Childhood neighborhoods and cause-specific adult mortality in Sweden 1939-2015

The socioeconomic health gradient has widened since the mid-21st century, but the role of childhood neighborhoods remains underexplored. Most neighborhood studies on health are cross-sectional, and longitudinal research is lacking. We analyze how socioeconomic neighborhood conditions in childhood influence cause-specific deaths in adulthood. We use uniquely detailed geocoded longitudinal microdata for the Swedish town of Landskrona, 1939-1967, linked to Swedish national registers, 1968-2015. We measure neighborhood SES by social class and use dynamic sizes of individual neighborhoods. Cox proportional hazards models are employed to estimate the impact of neighbor’s social class in childhood (ages 1-17) on mortality in ages 40-69. We control for class origin, class in adulthood, schools, and physical neighborhood characteristics. The class of the nearby, same-age, childhood neighbors had a lasting effect on male all-cause and preventable, but not non-preventable, mortality. Men who grew up with having 10% more children from white-collar families as close-proximity neighbors had an 8% lower mortality risk due to preventable causes of death in adulthood. The mortality for women was not affected by their childhood neighbors, although both a lower adult class and class origin increased their mortality. Because preventable causes of death are linked to lifestyle factors, this study suggests that childhood neighborhood peers had a strong and lasting influence on the health behavior of men growing up before the health gradient was fully established. Hence, our applied life-course perspective on childhood neighborhoods is crucial to better understand the mortality differentials by SES.