Heterogeneous Dose-Response, Alcohol Abuse and Dependence

Supplement Directors: Paul J. Gruenewald, Ph.D. and Christina Mair, Ph.D.


“You only drink as much as you can . . . but sometimes more than you should.”

Understanding why people drink, what causes them to drink as much as they do, and why some drinkers continue to drink despite the many problems that drinking may cause in their lives, remains a key concern of alcohol researchers.  We know that some drinkers “abuse” alcohol; their drinking leads to many family, work and social problems.  We know that other drinkers are alcohol “dependent”; they drink despite these problems, drink compulsively, drink to avoid or reduce other problems, or experience “alcohol withdrawal;” additional problems arise when they don't drink, causing them to start drinking again.  The “spiral of addiction” that this may cause can lead to significant health problems, liver cirrhosis, and death.

This project looks at what it means to be a “normal” drinker, someone who is “normally” responsive to alcohol's effects:  Most people “only drink as much as they can.”  When they start having problems they reduce how often or how much they drink so they don't have those problems in the future.  Some drinkers, however, do not appear to respond to the problems that occur when they drink.  We want to know why this is true and, if we can, find a way of measuring how different drinkers respond to problems related to their drinking.  To do this, we built a simple mathematical model of drinking behavior that allows us to conduct these measurements using available survey data.  We can then take a look at differences between “normal” vs. “dependent” drinkers.


Research Goals and Activities

We built the simplest mathematical model of drinking that we could in order to characterize “normal” drinking behaviors.  We just assume that whenever someone drinks, Qt, these lead to problems, Pt.  Then, when the same drinker drinks again she takes previous consequences of drinking into account; if she experiences more problems, she drinks less; if she experience no problems, she might drink the same amount or more.  Under normal circumstances, the problems she experiences will moderate the amount she drinks.  The diagram to the right pictures these relationships.

We use this model to take a look at what kinds of control over drinking “normal” drinkers exhibit.  “Normal” drinkers respond to the behavioral controls in their alcohol environments; if drinking leads to more problems, these drinkers drink less.  They only drink as much as they feel they can to avoid problems. 

Some drinkers, however, may drink a great deal more because they need to do so to continue to experience the pleasurable effects of intoxication (tolerance) or to avoid problems related to less use (withdrawal symptoms); these drinkers experience unusual benefits related to drinking, Bt, and drink more than they should.  This is one very naïve way of describing what happens when someone is addicted to alcohol or alcohol dependent.  The diagram also shows that these benefits will moderate subsequent drinking. 

We use this aspect of the model to take a look at what kinds of control over drinking “dependent” drinkers exhibit.  “Dependent” drinkers also respond to the behavioral controls in their alcohol environments, but much less soThey continue to sometimes drink much more than they should. 

In some ways all these observations may be common knowledge.  All we've done is come up with a mathematical way of showing these relationships.

That would be unimportant except for two curious observations: 

  • “Dependent” drinking occurs among drinkers who don't drink very much, those who may typically drink only 2 or 3 drinks. 
  • Many very heavy drinkers, those who typically drink up to 12 drinks are not “dependent.”
So we think we can use this model, or something like it, to begin to sort out who is and is not at risk for alcohol dependence regardless of average drinking levels.  This would help all of us.

Discoveries (Findings)

We are still working on this, but herehold-rsquo;s what we know so far:

  • Most college drinkers appear to show predictably “normal” levels of response to the problems that occur when they drink.  When they have more problems they drink less.
  • But heavier college drinkers, those who drink much more, are generally less responsive.  The problems they experience when they drink have less of an effect over their drinking levels. 
  • We call this “heterogeneous dose-response” because heavier drinkers respond less to the problems that occur when they drink.
  • It also appears that “dependent” drinkers, those drinkers who tell us they have many problems related to drinking but continue to drink, are much less dose-responsive; they do not appear to factor the consequences of their drinking into future decisions to drink.

What can we do with this information?

All this research is in a very early stage and quite technical, but we think there are a few things that you should keep in mind:

  • First, it looks like the amount you drink is not the best predictor of whether you abuse or are dependent on alcohol.
  • Some moderate and lighter drinkers have many problems and exhibit symptoms of alcohol dependence. 
  • Second, if we step back and look at all the drinkers in the US, most problems are related to moderate, not heavy, drinking.
  • Most drinkers drink at these levels most often, even the so-called “heavy drinkers.”  So that is where most problems arise.
The last point is critical for prevention:  If we are interested in reducing problems for individual drinkers, of course they should drink less.  If we are interested in reducing problems for all drinkers, moderating moderate use, where most of the drinking and most of the problems occur, would be very helpful.