Work stress and risk of death in men and women with and without cardiometabolic disease: a multicohort study
Mika Kivimäki , Jaana Pentti , Jane E. Ferrie , G. David Batty , Solja T. Nyberg , Markus Jokela , Marianna Virtanen , Lars Alfredsson , Nico Dragano , Eleonor I. Fransson , Marcel Goldberg , Anders Knutsson , Markku Koskenvuo , Aki Koskinen , Anne Kouvonen , Ritva Luukkonen , Tuula Oksanen , Reiner Rugulies , Johannes Siegrist , Archana Singh-Manoux , Sakari Suominen , Töres Theorell , Ari Väänänen , Jussi Vahtera , Peter J. M. Westerholm , Hugo Westerlund , Marie Zins , Timo Strandberg , Andrew Steptoe , John Deanfield
AbstractBackground Although some cardiovascular disease prevention guidelines suggest a need to manage work stress in patients with established cardiometabolic disease, the evidence base for this recommendation is weak. We sought to clarify the status of stress as a risk factor in cardiometabolic disease by investigating the associations between work stress and mortality in men and women with and without pre-existing cardiometabolic disease. Methods In this multicohort study, we used data from seven cohort studies in the IPD-Work consortium, initiated between 1985 and 2002 in Finland, France, Sweden, and the UK, to examine the association between work stress and mortality. Work stress was denoted as job strain or effort–reward imbalance at work. We extracted individual-level data on prevalent cardiometabolic diseases (coronary heart disease, stroke, or diabetes [without differentiation by diabetes type]) at baseline. Work stressors, socioeconomic status, and conventional and lifestyle risk factors (systolic and diastolic blood pressure, total cholesterol, smoking status, BMI, physical activity, and alcohol consumption) were also assessed at baseline. Mortality data, including date and cause of death, were obtained from national death registries. We used Cox proportional hazards regression to study the associations of work stressors with mortality in men and women with and without cardiometabolic disease. Results We identified 102 633 individuals with 1 423 753 person-years at risk (mean follow-up 13·9 years [SD 3·9]), of whom 3441 had prevalent cardiometabolic disease at baseline and 3841 died during follow-up. In men with cardiometabolic disease, age-standardised mortality rates were substantially higher in people with job strain (149·8 per 10 000 person-years) than in those without (97·7 per 10 000 person-years; mortality difference 52·1 per 10 000 person-years; multivariable-adjusted hazard ratio [HR] 1·68, 95% CI 1·19–2·35). This mortality difference for job strain was almost as great as that for current smoking versus former smoking (78·1 per 10 000 person-years) and greater than those due to hypertension, high total cholesterol concentration, obesity, physical inactivity, and high alcohol consumption relative to the corresponding lower risk groups (mortality difference 5·9–44·0 per 10 000 person-years). Excess mortality associated with job strain was also noted in men with cardiometabolic disease who had achieved treatment targets, including groups with a healthy lifestyle (HR 2·01, 95% CI 1·18–3·43) and those with normal blood pressure and no dyslipidaemia (6·17, 1·74–21·9). In all women and in men without cardiometabolic disease, relative risk estimates for the work stress–mortality association were not significant, apart from effort–reward imbalance in men without cardiometabolic disease (mortality difference 6·6 per 10 000 person-years; multivariable-adjusted HR 1·22, 1·06–1·41). Interpretation In men with cardiometabolic disease, the contribution of job strain to risk of death was clinically significant and independent of conventional risk factors and their treatment, and measured lifestyle factors. Standard care targeting conventional risk factors is therefore unlikely to mitigate the mortality risk associated with job strain in this population.
|Journal series||Lancet Diabetes & Endocrinology, ISSN 2213-8587, (A 50 pkt)|
|Publication size in sheets||0.5|
|ASJC Classification||; ;|
|Publication indicators||: 2017 = 5.226; : 2017 = 19.313 (2) - 2017=20.231 (5)|
* presented citation count is obtained through Internet information analysis and it is close to the number calculated by the Publish or Perish system.