RESUMEN
OBJECTIVE: To investigate preconceptual thyroid peroxidase antibody (TPO-ab) positivity and/or thyroid stimulating hormone (TSH) levels in the upper range of normal as risk factors for recurrent unexplained first-trimester miscarriage. DESIGN: A post-hoc study of a randomized trial, in which acetylsalicylic acid did not affect the risk of a new miscarriage. PATIENTS: Women (n = 483) with at least three unexplained recurrent first-trimester miscarriages investigated at a Swedish secondary referral center. MEASUREMENTS: The levels of TPO-ab and TSH were determined before pregnancy. The occurrence of a new first-trimester miscarriage was analyzed by logistic regression with adjustments when applicable, for age, number of previous miscarriages, obesity and the investigated covariates levels of TPO-ab and TSH. RESULTS: Including all first trimester miscarriages, odds ratio (OR) according to presence of TPO-ab was 1.60 (95% confidence interval [CI]; 0.99-2.57), after adjustment 1.54 (95% CI; 0.94-2.53). Very early (biochemical) pregnancy losses occurred more often in women with than without preconceptual TPO-ab (6.8% vs. 2.0%), OR 3.51 (95% CI; 1.15-10.71), after adjustment 2.91 (95% CI; 0.91-9.29). There was no association between TSH in the upper range of normal and a new miscarriage, adjusted OR 0.76 (95% CI; 0.32-1.83). A prediction model for a new miscarriage included number of previous miscarriages, woman's age and presence of TPO-ab. CONCLUSION: In women with at least three recurrent unexplained pregnancy losses, the presence of TPO-ab may contribute to an increased risk of a first-trimester miscarriage, possibly more pronounced in very early pregnancy. TSH levels 2.5-4.0 mU/L do not seem to increase the miscarriage risk.