The Third Lane: Notes from EDAC 2026 on Eating Disorder Care

I just got back from the biennial Eating Disorders Association of Canada EDAC 2026 conference in Halifax, and I'm still buzzing from all the inspiring talks and conversations. The theme this year was Harbouring Collaboration, and across two days of presentations, the words “collaboration” and “personalization” came up repeatedly as core missing ingredients in our country's eating disorder (ED) care practices. 

In an amazing presentation on longstanding EDs, Howard Steiger, Linda Booj, and Mimi Israel from McGill's Douglas Research Centre (DRC) quoted UK mental health campaigner, researcher, and expert by experience, James Downs: harm reduction cannot become a form of managed decline. It can be a meaningful step toward recovery for some, and recovery looks different for everyone, but harm reduction itself cannot be the final destination. The line between harm reduction as a means to initiate and maintain motivation to stay in treatment and harm reduction as an endpoint of treatment is the line the entire debate hinges on, and the system, however well-intentioned, does not always operate in the patient's best interests. 

Anorexia is a serious mental illness in which a person restricts what they eat, fears gaining weight, and sees their body in ways that do not match reality. It carries one of the highest mortality rates of any mental illness, with mortality roughly five to six times that of the general population. That sobering fact often overshadows what the long-term data actually shows. Among those who survive, most eventually recover. A 22-year follow-up found that 63 percent of people with AN had fully recovered, sometimes after a decade or more of illness. That said, for a meaningful minority, the illness lingers. These are the people who have done the multiple (re)admissions of inpatient stays, the tubes, the refeeding, myriad outpatient treatments, including the rounds of CBT, FBT, DBT, and the ED/AN still does not let go. For this group, medicine has tended to reach for a label, rather than a path forward: severe and enduring eating disorder (SEED) or SE-AN when referring specifically to the AN presentation. 

Harm reduction in AN means shifting our treatment approach from all-or-nothing “full recovery or failure” thinking to “what would reduce suffering and risk right now, even if some symptoms remain” in deference to longer-term recovery goals. Done well, it does not preclude receiving care, making changes, or working towards full remission. It instead widens the doorway in. A 2021 Filter Magazine profile citing Vancouver-based ED clinical psychologist Josie Geller's work describes a woman called Linda who had been a “revolving-door” admission for years. Geller learned Linda was a cyclist, skier, and runner, and built a “closet plan” where her bike, ski boots, running shoes, and roller blades stayed in the closet, and when Linda wanted them, she ate her way to them. With dietitian consultation, Linda never required hospitalization again nor reached a “normal” weight. She kept working and skiing. Her life got bigger. Geller, who developed British Columbia's eating disorder clinical practice guidelines, has spent two decades arguing that the strongest predictor of treatment success is not how sick someone is, but whether care effectively meets them where their readiness lives. 

Access matters too. Residential treatment can cost thousands of dollars per day, and Canadian waitlists are long. AN in particular remains stereotyped as a young, thin, white cisgender woman's illness, leaving people of colour, queer and trans folks, people in larger bodies, and men chronically under-diagnosed and under-treated. A system that treats recovery as a precondition for support rather than a goal of it has the order of operations backwards, and the people who pay the price are the ones with the least access to begin with. 

Some of the sharpest critiques of harm reduction come from people with lived experience. James Downs has argued that the SEED label can become a self-fulfilling prophecy, locating intractability in the patient rather than in protocols built on narrow, homogenous evidence and delivered in chronically under-resourced care settings. A 2024 meta-synthesis of 447 lived experience accounts across 45 studies found many people rejected the label entirely, calling it “hope-limiting”. Some described being turned away from care based on poor prognosis, and then watching that exclusion deepen and entrench their illness further. 

There is also the term “treatment-resistant,” which deserves more scrutiny than it usually receives. A growing literature on iatrogenic harm in ED care documents how coercive treatment environments, surveillance, supervised meals without consent, threats of care withdrawal, and discharge for “non-compliance” can replicate prior trauma and erode therapeutic trust. Many people who appear “resistant” are actually avoiding a system that has hurt them. That is system-induced avoidance, not patient pathology, and it is a distinction the field is only just learning to make. 

The recent charged debate over "terminal anorexia" sharpened all of this. In an often-cited and contested 2022 article, clinicians proposed criteria that would allow patients access to hospice treatment and, in some jurisdictions, medical assistance in dying (MAID). Rosiel Elwyn, a lived experience researcher who would have met the proposed criteria, has written about how prognoses of futility shaped her clinical reality and nearly cost her life, and about the enduring hope that the "terminal" label foreclosed. Dr. Jennifer Gaudiani, who initially proposed the framework, has since publicly stated that she is no longer using the term, recognizing how easily that language could be misused. Language in this field does not just describe people. In many ways, it proactively conditions what is possible for them. This raises questions around who benefits when these terms are invoked. In an under-resourced and exhausted system, all of these labels can become a convenient way to lower expectations instead of an authentic, collaborative step toward something better, making something like harm reduction too often in practice function like abandonment under a kinder name.

The clinicians I heard at EDAC are not arguing for less structure. They are arguing for the right structure, built around the person more than the diagnostic category. In her impactful keynote, Clinical Psychologist Cheri Levinson made the case that we have spent decades treating EDs as if every patient is the same person, when the data clearly shows they are not. Her work on personalized treatment finds that tailoring interventions to a person's specific symptoms, comorbidities, and life context improves adherence and reduces not only ED symptoms but also depression, worry, and anxiety. This is not the absence of evidence-based care. It is evidence-based care assembled around variables that standard manualized protocols often ignore: cross-cultural differences in how distress and embodiment are expressed, neurodivergence that changes how exposure and interoception work, intergenerational and complex trauma that shapes both the illness and the response to treatment, gender identity, body size, sexuality, and the many complex ways these various factors intersect. When we say a treatment "failed," we often mean a one-size-fits-all protocol failed a person it was never designed for. This is more aptly a misapplication or implementation issue than some type of inherent patient resistance.

Clinical Psychologist Anita Federici made a parallel case from a different angle. In a compelling workshop introducing the MED-DBT protocol (Multi-Diagnostic Eating Disorder Dialectical Behaviour Therapy), which Federici developed together with Lucene Wisniewski, she shared how this approach is built explicitly for people whose presentations have been treated as too complex for standard care, such as patients with severe affect dysregulation, trauma, suicidality, BPD features, neurodivergence, and varied comorbidities. MED-DBT integrates DBT skills, exposure work, and a stage-based collaborative contingency management framework that explicitly accounts for ambivalence rather than treating it as non-compliance. Its operating premise is clear: when treatment fails, it reflects the limits of our methods, not the limits of the patient. The model is built on collaboration, flexibility, and patient responsibility. Federici emphasized that working with a patient is what is most likely to produce durable change. 

This is where the most serious clinical innovation in the field is happening right now. Models like MED-DBT, the integrated trauma-informed approaches, the staged care frameworks being developed for SE-AN, and Levinson's personalized treatment work share a premise that the conversation about harm reduction has often obscured. The binary of “standard recovery-focused care versus harm reduction” misses a third lane, which is complexity-aware, evidence-based care designed explicitly for the populations that conventional protocols were never built to serve. Many people who have been called treatment-resistant have not had access to this new kind of personalized care. Many would likely defy the intractable labels given to them if they did. 

None of this harm reduction talk is happening in a vacuum. For the first time in over a decade, thinness has become aspirational again. Vogue Business's Spring/Summer 2025 size inclusivity report found that of 8,763 looks across 208 shows, less than 1 percent were plus-size, and nearly 95 percent fit the industry's smallest standard sample size, roughly US 0 to 4. Likewise, before TikTok banned the hashtag in June 2025, #SkinnyTok had amassed tens of thousands of posts and millions of views glorifying extreme thinness, and similar content continues to spread under altered tags. GLP-1 drugs like Ozempic and Wegovy are aggressively marketed for weight loss, and ED organizations have issued urgent warnings about their potential to trigger relapse. Recovery has always required moving against the cultural current, but right now that current is stronger than it has been in years. Someone in recovery is not just resisting a private illness. They are resisting an algorithm, a pharmacy, a fashion industry, and often a family system that has absorbed all of the above, often without examining any of it critically. 

After two days at EDAC, here is where I land. The clinicians doing the most thoughtful work in this field are not arguing for harm reduction instead of recovery, and they are not arguing for personalization as a synonym for unstructured deference to patients’ preferences. They are arguing for structured, evidence-based, integrative care that takes seriously what protocol-driven treatment has too often ignored, namely that people arrive with whole lives, intersecting identities, comorbid conditions, prior treatment traumas, and contexts that shape both the illness and their unique path through it. Coercion, rigid protocols, and one-size-fits-all care often produce the very “treatment resistance” they then blame on the patient and use as justification for throwing in the towel. Personalization done properly at the outset may appear as potentially more rigorous and demanding than standard care, but it is really just an efficient reallocation of resources to the front end that pays off down the road. 

The field does not yet have definitive answers about any of this, and that’s okay. What it does have, increasingly, is people willing to listen to patients more, revise when they get it wrong, sit with what is genuinely uncertain, and refuse the easy conclusions that have failed so many before. After Halifax, the field feels closer to where it needs to be than it has in a long time.