The PubMed, Embase and Cochrane Library were comprehensively searched. Data was extracted and effects were estimated from an odds ratio (OR) with 95% confidential intervals (95%CIs). Subgroup and sensitivity analyses identified sources of heterogeneity.
CTLA4 + 49A/G and − 318C/T polymorphisms were significantly correlated with pathogenesis of pediatric T1D (+ 49A/G: GG vs. AA + AG, OR:1.55; 95%CIs:1.37–1.75; P < 0.01; AA vs. GG, OR:0.52; 95%CIs:0.41–0.67; P < 0.01; AA vs. AG, OR:0.71; 95%CIs:0.59–0.86; P < 0.01; A vs. G, OR:0.70; 95%CIs:0.60–0.82; P < 0.01; − 318C/T: CC vs. CT + TT: OR: 5.77; 95%CIs: 2.94–11.32; P < 0.01). In subgroup analysis of ethnicity in + 49A/G, we also found similar results in Caucasian and Asian populations (Caucasians, OR: 0.83; 95%CIs: 0.75–0.92; P < 0.01; Asians, OR: 0.63; 95%CIs: 0.55–0.73; P < 0.01). No association was detected in subgroup analysis based on the T1D patients with/without autoimmune thyroid disease (AITD).
Our meta-analysis suggested that CTLA4 + 49A/G and − 318C/T genetic polymorphisms were significantly associated with the onset of T1D in children. A large-scale, multi-center case-control study is needed to confirm our findings.