Less rapid eye movement (REM) sleep is associated with an increased risk for death in middle-aged and older adults, new research suggests.
Investigators at the University of California, San Diego, found that, over a 12-year period, each 5% reduction in REM sleep was associated with a 13% increase in mortality rate. However, the investigators noted that this is only an association and does not indicate cause and effect.
“Determining causality can be difficult,” study investigator Sonia Ancoli-Israel, PhD, professor emeritus of psychiatry at the University of California, San Diego, said in an interview.
“It is therefore important that physicians and the public understand that our findings suggest an increased risk, but that does not mean that reduced REM will always result in shorter survival. With all the self-monitoring sleep gadgets available to the public, I would caution against any panic if one notices reduced REM. But mentioning it to a physician may be a clue to examine what else might be going on with that patient that could more easily be targeted,” Dr. Ancoli-Israel added.
The research was published online July 6 in JAMA Neurology.
Approximately 50-70 million Americans have problems with sleep. Such problems have a multitude of consequences for health, including cardiovascular disease; metabolic, psychiatric, and cognitive disorders; lower quality of life; and increased mortality.
The investigators noted that the aspects of sleep that may be driving this association remain unclear. Because decreased REM sleep has been associated with poor mental and physical health outcomes, the researchers hypothesized that decreased REM sleep may be associated with an increased risk for death.
To test this hypothesis, they conducted a multicenter, population-based, cross-sectional investigation using data from independent cohorts – the Outcomes of Sleep Disorders in Older Men (MrOS) Sleep Study and the Wisconsin Sleep Cohort (WSC). The MrOS cohort included 2,675 men (mean age, 76.3 years) who were recruited from December 2003 to March 2005 at six U.S. centers and were followed for a median of 12.1 years. The WSC cohort included 1,386 individuals (54.3% men; mean age, 51.5 years) and had a median follow-up of 20.8 years. Data from this study were used to replicate the findings from the MrOS study.
Primary outcome measures included all-cause and cause-specific mortality, which were confirmed using death certificates.
Participants in both cohorts underwent polysomnography and evaluation with the Epworth Sleepiness Scale. For MrOS participants, investigators calculated the total number of minutes per night spent in REM sleep and the corresponding percentage of total sleep time.
Less sleep, more death
Self-report sleep measures in MrOS participants were collected using the Pittsburgh Sleep Quality Index and the Functional Outcomes of Sleep Questionnaire
The investigators contacted participants in MrOS every 4 months to determine vital status. Cause of death was categorized by the ICD-9 as cardiovascular, cancer, and other. In WSC, the researchers identified deaths by matching participants’ social security numbers with national and state registries. The cause of death was categorized in the same manner as in the MrOS cohort.
Approximately half (53%) of the MrOS cohort died during follow-up. For each mortality category, the highest percentage of deaths occurred among those in the lowest quartile percentage of REM sleep. Adjusted analyses revealed that the MrOS participants had a 13% higher mortality rate for every 5% reduction in REM sleep (hazard ratio, 1.13; 95% confidence interval, 1.08-1.19). These findings were similar for cardiovascular and other causes of death but were not significant for cancer-related mortality. For all mortality categories, the mortality rate was higher for participants who had less than 15% REM sleep per night in comparison with individuals who had 15% or more.
The findings were similar in the WSC cohort despite its younger age, the inclusion of women, and longer follow-up (HR, 1.13; 95% CI, 1.08-1.19). Compared with MrOS participants, WSC participants were more likely to be obese and to use more antidepressants or sedatives. Overall, the mean percentage of REM sleep was 19.2%. Participants in the lowest quartile of REM sleep generally were older, had higher rates of antidepressant use, hypertension, heart attack, and transient ischemic attack, as well as engaging in less physical activity.