Analysis of the bereavement effect after the death of a spouse in the Amish: a population-based retrospective cohort study

Objective This study investigates the association between bereavement and the mortality of a surviving spouse among Amish couples. We hypothesised that the bereavement effect would be relatively small in the Amish due to the unusually cohesive social structure of the Amish that might attenuate the loss of spousal support. Design Population-based cohort study. Setting The USA. Participants 10892 Amish couples born during 1725-1900 located in Pennsylvania, Ohio and Indiana. All the participants are deceased. Outcome measures The survival time is age'; event is death'. Hazard ratios (HRs) of widowed individuals with respect to gender, age at widowhood, remarriage, the number of surviving children and time since bereavement. Results We observed HRs for widowhood ranging from 1.06 to 1.26 over the study period (nearly all differences significant at p<0.05). Mortality risks tended to be higher in men than in women and in younger compared with older bereaved spouses. There were significantly increased mortality risks in widows and widowers who did not remarry. We observed a higher number of surviving children to be associated with increased mortality in men and women. Mortality risk following bereavement was higher in the first 6months among men and women. Conclusions We conclude that bereavement effects remain apparent even in this socially cohesive Amish community. Remarriage is associated with a significant decrease in the mortality risk among Amish individuals. Contrary to results from previous studies, an increase in the number of surviving children was associated with decreased survival rate.