Everyday function in schizophrenia: The impact of aerobic endurance and skeletal muscle strength

Background: Patients with schizophrenia suffer from physical health conditions, culminating in reduced physical functioning with enormous costs for patients and society. Although aerobic endurance and skeletal muscle strength, typically reduced in this population, relate to cognition and function, no study has explored their respective contributions to performance of functional skills and everyday tasks. Methods: In a cross-sectional study, 48 outpatients (28/20 men/women; 35 ± 11(SD) years) with schizophrenia spectrum disorders (ICD-10; F20 – 25) were administered the UCSD Performance-based Skills Assessment-Brief (UPSA-B; functional skills), Specific Level of Functioning (SLOF; functional performance) and the Positive and Negative Syndrome (PANSS) scale. Peak oxygen uptake (V ˙ O 2peak ) was assessed along with leg press maximal muscle strength (1RM) and mechanical power. Results: UPSA-B performance was associated with V ˙ O 2peak ( r = 0.28, p < 0.05), accounting for 8 % (p < 0.05) of shared variance, but was unrelated to 1RM and mechanical power. The SLOF physical functioning domain was associated with V ˙ O 2peak ( r = 0.30, p < 0.05) and 1RM ( r = 0.24,p < 0.05), while SLOF personal care ( r = 0.27,p < 0.05) and activities (r = 0.30,p < 0.05) were related only to V ˙ O 2peak . Hierarchical regression analyses revealed that while V ˙ O 2peak and age combined to account for 20 % (p < 0.05) of the variance in physical functioning, the contribution of 1RM was eliminated after adjusting for age. V ˙ O 2peak and negative symptoms combined predicted 24 % and 35 % of the variance in personal care and activities, respectively. UPSA-B scores did not add to the prediction of SLOF scores. Conclusions: Although V ˙ O 2peak and 1RM both relate to functional outcomes, the combination of V ˙ O 2peak , age, and negative symptoms exert the greatest detrimental influence on functional performance beyond skills deficits.


Introduction
Schizophrenia spectrum disorders are one of the leading causes of disability globally (GBD, 2017 Disease andInjury Incidence andPrevalence Collaborators, 2018).Over the past few decades the burden of disease has been augmented due to population growth and aging (Whiteford et al., 2013).Schizophrenia, a neurodevelopmental condition, affects the patients' ability to successfully achieve functional milestones (Bowie et al., 2012).Unfortunately, these impairments are present in young (Ho et al., 1997) and middle-aged adults (McGlashan, 1988), leading to lifelong disability.The impairments also commonly predate onset of illness (Bowie et al., 2012;Patterson et al., 2001;Twamley et al., 2002).Whereas functional capacity refers to critical skills measured in a clinical setting, such as physical, social, or vocational abilities, functional performance quantifies the (non)use of skills in the real world (Leifker et al., 2009;Patterson et al., 2001).Although cognitive remediation and supplemental skills training have yielded promising results on everyday function, patients' functioning is seldom fully restored after treatment (Bowie et al., 2012).As such, improving our understanding of the range of determinants of poor everyday functioning in this vulnerable patient group is of great importance.
There are indications that physical capacity, both in terms of aerobic endurance and skeletal muscle strength, may be related to cognition and functional performance in patients with schizophrenia (Holmen et al., 2019;Kimhy et al., 2014;Strassnig et al., 2017).These two distinctly different physiological systems, relating to oxygen transport and neuromuscular force production, respectively, are elicited to various degrees during daily tasks which require different amounts of effort and time.In turn, peak oxygen uptake (V˙O 2peak ) is commonly considered the most important factor for aerobic endurance performance, i.e. the ability to sustain multiple low force contractions over a longer period (Bassett Jr. and Howley, 2000;Aagaard et al., 2010).One repetition maximum (1RM) and mechanical power, on the other hand, are often acknowledged as the measures of choice to determine a subject's skeletal muscle strength (Nygard et al., 2021;Aagaard et al., 2010), i.e., the ability to perform maximal or rapid high force muscle contractions, respectively, over a shorter period.However, studies investigating these important physiological measures' potential impact on cognitive function, functional skills, and performance are scarce.Differentiating their independent roles in everyday function in patients with schizophrenia is warranted.
The few studies that have investigated the effect of endurance training on cognition have yielded contrasting findings, with one showing enhanced social functioning (Kimhy et al., 2021) and another showing no additional gains when implemented as adjunct therapy to cognitive remediation (McGurk et al., 2021).Evidence regarding the effect of strength training in patients with schizophrenia appears to be more conclusive but limited, with an improvement in cognitive functioning as an outcome (Strassnig et al., 2015b).Reduced handgrip strength is also documented to associate with cognitive deficits in the patient group (Firth et al., 2018b), with cognitive impairments known to be a primary cause of functional impairments (Green et al., 2000;Velligan et al., 1997).However, it is less well understood if such a relationship exists for functionally relevant lower extremity musculature.Interestingly, several studies have shown that the impact of impaired cognition on functional performance is mediated through functional skills (Bowie et al., 2006;Galderisi et al., 2014;Leifker et al., 2009;Mausbach et al., 2010).The question arises as to whether V˙O 2peak , 1RM, and mechanical power potentially influence functional performance directly, or if their potential contributions to performance are mediated through an effect on functional capacity.If the physical inability to perform activities in real life is the limiting case, a functional skills assessment in the laboratory may not capture these influences.
Despite increasing interest in the role of aerobic endurance and skeletal muscle strength on functional outcomes in patients with schizophrenia, investigations examining their individual importance are clearly lacking.Therefore, the aim of the current study was to investigate the respective associations between directly assessed V˙O 2peak , lower extremity 1RM, and mechanical power with functional capacity (UCSD Performance-based Skills Assessment-Brief, UPSA-B) and functional performance (The Specific Level of Functioning, SLOF).We also sought to examine the predictive hierarchy for reduced functional performance in outpatients with schizophrenia.Specifically, we hypothesized that both V˙O 2peak , 1RM, and mechanical power would (a) independently be positively associated with UPSA-B and SLOF scores, and (b) exert their influence on SLOF scores directly beyond, rather than through, the mediating effect of UPSA-B.

Patients
Forty-eight outpatients diagnosed with schizophrenia spectrum disorders (International classification of diseases, ICD-10; F20-25) who volunteered to take part in long-term exercise training (The LEXUS trial; ClinicalTrials.govidentifier: NCT02743143) were included in the current study.Exclusion criteria were life-threatening or terminal medical conditions, current pregnancy, mothers <6 months post-partum, admittance to a psychiatric ward, or inability to perform training and testing.Patients were given both oral and written information about the study prior to inclusion, and informed consent was collected from all patients before testing commenced.Study eligibility was assessed through a medical examination by a physician, while diagnosis and medication status were confirmed by a senior psychiatrist through medical records.Both analogue and digital records were screened to obtain information regarding hospitalization.Test procedures were undertaken at the Exercise Training Clinic at Department of Psychosis and Rehabilitation, Psychiatry Clinic, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway.The study was performed in line with the Declaration of Helsinki and approved by the regional committee for medical and health research ethics (REC Southeast).

Physiological testing: Peak oxygen uptake, maximal skeletal muscle strength, and mechanical power
All patients undertook one familiarization session (involving both treadmill and leg press testing) on a separate day before any further physiological testing.All other physiological tests were done on the same day in our lab.Height (to the nearest cm) and weight were measured before the V˙O 2peak test.
The same metabolic test device (MetaMax II, Cortex Biophysik Gmbh, Germany) was used for all tests to obtain data on V˙O 2peak on an individualized incremental treadmill (PPS Med, Woodway Gmbh, Germany) test to exhaustion, previously utilized in both healthy participants (Helgerud et al., 2007) and participants with schizophrenia (Heggelund et al., 2011).A Polar monitor (V800, Polar Electro, Kempele, Finland) was used to measure heart rate.The treadmill test started with the patients instructed to walk at a predetermined workload of 4.5 km • h − 1 and 5 % incline for 5 min to warm up.The workload was increased (either 0.5-1.0km • h − 1 or 2-3 % inclination) every 2-3 min until the patients reached exhaustion.The patients were strongly encouraged by the test personnel to maximize their effort during the latter stages of the test.To determine V˙O 2peak , the mean of the highest oxygen uptake measurements over 30 s was used, with the highest measured heart rate collected to establish HR peak .
Lower extremity 1RM and mechanical power were measured in a 40 • incline leg press apparatus (Hammer Strength HSLLP, Life Fitness, Franklin Park, IL, USA).All lifts during the strength and mechanical power testing were done starting with the leg position near complete extension, down to 90 • knee joint angle with a short (< 1 s) stop, before attempting to lift the weight with maximal mobilization of force as fast as possible.Separated by 3-5 min rest periods, 1RM was attained through 4-8 attempts, with only 2-4 lifts of >80 % of 1RM.External load was increased by 5-20 kg until failure, and the lift with the highest completed load was used to determine 1RM.After the 1RM test was completed, patients proceeded to the lower extremity mechanical power examination.At a load corresponding to 70 % of 1RM, patients were given 3 attempts, interspersed with 3 min of rest.The Muscle Lab System (Ergo test Technology, Langesund, Norway) was used to measure the time used in the concentric lifting phase.External force of the lift, in combination with work distance and liftingtime, were used to calculate mechanical power in N • m • s − 1 (Storen et al., 2008): In which force (N) = weight lifted x 9.81 m ⋅ s − 2 .

The University of California san Diego (UCSD) performance-based skills assessment-brief
The Brief version of the UCSD Performance-based Skills Assessment (UPSA-B) is applied to directly measure functional skills in outpatients diagnosed with serious mental disorders (Mausbach et al., 2010).Through standardized role-playing scenarios, the test assesses functioning skills via two sub-tests relating to 1) finance, and 2) communication.For the finance sub-test, patients are asked to conduct a series of tasks, e.g., paying an invoice, counting change, and read a utility bill; the communication sub-test requires the patients to use an unplugged telephone to e.g., make an emergency call, change a medical appointment, and dial a number from memory.The test is completed within 10-15 min.Raw scores are then converted into scaled scores of 0-50 for each sub-test and 0-100 for the total score, with higher values indicating better performance.

The specific level of functioning
The Specific Level of Functioning (SLOF) questionnaire was used to examine the patients' real-world everyday functional performance (Schneider and Struening, 1983).The items are scored on a 5-point Likert scale relating to the frequency of a certain behavior or/and independence level over the past week.The SLOF assesses functioning through 43 items in the following six domains: 1) physical functioning (e.g., vision, hearing, walking, and use of hands/arms; scored from 5 to 25); 2) personal care skills (e.g., personal hygiene, dressing/grooming self, care of own possessions and living space; scored from 7 to 35); 3) interpersonal relationships (e.g., accepting and initiating contact with others, effective communication, and formation/maintenance of friendships; scored from 7 to 35); 4) social acceptability (e.g., verbal and physical abuse of self or others, destruction of property, repetitive behaviors; scored from 7 to 35); 5) activities (e.g., household responsibilities, communication/financial handling, public transportation, avoiding dangers, use of medical and community services; scored from 11 to 55); and 6) work skills (e.g., employable skills, work with minimal supervision, sustaining/completing work tasks; scored from 6 to 30).
The SLOF scale is typically rated by caregivers who are well familiar with the patients' functioning, preferably key caregivers, but also case managers and family members.The informants all indicated that they knew the outpatients "very well" on a 5-point Likert scale.None of the informants were aware of the patients' results from UPSA-B or the Positive and Negative Syndrome Scale (PANSS) assessment, and the SLOF was conducted within 30 days of the other measures included in this study.For each domain, results were summarized to provide both sub-scores and a total score (from 43 to 215), with higher values suggesting higher levels of functioning.

The positive and negative syndrome scale
Two experienced psychiatric nurses at the department certified with utilizing the PANSS (Kay et al., 1987) assessment conducted the interviews with the patients to examine symptom severity.The scale includes 30 items rated from 1 to 7, divided into three domains: the positive scale (seven items, scored from 7 to 49), negative scale (seven items, scored from 7 to 49), and general psychopathology (16 items, scored from 16 to 112), accompanied by a total score (30− 210), with increasing scores indicating greater symptom severity.

Statistical analyses
All statistical analyses were undertaken with SPSS Statistics version 29.0 (IBM Corp., Armonk, NY, USA).Figures were produced with GraphPad Prism version 9.3 (GraphPad Software Inc., Boston, MA, USA) or BioRender.com(Toronto, Canada).Data are reported as mean ± standard deviation unless otherwise stated.Normality was examined by visual inspection of P-P and scatter plots of standardized residuals and predicted values from the dependent variables.Consequently, the linear associations between variables were calculated using Pearson correlation coefficient.Hierarchical multiple regression analyses were conducted with data from UPSA-B and SLOF as dependent variables.No corrections for multiple testing were applied, but to minimize the number of statistical tests, and prevent overfitting of the models, regression analyses were done only when physiological variables (V˙O 2peak , 1RM, or mechanical power) were found to manifest zero-order correlations with UPSA-B or SLOF outcomes.Physiological variables were entered in block 1, demographic variables (age and sex) in block 2, with clinical variables (defined as daily dose of antipsychotic medication, and negative symptom score from the PANSS) entered in block 3. To assess if UPSA-B scores mediated the relationship between physiological variables and SLOF scores, UPSA-B total scores were included in block 4. Additional partial correlation analyses were done with SLOF scores as dependent variable to describe the individual contribution of each predictor to the overall model variance.Multicollinearity among predictor variables was assessed with variation inflation factor (VIF) estimates.VIF ranged from 1.003 to 1.100, indicating that multicollinearity was not an issue within the included data.A significance level of <0.05 for two-tailed tests was chosen for all analyses.

Patients
Forty-eight outpatients diagnosed with schizophrenia spectrum disorders (F20 Schizophrenia: 43; F25 Schizoaffective disorders: 5; 58 % men) were included in this study, with a mean age of 35 ± 11 years (Table 1).Of the 48 patients, 45 were currently receiving antipsychotic treatment.V˙O 2peak , 1RM, and mechanical power are presented in Table 1.

Associations between functional outcomes and peak oxygen uptake, maximal skeletal muscle strength, and mechanical power
UPSA-B total score was associated with V˙O 2peak (p < 0.05; Table 2; Fig. 1A), while UPSA-B total and subscale scores were unrelated to 1RM or mechanical power.(See Fig. 2.) The SLOF physical functioning domain was associated with V˙O 2peak (p < 0.05) and 1RM (p < 0.05).The personal care skills and activities domains were associated only with V˙O 2peak (both p < 0.05; Fig. 1B-C).

Predictors of functional outcomes
To assess the relative contribution of V˙O 2peak , 1RM, and mechanical power to functional skills and performance, hierarchical multiple regression analyses were undertaken with functional scores as dependent variables (Table 3).As no other demographical or clinical variable was found to associate with the UPSA-B total score, the overall model included only V˙O 2peak as predictor variable, accounting for 8 % (p < 0.05) of the variance.
Given that both V˙O 2peak and 1RM were found to associate with the SLOF physical functioning domain, separate analyses were conducted with age as additional predictor variable.The overall model, including V˙O 2peak and age, accounted for 20 % (p < 0.05) of the variance in physical function.Including UPSA-B total score did not change the overall fit of the model (p = 0.73).V˙O 2peak and age, respectively, accounted for 7 % and 12 % of the variance in physical functioning.
The overall model with 1RM and age as predictor variables accounted for 16 % (p < 0.05) of the variance in SLOF physical functioning, although the inclusion of age eliminated the unique contribution of 1RM.The overall fit of the model remained unchanged by including UPSA-B total score (p = 0.34).1RM and age, respectively, accounted for 2 % and 12 % of the variance in physical functioning.
For the SLOF personal care skills domain, V˙O 2peak (p < 0.05) and PANSS negative factor scores (p < 0.01) both contributed uniquely to the model, as the overall model explained 24 % of the outcome variance.Including UPSA-B total score did not change the overall fit of the model (r 2 = 0.29; Δ r 2 = 0.05; p = 0.09).V˙O 2peak and PANSS negative factor, respectively, explained 8 % and 18 % of the variance in personal care skills.
For the SLOF activities domain, the overall model explained 35 % of the outcome variance, with both V˙O 2peak (p < 0.05) and PANSS negative factor (p < 0.001) contributing uniquely to the model.UPSA-B total score did not change the model overall fit (p = 0.60).V˙O 2peak and PANSS negative factor, respectively, explained 11 % and 29 % of the variance in SLOF activities.

Discussion
Patients with schizophrenia typically suffer from multiple physical and mental health issues, culminating in severely attenuated daily functioning.Our study sought, for the first time, to quantify the independent contributions of aerobic endurance and skeletal muscle strength on functional skills and everyday functional performance.Our main findings were that: (a) Only V˙O 2peak was associated with the total score from the financial and communication skills assessment (UPSA-B); (b) V˙O 2peak and lower extremity 1RM were related to the physical functioning, personal care, and activities performance domains (SLOF); (c) Adjusting for demographic/clinical variables, 1) V˙O 2peak , 2) age, and 3) negative symptoms predict SLOF performance; and (d) these factors associate directly with SLOF performance beyond the effects of UPSA-B.Taken together, our results suggest that while both V˙O 2peak and lower extremity 1RM relate to functional outcomes, V˙O 2peak , age, and negative symptoms in combination appear to exert an additional, and substantial, detrimental influence on functional performance beyond skills deficits.

Functional skills assessment: The role of peak oxygen uptake, maximal skeletal muscle strength, and mechanical power
Our study revealed that out of V˙O 2peak , lower extremity 1RM, and mechanical power, only V˙O 2peak was positively associated with functional skills in outpatients with schizophrenia, accounting for 8 % of the variance in the UPSA-B.Although no study to date has previously documented the association between directly measured V˙O 2peak and functional skills in outpatients with schizophrenia, the relationship observed between these two factors are as expected.The association of V˙O 2peak with both cognitive abilities and functional performance in this patient group is previously established, and as such our results corroborate, and extend, the results from other studies (Holmen et al., 2019;Kimhy et al., 2014).V˙O 2peak is previously shown to account for 22 % of the variance in cognitive functioning (Kimhy et al., 2014), measured as the MATRICS Consensus Cognitive Battery composite score.Further, as cognition and functional skills are closely related (Harvey et al., 2007), albeit not completely overlapping, the strength in the relationship between V˙O 2peak and UPSA-B observed in the current study appears reasonable.
In contrast to our hypothesis, no relationship between lower extremity 1RM or mechanical power with UPSA-B was observed in our Values are mean ± standard deviation, or number of patients.Data from psychiatric services are median (and range) due to non-normal distribution of variables.ICD, International classification of diseases; PANSS, Positive and Negative Syndrome Scale; V˙O 2peak , peak oxygen uptake; HR peak , peak heart rate; 1RM, one-repetition maximum; m b , body mass.† External load lifted in a 40 • (incline) angle.

Table 2
Functional skills, performance, and correlations with peak oxygen uptake, maximal skeletal muscle strength, and mechanical power in outpatients with schizophrenia spectrum disorders.included outpatients.This is somewhat surprising, particularly considering that results from over 1100 individuals with schizophrenia have previously described an association between handgrip strength with working memory and processing speed (Firth et al., 2018b).Reversal of aging-related white matter degradation in the brain following strength training is suggested as a likely mediator of this association (Firth et al., 2018b).In addition, the relationship between cognition and functioning is well established, as poor performance in working memory and processing speed are prevalent and shown to increase the risk of attenuated social and occupational abilities in the patients (Hofer et al., 2005;Nuechterlein et al., 2011).However, one possible explanation of our contrasting findings may be related to test modality.As cognitive tests can be administered by computerized touchscreen interface, the association with handgrip strength may arguably be an expression of hand speed/dexterity (Firth et al., 2018a;Lyall et al., 2016), rather than skeletal muscle strength.However, it is also possible that the association between skeletal muscle strength and cognition (Firth et al., 2018b) does not uniformly extend to functional skills, which are influenced by additional factors (Harvey et al., 2007).A lack of statistical power (Firth et al., 2018b) may also be an issue in our study.Our cohort's representativeness is strengthened by the outpatients being recruited from outpatient clinics, supported housing services, and district psychiatric centers.However, including an even more heterogenous sample in terms of age and disability/functioning status may also have improved our ability to disseminate the contribution of skeletal muscle strength on functional skills.

Functional performance, peak oxygen uptake, maximal skeletal muscle strength, and mechanical power in outpatients with schizophrenia
V˙O 2peak and lower extremity 1RM, rather than mechanical power, were related to physically demanding functional performance domains, i.e., physical functioning, personal care skills, and daily activities, assessed by the SLOF.To the best of our knowledge, no other study has described the independent contributions of directly measured V˙O 2peak , lower extremity 1RM, or mechanical power on functional performance in this patient group.However, our results are in line with our hypothesis, given that reduced aerobic endurance and skeletal muscle strength previously are shown to be related to poor functioning (Nygard et al., 2019;Strassnig et al., 2017;Vancampfort et al., 2015;Vancampfort et al., 2013).
After adjusting for relevant demographic/clinical variables, V˙O 2peak and negative symptoms most strongly predicted SLOF performance, independently contributing 8-11 % and 18-29 % to the variance in personal care skills and daily activities, respectively.Notably, negative symptoms, which are related to reduced motivation, not only associate with reductions in active behaviors, but commonly persist over full days and as a consequence may in turn impact lifetime levels of functional milestone achievement (Martinuzzi et al., 2022;Perez et al., 2022).Several other, potentially complementary, pathways contributing to this relationship may also exist, directly or indirectly, such as depressive symptoms, self-esteem, or social withdrawal (Kirschner et al., 2017;Leifker et al., 2009;Strassnig et al., 2015a), thus the influence of these factors cannot be excluded.However, the finding that both V˙O 2peak and negative symptoms independently appear to contribute to functional performance suggests that a combination of low V˙O 2peak and motivated behavior are important aspects of poor functional outcomes in outpatients with schizophrenia.
In our data, the independent contribution of lower extremity 1RM in predicting physical functioning was eliminated after taking age into consideration.This may not be particularly surprising, given that physical activity/training volume commonly declines with age in the general population (Edvardsen et al., 2013;Westerterp, 2018), accompanied by a reduction in maximal skeletal muscle strength by 8-10 % per decade after the 4th decade of life (Lindle et al., 1997).As a consequence, it is difficult to describe to what extent lower extremity   1RM and age independently influence the outcome in physical functioning, as physical activity/training volume is a likely covariate influencing both factors.One possible explanation of the lack of association may be that the SLOF questionnaire is unable to discriminate between activities of daily living which require different amounts of force.This likely holds true also for the lack of association between mechanical power and functional performance.In fact, both lower extremity 1RM and mechanical power are previously documented to associate moderately (both r = 0.45-0.50)with weight-bearing physical functioning, e. g., expressed as stair climbing ability and chair raising, in outpatients with schizophrenia (Nygard et al., 2019).Further, functional performance is arguably influenced by personal, societal, or environmental factors, interfering with whether the patients' physical capacities are in fact deployed in the everyday real life context, as documented by the SLOF (Bowie et al., 2006).Importantly, our results showed that UPSA-B did not add to the prediction of SLOF outcomes when V˙O 2peak and lower extremity 1RM were considered, indicating that these parameters may influence functional performance directly rather than through skills.Of note, this finding contrasts with previous studies, where the relationship between cognition and SLOF is mediated through UPSA-B (Bowie et al., 2006;Leifker et al., 2009).Other studies have shown that physical capacity can be more important for certain real-world outcomes than cognitive functioning or symptoms (Strassnig et al., 2017).These data suggest that reduced competence in functional skills is less important in this sample than limitations in V˙O 2peak .Thus, a combination of reduced V˙O 2peak and motivation may be the pathway toward limitations in functional performance in certain domains.

Clinical implications
V˙O 2peak and lower extremity 1RM are documented to be attenuated in patients with schizophrenia (Brobakken et al., 2020;Brobakken et al., 2022;Nygard et al., 2021).As these factors are associated with functional skills and performance in the current study, future studies aiming to increase V˙O 2peak and lower extremity 1RM in clinical physical training studies should explore whether functional skills and performance are indeed also improved.This would also provide information on directionality/causality of the relationships.Importantly, a physical training intervention would seem to have significant potential to improve functional performance outcomes, as skills deficits, although related to some elements of endurance, are not predictive in the final equations.In addition, given that the collective contribution of V˙O 2peak and negative symptoms predict reduced functional performance most strongly, clinical treatment strategies targeting these factors may be an important and appropriate area of future research.It is possibly the case that motivation might be augmented by improved fitness, potentially through improvements in perceptions of abilities and increases in selfefficacy.

Conclusions
In conclusion, the current study provides evidence that attenuated aerobic endurance and skeletal muscle strength are associated with poor functional outcomes in outpatients diagnosed with schizophrenia.Specifically, while only V˙O 2peak was associated with functional skills, both V˙O 2peak and lower extremity 1RM were related to physically demanding functional performance domains.Notably, however, V˙O 2peak , age, and negative symptoms collectively appear to exert a substantial detrimental influence on functional performance, beyond skills deficits, in this patient group.

Role of the funding source
This study was supported by grants from the Norwegian Extra-Foundation for Health and Rehabilitation, The Norwegian Directorate of Health and The Liaison Committee for education, research and innovation in Central Norway.The funding sources did not partake in the design, collection, analysis, interpretation of data, writing the manuscript, or the decision to submit the article for publication.

Fig. 1 .
Fig. 1.The relationship between peak oxygen uptake and A) functional skills (UPSA-B Total score) and B -C) functional performance (from the SLOF).Lines represent best linear fit with Pearson correlation coefficients.Only relationships between physiological variables (V˙O 2peak , 1RM or mechanical power) and functional outcome measures (UPSA-B or SLOF) with p < 0.05 after adjusting for demographical and clinical predictor variables (Table3) are presented.V˙O 2peak , peak oxygen uptake; UPSA-B, The University of California San Diego Performance-Based Skills Assessment-Brief; SLOF, Specific Level of Functioning.

Table 1
Characteristics of all 48 included outpatients diagnosed with schizophrenia spectrum disorders.

Table 3
Functional skills, performance, and physiological, demographical, and clinical predictors in outpatients with schizophrenia spectrum disorders.