Schizophrenia and bipolar disorder (BD) are considered severe mental disorders. Cognitive impairments are commonly present at illness onset in both disorders and are related to poorer functional outcomes. This has given rise to questions about the role of cognition in the nature and development of these disorders. Findings of neuroanatomical and neurophysiological changes and more severe deficits in chronic samples have engendered neurodegenerative models of the disorders. Though schizophrenia is now generally regarded as a neurodevelopmental condition, the question of neuroprogression in BD remains disputed. However, several factors could explain the observed severe cognitive deficits in chronic stages of the illnesses, including confounding clinical factors and sampling bias. For this reason, long-term follow-up studies of recently diagnosed individuals are needed. The overarching aim of this study was to investigate the long-term intellectual and cognitive course in first-episode schizophrenia and first-treatment BD.
The aim of the first paper was to compare trajectories of intellectual functioning in schizophrenia, BD, and healthy controls (HC). Linear mixed models of intellectual course were built using a premorbid estimate as well as measurements of IQ at baseline and 10-year follow-up. Group comparisons showed the expected cognitive deficits in the clinical groups with more marked deficits in schizophrenia. Clinical groups also showed a drop from premorbid to post-onset IQ, again with a more marked decline in schizophrenia. Over the follow-up period, however, all three groups had parallel improvements, indicating long-term stability in intellectual functioning.
The second paper aimed to investigate the 10-year cognitive course in schizophrenia, providing an in-depth perspective on domain-level changes and individual differences. Eight cognitive domains and a cognitive composite score were used. Trajectories of stability or improvements parallel to HC were found in most domains in both clinical groups. However, attention and cognitive control declined in the schizophrenia group, as opposed to in the controls. Clinically significant cognitive deficits were prevalent in the schizophrenia group, although few individuals had clinically significant changes in cognition. The findings show domain-specific cognitive changes as well as a high degree of inter-individual variability in the course and level of cognition in schizophrenia.
In the third paper, the aim was to elucidate the long-term cognitive course in first-treatment BD. Eight cognitive domains and a cognitive composite were used to model cognition in the BD group and HC from baseline to 10-year follow-up. We found parallel changes in both groups, with stability and improvements in all domains. No correlations were found with symptom levels.
The results show stable cognitive deficits and are broadly inconsistent with neurodegeneration as a model of schizophrenia and BD. The studies were limited by high attrition leading to small sample sizes at follow-up, precluding complex statistical models and subgroup analyses. Future research should focus on subgroup-level trajectories, investigate mediating factors in the cognitive course, and explore the potential long-term benefits of early intervention on cognitive outcome.