Decreased serum potassium may disturb sleep homeostasis in essential hypertensives.

The aim is to investigate the association between alterations in the serum potassium (K+) concentration and sleep architecture parameters in essential hypertensives. Two hundred ninety-two hypertensives undergoing polysomnography and providing blood samples were recruited. The sleep architecture was composed of sleep stages 1 (N1), 2 (N2), 3 (N3), 4 (N4) and REM. The light sleep stage (LST) was composed of N1 + N2, and the deep sleep stage (DST) was composed of N3 + N4. The potentialrelationships between electrolytes and sleep parameters were determined via univariate and multivariate analyses. The subjects were divided into two groups via the serum K median (3.86 mmol/L). The K < 3.86 mmol/L group showed significantly decreased N1 (7.10 ± 4.55% vs 8.61 ± 5.23%, p = 0.002), LST (71.48 ± 11.33% vs 75.92 ± 17.08%, p = 0.013), and periodic leg movement during sleep related to microarousals (MA) /arousal (PLMS-A) [4 (110) vs 8 (315)/night, p < 0.001] and increased REM (17.38 ± 6.43% vs 15.37 ± 6.18%, p = 0.007) compared to the K ≥ 3.86 mmol/L group. A subdivided analysis by gender showed that these changes were more statistically significant in men than in women. Significant positive correlations were identified between K and N1 (r = 0.169, p = 0004), as well as PLMS-A (r = 0.222, p < 0.001) in subjects. Compared to women, a significantly strong correlation was identified between K+ and REM sleep in men (r = 0.158, p = 0.028 vs. r = 0.078, p = 0.442). Multiple linear regression analysis indicated that K+ is significantly associated with N1 in all subjects (p = 0.03) and with REM in men (p = 0.008), even after adjusting for confounders. Decreased K+ may disturb the homeostasis of the sleep architecture, and gender may interfere with their links in the hypertensive population.