qiao
B.
and
Zhan
Y.
(
2013
)
Association between genetic variations in MTNR1A and MTNR1B genes and gestational diabetes mellitus in Han Chinese women
.
Gynecol. Obstet. Invest.
76
,
221
227
[PubMed]
28.
Su
Y.
and
Zhang
Y.
(
2015
)
Identification of biological processes and genes for gestational diabetes mellitus
.
Arch. Gynecol. Obstet.
292
,
635
640
[PubMed]
29.
Vejrazkova
D.
,
Lukasova
P.
and
Vankova
M.
(
2014
)
MTNR1B Genetic Variability Is Associated with Gestational Diabetes in Czech Women
.
Int. J. Endocrinol.
2014
,
508923
[PubMed]
30.
Wang
X.
,
Sun
S.
and
Bao
L.
(
2014
)
Association between gene polymorphism of Melatonin receptor 1B and gestational diabetes mellitus
.
Chin. J. Diabetes
5
,
398
400
31.
Junior
J.P.
,
Frigeri
H.R.
and
Dos Santos-Weiss
I.C.
(
2015
)
The MTNR1B gene polymorphism rs10830963 is associated with gestational diabetes in a Brazilian population
.
Gene
568
,
114
115
[PubMed]
32.
Liu
Q.
,
Huang
Z.
and
Li
H.
(
2016
)
Relationship between melatonin receptor 1B (rs10830963 and rs1387153) with gestational diabetes mellitus: a case-control study and meta-analysis
.
Arch. Gynecol. Obstet.
294
,
55
61
[PubMed]
33.
Tarnowski
M.
,
Malinowski
D.
and
Pawlak
K.
(
2017
)
GCK, GCKR, FADS1, DGKB/TMEM195 and CDKAL1 Gene Polymorphisms in Women with Gestational Diabetes
.
Can. J. Diabetes
41
,
372
379
[PubMed]
34.
Popova
P.V.
,
Klyushina
A.A.
and
Vasilyeva
L.B.
(
2017
)
Effect of gene-lifestyle interaction on gestational diabetes risk
.
Oncotarget
8
,
112024
112035
[PubMed]
35.
Rosta
K.
,
Al-Aissa
Z.
and
Hadarits
O.
(
2017
)
Association Study with 77 SNPs Confirms the Robust Role for the rs10830963/G of MTNR1B Variant and Identifies Two Novel Associations in Gestational Diabetes Mellitus Development
.
PLoS One
12
,
e0169781
[PubMed]
36.
Stang
A.
(
2010
)
Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses
.
Eur. J. Epidemiol.
25
,
603
605
[PubMed]
37.
Lau
J.
,
Ioannidis
J.P.
and
Schmid
C.H.
(
1997
)
Quantitative synthesis in systematic reviews
.
Ann. Intern. Med.
127
,
820
826
[PubMed]
38.
DerSimonian
R.
and
Laird
N.
(
1986
)
Meta-analysis in clinical trials
.
Control. Clin. Trials
7
,
177
188
[PubMed]
39.
Begg
C.B.
and
Mazumdar
M.
(
1994
)
Operating characteristics of a rank correlation test for publication bias
.
Biometrics
50
,
1088
1101
[PubMed]
40.
Egger
M.
,
Davey Smith
G.
and
Schneider
M.
(
1997
)
Bias in meta-analysis detected by a simple, graphical test
.
BMJ
315
,
629
634
[PubMed]
41.
Wacholder
S.
,
Chanock
S.
and
Garcia-Closas
M.
(
2004
)
Assessing the probability that a positive report is false: an approach for molecular epidemiology studies
.
J. Natl. Cancer Inst.
96
,
434
442
[PubMed]
42.
He
J.
,
Wang
M.Y.
and
Qiu
L.X.
(
2013
)
Genetic variations of mTORC1 genes and risk of gastric cancer in an Eastern Chinese population
.
Mol. Carcinog.
52
,
E70
E79
[PubMed]
43.
Brok
J.
,
Thorlund
K.
and
Gluud
C.
(
2008
)
Trial sequential analysis reveals insufficient information size and potentially false positive results in many meta-analyses
.
J. Clin. Epidemiol.
61
,
763
769
[PubMed]
44.
Wetterslev
J.
,
Jakobsen
J.C.
and
Gluud
C.
(
2017
)
Trial Sequential Analysis in systematic reviews with meta-analysis
.
BMC Med. Res. Methodol.
17
,
39
[PubMed]
45.
Thorlund
K.
,
Imberger
G.
and
Wetterslev
J.
(
2010
)
Comments on ‘Sequential meta-analysis: an efficient decision-making tool’ by I van der Tweel and C Bollen
.
Clin. Trials
7
,
752
753
,
author reply 754
[PubMed]
46.
Silverman
B.L.
,
Rizzo
T.A.
and
Cho
N.H.
(
1998
)
Long-term effects of the intrauterine environment. The Northwestern University Diabetes in Pregnancy Center
.
Diabetes Care
21
,
B142
149
[PubMed]
47.
Whyte
K.
,
Kelly
H.
and
O’Dwyer
V.
(
2013
)
Offspring birth weight and maternal fasting lipids in women screened for gestational diabetes mellitus (GDM)
.
Eur. J. Obstet. Gynecol. Reprod. Biol.
170
,
67
70
[PubMed]
48.
Kwak
S.H.
,
Kim
S.H.
and
Cho
Y.M.
(
2012
)
A genome-wide association study of gestational diabetes mellitus in Korean women
.
Diabetes
61
,
531
541
[PubMed]
49.
Ekelund
M.
,
Shaat
N.
and
Almgren
P.
(
2012
)
Genetic prediction of postpartum diabetes in women with gestational diabetes mellitus
.
Diabetes Res. Clin. Pract.
97
,
394
398
[PubMed]
50.
Mao
H.
,
Li
Q.
and
Gao
S.
(
2012
)
Meta-analysis of the relationship between common type 2 diabetes risk gene variants with gestational diabetes mellitus
.
PLoS One
7
,
e45882
[PubMed]
51.
Lambrinoudaki
I.
,
Vlachou
S.A.
and
Creatsas
G.
(
2010
)
Genetics in gestational diabetes mellitus: association with incidence, severity, pregnancy outcome and response to treatment
.
Curr. Diabetes Rev.
6
,
393
399
[PubMed]
52.
Palmer
N.D.
,
Goodarzi
M.O.
and
Langefeld
C.D.
(
2015
)
Genetic Variants Associated With Quantitative Glucose Homeostasis Traits Translate to Type 2 Diabetes in Mexican Americans: The GUARDIAN (Genetics Underlying Diabetes in Hispanics) Consortium
.
Diabetes
64
,
1853
1866
[PubMed]
53.
Salman
M.
,
Dasgupta
S.
and
Cholendra
A.
(
2015
)
MTNR1B gene polymorphisms and susceptibility to Type 2 Diabetes: A pilot study in South Indians
.
Gene
566
,
189
193
[PubMed]
54.
Semiz
S.
,
Dujic
T.
and
Velija-Asimi
Z.
(
2014
)
Effects of melatonin receptor 1B gene variation on glucose control in population from Bosnia and Herzegovina
.
Exp. Clin. Endocrinol. Diabetes
122
,
350
355
[PubMed]
55.
Zhang
Y.
,
Sun
C.M.
and
Hu
X.Q.
(
2014
)
Relationship between melatonin receptor 1B and insulin receptor substrate 1 polymorphisms with gestational diabetes mellitus: a systematic review and meta-analysis
.
Sci. Rep.
4
,
6113
[PubMed]

Author notes

*

These authors contributed equally to this work.

This is an open access article published by Portland Press Limited on behalf of the Biochemical Society and distributed under the Creative Commons Attribution License 4.0 (CC BY).