HypertensionCohort study
Predicting high excess risk of hyponatraemia among thiazide users
- Population-based cohort study
- causal forest prediction model
- development and validation cohorts
Authors' conclusion
The population-level burden of thiazide-induced hyponatraemia can potentially be markedly reduced by identifying and prescribing alternative antihypertensive drugs to high-risk patient groups using a simple set of baseline information.
Study context
- Population
- Individuals aged ≥40 years, Denmark, 2014–2020development 185,699
- Exposure
- Thiazide vs non-thiazide antihypertensive drugs
- Outcome
- Moderate-to-severe hyponatraemia (plasma sodium <130 mmol/L)
- Follow-up
- Within 120 days of treatment
Main finding
As reported in the abstract
Excess risk of hyponatraemia with thiazide · 120 days
- Average excess risk among thiazide-treated
- 1.8% (95% CI, 1.3-2.2%)
- Individual-level excess risk range
- −1.6% to 15.9%
- Average excess risk in top 10% highest-risk thiazide-treated
- 7.4% (95% CI, 4.4-10.5%)
- Four-covariate model (age, baseline sodium, haemoglobin, C-reactive protein) concordance-for-benefit, validation cohort
- 0.66 (95% CI, 0.66-0.67)
- Reassigning high-risk group to non-thiazide drugs: reduction in excess risk
- 0.7% (95% CI, 0.4-1.0%); 42% relative reduction
Observational data: associations, not effects.
Original publication
Predicting high excess risk of hyponatraemia among thiazide users. Eur J Prev Cardiol. Published 30 July 2025. doi:10.1093/eurjpc/zwaf474
Summarised from the published abstract; every number is checked against it