Development of the Japanese version of the Manchester short assessment of quality of life (MANSA) among people with self-reported severe mental illness Authors: Masako Kageyama, Kayo Ichihashi, Momoko Kusaka, Keiko Yokoyama Published: 2026 ABSTRACT AbstractPurposeThe Manchester Short Assessment of Quality of Life (MANSA) is a quality-of-life (QOL) instrument developed for people with mental illness. This study aimed to develop a Japanese version of the MANSA for people with severe mental illness (SMI) and examine associations between individual MANSA domains and psychiatric diagnosis.MethodsThe MANSA includes 12 self-rated subjective QOL items; the overall score is their mean on a 1–7 scale. We prepared a Japanese translation and tested validity and reliability using cross-sectional survey data from 237 adults with SMI in Japan. Concurrent validity was tested in 38 respondents who also completed the WHOQOL26. The sample was randomly split for exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) in separate subsamples. Analyses included score distributions, validity, internal consistency, and domain-level ordinal logistic regression.ResultsThe MANSA items showed the expected associations with the relevant scales and items. Although a two-factor structure was supported by EFA, the CFA fit indices were CFI = 0.86, TLI = 0.82, and RMSEA = 0.12. In the one-factor EFA solution, all factor loadings exceeded 0.40. Cronbach’s alpha for all 12 items was 0.877. The schizophrenia-spectrum group had higher scores in several domains than the bipolar- and depressive-disorder groups, and symptom stability was associated with higher scores across all domains.ConclusionsThe Japanese version of MANSA showed acceptable concurrent validity and internal consistency. The proposed two-factor structure was not supported; total-score use remained reasonable, but further validation is needed. Diagnosis and symptom stability were associated with domain-specific QOL.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1007/s11136-026-04381-y.Keywords: Quality of life, Mental disorders, Scale and assessment, MANSA Introduction Quality of life (QOL) is an important outcome and needs indicator for people with mental illness [1]. The World Health Organization defines QOL as an individual’s perception of their position in life in the context of the culture and value systems in which they live and in relation to their goals, expectations, standards, and concerns [2]. A systematic review identified 44 QOL instruments for people with mental health problems: 16 general health-related measures, 9 mental health-specific measures, and 19 disease- or subgroup-specific measures [3]. General health-related QOL instruments were not designed specifically for mental illness, and their validity and sensitivity in this population have been questioned [3, 4]. In Japan, the WHOQOL26 [5] and SF-36 [6] are widely used, but no Japanese version of an international mental health-specific QOL instrument has been available. The Manchester Short Assessment of Quality of Life (MANSA), a mental health-specific instrument, has been used in more than 700 studies internationally [7]. A Japanese version could provide a measure better aligned with the experiences of people with mental illness, facilitate international comparisons, and support mental health research and practice in Japan. The MANSA was developed as a shortened version of the Lancashire Quality of Life Profile (LQLP) [8]. The LQLP is based on surveys of people with mental illness [9] and on a QOL model derived from interviews [10] developed by Lehman [8, 11]. This QOL model conceptualizes general well-being as a product of personal characteristics, objective life conditions in various domains, and satisfaction with those life conditions [10]. The validity and reliability of both the LQLP [10, 12] and the MANSA [7, 8, 13, 14] has been examined across a range of mental illnesses. Most QOL studies have emphasized overall or composite scores rather than individual QOL domains [15]. However, research focusing on QOL domains in schizophrenia [16, 17], bipolar disorder [18], and depression [19, 20] has reported that different sociodemographic and clinical factors are associated with each QOL domain. Furthermore, studies examining factors associated with QOL domains across diagnoses have shown that the associated factors differ depending on the diagnosis [15, 21]. Assessing QOL domains may help develop highly individualized rehabilitation programs that target the most affected areas, rather than relying on a standardized approach [16]. QOL domains that are poorly rated could be identified and targeted in treatment [15]. Domain-level analysis may identify diagnosis-related patterns and inform more targeted support. This study aimed to develop a Japanese version of the MANSA, evaluate its validity and reliability in people with self-reported SMI, and examine associations between MANSA domain scores and psychiatric diagnosis. Methods General description of the MANSA The MANSA comprises three sections. Personal characteristics are covered in Sects. “1” and “2”. Section “1” comprises four items, namely, date of birth, gender, ethnic origin, and diagnosis. Section “2” comprises nine items: education, employment status, occupation, number of working hours, income, social benefits, number of children, number of people living together, and type of residence. These items are descriptive and should be collected during the initial interview and updated when changes occur. Section “3” comprises 16 QOL items that can be answered in a multiple-choice format with the patient’s self-administered responses. Four of the 16 items are objective QOL items that are answered on a yes/no basis. These objective items assess the presence of a close friend, the number of contacts with friends per week, accusation of a crime, and victimization of physical violence. The remaining 12 items are subjective QOL items. These questions assess satisfaction with (1) life as a whole, (2) job/unemployment, (3) financial situation, (4) friendships, (5) leisure activities, (6) accommodation, (7) personal safety, (8) living situation, (9) sex life, (10) family, (11) physical health, and (12) mental health. Satisfaction is rated on a 7-point scale (1 = negative extreme, 7 = positive extreme), and the overall subjective QOL score is calculated as a mean value. Recent studies of the factor structure have shown that the 12 subjective QOL items comprise two factors: life and health-related aspects (items 1, 2, 3, 4, 5, 9, 11, and 12) and quality of environment (items 6, 7, 8, and 10) [7, 13]. The latter reflects QOL related to the living environment, although it may be interpreted somewhat differently from subjective QOL overall [7]. Development of the MANSA Japanese version The Japanese version was prepared with reference to the guidelines for the translation and cultural adaptation process [22] and the previous research on the development of a Japanese version of a psychological scale for patients with SMI [23]. (1) Preparation: We contacted the developer of the MANSA (SP) and obtained his permission to create a Japanese version of the scale. (2) Forward translation: Four mental health researchers (a psychiatrist and nurses) whose first language is Japanese translated the original text into Japanese. (3) Reconciliation: The four researchers discussed the wording until a consensus was reached, taking care not to undermine the original intent and ensuring that the wording reflected the Japanese system, situation, and culture. (4) Cognitive debriefing: We asked three people with SMI and one social worker about any difficulty they had in understanding the meanings of the scale items. (5) Review of cognitive debriefing results: The four researchers discussed the wording of the items and exchanged opinions until consensus was reached. They concluded that the ethnic origin in Sect. “1” was unfamiliar to Japanese respondents and difficult to understand; therefore, ‘(Japan, others)’ was included as an example. In Sect. “2”, state benefits and types of residence were considered difficult to imagine, so examples were provided. In Sect. “3”, the Japanese term for ‘friend’ was considered too narrow, so a similar Japanese word was added. We also considered adding online methods of meeting friends as a contemporary example. (6) Additional cognitive debriefing: We asked three other individuals with SMI to review the revised version to confirm that it was easy to understand. (7) Back translation: We commissioned a native speaker to perform the back translation. (8) Back-translation review: We asked SP to review the back-translation. SP responded that online services should not be included as a way of meeting friends because the coronavirus disease 2019 pandemic had subsided and the original version did not include this example. (9) Finalization: Based on SP’s advice, the four researchers agreed to keep the original wording without adding online services as a way of meeting friends, and the version was finalized with SP’s approval. Survey to assess validity and reliability To test the validity and reliability of the MANSA, we used data from an online cross-sectional survey of 300 adults (aged 18 years and older) undergoing treatment for mental illness in Japan [24]. The survey was conducted in the Japanese language and distributed to two regional peer groups and two national networks of people with mental illnesses to ensure broad coverage across Japan. Because this was a self-administered questionnaire survey, it did not capture symptom severity or cognitive ability. In this study, we analyzed data from 237 individuals with schizophrenia spectrum disorder (n = 88), bipolar disorder (n = 70), and depressive disorder (n = 79), all of which are included in the core category of SMI [25] and were represented in sufficient numbers for analyses. The sample size of 237 participants met the requirements for exploratory factor analysis (EFA) (4–20 times the number of items) [26]. Previous studies have examined the validity and reliability of the MANSA using the 12 subjective QOL items from the 16 items in Sect. “3” [7, 8, 13, 14]. All 16 items in Sect. “3” were included in the survey. The items used to test concurrent validity included the Japanese version of the Rosenberg Self-Esteem Scale [27, 28] (Cronbach’s alpha = 0.88), the WHOQOL26 (Cronbach’s alpha = 0.93), and two survey items: marital status and symptom stability. Because self-esteem is an indicator of recovery in people with mental illness [29], we hypothesized that respondents with higher MANSA scores would also have higher self-esteem. The WHOQOL26 was used to test the concurrent validity of nine MANSA items, but three MANSA items (8, 10, and 12) had no similar items in the WHOQOL26. Therefore, we used the two survey items for these domains. A previous literature review on marriage in SMI [30] reported positive aspects of marriage that may enhance life satisfaction among patients. Because a spouse is generally a family member chosen by the individual and often lives with that person, we hypothesized that married individuals would be more satisfied with their living situation (Item 8) and family (Item 10). Using self-reported items on psychiatric symptom stability, we hypothesized that people with more stable psychiatric symptoms would be more satisfied with their mental health (Item 12). Demographic data included items related to age, biological sex, residential style, sexual partners, self-reported main mental illness diagnosis, number of years since mental illness diagnosis, disability certification, social participation, and personal annual income. Concurrent validity was further examined in a subsample of 38 respondents who completed the WHOQOL26. Because a Japanese version of the LQLP has not been developed, we used the Japanese version of the WHOQOL26 [5]. Previous studies have reported correlation coefficients greater than 0.8 between the LQLP and MANSA [8] and greater than 0.40 between the MANSA and other validation measures [8, 14]. Under the administrator’s policy, WHOQOL26 items could not be embedded in the main online survey. All 300 survey respondents were invited on a first-come, first-served basis to complete a separate WHOQOL26 survey; 50 responded, of whom 38 met the SMI criteria for the present analysis. With an alpha error of 0.05 and a power of 0.80, and a recommended sample size of 44 calculated by G*Power, we set the target sample size at 50. Data analysis The validity and reliability of 12 of the 16 items in Sect. “3” of the MANSA were examined. First, the distribution of item scores was checked and floor and ceiling effects were assessed. To test concurrent validity, the association of the MANSA total score with the WHOQOL26 total score and the Rosenberg Self-Esteem total score was examined using Spearman correlation coefficients. Nine of the 12 MANSA items were examined for their association with similar items in the WHOQOL26. As a sub-analysis, we compared the demographic characteristics of 38 WHOQOL26 respondents with those of the 199 nonrespondents and calculated partial correlation coefficients using variables that showed significant differences as adjustment variables. The remaining two MANSA items (Item 8: living situation and Item 10: family) were examined for their association with marital status, and the remaining one MANSA item (Item 12: mental health) was examined for its association with symptom stability, using Wilcoxon rank-sum tests. Factor structure was examined using EFA and confirmatory factor analysis (CFA) in separate samples [31]. The sample was randomly split into two subsamples; EFA was conducted on Subsample 1 (n = 118), and CFA was conducted on Subsample 2 (n = 119). EFA was based on a polychoric correlation matrix with unweighted least-squares extraction and evaluated one- and two-factor solutions. CFA tested the previously proposed two-factor structure [7, 13], using diagonally weighted least squares (DWLS) with mean- and variance-adjusted test statistics (WLSMV). Model fit was evaluated using the comparative fit index (CFI; ideally > 0.95), Tucker–Lewis index (TLI; ideally > 0.95), and the root mean square error of approximation (RMSEA; ideally < 0.06) [32]. Internal consistency was evaluated using Cronbach’s alpha for all 12 items and for the two previously proposed subscales. MANSA domain scores and demographic characteristics were compared across diagnostic groups. Ordinal logistic regression was then performed separately for each domain, with diagnosis as the primary independent variable. All demographic variables were included as covariates except years since diagnosis, which was strongly correlated with age (r= 0.651). CFA was conducted in R version 4.6.1 (R Foundation for Statistical Computing, Vienna, Austria); all other analyses were conducted in SAS version 9.4 (SAS Institute, Cary, NC, USA). Results Respondent demographics Table 1 presents the demographic characteristics of the 237 respondents. All respondents were 20 years of age or older, and most were in their 30s (33.3%) or 40s (35.4%). The sample included 91 men (38.4%) and 146 women (61.6%). Regarding sexual partners, 87 (36.7%) had partner(s), 145 (61.2%) had no partner, and 5 (2.1%) selected ‘other.’ The mean duration since diagnosis of mental illness was 15.2 ± 11.1 years. A total of 162 respondents (68.4%) held disability certificates. A total of 180 individuals (76.0%) had an annual income of less than JPY 2 million or received welfare benefits. Of all respondents, 170 (71.7%) reported that their psychiatric symptoms were stable or rather stable. Table 1Respondent demographics and objective QOL itemsAll respondents (N=237)WHOQOL26 (n=38) p n % n %Mean ± SDMean ± SD Demographic data Age (years)41.4±9.644.3±8.70.037*20-292410.125.330-397933.3923.740-498435.41744.750-594318.1923.760-6673.012.6Sex (biological)Man9138.42052.60.049*Woman14661.61847.4Residential styleLive alone8134.21436.80.705Live with someone15665.82463.2Sexual partnersYes8736.71231.60.610†No14561.22565.8Other52.112.6Main diagnosis (self-reported)Schizophrenia spectrum disorder8837.11744.70.053Bipolar disorder7029.5513.2Depressive disorder7933.31642.1Years since diagnosis of mental illness (years)15.2±11.118.9±9.50.025*Disability certificateYes16268.43284.20.022*No7531.7615.8Social participationSpends most of the time at home8837.11129.00.349†Institutional day care for disabilities218.9615.8Job training3012.737.9Regular employment8335.01539.5School attendance31.300.0Others125.137.9Individual annual incomeLess than JPY 2 million or welfare benefits18076.03079.00.637JPY 2 million or more5724.1821.1 Items used for hypothesis testing Marital statusMarried (legal and de facto)8033.81129.00.494Unmarried15766.22771.1Stabilization of symptomsStable or rather stable17071.72976.30.493Not stable or rather unstable6728.3923.7 Objective QOL items Close friendsYes11749.41847.40.788No12050.62052.6Meet with friends within a weekYes9540.11847.40.317No14259.92052.6Being accused of a crime in the past yearYes20.812.60.189No23599.23797.4Victim of physical violence in the past yearYes62.500.00.278No23197.538100.0t-test, chi-square test, or Fisher’s exact test (†) were used, as appropriate*: p<0.05 Respondent demographics and objective QOL items t-test, chi-square test, or Fisher’s exact test (†) were used, as appropriate *: p<0.05 Among the four objective MANSA items, 117 respondents (49.4%) reported having close friend(s), 95 (40.1%) had met friend(s) within the past week, 2 (0.8%) reported having been accused of a crime during the past year, and 6 (2.5%) had been victims of physical violence in the past year. Table 1 also presents the demographics of the 38 respondents who completed the WHOQOL26. Compared with those who did not complete the WHOQOL26, this group was significantly older, more likely to be men, had a longer period since diagnosis, and had a significantly higher proportion of individuals with a disability certificate. Score distributions Table 2 provides the score distribution, with the mean scores ranging from 3.42 to 5.02. The lowest mean scores were as follows: financial situation (3.42), mental health (3.54), sex life (3.63), and physical health (3.78). Conversely, the highest mean scores were personal safety (5.02), living situation (4.70), and accommodation (4.47). None of the items exhibited floor or ceiling effects. Table 2Distribution of each item scoreNoMANSA itemsQuestionsResponse alternatives M SD M-SDM+SDN=2371234567Cronbach’s αn,%1Life as a wholeHow satisfied are you with your life as a whole today?9204066841444.071.262.815.333.88.416.927.935.45.91.72Job/unemploymentHow satisfied are you with your job (or sheltered employment, or training/education as your main occupation)? Or if unemployed or retired, how satisfied are you with being unemployed/retired?11304156722433.981.392.595.374.612.717.323.630.410.11.33Financial situationHow satisfied are you with your financial situation?22575040491363.421.531.894.959.324.121.116.920.75.52.54FriendshipsHow satisfied are you with the number and quality of your friendships?10242357823384.301.422.885.724.210.19.724.134.613.93.45Leisure activitiesHow satisfied are you with your leisure activities?6272658842884.281.362.925.642.511.411.024.535.411.83.46AccommodationHow satisfied are you with your accommodation?111929398144144.471.512.965.974.68.012.216.534.218.65.97Personal safetyHow satisfied are you with your personal safety?21016309564205.021.233.786.250.84.26.812.740.127.08.48Living situationHow satisfied are you with the people that you live with? Or if you live alone, how satisfied are you with living alone?101719368350224.701.523.196.224.27.28.015.235.021.19.39Sex lifeHow satisfied are you with your sex life?243323101411233.631.392.245.0210.113.99.742.617.35.11.310FamilyHow satisfied are you with your relationship with your family?211719627630124.241.552.695.798.97.28.026.232.112.75.111Physical healthHow satisfied are you with your health?10384861601913.781.342.445.124.216.020.325.725.38.00.412Mental healthHow satisfied are you with your mental health?14565639462243.541.492.055.045.923.623.616.519.49.31.7Factor1: Life and health-related aspects†3.880.990.859Factor2: Quality of environment‡4.611.140.790All items4.120.920.877†:Item number 1, 2, 3, 4, 5, 9, 11, 12, ‡: item number 6, 7, 8, 10M Mean, SD standard deviation. The correlation between Factor 1 and Factor 2 was r = 0.510 Distribution of each item score †:Item number 1, 2, 3, 4, 5, 9, 11, 12, ‡: item number 6, 7, 8, 10 M Mean, SD standard deviation. The correlation between Factor 1 and Factor 2 was r = 0.510 Concurrent validity The relationships between the MANSA total score and WHOQOL total score, or the Rosenberg Self-Esteem total score, and nine MANSA items with corresponding WHOQOL26 items are shown in Table 3. The Spearman’s correlation coefficients ranged from 0.339 to 0.815, indicating a moderate to relatively strong correlations. After adjustment for age (years since diagnosis was not included), sex, and disability certificate status, the partial correlation coefficients were 0.294 or higher, indicating moderate correlations. Table 3Concurrent validityNoMANSAMANSAWHOQOL26rs p Partial rs p MeanSDMeanSD WHOQOL26 item (n = 38) 1Life as a wholeOverall quality of life4.161.603.211.070.516< 0.0010.549< 0.0012Job/unemploymentWorking capacity4.181.432.681.020.560< 0.0010.561< 0.0013Financial situationFinancial resources3.241.632.550.980.619< 0.0010.637< 0.0014FriendshipsSocial support4.581.603.550.920.685< 0.0010.678< 0.0015Leisure activitiesParticipation in and opportunities for recreation and leisure4.131.603.160.950.3390.0370.2940.0876AccommodationHome environment4.131.863.111.230.815< 0.0010.831< 0.0017Personal safetyPhysical safety and security5.031.383.531.060.532< 0.0010.5300.0018Living situationn/a4.031.73n/an/an/an/an/an/a9Sex lifeSexual activity3.131.632.421.150.808< 0.0010.862< 0.00110Familyn/a4.001.82n/an/an/an/an/an/a11Physical healthOverall health3.741.452.871.090.4240.0080.5090.00212Mental healthn/a3.971.68n/an/an/an/an/an/aAll 12 itemsWHOQOL26 All items (n=38)0.757< 0.0010.793< 0.001All 12 itemsRosenberg’s self-esteem scale (N=237)0.509< 0.0010.522< 0.001rs Spearman’s rank correlation coefficient. Partial rs rs controlling for age, sex, and disability certificate Concurrent validity rs Spearman’s rank correlation coefficient. Partial rs rs controlling for age, sex, and disability certificate The results of the Wilcoxon rank-sum test for the three MANSA items (Item 8, 10, and 12) and two survey variables (marital status and symptom stability) are presented in Table 4. Significant associations were observed for all items, as hypothesized. Table 4Concurrent validityN = 237MANSA items (number)Living situation (8)Family (10)Mental health (12)Demographic data n Mean SD p Mean SD p Mean SD p Marital statusIn marriage805.161.28< 0.0014.781.32< 0.001Not married1574.471.583.961.59Stabilization of symptomsStable or rather stable1703.951.40< 0.001Not stable or rather unstable672.521.21p Wilcoxon rank-sum test Concurrent validity p Wilcoxon rank-sum test Factor structure In the randomly split sample, the EFA was conducted in Subsample 1 (n = 118), and the CFA was conducted in Subsample 2 (n = 119) (Table 5). EFA identified two factors with eigenvalues greater than 1. The two-factor EFA yielded a structure similar to that reported previously (Table 5). In the one-factor solution, all item loadings exceeded 0.40. The two-factor CFA showed inadequate fit (CFI = 0.86, TLI = 0.82, RMSEA = 0.12; Fig. 1). Table 5Factor components and item loadingsNoMANSA itemsSubsample1 (n=118)Two factorsOne factorFactor1life and health-related aspectsFactor2quality of environment2Job/unemployment 0.803 − 0.083 0.680 12Mental health 0.835 − 0.036 0.744 3Financial situation 0.756 0.090 0.776 1Life as a whole 0.743 0.246 0.887 11Physical health 0.718 − 0.097 0.593 5Leisure activities 0.641 0.098 0.676 4Friendships 0.435 0.333 0.660 9Sex life 0.412 0.124 0.481 8Living situation− 0.183 0.992 0.505 6Accommodation0.122 0.717 0.622 10Family0.078 0.569 0.483 7Personal safety0.144 0.551 0.534 Using polychoric correlation matrix using the unweighted least squares (ULS) extraction methodBolded numbers are factor loadings of 0.4 or greaterCFA in Subsample 2 (n=119) : CFI = 0.86, TLI = 0.82, and RMSEA = 0.12 Factor components and item loadings Using polychoric correlation matrix using the unweighted least squares (ULS) extraction method Bolded numbers are factor loadings of 0.4 or greater CFA in Subsample 2 (n=119) : CFI = 0.86, TLI = 0.82, and RMSEA = 0.12 Standardized parameters of the MANSA item loadings to two factors The model was estimated using the WLSMV estimator The Cronbach’s alpha coefficients were 0.877 for all items, 0.859 for Factor 1, and 0.790 for Factor 2, indicating good internal consistency (Table 2). MANSA domains and diagnosis Supplement 1 presents domain scores by diagnosis, and Supplement 2 compares demographic characteristics across diagnostic groups. Median MANSA domain scores in respondents with schizophrenia-spectrum disorders were equal to or higher than those in the other two diagnostic groups across all domains. Respondents with schizophrenia-spectrum disorders were also older and had a longer time since diagnosis than the other groups. In adjusted ordinal logistic regression models (Table 6), the schizophrenia-spectrum group had higher scores in several domains than the bipolar- and depressive-disorder groups, and no MANSA domain was significantly lower in the schizophrenia-spectrum group. Women had higher scores for life as a whole and accommodation. Respondents living with others had higher family scores, and those with a sexual partner had higher friendship and sex-life scores. Disability-certificate holders had higher scores for life as a whole and job/unemployment. Respondents with an annual income of JPY 2 million or more had higher job/unemployment and financial-situation scores. Married respondents had higher accommodation, living-situation, and family scores. Stable or rather stable symptoms were associated with higher scores across all 12 domains. Table 6Association between MANSA domains and diagnosis, adjusted for covariates123456789101112Life as a wholeJob/unemploymentFinancial situationFriendshipsLeisure activitiesAccommodationPersonal safetyLiving situationSex lifeFamilyPhysical healthMental healthOR95%CIOR95%CIOR95%CIOR95%CIOR95%CIOR95%CIOR95%CIOR95%CIOR95%CIOR95%CIOR95%CIOR95%CIMain diagnosisSchizophrenia spectrum disorderReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceBipolar disorder0.550.301.020.650.361.19 0.53 0.29 0.97 0.930.511.690.700.391.28 0.45 0.24 0.81 0.43 0.23 0.79 0.800.441.450.660.361.200.780.431.410.680.381.230.640.351.16Depressive disorder 0.42 0.22 0.80 0.630.341.180.690.371.280.650.351.21 0.41 0.22 0.78 0.38 0.20 0.72 0.41 0.21 0.77 0.580.311.090.680.361.28 0.44 0.23 0.82 0.730.401.370.540.291.01Age0.990.961.021.010.991.040.990.961.021.010.991.040.990.971.021.000.971.031.000.981.030.990.971.020.980.961.010.970.940.990.980.961.011.010.991.04Sex (biological)ManReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceWoman 1.78 1.06 3.00 1.500.912.501.540.932.541.620.972.681.310.792.18 1.69 1.02 2.80 1.180.701.961.340.812.221.410.852.360.780.471.290.720.441.190.970.591.61Residential styleLive aloneReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceLive with someone1.480.842.601.190.692.071.390.802.401.160.672.011.450.842.531.110.641.930.720.411.270.740.431.281.180.672.06 2.15 1.23 3.74 0.980.571.690.910.521.57Sexual partnersNoReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceYes0.940.442.021.670.793.540.990.472.07 2.21 1.04 4.69 0.890.421.880.770.371.621.130.532.410.870.411.83 3.77 1.74 8.17 0.600.291.271.520.733.201.450.693.06Disability certificateNoReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceYes 1.98 1.03 3.81 1.91 1.01 3.62 0.820.441.540.910.481.710.700.371.320.730.391.381.020.531.940.890.471.680.820.431.571.080.572.040.990.531.871.260.672.38Individual annual incomeLess than JPY 2 million or welfare benefitsReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceJPY 2 million or more1.430.772.64 2.08 1.13 3.82 2.78 1.52 5.09 1.150.632.090.910.501.661.590.872.901.060.581.951.270.702.301.650.893.041.460.802.660.940.521.701.150.632.09Marital statusUnmarriedReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceMarried (legal and de facto)1.530.663.550.760.341.721.410.633.160.580.261.320.950.422.14 2.58 1.14 5.84 2.291.005.28 2.76 1.21 6.26 0.630.281.45 3.03 1.33 6.93 0.990.442.210.840.371.90SymptomsNot stable or rather unstableReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceReferenceStable or rather stable 8.55 4.71 15.52 5.62 3.19 9.91 3.15 1.83 5.43 3.01 1.75 5.18 3.50 2.02 6.06 1.73 1.02 2.94 2.37 1.37 4.11 1.98 1.16 3.38 1.89 1.10 3.25 2.00 1.17 3.43 2.87 1.67 4.93 7.66 4.24 13.85 Ordinal logistic regression analysisBold indicates statistically significant differences Association between MANSA domains and diagnosis, adjusted for covariates Ordinal logistic regression analysis Bold indicates statistically significant differences Discussion We developed a Japanese version of the MANSA and evaluated its psychometric properties in adults with self-reported SMI. Concurrent validity and internal consistency were acceptable. The findings did not support the proposed two-factor structure and were compatible with use of the overall MANSA total score, although the factor structure requires further validation. Domain-level analyses also identified associations with diagnosis and symptom stability. Respondent demographics Most respondents reported stable or rather stable symptoms (71.7%), held a disability certificate (68.4%), and had an annual income below JPY 2 million or received welfare benefits (76.0%). These characteristics indicate substantial and persistent functional impairments, consistent with a SMI sample [33]. Because the WHOQOL26 subsample differed from nonrespondents in age, sex, time since diagnosis, and disability-certificate status, the concurrent-validity findings should be interpreted with caution despite statistical adjustment. Score distribution The distribution of the scores for the 12 subjective MANSA items showed no floor or ceiling effects. The lowest mean scores were observed for financial situation (3.42), mental health (3.54), sex life (3.63), and physical health (3.78). The low satisfaction with mental and physical health may reflect the fact that the respondents were undergoing treatment for mental illness. Satisfaction with financial situation was the lowest. In this study, 76.0% of respondents had an annual income of less than JPY 2 million or received welfare benefits. The average annual salary income in Japan is JPY 4.58 million and 20.5% of the population earns less than JPY 2 million [34]. According to a nationwide survey in Japan, 73.9% of individuals aged 19–64 years with a mental disability certificate had an annual income of less than JPY 1.8 million [35]. This suggests that one reason for the low level of satisfaction with financial situation is that people with SMI typically have annual incomes considerably lower than those of the general population in Japan. Satisfaction with sex life was also low. In this study, 61.2% of the respondents did not have a sexual partner, and many may therefore have had limited opportunities for sexual relationships. The low mean scores for financial situation [7, 15, 20, 36] and sex life [7, 15, 36, 37] are consistent with previous studies of SMI. Both domains have also been reported to score lower than other WHOQOL26 items in the general Japanese population [5]. Low satisfaction with financial situation and sex life may therefore reflect both the effects of mental illness and broader characteristics of the Japanese context. The highest mean scores were observed for personal safety (5.02), living situation (4.70), and accommodation (4.47). These three items were also rated highly in previous studies conducted in other countries [7, 20, 36, 37]. However, personal safety had the highest score in our survey. One possible reason is that crime rates are relatively low and personal safety is generally high in Japan [38]. Validity and reliability All items used to assess concurrent validity showed significant or moderate associations, as hypothesized, even after adjustment for variables that differed significantly between the main sample (N = 237) and the subsample of WHOQOL26 (n = 38). Although EFA supported a two-factor structure, the proposed two-factor CFA yielded CFI = 0.86, TLI = 0.82, and RMSEA = 0.12, indicating a weaker fit than that reported in previous studies (CFI = 0.90, RMSEA = 0.067 [7] and CFI = 0.96–0.98, RMSEA = 0.07–0.08 [13]). By contrast, all loadings in the one-factor EFA exceeded 0.40. Taken together, these findings did not support the proposed two-factor structure and were compatible with use of the overall MANSA total score; however, the factor structure requires further validation. MANSA domains and diagnosis In this study, even after adjusting for covariates, MANSA scores for schizophrenia-spectrum group were higher than or comparable to those for bipolar- and depressive-disorder groups across all domains, broadly consistent with previous findings [15, 21]. Depressive symptoms may have a particularly negative impact on QOL [39], which could partly explain lower scores in the mood-disorder groups. Unlike a previous diagnosis-based meta-analysis of MANSA domains [15], this study found no diagnostic-group differences in sex life or living situation. Difficulties with intimate relationships and living arrangements have also been described in bipolar disorder [40], and across SMI more broadly [41], suggesting these challenges are not unique to schizophrenia spectrum disorders. Symptom stability was associated with higher scores in every MANSA domain, consistent with previous research [15]. Although causality cannot be inferred from this cross-sectional study, the finding highlights the close relationship between symptom status and multiple aspects of QOL. Married respondents reported higher accommodation, living situation, and family scores. These domains are related to the proposed quality-of-environment factor. A review article on marriage among individuals with SMI indicates that marriage serves as a protective factor in enhancing social support and social functioning [30]. Because this study was cross-sectional, the direction and mechanisms of these associations cannot be determined. Holding a disability certificate was associated with higher job/unemployment satisfaction. In Japan, companies are required by law to employ people with disabilities [42]. These disability employment policies may increase access to supported or designated employment opportunities; however, the mechanism underlying this association could not be determined from the present data. Women reported higher scores for life as a whole and accommodation than men. However, the present data do not explain these differences, and they should not be interpreted causally. Replication and further investigation of relevant social, clinical, and environmental factors are required. Limitations This study has several limitations. First, diagnoses were self-reported rather than confirmed by psychiatrists. Therefore, it is possible that some respondents who did not meet the diagnostic criteria were included. However, because the respondents were currently receiving psychiatric treatment, it is likely that they were aware of their diagnosis. Second, because the online survey was self-administered, symptom severity and cognitive ability could not be assessed. Third, the sample size was limited for subgroup analyses, split-sample factor analyses, and adjusted correlations, reducing statistical precision. Fourth, the WHOQOL26 used to assess concurrent validity is not an instrument developed specifically for mental illness. However, no Japanese version of a QOL instrument specific to mental illness was available. Finally, reliability was assessed only in a cross-sectional study and could not be verified using a test-retest method. Further validation is needed in larger, clinically characterized samples. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 2 Acknowledgements We thank all participants. We thank MANSA developer Dr Stefan Priebe for his honest and precise support. We would also like to thank Mr. Sou Kuroda for his advice regarding statistics. Author contributions Design of the study, translation and manuscript preparation: MKa, KI, MKu, and KY; data management and analysis: MKa. All the authors have reviewed and confirmed the final manuscript prior to submission. Funding Open Access funding provided by The University of Osaka. This study was supported by the Japan Society for the Promotion of Science (JSPS) KAKENHI (grant number JP 23H03226, JP 23K27916). Data availability The Japanese version of the Manchester Short Assessment of Quality of Life (MANSA) is available from the corresponding author. The datasets generated and analyzed during the current study are not publicly available but are available from the corresponding author upon reasonable request. Declarations Conflict of interest The authors declare no competing interests. Ethical approval This study was approved by the Ethical Review Board Osaka University Hospital (approval no. 23362; 11 January 2024). Explanations were displayed online regarding the purpose and outline of the study; freedom to participate, refuse, or discontinue the survey, and guarantees of anonymity. Consent to participate Informed consent was obtained from all individual respondents by ticking the confirmation box. Consent for publication Not applicable. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1 Supplementary Material 2 Data Availability Statement The Japanese version of the Manchester Short Assessment of Quality of Life (MANSA) is available from the corresponding author. The datasets generated and analyzed during the current study are not publicly available but are available from the corresponding author upon reasonable request.