Moderate drinking: A destination or way station on the road to abstention?

My comments on “The False Gospel of Alcoholics Anonymous” in The Atlantic

Gabrielle Glaser’s lengthy article in the April 2015 issue of The Atlantic, “The False Gospel of Alcoholics Anonymous,” (online as “The Irrationality of Alcoholics Anonymous“) is something of a misnomer, as its focus is on the failings of the treatment industry as a whole, including over-reliance on AA’s 12 Steps, interspersed with anecdotes from people interested in reducing or moderating their drinking. The dramatic tension in the article is between abstinence-based models (AA and inpatient treatment) and harm-reduction or controlled drinking models for people who may or may not meet DSM-V diagnostic criteria for alcohol dependence.


This is a common theme that crops up regularly in the media and is promoted heavily on the web – alternatives to AA that either teach people how to drink “normally” or offer narcotic medications as alcohol substitutes. While Glaser identifies several valid criticisms of the current system, the author’s focus is on promoting alternatives to abstention for people who very likely should be putting the cork in the bottle and leaving it there.


Here’s my quick take:



Moderation/experimentation can be useful if there is an adequate assessment process (including drug testing and input from family members) and accountability in the event of failure. Unfortunately, Glaser overstates its benefits and understates its risks. In my view, it’s never appropriate for drinkers with severe disorders or co-occuring mental health concerns.



We need to be clear what kind of drugs we’re talking about here. Drugs such as naltrexone that are not cross-addicting (unlike, say, Valium) can be helpful to some people in reducing their drinking or abstaining and are available at many treatment centers. But replacing addiction to alcohol with addiction to benzodiazepines should not be considered a success.


Treatment outcomes

In terms of recovery rates, programs for pilots and doctors have far superior outcomes to all other programs, and we should be talking about implementing their programs for all. Arguing about AA success rates and other treatment models for which there’s no data seems at best subjective, at worst pointless.


Alternatives to AA

Yes! Patients/clients should be offered material without references to God so they can participate in treatment center activities with their peer group. They should also be offered alternatives to AA meetings, such as Smart Recovery and Women in Sobriety. But let’s not forget that AA (and 12-Step programs) has worked wonders for a lot of people; there’s no need to completely trash it out. Additionally, from what I’ve seen, the article exaggerates AA’s role in most treatment programs.


Range of options in community-based settings

In an ideal world, a patient’s learning skills and preferences would be matched with a menu of treatment resources. With the ACA, many more options are in place, including groups for people who are thinking about stopping drinking. For example, in my hometown, two miles down the road from Dr. Willenbring’s office, Hazelden is building a brand new outpatient megamall. The one-size-fits-all criticism is now outdated in many communities.

A framework for evaluating addictive behavior

Whenever information is presented about someone with an apparent addiction, it’s important to understand the framework used to diagnose and assess the concern. My approach is grounded in medically-based information on the effect of alcohol or drugs on the brain, the American Psychiatric Association DSM-V Diagnostic Manual, and an evidence- or facts-based evaluation model, including drug testing.

The starting point for evaluating articles about drinkers or drug addicts

The starting point for sorting through these advocacy pieces, then, is the disease model for substance use disorders and contemporary “best practices” as applied to the four anonymous prototypes in the article: 

  • G. who used to drink a liter of Jameson a day but now does not drink and is taking Baclofen and Valium, as needed for anxiety.
  • Claudia who takes naltrexone and now stops at one drink.
  • Jean who was drinking a bottle of red wine a day and now has an occasional drink.
  • P who had as many as 20 drinks at one time, and by using naltrexone binges two or three times a month.

Brain changes due to drinking

Consistent with the mainstream medical perspective, the article reports that addiction results from changes in brain structure due to long-term excessive drinking:


Each time [a person drinks] the endorphins released in the brain strengthen certain synapses. The stronger these synapses grow, the more likely the person is to think about, and eventually crave, alcohol—until almost anything can trigger a thirst for booze, and drinking becomes compulsive. (p. 55)


While alterations in the brain can be seen through imaging, scans are expensive and not usually used in the diagnostic process. Instead, the American Psychiatric Association has developed 11 diagnostic criteria to assess the degree of substance dependence (brain alteration) in the DSM-V.


The DSM-V assessment process uses 11 criteria with three levels of dependence based on symptomology: Severe (6+), Moderate (4-5) or Mild (2-3) alcohol use disorder. Keep these in mind when thinking about controlled drinking, for as the article reports:


Moderate drinking is not a possibility for every patient, and [Dr. Willingbring] weighs many factors when deciding to recommend lifelong abstinence. He is very unlikely to consider moderation as a goal for patients with a severe alcohol-use disorder. … Nor is he apt to suggest moderation for patients who have mood, anxiety, or personality disorders; chronic pain; or lack of social support.


The idea being that if the brain is altered sufficiently, it’s almost impossible to regain control at the 6+ level. These people are candidates for stop-drinking programs, whereas the moderate and mild drinkers are said to be appropriate for the doctor’s managed-drinking program using naltrexone.

Substance use diagnosis first, please

The first question anyone, including writers, should ask our four prototypes is:


What’s your DSM diagnosis? And then, what’s your status regarding related co-occurring disorders?


Current thinking is that people with a mild diagnosis may be able to control their drinking because their brains have not crossed the threshold to permanent alteration. However, people in the moderate category may not be able to consistently regain control, which brings up the harm-reduction/moderation debate in the treatment profession.

Harm reduction

We all know people who meet the moderate/severe category and want to continue drinking. P, for example, is using naltrexone to cut his binging down to three times a month. This type of harm reduction is viewed as a “win.”

  • However, for people like P, who meet the severe or moderate criteria, the risk is that they can revert to their old patterns, drink and drive, or hurt their family members.

Even with reduced drinking, they may still retain the emotional disorders associated with greater use (i.e., the dry drunk). Harm reduction then becomes a tolerated permanent status rather than a stopover on the way to abstention. Maybe not such a great win for P’s liver, spouse, children, or employer. But who knows?


J.G., who is now “not drinking,” is prescribed Valium for his anxiety. Anyone in the treatment field knows Valium is commonly used for alcohol detoxification because it hits many of the same brain receptors. It’s alcohol in a pill form and often used in place of alcohol by people who “don’t drink at all.” It’s odorless and easy to obtain over the Internet or through doctor shopping.


Replacing alcohol with Valium is simply substituting one addictive substance for another and is not considered abstinence as defined in the DSM-V or mainstream addiction medicine. But there are doctors and treatment centers that prescribe benzodiazepines (like Valium) to people with substance use disorders and claim their patients are in recovery – on maintenance therapy.


The author goes very light on this topic and does not even mention the very real possibility of cross-addiction. She also fails to note that Baclofen is associated with significant withdrawal symptoms and cautionary side effects and is in need of further study for use in treating alcoholism.

Addiction-certified doctors

The Valium script issue highlights the need for training on addiction in the medical profession. However, the number cited, 582, is incorrect, as there are over 300,000 members of the American Society of Addiction Medicine (ASAM), according to its website. With more addiction-savvy doctors, fewer patients will be prescribed benzos to help them control their drinking.

Assessment information: corroboration and BAC levels

One problem omitted from the article is the difficulty of assessing alcohol use when drinkers self-report in the absence of outside verification from family and friends or drug tests. Without such a system in place, moderation and experimentation can easily be a cover for continued excessive drinking and using. (For example, courts in Minnesota require three collateral sources.) “Trust but verify” through corroboration and random, supervised testing is state-of-the-art these days and would aid in determining whether Jean and Claudia meet the moderate, mild, or no diagnostic standard.

Experimentation/Controlled drinking

Drinkers in the moderate and mild category are said to be candidates for experimentation with or without the aid of naltrexone. By helping clients set behavioral expectations and then reviewing progress in meeting their goals, counselors can play a key role in building awareness regarding negative use patterns. This approach works particularly well with young adults.


This kind of “experimentation,” as discussed in the article, where people see if they can stop for a time, drink less, or drink on fewer occasions can be a good learning experience, as the author points out. It helps in breaking through the self-perception hurdle – over 90 percent of people with addictive disorder issues do not perceive themselves as having a problem – and it increases willingness to seek additional help when goals are not met.


As Glaser points out, these techniques, including prescribing Naloxone, should be part of the repertoire of outpatient clinics and individual addiction counselors these days. And, since most people who drink excessively are not alcohol dependent, an evaluation process is preferable to going directly to inpatient treatment, as can happen in the surprise intervention model.

Experimentation, to what end?

In my experience, people who “experiment” and fail don’t usually decide to increase their level or intensity of treatment, unless there is a written agreement in place or there is pressure from family or an income source (e.g., a trust or employer). In other words, insight alone is often insufficient to induce behavior change.


This leads to another concern: People don’t have alcohol and drug problems in isolation. It’s a family/community problem, and this larger group’s interests and opinions need to be taken into account as to how long they are willing to tolerate “failed experimentation.”


When the topic of treatment effectiveness is raised, my starting points are the highly successful programs run by airlines for pilots and medical boards for doctors.

Outcomes – Best practices

The programs for physicians and pilots have proven outcomes of 74 percent continuous abstinence at five years and 92 percent at two years. Proven because the participants are drug tested over that time period. No other programs match these outcomes. The real criticism of the treatment industry should be its failure to apply the physician/pilot programs to all groups.

Treatment compliance/effectiveness

Patient failure to comply with treatment recommendations is a major cause of relapse. Is this due to lack of programs that appeal to clients? The system lacks incentives to encourage compliance? Or is alcoholism simply very difficult to recover from?


For the author, it’s lack of appealing programming. From my view, it’s lack of incentive-based contingency management approach. Reward systems can be very effective in treating even hardcore crack smokers. Again, its common knowledge that the medical boards and airlines (also DUI/drug courts) use therapeutic leverage to encourage treatment compliance, but this approach and these state-of-the-art programs are never mentioned.

$35 billion a year on substance abuse treatment – we need better outcomes

This is one theme of the article: What are we getting for all this money?


The better question is why aren’t treatment centers adopting the pilot/physician model for all patients? The long answer is that inpatient treatment centers tend to be their own program and economic entities. Until they are held accountable for outcomes, they will have little reason to ensure quality, patient-centered options exist in the pre- and post-treatment environment. Again, the ACA appears to be creating incentives for accountability.

Inpatient treatment

J.G. says he went to a center where they offered little more than AA. Being from St. Paul, I am not certain what place he is referring to. It is clearly not a Minnesota Model treatment center like Hazelden, which offers a full range of modalities, including CBT therapy, acupuncture, meditation, education on the disease concept, groups, and mental health-related services (by the way, it also offers naltrexone for craving reduction). He might have attended The Retreat, an AA-only program located in Wayzata, Minn.

Treatment centers promotion of AA and alternatives

Inpatient treatment centers provide a variety of programming during the day (see above), but their residences are organized on the basis of self-regulated peer groups, usually around the principles of AA. Patients are also asked to participate in actual AA meetings. That can be a problem for people who don’t like AA or the God part. Some, like me, join in, as it seems to help and is benign. Others object.


Treatment centers should adjust their programming to meet the needs of their patients, rather than vice versa. It is possible to modify the 12 Steps and peer-based activities to eliminate references to God so patients have a choice when participating in group activities. Offering alternatives to AA, such as Smart Recovery and Women for Sobriety, makes sense as well, although it requires staff that is sufficiently talented to oversee and help implement these support groups as an AA alternative.

Post-treatment referrals

To clarify, patients are referred to a range of post-treatment resources, including AA meetings. These include therapy and outpatient or aftercare support groups. Some people who don’t like attending AA only go to their non-AA activities.


As noted in the article, the quality and content of AA programs vary greatly. In my experience, AA works pretty well for middle- and higher-income males, but not for others (e.g., minorities, women, LGBT, the affluent, and the poor) who fare better in more targeted groups. But only a few of these specialized groups exist. I agree that clients leaving treatment should be offered options to referrals to AA.

A range of evidence-based options needs to be offered people seeking help

In an ideal world, a patient’s learning skills and preferences would be matched with treatment resources. The treatment profession is heading in that direction. Until the enactment of the ACA and Wellstone parity implementation, insurance companies underfunded non-residential services. But that is now changing for the better.

AA concepts originally for chronic, severe drinkers, or not?

The in-depth historical analysis in the article helps the reader understand why the 12 Steps of AA are so prevalent in treatment today and why abstention is the only recommendation for those who are “truly” alcoholic. From this analysis, the author argues that AA is a one-size-fits-all program and inappropriate for those with less severe patterns of drinking.


Having interned at a treatment center, I know that staff can use a generic approach for everyone, without regard to their individual circumstances. However, the Big Book is more flexible and in fact has a set of stories entitled “They Stopped in Time,” describing how people came into recovery before they hit bottom. Many drinkers seek help with moderate use disorders and find the Big Book and 12-Step model useful in becoming sober. It’s all in how you approach it – open-minded or closed-minded.

Lack of trained, well-educated staff

Unfortunately, rigidity in implementation occurs because it is so often based on the personal opinions of staff in recovery and AA volunteers. It is this off-putting attitude that is so offensive to many people, like Jean in the article, who then recoils from what might be helpful concepts. It also highlights the lack of professionalism in treatment centers – another valid comment.

AA’s value

AA’s value lies in providing a support group for people with a common problem: how to stop drinking and stay stopped. AA focuses on the present moment – today – versus remorse over the past and anxiety over the future; shame reduction; emphasizing the disease concept; and being accepted for “who you are.” AA also provides a relationship network so people can share personal stories, make connections and feel better. As to the latter point, we now know that intimate group-sharing releases positive endorphins. These are valuable, if not critical elements, in sustaining long-term recovery, which the author overlooks in her critique of AA.

Addiction as a “good habit gone bad”

In his book The Power of Habit, Charles Duhigg explains that habits are formed over time and based on rewards from engaging in specific behaviors. Due to changes in brain structure from long-term excessive drinking, alcoholics respond to craving by drinking, and their reward is the alcoholic buzz. To stop this compulsive behavior, mainstream treatment is designed to teach new, healthy responses to cravings and new rewards. (This, by way, is the also the core function of AA.)


The tension between the moderation model and abstinence model has to do with the different responses to stimulus and the reward.

  • Drinkers practicing moderation remain in the same stimulus-reward system where they want to take a drink with the resulting effect on their body and emotions (but with craving synapses hopefully dampened through naltrexone).
  • In contrast, those of us in the abstention model learn an alternative response to picking up a drink – say exercising, with the reward being increased self-esteem or an exercise-induced increase in endorphins.

Abstaining usually involves “transformational change,” in that the person no longer believes alcohol is necessary to live and adopts a new value system.

Leave a Reply

Your email address will not be published. Required fields are marked *