Sleep Duration and Mortality Risk in Women

By Gregg D. Jacobs, Ph.D.

It is commonly believed, based on a few scientific studies and information disseminated by the media, that eight hours of sleep nightly is needed for optimum health. In February of 2002, Dr. Dan Kripke and his colleagues at the University of California San Diego studied the relationship between sleep duration and death risk in a sample of about one million people.

They found that people who slept 6-7 hours per night lived the longest, while those who slept 8 or more hours per night had the highest death rates over a six year follow-up period. People who slept 5 hours per night lived longer than those who slept 8, and the longer people slept, the higher their death rate. Because the Kripke study did not control for the effects depression, alcohol use, shift work, and sleep disordered breathing, it was unclear whether the relationship between sleep duration and mortality could be explained by one or more of these variables.

In a major study presented recently at the 2003 APSS national sleep meetings in Chicago, a group of researchers at Harvard Medical School, led by Dr. S.R. Patel, used the Nurses' Health Study cohort to assess whether mortality risk attributed to sleep time could be explained by confounding variables such as depression, alcohol use, sleep disordered breathing , or shift work.

They studied a sample of 82, 975 female nurses (one of the largest samples ever employed in a prospective study on sleep) who answered questions about how much they slept and then were followed for 12 years for causes of death. Nurses who slept 7 hours per day had the fewest deaths over the twelve year follow-up period, which replicated the findings of the Kripke group. For nurses who slept 5 hours per day (short sleepers), there was no increased death risk compared to 7 hour sleepers when controlling for age, smoking, weight, health problems such as diabetes and hypertension, alcohol use, depression, shift work, and snoring (sleep disordered breathing). People who slept 8, 9, or 10 or more hours per day had a significantly increased risk of death compared to 7 hour sleepers even when controlling for age, smoking, health problems, alcohol use, etc. For example, 8 hour sleepers had a 13% increases risk of death, 9 hour sleepers had a 40% increased risk of death, and 10 or more hour sleepers had a 70% increased risk of death compared to 7 hour sleepers.

The Patel group's findings confirm and extend the Kripke group's findings by again demonstrating that people who sleep 7 hours per night have the lowest death rates and those who sleep 8 or more hours per night have progressively increasing rates of death. Although several laboratory based studies using young healthy subjects and small sample sizes have reported that eight hours of sleep is required to maintain optimal cognitive and neurobehavioral functions, methodological limitations in these studies combined with the overwhelming evidence form the Kripke and Patel studies involving dramatically larger sample sizes from "real-life" populations suggests that 7 hours of sleep per day is the optimal amount for health.