Overtime doesn't pay?

“Britain's long working hours could be putting millions at risk of dementia” the Daily Mail has reported. The newspaper said that many workers regularly work over 55 hours a week, and that new research shows that long working hours could lead to poorer mental skills. The newspaper said the effect could be “as bad for the brain as smoking”.

This research surveyed civil servants on their working hours and lifestyle, comparing this data with cognitive (mental functioning) tests taken at the time and once more around five years later. In this second testing workers who had done the most overtime recorded slightly lower scores in two of five key brain function tests, those on reasoning and vocabulary. The researchers said “this study shows that long working hours may have a negative effect on cognitive performance in middle age”.

However, as only a short time had elapsed between data collections, this was not a long enough period to say that long working hours have a direct causal influence on cognitive function, let alone dementia. Further research would be needed to clear up this debate.

Where did the story come from?   

This article was written by Dr Marianna Virtanen from the Finnish Institute of Occupational Health in Helsinki and colleagues from University College London, France and Texas. Some authors were involved in the original Whitehall II study, which provided data for this research. The Whitehall II study was supported by grants from the British Medical Research Council, the US National Heart, Lung, and Blood Institute, the British Heart Foundation and other funding agencies.

The study was published in the peer-reviewed medical journal the American Journal of Epidemiology.

What kind of scientific study was this?   

This was an analysis of follow-up data from a prospective cohort study, looking for potential links between long working hours and cognitive function.

The researchers had access to data from a large-scale, prospective occupational cohort of British civil servants, the Whitehall II study. This original study included data on education, occupational position, physical health status, psychological and psychosocial factors, sleep problems and other health behaviours. The researchers in this subsequent study were able to use other data on brain function and psychological testing conducted on participants over a five-year follow-up period during the study.

Having access to this rich data source means that the researchers were able to take into account several factors that might confound or mediate any association between work hours and cognitive function. Adjusting their analysis in relation to these factors allowed them to focus more clearly on the links between working hours and brain function.

The Whitehall II study began in late 1985 and up to early 1988 recruited volunteers from all office staff aged 35–55 years working in 20 London-based civil service departments. The response rate in this original study was 73% (6,895 men and 3,413 women).

Since the start of the study there have been seven further data collection phases. In phase five (1997–1999) and phase seven (2002–2004) cognitive test scores and data on some of the other confounding factors were collected. It was this data on 2,214 participants (1,694 men and 520 women) who had completed these two phases that was included in this analysis. Participants’ cognitive ability was assessed through a range of tests in phases five and seven. These tests included tests of fluency, vocabulary, reasoning, comprehension and a 20-word recall test.

Working hours were determined at phase five (1997-1999) by asking two questions: ‘‘How many hours do you work per average week in your main job including work brought home?’’ and ‘‘How many hours do you work in an average week in your additional employment?’’. Participants were grouped by their answers into those who worked a total of 35–40 hours, those who worked a total of 41–55 hours and those who worked for more than 55 hours per week.

Complex statistical modelling was used to assess the relationship between hours worked and cognitive test results. During this analysis the researchers adjusted for several measures that might confound their results: age, sex, marital status, employment status during follow-up, occupational grade, education, income, physical health indicators, psychological distress, anxiety, sleep problems, health risk behaviours, social support, family stress and job strain.

What were the results of the study?   

A total of 853 (39%) participants reported working 35–40 hours per week, 1,180 (53%) reported 41–55 hours and 181 (8%) reported more than 55 hours of work per week.

Compared with employees who worked 35–40 hours, those who worked more than 55 hours were more likely to be male, married or cohabiting, have a higher occupational grade, have attended higher education, have higher income, experience more psychological distress, have shorter sleep, higher alcohol use and more social support. The researchers adjusted for these and other factors in their statistical analysis, and found three statistically significant results out of 10 comparisons reported.

These significant results were that employees working more than 55 hours had lower vocabulary scores at the first assessment and follow-up when compared with employees working 40 hours or less per week. At follow-up, they also had lower scores on the reasoning test.

No significant difference between groups was found in any other measures of cognitive function at follow-up.

What interpretations did the researchers draw from these results?   

The researchers say that this study shows that long working hours may have a negative effect on cognitive performance in middle age.

They also say that the findings can have clinical significance “as the 0.6- to 1.4-unit difference in aspects of cognitive functioning between employees working long hours and those working normal hours is similar in magnitude to that of smoking, a risk factor for dementia”.

What does the NHS Knowledge Service make of this study?    

There are two parts to the analysis reported. In the earlier cross-sectional analysis (where data on working hours were collected at the same time as the data on cognitive function), the researchers found a statistically significant difference of less than one unit in a score of reasoning that ranges from one to 33.  In the second part, where data on cognitive function were collected up to seven years later (average five years), there was a less than one point difference in the vocabulary scale and a less than two point difference in the 53-point reasoning scale.

There are limitations to interpreting this study as showing a causal link between hours of work and dementia:

• Most of the data analysed are cross-sectional or with only around five years between the measurement of work hours and cognitive function. A five-year interval is relatively short for looking for long-term cognitive impairment. This limits the ability to conclude that one factor necessarily follows the other. For example, people who worked longer hours might have had impaired cognitive function in the first analysis.
• The multiple adjustments for known risk factors of impaired cognitive function were necessary as the groups were not well balanced. Although the adjustments may remove the effects of some of the differences between the groups, the likelihood that other unknown differences exist between the groups limits the reliability of the results.
• The clinical significance of a few points’ change in these scores is not clear. Although the authors say “mild cognitive impairment predicts dementia and mortality”, it is a further step in logic to say that a two-point change in the reasoning scale after five years could also be linked to dementia later in life. This second link was not tested by the study.
• The study looked at data on civil servants in an office-based environment. The study’s findings might not be directly applicable to other types of worker.

Overall, this study has attempted to assess the link between working hours and cognitive impairment. However, to provide definitive answers for employers or employees would require a study with a longer period between the start of the study and collection of outcome data.