Skip to content
A Member of the Law Professor Blogs Network

Brits Healthier Than Americans? (Part III) Gail Heriot

In my last post, I expressed some skepticism about the recent JAMA study that finds middle-aged, white Americans are, as a group, in much poorer health than their English (not British) counterparts.

The problem is that American life expectancy at age 65 (the only age other than zero for which I could find comparable data) is as good or better than U.K. life expectancy.  (I should add here as I did in my last post that American life expectancy at birth is somewhat less good. I believe that this is largely due to higher infant mortality rates–not something to be proud of, but not the same thing as a higher adult mortality or morbidity rate.)  It seems implausible to me that the higher morbidity rates that the JAMA study points to have no effect on mortality after age 65.  Between the mortality rates after age 65 and the JAMA study, I am inclined to credit the mortality rates, since that kind of information is not subject to the same kinds of judgment calls, manipulations and sampling and other errors that the data in the JAMA study may be prone to.

The JAMA study presents both self-reported and clinical data for its conclusion.  I am inclined to discount the self-reported data entirely.  So what if American interviewees report more illness than English interviewees?  That might be because they’ve received better health care and hence have had more of their maladies diagnosed.  Alternatively, it might be because Americans are more inclined to complain. I doubt anyone would regard that as enough to outweigh the evidence of life expectancies at age 65.

The clinical data, however, are not as easy to write off.  According to the study, 6.4% of white Americans between the ages of 40 and 70 have glycosylated hemoglobin levels higher than 6.5 (apparently a marker for diabetes).  The rate among their English counterparts was only 3.8%.  Similarly, 34.5% of the American sample had hypertension; only 32.2% of the English sample did.  Indeed all of the clinical tests reported in the study–C-reactive protein, fibrinogin, and HDL cholesterol–show results that favor the English, sometimes by substantial margins.  For example, 21.6% of the American sample had low concentrations of HDL (“good”) cholesterol, while only 8.0% of the English sample did.  All this information was gleaned from major studies conducted by national governments using very large samples.

I don’t know why that the data came back the way it did.  I don’t see any evidence of manipulation.  The authors do not appear to be politically motivated.  Indeed, they state themselves that “it is … important to recognize that health insurance cannot be the central reason for the better health outcomes in England because the top SES tier of the US population have close to universal access but their health outcomes are often worse than those of their English counterparts.” (I suspect that some people will  erroneously try to use it to argue for state-run health care anyway.)

My first thought when I saw the clinical results was that the American sample must be top heavy with participants towards the higher end of the age group.  But the study reports that the average age of the English participants was actually a little higher than that of the American participants, so that’s unlikely to be what’s driving this.  Perhaps the tests were simply conducted differently in each country or the sample are not quite random.

One thing I noticed that I thought was curious was that the clinical data for the American sample was as much as four years older (1999-2002) than the data used for the English sample (2003).  That doesn’t sound like much, but life expectanices are changing a steady clip.  According the the 2004 Time Magazine Almanac, Americans have been adding about  7 months to their life expectancy every four years or so.  It doesn’t strike me as impossible for this to have had some effect on the middle-aged morbidity rates as well.

Another curiosity I noticed was that the authors of the study presented the self-reported morbidity data two ways–once straight and once controlling for obesity, heavy drinking and smoking.  In both cases, the data showed the Americans in a significantly worse position that their English counterparts.  When it came to the much more important clinical data, however, the authors did not present the data controlling for obesity, heavy drinking and smoking (though they said that they ran the data both ways and found that a gap remained).  Presumably the gap is less when those factors are controlled for.  The problem is that we aren’t told how much less and we could have been.

A third curiosity is the clinical definition of hypertension used by the study.  It includes anyone whose systolic blood pressure is 140 or higher or whose diastolic blood pressure is 90 or higher or who is taking medication for hypertension.  That’s a sensible definition for many purposes.  But it is not a useful definition if one is trying to suggest that one country’s medical care is better or worse than the other’s (as some readers will probably try to suggest.) A person whose blood pressure is under control on account of medication is well, not ill.  His situation reflects well on his medical care providers, not poorly.

I have a bit more to add later.