From Diseases to Behaviors: A New Way of Looking at Health

What is the leading cause of death in the United States?

Heart disease?

Cancer?

Smoking.

Smoking? Yes, depending on how you ask the question.

In the early 90s, McGinnis and Foege turned the age-old question of what people die of on its head by asking not what diseases people die of but rather what the causes of these are. Instead of chalking up the death of an older man to say lung cancer, they sought to understand the proximate cause of death, which in the case of lung cancer is largely smoking. Using published data, the researchers performed a simple but profound calculation — they multiplied the mortality rates of leading diseases by the cause-attributable fraction, that proportion of a disease that can be attributed to a particular cause (for example, in lung cancer 90 percent of deaths in men and 80 percent of deaths in women are attributable to smoking). Published in JAMA in 1993, their landmark study became a call to action for the public health community.

When looked at the conventional way the top 10 causes of death in America are represented in the top graph below (this is data from the 2004 update of the original study). Viewed this way heart disease, cancer, and stroke are the leading causes of death, respectively. This accounting may help us understand the nation’s burden of illness, but does little to tell us how to prevent these diseases and improve health. Through the lens of McGinnis and Foege, we get the bottom graph, which shows the top 10 causes of actual death (e.g., the major external modifiable factors that contribute to death). This analysis shows that the number 1 cause of death in America is tobacco use, followed closely by poor diet and physical activity, and then alcohol consumption.

For those of passionate about preventive health, we take heart that fully 6 of the 10 leading actual causes of death are controllable behaviors — all except microbial agents, toxic agents, motor vehicle, and firearms (though hand washing and seat belts can make a huge difference!) — and that 4 of the 10 are addressed by U.S. Preventive Services Task Force recommended clinical preventive services.

You may wonder why am I presenting the results of a 7-year-old study that is itself an update of a paper published 18 years ago.  Two weeks ago, I attended the American College of Preventive Medicine annual scientific session. Along with hundreds of my peers, I went to the conference excited to hear about the latest and greatest in preventive medicine. At highly anticipated keynote address, the opening speaker David Katz presented the first of what would be hundreds of slides of data shown throughout the three-day conference. And what did he kick off the conference with? The graphs above showing the top 10 actual causes of death in America.

Often when it comes to health it is hard to separate the signal from the noise. We get so many messages through the media, our social networks, and even our doctors. “Heart disease is the number 1 killer,” “Eat more omega-3,” “Don’t forget to check your breasts.” There are only so many hours in the day and so many competing agendas, it is hard to know what to make a priority. Sometimes to sort it all out it helps to just step back and let the data speak for itself.

So the question is: what does the data say to you?

- Shantanu Nundy, M.D.