Essays on mortality and the life course
Philadelphia, University of Pennsylvania (2009)
This dissertation explores determinants of mortality from the life course perspective. Chapter 1 provides an introduction. Chapter 2 analyzes how mortality conditions early in life affect cohort-level later mortality, and how these lagged influences compare with the effects of period mortality conditions. I use mortality data covering six countries and 1751-1915 cohorts and find that variation in adult mortality is mainly attributable to period conditions, suggesting that the importance of cohort’s early life mortality conditions on later mortality is limited.
Chapter 3 analyzes the effects of macro conditions at birth on mortality for Finnish 1900-1950 cohorts using individual-level data with 1980-2006 follow-up. I find that those born in times of above-trend stillbirth rates may have decreased adult mortality due to selection or acquired immunity. Contrary to research analyzing 19th to early 20th century cohorts, I find no evidence that being born in recession would increase adult mortality. Consistent with prior research I find that late spring-early summer babies have higher adult mortality, especially cardiovascular mortality, than late fall-early winter babies. The effect is smaller for richer versus poorer areas and for cohorts born in booms versus recessions. These gradients and the declining trends in cardiovascular mortality suggest that the month of birth effect may attenuate over time.
Chapter 4 focuses on adult factors and analyzes how weight change affects mortality. While extremes of weight status are known to affect mortality, it is not well known how weight changes affect mortality or how the relationship is modified by initial weight status. I find that weight losses are associated with excess mortality unless initial weight status is in the obese range, and that weight gains are not associated with excess mortality unless the gain is large and initial weight status is well in the obese range. These findings suggest that the potential benefits of a lower weight status may be offset by the negative effects associated with weight loss. The magnitude of the effects is larger than any of the early life effects estimated in Chapters 2-3, again suggesting that for adult mortality, period conditions may matter more than early life factors.