Almost heaven. (Photo by Linda Davidson / The Washington Post)

From the Summer of Love in the 60s to a thriving marijuana scene today, you might consider California the epicenter of American drug culture. But an interesting new study from Gallup finds that among the states, the use of "mood-altering drugs" is highest in West Virginia. California had the third-lowest rate of drug use on this measure.

Researchers posed the following question to at least 450 adult respondents in each state: "How often do you use drugs and medications, including prescription drugs, which affect your mood and help you relax -- almost every day, sometimes, rarely or never?" Crucially, the exact meaning of "drugs and medications" was left to the respondents. "While the question specifically refers to drugs that 'affect your mood or help you relax,' the interpretation of that description is left up to respondents and could include prescription drugs, recreational drugs, alcohol or nicotine," the Gallup researchers write.

According to the survey, 28.1 percent of West Virginians said they altered their mood with drugs almost every day, the highest percentage of any state. They were followed by Rhode Island, Kentucky, Alabama and Louisiana. Six of the top 10 drug use states are in the South. At the other end of the spectrum, Alaska had the lowest rate of frequent use at 13.5 percent, followed by Wyoming, California, Illinois and North Dakota.

Nationally, 18.9 percent of Americans say they take drugs to relax almost every day, while 62.2 percent say they never do.

But before we draw any broad conclusions let's go back to that question wording, because it's really important here. It's almost certain that a large number of respondents didn't consider alcohol or tobacco when answering the survey. We know, from the most recent National Survey of Drug Use and Health, that at least 71 percent of American adults drank in the past year, and 56 percent drank in the past month. So that 62.2 percent abstinence number doesn't jibe with what we already know about American drinking -- to say nothing about tobacco or illicit drug use.

So if the question doesn't measure drinking, exactly what kind of behavior is it measuring? That's harder to say. Over at the Fix, Philip Bump notes that there's some correlation between state drug use and median age, although the linkage isn't particularly tight. Gallup found that high rates of drug use are linked to lower measures of happiness, as measured in its Well-Being Index. The most likely explanation there? Americans who are less happy are more likely to be taking drugs -- prescription or otherwise -- to address that problem.

But illicit drugs like marijuana clearly aren't driving these numbers. Colorado and Alaska have some of the highest marijuana use rates, but among the lowest percentages on this question. And the Southern States scoring the highest on the Gallup question have some of the lowest overall illicit substance use rates.

My best guess is that the question is capturing some combination of prescription anti-depressant and anti-anxiety medication use, along with a fair share of recreational drug use. But that's only a guess, and it's impossible to tell for certain.

One quirk of our national drug policy is that it treats alcohol and tobacco as separate from "drugs." Hence we have a "Bureau of Alcohol, Tobacco and Firearms," and a "Drug Enforcement Administration." From a scientific and public health perspective, this is questionable: Alcohol and tobacco interact with your brain chemistry just like pot, heroin and oxycodone do. If crack-cocaine is a drug, so is an appletini.

It would be nice to have a survey question that captured the totality of this drug use. It would help us better understand the universe of American substance use, and it would also help inform the public that a drug is a drug, whether you inject it, smoke it, or enjoy a glass of it with dinner. This Gallup question is an interesting first step in that direction, but as its currently worded it leaves too much to the vagaries of individual interpretation.