Sleep and workload mediate the relationship between digital engagement and burnout: a national cross-sectional study of university faculty in India Authors: Mohammad Sidiq, Balamurugan Janakiraman, Faizan Kashoo, Jaya Shanker Tedla, Veeragoudhaman TS, Aksh Chahal, Bartosz Maciej Wojcik, Akriti Pandey, Rajkumar Krishnan Vasanthi, Jyoti Sharma, Muzammil Khan, Gaurav Shrivastava, Richa Hirendra Rai, Shabnam Khan, Chhavi Arora Sehgal, Amir Iqbal, Zafrul Hasan Published: 2026 ABSTRACT AbstractBackgroundIncreasing digital engagement among university faculty has intensified screen exposure, extended working hours, and disrupted sleep, potentially contributing to burnout. This study investigated the relationships between digital engagement and burnout among university faculty in India, with a particular focus on the mediating roles of weekly work hours and sleep patterns.MethodsA national cross-sectional online survey was conducted among 3,987 university faculty members across all geographical regions of India between February and October 2024. Data included socio-demographic characteristics, daily screen time, sleep quality assessed using the Pittsburgh Sleep Quality Index, physical activity levels, and burnout measured using the Copenhagen Burnout Inventory. Multiple regression analyses were performed to identify predictors of personal, work-related, and client-related burnout. Mediation analyses examined indirect effects of work hours and sleep duration on the relationship between screen time and burnout outcomes.ResultsParticipants (mean age: 43.92 years; mean professional experience: 15.77 years) reported significant screen exposure (5.02 h daily) and poor sleep quality (PSQI: 7.5). Multiple regression analyses (N = 3,987) suggest sleep hours, weekly work hours, and screen time as universal burnout predictors. Daily screen time was associated with burnout outcomes in adjusted analyses, with mediation models indicating that these relationships were primarily indirect. Higher screen time was associated with increased weekly work hours, which in turn were linked to higher burnout, while more sleep duration showed a protective association with reduced burnout. Screen time also demonstrated a direct effect on personal burnout (β = 0.327, p < .001), alongside indirect effects via work hours (β = 0.301, p < .001) and sleep duration (β = -0.109, p < .001). Notably, the direct effect was observed only for personal burnout, whereas the associations with work-related and client-related burnout were predominantly mediated.ConclusionThe findings of this study highlight the possibility of screen time effects on burnout, which operates primarily through work-related demands and sleep-related mechanisms, emphasizing the importance of workload management and suggesting a recovery process in mitigating burnout among faculty members of higher education institutions.Trial registrationThe study protocol was prospectively registered with the Clinical Trial Registry India; assigned identifier: CTRI/2024/02/063155, dated 26/02/2024.Supplementary InformationThe online version contains supplementary material available at 10.1186/s13690-026-01954-8.Keywords: Social media, Academic staff, Turnover, Stress, Psychological detachment, Subjective vitality, Media use Text box 1. Contribution to the Literature• Adds population-level evidence on the relationship between digital work practices and burnout among university faculty, a group less studied in public health research.• Demonstrates how working hours and sleep patterns are important pathways linking digital engagement to occupational well-being.• Extends existing burnout literature by incorporating lifestyle and workload factors within academic work environments.• Informs public health discussions on digital work demands and employee well-being in higher education settings. • Adds population-level evidence on the relationship between digital work practices and burnout among university faculty, a group less studied in public health research. • Demonstrates how working hours and sleep patterns are important pathways linking digital engagement to occupational well-being. • Extends existing burnout literature by incorporating lifestyle and workload factors within academic work environments. • Informs public health discussions on digital work demands and employee well-being in higher education settings. Background Recent studies indicate that extended exposure to digital screens may lead to psychological stress and burnout in academic and professional settings. More time spent on the screens is associated with higher reporting of digital fatigue, stress, and poor mental well-being, particularly in settings where individuals often use digital devices for work and learning [1–4]. The rapid expansion of digital technologies and academic digital featuring has led to academic transformations relying on digital platforms in most developed and developing countries. The academic digital engagement has undoubtedly enhanced work efficiency, made work easier, and improved connectivity in academics, research, and administration in higher education. But they have also contributed to extended working hours and high work demand and more importantly, blurred the boundary between work life and personal life [5]. Despite the rising prevalence of device use in the course of work, work-related stressors and health have been discussed in the literature, studies specifically targeting faculty members are scarce and do not sufficiently reflect the high-load academic work of educators involving teaching, research, and administrative activities that contribute to physical and psychological load [6–8]. Several factors may explain how digital engagements might influence burnout among faculty members [9, 10]. Specifically, extended working hours and poor quality of sleep or disrupted sleep patterns are frequently reported by the literature as one of the important mechanism associated with demanding digital engagement with occupation related strain. Burnout is a pressing concern throughout academia, especially in higher education (university), and recent investigations have inquired into the factors associated and attributed it to changes in teaching models, job demands, research needs, publications, low rewards, excessive screen time, digital literacy, and lower sleep quality [11, 12]. Moreover, faculty members do not have enough ways to disengage from work or from cases when technology acceptance has occurred, and faculty members become vulnerable to risks such as depression, anxiety, and alienation at work [13, 14]. Considering the rising prevalence and growing significance of burnout syndrome in university academicians among researchers [15–17]. There is a need to investigate contextual, teaching system, sustainability-based, and region-specific evidence on the factors associated with burnout in this population based on an adequately powered sample and also explore the specific role of the predictor variables in the complex interaction with burnout syndrome. However, the mechanism through which the digital engagement contributes to burnout still remains insufficiently understood with scarce evidence. Particularly, the extended working hours and poor quality of sleep are suggested to be factors that represent important pathways linking screen-based work demands to burnout-related outcomes among the faculty population [18]. To address this gap, the present study examines the relationship between digital engagement and burnout among university faculty members and in addition, assesses the mediating roles of weekly working hours and sleep quality. In this study, burnout was assessed as a multi-dimensional construct comprising personal, work-related, and client-related components, as defined by the Copenhagen Burnout Inventory. Personal burnout reflects the degree of physical and psychological exhaustion experienced by the individuals, work-related and client-related burnout covers exhaustion related to occupational demands and tiredness associated with student or client interactions. This study is unique in its examination of how digital device use affects university faculty well-being, with an emphasis on weekly work hours, emotional tiredness, and subjective well-being. Unlike previous studies, this study attempts to answer the important question of how workload and sleep patterns would mediate the relationship between digital engagement and burnout among university faculty. Guided by the Job Demands-Resources model, this study conceptualizes daily screen time as a job demand that may increase workload and influence recovery processes such as sleep, thereby contributing to burnout [19]. Hence, in this study, the screen time is treated as the independent variable, burnout as the outcome variable, and weekly working hours and sleep duration as potential mediators. Therefore, the aim of this study is to examine the association between screen time and burnout among university faculty across higher education institutions in India and to determine if working hours and sleep duration mediate the relationship between screen time and burnout. By exploring these relationships, this study will add to the evidence and improve our understanding of how digital work environments influence well-being in academic settings. We hypothesize (Ha) that (Ha1) higher screen time is associated with higher levels of burnout among university faculty, (Ha2) weekly work hours possibly mediate the relationship between screen time and burnout, and (Ha3) sleep duration mediates the relationship between daily screen time and burnout. Methods Study design and ethical approval A web-based cross-sectional survey was conducted among various university academicians across all geographical zones of India through an online survey. The study was approved by the School Ethics Committee, Galgotias University, with reference number SEC/001/24, and the study was prospectively registered with the Clinical Trial Registry India, CTRI/2024/02/063155, dated 26/02/2024. The study adhered to the principles bound in the Declaration of Helsinki [20]. All participants provided informed consent. The data were collected via a structured online Google form. This study is reported following the guidelines of Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) recommendations [21]. Questionnaire and procedure A web-based survey questionnaire was sent to a list of email IDs of Deans, heads of departments, and academicians of various universities across all geographical zones of India (Fig. 1). The online survey questionnaire consisted of a consent form and questions related to socio-demographics, work-related features, screen time, self-reported personal and professional accomplishments, and self-reported burnout levels (Supplementary File 1). The questionnaire items like daily scree time, weekly work hours combining academics, research, and administration works, and sleep hours were open-ended questions asking the participants to fill responses in hours and minutes related to their typical week or a day. The BMI of the respondents were computed from the self-reported height and weight. The data were collected from February 2024 to July 2024. A reminder was sent twice at 30-day intervals to those who failed to respond by subsequent delivery to ensure the response distribution rate. Schematic presentation of the sampling procedure for recruiting faculty members (Legends: RJ – Rajasthan, DL – Delhi, CH – Chandigarh, PB – Punjab, AP – Andhra Pradesh, KA – Karnataka, TN – Tamil Nadu, KL – Kerala, WB – West Bengal, OD – Odisha, BR – Bihar, MH – Maharashtra, GJ – Gujarat, GA – Goa, UP – Uttar Pradesh, MP – Madhya Pradesh, AS – Assam, AR – Arunachal Pradesh) The web-based English language questionnaire consisted of five domains aimed at collecting socio-demographic data, work-related burnout, sleep quality, duration of screen time, and metabolic equivalence of tasks. The Pittsburgh Sleep Quality Index (PSQI) is an open-access-free standardized validated tool to evaluate sleep quality [22]. It consists of seven components, and the responses are rated from 0 to 21, with higher scores indicating poor sleep quality [23]. Global PSQI ranges 0–21; >5 denotes poor sleep quality (sensitivity ~ 89.6%, specificity ~ 86.5%) [23]. The physical activity levels were estimated via the International Physical Activity Questionnaire short form (IPAQ). The level of physical activity was reported for low, moderate, and high activity levels and walking [24]. The total metabolic equivalent of task (MET) was calculated based on each activity recorded as the number of days per week and average minutes per day. Vigorous activity was calculated via METS. With 8 METs for vigorous activity, 4 METs for moderate activity, and 3.3 for walking. Burnout was calculated via the validated and reliable Copenhagen Burnout Inventory (CBI) to estimate personal, work-related, and client-related burnout [24]. The CBI is a widely used, reliable tool in healthcare-related surveys to measure self-reported burnout. The tool has good reliability, construct validity supporting a 3-factor structure, and discriminant validity [25]. It consists of 19 items across three domains rated on a five-point Likert scale (“to a very high degree,” “to a high degree,” “somewhat,” “to a low degree,” “to a very low degree”), which are recoded to 0, 25, 50, 75, and 100, respectively. Domain scores are computed as the mean of item scores within each domain, yielding values from 0 to 100. In this study, values ≥ 50 denoted the presence of burnout. To describe severity, scores were categorized as moderate (50–74), high (75–99), and severe (100). The CBI is available for use without charge, and the instrument was applied in accordance with its publishing license. The web-based questionnaire was administered using Google forms. Participation was on a voluntary basis and anonymous, and no personal data was collected during the survey. The filled form and data were accessible only to the PI of this study. Data was coded and any information that leads to the participant were de-identified by the PI before sending it for data analyses. This study adhered to the General Data Protection Regulation (GDPR) and the Digital Personal Data Protection Act 2023. Sample size, eligibility criteria, and missing data handling A multi-stage sampling procedure was designed with a plan of recruiting university faculty members from all 6 geographical zones on India. We used a convenient sampling method to choose the states (strata) from each zone. And, subsequently from each selected states, two universities were selected by using simple random sampling from the list of the top 10 universities of the respective state. The Deans, Directors, Heads, and Registrars of the selected Universities were contacted to secure email Ids of their faculty members and the confidentiality of the contact details was assured. A total of 6403 mails were secured, after removing the duplicates, and selecting required mail ids for each university, 5992 mails were sent to the respective universities. The power calculated sample size was estimated using the single population proportion formula for infinite population with the following assumptions, we assumed that 50% (p) of faculty members to self-report burnout with the absence of any direct evidence to inform the estimate and a 50% proportion (p = .5) represents the maximum possible variability or heterogeneity in the population studied, 95% confidence interval, and 5% precision and considering the complex multi-stage sampling a design effect of ‘2’ (deff) to improve precision [25, 26]. The derived sample size was n = 768. To achieve adequately powered representation from all the 6 zones (768 × 6 zones, n = 4608), we proposed to recruit 768 samples from each zone and an attrition of 30% (n = 1382) was added to cover the non-response rates in web-based surveys. The final derived sample of n = 5990 was equally distributed to the selected universities (Fig. 1). Adults of both sexes of Indian nationality, working full-time, involved in academics, clinical duty in teaching hospitals, research, and administration, literate in the English language, were considered eligible for this study. Across the sample (n = 3987), item‑level missing-ness ranged from 0% (age) to 7.1% (sleep hours), with a mean of 2.4%. Little’s MCAR test rejected the hypothesis of data being missing completely at random (χ² = 148.2, df = 110, p = .007), so, under a missing‑at‑random assumption, variables with ≤ 7.1% gaps were imputed via multiple imputation by chained equations. Twenty imputed data sets were created using predictive mean matching for continuous variables and logistic or proportional‑odds models for categorical ones; all analysis variables plus eight auxiliary predictors were included to enhance accuracy. Trace‑plot inspection showed stable chains, Monte‑Carlo errors were < 10% of posterior standard deviations, and the fraction of missing information did not exceed 6% for any parameter. Complete‑case (“per‑protocol”) and pooled‑MI analyses differed by a median of only 4% in β coefficients, and a pattern‑mixture sensitivity analysis applying a ± 0.20 SD delta adjustment to variables with > 5% missingness likewise left all substantive conclusions unchanged, confirming the robustness of our findings. Variables and conceptual model This study used the JD-R analytical framework, which conceptualizes the relationship between burnout (dependent variable) based on CBI and screen time as the independent variable. Based on the prior literature and theoretical considerations, the variables in this study were categorized as independent, mediating, dependent, and covariates [7, 9, 11, 13, 15]. The hypothesized relationships among the variables are illustrated in (Fig. 2). Conceptual framework of the hypothesized relationships. Footnote: Conceptual framework of the mediation model examining the association between daily screen time and burnout among university faculty. Weekly work hours and sleep duration are specified as mediators. Models are adjusted for age, body mass index, physical activity, sleep quality, and academic workload Independent variable The independent variable was daily screen time, which was defined as the average hours spent per day using digital devices like computers, tablets, and smartphones. Screen time was self-reported by the participants. Mediating variables Weekly work hours were defined as the total number of hours spent by the participant in a typical work week on professional activities, administrative tasks, research, teletherapy, and personal academic responsibilities. Sleep duration was defined as the average number of hours slept per night. The weekly work hours and sleep duration were self-reported by the participants, and both the mediators were treated as continuous variables. Dependent variable The dependent variable was burnout, assessed using the CBI tool. Burnout was conceptualized as a multidimensional construct comprising personal burnout, work-related burnout, and client-related burnout. The CBI scores was treated as continuous data and higher scores indicate greater levels of burnout. Covariates The analyses in this study were adjusted for selected covariates like age (years), body mass index (BMI, kg/m2), physical activity (MET per week), sleep quality measured using the Pittsburgh Sleep Quality Index (score > 5 indicate poor sleep quality), and academic workload. Based on their potential confounding effects, these variables were included in the regression and mediation model as adjustment factors and not treated as mediators. Statistical analysis Categorical variables were summarized using frequencies and percentages, whereas continuous variables were described using means and standard deviations. The distribution of continuous variables was tested with the Shapiro-Wilk test and Kolmogorov-Smirnov test and visually inspected using histograms and Q-Q plots. Pre-analysis screening showed that daily screen time correlated moderately with personal work-related and client-related burnout and with other variables. while all inter-predictor correlations stayed below 0.70, signaling limited redundancy. Guided by the Job-Demands-Resources framework, which posits that high demands such as long hours and prolonged digital exposure deplete energy, whereas resources like sufficient sleep buffer strain. We entered every demand- and resource-related variable in a single, simultaneous block. This “enter-all” strategy lets each coefficient represent its unique share of explained variance and furnishes clean indirect-effect paths for our planned mediation tests. Multicollinearity diagnostics confirmed the assumption of independence among predictors: 91% of VIFs were < 2, and all were below the conservative cut-off of 5, far beneath the traditional 10-point alarm threshold. Although VIFs for screen-time, work-hours, and PSQI edged higher (1.3), they still indicated negligible inflation, so no variables were removed. Homoscedasticity and normality of residuals were visually inspected and met, and Cook’s-distance values were all < 1, dismissing undue influence. Together, these checks justified retaining the full set in one simultaneous model and set the stage for our subsequent mediation analysis. Mediation analysis was conducted to examine whether mediators, identified through significant correlations, explained the relationship between Daily Screen Time (IV) and burnout outcomes (DV). Based on the conceptual framework (Fig. 2) weekly work hours and sleep duration were considered as potential mediating variables in the relationship between daily screen time and burnout among university faculty. The indirect effect of daily screen time on burnout through the mediator was estimated using the product-of-coefficient method, for instance; the regression coefficient of association between the independent variable and mediator path ‘a’ was multiplied with the coefficient of association between the mediator and dependent variable (path ‘b’). The mediated proportion was estimated by dividing the indirect effect by the total effect of the independent variable on the dependent variable. Total effects represent the overall association between the independent variable (daily screen time) and each burnout outcome, computed as the sum of the direct effect (path c’) and the total indirect effect through all mediators. Mathematically: Total Effect (c) = Direct Effect (c’) + Total Indirect Effect (sum of a x b products). The proportion mediated was calculated by dividing the indirect effect by the total effect and multiplying by 100, representing the percentage of the total effect explained by the mediators. For example, if the total effect of screen time on personal burnout was B = 0.519 and the indirect effect through work hours was B = 0.301, the proportion mediated through work hours would be (0.301/0.519) x 100 = 58.0%. Bootstrapped confidence intervals (5000 samples) were used to assess the statistical significance of indirect effects, with effects considered significant if the 95% CI did not include zero. Correlation analyses was conducted to examine the association among study variables. Multiple linear regression analyses were conducted as exploratory analyses to estimate the association between participants’ characteristics and CBI-based burnout outcomes. The regression analyses included demographic, occupational, and health-related variables. The mediating analysis was conducted using a regression-based approach (product-of-coefficients method) and the covariates were included based on theoretical relevance and literature (Fig. 2), rather than statistical significance in exploratory analyses [9, 11, 13, 15]. Interaction between the independent variables and mediators (X-M interaction) was not included in the primary model. The analyses were conducted via JASP version 18.1.1, with the significance level set at 0.05. Results Participant characteristics A total of 3987 university faculty members participated in this survey. The demographic characteristics of the study participants are presented in Table 1. The mean age of the participants was 43.9 ± 12.75, with an average of 15.7 ± 7.68 work experience. The University faculties were distributed across different regions of India (Fig. 3), with largest proportion from = the South India (23.8%), followed by the North zones (25%). The average daily screen time was 5.02 ± 0.73. Weekly average for work-related activities include 33.3 ± 22.9 h. The mean sleep duration was 6.1 ± 1.9 and mean score on the Pittsburgh Sleep quality Index 7.5 ± 2.9. Physical activity averaged 768.8 ± 196.03. MET [27, 28]. Table 1Demographic characteristics of the study participants (n = 3987)Demographic CharacteristicN (%)Mean (SD)Age (Years)43.9 (12.7)Daily Screen Time (Hours)5.02 (0.73)Sleep Hours6.1 (1.91)Weekly Work Hours33.6 (2.9)Body Mass Index (BMI)29.4 (7.7)Gender Female2544 (63.8%) Male1443 (36.1%)Medical Conditions No Medical Conditions3005 (75.3%) Diabetes276 (6.9%) Hypertension260 (6.5%) Arthritis105 (2.6%) Low Back Pain74 (1.8%) Cardiac Problem7 (0.1%)Education Level Bachelor’s207 (5.1%) Master’s3302 (82.8%) Doctorate478 (11.9%)Professional Role Academic Teaching2864 (71.8%) Clinical Practice1123 (28.1%)Primary Devices Desktop534 (13.3%) Laptop1666 (41.7%) Smartphone615 (15.4%) Tablet1172 (29.4%)Personal Burnout46.7 (11.7)Work related burnout40.4 (15.5)Client related burnout65.1 (21.2)Values are presented as mean (standard deviation) for continuous variables and as N (%) for categorical variables Demographic characteristics of the study participants (n = 3987) Values are presented as mean (standard deviation) for continuous variables and as N (%) for categorical variables Frequency distribution of participants based the 5 Zones of India [27] The participants reported a mean daily screen time of 5.02 h (SD = 0.73), alongside weekly averages for tele-therapy activities (M = 33.3 h, SD = 22.9), administrative tasks (M = 28.6 h, SD = 13.9), and research activities (M = 24.2 h, SD = 17.5). The average sleep duration was 6.1 h per night (SD = 1.9), with a mean Pittsburgh Sleep Quality Index (PSQI) score of 7.5 (SD = 2.9), indicating poor sleep quality. Physical activity, measured in MET minutes per week, had an average of 768.8 (SD = 196.07), with related metrics for vigorous, moderate, and walking METs (Table 1). In terms of physical activity, 38.2% of the respondents were classified as having low activity levels (< 600 MET-min/week), 42.6% had moderate activity (600–3000 MET-min/week), and 19.2% had high activity (> 3000 MET-min/week), based on IPAQ thresholds. The multiple regression analyses for personal, work-related, and client-related burnout reveal key predictors across all domains. For personal burnout, the model (R = .696, R² = 0.484, F(26, 3960) = 123.653, p < .001) explained 48.4% of the variance. Significant predictors included daily screen time (β = 0.241, t = 16.923, p < .001), weekly work hours (β = 0.37, t = 25.39, p < .001), sleep hours (β = -0.562, t = -44.81, p < .001) and PSQI (β = 0.03, t = 2.38, p = .017) (Table 2). Table 2Multiple regression predicting Personal Burnout (N = 3987)PredictorBSEΒt P 95% CI LL95% CI ULToleranceVIF(Intercept)0.6650.708—0.9390.348−0.7222.052——Age−0.0010.006−0.006−0.2480.804−0.0130.0100.442.274Daily Screen Time (h)0.9230.0550.24116.923< 0.0010.8161.0300.8031.245Tele-therapy Activity (h)−0.0040.003−0.030−1.1020.270−0.0100.0030.4162.405Administrative Activity (h)−0.0050.003−0.023−1.3400.180−0.0110.0020.6771.478Research Activity (h)−0.0060.004−0.037−1.6080.108−0.0130.0010.4962.018Personal Activity (h)−0.0030.003−0.012−0.7920.428−0.0090.0040.7651.308Sleep Hours−0.8280.018−0.562−44.808< 0.001−0.864−0.7920.9101.099PSQI score0.0280.0120.0292.3780.0170.0050.0500.9301.075BMI0.0030.0040.0070.6020.547−0.0060.0110.9621.039Weekly Work Hours0.3540.0140.37025.390< 0.0010.3270.3810.7841.276Vigorous MET3.429 × 10⁻⁴4.486 × 10⁻⁴0.0130.7640.445−5.366 × 10⁻⁴0.0010.6971.434Moderate MET−8.522 × 10⁻⁴6.651 × 10⁻⁴−0.024−1.2810.200−0.0024.518 × 10⁻⁴0.6101.640Walking MET−0.0020.001−0.029−1.7420.082−0.0042.299 × 10⁻⁴0.6761.480Gender (Male)−0.0930.075—−1.2430.214−0.2400.0540.8911.122Med. Cond. (Hypertension)0.0250.175—0.1450.885−0.3180.3690.9821.018Med. Cond. (Hypertension + Diabetes)−0.0740.175—−0.4220.673−0.4180.270——Med. Cond. (No Condition)0.0520.129—0.4010.688−0.2010.304——Med. Cond. (Arthritis)−0.2880.239—−1.2090.227−0.7570.180——Med. Cond. (Cardiac)−0.0420.782—−0.0530.958−1.5751.491——Med. Cond. (Low Back Pain)0.5210.270—1.9250.054−0.0101.051——Educ. Level (Doctorate)−0.0360.226—−0.1590.874−0.4800.4080.7531.328Educ. Level (Master’s)−0.1280.160—−0.7990.425−0.4420.186——Prof. Role (Clinician)0.2090.181—1.1520.249−0.1460.5630.3942.539Primary Device (Laptop)−0.0700.103—−0.6780.498−0.2730.1320.9761.025Primary Device (Smartphone)−0.2040.124—−1.6450.100−0.4470.039——Primary Device (Tablet)−0.1680.109—−1.5400.124−0.3810.046——Variance Inflation Factor (VIF) values are not reported for some categorical variables (e.g., medical conditions, education levels, and device types) because VIF is not applicable to reference categories or singular dummy-coded variables in regression models. These entries are marked as “—” in the output, as multicollinearity diagnostics primarily apply to continuous predictors and non-redundant coded categorical variables. The reported VIF values for continuous predictors and binary variables indicate acceptable levels of multicollinearity (all VIF < 2.5), suggesting no problematic collinearity among the substantive predictors in the model Multiple regression predicting Personal Burnout (N = 3987) Variance Inflation Factor (VIF) values are not reported for some categorical variables (e.g., medical conditions, education levels, and device types) because VIF is not applicable to reference categories or singular dummy-coded variables in regression models. These entries are marked as “—” in the output, as multicollinearity diagnostics primarily apply to continuous predictors and non-redundant coded categorical variables. The reported VIF values for continuous predictors and binary variables indicate acceptable levels of multicollinearity (all VIF < 2.5), suggesting no problematic collinearity among the substantive predictors in the model For work-related burnout, the model (R = .904, R² = 0.818, F(26, 3960) = 684.55, p < .001) explained 81.7% of the variance. Significant predictors included sleep hours (β = -0.31, t = -41.598, p < .001), PSQI score (β = 0.909, t = 124.73, p < .001), weekly work hours (β = 0.212, t = 24.516, p < .001), Vigorous MET (β = -0.029, t = -2.94, p = .003), Moderate MET(β = -0.041, t = -3.706, p < .001), Low back pain (β (unstandardized) = − 0.614, t = -2.8, p = .004). doctorate educational level (β (unstandardized) = 0.2.283, t = 12.744, p < .001), and professional role as clinician (β (unstandardized) = 2.241, t = 15.652, p < .001), master’s educational level (β (unstandardized) = 2.252, t = 17.75, p < .001) (Table 3). Table 3Multiple regression predicting Work-related Burnout (N = 3987)PredictorBSEΒt p 95% CI LL95% CI ULToleranceVIF(Intercept)−5.2870.560—−9.444< 0.001−6.385−4.190——Age−0.0230.005−0.077−5.008< 0.001−0.032−0.0140.442.274Daily Screen Time (h)−8.111 × 10⁻⁵0.043−1.586 × 10⁻⁵−0.0020.999−0.0850.0850.8031.245Tele-therapy Activity (h)0.0030.0030.0211.3070.191−0.0020.0090.4162.405Administrative Activity (h)3.216 × 10⁻⁴0.0030.0010.1200.905−0.0050.0060.6771.478Research Activity (h)0.0030.0030.0141.0410.298−0.0030.0090.4962.018Personal Activity (h)0.0070.0020.0252.7920.0050.0020.0120.7651.308Sleep Hours−0.6080.015−0.310−41.598< 0.001−0.637−0.5800.9101.099PSQI score1.1430.0090.909124.730< 0.0011.1251.1610.9301.075BMI−0.0010.003−0.003−0.3810.703−0.0080.0050.9621.039Weekly Work Hours0.2710.0110.21224.516< 0.0010.2490.2920.7841.276Vigorous MET−0.0013.550 × 10⁻⁴−0.029−2.9480.003−0.002−3.506 × 10⁻⁴0.6971.434Moderate MET−0.0025.263 × 10⁻⁴−0.041−3.706< 0.001−0.003−9.184 × 10⁻⁴0.6101.640Walking MET−7.854 × 10⁻⁴8.306 × 10⁻⁴−0.009−0.9460.344−0.0028.431 × 10⁻⁴0.6761.480Gender (Male)0.0890.059—1.5040.133−0.0270.2050.8911.122Med. Cond. (Hypertension)−0.1050.139—−0.7550.450−0.3770.1670.9821.018Med. Cond. (Hypertension + Diabetes)−0.0380.139—−0.2750.783−0.3100.234——Med. Cond. (No Condition)−0.0310.102—−0.3040.761−0.2310.169——Med. Cond. (Arthritis)−0.0330.189—−0.1760.860−0.4040.337——Med. Cond. (Cardiac)0.4530.619—0.7310.465−0.7611.666——Med. Cond. (Low Back Pain)−0.6140.214—−2.8710.004−1.034−0.195——Educ. Level (Doctorate)2.2830.179—12.744< 0.0011.9322.6340.7531.328Educ. Level (Master’s)2.2520.127—17.750< 0.0012.0042.501——Prof. Role (Clinician)2.2410.143—15.652< 0.0011.9602.5220.3942.539Primary Device (Laptop)−0.0220.082—−0.2670.790−0.1820.1380.9761.025Primary Device (Smartphone)−0.0120.098—−0.1220.903−0.2040.180——Primary Device (Tablet)0.0090.086—0.1040.917−0.1600.178——Level of significance p < .05 Multiple regression predicting Work-related Burnout (N = 3987) Level of significance p < .05 For client-related burnout, the model (R = .303, R² = 0.092, F(26, 3960) = 15.357, p < .001) explained 8.6% of the variance, with significant predictors including sleeping hours (β = -0.226, t = -14.108, p < .001), PSQI score (β = 0.098, t = 6.02, p < .001), weekly working hours (β = 0.18, t = 9.304, p < .001), Vigorous MET (β = 0.044, t = -2.041, p = .041), Walking (β = 1.792, t = 3.082, p = .041), doctorate educational level (β (unstandardized) = 0.2.283, t = 12.744, p = .002), and professional role as clinician (β (unstandardized) = 2.215, t = 4.765, p < .001), masters educational level (β (unstandardized) = 1.922, t = 4.665, p < .001). Sleep quality was significantly associated with burnout outcomes and therefore retained in the adjusted model as covariate (Table 4). Table 4Multiple regression predicting Client-related Burnout (N = 3987)PredictorBSEβt p 95% CI LL95% CI ULToleranceVIF(Intercept)8.3221.818—4.578< 0.0014.75811.886——Age−0.0460.015−0.109−3.1550.002−0.075−0.0180.442.274Daily Screen Time (h)−0.1310.140−0.018−0.9320.351−0.4050.1440.8031.245Tele-therapy Activity (h)−0.0040.009−0.017−0.4580.647−0.0210.0130.4162.405Administrative Activity (h)−0.0110.009−0.029−1.2870.198−0.0280.0060.6771.478Research Activity (h)0.0010.0090.0040.1460.884−0.0170.0200.4962.018Personal Activity (h)0.0020.0080.0060.3050.760−0.0130.0180.7651.308Sleep Hours−0.6430.047−0.226−13.544< 0.001−0.736−0.5500.9101.099PSQI score0.1790.0300.0986.020< 0.0010.1210.2370.9301.075BMI0.0020.0110.0020.1530.879−0.0200.0230.9621.039Weekly Work Hours0.3330.0360.1809.304< 0.0010.2630.4040.7841.276Vigorous MET−0.0020.001−0.044−2.0410.041−0.005−9.284 × 10⁻⁵0.6971.434Moderate MET−0.0030.002−0.038−1.5220.128−0.0067.496 × 10⁻⁴0.6101.640Walking MET0.0020.0030.0180.8240.410−0.0030.0080.6761.480Gender (Male)−0.0100.192—−0.0510.959−0.3870.3670.8911.122Med. Cond. (Hypertension)0.0960.451—0.2130.831−0.7870.9790.9821.018Med. Cond. (Hypertension + Diabetes)0.4540.451—1.0080.314−0.4291.337——Med. Cond. (No Condition)0.5340.331—1.6140.107−0.1151.183——Med. Cond. (Arthritis)1.1440.613—1.8650.062−0.0582.346——Med. Cond. (Cardiac)5.3222.009—2.6490.0081.3839.260——Med. Cond. (Low Back Pain)−0.6550.695—−0.9430.346−2.0170.707——Educ. Level (Doctorate)1.7920.582—3.0820.0020.6522.9330.7531.328Educ. Level (Master’s)1.9220.412—4.665< 0.0011.1142.730——Prof. Role (Clinician)2.2150.465—4.765< 0.0011.3043.1270.3942.539Primary Device (Laptop)0.0210.265—0.0810.936−0.4990.5420.9761.025Primary Device (Smartphone)0.0590.318—0.1840.854−0.5660.683——Primary Device (Tablet)−0.0720.280—−0.2560.798−0.6200.477—— Multiple regression predicting Client-related Burnout (N = 3987) Mediation analysis Because only sleep quantity (Sleep Hours), and weekly work hours emerged as significant predictors in every model of personal-, work-, and client-related burnout, we confined our mediation tests to this duo: weekly work hours (demand) ➔ sleep duration (resource) ➔ burnout. Daily screen time demonstrated a significant direct positive effect on personal burnout (B = 0.327, SE = 0.019, p < .001), but no significant direct effects on work-related (B = -0.005, SE = 0.014, p = .721) and client-related burnout (B = -0.028, SE = 0.026, p = .291). Significant indirect pathways emerged: (1) Sleep hours mediated protective effects across all burnout dimensions (personal: B = -0.109, SE = 0.012, p < .001; work-related: B = -0.059, SE = 0.007, p < .001; client-related: B = -0.042, SE = 0.006, p < .001), consistent with its negative path coefficients (e.g., personal: B = -0.563, SE = 0.013); (2) Weekly work hours mediated risk effects (personal: B = 0.301, SE = 0.013, p < .001; work-related: B = 0.153, SE = 0.009, p < .001; client-related: B = 0.128, SE = 0.016, p < .001), driven by screen time’s strong positive effect on work hours (B = 0.810, SE = 0.017, p < .001). Total effects confirmed net burnout increases (personal: B = 0.519, SE = 0.020, p < .001; work-related: B = 0.098, SE = 0.022, p < .001; client-related: B = 0.059, SE = 0.022, p = .006, with total indirect effects significant for all outcomes (personal: B = 0.193, SE = 0.017). The model explained substantial variance in work-related burnout (R² = 0.745) but minimal variance in client-related burnout (R² = 0.051), (Table 5; Fig. 4). Table 5Mediation analysis: path decomposition with 95% confidence intervals (Δ-method SEs, ML estimator)PathBSEz-value p Path a: Effect of Daily Screen Time on Mediators DST → Sleep Hours0.1930.0218.997< 0.001 DST → Weekly Work Hours0.8100.01746.465< 0.001Path b: Effect of Mediators on Personal Burnout (controlling for DST) Sleep Hours → PersBurn-0.5630.013-45.022< 0.001 Weekly Work Hours → PersBurn0.3720.01425.848< 0.001Path b: Effect of Mediators on Work-Related Burnout (controlling for DST) Sleep Hours → WorkBurn-0.3070.009-35.083< 0.001 Weekly Work Hours → WorkBurn0.1890.01018.728< 0.001Path b: Effect of Mediators on Client-Related Burnout (controlling for DST) Sleep Hours → ClientBurn-0.2200.017-13.026< 0.001 Weekly Work Hours → ClientBurn0.1580.0198.152< 0.001Path c’: Direct Effect of Daily Screen Time on Burnout (controlling for mediators) DST → Personal Burnout0.3270.01916.867< 0.001 DST → Work-Related Burnout-0.0050.014-0.3570.721 DST → Client-Related Burnout-0.0280.026-1.0560.291Indirect Effects (a × b): Specific Indirect Pathways DST → WorkHrsW → PersBurn0.3010.01322.588< 0.001 DST → SleepH → PersBurn-0.1090.012-8.823< 0.001 DST → WorkHrsW → WorkBurn0.1530.00917.371< 0.001 DST → SleepH → WorkBurn-0.0590.007-8.715< 0.001 DST → WorkHrsW → ClientBurn0.1280.0168.030< 0.001 DST → SleepH → ClientBurn-0.0420.006-7.403< 0.001Total Indirect Effects Total Indirect → Personal Burnout0.1930.01711.337< 0.001 Total Indirect → Work-Related Burnout0.1030.0205.075< 0.001 Total Indirect → Client-Related Burnout0.0870.0165.350< 0.001Total Effects (c = c’ + ab) Total Effect → Personal Burnout0.5190.02025.952< 0.001 Total Effect → Work-Related Burnout0.0980.0224.545< 0.001 Total Effect → Client-Related Burnout0.0590.0222.7420.006Proportion Mediated (Indirect / Total) Personal Burnout37.2%--- Work-Related Burnout105.1%*--- Client-Related Burnout147.5%*---DST Daily Screen Time, SleepH Sleep Hours, WorkHrsW Weekly Work Hours, PersBurn Personal Burnout, WorkBurn Work-Related Burnout, ClientBurn Client-Related Burnout. Path a effect of independent variable on mediator, Path b effect of mediator on outcome controlling for IV, Path c direct effect of IV on outcome controlling for mediators, Indirect effecta × b product, Total effect c’ + total indirect effect*Proportion mediated exceeds 100% because the direct effect (c’) is in the opposite direction of the total effect,indicating inconsistent mediation. Models are adjusted for age, BMI, physical activity, workload, and sleep quality as covariates Mediation analysis: path decomposition with 95% confidence intervals (Δ-method SEs, ML estimator) DST Daily Screen Time, SleepH Sleep Hours, WorkHrsW Weekly Work Hours, PersBurn Personal Burnout, WorkBurn Work-Related Burnout, ClientBurn Client-Related Burnout. Path a effect of independent variable on mediator, Path b effect of mediator on outcome controlling for IV, Path c direct effect of IV on outcome controlling for mediators, Indirect effecta × b product, Total effect c’ + total indirect effect *Proportion mediated exceeds 100% because the direct effect (c’) is in the opposite direction of the total effect,indicating inconsistent mediation. Models are adjusted for age, BMI, physical activity, workload, and sleep quality as covariates Structural-equation model linking screen-time, workload, sleep, and burnout domains (Legends: The path diagram image was generated using DeepSeek. All depicted pathways are statistically significant (p < .001). The notations such as b31 and c11 are standard labels used in mediation analysis to denote specific path coefficients, where b paths represent mediator-to-outcome effects, a paths indicate independent-variable-to-mediator effects, and c’ refers to direct effects, all expressed with unstandardized beta (β) coefficients. The paths are labeled according to standard mediation notation: the a paths represent the effect of the independent variable (Daily Screen Time) on the mediators (Weekly Work Hours and Sleep Hours), the b paths represent the effect of those mediators on the outcomes (Personal, Client-Related, and Work-Related Burnout) while controlling for the independent variable, and the c’ path represents the direct effect of Daily Screen Time on Personal Burnout. The exact proportions of variance explained (R²) for each outcome are displayed, and the significant residual covariance (r = .127) between the two mediators is also shown. The measurement model demonstrated excellent fit, χ²(6) = 145.32, p = .08, CFI = 0.97, TLI = 0.96, RMSEA = 0.04 (90% CI [0.00, 0.07]), and SRMR = 0.05) Discussion The key findings of this large sample-based study were that the average screen time exposure per day of the Indian faculties is high, and the participants also reported moderate sleep disturbances. This results also demonstrated that screen time, work hours, and sleep hours are associated with self-reported burnout among the respondents. Furthermore, the daily screen time of faculty demonstrated a significant direct effect on personal burnout, whereas for the work-related and client-related burnout dimensions, the associations were primarily indirect and mediated through weekly work hours and sleep duration. Although specific regional data related to digital screen exposure among university faculty members in India is limited which did not allow regional comparisons. Studies conducted in Europe and North America suggested higher occupational stress and burnout among academic staff, which is often reported to be associated with increasing teaching workloads, research works, and digital demands. For instance, Watts and Robertson (2011) found a strong correlation between institutional pressure and workload intensity and self-reported burnout among university faculty. Similarly, a study in the United Kingdom found that academic members in higher education institutions report high levels of work-related stress and emotional exhaustion, largely due to long working hours and work-life imbalance [7, 29, 30]. These findings are consistent with the present results, which suggest that extended screen time, longer work hours, and reduced sleep may contribute to burnout among university faculty population, though this study did not consider work-life imbalance among the university faculty members which shall be considered by Indian researchers in future studies. Mean sleep duration was 6.1 h/night, below the ≥ 7 h/night recommended for adults; coupled with a mean PSQI of 7.5 (above the > 5 poor-sleep threshold). Experimental and field studies further show that evening light from self-luminous screens suppresses melatonin and delays sleep, supporting the observed links between greater digital exposure and poorer sleep/burnout [31, 32]. Participants reported 5.02 h/day of screen time, which exceeds the ≥ 4 h/day level commonly treated as high exposure in adult cohorts and falls within the range associated with higher cardiometabolic and all-cause mortality risk from prolonged screen/TV viewing [33, 34]. Although the WHO adult sedentary-behaviour guideline does not prescribe a numeric limit, it advises reducing and breaking up sedentary [35, 36]. A review of 12 studies during the time period 2005 to 2020 reporting burnout among University professors found that 5% to 45% of the university faculties experienced burnout [37]. The authors suggest that the new scenario due to the new educational reality has resulted in an increase in work responsibility, digital education, e-teaching, and e-materials, research, publication tasks, technology inefficiency, difficulty in balancing personal and professional life being important predictors of burnout among University faculty. Given the contextual differences, sample variability, and use of different outcome measures has let to inconsistent data, however there is clear evidence that burnout syndrome is higher among university level teachers. The participant characteristics in this study revealed a mean age of 43.9 years and an average of 15.7 years of professional experience, with a mean daily screen time of 5.02 h and an average sleep duration of 6.1 h per night. These findings align with previous research indicating that university faculty members often experience significant screen time and sleep disturbances due to their demanding roles. ​ For example, a study by Slišković and Maslić Seršić [7] reported that university teachers reported high levels of work-related stress and long working hours, which contributed to sleep problems and overall well-being issues. Similarly, Kim and Asbury [38] highlighted the increased screen time and stress among teachers during the COVID-19 pandemic, further exacerbating sleep disturbances and emotional exhaustion. The findings of the present study on screen time and sleep quality are consistent with those of Pitta et al. [18], who reported strong associations between screen time, physical inactivity, and sleep quality among adults. The mean Pittsburgh Sleep Quality Index (PSQI) score of 7.5 in this study indicates poor sleep quality, which is comparable to the findings of Wrede et al. [2], who reported that digital stress significantly impacts sleep quality and physical complaints in home office settings. ​ This suggests that the high screen time and poor sleep quality observed among university faculty members are not isolated phenomena but are part of a broader trend affecting professionals engaged in extensive digital work. Studies, rationalize the association between excessive screen time and poor sleep quality could be due to melatonin suppression secondary to light emitted by the digital screens and radio-frequency electro-magnetic fields [39, 40] . Multiple linear regression analysis revealed significant predictors of personal, work-related, and client-related burnout among university faculty, emphasizing the roles of screen time, sleep hours, and weekly work hours. ​ This aligns with previous studies that have explored similar themes. For example, Tarafdar et al. [41] highlighted the dual nature of technostress, where excessive screen time can lead to techno-distress, contributing to emotional exhaustion and burnout. ​ Similarly, Derks et al. [42] reported that work-related smartphone use negatively impacts psychological detachment and increases exhaustion, supporting the findings that increased screen time is associated with greater emotional exhaustion. Furthermore, Sonnentag and Fritz [43] emphasized the importance of recovery from job stress through adequate sleep, which is consistent with the highlighted paragraph’s finding that more hours of sleep are inversely related to emotional exhaustion. ​ The study by Suh and Lee [44] also corroborates the negative impact of high work hours on job satisfaction and well-being, reinforcing the highlighted paragraph’s assertion that increased weekly work hours are linked to greater personal burnout. ​Although our regression model identified significant negative associations between MET levels and specific burnout domains (particularly work-related and client-related burnout), it is important to interpret these as observational associations. Our findings are broadly in line with the existing literature, including studies by Neczenski et al., which report that physical activity can buffer digital stress. However, given the cross-sectional nature of our study, we cannot infer causality [45]. Thus, the alignment is associative and supports the need for further longitudinal research. The findings of this study supports a mediation path, in which daily screen time of the faculty to be associated with burnout both directly and indirectly. Specifically, the weekly work hours act as a possible mediator, whereby higher screentime is associated with more work-related demands, which in turn contribute to higher levels of burnout. Likewise, sleep duration was observed to functions as a secondary mediating pathway, with increased sleep duration was observed to have protective association with burnout outcomes. The present findings extend the Job Demands-Resources model by empirically supporting that daily screen time influences burnout primarily through increased work demands and altered recovery processes. Particularly, the mediation analyses observed that weekly work hours significantly supported the effect of screen time on burnout, suggesting the role of workload as a central job demand. In parallel, sleep duration emerged as a secondary pathway, indicating that recovery-related processes may partially offset the adverse effects of prolonged screen exposure. Importantly, the direct effect of screen time was limited to personal burnout, suggesting that its influence on other burnout dimensions might predominantly through indirect mechanisms [46]. For example, Turel and Bechara [47] reported that excessive screen time, particularly related to social media use, can impair sleep quality and increase impulsivity, which in turn exacerbates emotional exhaustion and burnout. This aligns with the highlighted paragraph’s assertion that work hours mediate the effect of screen time on personal burnout.​. Moreover, the study by Bennett et al. [47] on videoconference fatigue during the COVID-19 pandemic supports the notion that increased screen time can lead to significant mental fatigue and burnout. Their findings suggest that prolonged screen exposure, especially in a work context, can diminish cognitive resources and contribute to emotional exhaustion, reinforcing the results of the mediation analysis. ​ Additionally, the research [48] conducted on the work-family interface during the pandemic highlights how increased work hours and screen time can disrupt sleep patterns, leading to poorer sleep quality and higher levels of burnout. ​ This is consistent with the highlighted paragraph’s finding that sleep hours mediate the effects of work hours on Personal, Work-Related, and Client-Related Burnout [49]. An unexpected finding was the positive association between daily screen time and sleep duration, contrary to the well-established literature suggesting that screen time typically delays sleep onset and reduces sleep duration. We offer several potential explanations for this counterintuitive finding. First, our measure of screen time captured total daily exposure across all activities (work-related, personal, and recreational), not specifically pre-s bedtime use. Faculty members with higher screen time may compensate by extending their total time in bed to achieve adequate sleep. Second, the cross-sectional design captures a snapshot that may reflect compensatory sleep behaviors rather than causal relationships. Third, the academic context involves irregular work schedules where faculty may sleep longer on certain days to compensate for high-workload periods. Fourth, our sample reported a mean sleep duration of 6.1 h, below the recommended 7–9 h, suggesting that even with increased screen time, participants may be attempting to maximize available sleep time. This finding aligns with recent research suggesting that the relationship between screen time and sleep may be context-dependent and moderated by factors such as timing of use, content type, and individual sleep need. Future longitudinal studies should examine temporal patterns of screen use and sleep to clarify this relationship. Limitations and future directions The study limitations that should be considered while interpreting the findings; the cross-sectional design of this study did not allow causality relationship among variables, and the mediation findings should be interpreted with caution, the relationships are potential pathways rather than definitive causal mechanism. This study used a regression-based mediation approach rather than an advanced counterfactual mediation framework due to assumption limitation related to observational data of this study. Although important covariates were included in the analyses based on theoretical approach, the possibility of residual confounding cannot be excluded. Self-reported screen time, sleep quality, and physical activity levels would have let to recall bias and social desirability bias. Though, self-reported height and weight are commonly used to compute BMI in online survey designs, and this is also regarded as a limitation due to social desirability bias [50]. In absence of a robust theoretical framework or model like the Job Demands-resources (JD-R) model, this study could not adequately explain the relationship between variables like workload, screen time, job crafting, work performance, personal resources, and burnout. Nevertheless, this research included a larger sample of faculties and from different part of the country. Hence, the findings established from this diverse sample would enhance the generalizability of the findings. Further, the use of validated outcome measures and statistical measures used in this study would have countered the possible biases. Future researches should consider conducting a longitudinal design explaining cause-effect relationships of predictor variables, and use of objective outcome measures to assess digital screen time, subject/area of discipline workload stress, work satisfaction, and organizational policies related to academic, administrative, and research related tasks. Conclusion The findings of this large sample study using mediating analysis suggest the existence of burnout among university faculty in India and the possibility of dual pathways through which screen time might influence burnout, emphasizing the role of work demands, and more importantly, sleep duration in shaping the burnout outcomes. Hence, this study contributes to the growing body of literature on digital device usage and its implications for professional and personal life, highlighting the need for targeted interventions to promote healthier work environments in academia. Advocacy of screen time management, recommended sleep time, and workload management should be considered in policies to promote wellbeing of university faculty. Supplementary Information Supplementary Material 1. Acknowledgements The authors would like to thank the Researcher Supporting Project Number (RSPD2025R729), King Saud University, Riyadh, Saudi Arabia. Authors’ contributions M.S. A.P. B.J. F.K. M.K. A.C. B.M.W. Z.H. J.S.T. V.T.S. and A.I. proposed the study concept and design. M.S. A.C. B.J. and A.P. J.S. R.K.V. F.K planned the methodology. A.P. J.S. R.K.V. F.K. M.K. B.M.W. and A.I. collected data. M.S. A.C. R.K.V. Z.H. A.I. and F.K. contributed to the data analysis and interpretation M.S. A.C. A.P. J.S. R.K.V. and F.K. prepared the manuscript’s initial draft. M.S. A.C. B.J. F.K. M.K. B.M.W. Z.H. J.S.T. V.T.S. and A.I. critically reviewed and edited the manuscript for intellectual content. All authors have read, understood, reviewed, and approved the manuscript’s final version to be submitted or published and take responsibility for the intellectual content of the same manuscript. Funding The study was funded by the Researcher Supporting Project Number (RSPD2025R729), King Saud University, Riyadh, Saudi Arabia. Data availability All data generated or analyzed during this study is available upon a reasonable request from the corresponding author. Declarations Ethics approval and consent to participate Ethical approval for this study was conducted by the School Ethics Committee, Galgotias University (SEC/001/24). The study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki. Additionally, the study was prospectively registered with the Clinical Trial Registry India; assigned identifier: CTRI/2024/02/063155, dated 26/02/2024. All participants were informed about the purpose of the study, assured of confidentiality, and provided written consent prior to participation. Participation was voluntary, and respondents could withdraw at any time without consequence. Consent for publication Not applicable. Competing interests The authors declare no competing interests. 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. Data Availability Statement All data generated or analyzed during this study is available upon a reasonable request from the corresponding author.