Examination of Rehabilitation Responsiveness in Schizophrenia using Growth Mixture Modeling

Background and Purpose. A growing body of research has shown that neurocognition predicts community functioning in schizophrenia. Neurocognition is theorized to be a therapeutic mediator between psychosocial treatments and functional outcomes. However, previous studies have not investigated differences in treatment responsiveness to community-based psychosocial rehabilitation which is seen as an essential step in developing and disseminating more effective community rehabilitation services. The present study aimed to: 1) identify heterogonous treatment response sub-groups over time in community-based psychosocial rehabilitation; 2) examine the predictors of the treatment response sub-groups and 3) explore neurological, social, and psychiatric characteristics of treatment response sub-groups.

Methods. The sample consisted of 130 individuals diagnosed with schizophrenia who were recruited upon admission to three community-based psychosocial rehabilitation interventions. The subjects were assessed at baseline and then prospectively at 6 and 12 months on the Role Functioning Scale which assesses work, social, and independent living functioning. Neurocognition was measured at baseline and 12 months later. Neurocognition was measured with five tests reflecting verbal fluency, working memory, episodic memory, sustained attention, and mental flexibility. Psychiatric symptoms were assessed using the Brief Psychiatric Rating Scale. Treatment intensity was measured as the number of days treatment was received during the one-year study period. Data were analyzed using Latent Growth Mixture Modeling.

Results. Growth Mixture Model analyses revealed two latent class trajectory groups which discriminated individuals who showed a significant relationship between treatment intensity and functional change during rehabilitation (treatment responders) and those who did not show a significant relationship between service intensity and functional change during rehabilitation (treatment non-responders). In the next step, we found that individuals with improvements in neurocognition were about five times more likely to be the treatment responders than those with non-improvements in neurocognition (OR = 4.99). Finally, exploratory logistic regressions revealed that Individuals at baseline who were younger (OR = .92), had more social contacts (OR = 1.56), had better neurocognition (OR = 1.19) and who had less symptomatology (OR = .94) were more likely to be in the treatment responder class.

Conclusion and Implications. Based on our findings we propose a neuro-psycho-social model for understanding responsiveness to intensive psychosocial rehabilitation in the community. This model suggests that a combination of neurological, psychiatric, and social variables predicts the latent class of treatment responsiveness in schizophrenia. This suggests that comprehensive and interdisciplinary theoretical approaches are needed to understand complex issues like treatment responsiveness in schizophrenia, and that translating findings across theoretical boundaries is essential. The findings also suggest that interventions that focus on impacting neuropsychological, symptom and functional variables could contribute to better psychosocial functioning but also to greater rehabilitative treatment gains over time by improving responsiveness to rehabilitative interventions. We recommended that clinicians should consider neurocognition an essential therapeutic mediator and not an obstacle to rehabilitative success in community-based psychosocial rehabilitation programs.