The hippocampal formation (HF), located within the medial temporal lobe, comprises several subfields with specialized cytoarchitecture, including the Cornu ammonis (CA), dentate gyrus (DG), presubiculum, and subiculum. The HF plays a vital role in memory, navigation, and cognition, particularly episodic memory. Subfields contribute distinctly to memory formation; rostral subfields (DG and CA2/3) are involved in encoding, while caudal subfields (subiculum) aid in retrieval. The HF is part of a tightly interconnected network. In schizophrenia, HF atrophy is well-established, with subfield-specific vulnerability, particularly in CA1, CA4, DG, and the subiculum. These morphological and physiological abnormalities are associated with cognitive deficits, especially verbal memory. Aerobic exercise is a promising intervention to counteract neurodegeneration and promote neuroregeneration. Meta-analyses suggest aerobic exercise prevents age-related HF volume decline in healthy individuals, with positive correlations between aerobic fitness and HF volume. However, studies in schizophrenia are scarce and inconclusive, often focusing on overall HF volume rather than subfields. This study aimed to investigate the interrelations between aerobic fitness, HF subfield structure, functional connectivity, and verbal memory in individuals with schizophrenia using a multiparametric research design. The hypothesis was that positive associations exist between aerobic fitness and HF subfield volumes, and that aerobic fitness is positively related to functional connectivity within the HF and between the HF and other brain regions involved in cognition. It also tested the hypothesis that these associations mediate verbal memory functioning.

Existing literature demonstrates hippocampal formation (HF) volume loss as a consistent finding in schizophrenia, impacting cognitive function. Specific subfields, such as CA1, CA4, DG, and the subiculum, show greater vulnerability. Studies in healthy populations show positive correlations between aerobic fitness and HF volume, suggesting exercise's neuroprotective potential. However, research on this relationship in schizophrenia is limited and inconclusive, lacking multiparametric MRI studies examining subfield volumes, functional connectivity, and cognitive outcomes.

This cross-sectional study analyzed baseline data from the Enhancing Schizophrenia Prevention and Recovery through Innovative Treatments (ESPRIT) C3 study. Fifty-three participants with schizophrenia underwent brain structural and functional magnetic resonance imaging (MRI) and assessments of aerobic fitness and verbal memory. Aerobic fitness was measured using a lactate threshold test on a bicycle ergometer. Structural T1-weighted images were processed with Freesurfer v7.2 to segment 38 HF subfields. Resting-state fMRI (rs-fMRI) data were pre-processed with fMRIPrep, and functional connectivity between HF seeds and other brain regions (parahippocampal gyrus, middle frontal gyrus, cingulate gyrus) was computed using Nilearn. Verbal memory was assessed using the Verbal Learning and Memory Test (VLMT). Multivariate multiple linear regressions (MMLR) were performed to determine associations between aerobic fitness and HF subfield volumes and functional connectivity, with mediation analyses exploring the relationship between aerobic fitness, HF variables, and VLMT scores.

Multivariate multiple linear regression analysis revealed a significant positive association between aerobic fitness and the volumes of most HF subfields (p=0.003). The strongest associations were observed in the cornu ammonis, dentate gyrus, and subiculum. No significant associations were found between aerobic fitness and HF functional connectivity. Mediation analyses did not show significant effects of HF subfield volumes on verbal memory, after correcting for false discovery rate. However, before correction, some hippocampal subfields showed a significant positive association with verbal memory performance.

This study provides evidence that higher aerobic fitness levels are associated with larger hippocampal subfield volumes in individuals with schizophrenia, particularly in areas most vulnerable to volume reduction in this population. These findings support the potential of aerobic fitness to mitigate neurodegeneration or promote neuroregeneration within the hippocampus. The lack of significant associations between aerobic fitness and functional connectivity or mediation effects on verbal memory requires further investigation, possibly due to the cross-sectional nature of the study and limited sample size. The results suggest that the relationship between aerobic fitness and hippocampal volume may be complex and potentially dependent on factors not included in the study.

Higher aerobic fitness levels are positively associated with greater hippocampal subfield volumes in schizophrenia, particularly in subfields known to be vulnerable in this disorder. This suggests potential neuroprotective effects of aerobic fitness. Future longitudinal studies with larger sample sizes are needed to confirm these findings and investigate causal relationships, potentially exploring the impact on different cognitive domains and accounting for individual genetic factors.

This cross-sectional study cannot establish causality. The study's statistical power was limited, potentially missing small effects. Data on menstrual cycle phase in women was not collected. Analysis was limited by the availability of high-resolution MRI data from only one study center. Genetic risk factors and subgroup characteristics that might moderate effects were not examined.