Three of four dance styles cut depression scores in a 142-student trial, a new review notes; jazz did not

A team from universities and a hospital in Wuhan, China, reviewed the clinical evidence on dance for depression and anxiety and settled on "cautious optimism". Their review, published in Frontiers in Psychiatry on October 19, 2026, backs dance as a complement to care and rules it out as a replacement.

A map of trials, not a verdict

The paper is a structured narrative review. The authors searched PubMed/MEDLINE, PsycINFO, Web of Science, Scopus and the Cochrane Library up to early September 2026, and describe the result as a narrative and translational synthesis rather than a systematic review or meta-analysis. There is no pooled effect size. There is no formal grading of how certain the evidence is.

Headlines that promise to dance away a low mood have an unwelcome reader here: the paper itself.

It splits the field in two. Rhythmic dance interventions are structured, music-led classes in a group, the kind a community hall or a studio runs. Dance/movement therapy is psychotherapy delivered inside a therapeutic relationship, with clinical aims and a trained facilitator. The authors write that the two "should therefore not be treated as interchangeable interventions".

The most direct clinical evidence, they say, concerns dance/movement therapy in adults with depression. The class-style research is broader but more heterogeneous, and the breadth should not be read as strength: much of it was done on people with elevated symptoms or in supportive settings, not on diagnosed disorders. The authors report no financial support for the work and no commercial conflict of interest.

Four styles, one trial, one exception

The review's evidence table lists the primary studies one by row, and one row shows how uneven results can look inside a single trial. Zhao and colleagues, in 2025, assigned 142 college students with depressed mood to modern dance, Chinese classical dance, jazz, ballet or a non-dance control. Sessions ran 20 minutes, four times a week, for 12 weeks.

Scores on the Zung Self-Rating Depression Scale fell after classical dance (p=0.029), ballet (p=0.008) and modern dance (p<0.001). They did not fall after jazz (p=0.241) or in the control group (p=0.591). Modern dance came out strongest.

Jazz is the interesting row.

Nothing in the review explains why the one style that did not move scores should be the one that sounds liveliest, and the table does not pretend to. The authors flag the study's limits instead: a college sample with low mood rather than diagnosed depression, relatively small groups, and no formal test of why the dancing worked.

The other rows are a mixed bag. Some are tiny. Xu and colleagues, in 2024, compared intensive dance/movement therapy with a wait-list in 36 university students with depression and anxiety symptoms, and the benefits held at three months. Benítez and colleagues, in 2025, followed adults over 40 who practised Argentine tango. Of 24 enrolled, 16 finished. Systolic blood pressure dropped (p=0.0019), but quality of life, sleep and depressive symptoms did not change significantly. A 2025 case series of dance/movement therapy for treatment-resistant depression had three participants.

Two more rows show how hard it is to isolate the dancing. In a 2025 pilot, Agres and Chen gave 76 university students a 10-week performing arts program that mixed facilitated drumming and dancing. Stress scores improved by week 10, but the review notes that dance-specific effects cannot be isolated from the combined program. In another 2025 trial, 70 policewomen with elevated anxiety finished 12 weeks of either mindfulness-based aerobic dance or traditional aerobic dance. The mindfulness version improved anxiety scores earlier and more strongly. The trial was rated at high risk of bias. It had no non-exercise control, so it compares two kinds of dancing and says little about dancing against nothing.

The review is blunt about what that adds up to. In its words, "Active or credible comparators remain uncommon, making it difficult to separate intervention-specific effects from the effects of therapeutic attention, group participation, physical activity, expectancy, or general engagement". A wait-list group, for instance, gets none of the attention that a weekly class hands out for free.

The authors list the general weaknesses in one sentence: "Many DMT and RDI trials have small samples, heterogeneous protocols, variable comparator conditions, and limited follow-up". On why dance might work, the paper lists candidate mechanisms: rhythmic entrainment, behavioral activation, reward engagement, emotion regulation, social connectedness. It also states that direct mediation evidence is still insufficient. On reward and activation specifically, it calls the direct dance evidence "currently sparse".

Six months of Jive and Cha-Cha

A study published in early October 2026 in Frontiers in Public Health gives the review's cautious reading a concrete case. Yuanmei He and Jialin Wu, of Guangzhou University, followed students who danced Jive and Cha-Cha to rock 'n' roll music. They recruited from four private Dancesport clubs and two university dance societies in a major Southern Chinese city. The study reports no funding.

At the start, 152 people took part. At three months 136 remained, at six months 131, and 120 completed all three rounds of measurement. The researchers compared those who attended three or more structured sessions a week with those who attended one or two.

The frequent dancers scored higher on social self-efficacy (a gap of 8.62 points on the scale used) and on general self-efficacy (3.60 points). They scored 2.80 points lower on the GAD-7 anxiety questionnaire and had higher resting heart rate variability, by 7.64 milliseconds. Over the six months, the abstract lists widening gaps for social self-efficacy, general self-efficacy and heart rate variability, but not for anxiety.

The most useful number for anyone nervous about turning up is the beginner count.

Of the 152 participants, 56 (36.8%) were complete beginners and 63 (41.4%) had under a year of experience. The authors lean on that: it weakens, they argue, the idea that only confident people join in the first place.

They also concede the obvious. In their words, "it remains plausible that participants with higher baseline self-efficacy and lower anxiety were more likely to maintain high dance frequency". There was no group of non-dancers. Frequency and mood were self-reported, the sample was purposive, and the GAD-7 measures symptoms, not a diagnosis. Attrition reached 21.1% across the three rounds.

The qualitative half rests on 32 interviews. The authors drew four themes from them: rhythmic entrainment as an affective catalyst, embodied mastery as a source of self-efficacy, synchronized partnership as a context for trust, and meaning-making as narrative integration. Mastery and trust are readings of what people said, not measured mechanisms.

So the picture is a group that keeps turning up, improves slowly at something it does badly at first, and reports feeling more able to approach other people. Whether that is the dance, the club, the friends or the kind of person who stays, the design cannot say.

A second review says the same, more quietly

Days earlier, Frontiers in Psychology published a systematic review from Jiangsu University of Science and Technology, registered in PROSPERO, the international register of reviews. It pulled together 22 trials with 1,645 participants, searched up to late August 2026. Some outcomes looked favorable: global cognition, memory and attention, plus some measures of depression, anxiety, stress and quality of life.

Then it undercut itself. The findings "were not consistent across all studies or outcome measures", and "Several favorable findings were derived from studies judged to have a high, serious, or critical risk of bias". The evidence came mostly from older adults, especially those with cognitive impairment, so no one could compare age groups with confidence.

Two reviews in two weeks, written by different teams, land on the same hedge.

Neither hands a studio owner a slogan.

Dancing is one way to do it together

For readers in the United States, a nearby result points at a plainer version of the same question. Brigham Young University researchers, writing in Family Relations, analysed survey answers from nearly 1,700 young married couples in a national study called CREATE, in a university release from early October 2026. Couples who exercised together reported better communication, more effective conflict resolution and less marital stress than couples who did not share physical activity.

Walking led the list. About 60% of husbands and wives said they exercised, the lead author Jeremy Yorgason said, "but only about a third exercised together". He added: "While the impact is not necessarily large, it is very consistent". The data come from a survey, so they show an association and not a cause, and the shared activity did not have to be dance. An earlier brief on this site, about an engaged couple whose dancing at a Weeknd concert in Singapore drew 245,300 views, covered another pair moving in step.

What the authors ask programs to do

For readers in the United States and the United Kingdom, where dance classes and community sessions are easy to find, the practical message is narrower than the marketing. The review treats classes as a possible supplement and says the evidence cannot yet rank styles against each other or against other active treatments. In its own words, the evidence is "not yet sufficient to establish superiority over other active interventions".

The review stops short of telling anyone what to sign up for. Good. It proposes a reporting scheme, FITT + 3R, which extends the standard measures of frequency, intensity, time and type with three further items, so that future studies describe their dance sessions in a comparable way. It suggests physical screening, a warm-up and cool-down, graded intensity, and low-impact or seated alternatives, plus referral routes when a participant shows signs of acute risk. These recommendations, the authors write, are "pragmatic safeguards rather than validated decision rules".

The framework itself is a proposal.

The authors state that its clinical relevance and the best settings for each dimension "require prospective empirical testing".

Sources

• Rhythmic dance interventions and dance/movement therapy for depression and anxiety: a structured narrative review, Frontiers in Psychiatry, Chen, Wang, Wang and Tan, October 19, 2026
• Dancing to wellbeing: a three-wave longitudinal mixed-methods study of Rock "n" Roll Dancesport and psychosocial outcomes in Chinese university students, Frontiers in Public Health, He and Wu, October 2026
• Effects of dance-based physical activity on cognitive and psychological health across age groups: a systematic review, Frontiers in Psychology, Fan and Liu, October 2026