Magnitude of the impact of hot flashes on sleep in perimenopausal women

Abstract

Objective: To quantify the impact of objectively recorded hot flashes on objective sleep in perimenopausal women.

Design: Cross-sectional study. Participants underwent 1-5 laboratory-based polysomnographic recordings for a total of 63 nights, including sternal skin-conductance measures, from which 222 hot flashes were identified according to established criteria. Data were analyzed with hierarchical mixed-effect models and Spearman's rank correlations.

Setting: Sleep laboratory.

Patient(s): Thirty-four perimenopausal women (age ± SD: 50.4 ± 2.7 years).

Intervention(s): None.

Main outcome measure(s): Perceived and polysomnographic sleep measures (sleep quality, amount of time spent awake after sleep onset, and number of awakenings). Subjective (frequency and level of bother) and objective (frequency and amount of hot flash-associated awake time) hot-flash measures.

Result(s): Women had an average of 3.5 (95% confidence interval: 2.8-4.2, range = 1-9) objective hot flashes per night. A total of 69.4% of hot flashes were associated with an awakening. Hot flash-associated time awake per night was, on average, 16.6 minutes (95% confidence interval: 10.8-22.4 minutes), which accounted for 27.2% (SD 27.1) of total awake time per night. Hot flash-associated time awake, but not hot flash frequency, was negatively associated with sleep efficiency and positively associated with waking after sleep onset. In addition, self-reported wakefulness correlated with hot flash-associated waking, suggesting that women's estimates of wakefulness are influenced by the amount of time spent awake in association with hot flashes during the night. Having more perceived and bothersome hot flashes was correlated with more perceived wakefulness and awakenings and more objective hot flash-associated time awake and hot-flash frequency.

Conclusion(s): The presence of physiological hot flashes accounts for a significant proportion of total objective time awake during the night in perimenopausal women.

Keywords: Hot flash; menopause; polysomnography; skin conductance; sleep.

Figure 1

Calculation of the amount of wake associated with the presence of hot flashes. In the top panel, the stages of sleep (wake, REM, N1, N2 and N3) are represented as a function of time (hypnogram), as obtained from the polysomnographic (PSG) recording from one participant. Above the hypnogram, three physiological hot flashes (defined as an increase in skin conductance >2 µmho in 30 s) are marked. Below the hypnogram, the skin conductance signal for the three hot flashes is plotted with a time resolution of 20 min (from 5 min before to 15 min after the onset of the hot flashes); the PSG sleep stages for each of the forty 30-s epochs within the 20-min window is displayed under each hot flash. Hot flashes HF#1 and HF#3 were included in the calculation of hot flash impact (HF-impact_[obj], min) as the sum of the minutes of PSG wake time associated with each hot flash (6.5 min [HF#1] + 11.5 min [HF#3]). For this night, the total amount of wake associated with hot flashes (HF-impact_[obj]) was 18 min. The hot flash impact index (wake time associated with hot flashes as a percentage of total wakefulness after sleep onset) was calculated as 32.17 %. HF#2 was not included in the calculation of HF-impact_[obj] since the wake period started more than 1 min before the start of a rise in skin conductance, making the association between the hot flash and the awakening uncertain.

Figure 2

Amount of wake associated with the presence of hot flashes. Percentage of wake after sleep onset (WASO) associated with hot flashes for each woman, showing the contribution of number of nights and hot flashes to each percentage value.