Intended for healthcare professionals

Observations Yankee Doodling

The key to longevity

BMJ 2012; 345 doi: (Published 25 September 2012) Cite this as: BMJ 2012;345:e6456
  1. Douglas Kamerow, chief scientist, RTI International, and associate editor, BMJ
  1. dkamerow{at}

Can we help people live longer with better medical care?

Almost six years ago, in my first Observations column for the BMJ, I wrote that the secret to a longer life is something that doctors can do very little to affect: patients’ level of education.1 It seems that not much has changed.

As I write this, the lead story on the front page of the New York Times announces (yet again) that poorly educated Americans die younger than those who have had more schooling; now, however, it also seems that the life expectancy of those lacking a high school diploma is actually getting shorter.2 This raises the previously unimaginable possibility that at least some part of a generation of children could actually die younger, on average, than their parents.

The Times article cited a recent study by S Jay Olshansky and colleagues that found that US men and women with less than 12 years of education in 2008 had life expectancies similar to US adults in the 1950s and 1960s.3 Further, when you include race as a factor, the disparities between the life expectancy of white people with 16 or more years of education and that of black people with 12 or fewer years are even greater: 14 years for men and 10 for women.

This brings up an interesting question. Is education actually a fundamental cause of better health and longer life, or is it just a marker for other, well described risk factors, such as smoking, poor diet, and inadequate healthcare? If education and other social factors are fundamental causes, then can we hope to improve longevity by focusing only on medical risk factors?

In medicine we are used to thinking of risk factors as the “behind the scenes” mediators of disease and death. We officially track the leading causes of death by disease, as recorded on death certificates. The US Centers for Disease Control and Prevention publishes the list of the leading causes of death every year, starting with heart disease and moving through cancer, chronic lung disease, stroke, injuries, and the rest. But the CDC also acknowledges that the “actual” causes of death are the risk factors for those diseases: smoking, poor diet, lack of exercise, and so forth.4

Some sociologists look at it differently. They argue that “risk factor epidemiology” is on the way out. They see social conditions—education, occupation, and income, the components of what used to be called “social medicine”—as the fundamental causes of health inequality and the keys to improving longevity.5

Obviously, people do die from diseases, and diseases don’t arise directly from poor social conditions. Poverty does not actually inflict people with heart disease or cancer. There have to be mediators, so the question really is which mediators are more explanatory—traditional medical ones (smoking, for example) or socioeconomics.

One telling argument in favor of socioeconomic status as a fundamental cause of health inequalities is the changing nature of the “medical” risk factors over the years despite the persistence of the association between health status and socioeconomic status. Overcrowding and poor sanitation led to deaths from infectious disease among poor people 200 years ago. Fixing these conditions ended deaths from cholera, typhoid, measles, and diphtheria but did not eliminate the relationship between socioeconomic status and health status.5 It is now mediated through smoking, bad diet, and other behavioral and environmental risk factors.

Similarly, poor access to healthcare is seen as an extremely important risk factor for health status and something that can explain much of the relationship between low socioeconomic status and curtailed longevity. Yet providing free healthcare for all, which has been implemented in many countries, has not necessarily reduced the relationship between socioeconomic status and mortality, which has persisted despite universal healthcare.6

To come back to the newspaper headlines that motivated this musing, what should our response be to yet more evidence that educational (and racial) differences account for increasing disparities in life expectancy? In the US, at least, we take a largely individual approach to improving health, trying to get people to stop smoking, exercise more, see their doctor to get flu shots, and so forth. Maybe it is time to heed the advice of sociologists and move beyond standard risk factors to try to get at the fundamental causes of ill health in the community.

This means providing population based interventions that do not depend on individual resources and actions. Immunizations in schools and workplaces, not in doctors’ offices. Fluoridation of tap water rather than toothpaste. Even passing and enforcing the dreaded regulations to ensure safe housing, cars, and environment rather than enjoining individuals to watch their children, drive safely, and clean up their individual spaces.

The big (and frightening) conclusion to all this is that without improving education and economic status we probably have little hope of closing the longevity gap. It is frustrating to those of us in individual healthcare, because it doesn’t seem to be anything we can do much about, as doctors at least. Maybe, however, we can affect change as citizens— or at least as voters.

This is a message of perhaps particular relevance in the US, as we look forward to the November election, just a few weeks away.


Cite this as: BMJ 2012;345:e6456



View Abstract