Friday October 4 2013
Regularly getting around 7-9 hours sleep is recommended
“People who get more than 10 hours a night have an increased risk of heart disease, diabetes and obesity,” the Mail Online warns. The study this news is based on also found that those who don’t get enough sleep have an increased risk of disease.
The study in question used survey data, collected via telephone, from more than 50,000 middle aged and older adults from 14 US states. The survey included questions on whether the person had ever been told they had heart disease, stroke or diabetes and how many hours sleep they normally got.
The researchers found either sleeping more or less than the recommended amount (seven to nine hours) was associated with increased likelihood of having these three chronic diseases.
A limitation of this study is its design; it was a cross sectional study where data is gathered at a single point in time. This means it cannot show a direct cause and effect relationship between sleep and disease risk. For example, it could be the case that the symptoms of heart disease were causing some people to sleep more, rather than sleeping more leading to heart disease.
The study also failed to assess the various other factors that could influence both chronic disease risk and sleep history, such as lifestyle (for example, smoking, alcohol, physical activity and diet), family history, and other diagnosed physical and mental health illness.
Overall, the study supports current recommendations on optimal sleep duration, but does not prove that less than or more than this directly causes chronic disease. So, occasionally having a long snooze is probably not something you should lose any sleep over.
Where did the story come from?
The study was carried out by researchers from the Centers for Disease Control and Prevention, Atlanta, US, and received no external funding.
The study was published in the peer-reviewed medical journal Sleep.
The Mail Online accurately reports the main findings of this study but does not discuss its inherent limitations – that it cannot prove any direct cause and effect relationship between sleep duration and disease risk.
What kind of research was this?
This was a cross sectional study which used survey data collected from more than 50,000 middle aged and older adults from 14 US states. The data examined their health and lifestyle factors, and the researchers aimed to look at the relationship between sleep duration, heart disease and diabetes, and to see how this relationship was influenced by obesity and mental health.
The researchers suggested that short sleep of six or fewer hours per night, due to our work and lifestyles, may be associated with several chronic diseases, though the underlying mechanisms are poorly understood. The researchers’ theory is that short sleep may influence our metabolism and insulin regulation and increase risk of weight gain. However, these are only theories.
The main difficulty with this study design is that it is cross sectional so cannot prove cause and effect and say that it is the duration of sleep that is directly causing the risk of these diseases. A multitude of biological, health and lifestyle factors may be confounding the relationship and having an influence on both a person’s sleep duration and their risk of the chronic diseases studied.
What did the research involve?
The research used data from the 2010 Behavioural Risk Factor Surveillance System survey, which uses random-digit dialling to survey people in all 50 US states. The overall response rate in 2010 was 52.7% of those invited to participate. In addition to interviewer-administered questionnaires about health-related behaviours and chronic diseases, 14 of the states surveyed in 2010 also completed the optional sleep module.
Presence of chronic disease was assessed through an affirmative ‘yes’ response to the question of whether they had ever been told by a health professional that they had a history of coronary heart disease (such as heart attack or angina) stroke or diabetes. People who said ‘don’t know’ or ‘not sure’ were classed as not having the conditions.
People who also said they had pre-diabetes or borderline diabetes (raised blood glucose but not meeting diagnostic criteria for diabetes) were not classed as having diabetes.
Because of the low prevalence of these diseases among adults younger than 44 years old, the researchers restricted their study to adults aged 45 years or older.
Sleep duration was ascertained by asking ‘On average how many hours of sleep do you get in a 24 hour period?’ Responses were rounded to the nearest hour. The optimal amount of sleep recommended varies by different organisations, but tends to be either seven to eight or seven to nine hours a night for an adult. Therefore the researchers considered short sleep duration to be six or fewer hours, and long duration to be 10 or more hours a night.
When analysing the relationship between sleep duration and the chronic diseases assessed, researchers took into account assessed variables of age, ethnicity, education, body mass index (BMI) (calculated from self-reported height and weight), and ‘frequent mental distress’ (FMD).
FMD was assessed by asking participants ‘about your mental health, which includes stress, depression and problems with emotions, for how many days during the past 30 was your mental health not good?’
Those who answered 14 or more days to this question were defined as having FMD.
What were the basic results?
Complete 2010 survey data was available for 54,269 adults aged 45 or older in the 14 states. A third of these people were 65 years or older, half were women and three quarters were of white ethnicity.
Around a third (31.1%) of participants reported sleeping for six or fewer hours each night, while only 4.1% slept for 10 or more hours a night.
Prevalence of the chronic diseases among all participants was:
- coronary heart disease: 10.9%
- stroke: 4.3%
- diabetes: 13.2%
Just under a third (28.8%) of participants were obese and 9.7% were defined as having FMD.
Compared to those having the optimal seven to nine hours sleep a night, both shorter duration and longer duration of sleep were associated with significantly higher prevalence of all three chronic diseases, FMD and obesity. The significant associations remained when adjusting for sex, age, ethnicity and education. The size of the risk association with the three diseases altered slightly but remained significant when adjusting separately for obesity, and then for FMD, though no model adjusted for both of these factors at the same time.
How did the researchers interpret the results?
The researchers conclude that their study demonstrates that compared with an optimal sleep duration of seven to nine hours per day, both shorter (six or fewer hours) and longer duration (10 or more hours) were associated with significantly increased risk of coronary heart disease, stroke and diabetes among adults aged 45 years and older.
This 2010 survey data from middle aged and older adults from 14 US states suggests an association between shorter and longer than optimal sleep duration and three chronic diseases. The optimal amount of sleep recommended varies by different organisations, but tends to be either seven to eight or seven to nine hours a night for an adult.
However, though the study benefits from its large sample size of over 50,000 adults it has significant limitations.
Cross sectional study design
Most importantly, the cross sectional study design which has assessed sleep duration and disease presence at the same time cannot prove cause and effect. It is not possible to say whether the shorter or longer sleep preceded or followed the onset of these conditions.
Self reported responses
All responses were self-reported. This included both the presence of diseases (which were not confirmed by medical records), sleep duration (which for many people may only be an estimate and may not remain the same all the time), and obesity (assessed though self-reported height and weight, which may be inaccurate).
Likely influence of confounding factors
It is possible that if a true relationship exists between sleep duration and these three chronic diseases, it is not a direct effect of sleep duration but is being influenced by confounding from other biological, health and lifestyle factors. The main factors that the researchers considered as potential confounders (aside from sex, age, ethnicity and education) were obesity and their measure of ‘frequent mental distress’.
As stated, obesity was from self-reported measures and may not be accurate, and similarly the researchers’ method of assessing FMD by a single question may not give a reliable indication of the person’s psychological health.
The researchers adjusted their analyses for obesity and FMD independently, though not together, but did not, or could not, measure the extent of other factors that may be confounding the relationship – for example, other lifestyle factors such as smoking, diet, alcohol and physical activity, family history, and presence of other diagnosed physical or mental health illnesses.
Possible selection bias
As the survey was conducted via landline telephone it may have been prone to a possible selection bias. For example, people on low incomes who cannot afford a telephone connection, people in institutions, or people with significant health problems who could not answer the telephone, would have been excluded.
And while this is a large sample size, it is only representative of middle aged to elderly adults in only 14 US states.
Overall the study supports current recommendations on optimal sleep duration, but does not prove that less than or more than this directly causes chronic disease.
Occasionally sleeping a few hours more or less a night is probably not going to lead to any problems. But if you have persistent pattern of over or under sleeping you should contact your GP for advice.
Analysis by Bazian. Edited by NHS Choices. Follow Behind the Headlines on Twitter.