News Release

Study shows that people with sleep apnea have a high risk of death

18-year follow-up study of a random sample of men and women produces striking new insights regarding the public health burden of sleep apnea

Peer-Reviewed Publication

American Academy of Sleep Medicine

WESTCHESTER, Ill. – A study in the August 1 issue of the journal Sleep shows that people with severe sleep apnea have a much higher mortality risk than people without sleep apnea, and this risk increases when sleep apnea is untreated.

Results show that people who have severe sleep apnea, which involves frequent breathing pauses during sleep, have three times the risk of dying due to any cause compared with people who do not have sleep apnea. This risk is represented by an adjusted hazard ratio of 3.2 after controlling for age, sex and body mass index. When 126 participants who reported regular use of continuous positive airway pressure (CPAP) therapy were removed from the statistical analysis, the hazard ratio for all-cause mortality related to severe sleep apnea rose to 4.3.

"We found that both men and women with sleep apnea in the general population - not patients - mostly undiagnosed and untreated, had poorer survival compared with persons without sleep apnea, given equal BMI, age and sex," said principal investigator and lead author Terry Young, PhD, professor of epidemiology at the University of Wisconsin-Madison.

According to Young, most previous studies of sleep apnea and mortality have involved patients referred for a clinical sleep diagnostic evaluation; the mortality risk for sleep apnea in the general population has not been previously reported.

The study was an 18-year follow-up of 1,522 participants in the ongoing Wisconsin Sleep Cohort Study, which was established in 1988 and involved a random sample of men and women from the community who were between the ages of 30 and 60 when the study began. After spending one night at the University of Wisconsin General Clinical Research Center for assessment by polysomnography, participants were categorized by apnea-hypopnea index (AHI), which is the average number of breathing pauses (apneas) and reductions (hypopneas) per hour of sleep. Sixty-three individuals (about four percent) had severe sleep apnea at baseline with an AHI of 30 or more and a range of 30 to 97 apneas and hypopneas per hour. About 76 percent of the study group (1,157 individuals) had no sleep apnea with an AHI of less than five.

For the follow-up study, state and national death records were reviewed up to March 1, 2008, to identify participants who had died and to note the causes of death listed on the death certificates. Eighty deaths were recorded, including 37 deaths attributed to cancer and 25 deaths attributed to cardiovascular disease and stroke.

About 19 percent of participants with severe sleep apnea died (12 deaths), compared with about four percent of participants with no sleep apnea (46 deaths). Although participants with mild sleep apnea (AHI of five to 14) or moderate sleep apnea (AHI of 15 to 29) had a mortality risk that was 50 percent greater than those with no sleep apnea, the results did not achieve statistical significance.

Hazard ratios for all-cause mortality remained high after further adjustments for other factors such as smoking, alcohol use, sleep duration and total cholesterol. Severe sleep apnea was associated with increased mortality whether or not participants experienced daytime sleepiness.

About 42 percent of deaths in people with severe sleep apnea (5 of 12 deaths) were attributed to cardiovascular disease or stroke, compared with 26 percent of deaths in people with no sleep apnea (12 of 46 deaths). When the 126 participants who reported regular CPAP use were removed from the analysis, the hazard ratio for cardiovascular mortality soared from 2.9 to 5.2 for people with severe sleep apnea. The results suggest that regular CPAP use may protect sleep apnea patients against cardiovascular death.

"I was surprised by how much the risks increased when we excluded people who reported treatment with CPAP," Young said. "Our findings suggest - but cannot prove - that people diagnosed with sleep apnea should be treated, and if CPAP is the prescribed treatment, regular use may prevent premature death."

Statistical adjustments show that high blood pressure, cardiovascular disease, stroke and diabetes may play a role in the association between sleep apnea and mortality, but the specific mechanisms by which sleep apnea contributes to mortality remain unclear.

The study was supported by grants from the National Institutes of Health.

According to the American Academy of Sleep Medicine, obstructive sleep apnea (OSA) involves a decrease or complete halt in airflow despite an ongoing effort to breathe. It occurs when the muscles relax during sleep, causing soft tissue in the back of the throat to collapse and block the upper airway.

The most common treatment for OSA is CPAP therapy, which provides a steady stream of air through a mask that is worn during sleep. This airflow keeps the airway open to prevent pauses in breathing and restore normal oxygen levels.

###

A media fact sheet about obstructive sleep apnea is available from the AASM at http://www.aasmnet.org/Resources/FactSheets/SleepApnea.pdf. Information from the AASM for patients and the public is available about OSA at http://www.sleepeducation.com/Disorder.aspx?id=7 and about CPAP at http://www.sleepeducation.com/CPAPCentral/.

Sleep is the official journal of the Associated Professional Sleep Societies, LLC, a joint venture of the American Academy of Sleep Medicine and the Sleep Research Society.

Dr. Terry Young will participate in a Media Conference Call at 3 p.m. CDT on Tuesday, July 29. To register for the conference call; for a copy of the study, "Sleep-Disordered Breathing and Mortality: Eighteen-Year Follow-Up of the Wisconsin Sleep Cohort;" or to arrange an interview with an AASM spokesperson, please contact Kathleen McCann, AASM director of communications, at (708) 492-0930, ext. 9316, or kmccann@aasmnet.org. Please register for the Media Conference Call by 12 p.m. CDT on Monday, July 28.

Journal Sleep Fact Sheet: Obstructive Sleep Apnea & Mortality
Comparing the Wisconsin Sleep Cohort with the Busselton Health Study
Journal Sleep – August 1, 2008

Two studies in the August 1 issue of the journal Sleep link obstructive sleep apnea (OSA) to mortality risk. This fact sheet compares the design and results of the two studies.

Study Design
Wisconsin: Random, population-based sample of adults who were recruited from several Wisconsin state agencies.

Busselton: Community-based sample of residents in the rural town of Busselton in the state of Western Australia.
Note: Previous studies linking OSA to mortality involved sleep-clinic patients instead of population samples.

Study Group
Wisconsin: 1,522 adults between the ages of 30 and 60 years

Busselton: 380 adults between the ages of 40 and 65 years

Sleep Assessment
Wisconsin: One night of in-lab, 18-channel polysomnography conducted at the University of Wisconsin General Clinical Research Center

Busselton: One night of at-home, 4-channel portable monitoring

Primary Measurement
Wisconsin: Sleep apnea or "sleep-disordered breathing," measured by the apnea-hypopnea index (AHI) – per hour of sleep, the average number of apneas (breathing pauses of 10 seconds or more) and hypopneas (breathing reductions with a decrease in blood oxygen saturation of four percent or more).
Severe (AHI 30 or more), Moderate (AHI 15 to 29), Mild (AHI 5 to 14), No sleep apnea (AHI less than 5)

Busselton: Obstructive sleep apnea, measured by the respiratory disturbance index (RDI) – per estimated hour of sleep, the average number of oxygen desaturations of three percent or more that were accompanied by an increased heart rate of 10 or more beats per minute and/or a burst of snoring at the beginning and end of the desaturation event.
Moderate-Severe (RDI 15 or more), Mild (RDI 5 to 14), No OSA (RDI less than 5)

Mortality Follow-Up
Wisconsin: Eighteen years, reviewing state and national death records up to March 1, 2008.
Mean follow-up: 13.8 years. Total observation: 20,963 person-years

Busselton: Fourteen years, reviewing state and national death records through 2004.
Mean follow-up: 13.4 years. Total observation: 5,073 person-years

Deaths
Wisconsin: Total: 80 of 1,522 individuals (5.3%). Severe sleep apnea: 12 of 63 (19%).
Moderate: 6 of 82 (7.3%). Mild: 16 of 220 (7.3%). No sleep apnea: 46 of 1,157 (4%).

Busselton: Total: 33 of 380 individuals (8.7%). Moderate-Severe OSA: 6 of 18 (33%).
Mild OSA: 5 of 77 (6.5%). No OSA: 22 of 285 (7.7%).

All-Cause Mortality Hazard Ratios
Wisconsin: Severe sleep apnea: 3.2. Moderate sleep apnea: 1.3. Mild sleep apnea: 1.5.
Adjusted for age, sex and body mass index. Hazard ratios were significantly increased with SDB severity. Adjusted hazard ratios remained high after further adjustments for smoking, alcohol use, general health status, educational status, neck girth, waist:hip ratio, sleep duration and total cholesterol.

Busselton: Moderate-Severe OSA: 6.24. Mild OSA: 0.47.
Adjusted for age, sex, body mass index, smoking, total cholesterol, HDL cholesterol, diabetes, angina and blood pressure. After controlling for all the same risk factors except blood pressure, the hazard ratio for moderate-severe OSA was 4.4. Only moderate-severe OSA was statistically significant.

Effect of Continuous Positive Airway Pressure (CPAP) Therapy
Wisconsin: When 126 participants who reported regular CPAP use were removed from the statistical analysis, the adjusted hazard ratio for all-cause mortality related to severe sleep apnea jumped from 3.2 to 4.3. The adjusted hazard ratio for cardiovascular mortality related to severe sleep apnea soared from 2.9 to 5.2. These results suggest that CPAP was protective particularly against cardiovascular death.

Busselton: The people living in Busselton did not have regular access to sleep medicine services, and treatment with CPAP was probably rare. The untreated all-cause mortality hazard ratio of 4.3 seen in the Wisconsin cohort is very similar to the hazard ratio of 4.4 observed in the most comparable analysis from the Busselton study.


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.