A healthy row at work?

A blazing row with your boss “may be good for your heart”, according to the Daily Mail. The newspaper also said that male workers who do not complain about unfair treatment double their risk of a heart attack.

The news is based on Swedish research that suggests there is a link between passive behaviour during workplace conflict and a risk of heart disease. However, the research has a number of limitations, including using very simple methods to assess the impact of complex factors, such as smoking and drinking. The study was also small and did not assess important factors including diet.

These limitations mean that this study cannot prove that a passive way of coping with conflict at work increases the risk of heart disease, nor does the study identify the best coping style. It is not advisable to shout at your boss on the basis of this research (even if you are right).

Where did the story come from?

This research was conducted by Dr Costanze Leineweber from the Stress Research Institute at the University of Stockholm and colleagues from other research centres in Sweden and the UK. The research was funded by the Swedish Council for Working Life and Social Research and the Academy of Finland. The study was published in the peer-reviewed Journal of Epidemiology and Community Health.

The story was covered by the Daily Mail, Daily Express and The Independent. The Express and Mail did not mention any of the limitations to this study, while The Independent said that no conclusions can be drawn about women.

What kind of research was this?

This was an analysis of data collected as part of the WOLF Stockholm study, a larger prospective cohort study looking at the health of people aged 19 to 70 working in Stockholm. This subanalysis of the WOLF study looked at whether using “covert coping” to deal with unfair treatment at work affects the risk of developing heart disease. The researchers defined covert coping as a person not showing that they feel unfairly treated.

This type of study (a prospective cohort) is the best way of looking at factors that cannot be controlled by researchers. However, the study must still be conducted in a careful way and should take into account factors that may affect results and the differences between comparison groups.

What did the research involve?

Between 1992 and 1995, the WOLF cohort study enrolled people who worked in the Stockholm area, assessing a number of factors including employees’ coping styles. This subsequent study followed participants until 2003, using national registers to identify any participants who had been hospitalised due to a heart attack or had died from heart disease.

Assessments of coping style involved a questionnaire that asked how participants usually reacted to unfair treatment or conflict within the workplace, both from superiors and workmates. The participants indicated how often they felt they used covert coping techniques or experienced negative effects that could be associated with these techniques including letting things pass without saying anything, going away, feeling bad (e.g. developing a headache or stomach ache), and getting into a bad mood at home. Their answers were used to assign them a covert coping score, and to split them into low (bottom 25%), high (top 25%) or medium (remaining 50%) scoring groups.

The current analysis only looked at the 2,755 male participants (average age 41.5 years) who had not been hospitalised for a heart attack before the study, and for whom complete data were available. The researchers looked at whether there was a relationship between how the participants coped with unfair treatment at work and their risk of heart attack or death from heart disease. They took into account factors that could affect results (confounders), such as the men’s age, socioeconomic factors (e.g. education), risk behaviours (e.g. smoking and alcohol problems), job strain including recent workplace conflicts, and biological risk factors such as diabetes, blood pressure, BMI and blood cholesterol.

Although it is appropriate to take into account factors that could affect results, this study assessed many of them using simple yes or no questions (e.g. “Have you sought help in the last 10 years because of drinking problems”, “Have you experienced conflicts in the workplace in the past 12 months”, and smoking status). Using such simple analysis methods when adjusting for these factors may not fully remove their influence. Relying on hospital records to identify heart disease based only on hospitalisation for heart attack or death from heart disease may miss some people with heart disease.

What were the basic results?

Men with more covert coping behaviour were more likely to be older, have lower income and poorer education, and less likely to have supervisory status in their job. There were also differences in their perceived job demands and they felt they had less ability to make decisions at work.

During the study, 47 men were hospitalised for a heart attack or died from heart disease. The researchers found that the more the men used covert coping behaviours and had related negative effects, the greater their risk of being hospitalised for a heart attack or dying of heart disease.

The researchers then performed their analyses having adjusted for all their measured confounders. They found that there was a just-significant association when comparing only men with high covert coping scores and men with low scores (hazard ratio 2.29, 95% confidence interval 1.00 to 5.29).

When analyses were restricted to just the covert coping behaviours (silently letting things pass or going away), the association between increasing behavioural score and risk of heart disease was significant. It remained significant even after adjusting for all potential confounders.

How did the researchers interpret the results?

The researchers concluded that their study showed that “covert coping is strongly related to increased risk of [hospitalisation or death due to] cardiovascular disease”.

Conclusion

This study suggests that there is a link between coping style at work and risk of heart disease. However, the study has a number of limitations:

• The results seen could be due to the influence of confounders. Although the researchers tried to take into account factors that could affect results, many of these factors were assessed using simple yes or no questions or through the participants’ self-reports. Quantitative factors (e.g. blood pressure, cholesterol and BMI) were measured, but only on one occasion. Using such simple analyses to take into account and adjust for these factors may not accurately measure their impact or fully remove their effects.
• Other unmeasured factors could also have a confounding effect. These include diet, depression or anxiety.
• This study identified heart disease only through records of hospitalisation for heart attack or death from heart disease. This may have missed some people with heart disease.
• Results of this all-male study may not apply to women.
• The study was relatively small and had a relatively short follow-up. Only 47 men had had heart disease events by the end of the study. This small number of events reduces the reliability of the results.
• The study carried out multiple statistical tests, which increases the possibility that a significant difference will be found by chance.
• The authors note that they only decided to analyse the two coping behaviours (silently letting things pass or going away) separately to the effects (feeling bad or getting into a bad temper at home) after they saw the results of an analysis of each item individually. They say that these results should be treated with caution and need confirmation from other studies.
• The authors also note that their study does not suggest what a healthy coping strategy might be, and say that they found no association between the open (“active”) coping strategies they assessed and heart attacks or cardiac death.

Overall, this study by itself does not provide robust evidence that a covert coping style directly increases the risk of heart disease. Its results will need to be assessed in the light of other research.