Parkinson’s disease (PD, OMIM: 168600) is the second most common neurodegenerative disorder that increases the social and economic burden on societies. PD affects ~2% of the population aged ≥65 years (1,2). The clinical features of PD are resting tremor, muscular rigidity, bradykinesia and postural instability (3). PD can cause pain (4), depression (5,6), visual hallucinations (7), dementia (8) and other non-motor symptoms (9–12).

Environmental and genetic factors may be involved in the pathogenesis of PD. The environmental hypothesis was dominant for much of the 20th century (3). Environmental factors such as encephalitis (13), oxidative stress (14), smoking and coffee (15) and environmental toxins (16) were found to be significantly associated with PD. Recent genetic studies have identified a few genetic markers of PD (17–19), however, the etiology of PD remains unclear.

Findings of previous studies suggested that the degeneration of dopamine neurons in the substantia nigra contributes to the pathogenesis of PD (20,21). Dopamine is a major modulatory neurotransmitter in the central nervous system (CNS) and thus affects neuroendocrine secretion (20), which was shown to be associated with smoking, a well-known risk factor of PD (22–24). There are five subtypes of G-protein-coupled dopamine receptors (DRD1–5) comprising D-1-like (DRD1 and DRD5) and D-2-like (DRD2–4). Among these subtypes, DRD2 (25–36) located on chromosome 11q23 and DRD3 (33,35–40) located on chromosome 3q13.3 are the most studied in the association tests of PD.

Previous case-control studies showed inconsistent results between PD and the two dopamine receptor genes (DRD2 and DRD3). These discrepancies may be due to different ethnic background, or sample size difference (41), or the uncorrected physiological status among various association studies (42). Meta-analysis is a robust method for application to enhance statistical power and to draw a more convincing conclusion through the pooling of data from each of the involved studies (43). The aim of this meta-analysis was to evaluate the contribution of DRD2 and DRD3 gene polymorphisms to PD and to determine the causes of the inconsistent results among various case-control association studies.

As shown in Fig. 1, 33 relevant studies of DRD2 and 19 studies of DRD3 were included from PubMed. In addition, we also retrieved two articles on DRD2 gene from the Chinese Wangfang database and one article on DRD3 gene from the China National Knowledge Internet. Following removal of three duplicates, 16 articles for the current meta-analysis (Table I) (25–40). Moreover, nine stages (36) (Table I) were excluded with significant deviation from HWE (P<0.05). Altogether there were 4,279 PD patients and 5,661 controls in the current meta-analyses of 9 polymorphisms (Tables II and III).

As shown in Tables II and III, no significant associations were observed in the meta-analyses of five DRD2 polymorphisms. A further subgroup study by ethnicity showed a borderline association between DRD2 rs1800497 polymorphism and PD in Europeans (P=0.05, OR=1.13, 95% CI: 1.00–1.27; Table III and Fig. 2). Significant results was found in the meta-analysis of DRD3 rs2134655 polymorphism (P=0.01, OR=1.17, 95% CI: 1.03–1.32; Table II and Fig. 2). No significant results were observed in other meta-analyses. There was no publication bias for any of the meta-analyses (Figs. 2 and 3).

As shown in Tables II and III, there was no statistically significant heterogeneity (I²<50%) in the meta-analyses of five DRD2 polymorphisms (rs1079597, rs6279, rs273482, rs1799732 and rs1076563) and two DRD3 polymorphisms (rs6280 and rs2134655). Significant heterogeneity was observed in DRD2 rs1800497 (I²=73%) and DRD2 rs6278 (I²=68%). The allele frequency of rs1800497-G in Asians (HapMap-HCB) is 0.512 which is much lower to the 0.805 in Europeans (HapMap-CEU). A further analysis of the rs1800497 polymorphism showed a moderate ethnic difference between Asians and Europeans (Fst=0.1). Since there existed different genotypic distribution of rs1800497 polymorphism between Asians and Europeans, we performed subgroup ethnic meta-analyses for the two populations. A significant heterogeneity was found in Asians (I²=76%, Table III), which may be explained by various types of allele frequency in Asians (HapMap-CHB=0.223, HapMap-JPT=0.349 and HapMap-CHD=0.310).

Power analyses showed that sufficient power (power>0.8) for the meta-analyses of five DRD2 polymorphisms (rs1800497, rs1079597, rs6279, rs273482 and rs1076563) and two DRD3 polymorphisms (rs6280 and rs2134655). The results also indicated that there was a lack of power for the meta-analyses of DRD2 rs6278 (power=0.637) and rs1799732 (power=0.313) polymorphisms.

In the current meta-analysis, we summarized the associations of DRD2 (n=7) and DRD3 (n=2) gene variants with PD from 16 studies (46 stages) among 4,279 cases and 5,661 controls. Our meta-analyses demonstrated a borderline significant association with PD for DRD2 rs1800497 in Europeans and a significant association with PD for DRD3 rs2134655. Previous studies have shown that the DRD2 rs1800497-T allele reduced DRD2 density in the postmortem brain (48) and decreased dopamine activity in the CNS (49), which may play an important role in the pathogenesis of PD. DRD3 rs2134655 was a C to T transition that showed significant associations with neurological diseases such as schizophrenia (50).

Our study was the only meta-analysis for DRD2 and DRD3 gene polymorphisms in PD with strict selection criteria. Although power analyses showed our meta-analyses have much larger power than that of each of the individual studies, we did not exclude the possibility of false-negative results in the meta-analyses for DRD2 rs6278 (power=0.637) and rs1799732 (power=0.313). We excluded 9 stages that failed to meet HWE in their control samples (36). We also performed subgroup analyses by ethnicity for DRD2 rs1800497, rs1079597 and DRD3 rs6280.

Our study has several limitations that need to be carefully considered. Firstly, although the power in most meta-analyses is sufficient, the number of involved stages in the Europeans was limited for polymorphisms (including rs2134655 of DRD3 gene; and rs1800497, rs1076563, rs1799732, rs6278, rs6279 and rs273482 of the DRD2 gene). Studies with a larger sample size are therefore needed. Secondly, factors such as gender, age and other different physiological status of PD patients may have influenced the result of our study and partially explain the discrepancies in the involved case-control studies. Future case-control studies may need to provide detailed information to better estimate the contribution of these factors to PD susceptibility. Thirdly, we have performed multiple association tests for the 9 polymorphisms, however, we did not provide correction of P-values that cause false-positive results in our findings. Additionally, genetic heterogeneity may exist for the polymorphisms of DRD2 and DRD3 genes. There are 134 polymorphisms in DRD2 and 88 polymorphisms in DRD3. Our meta-analyses only focused on 9 polymorphisms that might not fully represent the overall contribution of DRD2 and DRD3 gene polymorphisms. Analyses of other polymorphisms in DRD2 and DRD3 genes are required in future investigations to evaluate their contribution to PD.

In conclusion, our meta-analysis observed a borderline significant association of DRD2 rs1800497 polymorphism with PD in Europeans and a significant association of DRD3 rs2134655 with PD. Large-scale well-designed studies are required in future to evaluate the polymorphisms of DRD2 and DRD3 genes that might contribute to the risk of PD.