Thiazide-Associated Hyponatremia

In This Article

Abstract and Introduction

Abstract

Background Hyponatremia is one of the most common adverse reactions to thiazide diuretics. In the present study, we analyzed differences in thiazide-associated hyponatremia between men and women and between different categories of age, body mass index (BMI), daily thiazide dose, and estimated glomerular filtration rate.

Study Design Population-based cohort study.

Setting & Participants 13,325 individuals 45 years and older living in a suburb of Rotterdam, as part of the Rotterdam Study.

Predictor Exposure to thiazide diuretics.

Outcomes The association between thiazide exposure and hyponatremia (defined as sodium level ≤135 mmol/L; mild hyponatremia, 130-≤135 mmol/L; moderate, >125-<130 mmol/L; and severe, ≤125 mmol/L) was studied in a period covering more than 10 years using Cox proportional hazard regression analyses.

Results 718 participants used thiazides at baseline, and 2,738 participants started on thiazide therapy during follow-up. 522 participants developed hyponatremia, of whom 32.4% were exposed to thiazide diuretics at the time of hyponatremia. Thiazide exposure was associated with an almost 5 times higher risk of hyponatremia than no exposure (HR, 4.95; 95% CI, 4.12–5.96). The risk of mild hyponatremia was more than 4.5 times higher in thiazide-exposed individuals; risks of moderate and severe hyponatremia were both 8 times higher in individuals exposed to thiazides. Age and BMI (but not sex [P = 0.8] or estimated glomerular filtration rate [P = 0.2]) significantly modified this risk of thiazide-associated hyponatremia (P < 0.05).

Limitations Some cases of severe hyponatremia may have been missed if patients were admitted to the hospital without assessment of serum sodium in the general practitioner's laboratory. Nonproportionality of hazards in the first period was explained as possible "depletion of susceptibles" in this closed cohort.

Conclusions Thiazide use is associated with a substantially increased risk of hyponatremia. Age and BMI significantly influenced the thiazide-associated risk of hyponatremia.

Inroduction

Diuretics are a common cause of hyponatremia, as first was reported almost 50 years ago.^[1–3] Thiazide diuretics account for the majority of cases of diuretic-associated hyponatremia.^[4,5]

Various mechanisms have been identified for thiazide-associated hyponatremia;^[6,7] however, the 2 main causes are sodium deficiency and water overload. Thiazides inhibit sodium chloride reabsorption in the distal convoluted tubule, which leads to direct inhibition of urinary dilution capacity. Furthermore, due to direct diuretic action and an increase in vasopressin secretion, sodium loss may be relatively higher than water loss.^[6,8]

After the initiation of thiazide therapy, a new equilibrium in sodium-water balance is reached, usually within a period of 2 weeks.^[9] Hyponatremia is most likely to occur within this first period,^[4] but may occur at any time during treatment as a consequence of physiologic or environmental changes.^[10,11]

Women have a more than 3 times higher risk to be hospitalized due to diuretic-induced hyponatremia than men.^[12,13] The reason that hyponatremia is observed more frequently in women is unclear, but older age and lower body mass have been incriminated.^{[4,8,10,14,15,16]} Thiazides are prescribed more frequently in women,^[17] and there is an ongoing debate about whether a higher risk of hyponatremia in women is genuine or can be explained by more frequent use.

The present study was conducted in a population-based cohort. We analyzed differences in thiazide-associated hyponatremia between men and women and in different groups of age, body mass index (BMI), dose, and kidney function.

The National Kidney Foundation

Published by Elsevier

Table 1. Baseline Characteristics
	Study Population
No. of participants	13,325
Female sex	7,870 (59.1)
Age (y)	63.1 ± 13.9
BMI (kg/m²)	27.0 ± 4.1
eGFR (mL/min/1.73 m²)	72.5 ± 18.2
Diabetes mellitus	927 (7.4)
Prevalent MI	749 (8.0)
Prevalent HF	191 (2.0)
Systolic BP (mm Hg)	138 ± 22
Diastolic BP (mm Hg)	74 ± 12
Thiazide users	718 (5.4)
Other medication use
Cardiac therapy	888 (7.8)
Antihypertensives	785 (6.9)
β-Blockers	1,916 (14.4)
Calcium blockers	754 (5.7)
RAAS inhibitors	1,249 (9.4)
Lipid-lowering agents	1,417 (12.5)

Note: Values for categorical variables are given as number (percentage); values for continuous variables, as mean ± standard deviation. Due to missing values, some percentages do not reflect proportion of total number of participants.
Abbreviations: BMI, body mass index; BP, blood pressure; eGFR, estimated glomerular filtration rate; HF, heart failure; MI, myocardial infarction; RAAS, renin-angiotensin-aldosterone system.

Table 2. Risk of Thiazide-Associated Hyponatremia per Stratum
	Exposed Cases	HR^a (95% CI)	HR^b (95% CI)
Sex
Male	24/135	3.71 (2.38–5.77)	3.79 (2.42–5.93)
Female	145/387	5.31 (4.32–6.53)	5.23 (4.24–6.44)
Thiazide dose
DDD <1.0	39	3.42 (2.46–4.77)	3.49 (2.50–4.87)
DDD ≥1.0	130	5.72 (4.67–7.00)	5.61 (4.76–6.89)
Age category^c
≤55 y	2/5	NA	NA
>55–65 y	10/35	7.88 (3.77–16.48)	7.70 (3.67–16.17)
>65–75 y	33/96	6.18 (4.03–9.48)	5.44 (3.53–8.38)
>75–85 y	77/221	4.52 (3.41–5.99)	4.74 (3.58–6.28)
>85 y	47/165	3.83 (2.73–5.37)	3.99 (2.84–5.60)
BMI category^c
<25 kg/m²	36/115	7.91 (5.31–11.79)	7.06 (4.71–10.60)
25-<27 kg/m²	35/109	5.93 (3.95–8.92)	6.07 (4.01–9.20)
27–29 kg/m²	40/98	5.77 (3.82–8.70)	6.37 (4.21–9.61)
>29 kg/m²	40/133	2.98 (2.05–4.32)	3.31 (2.28–4.81)
eGFR category
<63 mL/min/1.73 m²	74/209	3.33 (2.50–4.44)	3.49 (2.61–4.66)
63–74 mL/min/1.73 m²	43/136	3.94 (2.73–5.67)	3.86 (2.67–5.58)
75–84 mL/min/1.73 m²	32/108	4.55 (3.00–6.90)	4.28 (2.81–6.53)
≥84 mL/min/1.73 m²	18/63	4.46 (2.37–7.74)	4.39 (2.51–7.68)

Note: Unless otherwise indicated, values are given as number/total number in category.
Abbreviations: BMI, body mass index; CI, confidence interval; DDD, defined daily dose; eGFR, estimated glomerular filtration rate; HR, hazard ratio; NA, not applicable.
a HRs represent the risk of hyponatremia associated with thiazide exposure, adjusted for sex and age.
b HRs represent the risk of hyponatremia associated with thiazide exposure, adjusted for sex, age, heart failure, and blood pressure (systolic and diastolic).
c P < 0.05 for multiplicative interaction with thiazides (calculated using interaction terms in the regression model); hyponatremia is defined as serum sodium level ≤135 mmol/L.