Congresso Brasileiro do Sono

Página Inicial » Inscrições Científicas » Trabalhos

Dados do Trabalho


Sleep fragmentation related to moon phases


There are ancient folk beliefs regarding the influence of moon phases on human behavior and physiology. Current literature data show conflicting results about the association between moon phases and sleep architecture. Most studies found less total sleep time (TST), lower sleep efficiency (SE), higher sleep latency (SL) and higher REM sleep latency (REML) in full moon, comparing to new moon.


The objective of this study was to compare sleep architecture with moon phase.


A retrospective analysis of all 30,159 polysomnography reports (PSG) performed between 2008 and 2016 in a large sleep center was conducted. Parameters were age, sex, TST, SE, SL, REML, arousal index (AI), and percent time of sleep stages (N1, N2, N3 and REM). The moon phase was defined by setting the new moon day as zero and full moon either as 14 or 15 (depending on the month). Then phases were divided as New Moon (days 0 to 4), Intermediary Moon (days 5 to 9) and Full Moon (days 10 to 14 or 15). Exclusion criteria were repeated PSG of the same patient, split-night or CPAP PSG studies, age below 18 years, use of sleep-inducing drugs in the night of the PSG, apnea-hypopnea index (AHI) > 15/hour, and periodic limb movement index (PLMI) > 15/hour. Thus, 13,131 eligible PSG studies were analyzed. First, three-way ANOVA with Tukey post-hoc tests were conducted, with moon phase, sex, and age-range (18–35, 36–60, and >60) as group variables. After that, multiple regression analyses were conducted, with moon phase, sex, age, AHI, and PLMI as independent variables. The dependent variables in each test were TST, SE, SL, REML, N3, REM, and AI.


ANOVA results showed greater AI in the elder group comparing to the other two groups (+2.6/hour) and greater AI between middle-age group and the younger group (+1/hour). However, moon phase did not showed difference. The multiple regression analyses showed that PLMI (beta=0.14), AHI (beta=0.13), and age (beta=0.10) were significant in predicting AI; age (beta=-0.27), AHI (beta=0.06), and PLMI (beta=-0.05) in SE; PLMI (beta=-0.04), age (beta=-0.03), and AHI (beta=0.03) in N3; and AHI (beta=0.07) and PLMI (beta=-0.03) in REM. In none of the analyses the moon phase had significant effects.


This is the only clinical-based, large study evaluating the moon influence in sleep architecture parameters. In contrast to the existing literature, moon phases had not a significant effect.


Sleep-wake cycle, moon phase, polysomnography


Área Básica


Universidade Federal do Paraná - Parana - Brasil


Vinícius Hideki Nakamura, Fernando Mazzilli Louzada, João Guilherme Fiorani Borgio