Risk of beta-cell autoimmunity presence for progression to type 1 diabetes: A systematic review and meta-analysis.

BACKGROUND

Islet autoantibodies have been applied for diagnosis of type 1 diabetes mellitus (T1DM) at an asymptomatic stage in individuals with high-risk genotypes. Evidence is insufficient to support a broad application of islet autoantibody screening for T1DM in clinical practice. The aim of this study was to assess the evidence of an association between islet autoantibodies and the development of T1DM in a pooled population of both genetically at-risk individuals and general people without definite genetic background.

METHODS

A comprehensive literature search was performed of Pubmed, Web of knowledge and Cochrane library. Prospective cohort studies evaluating the role of islet autoantibodies in prediction of T1DM progression were included. Risk ratios (RRs) were calculated and pooled to arrive at summary estimate. χ and I-values were calculated as measures of heterogeneity and subgroup analyses were performed to explore sources of heterogeneity.

RESULTS

Twenty-one studies matched the inclusion criteria. A total of 71,482 nondiabetic participants who were genetically at-risk individuals or from the general population were included, and 926 cases of T1DM were reported during a median follow-up of 7 years. Compared with people free of islet autoantibody, those positive for any type or number of islet autoantibody showed a significantly increased risk of developing T1DM (RR 150.42 [95% CI 87.34, 259.04]). Moreover, the risk for people with multiple islet autoantibodies was 8.59-fold higher than the risk for those with single islet autoantibody, although a moderate heterogeneity existed between studies. The subgroup analysis further revealed that RRs of multiple islet autoantibodies in at-risk population and general population were 7.17 and 13.72, respectively.

CONCLUSION

This study established the association between the seroconversion of islet autoantibodies and T1DM progression in nondiabetic people with or without definite genetic susceptibility, providing further evidence for an extensive application in routine clinical practice to identify individuals at risk of T1DM.