PTSD With Comorbid Alcohol Use Disorder: Concurrent or Sequential Treatment?
The older model treated substance use disorder first and PTSD later. Current evidence favors integrated, concurrent treatment of both — but the pharmacologic evidence for the most commonly used PTSD-specific agent in this exact comorbidity is itself disappointing.
A.G. is a 38-year-old woman, a hotel event coordinator, whose PTSD followed years of domestic violence in a relationship she left four years ago; she now lives alone and has primary custody of her two children on alternating weeks. Her drinking, which she describes as beginning as "the only thing that turned my brain off at night," has progressed to daily use, most heavily in the evenings, and she meets criteria for moderate alcohol use disorder alongside her PTSD. She has no other medical history and this is her first time seeking treatment for either condition.
The older clinical model would have asked her to achieve sustained sobriety before starting PTSD-specific treatment, on the theory that active substance use would undermine trauma-focused work and that treating PTSD symptoms directly could even worsen drinking by surfacing distress faster than she could tolerate it sober. Current evidence has moved firmly against that sequential model: integrated approaches treating both conditions concurrently, such as Seeking Safety, now show better real-world engagement and outcomes than asking a patient to resolve one condition as a gatekeeping step before addressing the other, which in practice often meant patients dropped out of care before ever reaching the PTSD treatment they came for. Where the evidence gets genuinely mixed is on the pharmacologic side specifically: the placebo-controlled trial testing sertraline in patients with comorbid PTSD and alcohol dependence found no significant overall benefit over placebo, but a post hoc cluster analysis split the sample in a way that matters here. Sertraline-treated participants with less severe alcohol dependence and early-onset PTSD had significantly fewer drinks per drinking day; it was the group with more severe dependence and later-onset PTSD in whom placebo did better. A.G. — moderate alcohol use disorder, PTSD clearly predating the drinking — falls on the side of that split where sertraline outperformed placebo, not the side where it underperformed. That is a post hoc subgroup finding from a 94-patient trial and cannot carry much weight on its own, but it is the opposite of a reason to rule the drug out for her.
At the intake visit
I want to be direct that the old "get sober first" model is not what the current evidence supports, and I don't want her leaving today thinking she has to fix her drinking on her own before we'll treat her PTSD. Integrated treatment, addressing both concurrently, has better real-world engagement precisely because it doesn't ask her to clear a gate she's already told us she couldn't clear alone before.
Agreed on the concurrent-treatment model, but I want to be precise about what the sertraline trial actually showed, because it is easy to remember it as simply negative. Overall, yes — no separation from placebo. But the subgroup that did benefit was the one with less severe alcohol dependence and early-onset PTSD, and her symptoms trace back to the relationship itself rather than emerging after the drinking started. If that post hoc split means anything at all, it points toward her, not away.
That negative pharmacologic finding is specific to the SSRI-in-comorbid-AUD question — it doesn't undercut the concurrent-treatment model itself, which is really about psychotherapy structure more than about which pill to add.
Then the honest plan is: Seeking Safety as the concurrent, integrated core of treatment, and naltrexone for the alcohol use disorder itself, which is the best-evidenced pharmacologic step available to her today. On sertraline I'd offer it rather than either push it or withhold it — the subgroup signal is genuinely in her favor but far too weak to lean on, and starting it at the same visit as naltrexone and a new therapy would make it hard to attribute any change to anything. My preference is to raise it with her explicitly as an available option now and revisit it at her next visit, not to leave the room having quietly decided against it on evidence that does not say that.
Seeking Safety initiated as the integrated core of treatment, naltrexone started for the alcohol use disorder, and sertraline explicitly offered and left on the table for the next visit rather than either started as a default add-on or ruled out.
The team agreed on the concurrent-treatment framework without reservation; the sertraline question was left open on its own separate track, deferred for sequencing reasons rather than evidentiary ones, with the plan to revisit it at the next visit once naltrexone and therapy are underway and the team can better judge which symptoms are still active independent of alcohol's own effects on her sleep and mood.