Purpose: Hypomagnesaemia has been associated with various adverse outcomes. Loop and thiazide diuretics promote urinary magnesium excretion. However, it is unknown if this links to hypomagnesaemia. We study if loop or thiazide diuretic use affects serum magnesium levels and if it associates with hypomagnesaemia. In addition, we study the effect of combining a potassium-sparing diuretic with a thiazide diuretic on the presence of hypomagnesaemia.
Methods: The study performed a cross-sectional analysis within 9820 participants from the prospective Rotterdam Study. Hypomagnesaemia was defined as a serum magnesium level ≤0.72 mmol/L. Participants were categorized by defined daily dose (DDD), and all analyses were adjusted for age, sex, BMI, eGFR, serum potassium levels, proton pump inhibitor use, and comorbidities.
Results: Loop diuretic use was associated with higher serum magnesium levels (<1 DDD: 0.004 mmol/L 95% CI: -0.008; 0.017; 1 DDD: 0.023 mmol/L 95% CI: 0.013; 0.032; >1 DDD: 0.043 mmol/L 95% CI: 0.028; 0.057). Thiazide diuretic use was associated with lower serum magnesium levels (<1 DDD: -0.013 mmol/L 95% CI: -0.023; -0.002; ≥1 DDD: -0.018 mmol/L 95% CI: -0.028; -0.010), resulting in an increased odds ratio of hypomagnesaemia of 3.14 (95% CI: 1.67; 5.92) and 2.74 (95% CI: 1.57; 4.77), respectively. These effects were predominantly seen in participants using diuretics for more than 390 days. Combining thiazide diuretics with a potassium-sparing agent was not associated with lower serum magnesium levels or hypomagnesaemia.
Conclusions: Thiazide diuretic use is associated with lower serum magnesium levels and an increased risk of hypomagnesaemia. This increased risk is not seen in participants using a combination of thiazide diuretics with a potassium-sparing agent. The use of loop diuretics is not associated with an increased risk of hypomagnesaemia.
Keywords: diuretics; epidemiology; hypomagnesaemia; magnesium; pharmacoepidemiology; population-based.
© 2018 John Wiley & Sons, Ltd.