Findings in the Cornell university ambulatory blood pressure worksite study: A review

Purpose: 'Job strain' (defined as high demands and low control) has been previously associated with increased risk of cardiovascular disease (CVD) and, cross-sectionally, with hypertension and elevated ambulatory blood pressure (AmBP). Our longitudinal cohort study was designed to investigate the hypothesis that exposure to 'job strain' is causally related to increases in mean AmBP. Methods: The sample consists of 285 healthy male employees, aged 30-60 at initial recruitment, at 8 New York City worksites, 195 of whom were restudied 3 years after (Time 2) and 194 of whom were restudied 6 years after (Time 3) their initial participation. Mean systolic (SAmBP) and diastolic (DAmBP) ambulatory blood pressure at work, home and during sleep were computed from 24-hour recordings and diary entries specifying location. The relationship of 'job strain' to AmBP was examined cross-sectionally at each round of data collection. In addition, to take advantage of our information on 'job strain' status at each assessment and to evaluate the impact of changes in exposure, a 'job strain' change variable was constructed with four categories: those defined as having no 'job strain' at either Time 1 or Time 2 (N=138), those reporting 'job strain' at both times (N=15), and two groups which changed 'job strain' status. We repeated this analysis comparing Times 1 and 3 as well as Times 2 and 3. Multiple regression analysis was used to examine the cross-sectional associations of AmBP with 'job strain', as well as to predict 3 year and 6-year change in AmBP (from Time 1 to Time 2, Time 2 to Time 3 and Time 1 to Time 3) with 'job strain' change, controlling for age, body mass, race/ethnicity, smoking status, alcohol consumption, education, sodium and physical exertion level of the job. Results: Cross-sectional analyses at Time 1, Time 2 and Time 3 showed consistent significant effects of 'job strain' on AmBP, even though the overlap between the high strain groups at different Times was less than 50%. In cross-sectional analyses, subjects with 'job strain' had work SAmBP/DAmBP which were 5-7/3-5 mm Hg higher than subjects without 'job strain'. Those men facing chronic 'job strain', that is working in high strain jobs at both Time 1 and Time 2 (or Time 1 and Time 3) had work SAmBP/DAmBP on average, 10-12/6-8 mm Hg higher than those with no 'job strain' at both times. The two crossover groups had intermediate levels of blood pressure. Effect sizes for chronic three-year exposure to 'job strain' are larger than the estimated effect of aging 25 years or gaining 50 pounds in weight. In longitudinal analyses, subjects who changed from exposure to 'job strain' to no exposure three or six years later had a decrease in SAmBP/DAmBP of about 5/3 mm Hg. Conclusions: The previously reported cross-sectional association between 'job strain' and AmBP was replicated twice during follow-up periods. Repeated exposure to 'job strain' was associated with the highest levels of AmBP's, but not with any further increases in AmBP. However, changes in 'job strain' status predicted change in AmBP over three and six year periods, 'job strain' emerges as a consistent and substantial risk factor for AmBP in men. It appears that frequent changes in exposure status over time to work stress among working men can obscure associations between stress and outcome and should be taken into account in designing studies to assess the impact of work stress on CVD.