Risk of antidepressant-induced hyponatremia in geriatric patients: a systematic review and meta-analysis

BACKGROUND: Hyponatremia is a frequent electrolyte disorder in clinical practice and is particularly prevalent among elderly populations. Antidepressants have been identified as an important pharmacological contributor. However, the event rate, overall risk, and differences across antidepressant classes remain incompletely characterized. This systematic review and meta-analysis aimed to quantify the event rate and comparative risk of antidepressant-induced hyponatremia in geriatric populations. METHODS: PubMed, Web of Science, and EMBASE were systematically searched from inception to February 5, 2026. A random-effects model was applied to estimate pooled event rates and risk ratios (RRs) among antidepressant users. Subgroup analyses were conducted according to antidepressant classes and individual compounds. Head-to-head comparisons were performed to evaluate differences in hyponatremia risk across antidepressant classes. RESULTS: A total of 21 cohort studies, comprising 2,115,187 geriatric antidepressant users, were included. The pooled event rates of any hyponatremia and clinically relevant hyponatremia were 0.0825 (95% confidence interval [CI]: 0.0566-0.1125) and 0.0287 (95% CI: 0.0117-0.0526), respectively. Serotonin and noradrenaline reuptake inhibitors (SNRIs) demonstrated the highest event rates (0.1040 and 0.0503, respectively), followed by selective serotonin reuptake inhibitors (SSRIs: 0.0981 and 0.0275, respectively). Antidepressant use was associated with a significantly increased risk of any hyponatremia (RR = 2.43, 95% CI: 1.42-4.14, P = 0.001) and clinically relevant hyponatremia (RR = 2.96, 95% CI: 1.57-5.57, P < 0.001), resulting in an absolute increase of 0.0037 (95% CI: 0.0011-0.0080) and 0.0046 (95% CI: 0.0013-0.0108) in event rate compared with non-users, respectively. This increased risk remained significant in subgroups of SSRIs (RR = 2.40 and 3.22, respectively) and SNRIs (RR = 2.90 and 2.91, respectively). Head-to-head comparisons indicated that SNRIs were associated with higher risks of both any and clinically relevant hyponatremia compared with SSRIs. In contrast, mirtazapine was associated with significantly lower risks of any hyponatremia (RR = 0.61, 95% CI: 0.39-0.61, P = 0.031) and clinically relevant hyponatremia (RR = 0.49, 95% CI: 0.39-0.61, P < 0.001) compared with SSRIs. CONCLUSION: Antidepressant use in geriatric patients is associated with an increased risk of hyponatremia, with clinically relevant differences observed among antidepressant classes. SNRIs appear to confer a relatively higher risk, whereas mirtazapine may represent a comparatively safer option. However, these findings should be interpreted considering substantial heterogeneity and potential residual confounding among observational studies.