Subclinical hypothyroidism in secondary female infertility: a critical narrative review of evidence identified by a structured search

Abstract

Abstract Background Subclinical hypothyroidism (SCH) is defined by a thyroid-stimulating hormone (TSH) concentration above the applicable nonpregnant reference interval with a normal free thyroxine concentration. Its relevance to infertility remains contested, and secondary-infertility evidence is often conflated with mixed infertility populations, assisted reproductive technology (ART), autoimmunity, or post-conception outcomes. This review evaluates whether the published literature supports an independent association between SCH and secondary female infertility and whether screening or levothyroxine improves fertility outcomes in this subgroup. Main text This structured critical narrative review searched Europe PMC through 20 July 2026 using seven prespecified strands and one targeted strand added during revision. Because the search used one platform, screening was performed by one reviewer, and English-language reports were prioritized for detailed appraisal, the assembled study set may be incomplete and the conclusions most directly describe the predominantly English-language evidence appraised. Among 488 unique records and 83 full-text appraisals, five reports provided secondary-infertility subgroup findings; a sixth, population-based adjusted study that did not distinguish primary from secondary infertility was treated as high-quality indirect evidence and found no association between TSH and infertility or time to pregnancy. The subgroup prevalence signals were inconsistent, based on very few SCH cases, and almost entirely unadjusted. The search did not identify a randomized trial of levothyroxine restricted to secondary infertility. In broader preconception populations, pooled randomized evidence has not demonstrated benefit for live birth, pregnancy, or miscarriage; however, the confidence intervals remain compatible with clinically meaningful effects, and a secondary-infertility-specific benefit has neither been demonstrated nor excluded. Guidelines diverge without addressing this subgroup: the 2024 American Society for Reproductive Medicine guideline advises against universal screening and routine levothyroxine to improve fertility; the 2021 European Thyroid Association guideline is more intervention-oriented in assisted reproduction above approximately 4.0 mIU/L; and the 2026 American Thyroid Association guideline emphasizes local reference intervals, repeat confirmation, and individualized decisions. Population variation in TSH cautions against fixed universal cutoffs but does not by itself establish an optimal preconception threshold. Conclusion The subgroup-specific evidence located by this single-platform search is too limited and confounded to support a causal claim. A secondary-infertility-specific treatment benefit has not been demonstrated, but the available evidence also does not exclude a clinically meaningful benefit. This is evidence of uncertainty rather than reassurance that thyroid status is irrelevant. Clinical decisions should avoid labeling TSH 2.5–4.0 mIU/L as SCH in nonpregnant women, distinguish SCH from overt hypothyroidism and euthyroid autoimmunity, and interpret repeat testing against an appropriate local reference interval. Individualized management of confirmed SCH may include observation rather than levothyroxine and should account for reproductive timing, ART context, treatment burden, and patient preferences. Prospective subgroup-specific studies are needed.