Self‐Criticism as a Mechanism of Change in Compassion‐Based Interventions: A Meta‐Analytic Structural Equation Modelling Approach and Systematic Review
Meta-analýza
Typy studií podle spolehlivosti
Od nejspolehlivějšího po nejméně spolehlivý.
-
Souhrn více studií (meta-analýza, systematický přehled) ← tahle studie
nejspolehlivější obraz, jaký dnes máme. -
Randomizovaný pokus (RCT)
lidé byli rozděleni náhodně, takže rozdíl lze přičíst zásahu. -
Porovnání skupin
skupiny se porovnaly, ale bez náhodného rozdělení; naznačuje účinek. -
Pozorování nebo rozhovory
popisuje, co se děje a co spolu souvisí; příčinu to neukáže. -
Názor odborníků, teorie nebo jeden případ
nic se neměřilo na skupině lidí, jde o zkušenost a výklad.
Typ studie říká, jak byla udělaná, ne jak velký byl účinek.
Nový výzkum
Jak nová je studie
Podle roku vydání; klasika stojí stranou.
-
Nový výzkum ← tahle studie
vyšla letos nebo loni. -
Zavedené
vyšla předloni nebo dřív. -
Klasika
práce, na které obor staví.
Stáří neříká, jak je studie spolehlivá. To říká typ studie.
Ve zkratce
Compassion-based therapy lowers self-criticism, and that drop partly explains why symptoms improve. Across 57 randomised trials with 4,877 adults, effects were small to moderate. Most trials had high risk of bias (76.4%), so treat these results as promising but uncertain.
Celý souhrn studie Context · Findings · Implications · Co nepřeceňovat
Context
Compassion-based interventions teach people to respond to themselves with kindness rather than harsh judgement. They include compassion-focused therapy and mindful self-compassion programmes. Research shows they help with many mental health problems. But how they help is less clear. The authors looked at self-criticism: the habit of attacking or judging yourself harshly. Self-criticism is common across depression, anxiety and other difficulties. The authors asked one specific question. Do compassion-based interventions work partly by reducing self-criticism? To answer it, they combined 57 randomised controlled trials with 4,877 adults. Studies were published between 2015 and 2025 and came from 24 countries. About one quarter used non-Western samples. Most measured depression and anxiety. Self-criticism was usually measured with the Self-Compassion Scale or the Forms of Self-Criticising/Attacking and Self-Reassuring Scale.
Findings
The meta-analysis reports moderate reductions in self-criticism after treatment (r = -0.20, 95% CI -0.27 to -0.13). Symptoms also fell moderately (r = -0.21, 95% CI -0.27 to -0.15). Self-criticism and symptoms were strongly linked (r = 0.48, 95% CI 0.42 to 0.54). The mediation analysis found a small but reliable indirect effect (indirect effect = -0.054, 95% CI -0.085 to -0.030). The direct effect stayed significant, so self-criticism explains only part of the change. Two factors moderated the results. More daily meditation minutes predicted a bigger drop in self-criticism. Longer programmes showed a weaker indirect effect. The type of self-criticism mattered. Interventions reduced inadequate-self criticism (r = -0.263) more than hostile hated-self criticism (r = -0.104). Effects were larger against waitlists (r = -0.262) than against active controls (r = -0.066). Depression improved more than anxiety. The evidence has serious limits. 76.4% of trials had high risk of bias, mostly from unclear treatment adherence and missing data. Only one trial directly tested self-criticism as a mediator, with mixed results. Most trials measured only before and after treatment, so the timing of change is unknown. The mediation is statistical, not causal.
Implications
Self-criticism is worth assessing and tracking during compassion-based work. Targeting it directly may help, though no trial has yet shown that doing so improves outcomes. Expect slower change in contemptuous, hated-self criticism. Set clear daily practice and check how well clients do it, not just how often. The evidence does not settle whether self-criticism must change before symptoms improve.
Co nepřeceňovat
Do not tell clients that compassion training is proven to work through self-criticism. The mediation is statistical, not causal, and most trials measured only before and after. Do not promise large effects: correlations of about 0.20 are small to moderate. Do not claim superiority over other active treatments, because the effect against active controls was close to zero (r = -0.066). Do not generalise to trauma, addiction, eating disorders or personality disorders, since few trials covered them. Mention that 76.4% of trials had high risk of bias.
Zdroje
- Aida Palacios, Desirée Colombo, Belén Fernández‐Castilla, Maja Wrzesien, Ausiàs Cebolla (2026). Self‐Criticism as a Mechanism of Change in Compassion‐Based Interventions: A Meta‐Analytic Structural Equation Modelling Approach and Systematic Review. Clinical Psychology & Psychotherapy. https://doi.org/10.1002/cpp.70347