The original paper is here. Map at Gene Expression via Marginal Revolution.
Tags: Gene Expression, life expectancy, Marginal Revolution
This entry was posted on Monday, August 3rd, 2009 at 10:32 am and is filed under Rankings. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.
Doesn’t look good for the good ol’ boys.
Hard to tell from the scale of the map, but it looks like most urban areas do well, while rural (outside the South) are literally all over the map.
I took a look at the source data. It’s a treasure trove of info, because the include the county code. Thus, you can compare it to all kinds of data collected by the Census Bureau.
Just for kicks, I regressed the white male life expectancy in 1999 vs. median household income in 1999. What do you know? p factor is zero (about as statistically significant as can be), correlation coefficient is 0.28. Not all that correlated, if you ask me.
Next, just because I had the data, and playing a hunch, I regressed WMLE to the percentage of the population of each county that was black in the 2000 census.
Amazingly, it was MORE statistically significant than income, and the correlation coefficient was 17% HIGHER than income.
Because cleaning up the data was non-trivial (who knew that there were so many counties with NO African Americans in 2000?), I didn’t do a multiple regression for the entire US. But just looking at Indiana, controlling for both factors can explain 48% of the variation in the data. That’s not bad for socioeconomic data. Richard Florida makes sweeping conclusions about society with less than that.
You’re ahead of me on statistics, but I’d guess that the heavy preponderance of deaths in the South and the number of Southern states with large Black populations are correlation, not causality. Could you do the stats for New England or West of the Mississippi, factor out the South, and see if the correlation holds?
My guess is lifestyle is more related to the death rate.• Here are the top ten states for smoking: Kentucky, West Virginia, Tennessee, Oklahoma, Ohio, Arkansas, Alaska, Alabama, Indianan and Mississippi. (If you look at the map, the high death rates in Ohio & Indiana are in the south, bordering Kentucky and West Virginia).• Here are the top 10 for obesity: Mississippi, West Virginia, Louisiana, South Carolina, Tennessee, Kentucky, Oklahoma, Arkansas, Michigan.• Interestingly, alcoholism is higher in Northern states so doesn’t fit this pattern.
The Gene Expression site implies that there’s a relationship with the proportion of Scotch-Irish in the White population. But my bet is still on lifestyle.
oops, left out Alabama in obesity.
Michael, like I said, I did it just for Indiana, my home state. The correlation is still very high.
What got me thinking about the connection was that my county, Lake, is 25% African American, and has a WMLE of only 71 years. Most of Indiana is much, much whiter, and has much higher WMLE.
I also noticed that Marion County, which contains Indianapolis, had a high black population and a low white male life expectancy.
This could just be a measure of general poverty, with the cofactors you mention. A person with a little time on his hands could easily do correlation after correlation on this data. As I said, having that county code there makes mashing up the data with all kinds of Census Bureau data really easy. There’s a dataset that gives % living in poverty I could probably regress, but one would think that median household income would capture that pretty well.
The Gene Expression site implies that there’s a relationship with the proportion of Scotch-Irish in the White population.
If you’ve read Thomas Sowell’s “Black Rednecks”, you’ll understand another reason that I regressed WMLE vs. % black population. Sowell’s thesis is that the black underclass has the same culture as Scotch-Irish of generations past. They picked it up from the whites that they lived among and never outgrew it (although Sowell says that the whites have).
Here’s another study that doesn’t separate by gender but does divide race and geography in interesting ways. It points to injuries and untreated chronic diseases, which I’d attribute to lifestyle, as leading causes of mortality differences.(I found this at http://health.dailynewscentral.com/content/view/0002418/42/)———————————————————–
Although the average life span in the US continues to rise, gaps in life expectancy have changed little from 1982 to 2001. There is a wide gulf — as much as 33 years — between those who enjoy the best health and those who are most likely to suffer from illnesses, according to a new study published in PLoS Medicine.
Researchers at Harvard University’s Initiative for Global Health and its School of Public Health divided the US into eight “Americas” based on factors including race, location, population density, income and homicide rates:
• Asians, 10.4 million popuulation, $21,566 average income, life expectancy 84.9 years;• Northland low-income rural whites, 3.6 million population, $17,758 average income, life expectancy 79 years;• Middle Americans, 214 million population, $24,640 average income, life expectancy 77.9 years;• Low-income whites in Appalachia and Mississippi, 16.6 million population, $16,390 average income, life expectancy 75 years;• Western Native American, 1 million, $10,029 average income, life expectancy 72.7 years;• Black middle America, 23.4 million population, $15,412 average income, life expectancy 72.9 years;• Southern, rural, low-income black, 5.8 million population, $10,463 average income, life expectancy 71.2 years; and• High-risk urban black, 7.5 million population, $14,800 average income, life expectancy 71.1 years.
The primary cause of the disparities between racial and geographic groups is early death from chronic disease and injuries, an analysis of data from the Census Bureau and the National Center for Health Statistics showed.
Asian-American women living in Bergen County, NJ, enjoy the greatest life expectancy in the US, at 91 years. American Indians in South Dakota have the worst, at 58 years.
The differences were attributed to a combination of injuries and such preventable risk factors as smoking, alcohol, obesity, high blood pressure, elevated cholesterol, diet and physical inactivity — particularly among people from 15 years to 59 years of age. They were not due to income, insurance, infant mortality, AIDS or violence, said the study’s lead investigator, Christopher J.L. Murray, director of the Harvard Initiative for Global Health.
Most public health initiatives target children and the elderly, he noted.
The study looked at life expectancy by geographical areas as well. Hawaii led the 50 states and Washington, DC, with an average life span of 80 years, while DC trailed at 72 years.
Personal choices could be more important than access to medical care in improving life expectancy, Dr. Murray noted. Half of the people who have high-blood pressure fail to get it controlled, two-thirds of those with high cholesterol do not get medication to lower it, and two-thirds of diabetics fail to manage the disease, in spite of the fact that 85 percent of the population overall has health insurance.
That study was in the back of my mind when I did the regressions.
Actually, for such a simple statistic, life expectancy is loaded with all kinds of statistical baggage. It is extremely sensitive to infant mortality, homicides, auto accidents, and anything else that causes death at an early age. It is relatively insensitive to things that increase life expectancy at the high end (little things like overly generous Medicare coverage, for example).
So be careful what you wish for. From what I’ve seen, US health insurers do a better job of getting people on statins and other preventive medicines than the socialized medicine countries do. And maybe the way to increase usage of these drugs is through… ADVERTISING. TO CONSUMERS. Perish the thought.
Being Harvard docs, Murray & company talk about managing chronic conditions with drugs, which is going to be expensive and have other problems. The change that has to take place to improve US health and longevity is behavioral rather than treatment oriented. As the article says, managing “preventable risk factors as smoking, alcohol, obesity, high blood pressure, elevated cholesterol, diet and physical inactivity.” Not to mention wearing seat belts, not texting at 70 miles per hour and OHSA compliance in low wage industries.
This will be more important in the long run than who pays for what, the public option debate, etc. If someone has bypass surgery because they eat lots of fat and don’t exercise, it’s going to be expensive however you pay for it.
But these changes will take much more time and yield fewer profits than just pushing pills and surgery, so they’re not the focus. Obama’s new initiative of researching effective practices is a start.
But these changes will take much more time and yield fewer profits than just pushing pills and surgery, so they’re not the focus.
Perhaps. Or maybe getting people to pay more of their healthcare themselves (less third party payment) is the… least nannyish… way to get people to change their lifestyles.
Also, seeing as how higher BMI’s are associated with higher longevity, I’m not so sure that moderate obesity is a problem.
I also think the jury is still out regarding lifesyle vs. drugs. The results of widespread statin use is pretty spectacular. As these drugs go offpatent and become generic, the costs for treatment become pretty minimal (thanks to Walmart, if noting else).
In fact, getting the FDA to allow more drugs to go OTC is a pretty unheralded cost savings measure.
“I also think the jury is still out regarding lifestyle vs. drugs.”
Sorry, No it’s not. As the Harvard article says, risk factors such as smoking, alcohol, obesity, high blood pressure and elevated cholesterol are all manageable by behavior, including diet and exercise.• Smoking is absolutely tied to emphysema, heart disease and lung cancer.• Heavy alcohol and drug use are absolutely tied to liver damage and disease.• Weight bearing exercise builds bone density and muscle strength.• Aerobic exercise builds cardiovascular health and lung capacity.• Maternal diet, fitness, diet and alcohol & drug use are tied to low birth weights and infant mortality.• Seat belt use reduces serious injuries and deaths.
Studies back to the ’90’s and beyond show that diet and exercise can reduce heart disease and prolong health.William Evans Biomarkers studies at Tufts. http://www.amazon.com/Biomarkers-10-Keys…/0671778986Dean Ornish’s studies on preventing and reversing heart disease. http://www.pmri.org/
A leading cause of injury and death in older people is falling, and the best prevention is good balance and strength, which result from exercise. And the problem with falls is often broken bones, which are less likely with strong bones from… diet and exercise.
I agree with you about the BMI and people’s obsession with weight. It’s healthier to be fit and heavy than unfit and skinny. But I’m not sure there’s such a thing as moderate obesity, as opposed to being moderately “overweight”. I think obesity implies and unhealthy amount of fat and low fitness.
Michael, it’s no longer the realm of science fiction to believe that, in the very near future, there will be pills to control not just heart disease, but osteoperosis, diabetes, and weight.
And it won’t be very long after that that those drugs are available over-the-counter for $4 a month at Walmart.
At that time, what will be the point of excercising or any of the other behaviors that you cite?
You know what pill we really need? One that improves future time orientation (maybe by boosting IQ?). Then people wouldn’t smoke, drink excessively, or drive without their seat belts.
I think we’re looking at this differently. I’m focusing on health and you’re focusing on disease — along with most of Western medicine, so I guess you’re in good company.
I would maintain that overall health and well being is the goal including such things as strength, balance, strong cardio, lung capacity and absence of preventable chronic diseases. In addition the mental, psychological and social benefits of exercise are increasingly recognized. Being fit is a holistic experience, not just the lack of certain diseases or conditions. When it comes right down to it, I don’t just exercise because it’s healthy, I do it because it’s fun and I feel good. We’re designed to be active, it’s the set point of being humans and animals.
That’s number one. Number two is that all drugs have side effects, every one of them right down to aspirin. Once you get on the track of using drugs for conditions you could prevent with healthy behavior, you end up with either new risk factors and/or additional drugs to treat the side effects, ad nauseam (literally). This isn’t to say that I don’t use prescription and over the counter drugs, they have their uses but I try to limit them as much as possible.
I’m sure you’ve noticed that we’re the only two commenters on this post. Maybe nobody else is reading this. Maybe Barack should have us over for a beer to talk this out.
By the way, I really don’t care if change comes from government, industry, individuals, families, churches, etc. But I do believe that our present course towards sloth and unhealthy behavior is very bad for us as a society, a nation and a species.
When it comes down to it, behavioral change happens on a neighborhood, individual and friendship group level. That’s why I was excited about the growth of farmers markets about a month ago.
I don’t think this is something the federal government can mandate. Oprah and the Stewarts (Jon & Martha) could probably have more influence than the entire HHS bureaucracy. Michele’s garden and Barack’s basketball as role models may be more important than any policy statements they could make.
Michael, spoken like a true SWPL.
Talking public policy, rather than personal philosophy, you’re never going to get people to adopt your lifestyle without being very heavyhanded.
The fight against smoking, while wrapped in the rhetoric of health, really came down to the rights of one individual (the non-smoker) not to be inconveinanced by another (the smoker). You’re going to have a lot harder time addressing habits that don’t directly negatively effect anyone but the person with the habit.
Maybe our basic difference is in where we draw the line between public and private good. I’m a strong believer in public health, that the society has the obligation to care for its members. If we could eliminate smoking and alcoholism and increase seat belt use it would probably pay the cost of universal health care, private or public. Here are some of the things that have led me this direction.
I don’t know if you consider seat belt laws heavy handed. The only serious car accident I’ve been in, I was forced off the road and lost control when I was thrown from the drivers seat — I think I could have kept control if I’d had a seat belt, and have been wearing them every since. I’ve been first on the scene for two fatal accidents and from looking at the cars I believe both drivers would have walked away if they’d been wearing shoulder and seat belts.
On smoking, the public education and taxation to reduce smoking I believe are justified in terms of the cost to society of caring for the sickness that results. I quit smoking after a visit to my first wife’s parents and listening to my father in law coughing his guts out with emphysema. I also know that addictions are by definition irrational. My uncle in Ohio was a coal miner and died of black lung — when he was bedridden and on oxygen he still smoked. Neither of these were stupid men in terms of IQ.
On the other hand, the banning of smoking in bars I think were probably a matter of rights and maybe heavy handed. In Portland smoking had disappeared from most restaurants before any ban, because as most people quit the restaurant owners realized it was better for business to eliminate smoking sections. But in the places that smokers hung out, it was probably their right to be left alone.
The health results of heavy alcohol (legal) and drug use (not) are significant. I’ve had two friends who were heavy long time alcohol and/or drug users who got clean and sober and died young of liver failure anyway. One after a liver transplant that eventually failed.
Unhealthy food is another matter. You can quit smoking or drinking, or never start, but you can’t stop eating. This will need to be a cultural change, as will exercise, rather than legislation.
I’m betting on the good sense of the American people, despite H.L. Menckin. If we’re still on this blog in 10 years, let’s compare notes.
I do think that seat belt laws that are primarily enforced (police can pull you over for it) are heavy handed and have unintended consequences. We just had a situation where the police caused an accident on the highway. They pulled over someone for the seatbelt, and there was a chain reaction crash from other drivers slowing down because of the flashing lights. A family of 6 was crushed in their minivan between two trucks.
I personally always use my seatbelt, I don’t smoke, have never done drugs in my entire life, and drink relatively moderately. But I’m also a college educated professional. I think that you underestimate how many of the negative lifestyle choices are correlated to poor future time orientation, and how difficult it will be to change those behaviors without being very heavy handed.
When I say heavy handed, I think of what they’re now proposing to do in England. They want to put closed circuit cameras in the homes of “troubled families” to make sure that they’re not hitting the kids, that the kids are eating healthy, etc. That’s what we’d need to do to really achieve what you’re talking about. It would never fly in the US.
Your site looks good. Have you thought about upgrading your design? I think that will make it loads better.
chillicothe website designer…
[...]Creative Class » Blog Archive » Life Expectancy Map – Creative Class[...]…