Humor therapy research spans laughter-based activities, humor interventions, psychotherapy, and medical clowning. The largest review summarized 86 included articles, while later reviews examined hospital clowning and humor in psychotherapy. Across these sources, the results are promising but uneven: depression, stress, and anxiety often improved after laughter-inducing interventions, while humor-only results were not consistently significant.
Contents
- How large is the evidence base?
- Depression outcomes
- Stress and anxiety outcomes
- Results for older and healthy adults
- Cancer-care findings
- Medical clowning in hospital care
- Humor interventions in psychotherapy
How large is the evidence base?
The review by van der Wal and Kok (2019) screened 98 potentially relevant articles and included 86 in its systematic review. Of those, 29 studies entered the meta-analysis. The review identified 83 intervention studies involving laughter therapy with humor, without humor, or with humor status unknown.
The study designs varied substantially. Fourteen studies were randomized controlled trials. Another 41 were quasi-experimental pre-test/post-test studies with a control group. The remaining 31 were pilot studies, field studies without a control group, or studies whose design was unknown. This mix matters when interpreting the reported effects because controlled trials and exploratory studies provide different levels of evidence.
The review also separated interventions by whether humor was clearly part of the treatment. That distinction makes it possible to compare laughter-inducing therapy in general with humorous and non-humorous approaches, although the source does not establish that one approach is universally better.
Depression outcomes
Depression was an outcome measure in 31 studies in the van der Wal and Kok (2019) review. Depression significantly decreased after laughter-inducing therapy in 26 of those 31 studies. The average corrected effect size across all depression studies was dppc2 = 0.85.
The result was smaller when the analysis was restricted to randomized controlled trials: dppc2 = 0.57. That difference illustrates why the overall estimate should not be read as a single guaranteed effect for every setting. The broader estimate includes multiple designs, while the randomized-trial estimate focuses on the subset with stronger experimental controls.
The review reported different estimates according to the role of humor. Humorous therapy had a depression effect size of dppc2 = 0.43, compared with dppc2 = 0.73 for non-humorous therapy. The later review of humor interventions in psychotherapy (2023) reported the same comparison and described the difference as 41%.
The psychotherapy review also reported results by intervention type. Laughter and humor interventions together improved depression with p = 0.001. Laughter interventions alone improved depression with p < 0.0001 across 6 studies. Humor interventions alone were not significant for depression, with p = 0.34 across 3 studies. These findings separate laughter-based interventions from humor-only interventions rather than treating them as interchangeable.
| Depression result | Reported finding | Source label |
|---|---|---|
| All laughter-inducing therapy studies | dppc2 = 0.85 | van der Wal and Kok (2019) |
| Randomized controlled trials | dppc2 = 0.57 | van der Wal and Kok (2019) |
| Humorous therapy | dppc2 = 0.43 | van der Wal and Kok (2019) |
| Non-humorous therapy | dppc2 = 0.73 | van der Wal and Kok (2019) |
| Laughter interventions alone | p < 0.0001 across 6 studies | humor interventions in psychotherapy (2023) |
Stress and anxiety outcomes
Stress was measured in 19 studies in the van der Wal and Kok (2019) review. Stress significantly decreased in 18 of those 19 studies. The average corrected stress effect size across all studies was dppc2 = 0.58, and the randomized-controlled-trial-only estimate was dppc2 = 0.51. The non-humorous stress effect size was dppc2 = 0.66.
Anxiety was measured in 15 studies. Fourteen of those 15 studies reported a significant decrease after laughter-inducing therapy. The average corrected anxiety effect size across all studies was dppc2 = 0.81, while the randomized-controlled-trial-only estimate was dppc2 = 0.92.
The humor distinction again produced different estimates. The humorous anxiety effect size was dppc2 = 0.51, compared with dppc2 = 1.00 for non-humorous therapy. In the psychotherapy review (2023), humorous therapies were reported at dppc2 = 0.51 versus 1.00 for non-humorous therapies, a stated 49% difference.
The psychotherapy review reported a significant anxiety result for laughter and humor interventions together at p = 0.01. Laughter interventions alone were significant at p = 0.02 across 5 studies. Humor interventions alone were not significant for anxiety, with p = 0.28 across 2 studies.
Taken together, these figures describe a recurring pattern in the supplied research: combined or laughter-centered interventions show significant outcomes in several analyses, while humor-only subgroups can be smaller and nonsignificant. The figures do not establish that humor is ineffective; they show that the reported results vary by intervention definition and analysis group.
Results for older and healthy adults
The elderly subgroup in van der Wal and Kok (2019) contained 21 studies reporting a positive effect of laughter or humor on mental, social, or physical health. Within the elderly depression subgroup, 8 studies reported that laughter-inducing therapy significantly decreased depression. The subgroup description classified 2 studies as using spontaneous laughter, 3 as using non-humorous laughter, and 3 as using humor therapy with unclear humor status in the paper summary.
In the elderly depression section, 8 out of 11 studies showed improved depression outcomes. For depression measured with the Geriatric Depression Scale, or GDS, 7 out of 10 studies had confidence intervals below zero. Four out of those 10 studies showed an expected reduction of at least 4 points on the 0-to-30 scale.
The healthy-adult subgroup included 13 studies reporting a positive effect of laughter-inducing therapy on mental, social, or physical health. Seven of those healthy-adult studies used spontaneous laughter. These subgroup counts cover different populations and outcomes, so they should not be combined into a single estimate.
Cancer-care findings
The cancer-patient subgroup in van der Wal and Kok (2019) included 8 studies reporting a significant positive effect of laughter therapy on mental and physical health. Laughter therapy decreased anxiety in 3 studies.
The pain results were mixed: laughter therapy decreased pain in 1 study but not in another, producing 1 positive result and 1 null result. One study reported improved immune response, and one reported decreased cortisol levels. These findings are narrower than the depression and anxiety results because each of the immune-response and cortisol findings came from a single study.
The subgroup therefore includes several distinct outcome types rather than one overall cancer-care effect. Anxiety, pain, immune response, and cortisol were not reported with the same number of studies, and the supplied figures do not provide a pooled estimate for all cancer-patient outcomes together.
Medical clowning in hospital care
The medical clowning in hospitalized children review (2023) included 18 studies. Its anxiety analysis covered 912 children and pooled 14 studies. The pooled child-anxiety effect was -0.76 on the anxiety score, with P < 0.001.
The preoperative-anxiety analysis covered 512 children and pooled 9 studies. Its pooled effect was -0.78, with P < 0.001. Because these are preoperative results, they describe anxiety around surgery rather than anxiety in every hospital situation.
The pain analysis covered 338 participants and pooled 6 studies. The pooled pain effect was -0.49, with P = 0.06. This reported result is not statistically significant under the conventional threshold implied by the source’s other reported findings.
Parents were also evaluated. The parental-anxiety analysis covered 489 participants and pooled 10 studies, producing a pooled effect of -0.52 with P = 0.001. The parental-preoperative-anxiety analysis covered 380 participants and pooled 6 of those 10 parental-anxiety studies. The pooled parental-preoperative-anxiety result was significant with P = 0.02.
These medical-clowning figures are relevant to humor therapy because clowning uses performance, play, and humor in a care setting. They should still be kept distinct from adult psychotherapy studies and from general laughter interventions: the population, setting, and outcomes differ.
Humor interventions in psychotherapy
The humor interventions in psychotherapy review (2023) lists study-level participant counts that show how varied the underlying samples were. Zhao et al. included 814 adults and elderly participants. Rudnick et al. included 32 participants, Ventis et al. included 40 participants, and Deutsch included 40 participants aged 21 to 64 years. Cai et al. included 30 participants, including 16 male participants. Gelkopf et al. included 22 participants aged 30 to 57 years. O’Brien included 20 participants, including 12 male participants aged 18 to 25 years. Panichelli et al. included 110 participants, including 40 male participants aged 20 to 70 years. Falkenberg et al. included 6 participants.
The review’s depression table recorded decreased depression in 26 of 31 studies, while its anxiety table recorded decreased anxiety in 14 of 15 studies. Those counts align with the broader laughter-therapy review, but the psychotherapy review also reports outcomes from specific behavioral and distress measures.
In systematic desensitization, both the humorous and non-humorous groups outperformed the control group on the Behavioral Approach Test, with F(2,36) = 5.95 and p = 0.006. Both groups also outperformed the control group on Spider Cognitive-Dimension ratings, with F(2,35) = 8.00 and p = 0.001.
One study reported a significant decrease in distress on the Hopkins Symptom Checklist-58 for the humor condition, with t(9) = 5.64 and p < 0.06 as reported in the review. The non-humor condition also showed a distress decrease, with t(9) = 3.21 and p < 0.06. Therapist- and participant-rated problem distress decreased only in the humor condition, with t(9) = 3.61 and p < 0.06.
The statistical record is therefore broad but not uniform. It includes large and small participant groups, older and younger adults, hospitalized children and parents, and outcomes ranging from depression scores to preoperative anxiety and behavioral tests. The most consistent interpretation of these supplied figures is that laughter-related and humor-related interventions have been associated with improvements in several measured outcomes, while the size and significance of the result depend on the population, treatment definition, study design, and outcome measure.