The need to develop personalized interventions to improve cognition in schizophrenia

Green et al1 provide a review of the evidence on neurocognitive and social cognitive deficits in schizophrenia. These deficits span the course of the disease, starting from the prodrome, and are stable over time. Impairments in neurocognition involve learning and memory, vigilance/attention, speed of processing, reasoning and problem solving, and working memory. Social cognition deficits affect psychological processes implicated in the perception, encoding, storage, retrieval and regulation of information about other people and ourselves. The underlying neurobiological disturbances have their origin in brain networks involving the hippocampus as well as temporal, parietal and prefrontal cortex. Here we discuss the central role of the hippocampus in cognitive processes and the impact of non‐pharmacological treatments on this brain structure in schizophrenia patients.

The hippocampus has been implicated in episodic and working memory. In schizophrenia patients, associations were recently detected between hippocampal subregion volumes and cognitive performance in visual and verbal memory as well as working memory domains2. Among the most prominently altered hippocampal subregions in schizophrenia are cornu ammonis 4 (CA4)/dentate gyrus (DG) and CA2/3. In post‐mortem brains of schizophrenia patients, along with reduced volumes of these subregions, we detected a reduced number of oligodendrocytes (the myelin‐forming glia cells) in the left CA4 and a reduced number of neurons in the DG3. The reduced number of oligodendrocytes in the left CA4 was related to cognitive deficits in these patients.

These changes might in part be a consequence of disturbed neuro‐regenerative mechanisms in the brain4. This hypothesis is supported by findings of reduced synaptic proteins and dysregulation of structural synaptic elements in the temporal lobes in schizophrenia. The converging lines of evidence suggest that episodic memory dysfunction in schizophrenia might well be caused by a disturbance of synaptic and neuronal plasticity and disconnectivity4.

Understanding the underlying neurobiology of cognitive dysfunction is critical to allow researchers to develop pathophysiology‐based innovative treatment strategies. So far, however, efforts to develop new pharmacological treatments have been disappointing. Among promising non‐pharmacological add‐on interventions for cognitive impairments, Green et al¹ propose aerobic exercise. This treatment has been suggested to promote neuroplasticity at the synaptic level and to improve neurogenesis, at least in animal models. Moreover, epigenetic mechanisms may also be involved.

Green et al¹ mention a recent meta‐analysis of controlled trials investigating cognitive outcomes of aerobic exercise interventions in schizophrenia. Meta‐regression analyses indicated that greater amounts of exercise were associated with greater improvements in global cognition. Among the cognitive domains, aerobic exercise improved working memory, social cognition, and attention/vigilance5. Effects on verbal memory were not among the significant results, but this subdomain was only measured in six studies, which limits the strength of findings in this meta‐analysis.

To achieve meaningful real‐world functional benefits, Green et al suggest to combine cognitive remediation with aerobic exercise. In fact, in a three‐month aerobic exercise study, in which bicycle ergometer training augmented with cognitive remediation was compared with table soccer plus cognitive remediation, we found improvement in everyday functioning of schizophrenia patients measured with the Global Assessment of Functioning (GAF) scale, and in social adjustment measured with the Social Adjustment Scale (SAS‐II). The ability to work was associated with improvement in verbal memory and processing speed6.

Short‐ and long‐term verbal memory scores and cognitive flexibility performance were increased in schizophrenia patients and healthy controls receiving the endurance training augmented with cognitive remediation at three months versus six weeks, but this was not observed in those receiving table soccer augmented with cognitive remediation6. This finding supports the need to perform long‐lasting training programs to improve cognitive deficits in this severely affected patient group. We previously detected an increase in hippocampal volume after a three‐month endurance training, but we could not replicate this finding in our second study.

On the basis of the hypothesis that the individual genetic risk load for schizophrenia – which contributes to neuroplastic processes in the brain – plays a role in the response to aerobic exercise, we calculated the schizophrenia polygenic risk score in our sample. Volume changes in the left CA4/DG at three months versus baseline were significantly influenced by polygenic risk score in schizophrenia patients performing aerobic exercise. A larger genetic risk burden was associated with a less pronounced volume increase or even a volume decrease over the course of the exercise intervention7.

Results of exploratory enrichment analyses reinforced the notion that genetic risk factors modulate biological processes that are tightly related to synaptic ion channel activity, calcium signaling, glutamate signaling, and regulation of cell morphogenesis7. Interestingly, the CA4/DG region was again most affected, which corresponds to our post‐mortem findings in schizophrenia3. We hypothesize that a high polygenic risk may negatively influence neuroplastic processes during aerobic exercise in schizophrenia, indicating a gene x environment interaction.

Besides the need to replicate these findings in independent samples, future studies are needed to identify those patients who benefit from aerobic exercise interventions and to assess the effects of individual genetic and environmental factors on treatment‐induced improvements in cognitive abilities. This would contribute to the development of a personalized approach to improve cognition in schizophrenia.