ABC | Volume 112, Nº3, March 2019

Original Article Muniz et al Ideal cardiovascular health and job strain Arq Bras Cardiol. 2019; 112(3):260-268 Figure 2 – Bar graph with percentages of intermediate and poor cardiovascular health according to tob strain categories are low strain (low demand and high control), active (high demand and high control), passive (low demand and low control), and high strain (high demand and low control). Ideal cardiovascular health is not represented because none individual was in this group. 100 80 60 40 20 0 Percentage 89 90 91 93 11 10 9 7 Poor Intermediate Low strain Active Passive High strain Table 3 – Association between job demand, control, or strain with poor cardiovascular health Poor Cardiovascular Health (Crude) Poor Cardiovascular Health (Adjusted*) n (%) Odds Ratio (95% CI) p Odds Ratio (95% CI) p Isolated parameters High demand 180 (40) 1.24 (0.64-2.39) 0.52 1.49 (0.76-2.94) 0.24 Low Control 229 (49) 1.29 (0.69-2.40) 0.43 1.09 (0.54-2.20) 0.81 Combined parameters Low strain 144 (32) Ref - Ref - Active 93 (21) 1.17 (0.49-2.76) 0.73 1.21 (0.49-2.97) 0.68 Passive 128 (28) 1.21 (0.55-2.66) 0.64 0.91 (0.38-2.18) 0.84 High strain 86 (19) 1.67 (0.63-4.44) 0.31 1.79 (0.65-4.92) 0.26 CI: confidence interval; *Adjusted for age, sex, income, education and profession. countries might influence to change the impact of job strain in CV health. Other authors explored individual CV health factors and found association between high job strain and physical inactivity, 15,16 diabetes, smoking and obesity. 16 In the current study, diet pattern was associated with job strain, perhaps due to the fact that people with higher demand and lower control usually have less time to prepare or buy healthy food. We also found an association between job strain and obesity, which may be a consequence of a poor diet; it is reasonable to hypothesize that a poor diet is a mediator factor for the association between job strain and obesity. However, we did not found mediator effect for this association (data not shown). In addition, the control‑demand model was built to evaluate psychosocial factors that affect mental health. 7 Participants with job strain are more prone to have eating disorders resulting in weight gain, which may explain our findings. In our questionnaires, we did not explore mental health disorders to add on this discussion. In the current study, physical inactivity was not associated with job strain, which is different from a previous publication that found it as the only risk factor associated with job strain. 15 Our data showed a trend towards worse CV health factors and job strain, albeit without statistical significance. This is likely due to the lack of sample power given the relatively small sample size. In addition, it is also possible that the busiest and healthiest employees may have declined to participate in the study, introducing a bias towards the null. On the other hand, job stress might determine less influence in CV health in an academic environment in the Amazon region because employees would work in a less competitive behavior. This study has some limitations. Data on diabetes and hypercholesterolemia were self-reported, which could introducemisclassification bias. On the other hand, self‑reported questionnaires have low sensitivity for detecting these 264