A pilot study of body–mind–spirit group therapy on sleep and quality of life in individuals with visual impairment Authors: Chuen-Ru Liu, Mei-Yun Tseng, Dipanshu Sharma, Wen-De Hsieh, Wei-Shu Liu, Chien-Huey Sophie Chang, Jwo-Huei Jou Published: 2026 ABSTRACT PurposeThis study evaluated the feasibility and preliminary effects of body–mind–spirit (BMS) group therapy—integrating Eastern gratitude philosophies with Western psychotherapy—on sleep and quality of life (QoL) among adults with visual impairment (VI).MethodsA quasi-experimental study recruited 30 adults with VI (mean age 61.5 ± 11.4 years, range 26–78; 53.3% female). The experimental group received 10 weekly 130-min BMS sessions; the control group received usual care. QoL was assessed using the WHOQOL-BREF, and sleep was measured through accelerometry (XA-5) and self-reports.ResultsCompared with the comparison group, the intervention group showed significant improvements in overall QoL (p = 0.01, d = 0.58), with the most robust effect observed in physical health (d = 0.71). While objective sleep metrics remained statistically unchanged, the experimental group reported moderate improvements in sleep latency (d = −0.32; 23% reduction) and increased satisfaction. Data stability was influenced by a perceived lack of security in wearable devices. Notably, longer sleep duration correlated with better psychological well-being (p = 0.01), while lower gains among younger participants were linked to significantly shorter baseline sleep duration associated with their occupations.ConclusionBMS therapy is feasible and may enhance QoL and subjective sleep through improved emotional regulation. Although “circadian hardwiring” may limit physiological reset, BMS offers a psychological buffer against sleep dissatisfaction. Future trials should integrate circadian-focused strategies and address cultural conservatism regarding sensitive QoL dimensions.Trial registrationIRB of Taipei City Hospital (TCHIRB-11107010); ClinicalTrials.gov (NCT07105761); Registration date: 13 August 2025.Supplementary InformationThe online version contains supplementary material available at 10.1007/s11136-026-04260-6. Background Visual impairment (VI) affects about 2.2 billion people globally and is expected to rise with aging populations and lifestyle changes like more screen time, less outdoor time, and prolonged blue light exposure [1–3]. In Taiwan, individuals with VI rose 40% from 38,747 in 2000 to 54,317 in 2022, highlighting a growing public health burden [1]. Beyond visual loss, VI entails physical, psychological, and social challenges that greatly reduce quality of life (QoL) [1, 4–6]. Individuals with VI report markedly lower QoL than healthy populations and those with other chronic conditions, with declines linked to reduced daily functioning, social isolation, mental health problems, and increased mortality [5, 7–9]. VI is linked to depression, anxiety, cognitive decline, and higher dementia risk [1, 10–12]. Poor sleep quality, circadian dysregulation, and insomnia are common and may worsen fatigue, emotional dysregulation, and impaired psychosocial functioning [13–17]. Current rehabilitation programs focus on functional independence like orientation and mobility training but often neglect the complex psychological needs and emotional distress of vision loss [12, 13]. Body–Mind–Spirit (BMS) group therapy is a holistic intervention combining Eastern philosophies and Western psychotherapy [18]. A recent review suggests 8–12 sessions of 120–180 min each are needed for deep holistic well-being improvements [18]. This 10-week, 130-min session framework was adapted to balance intensive therapy with the physical endurance and deep oral sharing needs of participants with VI, aiming to reduce anxiety and depression and improve QoL and sleep quality [19, 20]. Implementing psychosocial interventions for individuals with VI presents logistical and methodological challenges, including recruitment difficulties and transportation barriers. Empirical evidence on the feasibility of holistic group therapies in community settings remains limited. To address this gap, the study was a quasi-experimental pilot trial. The primary aim was to evaluate the feasibility and acceptability of a 10-week BMS program for adults with VI; secondarily, it explored preliminary QoL and sleep trends to inform future large-scale trials. Methods Study design This study employed a single-blind, quasi-experimental design with matched groups to evaluate the feasibility and preliminary outcomes of the BMS intervention. Recruitment occurred between September 2022 and June 2024 through social media and community organizations. Owing to logistical constraints and the specific mobility needs of individuals with VI in a community setting, participants were allocated to either the experimental group (10-week BMS therapy) or the control group (usual care) based on registration order and scheduling availability rather than formal randomization. To minimize selection bias, baseline assessments occurred before group allocation, and data analysts were single blinded to group assignments. Outcomes—including QoL (WHOQOL-BREF TW) and objective sleep quality measured by the XA-5 accelerometer—were assessed at baseline (Week 1) and post-intervention (Week 10). The study protocol was approved by the Institutional Review Board of Taipei City Hospital (TCHIRB-11107010), registered at ClinicalTrials.gov (NCT07105761), and conducted in accordance with the Declaration of Helsinki. To ensure accessibility for participants with VI, a researcher read the informed consent form aloud privately, allowing ample time for questions. Written informed consent was obtained from all participants and their legal representatives, using signature guides if needed, in the presence of an independent witness. To maintain methodological rigor despite the pilot design, recruitment and data collection were supervised by a doctoral-level psychiatric nurse using standardized protocols and weekly research meetings to monitor fidelity and address operational issues. Participants Participants were recruited from community organizations and social media. Inclusion criteria were: (1) possession of an official disability certificate indicating VI; (2) age between 26 and 73 years; (3) ability to communicate verbally; and (4) willingness to participate in group discussions and provide consent for audio recording. Exclusion criteria included hearing or speech impairments preventing verbal communication, concurrent participation in other psychological or group therapies, or cognitive impairment (Mini-Mental State Examination score < 3, a threshold adjusted for VI to ensure basic verbal communication) that limited the ability to express personal intentions. The target sample size was determined to be 44 participants (22 per group) using G*Power 3.1, with α set at 0.05, power at β = 0.2, and an effect size of f = 0.38, consistent with current systematic reviews of BMS interventions [18]. Of the 50 individuals invited, 36 met the eligibility criteria and were enrolled (n = 17 in the experimental group; n = 19 in the control group). Attrition was documented to assess feasibility. Six participants discontinued participation due to transportation or physical health issues (n = 4), depressive symptoms (n = 1), or religious conflict (n = 1). Consequently, 30 participants (n = 12 in the experimental group; n = 18 in the control group) completed the final QoL assessment. Regarding objective sleep analyses, two additional participants withdrew (one from each group): one due to a perceived lack of security with the wrist-worn XA-5 accelerometer, and the other due to dizziness potentially associated with their arrhythmia. The final analytical sample for sleep outcomes included 28 participants (n = 11 in the experimental group; n = 17 in the control group). Participant flow is detailed in Fig. 1. Fig. 1Flowchart of participant recruitment. BMS, body–mind–spirit; ITT, intention-to-treat Flowchart of participant recruitment. BMS, body–mind–spirit; ITT, intention-to-treat Intervention Logistical support and accessibility Trained volunteers accompanied participants from the nearest MRT station to the venue to mitigate barriers. This support was essential because the venue is 1.3 km from the station, a 15–20 min walk across busy urban intersections. Volunteers were trained to offer only logistical and administrative help, staying supportive but neutral to ensure outcomes were due to the BMS program, not extra social support. Volunteers assisted with travel and administrative procedures at the station. To enhance engagement, all materials were provided in auditory formats. BMS group therapy The experimental group participated in a 10-week BMS program comprising weekly 130-minute sessions. The BMS model integrates Eastern gratitude-based philosophies (e.g., observing merit, appreciating kindness) with Western psychotherapeutic concepts (e.g., emotional regulation, forgiveness). The intervention comprised three core domains: Physical: Hand flexibility exercises and gentle stretching to support circulation, along with diaphragmatic breathing to promote autonomic balance and muscle relaxation. Psychological: Practices focused on positive thinking, self-reflection, and emotional processing to enhance self-worth. Themes of loss, acceptance, and personal growth were explored to support participants in letting go of attachments. Spiritual and Relational: Group sharing of life stories was intended to facilitate emotional resonance, mutual support, and social connectedness. Physical: Hand flexibility exercises and gentle stretching to support circulation, along with diaphragmatic breathing to promote autonomic balance and muscle relaxation. Physical: Hand flexibility exercises and gentle stretching to support circulation, along with diaphragmatic breathing to promote autonomic balance and muscle relaxation. Psychological: Practices focused on positive thinking, self-reflection, and emotional processing to enhance self-worth. Themes of loss, acceptance, and personal growth were explored to support participants in letting go of attachments. Psychological: Practices focused on positive thinking, self-reflection, and emotional processing to enhance self-worth. Themes of loss, acceptance, and personal growth were explored to support participants in letting go of attachments. Spiritual and Relational: Group sharing of life stories was intended to facilitate emotional resonance, mutual support, and social connectedness. Spiritual and Relational: Group sharing of life stories was intended to facilitate emotional resonance, mutual support, and social connectedness. A core affirmation used to foster resilience was: “My VI limits my mobility, but it cannot confine my spirit.” To promote self-efficacy, role models with VI lived experiences, and a reward-based reflection system encouraged participation. Sessions were facilitated by a doctoral-level nurse trained in BMS intervention and Yalom’s interpersonal process group therapy, creating a safe and supportive environment. The thematic structure is summarized in Table 1. Table 1Protocol for body–mind–spirit group therapy sessionsWeekTopicOutline1What is HappinessMutual understanding of group objectives, expectations, and normsExploring keys to Happiness and well-being, recognizing emotions3Effective CommunicationBuild supportive networks and identify communication pitfallsEstablish healthy relationships for well-being3Gratitude—Transforming the MindUnderstand the benefits of gratitude and transform resentmentAppreciate others’ merits and learn from them4Gratitude—Turning Adversity into OpportunityIncrease awareness of thoughts and intentionPractice doing good, speaking well, and maintaining positivity5Treasuring People Around YouCultivate a group atmosphere that recognizes each individual’s strengthsExperience inner joy through gratitude6There’s No Such Thing as Bad LuckBuild supportive networks and cultivate a positive outlook7Another Truth About LifeReflect on life goals and meaningExplore perspectives on disability and identify life values following loss8Happiness is Self-CreatedAppreciate goodness and actively engage in good deedsTake responsibility for mood regulation and happiness9Rebuilding the Meaning of LifeLearn to utilize resources to transform mindsetRebuild meaning and value in life10I Have Direction in LifeBecome familiar with group members and appreciate others’ strengthsEstablish a life direction and enhance well-being Table 1Protocol for body–mind–spirit group therapy sessionsWeekTopicOutline1What is HappinessMutual understanding of group objectives, expectations, and normsExploring keys to Happiness and well-being, recognizing emotions3Effective CommunicationBuild supportive networks and identify communication pitfallsEstablish healthy relationships for well-being3Gratitude—Transforming the MindUnderstand the benefits of gratitude and transform resentmentAppreciate others’ merits and learn from them4Gratitude—Turning Adversity into OpportunityIncrease awareness of thoughts and intentionPractice doing good, speaking well, and maintaining positivity5Treasuring People Around YouCultivate a group atmosphere that recognizes each individual’s strengthsExperience inner joy through gratitude6There’s No Such Thing as Bad LuckBuild supportive networks and cultivate a positive outlook7Another Truth About LifeReflect on life goals and meaningExplore perspectives on disability and identify life values following loss8Happiness is Self-CreatedAppreciate goodness and actively engage in good deedsTake responsibility for mood regulation and happiness9Rebuilding the Meaning of LifeLearn to utilize resources to transform mindsetRebuild meaning and value in life10I Have Direction in LifeBecome familiar with group members and appreciate others’ strengthsEstablish a life direction and enhance well-being Protocol for body–mind–spirit group therapy sessions Mutual understanding of group objectives, expectations, and norms Exploring keys to Happiness and well-being, recognizing emotions Build supportive networks and identify communication pitfalls Establish healthy relationships for well-being Understand the benefits of gratitude and transform resentment Appreciate others’ merits and learn from them Increase awareness of thoughts and intention Practice doing good, speaking well, and maintaining positivity Cultivate a group atmosphere that recognizes each individual’s strengths Experience inner joy through gratitude Reflect on life goals and meaning Explore perspectives on disability and identify life values following loss Appreciate goodness and actively engage in good deeds Take responsibility for mood regulation and happiness Learn to utilize resources to transform mindset Rebuild meaning and value in life Become familiar with group members and appreciate others’ strengths Establish a life direction and enhance well-being Comparison group The comparison group served as a wait-list control receiving usual care, consisting of routine life-skills and orientation and mobility training. To ensure intervention purity, these participants did not receive any BMS spiritual or psychological components (e.g., gratitude or forgiveness exercises) during the 10-week study. Monthly telephone check-ins were conducted to monitor well-being and maintain retention. Following the post-test, a condensed BMS workshop was offered to this group. Data collection: quantitative inquiry Demographic data Baseline demographic and clinical characteristics—including sex, age, educational level, marital status, religion, employment status, and physiological comorbidities—were collected. VI severity was clinically assessed by certified ophthalmologists and classified as mild, moderate, or severe. QoL The primary outcome measure was the 28-item World Health Organization QoL Questionnaire—Taiwan version (WHOQOL-BREF TW). This instrument assesses health-related QoL by capturing multidimensional health status and subjective satisfaction with outcomes, making it suitable for individuals with VI. The questionnaire comprises four domains: physical health (seven items), psychological well-being (six items), social relationships (four items), and environmental factors (nine items). Items are rated on a five-point Likert scale, with higher scores indicating better perceived QoL. Domain scores are the sum of item scores multiplied by 4 and divided by the number of items, ranging from 4 to 20 [1]. The WHOQOL-BREF TW shows strong psychometrics, with overall internal consistency (Cronbach’s α) of 0.97, test–retest reliability of 0.86, and domain-specific internal consistency from 0.68 to 0.76. Confirmatory factor analysis indicates good model fit (CFI = 0.886), and the content validity index is 0.84 [1, 21, 22]. Prior studies have confirmed satisfactory psychometric performance across Taiwanese populations, including older adults and individuals with VI [1, 5, 23–26]. Subjective sleep quality Sleep outcomes were assessed using subjective and objective measures to capture a comprehensive profile. Overall subjective sleep satisfaction was assessed using WHOQOL-BREF TW Item 16 (“Are you satisfied with your sleep?”), reflecting subjective sleep satisfaction [1, 5, 21–24]. This single-item approach minimizes participant burden, as auditory-assisted completion of lengthy scales (e.g., PSQI) can fatigue individuals with VI. To improve sensitivity to specific sleep complaints, supplementary items assessed self-reported sleep latency and time to sleep onset, offering nuanced insights into individual sleep experiences. Objective sleep parameters Objective sleep patterns were monitored using the XA-5 wrist-worn accelerometer. Although actigraphy has known limitations—including low specificity (~ 0.55), potential overestimation of sleep duration due to misclassification of wakefulness as sleep, and limited sensitivity for detecting sleep latency [27–30]—the XA-5 was selected for its non-invasive design and feasibility for community-based monitoring. The device demonstrates acceptable reliability (ICC 0.54–0.80), supporting its use for estimating sleep duration and timing [30]. Despite these constraints, group-level estimation errors remain below 30 min, and sleep detection accuracy among older adults can reach 80% [31]. The measured parameters included total sleep time, sleep efficiency, and awakening time (operationally defined as the total minutes of nocturnal wakefulness, comprising the sum of sleep latency and minutes of wakefulness after sleep onset [31]). To standardize the assessment, the monitoring period was fixed from 21:00 to 07:00 [31]. Baseline and post-intervention data were recorded for at least three consecutive days to obtain a stable mean value of habitual sleep patterns and to mitigate confounding from atypical daily events. Data were extracted and analyzed using proprietary software from the KY Laboratory. Concurrently, average daytime physical activity and physiological signals were recorded from 09:00 to 18:00 to account for daytime influences on sleep quality. Statistical analyses All analyses were conducted using SPSS Statistics for Windows, Version 24.0 (IBM Corp., Armonk, NY, USA). Descriptive statistics, including demographic characteristics (age, occupation, religion, sex, and VI severity) and baseline data (week 1), were used to examine potential attrition effects. Results are reported as counts (n), percentages (%), means, and standard deviations. Primary analyses followed an intention-to-treat (ITT) approach. Chi-square tests were applied to categorical variables. Specifically, the Mann–Whitney U test was used to assess baseline (Week 1) differences between the intervention and control groups for continuous variables, including WHOQOL-BREF TW scores, total sleep time, sleep efficiency, and awakening time. To evaluate intervention effects over time (Weeks 1 and 10), a generalized estimating equation (GEE) model was applied to examine the main effects of group and time, as well as their interaction. A significant interaction indicated differential changes between the groups. To account for potential confounding effects due to demographic imbalances, sex and VI severity were included as covariates in the model. For sleep-related outcomes, average daytime physical activity was also controlled to improve model accuracy. The GEE provided robust estimates under missing-at-random assumptions to handle incomplete repeated-measures data without imputation. Furthermore, to evaluate the magnitude of the intervention effects, effect sizes (Cohen’s d) were consistently calculated across all outcomes based on between-group differences in mean change divided by the pooled baseline standard deviation (SD). This unified approach provides conservative and robust estimates of practical significance that are unaffected by intervention-induced changes in variance. Results Participants’ demographic and clinical characteristics Thirty-six participants completed the baseline assessment, with 17 assigned to the experimental group and 19 to the control group. At the final evaluation (week 10), attrition rates were 35% in the experimental group and 11% in the control group. The higher attrition in the experimental group was primarily due to the rigorous 10-week intervention schedule and transportation challenges faced by some participants with VI. To mitigate potential attrition bias, we employed an ITT approach and utilized GEE, which provide robust estimates for repeated-measures data with missing values. Specifically, 12 participants in the experimental group and 18 in the control group completed the WHOQOL-BREF TW, whereas 11 and 17 participants, respectively, completed the sleep quality assessment. Chi-square tests of key categorical variables (Table 1) indicated no significant between-group differences in educational level, marital status, employment status, or VI severity. While the difference in sex distribution did not reach statistical significance (76% in the experimental group vs. 47% in the control group), this substantial proportional imbalance was identified as a potential confounder due to limited statistical power. To address this, sex was included as a covariate in the subsequent GEE analyses to statistically control for its influence. Additionally, a significant difference was observed in religious affiliation (p = 0.02), with a higher proportion of Buddhist participants in the experimental group. The experimental group also tended to include older individuals and retirees, indicating some demographic variation. To further evaluate baseline equivalence, Mann–Whitney U tests were performed on continuous variables, including age, average daytime activity, WHOQOL-BREF TW scores, total sleep time, sleep efficiency, and awakening time. Although most indicators showed no significant between-group differences, average daytime activity was significantly lower in the experimental group before the intervention. This difference may be associated with the older age profile of the experimental group, which could have influenced activity levels. Overall, these findings support the comparability of the experimental and control groups at the onset of the study (Table 2). Table 2Participant characteristics (N = 36)CharacteristicsExperiment(n = 17)n (%)Control(n = 19)n (%)p a Educational level 0.29Elementary school (≤ 6 years)2 (11.8%)0 (0%)Junior high school (7–9 years)3 (17.6%)4 (21.1%)Senior high school or college (10–12 years)3 (17.6%)7 (36.8%)University or higher (≥ 13 years)9 (53%)8 (42.1%) Sex 0.74Male4 (23.5%)10 (52.6%)Female13 (76.5%)9 (47.4%)Marital status0.63Unmarried3 (17.6%)7 (36.8%)Married10 (58.9%)8 (42.1%)Divorced3 (17.6%)3 (15.8%)Widowed1 (5.9%)1(5.3%) Working status 0.06No14 (82.4%)10 (52.6%)Yes3 (17.6%)9 (47.4%) Degree of visual impairment 0.50Mild0 (0%)1 (5.3%)Moderate2 (11.8%)1 (5.3%)Severe15 (88.2%)17 (89.4%) Religion 0.02Taoism6 (35.3%)13 (68.4%)Buddhist7 (41.2%)0 (0%)Christian4 (23.5%)5 (26.3%)Nichiren Buddhism0 (0%)1 (5.3%)Mean (SD)p bAge, years64.44 ± 7.558.95 ± 13.40.21WHOQOL-BREF TW scores b56.37 ± 8.656.15 ± 9.00.86Daytime average activity23.64 ± 9.433.05 ± 14.10.02Total sleep time341.56 ± 100.5362.11 ± 94.10.57Sleep efficiency86.23 ± 5.386.46 ± 7.70.68Awakening time53.18 ± 18.054.17 ± 30.00.80Sleep latency71.7 ± 30.778 ± 25.20.70a The chi-squared test was used to analyze categorical variablesb The Mann–Whitney U test was performed to analyze continuous variablesWHOQOL-BREF, World Health Organization QoL Questionnaire—Taiwan version; SD, standard deviation Table 2Participant characteristics (N = 36)CharacteristicsExperiment(n = 17)n (%)Control(n = 19)n (%)p a Educational level 0.29Elementary school (≤ 6 years)2 (11.8%)0 (0%)Junior high school (7–9 years)3 (17.6%)4 (21.1%)Senior high school or college (10–12 years)3 (17.6%)7 (36.8%)University or higher (≥ 13 years)9 (53%)8 (42.1%) Sex 0.74Male4 (23.5%)10 (52.6%)Female13 (76.5%)9 (47.4%)Marital status0.63Unmarried3 (17.6%)7 (36.8%)Married10 (58.9%)8 (42.1%)Divorced3 (17.6%)3 (15.8%)Widowed1 (5.9%)1(5.3%) Working status 0.06No14 (82.4%)10 (52.6%)Yes3 (17.6%)9 (47.4%) Degree of visual impairment 0.50Mild0 (0%)1 (5.3%)Moderate2 (11.8%)1 (5.3%)Severe15 (88.2%)17 (89.4%) Religion 0.02Taoism6 (35.3%)13 (68.4%)Buddhist7 (41.2%)0 (0%)Christian4 (23.5%)5 (26.3%)Nichiren Buddhism0 (0%)1 (5.3%)Mean (SD)p bAge, years64.44 ± 7.558.95 ± 13.40.21WHOQOL-BREF TW scores b56.37 ± 8.656.15 ± 9.00.86Daytime average activity23.64 ± 9.433.05 ± 14.10.02Total sleep time341.56 ± 100.5362.11 ± 94.10.57Sleep efficiency86.23 ± 5.386.46 ± 7.70.68Awakening time53.18 ± 18.054.17 ± 30.00.80Sleep latency71.7 ± 30.778 ± 25.20.70a The chi-squared test was used to analyze categorical variablesb The Mann–Whitney U test was performed to analyze continuous variablesWHOQOL-BREF, World Health Organization QoL Questionnaire—Taiwan version; SD, standard deviation Participant characteristics (N = 36) a The chi-squared test was used to analyze categorical variables b The Mann–Whitney U test was performed to analyze continuous variables WHOQOL-BREF, World Health Organization QoL Questionnaire—Taiwan version; SD, standard deviation Preliminary effects of BMS group therapy on QoL A group-by-time interaction analysis revealed significantly greater improvements in overall QoL for the experimental group (p = 0.01, Cohen’s d = 0.58; Table 3). Subdomain analyses confirmed significant interaction effects with moderate-to-large effect sizes in physical health (d = 0.71), environmental factors (d = 0.52), and psychological well-being (d = 0.47). No significant effect was found in social relationships (p = 0.30, d = 0.26). While the experimental group showed slightly higher percentage gains in psychological well-being (12%) than in physical health (11%), physical health yielded a more robust effect size. This divergence is primarily due to the score decline observed in the comparison group’s physical domain, which widened the inter-group gap (Table 3; Fig. 2). Table 3Generalized estimating equation analysis for sleep quality and quality of lifeCharacteristicsWeek 1aWeek 10 bWald X 2 gmeanmeanExp cCom dExp cCom dDiff eES f p WHOQOL-BREF TW h56.456.161.255.09%0.585.820.01Physical health i1413.815.513.211%0.716.880.009Psychological well-being j14.514.616.314.712%0.474.130.04Social relationships k13.713.514.213.04%0.261.20.3Environmental factors l14.114.315.414.09%0.524.850.02Sleep time m351.3360.0374.0359.16%0.180.480.45Sleep efficiency n85.586.987.186.92%0.290.360.49Awakening time o55.752.360.360.18%−0.150.530.81Sleep latency p73.871.251.357.4−23%−0.321.450.22a Week 1: Baselineb Week 10: Post-test 1c Exp: Experimental groupd Com: Comparison groupe Diff: Percentage change in the experimental group, calculated as (post-test – pretest) / pretestf ES: effect size, based on between-group differences in mean change and the pooled baseline SD of the two groupsg Wald X2: Generalized estimating equation interaction effect (group × time point)h–l WHOQOL-BREF TW: World Health Organization Quality of Life Questionnaire: Taiwanese version. This scale assesses the absence of limitations in physical health, psychological well-being, social relationships, and environmental factorsm Sleep time: Calculated as sleep period time (time between sleep onset and sleep offset) minus wake after sleep onsetn Sleep efficiency = (total sleep time/time in bed) × 100o Awakening time: Minutes of wakefulness after sleep onset and sleep latencyp Sleep latency: Time (in minutes) from going to bed to falling asleep Table 3Generalized estimating equation analysis for sleep quality and quality of lifeCharacteristicsWeek 1aWeek 10 bWald X 2 gmeanmeanExp cCom dExp cCom dDiff eES f p WHOQOL-BREF TW h56.456.161.255.09%0.585.820.01Physical health i1413.815.513.211%0.716.880.009Psychological well-being j14.514.616.314.712%0.474.130.04Social relationships k13.713.514.213.04%0.261.20.3Environmental factors l14.114.315.414.09%0.524.850.02Sleep time m351.3360.0374.0359.16%0.180.480.45Sleep efficiency n85.586.987.186.92%0.290.360.49Awakening time o55.752.360.360.18%−0.150.530.81Sleep latency p73.871.251.357.4−23%−0.321.450.22a Week 1: Baselineb Week 10: Post-test 1c Exp: Experimental groupd Com: Comparison groupe Diff: Percentage change in the experimental group, calculated as (post-test – pretest) / pretestf ES: effect size, based on between-group differences in mean change and the pooled baseline SD of the two groupsg Wald X2: Generalized estimating equation interaction effect (group × time point)h–l WHOQOL-BREF TW: World Health Organization Quality of Life Questionnaire: Taiwanese version. This scale assesses the absence of limitations in physical health, psychological well-being, social relationships, and environmental factorsm Sleep time: Calculated as sleep period time (time between sleep onset and sleep offset) minus wake after sleep onsetn Sleep efficiency = (total sleep time/time in bed) × 100o Awakening time: Minutes of wakefulness after sleep onset and sleep latencyp Sleep latency: Time (in minutes) from going to bed to falling asleep Generalized estimating equation analysis for sleep quality and quality of life a Week 1: Baseline b Week 10: Post-test 1 c Exp: Experimental group d Com: Comparison group e Diff: Percentage change in the experimental group, calculated as (post-test – pretest) / pretest f ES: effect size, based on between-group differences in mean change and the pooled baseline SD of the two groups g Wald X2: Generalized estimating equation interaction effect (group × time point) h–l WHOQOL-BREF TW: World Health Organization Quality of Life Questionnaire: Taiwanese version. This scale assesses the absence of limitations in physical health, psychological well-being, social relationships, and environmental factors m Sleep time: Calculated as sleep period time (time between sleep onset and sleep offset) minus wake after sleep onset n Sleep efficiency = (total sleep time/time in bed) × 100 o Awakening time: Minutes of wakefulness after sleep onset and sleep latency p Sleep latency: Time (in minutes) from going to bed to falling asleep Fig. 2a–e After 10-week BMS therapy, the experimental group showed significant improvements in overall QoL (2a), physical health (2b), psychological (2c), and environment (2e) compared with the control group; no significant change was found in social relationships (2d) a–e After 10-week BMS therapy, the experimental group showed significant improvements in overall QoL (2a), physical health (2b), psychological (2c), and environment (2e) compared with the control group; no significant change was found in social relationships (2d) Preliminary effect of BMS group therapy on sleep quality Subjective and objective sleep indicators showed no significant between-group interaction effects (p = 0.22–0.81; Table 3). Subjectively, the experimental group demonstrated a non-significant improving trend (B = 0.44, p = 0.24). Objective accelerometer data revealed similar patterns: total sleep time increased from 351 to 374 min (6%), and sleep efficiency rose from 85.5% to 87.1% (Fig. 3a–b). Notably, the experimental group showed a reduction in subjective sleep latency of 22.5 min (a 23% decrease), although the between-group difference did not reach statistical significance (p = 0.22; Fig. 3d; Table 3). Fig. 3a–d Changes in sleep parameters from baseline to post-test. The experimental group showed slight improvements in sleep time (3a) and efficiency (3b), but no significant group differences were found in sleep time, efficiency, awakening time (3c), or sleep latency (3d) a–d Changes in sleep parameters from baseline to post-test. The experimental group showed slight improvements in sleep time (3a) and efficiency (3b), but no significant group differences were found in sleep time, efficiency, awakening time (3c), or sleep latency (3d) Exploratory analyses Exploratory analyses underscored the association between sleep and well-being. Participants averaging more than six hours of sleep per night reported significantly higher overall QoL than those with shorter durations (mean difference = 2.85; p = 0.006). Each one-level increase in sleep satisfaction corresponded to a 4.38-point increase in QoL (p < 0.001), with psychological health showing the strongest correlation with total sleep time (p = 0.001). Discussion This study demonstrated that BMS group therapy significantly improved overall QoL among individuals with VI, particularly in the physical, psychological, and environmental domains. While objective sleep improvements via actigraphy were not statistically significant, subjective reports showed a notable reduction in sleep latency. Longer sleep duration and higher sleep satisfaction were positively associated with better QoL, especially psychological health. These findings suggest that enhanced subjective sleep quality may be linked to the well-being promoted by BMS therapy. Overall, the results imply that BMS improves QoL primarily through psychological and emotional support. The observed trends in sleep and environmental adaptability warrant further investigation as potential contributors to QoL, while sociocultural norms may influence how intervention effects are reflected in social relationships. BMS group therapy on QoL Results from this pilot study provide encouraging preliminary evidence that BMS group therapy may support improvements in QoL among individuals with VI, particularly in psychological well-being, physical health, and environmental adaptability. These findings align with prior research supporting holistic, compassion-based approaches for individuals with disabilities [19, 32]. Improvements in psychological distress (12% increase) may be linked to the program’s emphasis on mindfulness and gratitude. As noted by Hsiao, mindfulness practice can alleviate psychological distress [33]. In this study, core BMS components—such as recognizing strengths, cultivating gratitude, and practicing forgiveness—likely fostered self-acceptance and resilience. This is consistent with the literature indicating that positive psychological traits can buffer depression and anxiety and facilitate adaptive coping among individuals with VI [12, 20, 33–35]. Regarding physical health, this domain demonstrated significant clinical value despite a slightly lower percentage gain (11%) compared to the psychological domain. This divergence reflects the score decline in the comparison group, suggesting that BMS therapy may effectively prevent the functional deterioration often observed with conventional care. Improvements in this domain were most pronounced in subjective sleep satisfaction (slope = 4.45, p < .001). However, heterogeneity was observed, with younger participants showing smaller gains, likely due to their significantly shorter sleep durations. Qualitative feedback suggests that this relates to the irregular late-night work schedules common among younger adults with VI employed in massage therapy. This finding aligns with evidence that sleep quality is a strong predictor of QoL [36, 37]. For future trials, these results suggest the value of stratifying participants by occupation or integrating tailored sleep hygiene education. In the environmental domain, the findings highlight that many participants initially avoided independent travel because of psychological anxiety rather than physical barriers [38]. Group-based encouragement appeared to reduce this avoidance, indicating that supportive psychosocial contexts are essential for translating infrastructure accessibility into perceived environmental satisfaction. By contrast, gains in the social relationships domain were modest. This domain includes sensitive indicators, such as satisfaction with sex life and support from friends. The limited improvement likely reflects cultural conservatism and a reluctance to disclose such private vulnerabilities within a short-term group setting [1]. Overall, the pattern suggests a potential sequential change process, in which initial improvements in mindfulness lead to better sleep and eventually extend to broader QoL. Taken together, these findings suggest a pathway whereby BMS group therapy may first enhance emotional regulation, subsequently improving subjective sleep experiences and overall QoL. Preliminary findings on sleep quality and future implications Baseline assessments confirmed substantial sleep impairment among participants, who averaged 351 min of sleep per night—markedly shorter than the 420 min typical of healthy older adults [39]. Awakening time (defined as minutes of wakefulness after sleep onset and sleep latency) averaged 55 to 60 min, indicating highly fragmented sleep consistent with the circadian vulnerability observed in individuals with VI [40]. Participants sleeping fewer than six hours per night also reported significantly lower QoL, reinforcing established associations between short sleep duration and adverse health outcomes [10, 14, 17]. Post-intervention analyses revealed clinically meaningful trends. Subjective sleep latency showed a moderate improvement (d = −0.32), with a reduction of 22.5 min (a 23% decrease). This suggests that BMS therapy facilitates sleep initiation by reducing pre-sleep cognitive and emotional arousal. However, objective metrics from the XA-5 accelerometer showed only small to negligible effects: total sleep time increased by 6% (from 351 to 374 min; d = 0.16), and sleep efficiency rose slightly (d = 0.08). The stability of these objective measures may have been compromised by the monitoring process itself. Despite the presence of trained volunteers to ensure safe navigation, the novelty and physical presence of the wearable device remained a psychological stressor for some participants. This was evidenced by two dropouts: one due to a perceived lack of security with the device, and the other due to dizziness potentially associated with arrhythmia. Such psychological and physiological sensitivities toward wearable technology may interfere with natural sleep patterns, thereby limiting observable improvements in objective sleep metrics. Furthermore, the discrepancy between improved QoL and inconsistent objective sleep may be explained by the “circadian hardwiring” of the suprachiasmatic nucleus (SCN). In the absence of photic entrainment, melatonin secretion remains dysregulated—a physiological barrier that psychological interventions may struggle to override within a 10-week timeframe [10, 16, 31, 41–43]. Overall, the effects of BMS therapy on sleep appear to be mediated primarily through emotional regulation rather than direct circadian mechanisms. Future trials should consider multimodal strategies that integrate BMS with circadian-focused approaches, such as structured light exposure, to optimize recovery pathways. Feasibility of the intervention model This pilot study demonstrated the high feasibility and acceptability of the BMS group intervention among individuals with VI, evidenced by a 91% attendance rate in the experimental group. Strong engagement reflects the effectiveness of incorporating reflective practices—particularly observing virtues and cultivating gratitude—in fostering emotional resonance and intrinsic motivation. Consistent with prior research, gratitude was positively associated with life satisfaction and served as a predictor of psychological well-being [44, 45]. This holistic approach diverges from traditional VI rehabilitation, which often focuses primarily on functional training, highlighting the value of addressing the whole person to renew a sense of purpose and interpersonal connection (“life impacting life”). However, although the intervention itself was highly feasible, challenges emerged in measuring feasibility. Despite wrist-worn accelerometers (XA-5) being widely regarded as safe and non-invasive [46, 47], several participants reported anxiety or somatic discomfort (e.g., chest tightness) related to the device. Because only individuals with pacemakers are required to avoid electromagnetic interference [48], these reactions were interpreted as anxiety-related rather than physiological. These findings highlight an important implication for future trials: while holistic interventions such as BMS support emotional health, data collection methods must also be user-centered. Future research should integrate psychological strategies with anxiety-reducing, accessible wearable technologies to ensure high compliance and accurate physiological assessments. Limitations This pilot study has several limitations. First, the non-randomized design and small analytical sample size (N = 28–30), necessitated by community recruitment barriers and a 16.7% attrition rate, limit generalizability. Participation was significantly hindered by the mobility limitations and difficulties with independent travel inherent to the VI population. Such logistical challenges contributed to the reduced sample size and represent a primary feasibility concern for community-based interventions. While the sample size was sufficient to detect significant improvements in QoL, it may have lacked the statistical power to capture subtle changes in secondary sleep parameters. Furthermore, baseline imbalances—notably, a higher proportion of Buddhist participants in the experimental group—may have increased receptivity to the spiritual components of the BMS intervention. Second, potential physiological confounders, such as light perception (LP) status and non-standardized medication use, were not fully controlled. Third, the XA-5 accelerometer lacked specificity for sleep architecture and, as noted in the feasibility section, caused discomfort or anxiety in some participants, representing a potential source of measurement error. Fourth, this study assessed outcomes only immediately after the 10-week intervention. Without longitudinal follow-up data (e.g., at 3 or 6 months), the long-term sustainability of the observed improvements in QoL and psychological resilience remains unclear. Despite these limitations, this study establishes the feasibility of the BMS model and provides encouraging preliminary evidence to inform future large-scale randomized controlled trials. Conclusions This pilot study demonstrates the feasibility and acceptability of BMS group therapy for individuals with VI. Preliminary results indicate significant improvements in subjective QoL, particularly in the psychological and physical domains. While objective sleep parameters remained stable—likely due to inherent circadian dysregulation of the SCN in this population—the intervention effectively enhanced psychological resilience and subjective well-being. Rather than demonstrating definitive efficacy, these findings establish a foundation for future randomized controlled trials. Further research with longitudinal follow-up is warranted to evaluate the sustainability of these psychological buffering effects in community care. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1