No Concurrent Association Found Between Maternal Thyroid Hormone Concentrations (TSH, FT4, FT3) and Antepartum Depression in Late Pregnancy: A Meta-Analysis Highlighting the Need for Categorical Risk Assessment

Background: The relationship between maternal thyroid function and psychiatric morbidity remains inconclusive, particularly regarding the association with antepartum depression (APD). This meta-analysis aimed to precisely quantify the association between the three primary maternal thyroid hormone concentrations—thyroid-stimulating hormone (TSH), free thyroxine (FT4), and free triiodothyronine (FT3)—measured in late pregnancy and in the presence of APD. Methods: We conducted a systematic review and meta-analysis of observational studies identified through comprehensive database searches (PubMed, Web of Science, Scopus). Four exploratory studies were ultimately included, enrolling a total of 689 participants. We used random-effects models to pool the mean difference (MD) in hormone concentrations between depressed and non-depressed cohorts. Subgroup analyses were performed based on the study population (general versus hypothyroid), and publication bias was assessed using Begg’s and Egger’s tests. Results: None of the pooled hormone concentrations demonstrated a statistically significant association with APD. The overall MDs were non-significant for TSH (MD = −0.07, 95% CI: [−0.32, 0.18], p = 0.59), FT4 (MD = −0.11, 95% CI: [−1.14, 0.92], p = 0.83), and FT3 (MD = 0.53, 95% CI: [−0.20, 1.25], p = 0.15). Substantial and significant heterogeneity was detected across all models (I2 ranging from 70% to 94%). This heterogeneity was largely driven by conflicting directional findings (some studies linking APD to hypothyroid trends, others to hyperthyroid trends), masking a potential non-linear or categorical effect. Statistical tests found no significant evidence of publication bias for TSH (p = 0.33), FT4 (p = 0.12), or FT3 (p = 0.33). Conclusions: The absolute mean concentrations of TSH, FT4, and FT3 in late pregnancy are not robust concurrent biomarkers for antepartum depressive symptoms. The high heterogeneity suggests that APD may be associated with categorical dysfunction (i.e., TSH levels at the extreme high or low ends of the reference range) rather than linear changes in hormone concentration. Future research should prioritize investigating categorical risks, the influence of thyroid autoimmunity, and employing gold-standard diagnostic interviews to better delineate the complex endocrinological risk factors for APD. Due to the limited number of studies, these results should be considered hypothesis-generating rather than confirmatory. PROSPERO registration: CRD420251233154.