Aerobic Exercise in People with Schizophrenia: Neural and Neurocognitive Benefits

Abstract

Schizophrenia is characterized by extensive neurocognitive deficits, which are linked to greater disability, poorer functional outcome, and have been suggested to impact daily functioning more than clinical symptoms. Aerobic exercise (AE) has emerged as a potential intervention. This review examines the impact of AE on brain structure and function along with neurocognitive performance in individuals with schizophrenia. Preliminary evidence indicates that AE can increase hippocampal volume and cortical thickness, in addition to exerting a neuroprotective effect against hippocampal volume decrease and cortical thinning. There is also evidence that AE is able to significantly increase serum brain-derived neurotrophic factor (BDNF) levels, which are implicated in neurogenesis, neuroplasticity, and cognitive improvement. Finally, evidence suggests that AE plays a significant role in improving overall cognition, including improvements in processing speed, working memory, and visual learning. The authors discuss the implications of the findings and provide recommendations for future research and areas of inquiry.

Introduction

Neurocognitive deficits are well documented in individuals with schizophrenia, including impairments in memory, attention, processing speed, and executive functioning [1, 2]. These deficits predate the onset of illness and remain stable over time [3, 4]. Such impairments have also been linked to greater disability and poorer functional outcome in this population [5, 6], including decreased ability to maintain daily routines [7], independent living [8], social functioning [9, 10], and employment [11, 12]. In a large cohort of stable outpatients (N=672), the most prevalent impairments were in the domains of executive functioning and processing speed, while working memory impairment predicted the degree of functional disability [13]. Similarly, in a longitudinal study examining individuals with first-episode schizophrenia over 7 years, changes in verbal memory, processing speed, and attention significantly predicted functional impairment in relationships, employment, and ability to engage and enjoy recreational activities [14]. Overall, it has been suggested that neurocognitive deficits play a larger role in functional impairment than positive symptoms of schizophrenia [15], and consequently, cognitive impairment has emerged as a critical target for treatment [16].

Cognitive deficits in individuals with schizophrenia have been linked to structural and functional brain abnormalities. A review examining the link between brain structure and cognitive deficits in this population found that enlarged ventricles correlated significantly with cognitive inflexibility and deficits in language, attention, and psychomotor speed, while abnormalities in the hippocampus were associated with impairments in declarative memory and executive functioning [17]. Abnormalities in white matter integrity have also been reported, particularly in frontal and temporal cortices [18], and abnormal myelination of tracts responsible for communication between these regions has been implicated [19]. Smaller gray matter volumes in older individuals with schizophrenia have been reported [20, 21], though the progressive nature of these deficits is controversial. Larger structural deficits have been associated with increased exposure to antipsychotic treatment [22], leading some researchers to suggest both structural and cognitive deficits inherent to the illness are relatively stable over time [23]. Complementing these findings, neuroimaging studies have shown evidence of cortical tissue loss, ventricular enlargements, smaller thalamus and temporal lobes (including the hippocampus), enlarged caudate nucleus, and reversed cerebral asymmetries [24]. Similarly, cognitive deficits in people with schizophrenia have also been linked with the regulation of neurotrophins. In particular, brain-derived neurotrophic factor (BDNF) has attracted much research interest. A member of the neurotrophin family, BDNF is the most prevalent neurotrophin in the brain and is heavily involved in neuronal growth, development, and maintenance [25]. A recent meta-analysis of 16 studies showed evidence of significantly lower serum BDNF levels in schizophrenia [26], which has been linked to memory impairment [27] and smaller hippocampal volumes [28].

Treatment approaches for cognitive deficits in schizophrenia have primarily focused on pharmacology and cognitive remediation. Initially, second-generation antipsychotic medication appeared promising in addressing not only symptoms, but also cognitive impairment in this population [29]. However, results from the NIH-sponsored CATIE study examining the effects of typical and atypical antipsychotic medications on cognition in schizophrenia found only minimal improvement [30]. Investigations of novel add-on pharmacological compounds have yielded mixed results, with compounds targeting single receptors resulting in limited to no improvement [31–34] and compounds that have anti-inflammatory and/or anti-oxidant properties showing more promise [35, 36]. Cognitive remediation was also beneficial among individuals with first-episode psychosis [37, 38], as well as chronically ill individuals [39, 40]. However, a meta-analysis including 2104 individuals with schizophrenia demonstrated only small to moderate effect sizes on global cognitive improvement [41], and some studies failed to find any significant impact of cognitive training [42]. As a result, there remains a need for novel approaches to treating neurocognitive deficits in schizophrenia, particularly focused on fostering synaptogenesis and neurogenesis.

Recent studies have indicated a potential benefit of aerobic exercise (AE). Evidence from animal studies suggests that AE can have a positive impact on cognitive functioning [43], as well as strengthen neuronal growth and development [44, 45]. Most recently, mouse models of schizophrenia demonstrated that voluntary AE significantly increased BDNF levels in the hippocampus, improved spatial working memory, and encouraged increased social interaction [46]. Yet, studies investigating the effects of AE in animal models have noted differential effects of voluntary (spontaneous wheel running) versus involuntary or forced exercise (treadmill running). Both types of AE appear to upregulate BDNF levels in the hippocampus [47], improve spatial learning and memory, and increase activity-dependent BDNF in the hippocampus [48]. Leasure and Jones [49] reported that forced exercise actually resulted in more hippocampal neurogenesis as evidenced by more surviving progenitor cells compared to voluntary exercise at the expense of increased anxiety and emotional defecation. The authors also noted that animals engaging in voluntary exercise ran faster and for less time, while those in the forced exercise condition ran slower and for longer durations, suggesting the former group demonstrated greater exertion. In contrast, Ke et al. [50] found that voluntary exercise resulted in significantly higher hippocampal BDNF levels as compared to forced exercise and control groups, and forced exercise may induce greater stress resulting in downregulation of BDNF. These findings were contradictory to those reported by Leasure and Jones [49], who found that corticosterone levels were not significantly different between groups at end of study, and any increase in initial stress may be accounted for by habituation [48].

Similarly, findings among non-clinical and other clinical populations also provide support for AE positively affecting health outcomes [51–53], serum BDNF levels [54, 55], and cognition [56, 57]. Specifically, a meta-analysis of randomized clinical trials with 2049 adult participants demonstrated that exercise could lead to improvement in executive functioning, processing speed, attention, and declarative memory [58]. This is notable, since individuals with schizophrenia are often sedentary [59], and decreased physical activity has been implicated in cognitive impairment [60] and brain volume deficits [61•]. AE improves cardiopulmonary fitness in people with schizophrenia [62, 63] and preliminary evidence suggests AE may have a beneficial effect on cognitive functioning in schizophrenia. However, a comprehensive review of the current research literature is not available.

To address this gap in the literature, the primary aim of the present review is to examine the feasibility and efficacy of AE as an intervention for cognitive deficits in individuals with schizophrenia and bring the reader up to date regarding the most recent empirical evidence of AE’s impact on: (1) neurobiological structure and function and (2) neurocognitive performance. The secondary aim is to identify areas in need of further study. In order to obtain the most current literature, a search was conducted through January 2016 on Medline, Web of Science, and PsycNET for clinical trials examining the impact of AE on brain structure and function and neurocognition. Search words and terms included the following: “exercise” or “aerobic exercise,” “schizophrenia” or “psychosis,” “neurocognition” or “cognition,” “neurotrophin” or “BDNF” or “IGF” or “NGF” or “NT-3,” “brain structure,” and “hippocampal volume.” Relevant articles were pulled and reference lists examined for additional studies. The search yielded 11 studies, 8 of which are summarized in Table 1 and discussed more thoroughly below. Three remaining studies were only briefly mentioned in the review due to lack of comparison group (N = 1) and presentation of additional outcomes for a sample already included in review (N = 2).

Table 1.

Experimental design and outcomes of studies assessing impact of aerobic exercise in schizophrenia from 2010 to 2015

Study	Participants			Exercise intervention*	Control intervention	Assessment measures	Neurobiological outcomes	Neurocognitive/clinical outcomes
	SZE SZC		HCE
Pajonk et al. [64]	8	8	8	12 weeks, 3 sessions/week, 30 min/session Type: AE via cycling	12 weeks, 3 sessions/ week, 30 min/session Type: tabletop football	Physical: VO2max Imaging: SMRI Cognitive: RAVLT, Corsi block-tapping test Clinical: PANSS	Intervention group: hippocampal volume increased by 12 % in SZE 16 % in HCE Control group: hippocampal volume decreased by 1 % in SZC	Intervention group: STM increased by 34 % in SZE, decreased by 17 % in HCE. PANSS scores decreased by 9 % in SZE Control group: STM decreased by 6 % in SZC. PANSS scores increased by 13 % in SZC
Scheewe et al. [69]	18	14	25/27 (HC)**	6 months, 2 sessions/week, 60 min/ session Type: AE via cycling, rowing, crosstrainer (40 min) and 3 sets of muscle strength exercises, 10–15 repetitions/set (20 min)	6 months, 2 sessions/ week, 60 min/session Type: occupational therapy (SZC) and life as usual (HC)	Physical: VO2max Imaging: SMRI Clinical: PANSS	Intervention group: no sig. changes in global brain volume, hippocampal volume, or cortical thickness in SZE or HCE Control group: no sig. changes in SZC or HC	No neurocognitive data Intervention group: PANSS scores decreased by 10.75 % in SZE Control group: no sig. changes in PANSS scores in SZC
Malchow et al. [72]	20	19	21	12 weeks, 3 sessions/week, 30 min/session Type: AE via cycling	12 weeks, 3 sessions/ week, 60 min/session Type: tabletop football	Imaging: SMRI and VBM Clinical: PANSS	Intervention group: no sig. increases in hippocampus or hippocampal substructure volumes in SZE and HCE. Increased gray matter volume of the left anterior temporal in SZE vs SZC (p < .05) only during intervention period; no sig. differences 3 months post-study Control group: during intervention, increase in gray matter volume in anterior cingulate cortex and motor cortex bilaterally (p < .05). Volume decrease in left hippocampus 3 months post-study (p = .001)	No neurocognitive data Intervention group: no sig. changes in PANSS scores in SZE Control group: no sig. changes in PANSS scores in SZC
Kuo et al. [75]	33	–	30	10 weeks, 5 sessions/week, 20–30 min/session and 1 session/ week, 60 min/session of group physical activity Type: house cleaning/slow walking (30 min) or faster walking/ cycling (20 min); group activity type not reported	None reported	Physical: weight, BMI, waist circumference Neurotrophins: serum BDNF	Intervention group: serum BDNF levels increased (p < .001); weight, BMI, waist circumference decreased (p < .001) Control group: none reported
Kim et al. [76]	22	12	–	12 weeks, 3 sessions/week, 60 min/session Type: combination of AE via moderate walking (25 min) and strength training (25 min)	12 weeks, 1 session/ week, 60 min/session Type: stretching, dancing, and recreational activity	Physical: weight, BMI, YMCA 3-min step test and other physical functioning tests Neurotrophins: serum BDNF	Intervention group: serum BDNF values increased (p = .007), and positively correlated with fitness (p = .035) Control Group: no changes.
Silva et al. [80]	19	13	–	20 weeks, 2 sessions/week, 60 min/session Type (two arms): Resistance training only Resistance (30 min) and AE on treadmill (25 min)	20 weeks, 2 sessions/ week, 60 min/session Type: same protocol as AE intervention but equipment kept at lowest setting	Physical: weight, BMI, VO2max Neurotrophins: serum IGF-1, IGFBP-3, and BDNF Clinical: PANSS, CDSS, SF-36	Intervention group (1): no changes Intervention group (2): no changes Control group: no changes	No neurocognitive data Intervention group (1): PANSS scores decreased in SZE (p < .001); SF-36 physical role limitation scores improved (p = .011). No sig. changes in CDSS Intervention group (2): PANSS scores decreased in SZE (p = .003); SF-36 physical role limitation scores improved (p = .014). No sig. changes in CDSS Control group: no sig. changes
Kimhy et al. [82]	16	17	–	12 weeks, 3 sessions/week, 60 min/session Type: AE via Xbox active-play video games focused on whole-body exercises, treadmills, bicycle, and elliptical	Treatment as usual	Physical: VO2max Cognitive: MCCB Neurotrophins: serum BDNF	Intervention group: serum BDNF increased by 11 %. BDNF change explained 14.6 % of variance in neurocognitive improvement Control group: serum BDNF increased by 1.9 % (group difference was not sig.)	Intervention group: improved neurocognition by 15.1 %. Aerobic fitness explained 25 % of variance in neurocognitive improvement. Control group: 2.0 % decline in neurocognitive performance
Oertel-Knöchel et al. [83]	8	21	–	4 weeks, 3 sessions/week, 45 min/ session Type: AE via boxing and circuit training	Two arms: 4 weeks, 3 sessions/ week, 45 min/session Type: relaxation training, including breathing and imagery exercises (N = 11) Waitlist (N = 10)	Cognitive: MCCB Clinical: PANSS, RHS		Intervention group: sig. improvement of performance in visual learning, working memory and processing speed, both compared to baseline and to control groups. Significant reduction of PANSS negative symptoms scores in SZE. No sig. change in RHS Control group (1): sig. reduction of PANSS negative symptoms scores in SZC Control group (2): no changes

AE aerobic exercise, SZE individuals with schizophrenia receiving exercise intervention, SZC individuals with schizophrenia receiving control intervention, HCE healthy individuals receiving exercise intervention, HC healthy individuals receiving control intervention, Sig significant, SMRI structural magnetic resonance imaging, RAVLT Rey Auditory Verbal Learning Test, PANSS the Positive and Negative Syndrome Scale, STM short-term memory, VBM voxel-based morphometry, BMI body mass index, BDNF brain-derived neurotrophic factor, IGF-1 insulin-like growth factor 1, IGFBP-3 insulin-like growth factor-binding protein 3, CDSS Calgary Depression Scale for Schizophrenia, SF-36 Short Form-36 Health Survey, MCCB MATRICS Consensus Cognitive Battery, SAPS Scale for the Assessment of Positive Symptoms, SANS Scale for the Assessment of Negative Symptoms, BDI Beck Depression Inventory, BAI Beck Anxiety Inventory, STAI State-Trait Anxiety Inventory, RHS Revised Hallucination Scale

^*All exercise interventions were in addition to treatment as usual in individuals with schizophrenia

^**In this study, 25 healthy participants participated in exercise intervention and 27 additional healthy participants were recruited as exercise controls and continued life as usual (HC)

Impact of Aerobic Exercise on Brain Structure and Functioning

Three randomized control trials examined the effects of AE on brain structures. Pajonk et al. [64] compared males with schizophrenia (N = 8) and age-matched healthy male controls (N = 8) who engaged in 3-month AE program (30 min per session, three times per week), along with a schizophrenia patients comparison group (N = 8) that played table football for a similar duration and frequency. Comparison of changes in hippocampal volumes assessed indicated significant increase in volume in both the schizophrenia and control AE groups (12 and 16 %, respectively) versus a 1 % decrease in hippocampal volume in the non-exercise schizophrenia group. No changes were found in total brain or gray matter volume in either group. A second article examining cortical thickness in the same sample found no significant changes in either group [65]. Overall, these findings demonstrate that despite neuronal atrophy [66, 67] and lower levels of presynaptic proteins [68] in the hippocampus of individuals with schizophrenia, some degree of neurogenesis in the region was possible through AE.

A second trial by Scheewe et al. [69] randomized 32 individuals with schizophrenia and 52 age-matched healthy controls to exercise and control conditions. The exercise program, which included both cardiovascular and strength training exercises, lasted for 6 months, but at reduced frequency compared to Pajonk et al. (twice per week for 60 min; 40 min of AE and 20 min of strength training). At baseline, participants with schizophrenia had smaller cerebrum, gray matter, and ventricle volumes, as well as reduced cortical thickness compared to healthy controls. Post-intervention, the authors failed to find significant differences in brain volume, hippocampal volume, or cortical thickness. There was a trend (p = .05) reported for hippocampal volume change; schizophrenia group demonstrating mild decrease in volume post-exercise, while volume did not change significantly in health individuals post-intervention. The study also reported a significant association between improved aerobic fitness and increased volume of cerebral cortex and ventricles, as well as cortical thickening. Supporting these results, a recent small pilot study conducted by Rosenbaum et al. [70] also reported no changes in hippocampal volumes post-exercise treatment, despite significant improvement in aerobic fitness. However, a study revisiting the sample examined by Sheewe et al. revealed that all participants who engaged in AE irrespective of clinical status demonstrated some improvement in white matter integrity within the motor tracts, while non-exercising participants showed a decrease in integrity in the same regions [71]. These findings indicate that overall, exercise improved brain connectivity in this sample and highlight the importance of outcome variable selection when evaluating the effects of AE. Taken together, results suggest that while exercise at the frequency implemented by Scheewe et al. may not directly promote change in brain structures, it may play a neuroprotective role through improvement in cardiorespiratory fitness and attenuating the expected decreases in brain volume and white matter integrity observed in schizophrenia.

The most recent trial exploring the link between exercise and brain structures appears to support this hypothesis. Malchow et al. [72] utilized the same exercise intervention as Pajonk et al. [64] with a larger sample per group (range = 19–21). However, they failed to find significant changes in hippocampal volumes in both the schizophrenia and control groups post-treatment, though the authors reported observing a temporary increase in gray matter volume in the temporal lobe for the duration of the intervention in the schizophrenia exercise group. The study also evaluated participants 3 months post-intervention to assess lingering effects of exercise. Interestingly, the gray matter volume increase was no longer found in the schizophrenia exercise group, and the schizophrenia control group displayed evidence of a statistically significant decrease in volume. This appears to be in line with Scheewe et al.’s finding and supports the hypothesis that exercise may exert a neuroprotective effect.

A possible explanation for the discrepancy in imaging findings is differences in the frequency and intensity of the provided intervention. Preliminary findings indicate that fidelity with target training intensity is significantly correlated with improvement in cognitive performance in this population [73]. Similarly, a recent meta-analysis of exercise interventions in schizophrenia found that trials that administered at least 90 min per week of moderate-to-vigorous exercise that saw significant improvement not only in fitness, but also in psychiatric symptoms [74•]. Additional data regarding frequency, duration, and in-session intensity would be helpful in determining if the dose-response relationship is a contributor to seeing significant structural and neurobiological changes.

Four additional studies examined brain functioning by exploring the link between AE and circulating neurotrophins, with three studies focusing on BDNF and one study focusing on insulin-like growth factor (IGF-1) and insulin-like growth factor-binding protein (IGFBP-3). Kuo et al. [75] led the first investigation of the link in comparing obese individuals with schizophrenia (N = 33) and age- and weight-matched healthy controls (n = 30) before and after a 10-week weight reduction intervention, which included AE, dietary changes, and nutritional counseling. BDNF serum levels in the schizophrenia group were significantly lower at baseline. Post-intervention, those individuals who participated in the program demonstrated 4.65 ng/ml (21 %) increase (p < .001) in serum BDNF. Of note, the primary aim of the study was to determine the effects of weight reduction on circulating BDNF. Given the multi-faceted approach of the program to reduce weight, our inferences regarding the impact of exercise alone are limited. However, the authors were able to demonstrate that methods to reduce weight may have a significant impact on serum BDNF, supporting a further investigation of AE.

Kim et al. [76] took this step further by implemented a 3-month exercise program that combined 25 min of AE with 25 min of strength training three times per week. The participants were all individuals with schizophrenia, randomized into exercise and non-exercise groups. Post-treatment, the authors found that the exercise group had significantly higher serum BDNF levels. In addition, the group also displayed a significant improvement in aerobic fitness, which positively correlated with BDNF levels. Given decreased serum BDNF levels in schizophrenia [77] and its implication in various cognitive processes [78, 79], this study was the first to explore directly a combined exercise program as a method of modulating serum BDNF levels in this population.

Silva et al. [80] examined whether resistance training plays a significant role in upregulating neurotrophins. Male individuals with schizophrenia were assigned randomly to three groups: resistance training, combination of resistance and AE, and control group. Intervention was 60 min, two times per week, for 20 weeks. Baseline assessments of neurotrophins BDNF, IGF-1, and IGFBP-3 (IGF’s binding protein) were reported; however, no data was provided regarding statistical significance between groups. Furthermore, the authors failed to find any significant differences in neurotrophin levels within or between groups post-intervention. Although these results are surprising, several limitation of the study could account for the variation in results. The authors did not assess for aerobic fitness, but noted that participants experienced no weight reduction. Both aerobic fitness and loss of body weight were significantly associated with elevations in BDNF, as reported earlier. Furthermore, antipsychotic medication dosages, which have been shown to impact serum BDNF levels [81], were significantly different between groups. Finally, similar to the imaging findings described earlier, the dose and intensity of the intervention could be impacting these findings.

A recent study by Kimhy et al. [82] examined individuals with schizophrenia who were randomized to attend AE (N = 16) or control (treatment as usual; N = 17). The intervention consisted of 60-min sessions, three times per week, for 12 weeks, and participants exercised from 60 to 75 % of their maximal heart rate, increasing gradually across the initial first 4 weeks of treatment. Post-intervention, changes in serum BDNF levels significantly contributed to the variance in neurocognitive functioning, accounting for 14.6 % of the variance in AE-associated cognitive improvement. The difference between groups was not significant (11 vs. 1.9 %, respectively); however, potentially due to the small sample size.

Overall, three out of four studies demonstrated that AE significantly impacts serum BDNF levels. In turn, higher BDNF levels were correlated with improved aerobic fitness and accounted for a significant portion of cognitive improvement in one of the trials. One study failed to find any differences in serum neurotrophin levels [80]. Interestingly, this study employed a similar exercise regimen to a trial that also failed to find significant changes in brain volume, hippocampal volume, or cortical thickness [69], suggesting that an insufficient weekly dose of intervention could potentially explain the difference in results.

Impact of Aerobic Exercise on Neurocognitive Performance

Three studies examined the link between AE and neurocognition. Pajonk et al. [64] assessed short- and long-term memory using Rey Auditory Verbal Learning Test (RAVLT) and visuospatial short-term memory using the Corsi block-tapping test. Results indicated that after a 12-week AE intervention, individuals with schizophrenia in the exercise group showed a 34 % increase in short-term memory while the scores of exercising healthy controls dropped by 17 %. In addition, non-exercising patients’ scores dropped by 6 %. Interestingly, the authors also examined associations between short-term memory scores and hippocampal volume, which were positively correlated, but only when the schizophrenia was combined to include both exercising and non-exercising participants. Results for long-term memory and visuospatial memory were more complex. The former decreased and the latter increased somewhat in the combined exercise group (healthy and schizophrenia participants), though the there was no significant group × time interaction in the combined schizophrenia group (exercising and non-exercising participants). Taken together, these results suggest that the mechanisms by which exercise exerts its effects on memory may differ between healthy controls and individuals with schizophrenia.

Oertel-Knöchel et al. [83] examined the effects of AE and cognitive training on neurocognitive functioning in schizophrenia. The study consisted of an AE + cognitive training group, relaxation + cognitive training group, and a waiting control group (no intervention), to which participants were randomly assigned. AE intervention involved 45-min sessions three times per week for 1 month. Cognitive training consisted of 30-min sessions three times per week. Cognition was assessed using the Measurement and Treatment Research to Improve Cognition in Schizophrenia (MATRICS) Consensus Cognitive Battery (MCCB). MCCB assesses eight major cognitive domains, including verbal and non-verbal working memory, processing speed, attention, verbal learning, visual learning, problem-solving, and social cognition. Oertel-Knöchel et al. [83] assessed five of the eight domains, excluding attention, problem-solving, and social cognition. The results from post-intervention assessment indicated that cognitive performance improved in both the relaxation and AE groups; however, the effect was much stronger in the combined treatment group. Specifically, participants demonstrated significant improvement in processing speed, working memory, and visual learning. This indicates that AE played a crucial role in improving cognitive functioning independent of cognitive training. Contrary to expectations, the authors failed to find significant post-treatment effects in verbal learning. This is surprising, as recent studies have found significant improvements in this domain with exercise intervention [84].

Our group has also examined neurocognitive functioning as part of the 12-week BDNF study described earlier [82]. In this protocol, neurocognition was also assessed using the MCCB pre and post-treatment phase. Individuals with schizophrenia in the exercise group demonstrated 15.1 % improvement in overall cognitive functioning, while performance in the treatment as usual group decreased by 2 %. The AE intervention produced a large effect (d = .93) with 77 % of the participants improving in performance by 5 % or more. Of note, improvement in aerobic fitness was significant in the exercise group, and it positively correlated with improvements in the social cognition and visual learning domains of the MCCB. Further analysis revealed that after controlling for demographic and clinical variables, aerobic fitness explained 25.4 % of the variance in neurocognitive improvement. Given that cognitive deficits assessed by the MCCB in individuals with schizophrenia have been found to be between one and two-and-a-half standard deviations below the norms for healthy individuals [85, 86], AE-associated changes in neurocognitive functioning accounted for 24 % in improvement in the current sample, and suggesting the elimination of approximately 16–40 % of neurocognitive deficit generally reported in this population. These results provide strong support for AE as a safe, feasible, and effective intervention.

Conclusion and Future Directions

Empirical research investigating the neural and neurocognitive effects of AE in schizophrenia remains limited, but the emerging evidence is promising. In terms of AE impacting brain structure, the literature is divided. One study observed lasting increases in hippocampal volumes of exercising individuals in schizophrenia, while another observed temporary changes in gray matter volume that disappeared shortly after the intervention ended. A third one failed to find any increases in gray matter volume but noted that participants who did not receive an exercise intervention displayed decreases in gray matter, suggesting that AE and improvement in aerobic fitness may play a neuroprotective role in schizophrenia. This is a plausible hypothesis, as research in neurodegenerative illnesses such as Alzheimer’s and dementia has come to similar conclusions about the neuroprotective role of AE in brain structure and function [87, 88].

The majority of studies reviewed also demonstrated evidence that AE is able to significantly increase serum BDNF levels. This is particularly crucial, since improvement in cognitive functioning is related to the brain’s ability to modify neural connectivity and function. Such neuroplasticity is dependent on the support of neurotrophins that regulate neuronal survival, differentiation, and growth [89, 90], and influence cellular energy metabolism [91]. Indeed, our group demonstrated an increase in serum BDNF accounted for 14.6 % of the variance in AE-associated cognitive improvement in schizophrenia [82]. Future research should confirm these findings and examine the effects of AE interventions on other neurotrophin serum levels, as well as the impact of neurotrophin levels on mediating the link between AE and cognition. In particular, assessing the impact of AE interventions on vascular endothelial growth factor (VEGF), is also indicated. AE has been shown to upregulate VEGF in the periphery [92] and increased VEGF has been linked to greater hippocampal neurogenesis in animal studies [93]. Other potential targets of investigation include nerve growth factor (NGF), neurotrophin-3 (NT-3), and neurotrophin-4 (NT-4).

Research has also examined potential mechanisms via which AE is able to upregulate neurotrophin levels, specifically BDNF. Cotman and colleagues [94, 95] suggest that IGF-1 is directly upregulated by exercise in the periphery, and increased IGF-1 increases BDNF signaling in the central nervous system and periphery by both increasing BDNF’s signaling during AE and increasing the availability of BDNF’s binding receptor, tropomyosin receptor kinase B (TrkB). Additionally, findings have linked lower BDNF levels with increased hypertension [96] and metabolic syndrome [97]. This is particularly relevant in schizophrenia, as this population is at increased cardiometabolic risk after initiating antipsychotic treatment [98]. AE has been shown to effectively reduce this risk [99], suggesting that AE may be responsible for reducing impact of systematic risk factors on BDNF levels. Further study of the processes by which AE is able to upregulate BDNF and other neurotrophins in the context of schizophrenia-related co-morbidities would clarify the mechanism of action of AE in this population.

Regarding cognitive performance, preliminary evidence supports AE as an effective intervention to significantly improve global cognitive functioning. However, there is some disagreement regarding which cognitive abilities are improved through exercise. Future studies can elucidate and confirm initial findings that AE can specifically improve processing speed, working memory, visual learning, and verbal learning in this population.

In addition, the degree to which AE is able to produce changes in neurocognition, as well as brain structure and functioning, appears to be governed by a dose-response relationship. Future studies should confirm these preliminary findings and include close monitoring of adherence to intensity targets that will improve assessment of interventions and allow for comparison between studies. To increase enjoyment and decrease attrition in future programs, employment of engaging activities as part of AE programs is encouraged. For instance, integration of active-play video game systems has recently been shown to be safe and effective in this population [100•]. Current literature also highlights the importance of aerobic fitness and its positive correlation with volume changes in brain structures. Implementation of a standardized measure of aerobic fitness, such as the maximal oxygen consumption fitness test (VO₂ max), is recommended in future trials.

Finally, the majority of studies in the current review also reported clinical and functional improvements for individuals participating in AE, though information was limited. This supports findings that increased sedentary behavior and lower aerobic fitness in schizophrenia are associated with functional deficits in the domains of physical functioning, interpersonal relationships, daily activities, and work skills [59]. Similarly, significant associations between higher body mass index (BMI) and more extensive cognitive impairment [59, 101] have been reported, and studies have suggested that improvement in aerobic fitness may lead to greater health-related quality of life [102]. Taking into account the evidence thus far, future research should include measures of daily, social, work, and community functioning to better capture the real-world impact of AE-associated neural and neurocognitive changes. Overall, the present review found preliminary evidence for the benefits of AE to enhance neural and neurocognitive outcomes among individuals with schizophrenia.