A multilayered model of work outcome in patients with schizophrenia: Examination of the relationships among positive symptoms, cognition, and daily living skills Authors: Satsuki Ito, Junya Matsumoto, Chika Sumiyoshi, Yuka Yasuda, Michiko Fujimoto, Hidenaga Yamamori, Harumasa Takano, Tomiki Sumiyoshi, Ryota Hashimoto Published: 2026 ABSTRACT Patients with schizophrenia often show impairments in work functioning, yet the precise relationships among positive symptoms, cognition, daily living skills, and work outcome remain to be elucidated.We analyzed data from 396 patients with schizophrenia and 1109 healthy individuals. Work outcome was quantified as the mean weekly working hours. Cognitive decline was defined as the difference between current and premorbid IQ, while social function was evaluated by the Social Functioning Scale. We also conducted group comparisons and correlational analyses among the clinical variables. Using structural equation modeling (SEM), we further examined pathways from positive symptoms through cognitive decline, and determined whether daily living skills account for work outcome.Patients showed impairments in work outcome, cognitive decline, and social function, including daily living skills, as represented by Independence–Performance subscale of the Social Functioning Scale. Consistent with previous studies, cognitive decline and social function were correlated with weekly working hours among patients. SEM revealed several pathways from positive symptoms to working hours via daily living skills with or without a mediational effect of cognitive decline.The findings suggest that work outcome is influenced by multiple clinical symptom domains, including positive symptoms, cognitive decline, and daily living skills. The observed relationships among these features may aid in the development of effective interventions for occupational function in patients with schizophrenia. Introduction Schizophrenia is a psychiatric disorder characterized by positive symptoms, negative symptoms, and cognitive impairment. For the treatment of the disease, it is important to consider several factors, including functional outcomes, the development of diagnostic tools (Okazaki et al., 2023), and the evaluation of pharmacotherapy (Hasegawa et al., 2023). In terms of functional outcomes, social function, consisting of work functioning, activities, interpersonal relationships, and so forth, is disturbed in most patients with schizophrenia. Specifically, productive activities (hereafter work outcome), such as paid work, household labor, and study (Nuechterlein et al., 2011), are generally impaired. The loss of ability to work prevents their independent living and social inclusion (Harvey and Strassnig, 2012); therefore, work outcome should be considered the primary target of rehabilitation and recovery support. Vocational functioning has been a topic of extensive research on the prognosis of shizophrenia (Harvey and Strassnig, 2012; Kaneda et al., 2010; McGurk and Meltzer, 2000; Meltzer, 1999; Sumiyoshi et al., 2018, Sumiyoshi et al., 2025). To date, negative symptoms (Sumiyoshi et al., 2025; Bowie et al., 2006, Bowie et al., 2008; Fujino et al., 2016), cognitive impairment (Sumiyoshi et al., 2018; Bowie et al., 2008; Mausbach et al., 2010), and social function (Sumiyoshi et al., 2018, Sumiyoshi et al., 2025) have consistently been identified as key factors for work outcome. In these studies, cognitive impairment has been defined as the discrepancy between current IQ and estimated premorbid IQ (Fujino et al., 2017; Matsumoto et al., 2023; Yasuda et al., 2020; Weickert et al., 2000). Related to this, the severity of cognitive decline by indexing it according to the distribution of IQ in patients. Those studies reported that greater discrepancies of −10 (Weickert et al., 2000; Badcock et al., 2005; Kremen et al., 2008; Leeson et al., 2011) or − 15 (Fujino et al., 2017) were associated with severe cognitive impairment in this population. Despite these findings, further research is needed to clarify how multiple determinants of work outcome are related to one another. In previous studies, the measure of work outcome was categorical (e.g., work = 1, not in work = 0) (Nuechterlein et al., 2011; Kaneda et al., 2010). Although some studies have used quantitative measures, such as working hours (Sumiyoshi et al., 2018, Sumiyoshi et al., 2025; Mausbach et al., 2010; Ito et al., 2023; Bell et al., 2014) or wages (Llerena et al., 2018), further evidence using objective measures is needed. Second, daily living skills should be focused on in predicting work outcome, as it has a stronger link to work outcome than other social function domains in patients with schizophrenia (Sumiyoshi et al., 2025; Bowie et al., 2006, Bowie et al., 2008, Bowie et al., 2010). For example, the Independence–Performance subscale of the Social Functioning Scale (SFS) (Sumiyoshi and Sumiyoshi, 2011a; Birchwood et al., 1990; Nemoto et al., 2008), which reflects daily living skills, has been reported to effectively predict work outcome (Sumiyoshi et al., 2025). In fact, the Independence–Performance subdomain is included in the abbreviated version of the SFS (Pawliuk et al., 2022). Adverse effects of positive symptoms on functional outcomes have been underestimated despite their negative effects on the attainment and maintenance of employment (Ang et al., 2020; Wang et al., 2020; Harvey et al., 1999). To date, a longitudinal study (Harvey et al., 1999) reported that positive, rather than negative, symptoms were risk factors for functional decline in patients with schizophrenia. These findings suggest the need for further examination of the negative impact of positive symptoms on work functioning. How the above-noted factors contribute to work outcome and how they are related to each other should also be clarified. Although this information is needed to establish effective interventions for recovery of patients with schizophrenia, few studies have investigated the sequential and structural relationships of positive symptoms, cognitive decline, daily living skills, and work outcome. Clarifying these relationships is essential for developing systematic and evidence-based support strategies. The aims of this study were to examine the relationships among positive symptoms, cognitive decline, social function, and work outcome in patients with schizophrenia, with evaluation using a model that includes pathways originating from positive symptoms. First, associations among positive symptoms, cognitive function, social function, and work outcome (defined by weekly work hours) were examined by comparing them with those of healthy controls. The relationships among these factors and their influence on work outcome were determinedusing structural equation modeling (SEM). The primary goal was to examine the validity of pathways originating from positive symptoms to work outcome that are mediated by daily living skills (Independence–Performance) in patients with schizophrenia. Methods Subjects A total of 396 patients diagnosed with schizophrenia and 1109 healthy controls were recruited from two cohorts: Osaka University Hospital (Osaka cohort) and the National Center of Neurology and Psychiatry (NCNP cohort). The patient group included both inpatients and outpatients. Two trained psychiatrists applied the DSM-IV criteria and DSM-5 criteria to the Osaka cohort and the NCNP cohort, respectively. The study protocol was approved by the Institutional Review Board of the National Center of Neurology and Psychiatry (approval number: B2023–080). Written informed consent was obtained from all participants in accordance with the Declaration of Helsinki. Measures Working hours were assessed using the Social Activity Assessment (SAA) (Subotnik et al., 2008; Sumiyoshi and Sumiyoshi, 2011b), an interview that evaluates engagement in productive activities, such as paid employment; housework, including childcare and caregiving; and school-related tasks (e.g., class attendance and assignment work), over the past 12 weeks. The mean weekly work hours were calculated for each domain (paid employment, housework, and school-related tasks) using data from the past 12 weeks. The total weekly working hours were obtained by summing the three domain averages. In the present study, we included household chores or studies at school/home, into work, as they were thought to be productive activities. Although these activities are not competitive employment, they were considered to be equivalent to paid work if they were outsourced. Positive symptoms were assessed by the Positive and Negative Syndrome Scale (PANSS) (Kay et al., 1987). As noted in the Introduction, we focused on positive symptoms because they manifest overtly at the behavioral level and may directly affect work capacity. Therefore, we took the maximum-item approach to the positive-symptom factor, which is based on the idea that the most severe symptom has the greatest impact on behavior (similar to the concept of the Brief Evaluation of Psychosis Symptom Domains; BE-PSD) (Takeuchi et al., 2016). The highest score among the four symptoms (Wallwork et al., 2012), i.e., delusions, hallucinatory behavior, grandiosity, and unusual thought content, was chosen to represent the most behaviorally salient positive symptom at the time of assessment. Cognitive function was assessed using the Wechsler Adult Intelligence Scale as an index of current cognitive ability. The third edition (WAIS-III) (Wechsler, 1997) was used in the Osaka cohort, while the fourth edition (WAIS-IV) (Wechsler, 2008) was used in the NCNP cohort. Estimated premorbid IQ was assessed using the Japanese Adult Reading Test (JART) (Matsuoka et al., 2002). The degree of cognitive decline was operationalized as the difference between the FSIQ and estimated premorbid IQ (FSIQ − JART) (Fujino et al., 2017); thus, more negative values indicate a greater decline. Social function was assessed using the Social Functioning Scale (SFS) (Birchwood et al., 1990; Nemoto et al., 2008). The SFS consists of the following seven subscales: 1) Withdrawal, 2) Interpersonal Behavior, 3) Independence–Performance, 4) Independence–Competence, 5) Recreation, 6) Prosocial Activities, and 7) Employment/Occupation. In this study, the original SFS was modified to derive an adjusted SFS. Specifically, the Employment/Occupation subscale was excluded to avoid overlap between the predictor and outcome variables. In addition, the Independence–Competence domain was not included in the total score, as the contents are the same as those of the Independence–Performance. The perceived differences between competence (Independence–Competence) and performance (Independence–Performance) for daily living activities were minor. Additionally, most items in the Independence–Performance subscale evaluated everyday routines (e.g., taking care of personal appearance, washing one's own clothes, going food shopping, performing repairs, and cooking a meal) rather than financial or economic capacity to live independently. Statistical analysis Working hours were log-transformed for group comparisons and correlation analyses, as previous studies have shown that working hours in patients with schizophrenia are not normally distributed (Ito et al., 2023). Group comparisons of demographic and psychosocial variables were performed using t-tests. Effect sizes were calculated using Cohen's d. Thresholds were interpreted as small (≈0.20), medium (≈0.50), or large (≈0.80) (Cohen, 1988). To examine the extent to which the target variables (i.e., positive symptoms, cognitive decline, and daily living skills) were associated with working hours, correlation analyses were performed. The coefficients (Pearson's r) were interpreted as weak (|r| < 0.3), moderate (|r| < 0.5), or strong (|r| ≥ 0.5) (Cohen, 1988). In addition to the correlational analysis, SEM was conducted to evaluate the hypothesized hierarchical pathway to working hours, with positive symptoms, cognitive decline, and daily living skills (Independence–Performance) as observed variables. The model specified several routes starting from positive symptoms to work outcome via cognitive decline and daily living skills (Fig. 1). We report standardized regression coefficients (β), significance levels (p values), and variance explained (R2) for working hours. R2 was interpreted as weak explanatory power when it was less than 0.13, moderate when it was less than 0.26, and strong when it was equal to or greater than 0.26 (Cohen, 1988). A saturated model was constructed in which all possible paths were estimated freely to explore the relationships among the variables.Fig. 1The path diagram shows the direct and indirect effects of psychotic symptoms, cognitive impairment, and independence-performance on working hours in patients with schizophrenia. Arrows represent causal relationship among variables with the standardized path coefficients. Arrow widths are proportional to the absolute magnitude of the standardized coefficients (thicker arrows indicate stronger effects). Asterisks indicate statistical significance (*p < 0.05, **p < 0.01).Fig. 1 The path diagram shows the direct and indirect effects of psychotic symptoms, cognitive impairment, and independence-performance on working hours in patients with schizophrenia. Arrows represent causal relationship among variables with the standardized path coefficients. Arrow widths are proportional to the absolute magnitude of the standardized coefficients (thicker arrows indicate stronger effects). Asterisks indicate statistical significance (*p < 0.05, **p < 0.01). All the statistical analyses were performed using SPSS version 28.0 (IBM Corp., Armonk, NY, USA) and R version 4.2.2 (R Foundation for Statistical Computing, Vienna, Austria). SEM was performed with the R packages lavaan (Rosseel, 2012) and semPlot (Epskamp, 2022). Results Demographic and functional differences between patients and healthy controls The demographic and clinical characteristics of the patients and healthy controls are summarized in Table 1. Compared with healthy controls, patients with schizophrenia exhibited significantly shorter working hours, lower cognitive function, greater cognitive decline, and lower social function across all the subscales of the SFS. All group differences were statistically significant, except for age. Effect sizes showed particularly large differences in working hours (d = 2.0), cognitive function (d = 1.9), cognitive decline (d = 1.6), Adjusted SFS (d = 1.7), and Independence–Performance (d = 1.7), and Interpersonal Behavior (d = 1.6). Furthermore, large effect sizes were also observed for Prosocial Activities (d = 1.3) and JART scores (d = 1.1).Table 1Group comparison of demographic, social, and cognitive function between patients with schizophrenia and healthy controls.SchizophreniaHealthy controlsp valuedMeanSDMeanSDN (male/female)194/202572/537Age (years)36.113.135.715.10.57−0.031Years of education13.52.514.71.91.1 × 10−190.63Working hours (h per week)11.416.937.717.71.0 × 10–89a2.0aAge of onset (years)23.510.3Duration of illness (years)12.610.3Daily dose of CPZ eq (mg)551.8564.1PANSS total score89.119.2PANSS positive score21.05.3PANSS negative score22.25.4PANSS general score45.810.1Social Functioning Scale Adjusted SFS66.622.196.916.16.7 × 10−931.7 Withdrawal9.62.611.22.13.6 × 10−250.70 Interpersonal behavior6.63.010.21.93.2 × 10−741.6 Independence–performance23.38.433.14.64.1 × 10−751.7 Recreation17.76.922.46.64.0 × 10−300.72 Prosocial activities9.57.620.08.63.2 × 10−891.3WAIS-III/IV full-scale IQ84.317.6110.812.91.2 × 10−1051.9JART100.610.6109.36.97.5 × 10−431.1Cognitive declineb−16.612.91.511.25.4 × 10−941.6Abbreviations: SD, standard deviation; CPZ eq, Chlorpromazine equivalent; PANSS, Positive and Negative Syndrome Scale; SFS, Social Functioning Scale; WAIS-III/IV, Wechsler Adult Intelligence Scale – Third/Fourth Edition, JART; Japanese Adult Reading Test.aComparison was made after logarithmic transformation of work hours to meet normality assumptions.bCognitive decline was calculated as WAIS-III/IV full-scale IQ − JART; more negative values indicate greater decline. Table 1Group comparison of demographic, social, and cognitive function between patients with schizophrenia and healthy controls.SchizophreniaHealthy controlsp valuedMeanSDMeanSDN (male/female)194/202572/537Age (years)36.113.135.715.10.57−0.031Years of education13.52.514.71.91.1 × 10−190.63Working hours (h per week)11.416.937.717.71.0 × 10–89a2.0aAge of onset (years)23.510.3Duration of illness (years)12.610.3Daily dose of CPZ eq (mg)551.8564.1PANSS total score89.119.2PANSS positive score21.05.3PANSS negative score22.25.4PANSS general score45.810.1Social Functioning Scale Adjusted SFS66.622.196.916.16.7 × 10−931.7 Withdrawal9.62.611.22.13.6 × 10−250.70 Interpersonal behavior6.63.010.21.93.2 × 10−741.6 Independence–performance23.38.433.14.64.1 × 10−751.7 Recreation17.76.922.46.64.0 × 10−300.72 Prosocial activities9.57.620.08.63.2 × 10−891.3WAIS-III/IV full-scale IQ84.317.6110.812.91.2 × 10−1051.9JART100.610.6109.36.97.5 × 10−431.1Cognitive declineb−16.612.91.511.25.4 × 10−941.6Abbreviations: SD, standard deviation; CPZ eq, Chlorpromazine equivalent; PANSS, Positive and Negative Syndrome Scale; SFS, Social Functioning Scale; WAIS-III/IV, Wechsler Adult Intelligence Scale – Third/Fourth Edition, JART; Japanese Adult Reading Test.aComparison was made after logarithmic transformation of work hours to meet normality assumptions.bCognitive decline was calculated as WAIS-III/IV full-scale IQ − JART; more negative values indicate greater decline. Group comparison of demographic, social, and cognitive function between patients with schizophrenia and healthy controls. Abbreviations: SD, standard deviation; CPZ eq, Chlorpromazine equivalent; PANSS, Positive and Negative Syndrome Scale; SFS, Social Functioning Scale; WAIS-III/IV, Wechsler Adult Intelligence Scale – Third/Fourth Edition, JART; Japanese Adult Reading Test. Comparison was made after logarithmic transformation of work hours to meet normality assumptions. Cognitive decline was calculated as WAIS-III/IV full-scale IQ − JART; more negative values indicate greater decline. Correlations between working hours and psychosocial variables Working hours were significantly correlated with Adjusted SFS in patients with schizophrenia (r = 0.43, p < 4.9 × 10−19). Similarly, all subdomains showed moderate to weak correlations. Among the subdomains, the association was the strongest in the Independence–Performance subscale (r = 0.47, p < 1.8 × 10−23). Working hours were also significantly correlated with both cognitive function (r = 0.35, p < 8.3 × 10−13) and cognitive decline (r = 0.33, p < 3.3 × 10−11). In contrast, positive symptoms were significantly negatively correlated with working hours (r = −0.30, p < 2.0 × 10−9). Although some variables (Withdrawal, Independence–Performance, and cognitive function) exhibited significant associations with working hours in the healthy control group, they were relatively minor compared with those in the patient group (Table 2).Table 2Associations between working hours and social or cognitive functioning in patients with schizophrenia and healthy controls.SchizophreniaHealthy controlsrp valuerp valueSocial Functioning Scale Adjusted SFS0.434.9 × 10−190.071.0 × 10−2 Withdrawal0.293.3 × 10−90.105.8 × 10−4 Interpersonal behavior0.281.7 × 10−80.072.1 × 10−2 Independence–performance0.471.8 × 10−230.204.9 × 10−11 Recreation0.275.8 × 10−80.024.5 × 10−1 Prosocial activities0.261.0 × 10−7−0.025.7 × 10−1WAIS-III/IV full-scale IQ0.358.3 × 10−13−0.181.9 × 10−9JART0.178.9 × 10−4−0.113.5 × 10−4Cognitive declinea0.333.3 × 10−11−0.142.3 × 10−6Positive symptoms−0.302.0 × 10−9Correlations were computed with logarithmic transformed work hours. Raw p-values are shown. Bold and underlined values indicate correlations that remained statistically significant after Bonferroni correction (adjusted p < 0.05/19 = 0.0026).Abbreviations: SFS, Social Functioning Scale; WAIS-III/IV, Wechsler Adult Intelligence Scale – Third/Fourth Edition, JART: Japanese Adult Reading Test.aCognitive decline was calculated as WAIS-III/IV full-scale IQ − JART; more negative values indicate greater decline. Table 2Associations between working hours and social or cognitive functioning in patients with schizophrenia and healthy controls.SchizophreniaHealthy controlsrp valuerp valueSocial Functioning Scale Adjusted SFS0.434.9 × 10−190.071.0 × 10−2 Withdrawal0.293.3 × 10−90.105.8 × 10−4 Interpersonal behavior0.281.7 × 10−80.072.1 × 10−2 Independence–performance0.471.8 × 10−230.204.9 × 10−11 Recreation0.275.8 × 10−80.024.5 × 10−1 Prosocial activities0.261.0 × 10−7−0.025.7 × 10−1WAIS-III/IV full-scale IQ0.358.3 × 10−13−0.181.9 × 10−9JART0.178.9 × 10−4−0.113.5 × 10−4Cognitive declinea0.333.3 × 10−11−0.142.3 × 10−6Positive symptoms−0.302.0 × 10−9Correlations were computed with logarithmic transformed work hours. Raw p-values are shown. Bold and underlined values indicate correlations that remained statistically significant after Bonferroni correction (adjusted p < 0.05/19 = 0.0026).Abbreviations: SFS, Social Functioning Scale; WAIS-III/IV, Wechsler Adult Intelligence Scale – Third/Fourth Edition, JART: Japanese Adult Reading Test.aCognitive decline was calculated as WAIS-III/IV full-scale IQ − JART; more negative values indicate greater decline. Associations between working hours and social or cognitive functioning in patients with schizophrenia and healthy controls. Correlations were computed with logarithmic transformed work hours. Raw p-values are shown. Bold and underlined values indicate correlations that remained statistically significant after Bonferroni correction (adjusted p < 0.05/19 = 0.0026). Abbreviations: SFS, Social Functioning Scale; WAIS-III/IV, Wechsler Adult Intelligence Scale – Third/Fourth Edition, JART: Japanese Adult Reading Test. Cognitive decline was calculated as WAIS-III/IV full-scale IQ − JART; more negative values indicate greater decline. Hierarchical pathways among the psychosocial variables in SEM A path diagram based on the SEM results is shown in Fig. 1. The model explained 23% of the variance in working hours (R2 = 0.23). The direct effects of the predictors on work outcome were significant (Positive symptoms; β = −0.15; Cognitive decline; β = 0.20; and Independence–Performance; β = 0.30). The indirect effects of positive symptoms on work outcome were also significant; one route followed Independence–Performance, and the other two routes followed cognitive decline before reaching Independence–Performance (Fig. 1). Additional analyses were conducted to confirm the generalizability and to examine the effect of negative symptoms (the data are not shown). First, we constructed separate models for inpatients and outpatients. These models were highly similar, with slight differences in path coefficients. Second, a model using negative, instead of positive, symptom variables was constructed. As a result, R2 was slightly higher in the positive model, and the differences in the path coefficients were minor. Discussion This study demonstrated that occupational functioning is determined by a system that consists of positive symptoms, cognitive decline, and work outcome, in which daily living skills (Independence–Performance) interact hierarchically in patients with schizophrenia. In this model, positive symptoms were shown to affect work outcome both directly and indirectly through cognitive impairment and daily living skills. This hierarchical structure provides important insight into evaluation and interventions to improve functional prognosis in patients with schizophrenia. Compared with healthy controls, patients with schizophrenia had significantly shorter work hours and lower levels of cognitive and social functioning, consistent with the findings of previous studies (Sumiyoshi et al., 2018, Sumiyoshi et al., 2014; C. Sumiyoshi et al., 2016). Across domains, poorer work functioning was broadly associated with more pronounced psychosocial and cognitive difficulties. These patterns suggest that reduced work functioning in patients with schizophrenia arises from a combination of cognitive ability and social function. This relationship indicates the importance of considering multiple clinical dimensions when evaluating functional difficulties in patients with schizophrenia (Sumiyoshi et al., 2018, Sumiyoshi et al., 2025). Because we quantified work outcome as a continuous variable (i.e., weekly working hours), we were able to construct a path diagram that directly reflected work outcome. This approach allowed us to examine not only whether patients were working but also how much they were able to work, thereby providing more detailed information than a dichotomous measure such as employment status. In addition, the model was found and valid when the outcome domains were extended to include domestic and academic activities. These findings are consistent with observations that productive activities, irrespective of earning, alleviate psychiatric symptoms (Bell et al., 1996) and enhance self-esteem and efficacy (McGurk and Mueser, 2004) in patients with schizophrenia. Using the path diagram from SEM, we demonstrated a direct path from positive symptoms to work hours and three hierarchical indirect pathways that originate from positive symptoms (Fig. 1). Two of these indirect routes pass through daily living skills: in one route, positive symptoms influence work hours via daily living skills alone, whereas in the other, cognitive decline occurs before subsequent factors are reached. In either route, positive symptoms and daily living skills play important roles in predicting work outcome in patients with schizophrenia. In general, patients who experience severe hallucinations or delusions cannot engage in stable work; thus, reducing positive symptoms is a prerequisite for participation in productive roles. Our findings suggest that a model including pathways from positive symptoms through cognitive decline and daily living skills helps explain work outcome in patients with schizophrenia. Furthermore, the present approach used domains of positive symptoms that showed the highest scores, as the parameters. This method accords with the concept of the BE-PSD, which allowed us to effectively capture the early-stage influence of positive symptoms. Such effects might be obscured when positive symptoms are represented by summary scores. Moreover, some previous studies have used outpatients, who typically have relatively mild positive symptoms (Bowie et al., 2006, Bowie et al., 2008; Ang et al., 2020). This sampling characteristic may have influenced how the role of positive symptoms in functional outcomes has been interpreted, with greater emphasis placed on other symptom domains such as negative symptoms. In contrast, our study included patients with a broad symptom severity, allowing us to examine a pathway originating from positive symptoms. The wide-ranged sample in our study enabled us to show the effect of positive symptoms on work outcome. The structural pathway identified in our study has important implications for clinical practice. Improving positive symptoms is often a primary therapeutic target. Once positive symptoms are adequately controlled, improving psychosocial outcomes becomes more crucial. Our findings suggest that addressing these domains in an appropriate sequential order is critical for effective recovery. Specifically, considering the hierarchical relationships among symptoms, cognitive decline, and daily living skills may provide a useful framework for considering the timing and focus of interventions in the recovery process. This study has several limitations. First, we analyzed only working hours as an outcome measure; therefore, qualitative aspects such as employment type, job content, or work quality were not captured. The use of standardized assessment tools for social functioning, e.g., the Specific Level of Functioning Scale (T. Sumiyoshi et al., 2016), would have comprehensively captured the functionality of patients. Second, the data in this study were obtained cross-sectionally, precluding a thorough examination of the causal direction of the observed relationships. Third, environmental factors (e.g., family support, welfare services, and economic status) were not included in our model, which might have reduced the predictive power. Models that include extrinsic factors will help strengthen the understanding of real-world work functioning in patients with schizophrenia. In conclusion, this study demonstrated that work outcome is predicted by a multilayered model consisting of positive symptoms, cognitive decline, and daily living skills in patients with schizophrenia. Importantly, this hierarchical structure may originate from positive symptoms, which is mediated by daily living ability. These results provide insight into customizing vocational rehabilitation for patients with schizophrenia. CRediT authorship contribution statement Satsuki Ito: Writing – review & editing, Writing – original draft, Resources, Investigation, Formal analysis, Data curation, Conceptualization. Junya Matsumoto: Writing – review & editing, Resources, Methodology, Investigation, Data curation. Chika Sumiyoshi: Writing – review & editing, Writing – original draft, Formal analysis, Data curation, Conceptualization. Yuka Yasuda: Writing – review & editing, Validation, Resources, Methodology, Investigation, Data curation. Michiko Fujimoto: Writing – review & editing, Resources, Methodology, Investigation, Data curation. Hidenaga Yamamori: Writing – review & editing, Resources, Methodology, Investigation, Data curation. Harumasa Takano: Writing – review & editing, Resources, Methodology, Investigation, Data curation. Tomiki Sumiyoshi: Writing – review & editing, Validation, Conceptualization. Ryota Hashimoto: Writing – review & editing, Validation, Supervision, Resources, Methodology, Investigation, Data curation, Conceptualization. Funding This work was supported by the 10.13039/100009619Japan Agency for Medical Research and Development (AMED) under Grant Numbers JP18dm0307002, JP21dk0307103, JP21wm0425012, JP22tm0424222, JP24dk0307132, and JP25wm0625521; 10.13039/501100001691JSPS KAKENHI Grant Numbers JP19H05467: Grant-in-Aid for Specially Promoted Research, JP20H03611: Scientific Research B, JP20K07945, JP23K07001 and JP25K06796: Scientific Research C, JP23H00395: Scientific Research A, and JP25K17172: Early-Career Scientists; Intramural Research Grants for Neurological and Psychiatric Disorders of 10.13039/501100009438NCNP (Grants 3-1, 6-1, and 7-5); and the Osaka Psychiatric Clinics Association (Public Interest Incorporated Association). Declaration of competing interest The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Ryota Hashimoto reports financial support was provided by Japan Agency for Medical Research and Development. Junya Matsumoto reports financial support was provided by Japan Agency for Medical Research and Development. Ryota hashimoto reports financial support was provided by Japan Society for the Promotion of Science. Junya Matsumoto reports financial support was provided by Japan Society for the Promotion of Science. Satsuki Ito reports financial support was provided by Japan Society for the Promotion of Science. Chika Sumiyoshi reports financial support was provided by Japan Society for the Promotion of Science. Ryota Hashimoto reports financial support was provided by National Center of Neurology and Psychiatry. Junya Matsumoto reports financial support was provided by National Center of Neurology and Psychiatry. Tomiki Sumiyoshi reports financial support was provided by National Center of Neurology and Psychiatry. Yuka Yasuda reports financial support was provided by the Osaka Psychiatric Clinic Association (Public Interest Incorporated Association). Ryota Hashimoto reports a relationship with Takeda Pharmaceutical Company Limited that includes: speaking and lecture fees. Ryota Hashimoto reports a relationship with Viatris Inc. that includes: speaking and lecture fees. Ryota Hashimoto reports a relationship with Otsuka Pharmaceutical Co Ltd that includes: consulting or advisory, funding grants, and speaking and lecture fees. Ryota Hashimoto reports a relationship with EA Pharma Co Ltd that includes: consulting or advisory and speaking and lecture fees. Ryota Hashimoto reports a relationship with Eisai Co Ltd that includes: speaking and lecture fees. Ryota Hashimoto reports a relationship with Mochida Pharmaceutical Co Ltd that includes: speaking and lecture fees. Ryota Hashimoto reports a relationship with Sumitomo Pharma Co Ltd that includes: speaking and lecture fees. Ryota Hashimoto reports a relationship with Future Corporation that includes: speaking and lecture fees. Ryota Hashimoto reports a relationship with Ono Pharmaceutical Co Ltd that includes: consulting or advisory. Ryota Hashimoto reports a relationship with Nippon Boehringer Ingelheim Co Ltd that includes: consulting or advisory and funding grants. Tomiki Sumiyoshi reports a relationship with Nippon Boehringer Ingelheim Co Ltd that includes: consulting or advisory, funding grants, and speaking and lecture fees. Tomiki Sumiyoshi reports a relationship with Novartis Pharmaceuticals Corporation that includes: consulting or advisory. Tomiki Sumiyoshi reports a relationship with Takeda Pharmaceutical Company Limited that includes: speaking and lecture fees. Tomiki Sumiyoshi reports a relationship with Lundbeck that includes: speaking and lecture fees. Tomiki Sumiyoshi reports a relationship with Sumitomo Pharma Co Ltd that includes: speaking and lecture fees. Tomiki Sumiyoshi reports a relationship with Otsuka Pharmaceutical Co Ltd that includes: speaking and lecture fees. Hidenaga Yamamori reports a relationship with Eisai Co Ltd that includes: speaking and lecture fees. Hidenaga Yamamori reports a relationship with Otsuka Pharmaceutical Co Ltd that includes: speaking and lecture fees. Hidenaga Yamamori reports a relationship with Eli Lilly Japan KK that includes: speaking and lecture fees. Yuka Yasuda reports a relationship with Otsuka Pharmaceutical Co Ltd that includes: speaking and lecture fees. Yuka Yasuda reports a relationship with Viatris Inc. that includes: speaking and lecture fees. Yuka Yasuda reports a relationship with Towa Pharmaceutical Co Ltd that includes: speaking and lecture fees. Yuka Yasuda reports a relationship with Nobelpharma Co Ltd that includes: speaking and lecture fees. Yuka Yasuda reports a relationship with Sumitomo Pharma Co Ltd that includes: speaking and lecture fees. Junya Matsumoto reports a relationship with Sumitomo Pharma Co Ltd that includes: speaking and lecture fees. Harumasa Takano reports a relationship with Eisai Co Ltd that includes: funding grants. Harumasa Takano reports a relationship with Biogen Japan Ltd that includes: funding grants. Harumasa Takano reports a relationship with Sumitomo Pharma Co Ltd that includes: speaking and lecture fees. Harumasa Takano reports a relationship with Eli Lilly Japan KK that includes: speaking and lecture fees. Harumasa Takano reports a relationship with Otsuka Pharmaceutical Co Ltd that includes: speaking and lecture fees. Harumasa Takano reports a relationship with Kowa Company Ltd that includes: speaking and lecture fees. Chika Sumiyoshi and Tomiki Sumiyoshi are Editorial Board members of Schizophrenia Research: Cognition and co-authors of this article. Given their role as, had no involvement in the peer review of this article and had no access to information regarding its peer review. Full responsibility for the editorial process for this article was delegated to another journal editor. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.