Wednesday September 1 2010
It is unclear whether insomnia actually causes early death
“Men who find themselves unable to sleep during the small hours of the night may end up dying younger,” reported the Daily Mail.
This study looked at people’s insomnia and their risk of dying over a 14-year period. At the start of the study, people filled out a questionnaire on their history of insomnia and were observed for one night in a sleep laboratory. Men who reported a history of insomnia and slept for less than six hours in the lab were four times more likely to die in the follow-up period than those without insomnia who slept for six hours or more in the lab.
These findings require careful interpretation and do not prove that insomnia increases the risk of early death. Sleep duration was only objectively measured once, so it may not represent a typical sleep pattern or confirm that a person had insomnia. In addition, the study’s middle-aged participants were originally enrolled to investigate sleep disordered breathing, so they were not randomly selected and are unlikely to represent the general population.
In short, this research does not provide strong evidence that insomnia is linked to an early death, and it sheds no light on possible reasons behind a link. Further research is needed.
Where did the story come from?
The study was carried out by researchers from Pennsylvania State University College of Medicine in the US, and was funded by the National Institutes of Health. It was published in the peer-reviewed scientific journal Sleep.
The study was widely reported in the media. Few reports looked at the study’s limitations.
What kind of research was this?
This prospective cohort study examined whether insomnia and getting less than six hours a night affected the risk of death from any cause. The participants reported their insomnia themselves, and sleep duration was measured in a single night’s observation in a sleep laboratory.
This type of study, in which large groups of people are followed over time, are useful in assessing whether conditions or circumstances (in this case, insomnia and objectively measured sleep duration) are associated with later events (here, mortality). However, this cohort is limited in that it is a secondary analysis of a group of participants who were originally enrolled to investigate the age distribution of people with sleep disordered breathing.
The researchers point out that insomnia has never been linked to any serious medical disorders, such as cardiovascular problems. However, recent research has associated it as a risk factor for high blood pressure and diabetes. They speculate that severe insomnia is likely to be associated with higher mortality, saying this theory is supported by studies showing that insomniacs suffer from increased heart and metabolic rates and impaired heart rate variability.
The researchers say that previous findings relating to insomnia and mortality have been inconsistent. They point out that these studies relied only on self-reported sleep disturbance, did not measure sleep duration objectively, and did not always control for confounders. Here, they aimed to examine the association between insomnia and mortality while taking these factors into account.
What did the research involve?
The research was the secondary analysis of a larger study on sleep-disordered breathing. That larger study interviewed 16,583 people by telephone, asking them questions about their sleep habits. From this cohort, 741 men of average age 50, and 1,000 women of average age 47, agreed to take part in the sleep analysis study (representing 67.8% of men and 65.8% of women who had been asked to take part).. This selection was not random, and the researchers say that they had chosen a larger-than-usual proportion of people with a high BMI and who were at greater risk of sleep-disordered breathing.
All participants completed a comprehensive sleep history questionnaire and physical examination. Their sleep was evaluated for one night in the sleep laboratory, using polysomnography, a comprehensive recording of all biophysical changes that occur during sleep. They were then divided into two categories according to how long they had slept. Those who slept six hours or more were put in the normal sleep duration group, while those who slept less than six hours were in the short duration group.
On the same evening as the laboratory visit, the group also filled in a standardised questionnaire covering demographics, sleep-related questions (including questions on sleep disorders) and general health questions. The presence of insomnia was defined as insomnia that had lasted for at least one year.
The men in the study were followed up for 14 years, and the women for 10 years. People who died were identified using social security numbers matched to federal and state death record services. The possible association between insomnia, objectively measured sleep duration and the risk of mortality was assessed using standard statistical methods. The findings were adjusted to take account of possible confounders, such as age, race, education, body mass index, smoking, alcohol, depression and sleep-disordered breathing. The participants were also asked if they were being treated for diabetes or high blood pressure.
What were the basic results?
Overall, during the study period, 21% of men and 5% of women died. The main findings are as follows:
- In men, the risk of dying in the 14-year follow-up was increased in those who slept less than six hours in the lab and had also reported a history of insomnia, compared to men who had normal sleep duration and no insomnia. This analysis was adjusted for diabetes, high blood pressure and other potential confounders (OR 4.00, CI 1.14-13.99).
- Further analysis of these high-risk males (those reporting insomnia and with short sleep duration in the lab) revealed that men who also had diabetes or high blood pressure had the highest risk of death during follow-up (OR 7.17, CI 1.41-36.62) compared to men with no reported insomnia and normal sleep duration in the lab.
- Men with insomnia and short sleep duration who were not affected by diabetes or high blood pressure no longer had a significantly increased risk of death compared to males with ‘normal sleep’ (OR 1.45 CI 0.13-16.14) – i.e. diabetes and blood pressure modified the effect of insomnia upon mortality.
- There was no increased risk in men who reported insomnia but whose objectively measured sleep duration was six hours or more. Nor was there any increased risk in men who had not complained of insomnia, but whose sleep duration was less than six hours.
- Women had no association between insomnia, short sleep duration and a higher mortality.
How did the researchers interpret the results?
Men with chronic insomnia and objectively measured short sleep duration had a higher risk of dying early, say the researchers, independent of other factors associated with mortality. People who had diabetes or high blood pressure showed a far stronger association between insomnia and short sleep duration. They say that the diagnosis and treatment of insomnia should be targeted by public health policy.
This study has found that in middle-aged men, self-reported insomnia and objectively measured short sleep duration was associated with a greater risk of death during the 14-year follow-up period, compared with men who did not have insomnia or short sleep duration. However, these findings require careful interpretation and do not prove that insomnia increases the risk of early death:
- The study has an important limitation in that it is a secondary analysis of a study set up to assess the age distribution of people with sleep-disordered breathing. As such, the participants were not selected at random. All of them had a higher risk of sleep-disordered breathing, and the women had markedly higher BMIs. These factors may affect both mortality risk and insomnia. Therefore, the results should be interpreted with caution and cannot easily be generalised to the wider population.
- Sleep duration was only objectively measured in the laboratory once, at the start of the study, so the results may not have been typical or accurate. In men who reported a history of insomnia (self-reported sleep problems lasting at least a year) only those who slept for a shorter duration in the lab had an increased risk. However, a single night’s observation in this artificial environment does not necessarily ‘confirm’ that the person had insomnia. Those men who just reported a history of insomnia had no greater mortality risk than those who did not report sleep problems. Of note, the increased risk calculated for the men with insomnia and who slept for less than six hours in the lab had a wide confidence interval, which questions the reliability of this finding.
- The study found that men with diabetes or high blood pressure who suffered insomnia and slept for less than six hours in the lab were at an even greater risk of dying during follow-up than those without these conditions. Again, however, the very wide confidence intervals suggest the need for caution with these results.
- Although the researchers tried to adjust their findings for other factors that could have influenced mortality and sleep, it is possible that other confounding factors influenced the results. Insomnia may be related to numerous medical or psychological conditions that can also affect mortality risk.
To sum up, this research is not strong evidence that insomnia is linked to an early death, and it sheds no light on possible reasons behind a link. Further research is needed.