A New Study Asked People Why They Relapsed. The Answer Wasn’t Willpower.

Nine people sat down for interviews they didn’t have to give. All of them had relapsed after a real stretch of sobriety — not a slip in week two, but a return to use after they’d already done the hard part once. Researchers Mehari, Melese, and Reshid weren’t trying to catalogue their failures. They were trying to answer a narrower, harder question: what does relapse actually feel like from the inside, in the days before it happens, not the moment it happens?

Relapse is not a decision. By the time it looks like one, it’s already the last step in a process that started weeks earlier and went unnoticed by almost everyone, including the person living it.

That’s the finding underneath the study, published in Scientific Reports in February 2026, and it matters because so much of addiction treatment — and so much of the shame that follows relapse — is still built on the opposite assumption: that someone who relapses made a choice at the moment of using, full stop, and the job of treatment is to strengthen the willpower to make a different choice next time. If you’ve relapsed, you’ve probably had that framing handed to you by someone well-meaning, and you’ve probably felt the specific, private failure of hearing “you just need to want it more” about something that had almost nothing to do with wanting.

What “interpretative phenomenological analysis” actually means

The method has a clinical name — interpretative phenomenological analysis, or IPA — but the plain-language version is simple: instead of running a survey and counting answers, researchers sit with a small number of people and let them describe an experience in their own words, then look across all nine accounts for the patterns that repeat. It’s slower and smaller than the large cohort studies that usually make headlines in this field, and that’s the point. A nine-person interview study can’t tell you what percentage of people relapse for a given reason. It can tell you, in detail no spreadsheet captures, what relapse is actually made of — which is exactly the layer most large studies skip past on their way to a p-value.

What repeated across all nine accounts, according to the study, was a complex interplay of personal and environmental factors that had usually been building for some time before the actual use — not a single trigger, not one bad night, but an accumulation: a coping strategy that used to work and quietly stopped getting used, a support relationship that had thinned out without anyone deciding to let it thin, an environment that hadn’t changed enough between treatment and re-entry into daily life. The people in this study weren’t blindsided by relapse in the way the culture around addiction likes to imagine it — a moment of temptation, resisted or not. They described something closer to a slow leak.

The part that should change how programs are built

The study’s more useful finding, for anyone actually running a treatment program, is what the same nine people said helped them get back to stable ground afterward — because relapse, in their accounts, wasn’t the end of the story either. The coping strategies participants named as effective weren’t exotic: self-motivation, physically relocating away from a using environment, rebuilding or leaning harder on supportive social networks, deliberately working on mental resilience, making concrete plans for future stability, and for several participants, an explicitly spiritual or meaning-making practice. None of that is new to anyone who has worked a program. What’s notable is that it came from people describing what actually got them through it, not from a curriculum written in advance of knowing them.

This lines up with a broader, better-established literature on relapse mechanics. StatPearls’ clinical relapse-prevention review — the reference clinicians actually use — describes craving as the hinge between emotional distress and a return to use, with negative mood consistently identified as the strongest predictor that shows up before relapse across substances, including alcohol and cocaine. A separate 2026 qualitative study of addiction therapists, published in a SAGE journal using the same IPA method, found something adjacent from the clinician’s side: therapists supporting clients through repeated relapse described the experience as something that could feel, in their own words, “almost as if I’ve relapsed” — a secondary strain on the people doing the supporting that treatment programs rarely account for when they staff and pace this work.

What’s still unproven

Nine interviews is nine interviews. The authors aren’t claiming these six coping strategies are the definitive list, or that they’ll generalize to every population, every substance, or every treatment setting — the sample is small, self-selected, and drawn from people willing to revisit a difficult period on the record, which is its own kind of selection bias. What a study like this earns is not statistical power. It earns detail, and permission to describe an experience the field has mostly reduced to a single word on a chart.

It earns detail, and permission to describe an experience the field has mostly reduced to a single word on a chart.

If you’re reading this after a relapse, the study’s plainest finding is this: the leak was probably there before you noticed it, and noticing it late doesn’t mean you weren’t paying attention — it means the thing that leaked was quiet by design, the way a coping habit you stop using doesn’t announce its own absence. That’s not an excuse. It’s a more accurate map than “you didn’t want it enough,” and a more accurate map is the thing that actually helps you catch the next leak sooner. Explore recovery support and relapse-prevention resources →