Volume 31 - Article 31 | Pages 941–974  

HealthPaths: Using functional health trajectories to quantify the relative importance of selected health determinants

By Michael Wolfson, Geoff Rowe

Abstract

Background: It is widely known that that there is a significant dispersion in health status, as well as a strong correlation between health status and socio-economic status. But considerable uncertainty remains as to the quantitative importance of the various explanatory factors typically cited in this context. As health status is intrinsically a reflection of co-evolving dynamic processes, it is important to take a lifetime perspective when seeking to understand its determinants. The "bottom line" measure of overall population health is, though, health-adjusted life expectancy (HALE), which is an aggregation of individuals' health-adjusted life lengths (HALLs).

Objective: In an exploratory mode, we intend to provide a realistic assessment of the relative importance of selected health determinants of HALE.

Methods: This paper first draws on very detailed estimates of the covariates of vector-valued functional health trajectories, using the National Population Health Survey (Statistics Canada). We then use longitudinal microsimulation to draw out their implications by synthesising first a realistic base case -- specifically, a representative longitudinal population sample -- and then a series of exploratory counterfactual populations. Comparisons between and among counterfactuals and the base case are then used to estimate the quantitative importance of various factors in accounting for HALE.

Results: Several surprising results emerged. Of the four risk factors explicitly examined, obesity had the smallest impacts on HALE: moving from the fifth to the 95th percentiles of BMI increased HALE 1.5 and 2.5 years for men and women, respectively. Eliminating smoking increased HALE by five and four years, while moving from the lowest to the highest levels of education had similar effects of about five years for both men and women. Perhaps surprisingly, moving from the fifth to the 95th percentiles of the psycho-social factor, Antonovsky's sense of coherence, had the greatest impact on estimated HALE, with increases of 5.5 to six years.

Conclusions: While the analysis is exploratory, the results show the analytical potential of tightly coupling detailed empirical analysis of complex longitudinal data with microsimulation modelling. Substantively, the results of the analysis contribute to the evidence that the widespread public health preoccupation with the "obesity epidemic" is over-stated, while psycho-social and mental health problems continue to be seriously under-emphasised.

Author's Affiliation

Other articles by the same author/authors in Demographic Research

US regional and national cause-specific mortality and trends in income inequality: descriptive findings
Special Collection 2 - Article 8

Most recent similar articles in Demographic Research

The contributions of stochastic demography and social inequality to lifespan variability
Volume 49 - Article 13    | Keywords: heterogeneity, inequality, lifetime reproduction, longevity, stochasticity, variance decomposition

A test of the predictive validity of relative versus absolute income for self-reported health and well-being in the United States
Volume 48 - Article 26    | Keywords: absolute income, health, inequality, measurement, relative income, well-being

Female sterilization in the life course: Understanding trends and differentials in early sterilization
Volume 47 - Article 18    | Keywords: contraception, female sterilization, fertility, inequality, reproduction, sterilization regret, United States of America

Exploring the mortality advantage of Jewish neighbourhoods in mid-19th century Amsterdam
Volume 46 - Article 25    | Keywords: 19th century, Amsterdam, cause-specific mortality, historical demography, inequality, Jewish neighborhoods

Socioeconomic differentials in fertility in South Korea
Volume 44 - Article 39    | Keywords: education, family, fertility, gender, housing, inequality, Korea, labor market, nonstandard work, parity