Does Positive Affect Treatment increase positive affect and reward hyposensitivity?
Positive Affect Treatment (PAT) is a version of cognitive behaviour therapy, heavy on behaviour activation of enjoyable experiences, souped up with exercises from positive psychology. Michelle Craske and colleagues (2023) persevered through COVID to again compare PAT with Negative Affect Treatment (NAT), an admittedly contrived-for-research version of behavioural activation which, basically, constrains activation tasks to exposure to stressful situations and includes breathing retraining to help cope. I’ve blogged about their previous study of PAT v NAT before (https://www.drmatthewsmout.com/blog/2019/9/8/focus-on-the-positive-to-counter-anhedonia ) so I refer you to that article for an overview of the PAT protocol. This new study differed from the previous one by: 1) including more severe cases (lower positive affect and more self-reported disability); and 2) assessing whether PAT impacts reward sensitivity via a range of measures. It’s in this second department that the study really “goes to too much trouble” in the best possible way. In addition to several self-report scales, the study used physiological and behavioural task measures of reward processing, which require some description to appreciate.
In true scientist-practitioner ambition, PAT aims to reduce the reward hyposensitivity frequently demonstrated in people who experience anhedonia. Reward processing can be broken down into various components. One is reward anticipation/motivation. For example, healthy people get excited when anticipating a reward; depressed - especially, anhedonic - people, do not. The Monetary Incentive Delay Task (MIDT) is a reaction time task that could also be a game show. Participants are given multiple trials and before each trial they receive a signal that they will either be playing to win money or avoid losing money. They are then made to wait a beat. Then, they must press a button within a short amount of time when they see the target symbol appear. Healthy people get excited when there’s a chance to win money. In Craske and colleagues’ study, heart rate acceleration during the ‘wait a beat’ phase of opportunity-to-win trials was a physiological measure of reward anticipation. Reward motivation was also measured by an effort expenditure behavioural task. People with depression make fewer high effort choices than healthy people and the more anhedonic they are, the less effort they make. Willingness to expend effort for reward can be measured experimentally by the impeccably named Effort Expenditure for Rewards Task (which can be amusingly abbreviated to EEfRT – get it?). In this segment, participants must make a bar on a screen go to the top by repeatedly pressing a button, much like the strength-tester at a carnival sideshow if it only needed one finger instead of wielding a mallet. Before each trial, participants can choose an ‘easy’ way (30 button presses with dominant index finger in 7 seconds) or a hard ‘way’ (100 button presses with non-dominant little finger within 21 seconds). When participants choose the ‘easy’ way, they earn $1 but if they choose the ‘hard’ way, they can play for between $1.24 and $4.30. One more twist: they can’t win every time, but they are given accurate information about their odds of success on each trial which varies between high (88%), medium (50%) and low (12%). The score is the ratio of hard trial: total trial choices. Now, I know what you’re thinking, “if this doesn’t generalise to clinically depressed clients then I don’t know what will”. But, as far as experimental psychology paradigms go, I actually think it’s quite clever.
Individuals with anhedonia also experience weaker positive emotional responses to positive emotional stimuli. In Craske’s study, response to reward attainment (capacity to experience pleasure) was measured (besides self-report) via two tasks: 1) a dot probe task measuring response time to disengage from pictures of sad faces and prolong engagement with happy faces (compared to neutral faces) measured positive attention bias; and 2) heart rate acceleration to positive images (compared to neutral images) measured positive responsivity to positive stimuli.
A third component of reward processing is reward learning meaning learning which actions lead to rewards and learning which stimuli are rewarding. In signal detection tasks where participants report on which of two stimuli was just presented in each of a number of trials (A or B), you can induce a bias in healthy people. If, of all the times you correctly select A and B, I tell you you’re correct when you correctly choose “A” three times as often as I tell you you’re correct when you correctly choose “B”, you’ll come to choose A more often (than you should) in future trials. Depressed people don’t acquire this bias as strongly or as quickly as healthy people. This Probabilistic Reward Task was one behavioural measure of reward learning used by Craske and colleagues. The other was a Pavlovian Instrumental Transfer Task. Participants are: 1) trained in an instrumental association between squeezing a handgrip to a required force and receiving money; 2) trained in Pavlovian association by watching a series of colour background-sound combination stimuli paired with either a reward (20c) or absence of reward; 3) then had their grip strength responses tested when the Pavlovian stimuli were presented again in a partial extinction phase (no reward is available for most trials). During the final phase, healthy people show higher grip strength responses to the stimuli that were paired with rewards during phase two than the stimuli not paired with rewards, indicating they had learnt which stimuli were rewarding. To my knowledge this was the first study to test whether appetitive Pavlovian transfer was impaired in people with low positive affect.
All in all, a lot of inventive measures were used to assess the strength of effect on reward sensitivity plus several self-report versions of affect, mood and reward sensitivity and these multiple statistical comparisons were appropriately adjusted for. So, was it worth it?
Once again, those randomized to PAT made greater improvements in self-reported positive affect, interviewer-assessed (and self-report) anhedonia and self-reported depression and anxiety. This replicates the results of the earlier study, this time with treatment-seeking adults whose positive affect was at or below the 8th percentile compared to general community samples. Although I had planned to base a hilarious joke at the start of this blog about what an obvious finding it would be for PAT to increase positive affect, the use of the PANAS Positive Affect scale kind of ruins that. Harmon-Jones, Bastian and Harmon-Jones (2016) persuasively argue that the Positive Affect scale of the PANAS doesn’t actually contain many words representing positive emotions; for example, “happy” and “joy” are not on there, but “alert”, “active” and “attentive” are. So even the basic conclusion that PAT increases positive affect isn’t entirely clear cut. One self-report measure of anticipatory reward motivation was higher among PAT than NAT recipients but the other was no different. The same thing occurred for the two self-report measures of reward attainment (only one significantly higher for PAT recipients). There was no difference between PAT and NAT on self-reported reward learning.
The pattern with the fancier measures was even less consistent. Those receiving NAT showed a slight reduction in cardiac acceleration in anticipation of reward across the treatment period, whereas PAT maintained relatively stable patterns, but this wasn’t evidence that PAT improved reward anticipation. In terms of reward attainment, those receiving PAT showed a slight reduction in their time to disengage from sad faces in the dot probe task but there was no difference between PAT and NAT in heartrate acceleration in response to positive images. There were no differences between PAT and NAT on measures of reward learning.
What should we make of this? The bigger picture is that both NAT and PAT were very good treatments and took participants from the moderate range of the DASS into the normal range by end of treatment. PAT achieved this about 3 sessions earlier on average than NAT. Logically, there doesn’t seem to be any obvious disadvantage to tweaking your behavioural activation intervention to be consistent with PAT. PAT may also produce some subtle improvements in mechanisms likely to sustain mental health in the longer term, but this requires more research to confirm. If that takes some time, I hope this look under the hood at basic mechanism research can help us appreciate how much work goes into trying to mine incremental improvements in clinical outcomes and we can be patient with our research progress!
Clinician implications
This week I’m highlighting practice tips that stood out to me from the PAT treatment manual (see below):
After keeping an activity-mood diary for a week as baseline, aim to schedule 3-5 positive activities per week
No matter how simple the activity, have the client break it into smaller steps (on a worksheet) with a difficulty rating for each step
Use role-plays to help clients overcome their fears of following through on plans to contact other people
Explicitly teach clients that social activities boost mood more than non-social activities even if it doesn’t feel that way going into it
Teach the client to schedule a variety of activities so that hedonic habituation to a single heavily-relied upon activity doesn’t end up feeding into hopelessness
Remind clients “we may not have control over many things in life but we almost always have control over our own actions”
In generating positive activities, encourage clients to think far back to activities in their childhood that they enjoyed (foods, outdoor activities, travel, social activities)
Have clients rate the difficulty of the actions they consider scheduling on a scale of 0-10 (10 = most difficult). Ensure that at least some of the activities scheduled for the week are rated <5.
If clients exhibit strengths in some areas of the program, adapt the standard delivery to spend more time on the areas they find more difficult
For those who struggle to anticipate reward, spend more time on designing positive activities and imagining the positive
If clients struggle to learn which activities are rewarding, focus on taking ownership, gratitude and generosity earlier in treatment
Teach clients to savour positive activities (see Box) as part of the review of activity scheduling homework, at least during the first 3 sessions following the first week of activity scheduling.
Remind clients not to put off their homework tasks to the end of the day when they'll least feel like doing them
For the original article, go to: https://psycnet.apa.org/fulltext/2023-52159-001.html
REFERENCES
Craske, M. G., Meuret, A. E., Echiverri-Cohen, A., Rosenfield, D., & Ritz, T. (2023). Positive affect treatment targets reward sensitivity: A randomized controlled trial. Journal of Consulting and Clinical Psychology, 91(6), 350–366. https://doi.org/10.1037/ccp0000805
Harmon-Jones, C., Bastian, B., & Harmon-Jones, E. (2016). The Discrete Emotions Questionnaire: A New Tool for Measuring State Self-Reported Emotions. PLoS ONE 11(8): e0159915. doi:10.1371/journal.pone.0159915