Improving mental health by training the suppression of unwanted thoughts

Anxiety, posttraumatic stress, and depression markedly increased worldwide during the COVID-19 pandemic. People with these conditions experience distressing intrusive thoughts, yet conventional therapies often urge them to avoid suppressing their thoughts because intrusions might rebound in intensity and frequency, worsening the disorders. In contrast, we hypothesized that training thought suppression would improve mental health. One hundred and twenty adults from 16 countries underwent 3 days of online training to suppress either fearful or neutral thoughts. No paradoxical increases in fears occurred. Instead, suppression reduced memory for suppressed fears and rendered them less vivid and anxiety provoking. After training, participants reported less anxiety, negative affect, and depression with the latter benefit persisting at 3 months. Participants high in trait anxiety and pandemic-related posttraumatic stress gained the largest and most durable mental health benefits. These findings challenge century-old wisdom that suppressing thoughts is maladaptive, offering an accessible approach to improving mental health.


This PDF file includes:
Figs. S1 to S4 Tables S1 to S12 Fig. S1.Impact of imagination and suppression on event memory.(A) Imagining future events increased the probability of recalling events' key details (dark green bars), relative to the probability of recalling the key details in matched Baseline events that did not receive repeated imagination trials during the 3-days of training (light green bars).In contrast, suppressing thoughts of future events reduced the probability of recalling events' key details (dark red bars) relative to the probability of recalling key details for matched Baseline events (light red bars).Half of the participants suppressed feared events (left half), and half suppressed neutral events (right half).Notably, within the Negative group (left half), the events in the No-Imagine and Baseline-No-Imagine conditions are uniformly negative, whereas within the neutral group (right half), those same conditions concerned uniformly neutral events.In contrast, within the Negative group (left half), half of the participants imagined Positive events and half, Neutral events (with corresponding Positive and Neutral Baseline-Imagine events); in the Neutral group, the same was true (see Fig. 1B for a design overview).(B) The increase in the vividness of events post training, compared to pre-training was greater when participants repeatedly imagined those events during training (dark green bars) compared to the vividness changes observed for Baseline events that were not repeatedly imagined during training (light green bars).In contrast, the reduction in vividness of events post-training, compared to pre-training was greater when participants repeatedly suppressed those events during training (dark red bars), compared to changes observed for Baseline events that were not repeatedly suppressed (light red bars).The organization of Panel B is analogous to Panel A, with half of the participants suppressing Negative events (left half) and half suppressing Neutral events (right half).Instead of plotting changes in the vividness rating scale directly, scores in Panel B are computed as POMP scores (percentage of maximum point; see section on Statistical Analyses).Error bars reflect standard errors.improvement (postpre) on some negative mental health indices (negative affect and depression) and positive indices (wellbeing) on the immediate test (bars below solid lines) and the 3-month follow-up (follow-uppre; bars below dotted lines).Such correlations were significant after suppressing fears (Suppress-Negative), but generally not after suppressing neutral events (Suppress-Neutral).

C. Event Characteristics
Note.All reported event rating measures were collected prior to suppression training and were used algorithmically to assign events so as to match Baseline and Imagine or Baseline and No-Image conditions prior to training.Current Source of Concern: the extent to which the event is a source of current focus of one's thoughts, where 1 is not at all, 3 is to some extent, and 5 is most definitely.Distance in Future: how far in the future one imagines the event happening, with 1 being very soon and 5 being further in the future, with the maximum being 2 years away.
Emotional Intensity: how strongly the event makes one feel, where 1 is very unpleasant and 9 is very pleasant (Manikin scale wherein 5 is neutral).Frequency of Thought: how often one tends to think about the event, with 1 being rarely and 5 being very often.Likelihood of Occurrence: how likely the event is to actually happen to the person, with 1 being a very unlikely event and 5 being almost certain to happen.Long-Term Impact: how much long-term impact (aversive or beneficial) the event will have on one's physical and/or psychological wellbeing if the event were to come true, such that: 1-none at all, 2-negligible, 3-small, 4-moderate, 5-large/formidable.Significance indicated by independent t-test between the Baseline and the manipulation condition (No-Imagine or Imagine).No significant differences were found between Baseline and No-Imagine (or Baseline and Imagine) conditions in any of the rating scales for any group.
To-Be Suppressed Events Relative risks are italicised if statistically significant.If the relative risk = 1, or if the 95% confidence interval includes 1, then there is no significant difference between the groups.Table S11.Usefulness: Do you think that the suppression (mind-blanking) strategy you learned through the course of the training was useful or helpful to use outside the lab in your own life?

A. Training Benefits on our
Reusability: How likely are you to re-use the suppression (mind-blanking) strategy you learned to decrease your anxiety or thoughts about the negative events in your life?Rating Scale: Very unlikely -1; Unlikely -2; Neutral -3; Likely -4; Very likely -5 Impact: In your own words, please elaborate on your opinion about re-using the suppression (mind-blanking) strategy in the previous question.

Fig. S2 .
Fig. S2.Impact of imagination of vividness and affect for hopeful events.Imagining hopeful events increased their vividness on an immediate test for nearly all participants, irrespective of the initial emotional intensity of those hopes and post-traumatic stress status; horizontal bars are individual participants' average change in vividness (postpre, POMP scores).Participants are sorted by the average rated emotional intensity of their hopes pre-training from most pleasant (bottom bars, darker green background) to neutral (top bars, yellower background); yellow highlights on bars indicate participants with probable PTSD.Vividness facilitations are significantly greater for Imagine than for Baseline items (mean changes indicated by solid black vs dotted vertical bars).In the Imagine-Positive group, imagination increased positive affect for the imagined hopes, as seen by the greater increase in affect (postpre) for imagined events (solid black bar) compared to baseline events (dotted bar).

Fig. S3 .
Fig. S3.Pre-Training Trait Anxiety and Post-Traumatic Stress status predicted improvement on individual mental-health measures, but mainly when participants suppressed fears and not neutral events.(A) Correlations (r values) are plotted on the y-axis; significance is indicated by asterixis (*** < 0.001; ** < 0.01; * < 0.05).Trait anxiety levels prior to training predicted whether participants showed significant improvements on negative mental health indices (extending downward) and positive indices (extending upwards) arising on the immediate test (postpre; bars below solid lines) and the 3-month follow-up (follow-uppre; bars below dotted lines).Trait anxiety predicted improvement (i.e., for all six measures of mental health state) for those who suppressed fears (Suppress-Negative), more so than for those who suppressed neutral events (Suppress-Neutral).(B) Post-traumatic stress scores predicted

Fig. S4 .
Fig. S4.Mental health changes after a 3-month delay in our whole sample.Training at suppressing fears (upper half, red label) significantly reduced depression after 3 months compared to pre-training; training at suppressing neutral events (lower half, blue label) reduced worry after 3 months.Individual participants are indicated by dots connected by white lines; boxes reflect interquartile range, and lines reflect median scores.
Delayed Test for Participants with High Trait Anxiety Note.Participants are included as having high anxiety if they scored > 44 on the pre-training baseline measure of the Trait portion of the State Trait Anxiety Inventory.Scores reflect the change in mental health state 3 months after training relative to baseline measures prior to training (followuppre, POMP scores).Significance of Time is derived from paired samples student t-tests between pre-training and follow-up.Significance of Group is derived from independent samples student t-tests between Suppress-Negative and Suppress-Neutral groups.The p-value is indicated as such: < 0.001(***), < 0.01(**), < 0.05(*), <= 0.1( †), > 0.1(ns).B. Training Benefits on our Delayed Test for Participants with Likely PTSD Note.Participants are included as having likely PTSD if they scored > 32 on the pre-training baseline measure of the Impact of Event Scale Revised.Scores reflect the change in mental health state 3 months after training relative to baseline measures prior to training (follow-uppre, POMP scores).Significance of Time is derived from paired samples student t-tests between pre-training and follow-up.Significance of Group is derived from independent samples student t-tests between Suppress-Negative and Suppress-Neutral groups.The p-value is indicated as such: < 0.001(***), < 0.01(**), < 0.05(*), <= 0.1( †), > 0.1(ns

Table S1 .
A. Demographic Characteristics of ParticipantsNote.N = 120.Participants were on average 27.41 years old (SD = 10.21), and participant age did not differ by group.Positive and Negative Mental Health Indices contributed to the main indices of mental health and were also measured post-training and at followup.Here we display raw questionnaire scores, though in most figures and analyses we used POMP scores.Significance indicated is from independent t-test between Suppress Negative and Suppress Neutral groups.Non-significance is indicated by "ns".

Table S2 .
Immediate Effects of Suppression & Imagination on Key Detail Recall and VividnessNote.SIF: suppression induced forgetting, IIF: imagination induced facilitation.Key Detail recall reflects the percentage of items correctly recalled.Vividness reflects the change in rated vividness after suppression compared to before it (postpre).

Table S3 .
A. Relative Risk of Increased Accessibility (Reversals) after Training (postpre) for SuppressedEvents Compared to Baseline Events in the Suppress-Negative and Suppress-Neutral groups.Note.Relative risks are "undefined" if a calculation resulted in division by zero.If the relative risk = 1, or if the 95% confidence interval includes 1, then there is no significant difference between the groups.
B. Average Number of Fears Per Participant that Show Increased Accessibility (Reversals)After Training Compared to Before (postpre) for the No-Imagine and Baseline Conditions, separately for the Suppress-Negative and Suppress-Neutral Groups.Reversals Shown for Both Key Detail Recall and Vividness.

Table S4 .
Delayed Effects of Suppression & Imagination on Key Detail Recall and Vividness at 3 Months Note.SIF: suppression induced forgetting, IIF: imagination induced facilitation.Key Detail recall reflects the percentage of items correctly recalled.Vividness reflects the change in rated vividness after suppression compared to before it (postpre).

Table S5 .
A. Relative Risk of Increased Accessibility (Reversals) after Training (followuppre) forSuppressed Events Compared to Baseline Events in the Suppress-Negative and Suppress-Neutral groups.Relative risks are "undefined" if a calculation resulted in division by zero.If the relative risk = 1, or if the 95% confidence interval includes 1, then there is no significant difference between the groups.
B. Average Number of Fears Per Participant that Show Increased Accessibility (Reversals)After Training Compared to Before (followuppre) for the No-Imagine and Baseline Conditions, separately for the Suppress-Negative and Suppress-Neutral Groups.Reversals Shown for Both Key Detail Recall and Vividness.

Table S6 .
Effects of Suppression & Imagination on Subjective AffectNote.SIF: suppression induced forgetting, IIF: imagination induced facilitation.Scores for subjective affect reflect the change in affective rating after suppression (either Post -Pre in the Immediate condition, or Followup -Pre in the Delayed condition).

Table S7 .
RelativeRisk of Increased Affect (Reversals)after Training (Post-Pre or Followup -Pre) for Suppressed Events Compared to Baseline Events in the Suppress-Negative and Suppress-Neutral Groups.

Table S8 .
Summary of Relative Risks, Based on Calculations in Detailed Table Below.Significant reductions in risk of ironic worsening were observed for depression and well-being on the immediate test (bold).

Table S9 .
A. Training Benefits on our ImmediateTest for Participants with High Trait Anxiety.Note.Participants are included as having high anxiety if they scored > 44 on the pre-training baseline measure of the Trait portion of the State Trait Anxiety Inventory.Scores reflect the change in mental health state after training relative to baseline measures prior to training (postpre, POMP scores).Significance of Time reflects a paired samples student t-test between pre and post training.Significance of Group is derived from independent samples student t-tests between Suppress-Negative and Suppress-Neutral groups on the post -pre values.The p-value is indicated as such: < 0.001(***), < 0.01(**), < 0.05(*), <= 0.1( †), > 0.1(ns).Participants are included as having likely PTSD if they scored > 32 on the pre-training baseline measure of the Impact of Event Scale Revised.Scores reflect the change in mental health state after training relative to baseline measures prior to training (postpre, POMP scores).
B. Training Benefits on our Immediate Test for Participants with Likely PTSDNote.

Table S10 .
A. For High Trait Anxious Participants Only, the Relative Risk of Worsening Mental Health AfterTraining (Post-Pre or Followup -Pre) in the Suppress-Negative Condition Compared to the Suppress-Neutral Condition on Immediate and 3-Month Assessments (significant reductions in bold).Relative risks are italicised if statistically significant.If the relative risk = 1, or if the 95% confidence interval includes 1, then there is no significant difference between the groups.

Table S12 .
).Subjective Reports About the Experience of Suppressing Fearful Thoughts Over 3 Months, as a Function of Post-Traumatic Stress Symptoms.Participants are sorted from highest pandemic-related IES Score to lowest, and divided into quartiles to illustrate that perceptions of usefulness and likelihood of future use were greatest in the most symptomatic participants.
is the issue...I need to connect the anxiety triggers with the resource of blanking and putting that in place.203 15.91 10.00 20.00No 3 Potentially it could help but I have forgotten to try it!I will try to use it more and see.'s a great way to learn to control bad thoughts you have about life and to help improve mood.111 12.50 11.67 18.33 Yes 4 I think it is a useful strategy and it has been good to have this reminder as I will now use it more t really thought about it over the past 3 months.I don't tend to dwell on too many negative events that much, and when I do I tend to think through them quite a lot rather than blanking them out.If in the future my worries are becoming more problematic (constant), then I might use the suppression strategy.Useful for worries for things for which I have no control over such as worrying about things that haven't happened yet (e.g.deaths in the family) or things that might not happen (e.g.another lockdown), however potentially not productive to completely mind-blank on the issue because there could still be actions that I could take -such as spend more time with my grandparents, try to look for another job/take provisions for possible lockdown etc.So, I think I could use it to some extent but not completely mind-blank on issue/subject of worry.167 0.00 46.67 73.33No 3 I find it wouldn't be very helpful for me since I prefer to think through problems rather than blank them out.
234 69.32 53.33 71.67 Yes 3 Need to remember to do it in the moment 158 68.18 28.33 38.33 Yes 3 I like the method but often I find it difficult to remember that I can use it for situations outside of the ones I wrote down 80 67.05 33.33 66.67 Yes 5 The mind-blocking is able to prevent someone from concentrating on negative events and just focus on positive and neutral events 128 55.68 35.00 71.67 Yes 5 it's very useful when trying to get to sleep, because I can easily stop over-thinking 18 51.14 25.00 30.00 Yes 4 Sometimes it just feels impossible to stop thinking if something is worrying you, or you are taken by surprise by a bad event.But if something just begins to creep in you can actively decide to stop.t believe how effective it was and it made me realise how powerful my brain can be.I always thought that just dismissing thoughts would make things worse, like 'brushing it under the carpet' however, it's a useful tool as it can put some distressing irrelevant thoughts to bed.PTSD: 7.60 Depression: 11.46 Anxiety: 27.92 Usefulness: 69% Reusability: 3.50