Cutler and his colleagues consider how past trends in health behaviors (like the long term decline in U.S. smoking rates) have affected longevity, and what current trends (like rising obesity rates) portend for future health.
In this paper, we consider what has happened to the population's health behaviors over time, and consider various scenarios for trends in the future.
Past trends in behavioral risk factors have not been in a common direction. Some measures of population risk have improved markedly, while others have deteriorated. Smoking rates have fallen by more than a third since 1960 (Anonymous, 1999) and alcohol consumption has declined by 20 percent since 1980 (Lakins, Williams, and Yi, 2006), both leading to better health.
Demographically, the population is better educated, and better educated people live longer than less educated people (Elo and Preston, 1996). On the other hand, obesity rates have doubled in the past two decades (Flegal et al., 2002) and diabetes has increased as a result (Gregg et al., 2005). Further, the population has a higher share of minority groups, for whom life expectancy is lower. The net impact of these risk factor trends on population health expectations is uncertain (Preston, 2005).
Our analysis has two parts. We start by aggregating these different health trends into a single measure of population risk. We focus on the most common risk factors: smoking, drinking, obesity, hypertension, high cholesterol, and diabetes. We weight the different risk factors by their impact on predicted 10 year mortality, as determined through multiple regression analysis. We show that overall health trends in the past three decades have improved markedly. For the entire population aged 25 and older, the age adjusted probability of dying in 10 years, conditional on the same level of medical care, fell from 9.8 percent in the early 1970s to 8.4 percent around 2000, a 14 percent reduction. The largest contributors to this trend were reductions in smoking and improved blood pressure control.
The second part of our analysis considers the impact of a continuation of future trends. Our conclusions here are not as rosy. We show that if current obesity trends continue, the population mortality risk could increase, even with continued reductions in smoking. We estimate that about a third of the past gains would be reversed within 20 years. The increase in obesity is the proximate cause of this. But even given the increase in obesity, the health impact would be substantially blunted if more people took medication to control blood pressure, cholesterol, and diabetes.