Dose-Response Meta-analysis of Cognitive Behavioral Therapy for Parental Stress in Parents of Children with Special Needs.

Abstract

To assess how cognitive behavioral therapy (CBT) influences stress levels in parents of children with special needs, while also exploring the relationship between therapy intensity and possible influencing factors. We followed PRISMA guidelines and searched PubMed, Web of Science, PsycINFO, and the Cochrane Library through June 15, 2025, for randomized controlled trials. We defined PICOS explicitly as Population (parents of children with special needs), Intervention (CBT), Comparator (any control condition), Outcomes (parental stress measured by validated scales), and Study design (randomized controlled trials). Continuous outcomes were pooled as standardized mean differences (SMDs; differences in parental stress scores between the intervention and control groups). A three-level random-effects meta-analytic model was fitted to account for dependent effect sizes; heterogeneity was assessed with I² and the Q test. Sensitivity analyses and publication-bias assessments were performed. Subgroup and dose-response analyses were conducted to explore moderators. Sixteen trials met inclusion criteria; after exclusion of two outliers, 14 trials (n = 1007) were included in the primary analysis. There was substantial heterogeneity (I² = 92.3%). Pooled results indicated that CBT significantly reduced parental stress (SMD = - 0.57; 95% CI - 1.00 to - 0.14; p = 0.0095). Interventions with moderate session duration (≈ 90 min) delivered once weekly over several weeks (total ≈ 11-12 weeks) were associated with larger pooled effects in our subgroup analyses. Larger effect estimates were also observed when stress was measured with the Depression Anxiety Stress Scales (DASS), when control groups received no active intervention, and among parents of children with neurodevelopmental disorders (NDDs). Trials with predominantly mothers (> 90%) showed significant pooled effects within that subgroup. Dose-response analyses suggested diminishing returns when session frequency exceeded twice weekly; a total intervention time of approximately 14.42-17.42 h was associated with the most significant average reductions in stress. CBT was associated with a moderate pooled reduction in parental stress across included trials. The intervention protocol most strongly associated with larger average effects in our dataset was "≈ 90 min per session × once weekly × 11-12 weeks"; however, this should not be interpreted as a universally optimal protocol given the high between-study heterogeneity and uncertainty in some subgroup estimates. Future RCTs with larger samples, more consistent outcome measurement, prespecified dose-finding designs, and long-term follow-up are warranted to refine CBT dosage thresholds and to validate generalizability across different cultures and disability types.