Now Reading: Causes of death mapped to social factors instead of individual characteristics

I referred to this paper in a previous blog post but decided to give it its own, based on recent interest.

A lot of people are familiar with the often-cited breakthrough analysis, “Actual Causes of Death in the United States,” by McGinnis and Foege in 1993, which was followed by another, impactful analysis in 2002. Both changed ideas about what could be done to promote health and wellness, from looking at “heart disease,” to instead looking at “smoking,” (1993), to instead looking at “behavioral patterns,” (2002).

In the 2002 analysis, “social circumstances,” is listed as being responsible for 15 percent of deaths, with an accompanying editorial that said:

the data are still not crisp enough to quantify the contributions [of social circumstances] in the same fashion as many other factors. (reference)

Well, it’s almost 10 years later, and there is more crisp data, and greater awareness of the “causes of the causes” of poor health. I created a pie chart from that data:

And below are the three studies laid out the same way, 1993, 2002, 2011. We have come a long way in understanding what we can do to improve health. What do we believe the balance is now compared to 2002, and where should we act, and in what proportion, at the level of the individual, family, community, society? Quantified self and/or quantified community? An app for improved social health or one for individual behavior change?

Errors: Thumbnail path invalid ./../wp-content/thumb-cache/ab48a09d0826f4071ab2eb274003f959.jpgThumbnail path invalid ./../wp-content/thumb-cache/1d717633001067d58e112f3f2ac1a8b1.jpgThumbnail path invalid ./../wp-content/thumb-cache/9ad47e270769ce3b6cec567abec16913.jpg


Forgive me for not knowing explicitly how to connect the dots, but – how do we know (with reasonable rigor) that, for instance, education is the attributable cause of those 244,000 deaths?
Perhaps the answer is in that paper, but can you give me a concrete example, just so I understand the method?

 @ePatientDave Hey Dave, The answer is in the paper, and the data is as rigorous as the data used to generate the 1993 and 2002 studies :).
To derive the statistic that 245 000 deaths in the United States were attributable to low education ( higher than the number attributed to heart attacks ), the 2000 Census was used to get the % of Americans with low education (16 % of 24-64 year olds, 34.5 % of 65+ year olds) combined with death rates of Americans at each level of the education spectrum, expressed as the # of deaths over what would be expected in a population that did not have low education. Check out page 1461 of the paper, it’s explained well.
A useful “so what” look at this data is in the Marmot report, which I have blogged on previously -> . Look at the table toward the bottom of the post, which shows that among OECD countries, the investment in children is skewed toward late childhood, when it should be skewed the other way. Hungary is the only OECD country that invests more in a child’s early years in education compared to their later years. 
Glad I could pique your curiosity,

 @tedeytan You ALWAYS pique my curiosity. That’s why as life started to permit it, I’ve been re-engaging with your blog. 
On my “blog during summer vacation” mental list is to recount the things you taught me about how blogging can be about sharing everything you know that might be useful to others, not just about spouting one’s views. That was truly transformational.
Two examples are your past series “my own CIO,” which gave me a glimpse of what today is every single thing I do in my business technologically, and your occasional item “Someone asked me to explain this and it’s too long to do as a one-off email so I’m going to post it for posterity.” 

 @ePatientDave Well I am honored both about you spending time here and by the fact that you “get” me, everyone needs a group of those people, right? right,

 @ePatientDave Well I am honored both about you spending time here and by the fact that you “get” me, everyone needs a group of those people, right? right,

Hey Ted–was scanning your blog and this intrigued me. Unless we figure out how to move the discussion much further upstream and put more focus on the social determinants, I don’t believe we’ll make the progress we need to. The racial/social/educational correlations are striking.

It’s interesting that when I went to Peter Jennings’/ABC News Summit on Obesity in your neck of the woods many years ago, the one idea I brought home was how important the milieu is in the health of a population.

Thanks for posting this.

 @David Quam Hi Dave,
Thanks for stopping by! I am hopeful that our abilities and the creativity that people like you bring to this discussion as an aware physician bodes well for our communities. Looking forward to walking and talking soon,

Ted Eytan, MD