« In case you had forgotten | Main | Economists Speak »
How does socioeconomic status cause health?
Probably most of you know the familiar result that social status is one of the best predictors of personal health, even when adjusting for other measurable variables. David Cutler, Adriana Lleras-Muny and Tom Vogl have looked at the evidence more carefully and come up with the following:
This paper reviews the evidence on the well-known positive association between socioeconomic status and health. We focus on four dimensions of socioeconomic status -- education, financial resources, rank, and race and ethnicity -- paying particular attention to how the mechanisms linking health to each of these dimensions diverge and coincide. The extent to which socioeconomic advantage causes good health varies, both across these four dimensions and across the phases of the lifecycle. Circumstances in early life play a crucial role in determining the co-evolution of socioeconomic status and health throughout adulthood. In adulthood, a considerable part of the association runs from health to socioeconomic status, at least in the case of wealth. The diversity of pathways casts doubt upon theories that treat socioeconomic status as a unified concept.
In other words, "we don't know." My simplistic view has long been that high status simply helps "keep the juices flowing," in Roissy-like fashion, and that's good for you all over.
Can any of you high-status people find an ungated copy?
Posted by Tyler Cowen on September 23, 2008 at 04:25 PM in Medicine | Permalink
Comments
What gives? I thought you weren't linking to Roissy.
Posted by: d.cous. at Sep 23, 2008 4:33:19 PM
I think there is a lurking cause behind both status and health: discipline. A disciplined person can stay in school. He can refrain from both unhealthy actions such as gluttony, drug use, or smoking. And at the same time he can refrain from unwise financial decisions such as binging on credit purchases. (Obviously, race and ethnicity are not affected by discipline, but can coincide with other behaviors that are.)
Posted by: Brian at Sep 23, 2008 4:50:11 PM
sir,
for shame for that link! would be far better for the world if gentlemen such as "roissy" did not make their life's work a mockery of honor and gallantry!
yours truly
c. v. snicker
Posted by: chet snicker at Sep 23, 2008 4:51:36 PM
I am almost finished reading a book titled Worried Sick, by Dr. Nortin Hadler. Hadler contends that many procedures like bypass surgery, stents, angioplasty, colonoscopy, mammography, prostate cancer and cholesterol screening, among others, ultimately do very little for the patient and a lot for the medical and pharmaceutical industry. He claims that the biggest predictor of health is SES. He proposes a health insurance scheme based on proven effectiveness of procedures and pharmaceuticals, with medical care incorporating SES questions into the history and diagnosis. His contention is that we have "medicalized" conditions that have always been the bumps and bruises of life, with this medicalization resulting eventually in health insurance coverage and expansion of definitions that captures more people in these conditions and thereby expands the pool of patients. I'd like to hear the take from economists on Hadler's views and recommendations.
Posted by: bronxilla at Sep 23, 2008 4:57:21 PM
nber15.nber.org/essentials/paper7.pdf
Posted by: at Sep 23, 2008 5:25:23 PM
A quick scan of the paper shows they have no control for IQ. Gottfredson a while back showed that correcting for IQ makes the effect of SES pretty much vanish: simply put, smart people can understand a doctor's instructions, earn more money, and achieve higher status. Low IQ folks cannot.
Gottfredson, Linda. 2004. "Intelligence: Is It the Epidemiologists’ Elusive “Fundamental Cause” of Social Class Inequalities in Health?". Journal of Personality and Social Psychology.
Posted by: Eric Crampton at Sep 23, 2008 5:26:06 PM
Sorry, the "nber15.nber.org/essentials/paper7.pdf" link turns out to be incorrect.
Posted by: at Sep 23, 2008 5:33:21 PM
http://cc.msnscache.com/cache.aspx?q=74061330854280&mkt=en-US&setlang=en-US&w=f530a1c2,58d147e7
Posted by: burger flipper at Sep 23, 2008 7:07:27 PM
I recommend Sir Michael Marmot's _The Status Syndrome_ for a book length treatment of this. IIRC Sir Michael said that stress pathways from low status was the cause of higher mortality associated w/ low status. I'm just a layman but I thought the book was pretty well argued.
Posted by: Scot at Sep 23, 2008 8:23:04 PM
Perhaps health causes high status.
Posted by: Don at Sep 23, 2008 9:00:36 PM
I agree that you economic status has something to do with your health. I think that sometimes when your economic status is low it increases your risk of stress. A lot of this comes from not having enough money to support you family. I also believe that your health can be lowered when your economic status is high because of the high levels of stress someone receives from working a intensive labor job. Therefore your economic status can effect your health in positive and negative ways.
Posted by: at Sep 23, 2008 9:07:23 PM
bronxilla if you have not already you should look at Robin Hanson's posts on medical care at http://www.overcomingbias.com.
Tyler thank you for this post. I am very intested in what the SES and health relationship means and what is causal and how.
Posted by: Floccina at Sep 23, 2008 9:37:14 PM
I believe Eric Crampton actually meant to cite to The Bell Curve.
There's a pretty long string of research linking geography, poverty, and adverse health outcomes. Just to touch the tip of the iceberg, poor people live in areas exposed to more industrial pollutants, further from good hospitals and clinics, further from jobs, closer to crime, and even further from grocery stores (which means they have to rely on 7 Eleven--not exactly known for its produce--for meals). If you don't understand how ghettos were constructed by segregation and deindustrialization, then you should consult T. Sugrue's Origins of the Urban Crisis. For a look at these phenomena in Oakland (Alameda Co.), see the works of Dr. Anthony Iton.
Posted by: at Sep 23, 2008 10:00:40 PM
I, too , "think there is a lurking cause behind both status and health" -- stress.
speaking from my own person experience of prep school kids, high status reduces stress because the respect you get from other people increases the amount of control that you have over your life, enlarges the number of social networks that you will willingly interact with, and increases your personal happiness, thereby reducing your stress.
In our society, if you come from a high-net-worth family (which buys you a prestigious education), you have financial resources and high-status social networks independent from your job. You are much more free to screw up your career and your life with such a safety net, decreasing your stress and increasing your happiness.
This would only apply to inherited high status, however...
Posted by: Diana at Sep 23, 2008 10:11:04 PM
This is pretty irrelevant, but I feel honor-bound to defend 7-11 from this scurrilous attack. I quasi-frequently purchase very healthy meals from 7-11, including bananas, apples, hard-boiled eggs, fruit smoothies, milk, yogurt, and sandwiches on whole-wheat bread. They also have other healthy foods that require preparation and that I don't buy. Of course the selection of unhealthy foods is greater, but if you try you can get good meals at 7-11.
Posted by: Cliff at Sep 23, 2008 10:19:12 PM
Eric Crampton is correct: IQ is probably the underlying explanatory factor behind social class health differences.
Ian Deary's group http://en.wikipedia.org/wiki/Ian_Deary have created a retrospective cohort study based on one morning in 1932 when all Scottish 11 year olds had an IQ test. This single measurement on one day of a childs life, just a pencil and paper test, was predictive of a wide range of diseases and lifespan - as well as eventual social class and educational attainment. When the subjects were retested decades later, their IQ was highly stable.
See the various publications on: http://www.psy.ed.ac.uk/people/iand/publications.html
Posted by: bgc at Sep 24, 2008 1:44:26 AM
"In adulthood, a considerable part of the association runs from health to socioeconomic status, at least in the case of wealth."
This means health causes wealth, right? I'm not sure, the wording is confusing. Maybe I'd better go get a checkup.
If so, I'd tend to agree. I don't think people would like to admit to themselves how much they tend to shun the sick (and people with problems generally). We talk about things as if the world is an accepting place, but in reality, people are a lot more self-interested than they think they are. If you can't do something for them, you are rejected, in big or little ways. Reverse snowball ensues. As Bono said, "the rich stay healthy and the sick stay poor."
Posted by: Andrew at Sep 24, 2008 1:58:26 AM
Generally speaking, people who are broke finanically are that way for a reason. That is they think like a poor person, often coming from a point of "not enough". Often people who have greater finanicial resources have overcome major obstacles. They train themselves to be successful, and health would naturally follow in that line of thinking.
Posted by: Sports Fitness Nutrition at Sep 24, 2008 9:31:39 AM
The problem with attributing health disparities to IQ differences is that IQ is also dependent on environmental factors:
http://www.nytimes.com/2007/12/09/opinion/09nisbett.html
Posted by: Ricardo at Sep 24, 2008 3:23:05 PM
Ricardo is right: SES is a determinant of IQ, so the relationship is complicated. The main important point of the Cutler et al paper seems to me to be that SES isn't a unitary causal factor, but a syndrome of correlated traits, which should be sort of obvious, but isn't treated like it. It's more useful to break SES into its constituent variables.
Posted by: Will Wilkinson at Sep 24, 2008 3:34:12 PM
No, SES is not an important determinant of IQ; family environment has no great influence on the individual and their later life outcomes as study after study show.
80% of the variance in IQ is genetic, the rest comes from nonshared environment (environments unique to the individual -- not shared with siblings).
Ian Deary tracked people over their whole lifespan. The IQ scores measured at age 11 were the same as IQ scores at age 77, and these early differences are what predicted later health.
Posted by: Jason Malloy at Sep 24, 2008 4:07:04 PM
No commenter here has yet mentioned the connection between "good looks" and health. Beauty is, in part, symmetry, and, in part, resistance to disfiguring health problems. (e.g., development, the immune system).
Both IQ and "good looks" have genetic components, and both are sought qualities in a partner. What else is needed to explain the correlation of health with SES?
Posted by: NeedleFacatory at Sep 25, 2008 12:37:04 PM
It would be interesting to know how much the causation goes from IQ to health (you understand the doctor's instructions better, you correctly work out that these symptoms require an immediate trip to the hospital, etc.), and how much from early childhood health (caused by genetics and environment) affecting both IQ and health later in life. In the extreme case, if you're malnourished and sick for most of your childhood, that might depress both your IQ and your expected lifespan.
Does anyone know if there's data that addresses that?
Posted by: albatross at Sep 25, 2008 2:11:31 PM
No, SES is not an important determinant of IQ; family environment has no great influence on the individual and their later life outcomes as study after study show.
80% of the variance in IQ is genetic, the rest comes from nonshared environment (environments unique to the individual -- not shared with siblings).
Ian Deary tracked people over their whole lifespan. The IQ scores measured at age 11 were the same as IQ scores at age 77, and these early differences are what predicted later health.
that just means that IQ is pretty much set by the age of 11. it says nothing about what happens before the age of 11.
it can also be very difficult to differentiate genetic causes from shared environmental causes especially because genetic links are usually determined based on twin studies who...surprise surprise...share the same environment during the formative years of their neuro systems.
Posted by: BK at Sep 25, 2008 9:36:02 PM
BK,
Look, if you don't know the behavioral genetic literature, or its methods, then please don't spout off about it as if you do. IQ in preschool, and even in infancy, predicts adult IQ almost as well. You can take any kinship comparison you want and get the same results (step-siblings, half-siblings, cousins, mis-identified twins, "virtual" twins, DZ twins, adopted siblings); shared environment in the "formative years" doesn't matter very much. Similarities that can't be accounted for by genes fade and disappear with age, regardless of proximity. Meanwhile genetic relatives become more alike over time, whether or not they even know eachother.
Posted by: Jason Malloy at Sep 25, 2008 10:39:25 PM
