Endocrinology

Endocrinology journals mention AI but lack standardized reporting and reproducibility safeguards

Among the top 100 endocrinology journals, 84% mention AI in author instructions and 79% require disclosure of AI use. No journal allowed AI authorship, and only 64% permitted AI use in manuscript writing. Only 1% required AI-specific reporting guidelines and few endorsed IMCJE or COPE statements. Implication: editors and authors should expect variable journal requirements and prioritize transparent reporting and protocol registration. 1

Meta-analysis: HRT yields a small reduction in perimenopausal depressive symptoms

Twelve randomized trials showed HRT reduced depressive symptom severity versus placebo with SMD -0.23 (95% CI -0.43 to -0.03). Subgroup signals favored tibolone or selective tissue estrogenic activity regulator regimens but lacked statistical confirmation. Adverse events were generally mild to moderate across trials. Implication: consider HRT for symptomatic perimenopausal depression only after discussing modest benefit and low-certainty evidence. 2

Cluster RCT: culturally tailored educational films did not lower GDM-related adverse perinatal outcomes

In parallel trials in Uganda and India, a film-based package did not change the composite adverse perinatal outcome (pooled PR 1.03, 95% CI 0.97–1.10). Country-specific PRs were neutral: Uganda PR 1.00 (95% CI 0.87–1.14); India PR 1.04 (95% CI 0.96–1.11). High participation but variable loss to follow-up suggests implementation feasibility without clinical benefit. Implication: educational films alone are insufficient; pair with more intensive care or system changes to affect outcomes. 3