Andrew Sullivan is concerned that the new DSM-V psychiatric classification system will define the 40% of college students who sometimes binge drink as alcoholic. He links to an angry critique by Russell Blackford, who sees the potential over-extension of the alcoholic label as misguided and oppressive.
As I outlined a few days ago, and Dr. Deni Carise elaborates on today at Huffington Post, the alarming 40% figure that keeps being quoted is misleading. If criteria were broadened, the purpose would be specifically to identify problem drinkers who are not what is commonly understood as alcoholic. Deni writes:
…never fear, the new DSM will not cause more people to be diagnosed with addiction. Instead, more people who may not yet be addicted (but whose drug use is nonetheless problematic and unhealthy) will be able to access very inexpensive but proven effective treatment earlier and easier. Treating these folks is no different from treating those in the early stages of diabetes — it requires minimal professional help, some education, and simple lifestyle changes. We wouldn’t wait until a pre-diabetic started experiencing the symptoms of full-blown diabetes before we offered him or her help. Instead, we would intervene early in hopes of preventing such a difficult future. The same should apply for those with early substance misuse.
In short there is no chance that the new DSM will lump together unhealthy drinkers with people who have drunk very heavily for years and are now physically dependent (what we commonly think of as “alcoholics”). So if that worry is keeping you awake at night, roll over and sleep the sleep of the just.
This reality isn’t germane to Blackford’s other criticism, which is that by labeling certain patterns of drinking as unhealthy, medicine is inappropriately condemning particular lifestyles. I have heard this argument many times before. It rests on the false assumption that a statistical relationship is equivalent to a moral judgment imposed by outsiders.
A typical debate of this sort runs roughly as follows. An epidemiologist gets up at a conference and says that people who drink at level X have, say, twice as much chance of dying in a car accident than people who drink at level Y. Then someone in the audience says “My friends and I drink like that. We all have jobs and are responsible parents, how dare you call the way we drink risky! Who are you, doctor, to decide that my drinking is wrong?”.
But the people who decide that level X of drinking raises risk of an auto accident aren’t epidemiologists or any other elite professionals. Rather, the statistical relationship between the behavior and risk is set in the most democratic way imaginable: Tens of millions of people drink and then decide whether or not drive and then some of them have accidents and some don’t. The epidemiologist just counts the ballots.
The truth value of a statistical relationship between drinking (or any other behaviour) and bad outcomes is thus independent of whether anyone feels judged by it. Plenty of doctors drink in a risky fashion and may not like the relationship between heavy drinking and bad outcomes either. But it just doesn’t matter what those doctors (or anyone else) wish were true; the relationship between unhealthy drinking and bad outcomes reflects what millions of drinkers have actually done and the collective consequences of their behaviour.