Is the U.S. Population Behaving Healthier?
" Reduced smoking, better control of medical risk factors such as hypertension and cholesterol, and better education among the older population have been more important for mortality than the substantial increase in obesity."
Knowing whether health behaviors are improving over time is important in forecasting medical costs. And, a population that behaves in a healthier way will have a higher quality of life than one with a more adverse behavioral profile, even given length of life. In Is the U.S. Population Behaving Healthier? (NBER Working Paper No. 13013), authors David Cutler, Edward Glaeser, and Allison Rosen consider what has happened to the population's health behaviors over time and what the future may hold.
They find that the impact on longevity of trends in health behavior has not been uniform across different behaviors over the past three decades. For example, while fewer people smoke than used to, more people are obese. Examining these factors as a whole, the authors find significant improvement in the health-risk profile of the U.S. population between the early 1970s and the early 2000s. Reduced smoking, better control of medical risk factors such as hypertension and cholesterol, and better education among the older population have been more important for mortality than the substantial increase in obesity.
The results suggest substantial caution about the future, though. Where reductions in smoking can be expected to have a continued impact on improved health, future changes in obesity might more than overwhelm this trend. Two-thirds of the U.S. population is now overweight or obese. As a result, continued increases in weight from current levels will have a bigger impact on health than did increases in weight from lower levels of Body Mass Index (BMI).
A large part of the impact of BMI is moderated through its effect on hypertension and high cholesterol. Given that not everyone with these conditions takes medications, or is controlled by the medication they do take, the resulting impact of rising weight on health can be significant. The optimistic side of this picture, however, is the potential for better control of obesity. If the effectiveness of risk-factor control can be increased, through more people taking medication and those taking it using it more regularly, much of the impact of obesity on mortality risk can be blunted, according to the authors.
Understanding how to improve utilization of and adherence to recommended medications are key issues in health outcomes. The research to date has focused on two possible avenues. The first is performance-based payment: physicians are now paid for office visits, but not for ensuring follow-up with their recommendations. The idea behind pay-for-performance systems is to reward physicians (or insurance companies) for successful efforts to increase utilization and possibly adherence. Such efforts might involve having nurse outreach, automatic medication refills, or more convenient office hours to monitor side effects.
The second strategy involves use of information technology. Patients can receive electronic reminders about medication goals, information such as blood pressure can be transmitted and monitored electronically, and automated decision tools can help with dosing and medication switches. Whether these or other strategies offer the greatest promise of improved adherence is uncertain. The authors' results suggest that evaluating these strategies in practice is a high research priority.
The authors use as their primary data source the National Health and Nutrition Examination Survey (NHANES). In the United States, it is the leading survey and includes both physical examination and laboratory measurements. The authors use two NHANES surveys, the first from 1971-5 (NHANES I) and the second from 1999-2002 (NHANES IV). Their analysis begins with NHANES I because it is the first population health survey that asked about smoking status, a key variable in health risk.
-- Les Picker
The Digest is not copyrighted and may be reproduced freely with appropriate attribution of source.