Effect of anti-thyroid antibodies on recurrent miscarriage: A meta-analysis.

Setting: Previous studies addressed the association between anti-thyroid antibodies and recurrent miscarriage (RM), however, the role of anti-thyroid antibodies in RM patients is debatable.

Objectives: Therefore, we conducted this meta-analysis and the aim of this current study was to assess whether anti-thyroid peroxidase (anti-TPO) and/or anti-thyroglobulin (anti-TG) antibody positivity was associated with RM.

Design: A meta-analysis was conducted.

Participants: Recurrent miscarriage patients.

Methods: STATA 12.0 software were applied to compute odds ratios (ORs)/relative risks (RRs) and 95% CIs regarding association between anti-TPO and anti-TG antibodies and the prevalence of RM.

Results: N = 28 studies (8875 participants) explored effect of anti-thyroid antibodies on RM. Analysis of the 28 studies revealed significant association between anti-TPO, anti-TG antibodies and the prevalence of RM with a random effects model (OR/RR = 2.02; 95% CI: 1.63-2.51, p < 0.001; I = 44.3%, p value for Q test = 0.004). Analysis of the 20 studies revealed significant association between anti-TPO antibodies and the prevalence of RM with a random effects model (OR/RR = 1.59; 95% CI: 1.25-2.03, p < 0.001; I = 43.1%, p value for Q test = 0.022). Analysis of the 14 studies revealed significant association between anti-TG antibodies and the prevalence of RM with a random effects model (OR/RR = 2.25; 95% CI: 1.56-3.23, p < 0.001; I = 49.2%, p value for Q test = 0.019).

Conclusions: Based on the currently available analysis, our findings suggest that women with anti-TPO and/or anti-TG antibodies have a higher risk of RM than that in negative antibody women. Further investigation is needed to better clarify the exact role of the anti-thyroid antibodies in RM and whether treatment is of benefit.

Limitations: First, differences from various detection methods and reagents used in different studies may affect the diagnostic interpretation of anti-thyroid antibodies, which might influence the accuracy of this meta-analysis. Second, positive anti-thyroid antibodies seem likely to be part of a more general disorder of maternal immune system, due to restrictions of funding and condition, a complete autoantibody screening investigation is hardly to conduct in all participants, and this could be a possible limitation of all included studies. Third, there is no mention of thyroxine therapy on RM, making the meta-analysis even more limited.