196Association of longitudinal lifestyle patterns with diabetes risk among sedentary and not-sedentary workers

Abstract

Abstract Background Evidence from observational studies is limited regarding longitudinal patterns of lifestyles and diabetes onset. In occupational settings, physical activity at work is an important factor to consider when implementing health promotion. Methods This is a cohort study among 25,646 workers without diabetes in Japan. We calculated lifestyle scores using 5 low-risk lifestyle factors: no-smoking, avoiding heavy alcohol use, engaging exercise, avoiding sleep deprivation, and keeping healthy weight, yielding the total score of 0 to 5 each year. We identified lifestyle patterns over 3 years by group-based trajectory modelling. Diabetes was defined by blood glucose, HbA1c, or self-report. Hazard ratios (95% confidence intervals) for incident diabetes were calculated using Cox regression. Results We detected five lifestyle patterns from a whole sample. During a mean follow-up of 6.6 years, 2,223 developed diabetes. Overall, healthier lifestyles were associated with a lower diabetes risk. Among sedentary workers, the multivariable-adjusted hazard ratios (95% confidence intervals) were 1.00 (reference) for persistently unhealthy pattern, 0.70 (0.60, 0.80) for persistently moderately healthy pattern, 0.45 (0.38, 0.53) for persistently mostly healthy pattern, 0.38 (0.28, 0.52) for improved to completely healthy pattern, and 0.24 (0.17, 0.32) for persistently completely healthy pattern. Similar results were obtained among not-sedentary workers. Conclusions Healthier lifestyles were associated with a lower risk of developing diabetes among participants who were sedentary at work and those who were not sedentary at work, respectively. Key messages Adherence to healthy lifestyles is important to prevent diabetes for working adults regardless of occupational physical activity level.