Strength training integrated in long term collaborative care of patients with schizophrenia

Introduction: Skeletal muscle strength is reduced in patients with schizophrenia, contributing to their impaired physical health, functional performance, and potentially mental health challenges. Although short-term training programs have shown promising results, improving muscle strength and functional performance, it is unknown how exercise can be successfully integrated into the long-term clinical care of outpatients with schizophrenia. Objective: To investigate effects of strength training with adherence support in a collaborative care model. Methods: We randomized 28 men and 20 women (mean ± SD, 35 ± 11 years) to leg press maximal strength training (MST) with 4 sets at 90 % of one repetition maximum (1RM) 2 × week, facilitated by municipal service and professional supervision (TG), or a control group (CG). Results: The TG increased scaled leg press 1RM (0 – 3 months: 19 %; 0 – 6 months: 31 %, 0 – 12 months: 40 %, all p < .001, and 3 – 12 months: 18 %, p < .05) and power (0 – 3 months, 11 %; 0 – 6 months: 22 %, 0 – 12 months: 26 %, all p < .001, and 3 – 12 months: 13 %, p < .05) throughout the 1-year period compared to the CG. The increased muscle strength was accompanied by improved sit-to-stand performance (20 %) after 12 months ( p < .001). Both groups also exhibited within-group improvements in walking work efficiency after 6 months (TG: 13 %; CG: 23 %) and 1 year (TG: 11 %; CG: 21 %, p < .01 – 0.05), but with no evident differences between the groups. Stair climbing performance remained unchanged. Conclusion: Our results reveal that strength training can successfully be integrated as a part of long-term clinical care of outpatients with schizophrenia, contributing to improved functional performance.


Introduction
Patients with schizophrenia have attenuated physical health (Heggelund et al., 2011;Brobakken et al., 2019aBrobakken et al., , 2019b;;Nygard et al., 2019), likely contributing to their 15-20 years lower life expectancy (Hjorthoj et al., 2017), increased risk of cardiovascular disease (Osborn et al., 2007), and mental health challenges (Tandon et al., 2009).Skeletal muscle strength and (aerobic) endurance are key components of physical health and both contribute independently to the patients' attenuated physical health.Skeletal muscle strength is often assessed as maximal strength in a given exercise; one repetition maximum (1RM).Another relevant measure of skeletal muscle strength is how fast an individual can develop skeletal muscle force.This is of particular interest for functional performance when time to develop maximal muscle strength is not available, such as in rapid movements and balance adjustments (Maffiuletti et al., 2016).Power (force × velocity) and rate of force development (RFD) are expressions of skeletal muscle force development.
To improve skeletal muscle strength, strength training is the modality of choice rather than physical activity or endurance training, which have little or no effect (Grandys et al., 2008;Unhjem et al., 2016aUnhjem et al., , 2016b;;Unhjem et al., 2017).Strength training should be carried out with loads >70 % of 1RM, resulting in <12 repetitions per set (Schmidtbleicher and Haralambie, 1981;McDonagh and Davies, 1984).A few previous studies applying this principle have examined training programs of relatively short (8-20 weeks) duration in patients with schizophrenia and have shown promising results (Heggelund et al., 2012;Leone et al., 2015;Silva et al., 2015;Nygård et al., 2021), with increases in 1RM ranging between 20 and 40 %.Interestingly, Nygård et al. (2021) showed that the patients' muscle strength was restored to a level similar to healthy references.The increased leg press 1RM was accompanied by increased leg press power, which in turn was strongly associated with the 30-second sit-to-stand test (30sSTS) functional performance.It is, however, a timely question how strength training can be included in long-term clinical care of patients with schizophrenia as they often suffer from psychotic and negative symptoms, depression, apathy, lack of initiative, and cognitive impairment (Firth et al., 2016), making exercise adherence challenging.
Globally, people are becoming less physically active, damaging their physical and mental health (Guthold et al., 2018).When even the general population struggles with performing regular exercise, it is an even greater challenge to help patients with additional mental barriers to engage in long-term exercise.Although it appears that patients with schizophrenia tolerate heavy strength training well (Heggelund et al., 2012;Nygård et al., 2021), a lack of supervision will likely result in low adherence or termination of exercise (Firth et al., 2015).To meet these challenges, the aim of this study was to investigate if patients with schizophrenia could maintain exercise participation, including maximal strength training (MST) with heavy loads (~90 % of 1RM), for 1 year, when receiving transportation service and training supervision.The outcome was compared to controls introduced to the same training and given advice to continue exercising without further support.We hypothesized that after 1 year of MST patients randomized the training group (TG) would have higher 1) leg press 1RM, 2) leg press power and 3) better functional performance (walking work efficiency; sit-to-stand performance; stair climbing) compared to controls.

Design
This study is based on the 1-year results of the Long-Term Exercise Training Therapy versus Usual Care in Patients with Schizophrenia trial (LEXUS).This randomized controlled trial aimed at investigating the 1year effects of combined aerobic interval training and MST alongside outpatient adherence support.All included patients were given training gear (a set of clothes, shoes and water bottle), transportation to appointments at the clinic and a gift certificate at the local sports store after all test periods.A municipal healthcare professional, acting as a project coordinator, remained in close contact with study participants throughout its duration to schedule appointments, offer transportation and motivation as needed.
Testing was conducted on three separate days; day 1: ensuring eligibility, a physician conducted a medical examination, and patients completed the 36-Item Short Form Health Survey (SF-36); day 2: examination of functional performance tests and familiarization on treadmill and leg press device; day 3: aerobic endurance and strength testing.All functional and physical performance tests were conducted by the same exercise physiologists at 3 months, 6 months and 1 year, while the SF-36 was completed at 3 months and 1 year.
Testing and training sessions were done between January 2016 and February 2019 at the Exercise Training Clinic at the Department of Psychiatry, St. Olavs Hospital, Trondheim University Hospital, Norway.The trial was conducted in line with the Declaration of Helsinki (ClinicalTrials.govidentifier: NCT02743143) and the Regional Ethics Committee (REC South-East) approved the study prior to patient inclusion.

Participants
Of 68 patients assessed for eligibility, 20 were excluded (Fig. 1).Forty-eight outpatients participated in the trial and were included in this study (Table 1).Inclusion criteria were outpatients with a schizophrenia spectrum disorders diagnosis (International Statistical Classification of Diseases (ICD)-10, F20-29) between 18 and 65 years.Exclusion criteria were admittance to acute psychiatric ward, presence of ≥1 contraindications to exercise testing and training (life-threatening or terminal medical conditions, pregnancy, <6 months postpartum) or inability to perform testing/training.Patients were recruited in local health centers through therapists or other municipal healthcare professionals or supportive housing services.Before inclusion and testing, written informed consents were obtained.Patients' study eligibility was examined by a physician, and an experienced psychiatrist assessed patient diagnoses and medication status via medical records before enrolment.

Interventions
Patients in the TG performed aerobic interval training (Brobakken et al., 2020) and leg press MST 2d/week for 1 year.All MST sessions were done with the same leg press apparatus as during testing.One MST session included 4 × 4 repetitions at ~90 % of 1RM with 3-4 min passive rest in-between sets.During the eccentric movement phase, the weight was lowered slowly and controlled down to a 90 • knee angle, with a short stop (<1 s) before the concentric phase applying maximal intended velocity and force production back to the starting position.If a patient was able to complete 5 repetitions during the last set, the load was increased by 5 kg the next training session to ensure prescribed intensity.Qualified and experienced personnel at the training clinic and/ or exercise physiologists supervised each session.
The CG got two introductory training sessions in accordance with the MST protocol under close supervision to learn the procedure, advice on the benefits of regular exercise and encouragement to self-administer training on their own.Patients were also asked to report their exercise routines in a journal throughout the 1-year intervention.

Maximal muscle strength and power
Lower extremity 1RM and power were measured in a 40 • incline leg press apparatus (Hammer Strength HSLLP, Schiller Park, IL, USA).Patients used 4-8 attempts, of these 2-4 > 80 % of 1RM with 3 min rest between sets, to ascertain 1RM.Load was increased by 5-20 kg with each successful attempt until failure, with 1RM attained from the highest load.Each attempt was performed as a MST lift detailed in Section 2.3.
Lower extremity power was measured after the 1RM test following 3-5 min of rest.Lifts during power measurements were conducted as in the 1RM testing, with maximal mobilization of force in the concentric movement.All patients were given 3 attempts interspersed with 3-minute rest periods, with the best attempt utilized to calculate power.Load was set at 70 % of baseline 1RM at all test points.The Muscle Lab System (Ergo test Technology, Langesund, Norway) measured the duration of the concentric movement.External force, work distance and lifting time were used in the following equation to calculate power (Nm⋅s − 1 or Watt) (Storen et al., 2008): for which force (N) = weight lifted (kg) × 9.81 m⋅s − 2 .

Allometric scaling
Recommended procedures for allometric scaling were used for 1RM and power assessments due to variation in body mass (m b ) within groups (Hoff et al., 2005;Crewther et al., 2009).As 1RM and power increase in accordance with muscle cross-sectional area (L 2 ) rather than linearly with body mass (L 3 ), strength and power results are also reported raised to the power of 0.67 (2/3).

Functional performance
Functional performance assessment started with 30sSTS (Tveter et al., 2014).The patients started seated, arms across the chest and feet at approximately shoulder width.Patients were instructed to perform the test without shoes and, for 30 s, complete as many full stands as possible with a chair 45 cm high placed by the wall.Total number of completed stands were documented.
Patients proceeded to the stair test, where they were instructed to walk/run up and down a two-flight staircase consisting of 18 steps as fast as possible three times (Tveter et al., 2014).The staircase included a 180 • angle landing requiring an extra step.The patients were required to use all steps and allowed to use the bannister only to regain balance.The result was attained from the amount of time used to complete the test.

Walking work efficiency
Walking efficiency was measured with a portable metabolic device (Metamax II, Cortex Biophysic Gmbh, Leipzig, Germany) while the patients were walking on a treadmill (PPS Med, Woodway Gmbh, Weil am Rhein, Germany).Patients walked for 5 min at 4.5 km⋅h − 1 at 4 % inclination, and pulmonary measurements of rate of oxygen consumption in L⋅min − 1 (V˙O 2 ) were collected every 10 s.If a patient was unable to complete the 5 min at this workload, treadmill velocity was reduced by 1 km⋅h − 1 .V˙O 2 measurements were averaged over the last minute and utilized for energy expenditure calculation with the following equation (Wang et al., 2017): × 100

Health-related quality of life
Health-related quality of life was assessed with the SF-36 questionnaire (Ware et al., 2000).The SF-36 consists of eight subcomponents relating to functional performance, role limitations due to physical health problems, bodily pain, general health, vitality (energy/fatigue), social functioning, mental health (well-being) and role limitations due to emotional difficulties.These are summarized into the two main physical and mental component scores.Results are subsequently transformed into a score from 0 to 100, ranging from the worst to best possible health-related quality of life.

Symptom severity
Two certified psychiatric nurses assessed symptom severity in the patients at baseline using the positive and negative syndrome scale (PANSS) (Kay et al., 1987).

Randomization procedure
Patients were randomized to either the TG or CG in a 1:1 ratio.This trial utilized a block randomization procedure, with 20 patients in the first, and four patients in all the following blocks.This ensured that of the first 20 patients, 10 patients would be allocated to the TG and CG respectively, two in each group for the next block, and so forth.The University Hospital Clinical Trial Unit provided a randomization computer program to blind the researchers from treatment allocation.

Statistical analyses
To obtain a power of 80 % at a significance level of 0.05, expecting a ⁓30 % improvement with a SD half that size, 11 patients were required in each group when allowing for a ⁓30 % drop-out rate (Heggelund et al., 2012;Firth et al., 2015).Descriptive statistics are reported as mean ± SD where relevant.We used linear mixed models with the outcome variables one at a time as dependent variable, time point and training group and their interaction as covariates, and patient as random effect, using restricted maximum likelihood (REML) estimation.We adjusted for baseline as recommended by Twisk et al. (2018).Normality of residuals was checked by visual inspection of QQ-plots.Ninety-five percent confidence intervals (CI) are reported where relevant.Twosided p-values below 0.05 are regarded to indicate statistical significance, due to multiple hypotheses p-values between 0.01 and 0.05 should be interpreted with caution.Pearson correlation analyses were performed to investigate potential associations between improvements in 1RM and power with improvements in the functional performance tests, illness duration and baseline scores on the positive and negative syndrome scale (PANSS).The software packages IBM SPSS 27 (Chicago, USA) and Stata 17 were used for statistical analyses and figures were made using SigmaPlot version 14 (Systat Software Inc., USA).

Results
Protocol adherence is presented in Fig. 1 and baseline characteristics in Table 1.

Functional performance
The TG improved 30sSTS-performance more than the CG at 1 year (p = .038)while there was a trend at 6 months (p = .068;Table 3).The TG improved 30sSTS-performance by 14 % at 3 months, 17 % at 6 months and 20 % at 1 year (all p < .001).The CG improved performance by 11 % at 6 months (p = .050)and tended to improve at 3 months (p = .053)and 1 year (p = .077).1-year improvements in 30sSTS test tended to associate with improvements in scaled 1RM (r = 0.33, p = .067)and scaled power (r = 0.33, p = .063)in all included patients.
Both the TG (11 %, p = .024)and CG (21 %, p = .005)improved walking efficiency from baseline to 1 year (Table 3).This change, however, not apparent as a between-group difference.There was no significant change in the stair test.The first two columns show descriptive data for the training group and control group with p-values for within group change from baseline from mixed models.The difference estimates are from linear mixed models.a p < .05,aa p < .01change from 3 months, b p < .05,bb p < .01change from 6 months.m b , body mass in kg.

36-Item Short Form Health Survey
No statistically significant change in either the physical or mental component score was observed in the patient groups (Table 4).

Adherence
Adherence is reported in Table 5.

Discussion
Patients with schizophrenia have reduced muscle strength, power, and impaired functional performance, exacerbating their attenuated physical and mental condition.Strength training can counteract the reduced muscle strength; however, it is uncertain how effective training programs may be in long-term clinical care.We therefore aimed to investigate how effective MST, previously applied in short-term training interventions with excellent outcomes, could be integrated in clinical care for outpatients with adherence support and in-clinic training supervision.Main findings were that 1) 1RM and muscle power increased more in the TG compared to controls, 2) 1RM and muscle power increase throughout the whole study period, 3) the TG increased functional performance in the force-demanding 30sSTS test, and 4) adherence to the protocol was good throughout the period with 15 out of 25 patients completing the in-clinic MST, with only 2 patients dropping out from the study between 3 months to 1 year.Together our results show that MST can be integrated in standard clinical care of outpatients with schizophrenia, restoring muscle strength and power, contributing to improved functional performance.

One year of MST, maximal muscle strength, and power
After 3 months of leg press MST, 1RM and power increased by 19 % and 11 %, respectively.This is in line with other studies investigating heavy strength training in psychiatric populations, typically lasting ≤3 months (Heggelund et al., 2012;Leone et al., 2015;Silva et al., 2015;Unhjem et al., 2016aUnhjem et al., , 2016b;;Nygard et al., 2018).To the best of our knowledge, the present study is the first with a substantially longer intervention period in patients with schizophrenia and, notably, 1RM and power continued to increase from 3 months to 1 year.In fact, the improvements in these variables after 1 year were about twofold (1RM: 40 %; power: 26 %) compared to the improvements after 3 months.This is encouraging considering the patients' potential for lifelong benefits from the exercise intervention.Interestingly, as the age-related decline of 1RM is ~10 % per decade in the general population, the TGs 40 % The first two columns show descriptive data for the training group and control group with p-values for within group change from baseline from mixed models.The difference estimates are from linear mixed models.30sSTS, 30-second sit-to-stand test.

Table 4
The 36-item Short Form Health Survey following 3 months and 1 year of maximal strength training in patients with schizophrenia spectrum disorders.The first two columns show descriptive data for the training group and control group with p-values for within group change from baseline from mixed models.The difference estimates are from linear mixed models.SF-36, 36-item Short Form Health Survey.Values are mean ± SD. n, number of training sessions.
M. Nygård et al. improvement in 1RM can be interpreted as negating 40 years of aging (Lindle et al., 1997;Unhjem et al., 2017).Arguably, it could be beneficial to include more strength training exercises for a broader involvement of muscle groups.However, for simplicity and feasibility only one exercise, leg press, was chosen.Since leg press involves the major muscle groups in the lower extremities, relevant for weight-bearing physical activity, and is relatively easy to perform, it has great relevance for functional performance during activities of daily living.The patients' increase in 1RM was strongly correlated (r = 0.79) with the increase in power.The close association, and large increase in power, may be explained by the type of strength training performed, MST.MST is performed at a training intensity of ~90 % of 1RM combined with maximal intended velocity in the concentric phase, and is tailored to target neural adaptations, facilitating fast muscle force development (Unhjem et al., 2016a(Unhjem et al., , 2016b;;Nygard et al., 2018).Thus, it is likely that the aimed "explosive" execution of the repetitions during MST contributed to the large improvement in power in the present study.Of notice is that the actual MST movement velocity is slow because of the heavy loads lifted, but it is previously shown that the intended velocity, rather than the actual velocity, is important for the training-induced improvement (Behm and Sale, 1993).Despite the high intensity applied in MST, it can be safely performed, even in frail patient populations (Mosti et al., 2011;Mosti et al., 2013;Cešeiko et al., 2020;Berg et al., 2021) because the potentially more harmful eccentric phase of movement (Tøien et al., 2018) is carried out in a slow, controlled, fashion.This is also supported by the data from this study as no injuries were reported throughout the 1-year period.Importantly, the traininginduced power improvements and probable neural adaptations may be particularly important for patients with schizophrenia as their antipsychotic treatment may reduce the ability for "explosive" movements.Indeed, a negative association between power and defined daily dose of medication has previously been reported in this patient population (Nygard et al., 2019;Nygård et al., 2021).Illness duration and symptom scores on the other hand do not seem to affect muscle strength improvements in this study and supports previous findings from our lab (Nygård et al., 2021).

Muscle strength: implications for functional performance and healthrelated quality of life
The 30sSTS performance improved following 1 year of MST.Again, the increase was larger after 1 year (20 %) compared with 3 months.The 30sSTS increase also exhibited a clear tendency to be associated with 1RM (p = .067)and power (p = .063).This is unsurprising given the similarities between the leg press exercise, carried out down to 90 • in the knee joint, and the relatively simple sit-to-stand movement.Yet, the 30sSTS performance after 1 year of training is still markedly reduced in the patients, corresponding to performance observed in healthy individuals in their seventh decade (Tveter et al., 2014).Walking is a more complex movement pattern than sit-to-stands, and is, in addition to muscle strength, likely also influenced by other factors such as reduced postural control (Marvel et al., 2004) and balance (Nygard et al., 2019;Tsuji et al., 2019) in patients with schizophrenia.Yet, the patients in the TG increased walking work efficiency from ~22 % to ~25 %.This was however not different from the CG who exhibited a similar improvement.As the end point in both groups is close to the walking work efficiency seen in healthy individuals (Wang et al., 2017;Nygard et al., 2019), it may be close to a ceiling, limiting further increases.The 30sSTS and walking work efficiency improvements seen also in the CG might indicate a certain learning effect.The only physical performance that remained unaltered in the current study, against our hypothesis, was the stair test.The large variance between individuals in stair climbing performance, may partly explain the results.Another explanation may be the specificity of the test.While sit-to-stand and walking are both frequently practiced throughout the day for most patients in our study, the three times walk/run up and down a two-flight staircase is likely the test that deviates most from the combination of strength training and daily practice.Applying appropriate tests of functional performance is somewhat challenging as the tests may not directly transfer to the relevant activities of daily living often performed.Nevertheless, taken together, our results suggest that patients certainly improved their functional performance after 1 year in the present study.
The patients scored lower on health-related quality of life compared to normative values from the general Norwegian population (Garratt and Stavem, 2017), and this did not change throughout the 1-year period.In the physical component score the patients scored 8-22 % lower than average, while for the mental component score values were 23-56 % lower, clearly indicating that the patients feel they struggle more with their mental health compared to their physical health.It is concerning that such large improvements and normalization of muscle strength and power is not detected in the physical component score, however, a possible explanation could be the low impairment in scores compared to normative values at baseline.Of notice, although the SF-36 questionnaire has been documented to provide good validity, and detect differences between groups such as patients with schizophrenia and general population samples, evidence is lacking regarding its responsiveness, i.e., the ability to detect clinically important changes over time (Papaioannou et al., 2011).Yet, the observations in the present study indicate that high intensity strength training is feasible without mentally harmful side-effects.

Adherence, feasibility, and clinical implications
Our data show a drop-out rate of 40 % after 1 year of exercise, with 8 out of 10 patients dropping out during the first three months.Notably, about the same number of patients dropped out due to lack of motivation in each group, 6 in the TG and 5 in the CG, indicating that the training intervention was not necessarily the cause of drop-out.This result is encouraging given the apparent motivational difficulties the patients often suffer from (Vancampfort et al., 2012).Four additional patients dropped out of the TG due to other reasons than motivation: Two were lost to follow-up due to issues related to substance use disorders (overdose and psychiatric hospitalisation, respectively) preexisting at the time of inclusion; one due to somatic hospitalisation unrelated to the musculoskeletal or cardiorespiratory systems; one due to re-admittance to a psychiatric ward shortly following discharge.It is possible that the additional attrition from the TG, particularly due to hospitalisation, may be related to the more frequent contact with healthcare services, as patients in psychiatry are often underserved in terms of monitoring and treatment of various conditions (Nasrallah et al., 2006), or it simply may be due to random variation.Although we cannot fully discount the possibility, it appears unlikely that the additional patients lost to follow-up from the TG were caused by issues related to the training intervention applied in the current study.The low drop-out rate, combined with the continuous improvement throughout the study, and no reports of injuries strengthens the confidence with which we can state that implementing exercise in clinical treatment is feasible at a relatively low cost.Conversely, the CG is a timely reminder of the consequences when patients lack continuous support.
Importantly, strength training should be combined with (aerobic) endurance training targeting the maximal oxygen uptake and cardiovascular health of the patients (Brobakken et al., 2019a(Brobakken et al., , 2019b;;Brobakken et al., 2020).Clinical practice often appears random, and exercise may fall into a potpourri of physical activities.Careful administration and adjustments of strength and endurance training intensity, volume, and frequency is paramount to ensure efficacy of treatment.Especially the intensity of training is of fundamental importance for optimal effects (Helgerud et al., 2007;Heggelund et al., 2013).While patients in the current study trained at the clinic 2d/week, it may be possible that a higher frequency could be beneficial.Future investigations should examine how effective strength and endurance training could be combined with further sustainable lifestyle interventions affecting the poor diet and smoking habits in this patient group (Strassnig et al., 2003;Bobes et al., 2010) for additional improvement in physical health.Finally, of importance, long-term concurrent strength and endurance training should not only be limited to patients with schizophrenia, but also recommended for other populations within psychiatry (Flemmen and Wang, 2015;Silva et al., 2015;Mosti et al., 2016;Stubbs et al., 2018;Firth et al., 2020).

Conclusion
Strength training, integrated in long-term collaborative care of outpatients with schizophrenia, resulted in a continuous increase in 1RM and power which was accompanied by improved functional performance.In contrast, patients receiving an introduction to the same training, with encouragement to continue on their own, did not exhibit the beneficial outcomes.Given the feasibility, low-cost, and low risk of the model, combined with increased muscle strength and functional performance, our results suggest that integrating high intensity strength training in clinical care of patients with schizophrenia is highly beneficial.
CRediT authorship contribution statement J. Heggelund and G. Morken designed the study with assistance from E. Wang and J. Heggelund wrote the study protocol.M. Nygård, M. F. Brobakken, I. C. Güzey, J. Heggelund and E. Wang collected data.S. Lydersen and M. Nygård undertook the statistical analyses and interpreted the data.M Nygård, M. F. Brobakken, E. Wang and S. Lydersen prepared the manuscript.All authors contributed to and have approved the final manuscript.

Declaration of competing interest
The authors declare no conflict of interest for the publication of this study.

Fig. 2 .
Fig. 2. Change (Δ) in scaled leg press one repetition maximum (A) and power (B) following 3 months, 6 months and 1 year of maximal strength training in the training group and control group.Values are mean ± SEM. p, difference in change compared to control group.

Table 1
Patient characteristics at inclusion.
Values are mean and (SD) or number of patients.For psychiatric services values are median (range) due to non-normal distribution of variables.ICD, International Statistical Classification of Diseases; PANSS, positive and negative syndrome scale.

Table 2
Leg press one repetition maximum and power after 3 months, 6 months and 1 year of maximal strength training in patients with schizophrenia spectrum disorders.
Training group (TG) Control group (CG) Difference (group × time) TG and CG

Table 3
Functional performance following 3 months, 6 months and 1 year of maximal strength training in patients with schizophrenia spectrum disorders.

Table 5
Adherence to training sessions in patients with schizophrenia spectrum disorders in the collaborative care group.