I Used to Think Depression Was About Sadness. This Study Changed My Mind

I have been thinking about this study for two days now. What stays with me is not the data — though the data is striking — but the implication. For most of my life, I assumed depression was a presence. A heaviness that arrived and needed to be lifted. Sadness that needed to be talked through, medicated, or simply waited out. The treatment goal, in my mental model, was always subtraction: remove the dark, and the light returns on its own.

A 2026 paper in JAMA Network Open suggests this model may have the order wrong. The researchers — Alicia Meuret and Thomas Ritz at Southern Methodist University, and Michelle Craske at UCLA — argue that for the majority of people with major depression, the disabling feature is not the sadness itself. It is the absence of joy. The technical name is anhedonia, and the research suggests it affects nearly 90 percent of people with major depressive disorder (Meuret et al., 2026).

Reading that statistic stopped me. Ninety percent.

What anhedonia actually does

Anhedonia is the reduced ability or inability to feel positive emotions. It is not the same as feeling sad. A person can be free of sadness and still be unable to take pleasure in food, music, conversation, or accomplishment. Some researchers describe it as a kind of emotional flatness — not painful, exactly, but emptied out.

The clinical evidence is sobering. Anhedonia predicts a longer and more severe course of illness (Serretti, 2023). It undermines recovery and is a substantial predictor of suicidal behavior (Ducasse et al., 2018; Gillissie et al., 2023). It appears not only in depression but in anxiety disorders, post-traumatic stress disorder, substance use disorders, and schizophrenia (Taylor et al., 2022; Vinograd et al., 2022; Garfield et al., 2014). Conventional treatments, the authors note, demonstrate limited efficacy in achieving remission of anhedonia, contributing to elevated risk of relapse (Cao et al., 2019; Alsayednasser et al., 2022).

What I find most striking is what patients themselves say. When asked what matters most in being cured from depression, they consistently rate restoring positive emotions higher than reducing negative ones (Demyttenaere et al., 2015). The thing they want back is the capacity to feel good. Yet that is not, traditionally, what treatment has been organized around.

The frame shift

I think this is what I keep returning to. If anhedonia is the more disabling feature, and if patients themselves prioritize the return of joy over the removal of sadness, then a treatment model that primarily targets negative affect may be working on the wrong side of the equation. Not the wrong problem entirely, but a secondary one.

The new study tests whether targeting positive affect directly produces better outcomes than the traditional approach. The next chapter walks through what they found.

References

Alsayednasser, B., et al. (2022). Behaviour Research and Therapy, 159, 104185.

Cao, B., et al. (2019). Progress in Neuro-Psychopharmacology and Biological Psychiatry, 92, 109-117.

Demyttenaere, K., et al. (2015). Journal of Affective Disorders, 174, 390-396.

Ducasse, D., et al. (2018). Depression and Anxiety, 35(5), 382-392.

Garfield, J. B., et al. (2014). Australian and New Zealand Journal of Psychiatry, 48(1), 36-51.

Gillissie, E. S., et al. (2023). Journal of Psychiatric Research, 158, 209-215.

Meuret, A. E., Ritz, T., Craske, M. G., et al. (2026). JAMA Network Open.

Serretti, A. (2023). Clinical Psychopharmacology and Neuroscience, 21(3), 401-409.

Taylor, C. T., et al. (2022). Current Topics in Behavioral Neurosciences, 58, 201-218.

Vinograd, M., et al. (2022). Current Topics in Behavioral Neurosciences, 58, 185-199.

I want to walk through this carefully because the design matters. The study, published in JAMA Network Open in 2026, was a randomized clinical trial with 98 adults (Meuret et al., 2026). All participants had severely low positive affect along with moderate to severe depression or anxiety. They were randomly assigned to one of two treatments, each delivered as 15 weekly individual therapy sessions over telehealth.

One arm received Positive Affect Treatment, or PAT. The other received Negative Affect Treatment, or NAT — a comparison condition built from established components of conventional therapy for anxiety and depression.

What PAT actually looks like

PAT was developed over more than a decade by Craske, Meuret, and colleagues to directly target the brain's reward system (Craske et al., 2019; Craske et al., 2023). Where most therapies focus on reducing what hurts, PAT focuses on rebuilding what feels good. The therapy is organized around three components of reward processing.

First, reward anticipation and motivation. Patients plan pleasurable activities and practice envisioning positive future experiences. The goal is to re-engage the part of the mind that looks forward to things.

Second, reward consumption. Patients are guided through active engagement in enjoyable activities, then trained in vivid present-focused mental rehearsal — the technical term is savoring. Cognitive strategies redirect attention toward positive aspects of experience. Structured practices cultivate positive emotions, including loving-kindness and generosity.

Third, reward learning. The therapy strengthens the association between positive behaviors and improved mood, and encourages self-attribution for positive outcomes. Patients are taught to recognize that they themselves caused something good.

I notice that none of these techniques would feel out of place in a journaling practice. Savoring, gratitude, attention redirection toward positive experience, self-attribution — these are reflective practices, not pharmacological interventions. That is part of why this study interests me.

What NAT looked like

The comparison treatment, NAT, used established evidence-based components for anxiety and depression: exposure to feared situations, cognitive restructuring of distorted appraisals, respiratory training to reduce physiological arousal. It deliberately excluded any positive-affect components to maintain its mechanistic distinction. This is the kind of therapy most people receive when they walk into a clinic with depression or anxiety today.

What they found

PAT produced greater improvements in overall clinical status than NAT (Meuret et al., 2026). The advantage held at one-month follow-up. Patients in the PAT arm showed significant reductions in depression and anxiety symptoms — even though the treatment never explicitly targeted negative affect at all.

That last detail is what stopped me. A therapy that focused exclusively on rebuilding the capacity for joy produced larger reductions in depression and anxiety than a therapy that focused exclusively on reducing them. The work of cultivating positive emotion appeared to do something the work of reducing negative emotion did not.

The researchers also identified, through extensive mediation analyses, that modulation of reward and threat processes was the central mechanism driving these gains. Of seven self-reported reward and threat measures, six mediated clinical outcomes.

The next chapter is about a single sentence from one of the researchers that has stayed with me.

References

Craske, M. G., Meuret, A. E., Ritz, T., Treanor, M., Dour, H., & Rosenfield, D. (2019). Journal of Consulting and Clinical Psychology, 87(5), 457-471.

Craske, M. G., Meuret, A. E., Echiverri-Cohen, A., Rosenfield, D., & Ritz, T. (2023). Journal of Consulting and Clinical Psychology, 91(6), 350-366.

Meuret, A. E., Ritz, T., Craske, M. G., et al. (2026). JAMA Network Open.

The Distinction That Stays With Me: Helpless vs Hopeless

In the press coverage of the study, Alicia Meuret offered a sentence that I have not been able to set down. "There is a difference between feeling helpless and feeling hopeless. When you feel helpless, you still have the drive and the will to want to change things. When people feel hopeless, they don't believe anything will change. That's what anhedonia can look like, and taking away negative emotions doesn't fix it" (Meuret, 2026, as quoted in SMU press release).

I keep turning this over. Helplessness has motion in it. Hopelessness does not.

The texture of each state

When I feel helpless, I am still oriented toward something better. I want to fix the situation. I cannot, perhaps, but the wanting is intact. The reward system, in the technical sense the researchers use, is still active. I can imagine relief. I can picture what improvement would look like. The problem is mostly external — circumstances I cannot control.

Hopelessness is structurally different. The capacity to want is itself dimmed. It is not that I cannot think of solutions. It is that solutions have lost their pull. The future does not look like a place where things might be better. It looks like more of this. That is what the research is describing when it talks about anhedonia — not just an absence of pleasure in the present, but an absence of the anticipation that pleasure could return.

I think this is why the standard treatment frame can feel mismatched. A therapy that helps me reduce anxiety or sadness assumes I still want to feel better. It works on the friction between where I am and where I would prefer to be. But anhedonia, at its more severe end, can flatten that preference itself. There is no friction because there is no pull.

Why this matters for how depression is described

When I think back on the cultural language we use for depression, almost all of it is about pain. The dark cloud. The weight. The sadness that won't lift. These metaphors point at the negative side. They imply something pressing down that needs to be removed.

What I am taking from the new research is that this language may have been describing the wrong half of the experience for many people. The disabling feature is often not the presence of darkness but the absence of light. Not the cloud, but the missing sun behind it. Not the heaviness, but the loss of the capacity to feel buoyant in the first place.

I do not say this lightly. I am not trying to redefine other people's experience for them. But I notice that when I read first-person accounts of depression from people who have been through it, the most haunting passages are almost always about the flatness, not the pain. The inability to want what one used to want. The strange neutrality where the love of music or food or company should be. That is the territory the new research is mapping.

The next chapter is about what this might mean for my own writing practice, and for anyone whose journaling has felt insufficient to reach what is actually low.

References

Meuret, A. E. (2026). Quoted in SMU Research News press release accompanying Meuret, A. E., Ritz, T., Craske, M. G., et al. (2026), JAMA Network Open.
Khazanov, G. K., Xu, C., Dunn, B. D., Cohen, Z. D., DeRubeis, R. J., & Hollon, S. D. (2020). Behaviour Research and Therapy, 125, 103507.
Craske, M. G., Dunn, B. D., Meuret, A. E., Rizvi, S. J., & Taylor, C. T. (2024). Nature Reviews Psychology, 3, 665-685.

I want to be careful here. The study I have been describing is a clinical trial of a structured psychotherapy. It is not a journaling study. PAT is delivered by trained therapists across 15 sessions, with assessments and protocols I cannot reproduce on my own. Nothing in this article is a substitute for professional care, and the research itself is explicit that mediation analyses do not establish causality (Meuret et al., 2026).

What I am doing instead is noticing how closely the building blocks of PAT resemble practices I already do — and could do more deliberately — when I sit down to write.

Savoring as a writing practice

The PAT protocol uses something the researchers call savoring: vivid, present-focused mental rehearsal of a positive experience (Craske et al., 2023). I notice that this is something I do, sometimes, when I write about a good day. Not always. Often I write about what went wrong, what I am worried about, what I am trying to figure out. Those entries have their place. But the research suggests that the work of returning, in writing, to a moment that felt good — and lingering there long enough to actually re-experience it — may be doing something distinct from venting or problem-solving.

There is supporting evidence in the broader literature. A 2024 randomized trial by Kumar and colleagues, using brief behavioral activation plus savoring with university students, found measurable improvements in positive affect dysregulation (Kumar et al., 2024). A 2023 trial by LaFreniere and Newman showed that savoring positive emotions reduced contrast avoidance in generalized anxiety disorder (LaFreniere & Newman, 2023). These are not the same as the PAT trial, but they point in the same direction. Writing about positive experience, with attention, appears to do real work.

Gratitude as a deliberate redirection

PAT includes structured practices around gratitude and self-attribution for positive outcomes (Craske et al., 2019). I notice that gratitude journaling, as it is commonly recommended, can become rote — three things you are thankful for, written quickly before bed, without much inhabitation. The research seems to suggest that the value is not in the listing but in the dwelling. The redirection of attention toward what is good, sustained long enough that the brain registers the redirection.

This is one of the questions I find myself sitting with. When I journal, am I redirecting attention toward what feels good, or am I mostly auditing what feels bad? The two are not opposites. Both have value. But if the new research is right, the first kind of writing may be carrying weight I had not given it credit for.

Loving-kindness and the care for others

PAT also incorporates practices that cultivate positive emotions through care directed at others — loving-kindness, generosity, the deliberate noticing of moments of connection. Writing toward another person in this register, rather than only writing about myself, may be a practice worth more frequent attention.

What I am taking from this

I am not changing my whole practice based on a single study. The research is preliminary in some respects, and the authors are careful about what their findings can and cannot establish. Mechanistic conclusions about reward and threat processing should be interpreted cautiously, since the behavioral and physiological mediators did not reach significance — only the self-reported ones did (Meuret et al., 2026). The picture is incomplete.

But I am taking something. The next time I sit down to write, I want to ask a different question than the one I usually ask. Not only what am I struggling with right now, but also what felt genuinely good today, and can I stay with it long enough to feel it again. If the research is gesturing in the right direction, that second question may be doing more than I have been giving it credit for.

If you are reading this and any of it resonates, please remember that this is not therapeutic guidance. It is a reflection on research, written from outside the clinic looking in. If anhedonia is something you recognize in yourself, especially if it is severe or persistent, the people qualified to help with it are mental health professionals. The study referenced here is one of many; a recent review by Craske and colleagues offers a much fuller picture of where this field is moving (Craske et al., 2024).

References

Craske, M. G., Meuret, A. E., Ritz, T., Treanor, M., Dour, H., & Rosenfield, D. (2019). Journal of Consulting and Clinical Psychology, 87(5), 457-471.

Craske, M. G., Meuret, A. E., Echiverri-Cohen, A., Rosenfield, D., & Ritz, T. (2023). Journal of Consulting and Clinical Psychology, 91(6), 350-366.

Craske, M. G., Dunn, B. D., Meuret, A. E., Rizvi, S. J., & Taylor, C. T. (2024). Nature Reviews Psychology, 3, 665-685.

Kumar, D., Corner, S., Kim, R., & Meuret, A. (2024). Behaviour Research and Therapy, 177, 104525.

LaFreniere, L. S., & Newman, M. G. (2023). Journal of Anxiety Disorders, 93, 102659.

Meuret, A. E., Ritz, T., Craske, M. G., et al. (2026). JAMA Network Open.