Background: Expressed emotion (EE) is a measure of the family environm
ent that has been demonstrated to be a reliable psychosocial predictor
of relapse in schizophrenia. However, in recent years some prominent
nonreplications of the EE-relapse relationship have been published. To
more fully address the question of the predictive validity of EE, we
conducted a meta-analysis of all available EE and outcome studies in s
chizophrenia. We also examined the predictive validity of the EE const
ruct for mood disorders and eating disorders. Methods: An extensive li
terature search revealed 27 studies of the EE-outcome relationship in
schizophrenia. Using meta-analytic procedures, we combined the finding
s of these investigations to provide an estimate of the effect size as
sociated with the EE-relapse relationship. We also used meta-analysis
to provide estimates of the effect sizes associated with EE for mood a
nd eating disorders. Results: The results confirmed that EE is a signi
ficant and robust predictor of relapse in schizophrenia. Additional an
alyses demonstrated that the EE-relapse relationship was strongest for
patients with more chronic schizophrenic illness. Interestingly, alth
ough the EE construct is most closely associated with research in schi
zophrenia, the mean effect sizes for EE for both mood disorders and ea
ting disorders were significantly higher than the mean effect size for
schizophrenia. Conclusion: These findings highlight the importance of
EE in the understanding and prevention of relapse in a broad range of
psychopathological conditions.