(221) Incidence of moderate and profound hyponatremia during treatment with selective serotonin reuptake inhibitors (SSRIs) – evidence from the Stockholm Sodium Cohort (SSC).

Background: Hyponatremia is a known side effect of SSRIs but evidence is largely based on case reports and knowledge about the incidence during early phases of therapy is limited.

Objectives: To determine the incidence of hyponatremia during the first 10 weeks of SSRI treatment.

Methods:

Design: Registry-based cohort study.
Setting and exposure: Using the new Stockholm Sodium Cohort (SSC) encompassing longitudinal clinical and laboratory data on 1.6 million Stockholm inhabitants, individuals with a verified normal plasma sodium (P-Na) level when starting first-ever (since 2005) treatment with an SSRI were identified (n=12,352).
Outcome: Moderate (P-Na < 130 nmol/L) or profound (P-Na < 125 nmol/L) hyponatremia, based on routine clinical chemistry data.
Statistical analysis: Proportion of patients presenting with each outcome was calculated week-by-week during weeks 1-10 of SSRI treatment, along with 95% confidence intervals. Subgroup analyses were performed in men, women and patients >65 years-of-age.

Results: The risk of moderate hyponatremia peaked during week 4 with 19.4 per 10,000 treated patients (95% ci 12.7-29.4) having a P-Na < 130 mmol/L during this week, thereafter declining rapidly. For profound hyponatremia, the peak was reached week 3 (5.7 per 10,000, 95% ci 2.5-12.2). In patients aged 65 or older (n=3497), the pattern was similar, but the incidence 3 times higher (moderate 57.2 per 10,000, 95% ci 35.9-90.0, profound 20.0 per 10,000, 95% ci 8.8-43.2). In women (n=7351), moderate hyponatremia peaked already during week 2 (25.8 per 10,000, 95% ci 16.0-41.2) compared to week 4 in men (n=5001, 24.0 per 10,000, 95% ci 13.0-43.2) and profound hyponatremia was very rare in men (affecting only 3/5001 individuals over the full 10 week period).

Conclusions: In this large, population-based cohort, the incidence of hyponatremia during the early phases of SSRI treatment was highest during the first four weeks, possibly earlier in women. The risk was highest in the elderly, but even in those it did not exceed 0.6% during any given week. The incidence could be under-estimated due to incomplete sampling, but at least profound hyponatremia is expected to be symptomatic and prompt measurement of sodium, reducing the impact of such bias.