Parkinson’s: CBT appears to reduce anxiety and depressive symptoms, with variable results

A review of 21 randomised trials

The researchers pooled 21 randomised trials conducted in adults with Parkinson’s disease: 465 participants were in the CBT groups and 427 in the control groups. Their literature search covered four scientific databases through October 2025.

The included interventions involved at least four sessions spread over at least four weeks. Comparators could be usual care, a waiting list or an active control. These different conditions do not provide a single point of comparison: the overall result therefore does not describe the advantage of CBT over any one specific alternative.

An average benefit in scores, not a cure rate

For anxiety symptoms, assessed in 16 trials, the standardised mean difference was −0.66, with a 95% confidence interval ranging from −0.83 to −0.50. For depressive symptoms, assessed in 14 trials, it was −0.61, with an interval ranging from −0.78 to −0.44. In both cases, the results favour CBT.

These values allow results obtained using different scales to be pooled. They represent neither a percentage of patients cured nor a reduction directly interpretable as points on a single questionnaire. The abstract provides no absolute result that would allow individual chances of improvement to be quantified.

The authors describe the reductions as clinically meaningful. Without the full text or the thresholds used for this assessment, however, this conclusion cannot be examined in detail.

Variability limits what the result tells us

Heterogeneity between trials is substantial: I² reaches 79.1% for anxiety and 86.6% for depressive symptoms. The average benefit therefore masks markedly different results across studies.

Subgroup analyses suggest larger effects for interventions lasting more than ten weeks and with certain scales adapted to the population studied. They are not sufficient to establish an optimal treatment duration.

Removing each trial in turn did not undermine the overall estimates. Statistical tests detected no significant publication bias, but this does not rule out its existence.

What we can take away

This review supports the potential value of CBT for anxiety and depressive symptoms associated with Parkinson’s. It demonstrates neither an effect on disease progression nor the possibility of replacing other care.