Demographic Models of Health and Mortality at Both Extremes of the Lifespan

Julio E Romero Prieto, University of Pennsylvania


This dissertation consists of three essays. The first chapter proposes a model life table to investigate the human mortality at early ages. The model was estimated from the vital records, observing the experiences of 24 countries, which in some cases are at end of the nineteenth century and much of the twentieth century. Using few input values, the model predicts a mortality schedule for the first days, weeks, months, and years of life. Furthermore, the model is flexible to represent age patterns in conditions of either low or high mortality. Thus, the main application is as a method for indirect estimation, in contexts where vital records are incomplete, imperfect, or non-existent. In this direction, the second chapter takes advantage of the model to investigate the mortality patterns and the quality issues of the mortality estimates from self-reported data. To this end, a total of 252 Demographic and Health Surveys were analyzed in light of the predictions of the model, in order to identify particular characteristics of these populations. These comparisons lead to the conclusion that populations with high levels of mortality are more likely to show late patterns of under-five mortality. The model was also used to examine data quality issues regarding misreported ages at death. Particularly, this chapter proposes a simple solution to the problem of heaping at the age of 12 months and the underestimation of the infant mortality. The third chapter investigates the relationship between health status and survival expectations on a sample of mature adults aged 45+, who participated in the Malawi Longitudinal Study of Families and Health between 2006 and 2012. In particular, structural equation models were estimated assuming intertemporal relationships between physical health, mental health, and the formation of survival expectations. These models identify different pathways that have been discussed from theoretical and empirical approaches showing evidence of the concomitancy of physical and mental health issues, and the relevance of the expectations about life. This paper quantifies a significant impact of mental health on the prospective physical health and provides evidence on the differentiated adaptation pathways for men and women.

Subject Area


Recommended Citation

Romero Prieto, Julio E, "Demographic Models of Health and Mortality at Both Extremes of the Lifespan" (2017). Dissertations available from ProQuest. AAI10261137.