Factors of importance for discontinuation of thiazides associated with hyponatremia in Sweden: A population-based register study

Association between newly initiated thiazide diuretics and hospitalization due to hyponatremia

Purpose

Thiazide diuretics are the most common origin of drug-induced hyponatremia. However, population-based studies on clinical outcomes are lacking. We therefore explored the time course and absolute risk of thiazide-associated hospitalization due to hyponatremia in Sweden.

Methods

Population-based case-control study including patients hospitalized with a principal diagnosis of hyponatremia (n = 11,213) compared with controls (n = 44,801). Linkage of registers was used to acquire data. Multivariable regression was applied to explore time-dependent associations between thiazide diuretics and hospitalization due to hyponatremia. Attributable risks were calculated assessing the disease burden attributable to thiazides.

Results

Individuals initiating thiazide treatment were exposed to an immediate increase in risk for hospitalization with adjusted odds ratio (aOR) (95% CI) of 48 (28–89). The associations gradually declined reaching an aOR of 2.9 (2.7–3.1) for individuals treated for longer than 13 weeks. The attributable risk of hyponatremia-associated hospitalization due to thiazides of any treatment length was 27% (3095/11,213). Among 806 patients initiating treatment < 90 days before hospitalization, hyponatremia could be attributed to thiazides in 754. Based on nationwide data, 616,678 individuals were initiated on thiazides during the 8-year study period suggesting an absolute risk of 0.12% (754/661,678) for subsequent hospitalization with a main diagnosis of hyponatremia.

Conclusions

Thiazide diuretics attributed to more than one in four individuals hospitalized due to hyponatremia. The risk increase was very pronounced during the first month of treatment and then gradually declined, without returning to normal. However, the absolute risk for the development of hyponatremia demanding hospitalization may for most individuals be modest.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00228-020-03086-6.

Associations Between Antihypertensive Medications and Severe Hyponatremia: A Swedish Population-Based Case-Control Study

BACKGROUND

Calcium channel blockers (CCBs), beta-receptor blockers (BBs), angiotensin-converting enzyme inhibitors (ACEIs), and angiotensin II receptor blockers (ARBs) have occasionally been reported to cause severe hyponatremia. The aim was to explore the association between CCBs, BBs, ACEIs, and ARBs and hospitalization due to hyponatremia.

METHODS

Patients hospitalized with a principal diagnosis of hyponatremia (n = 11 213) were compared with matched controls (n = 44 801). Linkage of national population-based registers was used to acquire data. Multivariable logistic regression adjusting for co-medications, diseases, previous hospitalizations, and socioeconomic factors was used to explore the association between hospitalization for severe hyponatremia and the use of different CCBs, BBs, ACEIs, and ARBs. Furthermore, newly initiated (≤90 days) and ongoing use were examined separately.

RESULTS

Adjusted odds ratios (aORs) (95% confidence interval) for the investigated 4 drug classes ranged from 0.86 (0.81-0.92) for CCBs to 1.15 (1.07-1.23) for ARBs. For newly initiated drugs, aORs spanned from 1.64 (1.35-1.98) for CCBs to 2.24 (1.87-2.68) for ACEIs. In contrast, the corresponding associations for ongoing therapy were not elevated, ranging from 0.81 (0.75-0.86) for CCBs to 1.08 (1.00-1.16) for ARBs. In the CCBs subgroups, aOR for newly initiated vascular CCBs was 1.95 (1.62-2.34) whereas aOR for ongoing treatment was 0.82 (0.77-0.88).

CONCLUSIONS

For newly initiated CCBs, BBs, ACEIs, and ARBs, the risk of hospitalization due to hyponatremia was moderately elevated. In contrast, there was no evidence that ongoing treatment with investigated antihypertensive drugs increased the risk for hospitalization due to hyponatremia.

Inverse association between glucose-lowering medications and severe hyponatremia: a Swedish population-based case-control study

CONTEXT

Glucose-lowering medications have occasionally been reported to cause hyponatremia, but the evidence is scarce.

OBJECTIVES

To explore the association between glucose-lowering medications and severe hyponatremia.

DESIGN, SETTING, AND PARTICIPANTS

Subjects hospitalized with a principal diagnosis of hyponatremia (n = 14,359) were compared with matched controls (n = 57,383). Data were derived by linkage of national population-based registers. Multivariable logistic regression adjusting for co-medication, diseases, previous hospitalizations, and socioeconomic factors was used to explore the association between hospitalization for hyponatremia and the use of different glucose-lowering medications. Furthermore, newly initiated (≤90 days) and ongoing use was investigated separately.

MAIN OUTCOME MEASURES

Hospitalization due to hyponatremia.

RESULTS

The unadjusted ORs (95% CI) for hospitalization due to hyponatremia were 1.41 (1.29-1.54) for insulins, 1.38 (1.27-1.50) for metformin, and 1.22 (1.07-1.38) for sulfonylureas. However, after adjustment for confounding factors the association was consistently reversed. Thus, for any glucose-lowering medication the adjusted OR was 0.63 (0.58-0.68). For insulins, metformin and sulfonylureas, adjusted ORs (95% CI) were 0.58 (0.52-0.65), 0.81 (0.72-0.90) and 0.81 (0.69-0.94), respectively. Odds ratios for newly initated medications were overall higher while those for ongoing treatment were further decreased. Thus, adjusted ORs (95% CI) for ongoing treatment with insulins, metformin, and sulfonylureas were 0.54 (0.48-0.61), 0.82 (0.73-0.91) and 0.78 (0.66-0.92).

CONCLUSIONS

Glucose-lowering medications did not increase the risk for hospitalization due to severe hyponatremia. In fact, the association was inverse across all investigated drugs. The association may be mediated by pharmacologic mechanisms, but the uniform effects across drug-classes suggest properties of the diabetic disease are of importance.