Polymorphism in the 11q24.1 genomic region is associated with myopia: A comprehensive genetic study in Chinese and Japanese populations

Search strategy

To assess the total evidence of association between the CTNND2 gene and myopia disease, and between 11q24.1 genomic region and myopia, we performed the present comprehensive meta-analysis of published studies. A total of 11 studies focusing either on the association between the CTNND2 gene and myopia or between the 11q24.1 genomic region and myopia were identified according to our inclusion criteria; these publications comprised 6,954 cases and 9,346 controls [11,15-18]. The main characteristics of these studies are given in Table 1. We considered all studies that examined the association of the rs577948, rs11218544, rs12716080, and rs6885224 polymorphisms with myopia. Sources included MEDLINE and EMBASE (search last updated in Aug. 2012). The search strategy was based on combinations of the terms “CTNND2,” “T-cell delta-catenin,” “catenin delta-2,” “11q24.1 genomic region,” and “myopia.” Reference lists in the retrieved articles were also screened.

Inclusion and exclusion criteria

All the studies included satisfied all the following criteria: (1) They were association studies either between the rs12716080 and rs6885224 polymorphisms in the CTNND2 gene and myopia disease or between the rs577948 and rs11218544 polymorphisms and myopia; (2) they used myopia-free people as controls; (3) they provided genotype or allele distributions in both the case and control groups; and (4) they were independent studies and the subject groups investigated did not overlap with each other. Authors were contacted where clarification was required.

Only studies published in English were included. Studies presenting nonoriginal data were excluded, such as reviews, editorials, opinion papers, or letters to the editor. Studies using nonhuman subjects or specimens were excluded. Studies with no extractable numerical data were excluded. Only those articles that had some measure of diagnostic performance were included. Any duplicates that came up in the preliminary search were excluded.

Data extraction

The following information was independently extracted from the identified studies by two participants in the meta-analysis: first author, journal, year of publication, study design, ethnicity of the study population, gender, clinical characteristics, genotyping method, the number of cases and controls or odds ratio (OR) and 95 percent confidence interval (95%CI), country in which the study was conducted, and confirmation of diagnosis. The results were compared and any disagreement was discussed and resolved by consensus.

Statistical analysis

We used a comprehensive meta-analysis to analyze the ORs using Carlin’s method [19]. To determine whether the results of the meta-analysis were unduly influenced by any one study, we recomputed the meta-analysis statistics after deleting each study one at a time. We assessed publication bias using Egger et al.’s approach [20]. This method is based on the fact that the precision of the OR increases with larger study groups. It regresses the standard normal deviation of the OR (the OR divided by its standard error) against the precision of the OR (the inverse of its standard error). In the absence of bias, Egger et al. [20] showed that the regression of the standard normal deviate on the precision of the OR should run through the origin (i.e., small study groups with low precision have large standard errors and therefore small standard normal deviates; large study groups have higher precision, smaller standard errors, and large standard normal deviates). The publication bias statistic of Egger et al. [20] is the intercept of the regression, which will be significantly greater than zero in the presence of publication bias.

The random effects model using the DerSimonian and Laird method was employed, and the estimate of heterogeneity was determined using the Mantel–Haenszel model [21]. Heterogeneity among studies was tested using the Q and I² statistics. The effect size was represented by an OR with 95%CI. Sensitivity analysis was conducted by removing each study and analyzing the others to ensure that no single study was totally responsible for overall results. Every time, each study was removed in turn from the total, and the remainder was reanalyzed. This procedure was used to ensure that no individual study was entirely responsible for a finding. Between-study heterogeneity was tested using Cochran’s Q statistic, which is considered significant at p<0.10. The extent of inconsistency across studies was quantified with the I² statistic. The I² ranges between 0 and 100%. When there was very large or large (>50%) between-study heterogeneity, we used a simulation algorithm to evaluate how many studies had to be removed for the I² to reach <25%.

The significance level was set at 0.05, and all p values were two-tailed. The comprehensive meta-analysis was performed using Comprehensive Meta Analysis software (Version 2.2.046, BIOSTAT, Englewood, NJ) [22-24].

Eligible studies

The combined search yielded seven references. These were then filtered to ensure conformity with the inclusion criteria. One reference was discarded because it was a family-based study. Ultimately, six references met our criteria (Table 1, Figure 1).

Synthesis of quantitative data

The eligible studies for analysis included a total of 6,954 cases with myopia and 9,346 controls (Table 1). For the 11q24.1 genomic region, the rs11218544 polymorphism showed significantly association with myopia [OR and 95% CI: 1.167 (1.032–1.319), p=0.013], and rs577948 showed no association with the disease [OR and 95% CI: 0.988 (0.727–1.342), p=0.936]. For the CTNND2 gene, neither the rs6885224 nor rs12716080 polymorphism was significantly associated with myopia {rs6885224: [OR and 95% CI: 1.051 (0.795–1.391), p=0.725], rs12716080: [OR and 95% CI: 1.173 (0.990–1.390), p=0.065; Figure 2A-D).

Publication bias and heterogeneity

The sensitivity analysis showed that when any one of the studies was removed, the heterogeneity of the population was not changed deeply; namely, when one study was removed, the result was still significant (Figure 3A-D). This indicated that no heterogeneity existed in the population. There was no evidence to suggest that the magnitude of the overall OR estimates changed in the same direction over time. Moreover, the Egger funnel plots of publication bias analysis for four polymorphisms were shown (Figure 4A-D).