Posted by
Andrews on Monday, June 22, 2009 2:37:19 PM
Best of the Web Friday had an interesting article (just got it as I was on vacation this weekend):
A Toast for the Troops
"The rate of Army soldiers enrolled in treatment programs for alcohol
dependency or abuse has nearly doubled since 2003--a sign of the
growing stress of repeated deployments in Iraq and Afghanistan,
according to Army statistics and interviews," according to the lead
story in USA Today:
Soldiers diagnosed by Army substance abuse counselors with
alcoholism or alcohol abuse, such as binge drinking, increased from 6.1
per 1,000 soldiers in 2003 to an estimated 11.4 as of March 31,
according to the data. The latest data cover the first six months of
the fiscal year that began in October.
"We're seeing a lot of alcohol consumption," Gen. Peter
Chiarelli, the Army's vice chief of staff, told top officers during a
briefing on the Army's growing number of suicides.
This is pretty much a cookie-cutter story: Note the rising
prevalence of some sort of personal difficulty among servicemen or
veterans since 2003, and attribute it to the "stresses of combat." But
buried in the USA Today story is another statistic that does not fit
the mold:
Enrollments in drug abuse treatment programs have remained
largely unchanged in the Army during the war, rising from 3.7 per 1,000
in 2003 to an estimated 4.2 as of May.
Could it be that the abuse of booze is on the rise not because of
generic "stresses of combat" but from one particular stress of combat
in Muslim countries--to wit, the inaccessibility of alcohol? Perhaps
the inability to drink moderately while on deployment prompts some
soldiers to overdo it when they come home.
What makes this interesting is not their conclusion, valid as it may be*, but that it provides another interesting way that measures may mislead. Recently, I mentioned how using subjective numbers to compare two nations may lead to incorrect conclusions. And earlier, I argued that the number of cases of autism may be rising not because of any real rise, but because of an increase in diagnoses.
And this is another case in point. While most physical disorders are diagnosed based on very definite criteria**, psychological disorders are diagnosed based on largely subjective criteria. And alcoholism is a case in point. Where, at one time, it was largely confined to the traditional "drunk", the guy who never stopped drinking, it now includes "binge drinking", "patterned drinking" and even those who just "feel they can't control their drinking". People we now consider alcoholics would never have been seen as such a few decades ago.
So, is it not possible that this increase in numbers has nothing to do with the number of alcoholics and instead reflects a shift in the frequency with which alcoholism is diagnosed?
-------------------------------------------------------------------------------
* It is interesting as I argued in "Unintended Consequences I" that restrictive alcohol rules encourage binge drinking, which is rather similar to the argument they make here.
** Even objective measures are not immutable. What qualifies as "pre-diabetic" has been gradually lowered, making those who were once "normal" now sick. Similarly what is "obese" has gradually changed (as I mentioned in "Twice in a Row"), and even what is "healthy" cholesterol has changed with time. However, as we have numeric values in those cases, we can still compare some objective measure, as opposed to psychological diagnoses which are wholly subjective. (And before someone objects that diagnoses are based on set criteria, I would argue those criteria are still subjective. What is the objective measure of "paranoid ideation" or "lack of affect"? They are still judgment calls at some level. See "Finding What You are Looking For")
-------------------------------------------------------------------------------
POSTSCRIPT
I am not suggesting this is entirely related to the changing criteria for a diagnosis of alcoholism, or even that doctors and others are more willing to identify alcoholism. But that does need to be considered when dealing with diagnoses based on relatively subjective criteria.
Actually, there is one other question it would be interesting to have answered. Between 2003 and the present, has there been any push to help identify and treat alcohol issues among the troops? Such awareness programs can often have a dramatic effect on the number of cases reported, even when the underlying situation has not changed.
POSTSCRIPT II
The following relate to the topic of changing measurements leading to mistaken conclusions:
A Thought on Oil Reserves
Knowing Our Limits
Transfats?
Statistical Artifacts
Endangered Species
The World's Most Stupid Bureaucrat
A Brief Comment on Oil
Why I Doubt Peak Oil Predicitons
Some Global Warming Links
Several Convenient Untruths
Rejecting "Peak Oil"
Safe Nuclear Power
Organic Absurdities
Sampling Changes and Fictional Trends
Pet Peeve
Once Again, Confused by Our Own Data
Twice in a Row
Historic Myopia
More About the Hockey Stick Graph
The Failure of Peer Review
Allergies
A Dearth of Common Sense
Again Improving Science Misleads
The Perfect Model
How Green is Green Energy?
Debunking "Debunking Global Cooling"
More On Biodeisel
Interesting Evidence
Bad Science and Environmentalism
An Interesting Article
Two Questions About Health Care
A Useless Measure
I would also recommend readers check out "
Mental Illness", as it is my best presentation of my objection to modern conceptions of alcoholism.