Monday July 28 2008
Research has found that “men who suffer sleepless nights run double the risk of contracting diabetes”, the Daily Mail reported. It said the 10-year study of 5,000 middle-aged Swedish men and women, found that men who suffered the most stress were more than twice as likely to develop type 2 diabetes. The link remained when risk factors such as smoking and body mass index (BMI) were taken into account. It was suggested that the link only applies to men because they “bottle up their feelings more than women”.
This study found a link between levels of self-reported psychological distress symptoms and the development of type 2 diabetes 10 years later. However, it has some limitations, including the fact that diet, which affects diabetes risk and may also be related to stress, was not taken into account. This suggestion of a link is not new, and the more interesting finding here is that it did not find it in women. Given that other studies have also found a link in women, more research that accounts for important confounders such as diet is needed before it is possible to fully understand what is happening.
Where did the story come from?
Dr Anna-Karin Eriksson and colleagues from the Karolinska Institutet carried out the study. The research was funded by Stockholm County Council, the Swedish Council of Working Life and Social Research, Novo Nordisk Scandinavia and Glaxo Smithkline in Sweden. The study was published in the peer-reviewed medical journal Diabetic Medicine.
What kind of scientific study was this?
This cohort study was part of the Stockholm Diabetes Prevention Program. Participants to this large study were recruited by sending an invitation to all men born between 1938 and 1957, and women born between 1942 and 1961 in five municipalities in Stockholm. The questionnaire asked about the participant’s country of birth and if they, or any members of their family, had diabetes. All Swedish-born people who responded and who did not have diabetes themselves, but who reported a family history of the condition, were invited to attend a health examination. A random sample of people without a family history of diabetes (who were matched to the other group by age and municipality) were also invited to attend a health examination.
During this examination, the participants had an oral glucose tolerance test (to determine whether they had impaired glucose metabolism), and their body measurements were taken. They also responded to a questionnaire which assessed a variety of lifestyle factors, including smoking, physical activity, and socioeconomic status. From these assessments, 3128 men and 4821 women had information available for the study. Over the course of the study, a diabetes prevention program was implemented in three of the five municipalities. The programme encouraged increased physical activity, dietary improvement and smoking reduction.
Eight to 10 years after the study began, the participants were invited to attend a follow-up health examination. The researchers excluded anybody who already had diabetes when they originally enrolled, had missing data, or who had left Stockholm or died in the intervening period. This left a total of 2383 men and 3329 women for follow-up (76% and 69% of the original study group). At follow-up, the participants were asked if they had been diagnosed with diabetes since the first health exam, and those who had not were given the oral glucose tolerance test again. People who had impaired fasting glucose, impaired glucose tolerance or both, were identified as having ‘pre-diabetes’.
All participants were asked questions about their lifestyle and had their BMI measured. They also completed a questionnaire to assess ‘psychological distress’. This asked them whether they had experienced any of the following symptoms in the previous 12 months: insomnia, anxiety, apathy, depression or fatigue. The frequency that the symptom was experienced was given a score of one to four, according to whether it had been experienced ‘never’, ‘occasionally’, ‘sometimes’, or ‘frequently’. Depending on their total score, the participants were then divided into four groups, each group containing 25% of the participants. The first group contained those who had ‘never’ experienced psychological distress in the previous 12 months; the second contained those who had ‘occasionally’ experienced symptoms, and so on. Again, there was missing data at this follow-up stage, and this further reduced the final group to 2127 men and 3100 women for analysis (68% and 69% of the original baseline population).
The researchers then assessed the risk of ‘pre-diabetes’ and type 2 diabetes according to symptoms of psychological distress in the previous 12 months, taking into account other factors (age, smoking, socioeconomic status, activity, family history of diabetes, etc). For this analysis, they combined the groups of people who had ‘occasionally’ and ‘sometimes’ experienced symptoms. They also took into account the effects of the intervention that was being delivered.
What were the results of the study?
People who were included in the follow-up assessment were less likely to be smokers than those who were not included. In addition, women who were not included were more likely to be obese, have low socioeconomic status, and be psychologically distressed. They were also less likely to exercise. At follow up, 103 of the 2127 men in the analysis had developed type 2 diabetes, as had 57 of the 3100 women.
A family history of diabetes, smoking, low physical activity and low socioeconomic status were all more common in people with higher psychological distress than those with lower. When these factors were taken into account, men who had experienced psychological distress most frequently were more than twice as likely to develop type 2 diabetes than the men who had suffered it least frequently. This increase in risk with distress levels was not apparent in women. Risk of pre-diabetes was increased in both men and women with increased distress.
What interpretations did the researchers draw from these results?
The researchers concluded that self-reported psychological distress (including symptoms of anxiety, apathy, depression, fatigue and insomnia) was associated with development of type 2 diabetes in Swedish middle-aged men. This did not apply to women, although there was an association between distress and pre-diabetes.
What does the NHS Knowledge Service make of this study?
This cohort study suggests a link between stress and diabetes in men, but not in women. This is in contrast to other studies, which have found a link between depressive symptoms and type 2 diabetes in both men and women. When interpreting the results of this kind of study, it is important to keep in mind any limitations they may have. The researchers highlight some of these:
- The study relies on a self-report of psychological distress, using an instrument that was not fully validated (i.e. a questionnaire that has not been fully tested in other populations). It is possible that men and women report their distress symptoms differently. Men may be less likely to report that they are distressed unless the symptoms are very severe. Women on the other hand may over-report symptoms. If this stereotype were true, the dilution of the effect of distress in women and the concentration of it in men may be responsible for the discrepant results between the genders.
- Importantly, the study related the participants’ stress levels when they were first enrolled with the development of diabetes 10 years later. It does not consider any changes in the participants’ levels of stress during this follow-up period.
- There were significant differences between those who participated in follow-up and those who declined, with the non-participants generally having more risk factors for diabetes. Had these people been included, the results may have been different.
- Although the researchers accounted for some factors that can affect the link between distress and diabetes (age, physical activity, socioeconomic status), they did not account for diet. This is an important factor, and differences in diet or dietary responses to stress between men and women may account for the results seen here. It is unusual to see an effect on pre-diabetes, but not on diabetes itself. Clearly more research is needed to unpick this.
The suggestion that depression is linked to the development of diabetes is not new, and the more interesting finding from this study is the absence of this link in women. Given that other studies have found a link between depression and risk of diabetes in women, more research is needed to achieve a better understanding.
Sir Muir Gray adds...
We know that living in a stressful environment, in severe poverty for example, increases the risk of heart disease more than can be explained by higher levels of conventional risk factors such as smoking. The environment, both physical and social, affects your health and steps to improve the environment in which people live need to be taken to complement the provision of good information.