Population Health and the Influence of Medical and Scientific Advances

• Author: J. Michael McGinnis
• Position: M.D., M.P.P., Institute of Medicine
• Pages: 9-20

Page 9

M.D., M.P.P., Institute of Medicine, The National Academies. This work is based upon a live presentation made on February 4, 2004 and does not necessarily reflect events and changes thereafter.

Ours is an extraordinary time in the evolution of human health, with unprecedented insights on the nature of the determinants of health, how important they are, how they work, how they interact, and how they might be influenced. It is important therefore to place consideration of the role of genetics into the broader context, to assess how the tools emerging from the genomics revolution may shape the dynamics within and among the various domains of influence on the health of populations.

Some of that context is provided by data about various aspects of the health care delivery system. In 2005, approximately two trillion dollars will be spent for health care in the United States.¹ That amounts to about $7,000 for each of the nation's inhabitants, and represents sixteen percent of the Gross Domestic Product.² Although this level of investment puts us far ahead of other nations, our returns on investment, in health terms, are limited. In 2004, we ranked twenty-third among nations in life expectancy, far behind countries like Japan, Sweden, and Iceland.³

There are some explanations for this discrepancy. Our technologic superiority in medicine has its downsides in higher rates of medical errors, accounting for as many as 98,000 deaths a year.⁴ But more importantly, we are a nation of great social heterogeneity, different cultures, various ethnicities, and Page 10 geographically dispersed. We have some groups with chronically lower income streams, newly arrived groups with more acute needs, and a recently widening gap among the various income tiers. Except for our older people, our social support system is a patchwork of policies at the local, state, and federal levels. We have wide discrepancies in access to care and substantial instabilities in the continuity of care. People receive care, but it is expensive, and it is a scramble. Moreover, we are in the midst of a full-blown obesity epidemic in this country. Two-thirds of the population is overweight; the number of overweight adults and children has doubled in the last twenty years.⁵ Finally, we have high rates of low birth weight--which are in part derivative of higher rates of prenatal substance abuse.⁶ And nothing shortens the horizon of life expectancy like an infant death.

How then, in the midst of the challenges, do we identify our primary opportunities for improving the health of Americans? A reasonable starting point is to look at those factors that determine the health of populations, about which we now have a clearer understanding. It was not very long ago that we had a rather fatalistic approach to life--acceptance of accidents, chronic diseases, and other challenges as merely the inevitable consequences of life and the aging process.

We now know that the health of populations is determined by the dynamics in five domains of influence: our genetic predispositions, our social circumstances, our environmental exposures, our behavior patterns, and the medical care we receive.⁷

The topic of genetic predispositions is well covered elsewhere in this volume. Suffice it to say that the human genome numbers three billion base pairs, with around 20,000 to 25,000 genes under current estimates.⁸ These genes exert their influence on health in Page 11 various fundamental ways. Each of us has embedded in our genetic platform the assembly line blueprint for construction of the proteins which give form to our sizes, our shapes, our personalities, even to the biologic limit of our life expectancies-- somewhere between eighty-five and 110 years, if fibroblast tissue cultures are to be believed.⁹ Eons of adaptation has given an approximate form to these instructions, which are essentially similar across individuals, and sets us on a course that we call normal--that is, one which is not markedly disadvantaged from the environments in which we find ourselves.

Under certain circumstances, inborn variance of the code confers some discernable measure of disadvantage. This can be the result of even a single abnormal gene inherited from a parent, as with recessive conditions like cystic fibrosis or dominant conditions like Huntington's disease, or it can be the result of an entire chromosomal region being affected, as in Down's syndrome.

Changes also can occur in the codes of certain cells as a result of epigenetic influences, including interactions with other genes and gene products or various exposures throughout the life cycle. In some carcinogenesis or neural tube defects, for example, environmental triggers can alter the genetic coding signals, resulting in abnormal growth by certain cells or tissues. Knowledge in each of these areas is still primitive, despite the explosion of activity at centers such as the Pennington Biomedical Research Center, but that knowledge is growing rapidly with the completion of the sequencing of the human genome far ahead of schedule. Progress on identifying the locations of disease-linked genes is accelerating, and the number in 2004 had already reached some 1,400.

Still, sequencing the genome leaves open the question of the nature of its influence, which is, in part, our charge here: to understand how the genome plays in complex interaction with these other domains of influence. Although only about two percent of mortality in the U.S. can be attributed to purely genetic Page 12 diseases,¹⁰ a sizeable proportion of disorders of late onset--like diabetes, cardiovascular disease, and cancer--have an important genetic component. The significance of that component is still uncertain for most conditions. The BRCA1 gene accounts for only about five to ten percent of breast cancers in the United States.¹¹

Only about ten percent of colon cancers may be explained by these yet unidentified genes,¹² and only about one case in twenty of those with elevated serum cholesterol levels may be explained by familial hyperlipidemia.¹³ Studies of monozygotic twins focusing on the occurrence of schizophrenia, and other similar twin studies looking at mental alertness in older people, have found that about half of each could be explained by genetic factors.¹⁴ Some say two-thirds of the risk of obesity may be genetic but, like most predispositions, that risk is expressed only with exposure to lifestyle factors which are controllable. At this point, the best we can say is that the burden of purely genetic diseases is limited, and the vast majority of the genetic contribution is played out in those complex interactions with the other domains of influence.

At birth, our first encounter is with the domain of our social circumstances, about which a great deal more has been learned in recent years. From cradle to grave, our interpersonal linkages matter. Studies consistently have shown that infant nurturing enhances socialization and survival.¹⁵ Adults, including older people, who are socially isolated, have a two to fourfold higher death rate than others.¹⁶ We have learned that prenatal home visits Page 13 to at-risk mothers can reduce the likelihood of both risky health behaviors and criminal activities some fifteen years hence.¹⁷ For the population as a whole, the most consistent predictor of the likelihood of death in any given year is level of education, with those ages forty-five to sixty-four in the highest levels of education having death rates that are two to five times lower than those in the lowest level.¹⁸ Poverty has been said to account for six percent of U.S. mortality.¹⁹

The observation also has been made that each one percent rise in income inequality, that is, the differential between rich and poor, is associated with approximately a four percent increase in deaths among those on the low end.²⁰ It is difficult to sort out the pecuniary elements of deprivation from the biological, behavioral, and psychological consequences of place. For example, a study of British government workers, virtually none of whom were in poverty, found a threefold difference in death rates between the highest and the lowest positions.²¹ Interesting work on the physiologic mediators of the stress reaction suggests that there may be commonalities in the pathways involved...