Positive Affect Treatment
An evidence-based protocol for the part of depression and anxiety that ordinary CBT does not touch, and a re-reading of the positive-psychology arc of the curriculum as the practices its three modules already contain.
6 to 8 minWhat this protocol is, and why it exists
Positive Affect Treatment is a fifteen-session, manualized protocol developed by Michelle Craske and colleagues at UCLA, published in Journal of Consulting and Clinical Psychology in 2019 and replicated in a second randomized trial published in 2023. It was built to address a specific clinical gap. Standard cognitive-behavior therapy is good at reducing negative affect — fear, threat-monitoring, rumination, the loud parts of depression and anxiety — and it does this well enough that the trials have been replicated for forty years. What it does not do is restore positive affect. A patient can finish a course of CBT with their panic attacks down by eighty percent and their anticipatory worry quieted and still describe their life as joyless, dim, and not worth attending to. The clinical name for the joylessness is anhedonia, and it is the symptom that predicts relapse, suicidality, and treatment dropout more reliably than any other.
PAT was designed to do for the positive-affect system what CBT does for the negative-affect system. The protocol does not assume the patient has any reservoir of positive affect to access; it assumes the reward system is functionally hyposensitive, names the three places that hyposensitivity tends to show up, and trains them one at a time. The 2019 trial found that PAT outperformed an active comparator on positive affect, depression, anxiety, and suicidality at six-month follow-up. The 2023 trial localized the mechanism: PAT was changing reward sensitivity, not just symptom counts.
This page is not the protocol. The protocol requires a clinician trained in PAT, and the workbook and therapist guide are published by Oxford University Press. What this page does is name the architecture and map the existing practices in the positive-psychology arc of the wheel onto the three modules, so a reader can see what work each individual practice is doing inside the larger clinical frame and which clinician they would need to do the protocol properly.
If your prior treatment quieted your worst hours but did not return color to the rest, the gap PAT was built to address is the gap you are describing. The persistence of low positive affect after standard treatment is not personal failure; it is the documented limitation of the treatments most outpatient settings deliver.
Before you read further
These are calibration questions, not warm-up. Sit with each one for a moment before reading the next.
10 to 12 minThree places the reward system gets quiet
The mechanistic claim of PAT is that positive affect is not a single thing the brain either generates or fails to generate. It is the output of three separable subsystems, each of which can be functionally hyposensitive on its own, and each of which responds to a different intervention. Craske’s 2023 trial measured all three at baseline and post-treatment; the gains in positive affect were predicted by movement on those three measures, not by a generic improvement.
Anticipation is the upstream end of the reward loop, the small lift the body produces when a pleasurable event is on the calendar. The dopaminergic substrate is well-mapped: ventral striatum activity to reward cues is the standard imaging marker, and it is depressed in anhedonic samples even when the consummatory experience itself is intact. The behavioral correlate is what Craske calls the loss of motivated approach: the patient knows the dinner with friends will probably be fine, but the loop that translates that knowledge into the energy to put on shoes is no longer firing.
Module 1 of PAT works directly on anticipation. Pleasant events are scheduled and labeled in advance, and the labeling itself is part of the work, because naming the upcoming pleasure is one of the inputs the anticipation loop requires. The classical behavioral-activation literature treats this scheduling as an end in itself; PAT treats it as a primer for the in-session work that follows.
Attainment is the in-the-moment registration of pleasure, the part of the reward arc that occurs while the good thing is occurring. In anhedonic samples this is often the most paradoxical reading: the meal was, by report, fine; the conversation was, by report, fine; the body did not seem to register that any of this was happening. The phenomenology is one of muffled contact rather than absent contact.
The PAT response to attenuated attainment is imaginal recounting. After a pleasurable event, the patient is walked through the event again in detail — what was seen, heard, smelled, said — and asked to remain inside the recounting long enough for the body to catch up to what the day had already produced. The recounting is the active ingredient; the scheduling without the recounting produces small effects that fade.
Learning is the downstream end of the loop, the encoding work the brain does between events so that the reward becomes available for retrieval and weighting in future decisions. In a calibrated system, a pleasant event raises the predicted value of similar events; in an anhedonic system, the predicted value remains flat regardless of how the event went. The behavioral output is a person who keeps repeating the same low-reward schedule because their internal model has not been updated by what actually happens when they leave the schedule.
PAT addresses learning through the cognitive and cultivation modules: written attention to what went well, self-attribution for the role the patient played in producing the outcome, and the structured generation of positive content through gratitude, generosity, loving-kindness, and appreciative-joy practices. The cultivation work is doing what the clinical literature calls memory consolidation under a different name; the practices are old, the framing is new.
Craske, M. G., Meuret, A. E., Ritz, T., Rosenfield, D., Treanor, M., & Dour, H. (2019). Positive affect treatment for depression and anxiety: A randomized clinical trial for a core feature of anhedonia. Journal of Consulting and Clinical Psychology, 87(5), 457–471. PAT was compared to Negative Affect Treatment, a manualized form of CBT focused on threat reduction. At post-treatment and at six-month follow-up, PAT produced larger gains in positive affect, larger reductions in depression and anxiety, and lower suicidal ideation. The 2023 follow-up trial in Behaviour Research and Therapy localized the change to reward-system measures rather than generic symptom improvement.
15 to 18 minWhat the fifteen sessions actually do
PAT is delivered in three sequential modules across fifteen sessions. The modules are not interchangeable: each one rests on what the previous one trained, and the order is structured so that the patient is not asked to cultivate generosity before they have re-acquired the in-the-moment registration that gives generosity something to cultivate.
The patient and therapist build a list of activities the patient previously found pleasurable, or expects might be pleasurable, and schedule them across the week. The departure from standard behavioral activation is what happens after the activity. The patient returns to session and is walked through the event in detail — sensory, relational, temporal — and asked to remain inside the recounting until something registers. The recounting is repeated until it stops being a report of the event and starts being a re-experience of it.
Maps to the wheel. The first module is doing the work that Savoring (Module 3, Positive Psychology) was named for, in a more clinical register. Active Constructive Responding (Module 4) extends the same mechanism into the dyadic case: the recounting is shared, and the partner’s response either consolidates the encoding or interferes with it.
The cognitive module of PAT is the structural inverse of the cognitive module in standard CBT. Where CBT cognitive work targets the over-weighting of threat and the catastrophizing of ambiguity, PAT cognitive work targets the under-weighting of accomplishment and the externalization of pleasant outcomes. Three exercises run through the module: attending to positive aspects of an experience that the patient initially read as neutral or negative, claiming the role the patient played in producing a good outcome, and imagining one realistic positive event a week ahead in concrete sensory detail.
Maps to the wheel. The self-attribution work overlaps directly with Character Strengths (Module 2): the VIA framework is what one looks at when asking what trait the patient was using when the good outcome occurred. The future-imagining work is the protocol-grade version of Best Possible Self (Module 7). Hope Theory (Module 5) supplies the agency-and-pathways vocabulary that Snyder developed for the same loop PAT is training.
The third module trains the four cultivation practices Craske draws from the contemplative literature: gratitude, generosity, loving-kindness, and appreciative joy. Each is operationalized as a behavioral exercise rather than a contemplative one. Gratitude is written for specific recipients with specific causes; generosity is performed and then recounted; loving-kindness is practiced in daily fragments rather than as a formal sit; appreciative joy is rehearsed against the comparison-and-envy reflex by working with another person’s good news directly.
Maps to the wheel. Three Good Things (Module 1, the first practice published in the wheel) is the protocol-grade gratitude exercise. The cultivation practices in this PAT module overlap with the Equanimity page (Module 2, Fundamental Well-Being) where the Brahmavihara framework supplies the same four trainings under their Buddhist names. The convergence of two literatures on the same four trainings is the strongest argument for putting them late in the protocol rather than early.
The published outcome data are the post-treatment and six-month-follow-up windows; longer-term data are still accruing. The clinical observation across both trials is that the gains are largest where the patient is able to keep one or two of the practices running on their own after sessions end. The most-continued practices are the gratitude exercise from module three and the recounting habit from module one, both of which are short enough to run on a tired Tuesday.
This is also the point at which the wheel becomes useful in a way it was not before the protocol. The thirty-two practices supply continuation material for a person who has finished PAT and wants to keep working on the same machinery without re-entering weekly therapy. The positive-psychology column is the directly continuous one; the other three columns supply complements that PAT does not address — relational depth, existential reckoning, contemplative ground.
10 to 12 minWhere the protocol tends to fail, and what those failures show
The failure patterns below are documented in the trial data and in the therapist guide. Each one is most useful read as diagnostic information rather than as a problem to be solved by exhortation.
The patient completes the action exercises competently and then stalls when asked to claim authorship of a good outcome. The reluctance reads as superstition: if the credit is taken, the outcome will be revoked.
What is happening. The fear-of-positivity literature, formalized by Yamasaki and others, names a subset of patients for whom positive affect itself is threatening, usually because the patient learned early that a parent’s good mood was unstable and that being seen as happy was a precondition for the next loss. The reluctance to claim a good outcome is not modesty; it is a defense against the predicted retraction.
The reframe. Treat the reluctance as data the protocol surfaces, not as resistance to it. Module 2 cognitive work is exactly where this surfaces, and the therapist guide is explicit that the reframe of fear-of-positivity is part of the protocol rather than a deviation from it. For the reader without a clinician, this is the point at which a clinician is indicated.
The cultivation module produces lists that are technically gratitude exercises and functionally not. Family. Health. Job. Repeated nightly. Nothing changes.
What is happening. The exercise has collapsed into a category statement — gratitude as concept rather than as a particular event with a named cause. Generic phrases bypass the encoding work the practice was designed to perform; the consolidation subsystem cannot operate on a category.
The reframe. Specificity is the active ingredient. Not the family, but the four-minute phone call at 8:14 PM with the sister who returned the message. Not the job, but the eleven-minute window between meetings when the work felt clean. The why-step in the original Three Good Things protocol is what protects the practice from this collapse, and it is on every cultivation exercise in PAT for the same reason.
The patient improves through Module 1, declares themselves better, and stops attending. Three months later the gains have eroded.
What is happening. The action work is the most rewarding of the three modules to perform, and it produces the most visible early gains. The cognitive and cultivation modules do the consolidation work that makes the gains durable, and they are the modules that get truncated when the patient declares themselves done before the full course is delivered.
The reframe. The trial protocol is fifteen sessions for a reason. The mechanistic data show that the gains in reward learning — the consolidation subsystem — are produced by the second and third modules, not by the first. Stopping at Module 1 trains the wanting and liking subsystems but leaves the encoding system uncalibrated, which is precisely the configuration most prone to relapse under stress.
The patient was referred for anhedonia and the actual presentation is not anhedonia. The protocol is delivered competently and the patient does not improve.
What is happening. Several presentations look like anhedonia and require different treatment. Persistent dissociation can present as the muffled-contact phenomenology PAT was built for, and the cultivation module can deepen the dissociation rather than repair the contact. Untreated mania or hypomania can present as hyper-engagement that masks reward-system dysregulation in the opposite direction; PAT is contraindicated in active mania. Severe trauma without a stabilization phase can find the recounting work intrusive in a way the protocol cannot hold.
The reframe. PAT assumes the diagnosis. The clinical work that distinguishes anhedonia from these other presentations is not in the workbook because it is upstream of the protocol. For the reader, this is the strongest single argument for entering PAT through a clinician trained in differential assessment rather than through the workbook alone.
If the action exercises produce panic, a sense of fraudulence, or a recursive worry about whether one is doing the practice correctly, that is information about which subsystem is dominant rather than evidence the practice is wrong. Anxiety-prominent presentations sometimes need a brief threat-reduction stabilization before the positive-affect work can land, and the clinical literature is explicit that PAT is not a stand-alone for severe anxiety.
10 to 12 minWhen PAT is the right protocol, and when it is not
PAT is one of several treatments that work; the question for the reader is whether it is the one that fits the presentation. The decision tree below is condensed from the inclusion and exclusion criteria of the published trials, and is not a substitute for an evaluation.
For readers within reach of a UCLA-trained clinician, the protocol is increasingly available in metropolitan areas; outside those areas the workbook and therapist guide are sometimes used adjunctively in standard cognitive-behavioral practice. Brian’s clinical work in Pittsburgh draws on the PAT framework where appropriate, with the differential-assessment work running upstream of the protocol rather than inside it.
Week one of an attempted self-directed run
A reader has read the Oxford workbook, decided to attempt a self-directed version, and is sitting down with the first scheduled pleasant event of the week. Tap to reveal one possible run, and the readings of when the self-directed version is enough and when the clinician becomes necessary.
The activity was a walk along the river at five PM. It was, by report, fine. The recounting feels artificial, and the body has not caught up to anything. What does the practice look like, and at what point is the clinician necessary?
1. Run the recounting anyway. Sensory detail first — the angle of the light, the temperature of the air on the back of the neck, the sound of the bridge from below. Stay inside the recounting for the full ten minutes the workbook prescribes, even if nothing arrives.
2. Repeat across the week. The first three or four recountings are often dry; module-one effects in the trial were measured across seven sessions for a reason.
3. If by week three nothing has shifted at all, the differential question is whether the diagnosis is correct. Anhedonia responds within three weeks in the published data; persistent flatness across three weeks is a signal that the presentation may be something else and that a clinician is now upstream of any further self-direction.
4. If module two surfaces the fear-of-positivity pattern, that is the threshold on the other side. The cognitive work of PAT is the most clinician-dependent module; running it self-directed past the wall it surfaces is not advised.
6 to 8 minHow the wheel is the continuation
Three questions are worth sitting with after a course of PAT or a serious self-directed engagement with the workbook.
The closest single page to PAT in the curriculum is Three Good Things, the protocol-grade gratitude exercise that the third module of PAT trains. Beside it sit Savoring, Active Constructive Responding, Character Strengths, Hope Theory, and Best Possible Self, each of which is a component of one of the three modules and each of which the wheel publishes as a standalone practice for readers between courses of treatment.
Craske, M. G., Meuret, A. E., Ritz, T., Rosenfield, D., Treanor, M., & Dour, H. (2019). Positive affect treatment for depression and anxiety: A randomized clinical trial for a core feature of anhedonia. Journal of Consulting and Clinical Psychology, 87(5), 457–471.
Craske, M. G., Meuret, A. E., Ritz, T., Treanor, M., Dour, H., & Rosenfield, D. (2023). Positive affect treatment targets reward sensitivity: A randomized controlled trial. Journal of Consulting and Clinical Psychology, 91(6), 350–366.
Craske, M. G., Dour, H., Treanor, M., & Meuret, A. E. (2022). Positive Affect Treatment for Depression and Anxiety: Therapist Guide and Workbook. Oxford University Press.
Treadway, M. T., & Zald, D. H. (2011). Reconsidering anhedonia in depression: Lessons from translational neuroscience. Neuroscience & Biobehavioral Reviews, 35(3), 537–555.