Initial Commentary on SSRI and Alcohol Use Disorder Recommendation

Note: this is my initial reaction to one very narrow part of the guidelines. There are definite risks in adding alcohol to antidepressants and vice versa! I’m also not a doctor, clinician, or health researcher, nor do I play one on the Internet. I’m raising what I believe to be legitimate questions about the validity of one recommendation that seem to warrant additional objective consideration and further discussion.

As you’d expect, the recommendation from the new alcohol guidelines published in CMAJ getting the largest press is the one about antidepressants.

See: Canadian guideline for the clinical management of high-risk drinking and alcohol use disorder.

The (strong) recommendation reads: “Prescribing SSRI antidepressants (ADs) is not recommended for adult and youth patients with alcohol use disorder (AUD) and a concurrent anxiety or depressive disorder.”

While it says it’s based on “moderate” evidence, this may be the weakest part of the paper.

The main data is an incidental finding in a RCT (n=265) relevant to a large subset (60%) of participants. The study looked at citalopram (an SSRI) and AUD.

The other main ref was a study about a non-SSRI (trazodone, a tricyclic antidepressant useful for insomnia but not depression).

A few other studies were cited, comparing citalopram/escitalopram or sertraline with medications designed to treat AUD, e.g., naltrexone, while people were also undergoing CBT, alcohol treatment, etc. Depression generally improved with all groups, btw.

The evidence has been clear for a while that ADs aren’t a useful treatment for AUD. No question. The finding that they MAY increase alcohol use is not well known but important.

But does that justify not using ADs when AUD is present?

It does argue that alcohol use be monitored during treatment (duh) and if it’s increased that AD use may be a factor, and if so, should be addressed.

Again, non-obvious or well known so worth highlighting.

Not using at all? The argument would work if there was no evidence that ADs improved depression when AUD was present.

The incidental finding from the n=265 RCT found this. Unfortunately, the evidence is very mixed.

I did’t see studies about SSRIs/anxiety and AUD at all.

Other studies, even on the same medication (citalopram) have found different results, e.g. in STAR*D (much, much larger data set).
See e.g. Does comorbid substance use disorder impair recovery from major depression with SSRI treatment? An analysis of the STAR*D level one treatment outcomes.

It showed same response with or without substance use (includes breakdown of AUD/other).

There are always a huge number of potential confounds in mental health studies involving study population, other concurrent treatments, etc. And AUD has an effect on other factors, e.g. nutrition, that are known to affect depression treatment. (I’ve talked about this before, e.g., here).

Very weakly controlled for here in these studies.

In summary, is there evidence to conclude that SSRI’s shouldn’t be used for anxiety or depression when AUD is present?

Not based on one study of one SSRI (with others studies showing conflicting results) and one non-SSRI (trazodone, not used for depression).

This does not take away from the other guidelines around diagnosing and treating substance use. In this context, taking a substance use history as part of a mental health history (should be standard) is important. As is monitoring treatment and addressing setbacks (duh).

But this will be interpreted as “if I drink, I shouldn’t be prescribed an SSRI AD.” This will have widespread clinical implications in practice.

Based on what I’ve seen so far, which is doubtless incomplete, that conclusion is not supported.

Use of any AD requires monitoring to determine its effectiveness (trial and error is involved) and side effects, including on substance use. History and followup are critical. Again, duh.

But a blanket recommendation against AD use when AUD present? Doesn’t make sense.

Editorial: Medicine has its tribes like every other field, and I suspect that was a factor here. It will be interesting to see how other tribes (e.g. psychiatrists) respond.

And yes, I’m holding my breath waiting for my fat cheque from big pharma now.