Pharmacological management of gambling disorder: A systematic review and network meta-analysis
Ioannidis, K., Del Giovane, C., Tzagarakis, C., Solly, J. E., Westwood, S. J., Parlatini, V., Bowden-Jones, H., Grant, J. E., Cortese, S., & Chamberlain, S. R. (2025). Pharmacological management of gambling disorder: A systematic review and network meta-analysis. Comprehensive psychiatry, 137, 152566. https://doi.org/10.1016/j.comppsych.2024.152566
Woohoo — happy New Year. I’m personally pretty glad 2025 is over and done with. It was a tough one for me, and I’m not fully buying the whole new year, new you thing… but I am buying the new year, same journal club blog thing. Today’s paper felt like a good pick because gambling disorder is a blind spot for a lot of clinicians (and, honestly, sometimes for me too). It isn’t a “chemical” addiction in the way people usually imagine drugs or alcohol. It’s a behavioural addiction, and that can make it harder to recognise, harder to explain, and sometimes harder to treat.
I also can’t talk about gambling disorder without naming the uncomfortable context: this isn’t just about individual vulnerability. Modern gambling products are engineered to keep people playing, and many communities carry the harm while industries collect the profits. I’m increasingly uneasy about how “gambling-like” design shows up in everyday life too — variable rewards, near-misses, flashing notifications, timed “deals,” and a constant nudge to “try again.” Whatever language we use diagnostically, I don’t want to lose the ethical baseline: people with gambling disorder are not morally weak, and they’re not “the perpetrator.” They’re often the ones being targeted — and harmed.
The article I read is “Pharmacological management of gambling disorder: A systematic review and network meta-analysis,” published in Comprehensive Psychiatry (Ioannidis et al., 2025). I always scan conflicts of interest early because it helps me interpret tone and emphasis. The authors disclose various professional and editorial relationships, and one author (Dr Grant) reports research support from pharmaceutical companies and other academic income streams. Importantly, they explicitly state no conflicts related to the gambling or gaming industry, and no voluntary donations from those industries.
So what is a network meta-analysis, in normal language? Most people have heard of a meta-analysis, where researchers combine results from multiple studies to get a clearer picture than any single study can provide. A network meta-analysis goes a step further. It’s like running a tournament: instead of only comparing Drug A to placebo and Drug B to placebo in separate lanes, it uses both direct and indirect comparisons to estimate how A stacks up against B, even if there aren’t many head-to-head trials. The upside is that it can make sense of a scattered evidence base. The downside is that it relies on assumptions and the quality of the original studies still matters a lot. The authors used a pre-registered protocol, searched both published and unpublished trials, and updated their search through February 2024.
In the end, they included 22 randomised controlled trials in the systematic review, and 16 trials (977 participants) in the network meta-analysis itself. They focused on adults with gambling disorder or pathological gambling as the primary diagnosis, which is scientifically tidy but clinically a bit frustrating because in the real world many people have comorbid depression, anxiety, ADHD, substance use, trauma histories, and other factors all tangled together. Still, this approach helps answer the question: if we isolate “gambling disorder” as the main problem, what medications actually seem to help?
The medications they examined included opioid receptor antagonists such as naltrexone, nalmefene, and naloxone; SSRIs like paroxetine and fluvoxamine; topiramate; bupropion; olanzapine; and a plant-based antioxidant (silymarin). Other medications appeared in the broader review but couldn’t be analysed in the network meta-analysis due to limitations in the available data, including lithium, sertraline, clomipramine, baclofen, acamprosate, and N-acetylcysteine.
The headline result, in plain English, was that nalmefene showed the strongest reduction in gambling severity compared with placebo, with naltrexone next. What mattered to me clinically is that they didn’t just look at symptom scales — they also looked at quality of life, and both nalmefene and naltrexone were associated with improved quality of life compared with placebo. That’s a meaningful outcome because “feeling less driven to gamble” is important, but so is “Is your life getting better? Are your relationships stabilising? Are you sleeping? Are you less panicked about money? Are you reconnecting with who you were before this took over?”
But there’s a real trade-off here: people were much more likely to drop out due to side effects on nalmefene and naltrexone than on placebo. That doesn’t mean these medications are “bad,” but it does mean that if I’m using them, I have to do it thoughtfully — slower titration where appropriate, careful expectation-setting, close follow-up, and a shared plan for what we’ll do if side effects show up. Interestingly, despite previous interest, this analysis did not find olanzapine or topiramate to be more effective than placebo for gambling severity. Given olanzapine’s side-effect profile, I’m not upset by that finding.
My take-home is that if medication is being considered for gambling disorder, the best current evidence sits with opioid antagonists (especially nalmefene and naltrexone), but tolerability can limit their usefulness, and the evidence base still isn’t as robust as we’d want for such a major public health problem. Medication also isn’t a standalone fix: the broader ecosystem matters — psychological treatment, practical harm-reduction, financial counselling, social supports, and addressing comorbid conditions that amplify risk. And I keep coming back to the same ethical point: we can and should treat gambling disorder seriously as a psychiatric condition, but we should also name the systems that profit from human vulnerability. You can support recovery without blaming the person who got harmed.
If you or someone you love needs support in Australia, the National Gambling Helpline is 1800 858 858, and if things feel unsafe or overwhelming, Lifeline is 13 11 14.