Background & Aims : This study examined the effects of education, occupation, social networks and psychological factors on all-cause mortality in a middle-aged Japanese cohort.
Methods : The cohort study (n : 11,565) was conducted from 1993 until 2011. The subjects were middle-aged males and females living in rural (n : 4,501 ; deaths : 669) and urban (n : 7,064 ; deaths : 1,190) areas in Gunma, Japan. Cox proportional hazards model was applied to investigate how various factors affect mortality risk.
Results : In the age-adjusted model, higher educated rural males had significantly lower risks (junior high school : 1.00 ; high school : 0.76 ; junior college : 0.57), though rural females had a U-shaped risk pattern with the lowest significant risk (0.63) in the high school group. Also, males who were non-married or non-participated or heavily drinking or never exercised or BMI less than 18.5 or having chronic diseases or depression had significantly higher mortality risks. Current smoking was only one universal factor that was still significantly affected increased risk of mortality across both areas and genders.
Conclusions : In the rural area, educational levels were related to mortality risk, but the risk patterns were different from genders. Males of poorer social networks had higher mortality risks.
View full abstract