Residence in mountainous compared with lowland areas in relation to total and coronary mortality. A study in rural Greece.

STUDY OBJECTIVE To investigate the association of residence in mountainous or lowland areas with total and coronary mortality, in a cohort with 15 years of follow up. DESIGN AND SETTING Prospective study, based on the adult population of two lowland and one mountainous village in rural Greece. Baseline measurements on sociodemographic, lifestyle, somatometric, clinical, and biochemical variables were recorded in 1981 during a health survey and total and coronary deaths were ascertained on the basis of death certificates up to 1996. PARTICIPANTS 1198 men and women, who had participated in the 1981 survey. Analyses are based on 504 men and 646 women with complete data. MAIN RESULTS There were 150 deaths among men and 140 among women (coronary deaths: 34 and 33 respectively). In multivariate analysis with Cox regression, after adjustment for age, education, body weight, smoking, alcohol consumption, systolic blood pressure, serum total cholesterol, blood glucose, serum triglycerides, and serum uric acid, total and coronary mortality were lower for residents of the mountainous village in comparison with residents of the lowland villages (hazard ratios (95% confidence intervals) for men and women, respectively: total mortality, 0.57 (0.38 to 0.84) and 0.69 (0.47 to 1.02); coronary mortality, 0.39 (0.16 to 0.98) and 0.46 (0.20 to 1.05)). CONCLUSIONS Residence in mountainous areas seems to have a "protective effect" from total and coronary mortality. Increased physical activity from walking on rugged terrains under conditions of moderate hypoxia among the mountain residents could explain these findings.