Associations of mindful eating and physical activity with mental workload and sleep quality in medical interns Authors: Mina Babashahi, Niloofar Rahgoshay, Najme Setoude, Hadi Daneshmandi Published: 2026 ABSTRACT This study aimed to investigate mindful eating practices and physical activity levels among medical interns and to examine their associations with mental workload and sleep quality. This cross-sectional study was conducted in 2025 among 165 medical interns at Shiraz University of Medical Sciences, Iran. Participants were recruited through convenience sampling. Data were collected using the demographic characteristics questionnaire, Persian Version of the Mindful Eating Questionnaire (P-MEQ), Persian version of the International Physical Activity Questionnaire (P-IPAQ), Persian version of the National Aeronautics and Space Administration Task Load Index (P-NASA-TLX), and Persian version of the Pittsburgh Sleep Quality Index (P-PSQI). Multivariate linear regression analyses were conducted to examine the independent associations of mindful eating and physical activity with mental workload and sleep quality outcomes. Outcomes were evaluated using SPSS v17 with statistical tests (p < 0.05). The mean total mental workload score was 68.48 ± 13.55 out of 100, and the mean PSQI score was 14.83 ± 3.24, indicating poor sleep quality. Multivariate regression analyses revealed that mindful eating awareness, distraction, total mindful eating score, and physical activity were associated with total mental workload (R2 = 0.40), with standardized beta coefficients ranging from − 0.235 to − 0.350 (all p < 0.001). Similarly, mindful eating awareness, distraction, and physical activity were associated with total sleep quality (R2 = 0.30), with beta coefficients of − 0.245, − 0.235, and −0.235, respectively (all p < 0.015). Mindful eating practices and physical activity are associated with lower mental workload and better sleep quality among medical interns. Given the cross-sectional design, causal inferences cannot be made. These findings highlight that these modifiable lifestyle behaviors may be relevant targets for further research aimed at supporting trainee well-being, though comprehensive approaches addressing both individual and organizational factors are necessary for optimal outcomes. Introduction Eating behavior encompasses nutrition-related factors, eating patterns, food choice, and eating disorders1,2. Mindful eating applies mindfulness practices to maintain complete attention to eating experiences, cravings, and physical signals during food consumption, and has been used in treating eating disorders, depression, and anxiety3–5. Regular physical activity enhances longevity, reduces risk of cardiovascular disease, diabetes, and certain cancers, while supporting healthy Body Mass Index (BMI) maintenance6. The internship year—the final stage of medical school marking transition from student to physician—is highly stressful due to long shifts, heavy workloads, clinical decisions, and exposure to disease and death7–9. More than half of medical students suffer from psychological distress, resulting in diminished quality of life and elevated attrition rates10. This combination of pressures places interns at heightened risk for mental health challenges, including eating disorder11,12. Although medical students are presumed to understand health-promoting behaviors better than other students, empirical evidence confirming translation of this knowledge into practice remains absent13,14. Rotating shift schedules and time limitations serve as primary impediments to regular physical activity among residents15. To provide a theoretical basis for this study, we draw upon self-regulation theory16,17. Self-regulation is the capacity to manage thoughts, emotions, and behaviors in alignment with goals and demands. Within this framework, mindful eating serves as a self-regulatory strategy that enhances awareness of hunger/satiety cues and reduces stress-driven eating18. Regular physical activity reduces stress reactivity and stabilizes circadian rhythms19. Conversely, high mental workload depletes self-regulatory resources, making mindful eating and physical activity harder to maintain20. Poor sleep quality further impairs executive function and metabolic regulation21,22. Recent evidence supports these links: perceived stress influences eating behaviors, which mediate obesity-related measurements23, and emotional eating is associated with BMI and physical activity levels24. Thus, integrating mindful eating and physical activity within a self-regulation framework provides a coherent model for understanding their associations with mental workload and sleep quality in medical interns. Mental workload refers to the cognitive capacity used for understanding, predicting, and deciding during task completion25,26. Increased mental workload causes burnout and reduces care quality by raising patient safety risks27,28. From a self-regulation perspective, elevated mental workload consumes attentional resources, reducing capacity for mindful eating and intentional physical activity29–32. Poor sleep quality impairs cognition, performance, and physical health, causing stress, depression, and reduced quality of life33–36. Evidence suggests bidirectional relationships between sleep, eating behaviors, and physical activity: poor sleep disrupts appetite hormones (ghrelin/leptin), while physical activity and mindful eating improve sleep via stress reduction and circadian alignment21,22,37–39. Despite these theoretical and empirical links, the interrelationships among mindful eating, physical activity, mental workload, and sleep quality remain underexplored in medical interns. Based on our self-regulation framework, we hypothesized that higher levels of mindful eating and physical activity would be associated with lower mental workload and better sleep quality. Therefore, this study aimed to examine whether mindful eating and physical activity are independently associated with mental workload and sleep quality in medical interns. Materials and methods Study design, setting, and population This cross-sectional study was conducted in 2025 among medical interns at Shiraz University of Medical Sciences, Iran. Participants were recruited through convenience sampling from clinical medical internship students at affiliated hospitals. Sample size was determined using Power Analysis and Sample Size (PASS) software, based on a correlation coefficient of r = 0.269 from a previous study examining the relationship between mindful eating and job stress40. With α = 0.05, power (1-β) = 90%, and the formula n ≥ [(z₁₋α/₂ + z₁₋β) / (½logₑ (1 + r) / (1-r))]2, a minimum of 141 participants was required. A total of 165 medical interns completed the study. Inclusion criteria were: (1) Iranian students in clinical medical internship; (2) primary place of study at Shiraz University of Medical Sciences; (3) willingness to participate; (4) absence of specific diseases or physical/mental disorders; (5) not following a specific diet; (6) female students not in their menstrual period; and (7) not being professional athletes. Exclusion criteria were: (1) presence of any specific disease or physical/mental disorder; (2) following a specific diet; (3) female students in their menstrual period; (4) being a professional athlete; and (5) incomplete questionnaire responses. The study protocol was approved by the Ethics Committee of Shiraz University of Medical Sciences (Ethics Code: IR.SUMS.SCHEANUT.REC.1403.105). Data gathering tools Demographic characteristics questionnaire: A questionnaire was developed to collect demographic information, including sex, age, height, weight, daily working hours, and marital status.Persian version of the mindful eating questionnaire (P-MEQ): Mindful eating was assessed using the Mindful Eating Questionnaire (MEQ) developed by Framson et al.41 in 2009, which consists of 28 self-report items measuring five dimensions: disinhibition, external cues, awareness, emotional response, and distraction. Items are scored on a 4-point Likert scale ranging from 1 (never/rarely) to 4 (usually/always), with higher scores indicating better mindful eating. The validity and reliability of this questionnaire in the Iranian population were examined by Rezaei and Jahanbin in 202042; the test–retest reliability for the P-MEQ total score was 0.59, and subscale reliabilities ranged from 0.58 to 0.79.Persian version of the international physical activity questionnaire (P-IPAQ): Physical activity level was assessed using the International Physical Activity Questionnaire (IPAQ), an international standard instrument for measuring physical activity developed in 1998 by the World Health Organization (WHO) and the Centers for Disease Control and Prevention for the age group of 15 to 69 years43. The IPAQ is a standardized 7-item questionnaire for measuring physical activity levels. In this questionnaire, the first and second questions assess the number of days and duration of vigorous physical activity, the third and fourth questions assess the number of days and duration of moderate physical activity, the fifth and sixth questions assess the number of days and duration of walking, and the seventh question assesses the amount of sitting time during the last 7 days. Physical activity levels were calculated as Metabolic Equivalent of Task (MET)-minutes per week and classified into three categories (low, moderate, and high physical activity) according to the IPAQ scoring protocol. The validity and reliability of the Persian version of this questionnaire were examined by Gholamnia Shirvani et al. in 201544, and the intra-class correlation coefficient was 0.85.Persian version of the national aeronautics and space administration task load index (P-NASA-TLX): Mental workload was assessed using the Persian version of the National Aeronautics and Space Administration Task Load Index (P-NASA-TLX), a well-established multidimensional tool for evaluating subjective mental workload developed by Hart and Staveland in 198845. This method provides an overall mental workload score based on six subscales: mental demand, physical demand, temporal demand, effort, performance, and frustration level, each rated on a scale from 0 to 100. The raw NASA-TLX score was calculated as the unweighted average of the six subscale scores. Numerous studies have confirmed the reliability and validity of this method for mental workload assessment. Mohammadi et al.46 evaluated the validity and reliability of this questionnaire for Iranian nurses. The Cronbach’s alpha coefficient for the NASA-TLX questionnaire was 0.847. Additionally, based on the test–retest method, the Spearman correlation coefficient was 0.746 with a significance level of less than 0.01.Persian version of the Pittsburgh sleep quality index (P-PSQI): Sleep quality was assessed using the Pittsburgh Sleep Quality Index (PSQI), a widely used self-report questionnaire that assesses sleep quality and disturbances over a 1-month time interval. The PSQI was developed by Buysse et al.47 at the University of Pittsburgh in 1989. The validity and reliability of this questionnaire in the Iranian population were previously examined by Farrahi Moghaddam et al. in 201248. The questionnaire originally contains 9 items; however, since question 5 includes 10 sub-items, the total questionnaire comprises 19 items scored on a 4-point Likert scale ranging from 0 to 3. The PSQI consists of seven component scores: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleeping medication, and daytime dysfunction. In all components, a score of 0 indicates no problems and 3 indicates severe problems. The scores of these seven components are summed to yield a global PSQI score ranging from 0 to 21, where higher global scores indicate poorer sleep quality. A global PSQI score greater than 5 indicates poor sleep quality. Demographic characteristics questionnaire: A questionnaire was developed to collect demographic information, including sex, age, height, weight, daily working hours, and marital status. Demographic characteristics questionnaire: A questionnaire was developed to collect demographic information, including sex, age, height, weight, daily working hours, and marital status. Persian version of the mindful eating questionnaire (P-MEQ): Mindful eating was assessed using the Mindful Eating Questionnaire (MEQ) developed by Framson et al.41 in 2009, which consists of 28 self-report items measuring five dimensions: disinhibition, external cues, awareness, emotional response, and distraction. Items are scored on a 4-point Likert scale ranging from 1 (never/rarely) to 4 (usually/always), with higher scores indicating better mindful eating. The validity and reliability of this questionnaire in the Iranian population were examined by Rezaei and Jahanbin in 202042; the test–retest reliability for the P-MEQ total score was 0.59, and subscale reliabilities ranged from 0.58 to 0.79. Persian version of the mindful eating questionnaire (P-MEQ): Mindful eating was assessed using the Mindful Eating Questionnaire (MEQ) developed by Framson et al.41 in 2009, which consists of 28 self-report items measuring five dimensions: disinhibition, external cues, awareness, emotional response, and distraction. Items are scored on a 4-point Likert scale ranging from 1 (never/rarely) to 4 (usually/always), with higher scores indicating better mindful eating. The validity and reliability of this questionnaire in the Iranian population were examined by Rezaei and Jahanbin in 202042; the test–retest reliability for the P-MEQ total score was 0.59, and subscale reliabilities ranged from 0.58 to 0.79. Persian version of the international physical activity questionnaire (P-IPAQ): Physical activity level was assessed using the International Physical Activity Questionnaire (IPAQ), an international standard instrument for measuring physical activity developed in 1998 by the World Health Organization (WHO) and the Centers for Disease Control and Prevention for the age group of 15 to 69 years43. The IPAQ is a standardized 7-item questionnaire for measuring physical activity levels. In this questionnaire, the first and second questions assess the number of days and duration of vigorous physical activity, the third and fourth questions assess the number of days and duration of moderate physical activity, the fifth and sixth questions assess the number of days and duration of walking, and the seventh question assesses the amount of sitting time during the last 7 days. Physical activity levels were calculated as Metabolic Equivalent of Task (MET)-minutes per week and classified into three categories (low, moderate, and high physical activity) according to the IPAQ scoring protocol. The validity and reliability of the Persian version of this questionnaire were examined by Gholamnia Shirvani et al. in 201544, and the intra-class correlation coefficient was 0.85. Persian version of the international physical activity questionnaire (P-IPAQ): Physical activity level was assessed using the International Physical Activity Questionnaire (IPAQ), an international standard instrument for measuring physical activity developed in 1998 by the World Health Organization (WHO) and the Centers for Disease Control and Prevention for the age group of 15 to 69 years43. The IPAQ is a standardized 7-item questionnaire for measuring physical activity levels. In this questionnaire, the first and second questions assess the number of days and duration of vigorous physical activity, the third and fourth questions assess the number of days and duration of moderate physical activity, the fifth and sixth questions assess the number of days and duration of walking, and the seventh question assesses the amount of sitting time during the last 7 days. Physical activity levels were calculated as Metabolic Equivalent of Task (MET)-minutes per week and classified into three categories (low, moderate, and high physical activity) according to the IPAQ scoring protocol. The validity and reliability of the Persian version of this questionnaire were examined by Gholamnia Shirvani et al. in 201544, and the intra-class correlation coefficient was 0.85. Persian version of the national aeronautics and space administration task load index (P-NASA-TLX): Mental workload was assessed using the Persian version of the National Aeronautics and Space Administration Task Load Index (P-NASA-TLX), a well-established multidimensional tool for evaluating subjective mental workload developed by Hart and Staveland in 198845. This method provides an overall mental workload score based on six subscales: mental demand, physical demand, temporal demand, effort, performance, and frustration level, each rated on a scale from 0 to 100. The raw NASA-TLX score was calculated as the unweighted average of the six subscale scores. Numerous studies have confirmed the reliability and validity of this method for mental workload assessment. Mohammadi et al.46 evaluated the validity and reliability of this questionnaire for Iranian nurses. The Cronbach’s alpha coefficient for the NASA-TLX questionnaire was 0.847. Additionally, based on the test–retest method, the Spearman correlation coefficient was 0.746 with a significance level of less than 0.01. Persian version of the national aeronautics and space administration task load index (P-NASA-TLX): Mental workload was assessed using the Persian version of the National Aeronautics and Space Administration Task Load Index (P-NASA-TLX), a well-established multidimensional tool for evaluating subjective mental workload developed by Hart and Staveland in 198845. This method provides an overall mental workload score based on six subscales: mental demand, physical demand, temporal demand, effort, performance, and frustration level, each rated on a scale from 0 to 100. The raw NASA-TLX score was calculated as the unweighted average of the six subscale scores. Numerous studies have confirmed the reliability and validity of this method for mental workload assessment. Mohammadi et al.46 evaluated the validity and reliability of this questionnaire for Iranian nurses. The Cronbach’s alpha coefficient for the NASA-TLX questionnaire was 0.847. Additionally, based on the test–retest method, the Spearman correlation coefficient was 0.746 with a significance level of less than 0.01. Persian version of the Pittsburgh sleep quality index (P-PSQI): Sleep quality was assessed using the Pittsburgh Sleep Quality Index (PSQI), a widely used self-report questionnaire that assesses sleep quality and disturbances over a 1-month time interval. The PSQI was developed by Buysse et al.47 at the University of Pittsburgh in 1989. The validity and reliability of this questionnaire in the Iranian population were previously examined by Farrahi Moghaddam et al. in 201248. The questionnaire originally contains 9 items; however, since question 5 includes 10 sub-items, the total questionnaire comprises 19 items scored on a 4-point Likert scale ranging from 0 to 3. The PSQI consists of seven component scores: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleeping medication, and daytime dysfunction. In all components, a score of 0 indicates no problems and 3 indicates severe problems. The scores of these seven components are summed to yield a global PSQI score ranging from 0 to 21, where higher global scores indicate poorer sleep quality. A global PSQI score greater than 5 indicates poor sleep quality. Persian version of the Pittsburgh sleep quality index (P-PSQI): Sleep quality was assessed using the Pittsburgh Sleep Quality Index (PSQI), a widely used self-report questionnaire that assesses sleep quality and disturbances over a 1-month time interval. The PSQI was developed by Buysse et al.47 at the University of Pittsburgh in 1989. The validity and reliability of this questionnaire in the Iranian population were previously examined by Farrahi Moghaddam et al. in 201248. The questionnaire originally contains 9 items; however, since question 5 includes 10 sub-items, the total questionnaire comprises 19 items scored on a 4-point Likert scale ranging from 0 to 3. The PSQI consists of seven component scores: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleeping medication, and daytime dysfunction. In all components, a score of 0 indicates no problems and 3 indicates severe problems. The scores of these seven components are summed to yield a global PSQI score ranging from 0 to 21, where higher global scores indicate poorer sleep quality. A global PSQI score greater than 5 indicates poor sleep quality. Implementation of the study Before completing the questionnaires, participants were provided with detailed explanations about the research objectives and instructions on how to complete the questionnaires. Participants were informed about the confidentiality of their information and their right to withdraw from the study at any time without consequences. After providing written informed consent, participants completed a demographic characteristics questionnaire, followed by the P-MEQ, P-IPAQ, P-NASA-TLX, and P-PSQI. All questionnaires were self-administered and completed in a single session under the supervision of trained research staff who were available to answer any questions. Statistical analysis Data analysis was performed using the Statistical Package for the Social Sciences (SPSS) version 17. Descriptive statistics, including means, standard deviations, frequencies, and percentages, were calculated for all demographic and study variables. Multivariate linear regression analyses were conducted to examine the independent associations of mindful eating dimensions and physical activity (measured in MET-minutes per week) on mental workload and sleep quality outcomes, while controlling for potential confounders including gender, marital status, age, BMI, and daily working hours. Separate regression models were constructed for total scores and subscales of mental workload (NASA-TLX) and sleep quality (PSQI). Model assumptions including linearity, homoscedasticity, independence of residuals, and absence of multicollinearity were verified. In analyses, Variance Inflation Factor (VIF) values were examined (all VIF < 5), indicating no severe multicollinearity. For all statistical tests, a P-value < 0.05 was considered indicative of an association worth noting, though given the exploratory nature and multiple comparisons, these p-values are interpreted as descriptive rather than confirmatory. Results Table 1 presents the demographic characteristics of the medical interns. The mean age was 25.83 years, the mean BMI was 23.66 kg m-2, and the mean number of working hours per day was 6.43. Among the participants, 53.3% were male and 92.7% were single.Table 1Demographic characteristics of the medical interns (n = 165).Quantitative variableMean ± SDAge (year)25.83 ± 1.65Height (cm)172.41 ± 9.37Weight (kg)70.88 ± 4.05BMI (kg.m−2)*23.66 ± 3.36Working hours per day6.43 ± 2.10Qualitative variableNo. (%)GenderMale88 (53.3%)Female77 (46.7%)Marital statusSingle153 (92.7%)Married12 (7.3%)*Body Mass Index. Table 1Demographic characteristics of the medical interns (n = 165).Quantitative variableMean ± SDAge (year)25.83 ± 1.65Height (cm)172.41 ± 9.37Weight (kg)70.88 ± 4.05BMI (kg.m−2)*23.66 ± 3.36Working hours per day6.43 ± 2.10Qualitative variableNo. (%)GenderMale88 (53.3%)Female77 (46.7%)Marital statusSingle153 (92.7%)Married12 (7.3%)*Body Mass Index. Demographic characteristics of the medical interns (n = 165). *Body Mass Index. Table 2 presents the mean, standard deviation, minimum, and maximum scores for total mental workload and its subscales among participants. The mean total mental workload score was 68.48 ± 13.55 out of 100, indicating a moderately high level of perceived mental workload among medical interns. Among the subscales, mental demand showed the highest mean score (80.48 ± 18.68), followed by temporal demand (79.72 ± 19.33) and effort (77.21 ± 18.55), while performance showed the lowest mean score (35.78 ± 21.55).Table 2Mean, standard deviation, minimum, and maximum scores for total mental workload and its subscales among participants (n = 165).Mental workload subscaleMean ± SDaMinb–MaxcMental demand80.48 ± 18.680–100Physical demand74.15 ± 22.170–100Temporal demand79.72 ± 19.330–100Performance35.78 ± 21.550–100Effort77.21 ± 18.550–100Frustration63.54 ± 28.270–100Total68.48 ± 13.550–100aStandard deviation; bMinimum; cMaximum. Table 2Mean, standard deviation, minimum, and maximum scores for total mental workload and its subscales among participants (n = 165).Mental workload subscaleMean ± SDaMinb–MaxcMental demand80.48 ± 18.680–100Physical demand74.15 ± 22.170–100Temporal demand79.72 ± 19.330–100Performance35.78 ± 21.550–100Effort77.21 ± 18.550–100Frustration63.54 ± 28.270–100Total68.48 ± 13.550–100aStandard deviation; bMinimum; cMaximum. Mean, standard deviation, minimum, and maximum scores for total mental workload and its subscales among participants (n = 165). aStandard deviation; bMinimum; cMaximum. A higher score indicates a higher mental workload. Table 3 displays the multivariate linear regression results for the associations of mindful eating and physical activity on mental workload and subscales, adjusted for gender, marital status, age, BMI, and working hours per day. The overall model explained 40% of the variance in total mental workload (R2 = 0.40, Adjusted R2 = 0.385). Variables associated with total mental workload included mindful eating awareness (β = − 0.320, p < 0.001), distraction (β = − 0.235, p = 0.001), total mindful eating score (β = − 0.350, p < 0.001), and physical activity (β = − 0.300, p < 0.001 per 1000 MET-minutes).Table 3Multivariate linear regression analyses examining the effects of mindful eating and physical activity on mental workload and subscales (n = 165).VariableBaSEbBetaTcp-value95% CIdMental demand Mindful eating awareness Mindful eating disinhibition Physical activity (per 1000 METse)-6.50-9.35-3.052.102.580.74-0.271-0.320-0.290-3.096-3.632-4.1200.002 < 0.001 < 0.001-10.64 to -2.36-14.44 to -4.26-4.51 to -1.59Physical demand Mindful eating distraction Mindful eating total Physical activity (per 1000 METs)-4.85-7.35-2.101.782.620.70-0.180-0.235-0.210-2.732-2.801-3.0000.0070.0060.003-8.36 to -1.34-12.52 to -2.18-3.48 to -0.72Temporal demand–––––––Performance Mindful eating disinhibition Mindful eating total Physical activity (per 1000 METs)-12.00-11.00-4.003.313.041.14-0.340-0.310-0.280-3.629-3.615-3.500 < 0.001 < 0.001 < 0.001-18.54 to -5.46-17.01 to -4.99-6.25 to -1.75Effort Mindful eating disinhibition-9.102.48-0.295-3.667 < 0.001-14.00 to -4.20Frustration Mindful eating emotional response Mindful eating total-6.50-8.552.192.68-0.245-0.280-2.961-3.1870.0030.002-10.83 to -2.17-13.85 to -3.25Total mental workload Mindful eating awareness Mindful eating distraction Mindful eating total Physical activity (per 1000 METs)-15.05-11.00-18.05-4.753.843.394.411.11-0.320-0.235-0.350-0.300-3.923-3.241-4.097-4.286 < 0.0010.001 < 0.001 < 0.001-22.63 to -7.47-17.70 to -4.30-26.77 to -9.33-6.94 to -2.56aUnstandardized beta coefficient; bStandard error; cT-distribution; dConfidence interval; eMetabolic Equivalent of Task; fVariance Inflation Factor. Table 3Multivariate linear regression analyses examining the effects of mindful eating and physical activity on mental workload and subscales (n = 165).VariableBaSEbBetaTcp-value95% CIdMental demand Mindful eating awareness Mindful eating disinhibition Physical activity (per 1000 METse)-6.50-9.35-3.052.102.580.74-0.271-0.320-0.290-3.096-3.632-4.1200.002 < 0.001 < 0.001-10.64 to -2.36-14.44 to -4.26-4.51 to -1.59Physical demand Mindful eating distraction Mindful eating total Physical activity (per 1000 METs)-4.85-7.35-2.101.782.620.70-0.180-0.235-0.210-2.732-2.801-3.0000.0070.0060.003-8.36 to -1.34-12.52 to -2.18-3.48 to -0.72Temporal demand–––––––Performance Mindful eating disinhibition Mindful eating total Physical activity (per 1000 METs)-12.00-11.00-4.003.313.041.14-0.340-0.310-0.280-3.629-3.615-3.500 < 0.001 < 0.001 < 0.001-18.54 to -5.46-17.01 to -4.99-6.25 to -1.75Effort Mindful eating disinhibition-9.102.48-0.295-3.667 < 0.001-14.00 to -4.20Frustration Mindful eating emotional response Mindful eating total-6.50-8.552.192.68-0.245-0.280-2.961-3.1870.0030.002-10.83 to -2.17-13.85 to -3.25Total mental workload Mindful eating awareness Mindful eating distraction Mindful eating total Physical activity (per 1000 METs)-15.05-11.00-18.05-4.753.843.394.411.11-0.320-0.235-0.350-0.300-3.923-3.241-4.097-4.286 < 0.0010.001 < 0.001 < 0.001-22.63 to -7.47-17.70 to -4.30-26.77 to -9.33-6.94 to -2.56aUnstandardized beta coefficient; bStandard error; cT-distribution; dConfidence interval; eMetabolic Equivalent of Task; fVariance Inflation Factor. Multivariate linear regression analyses examining the effects of mindful eating and physical activity on mental workload and subscales (n = 165). Mindful eating awareness Mindful eating disinhibition Physical activity (per 1000 METse) -6.50 -9.35 -3.05 2.10 2.58 0.74 -0.271 -0.320 -0.290 -3.096 -3.632 -4.120 0.002 < 0.001 < 0.001 -10.64 to -2.36 -14.44 to -4.26 -4.51 to -1.59 Mindful eating distraction Mindful eating total Physical activity (per 1000 METs) -4.85 -7.35 -2.10 1.78 2.62 0.70 -0.180 -0.235 -0.210 -2.732 -2.801 -3.000 0.007 0.006 0.003 -8.36 to -1.34 -12.52 to -2.18 -3.48 to -0.72 Mindful eating disinhibition Mindful eating total Physical activity (per 1000 METs) -12.00 -11.00 -4.00 3.31 3.04 1.14 -0.340 -0.310 -0.280 -3.629 -3.615 -3.500 < 0.001 < 0.001 < 0.001 -18.54 to -5.46 -17.01 to -4.99 -6.25 to -1.75 Mindful eating emotional response Mindful eating total -6.50 -8.55 2.19 2.68 -0.245 -0.280 -2.961 -3.187 0.003 0.002 -10.83 to -2.17 -13.85 to -3.25 Mindful eating awareness Mindful eating distraction Mindful eating total Physical activity (per 1000 METs) -15.05 -11.00 -18.05 -4.75 3.84 3.39 4.41 1.11 -0.320 -0.235 -0.350 -0.300 -3.923 -3.241 -4.097 -4.286 < 0.001 0.001 < 0.001 < 0.001 -22.63 to -7.47 -17.70 to -4.30 -26.77 to -9.33 -6.94 to -2.56 aUnstandardized beta coefficient; bStandard error; cT-distribution; dConfidence interval; eMetabolic Equivalent of Task; fVariance Inflation Factor. Model adjusted for gender, marital status, age, BMI, and working hours per day; R2 = 0.4, Adjusted R2 = 0.385. Table 4 presents the mean, standard deviation, minimum, and maximum scores for total sleep quality and its subscales among participants. The mean global PSQI score was 14.83 ± 3.24, which is substantially above the clinical cut-off of 5, indicating poor sleep quality among the majority of medical interns. Among the seven components, duration of sleep showed the highest mean score (2.91 ± 0.65), followed by day dysfunction due to sleepiness (2.38 ± 0.68) and sleep latency (2.29 ± 0.85), while need for sleep medication showed the lowest mean score (1.39 ± 0.37).Table 4Mean, standard deviation, minimum, and maximum scores for total sleep quality and its subscales among participants (n = 165).Sleep quality subscaleMean ± SDaMinb–MaxcSleep latency2.29 ± 0.850–3Duration of sleep2.91 ± 0.650–3Sleep efficiency1.44 ± 0.440–3Sleep disturbance2.16 ± 0.450–3Need meds to sleep1.39 ± 0.370–3Day dysfunction due to sleepiness2.38 ± 0.680–3Subjective sleep quality2.24 ± 0.460–3Total14.83 ± 3.240–21aStandard deviation; bMinimum; cMaximumA higher score indicates poor sleep quality Table 4Mean, standard deviation, minimum, and maximum scores for total sleep quality and its subscales among participants (n = 165).Sleep quality subscaleMean ± SDaMinb–MaxcSleep latency2.29 ± 0.850–3Duration of sleep2.91 ± 0.650–3Sleep efficiency1.44 ± 0.440–3Sleep disturbance2.16 ± 0.450–3Need meds to sleep1.39 ± 0.370–3Day dysfunction due to sleepiness2.38 ± 0.680–3Subjective sleep quality2.24 ± 0.460–3Total14.83 ± 3.240–21aStandard deviation; bMinimum; cMaximumA higher score indicates poor sleep quality Mean, standard deviation, minimum, and maximum scores for total sleep quality and its subscales among participants (n = 165). aStandard deviation; bMinimum; cMaximum A higher score indicates poor sleep quality Table 5 displays the multivariate linear regression results for the associations of mindful eating and physical activity on sleep quality and subscales, adjusted for gender, marital status, age, BMI, and working hours per day. The overall model explained 30% of the variance in total sleep quality (R2 = 0.30, Adjusted R2 = 0.26). Variables associated with total sleep quality included mindful eating awareness (β = − 0.245, p = 0.010), distraction (β = − 0.235, p = 0.014), and physical activity (β = − 0.235, p = 0.015 per 1000 MET-minutes).Table 5Multivariate linear regression analyses examining the effects of mindful eating, and physical activity on sleep quality and subscales (n = 165).VariableBaSEbBetaTcp-value95% CIdSleep latency Mindful eating emotional response Mindful eating external cues Physical activity (per 1000 METse)-0.45-0.35-0.200.200.150.08-0.180-0.185-0.200-2.235-2.308-2.5000.0280.0240.015-0.84 to -0.06-0.65 to -0.05-0.36 to -0.04Duration of sleep Mindful eating disinhibition Physical activity (per 1000 METs)-0.28-0.150.120.06-0.195-0.210-2.400-2.5000.0190.015-0.51 to -0.05-0.27 to -0.03Sleep efficiency Mindful eating external cues Physical activity (per 1000 METs)-0.40-0.220.190.10-0.190-0.200-2.079-2.2500.0400.027-0.78 to -0.02-0.42 to -0.02Sleep disturbance Mindful eating awareness Mindful eating distraction Physical activity (per 1000 METs)-0.65-0.42-0.180.280.180.07-0.210-0.200-0.220-2.333-2.308-2.4000.0220.0240.019-1.20 to -0.10-0.78 to -0.06-0.32 to -0.04Need meds to sleep Physical activity (per 1000 METs)-0.300.12-0.230-2.5000.015-0.54 to -0.06Day dysfunction due to sleepiness Mindful eating emotional response Mindful eating disinhibition-0.52-0.700.230.29-0.205-0.225-2.245-2.4000.0270.019-0.97 to -0.07-1.27 to -0.13Subjective sleep quality Mindful eating emotional response Mindful eating disinhibition-0.82-0.990.340.40-0.220-0.235-2.414-2.4710.0180.016-1.49 to -0.15-1.78 to -0.20Total sleep quality Mindful eating awareness Mindful eating distraction Physical activity (per 1000 METs)-3.00-2.20-0.801.130.870.32-0.245-0.235-0.235-2.656-2.531-2.5000.0100.0140.015-5.23 to -0.77-3.91 to -0.49-1.43 to -0.17aUnstandardized beta coefficient; bStandard error; cT-distribution; dConfidence interval; eMetabolic Equivalent of Task.Model adjusted for gender, marital status, age, BMI, and working hours per day; R2 = 0.30, Adjusted R2 = 0.26. Table 5Multivariate linear regression analyses examining the effects of mindful eating, and physical activity on sleep quality and subscales (n = 165).VariableBaSEbBetaTcp-value95% CIdSleep latency Mindful eating emotional response Mindful eating external cues Physical activity (per 1000 METse)-0.45-0.35-0.200.200.150.08-0.180-0.185-0.200-2.235-2.308-2.5000.0280.0240.015-0.84 to -0.06-0.65 to -0.05-0.36 to -0.04Duration of sleep Mindful eating disinhibition Physical activity (per 1000 METs)-0.28-0.150.120.06-0.195-0.210-2.400-2.5000.0190.015-0.51 to -0.05-0.27 to -0.03Sleep efficiency Mindful eating external cues Physical activity (per 1000 METs)-0.40-0.220.190.10-0.190-0.200-2.079-2.2500.0400.027-0.78 to -0.02-0.42 to -0.02Sleep disturbance Mindful eating awareness Mindful eating distraction Physical activity (per 1000 METs)-0.65-0.42-0.180.280.180.07-0.210-0.200-0.220-2.333-2.308-2.4000.0220.0240.019-1.20 to -0.10-0.78 to -0.06-0.32 to -0.04Need meds to sleep Physical activity (per 1000 METs)-0.300.12-0.230-2.5000.015-0.54 to -0.06Day dysfunction due to sleepiness Mindful eating emotional response Mindful eating disinhibition-0.52-0.700.230.29-0.205-0.225-2.245-2.4000.0270.019-0.97 to -0.07-1.27 to -0.13Subjective sleep quality Mindful eating emotional response Mindful eating disinhibition-0.82-0.990.340.40-0.220-0.235-2.414-2.4710.0180.016-1.49 to -0.15-1.78 to -0.20Total sleep quality Mindful eating awareness Mindful eating distraction Physical activity (per 1000 METs)-3.00-2.20-0.801.130.870.32-0.245-0.235-0.235-2.656-2.531-2.5000.0100.0140.015-5.23 to -0.77-3.91 to -0.49-1.43 to -0.17aUnstandardized beta coefficient; bStandard error; cT-distribution; dConfidence interval; eMetabolic Equivalent of Task.Model adjusted for gender, marital status, age, BMI, and working hours per day; R2 = 0.30, Adjusted R2 = 0.26. Multivariate linear regression analyses examining the effects of mindful eating, and physical activity on sleep quality and subscales (n = 165). Mindful eating emotional response Mindful eating external cues Physical activity (per 1000 METse) -0.45 -0.35 -0.20 0.20 0.15 0.08 -0.180 -0.185 -0.200 -2.235 -2.308 -2.500 0.028 0.024 0.015 -0.84 to -0.06 -0.65 to -0.05 -0.36 to -0.04 Mindful eating disinhibition Physical activity (per 1000 METs) -0.28 -0.15 0.12 0.06 -0.195 -0.210 -2.400 -2.500 0.019 0.015 -0.51 to -0.05 -0.27 to -0.03 Mindful eating external cues Physical activity (per 1000 METs) -0.40 -0.22 0.19 0.10 -0.190 -0.200 -2.079 -2.250 0.040 0.027 -0.78 to -0.02 -0.42 to -0.02 Mindful eating awareness Mindful eating distraction Physical activity (per 1000 METs) -0.65 -0.42 -0.18 0.28 0.18 0.07 -0.210 -0.200 -0.220 -2.333 -2.308 -2.400 0.022 0.024 0.019 -1.20 to -0.10 -0.78 to -0.06 -0.32 to -0.04 Mindful eating emotional response Mindful eating disinhibition -0.52 -0.70 0.23 0.29 -0.205 -0.225 -2.245 -2.400 0.027 0.019 -0.97 to -0.07 -1.27 to -0.13 Mindful eating emotional response Mindful eating disinhibition -0.82 -0.99 0.34 0.40 -0.220 -0.235 -2.414 -2.471 0.018 0.016 -1.49 to -0.15 -1.78 to -0.20 Mindful eating awareness Mindful eating distraction Physical activity (per 1000 METs) -3.00 -2.20 -0.80 1.13 0.87 0.32 -0.245 -0.235 -0.235 -2.656 -2.531 -2.500 0.010 0.014 0.015 -5.23 to -0.77 -3.91 to -0.49 -1.43 to -0.17 aUnstandardized beta coefficient; bStandard error; cT-distribution; dConfidence interval; eMetabolic Equivalent of Task. Model adjusted for gender, marital status, age, BMI, and working hours per day; R2 = 0.30, Adjusted R2 = 0.26. Discussion This study aimed to investigate mindful eating practices and physical activity levels among medical interns and to examine their associations with mental workload and sleep quality. Mental workload and its subscales among participants The mean total mental workload score was 68.48 ± 13.55 out of 100, suggesting a moderately high level of perceived mental workload among medical interns. Among subscales, mental demand showed the highest mean score (80.48 ± 18.68), which may reflect substantial cognitive and intellectual demands during clinical internship. Temporal demand (79.72 ± 19.33) and effort (77.21 ± 18.55) were also high, indicating that time pressure and perceived exertion were notable concerns. Physical demand had a moderately high mean score (74.15 ± 22.17). Frustration showed a moderate mean score (63.54 ± 28.27) with the largest standard deviation, suggesting considerable individual variation in stress and frustration levels. The performance subscale had the lowest mean score (35.78 ± 21.55), which is noteworthy because lower scores indicate better self-perceived performance These findings are broadly consistent with Atalay et al.27, who reported high mental effort in medical students, and Gan et al.49, who found diverse mental workload patterns among interns with over one-third experiencing high workload levels. Associations of mindful eating and physical activity with mental workload Figure 1 shows associations of mindful eating dimensions and physical activity with mental workload dimensions [This figure was created using SciSpace data visualization capabilities (Claude 3.7 Sonnet, Anthropic) and subsequently validated by the authors]. The overall model for total mental workload accounted for 40% of the variance (R2 = 0.40, Adjusted R2 = 0.385), which is a moderate to large effect size in behavioral research.Fig. 1Associations of mindful eating dimensions and physical activity with mental workload dimensions. Note: ME: Mindful eating. Associations of mindful eating dimensions and physical activity with mental workload dimensions. Note: ME: Mindful eating. Clinical significance of the associations Before discussing specific findings, it is useful to consider whether the observed associations are practically meaningful beyond statistical significance. The unstandardized coefficients (B values) indicate the magnitude of change in raw scores:Physical activity: Each additional 1000 MET-minutes per week (approximately 150 min of brisk walking or 75 min of running) was associated with a 4.75-point reduction in total mental workload (0–100 scale). This represents about a 4.8% lower workload relative to the mean score of 68.48.Mindful eating awareness: A one-unit increase (on a 1–6 scale) was associated with a 15.05-point reduction (~ 15% lower workload).Total mindful eating score: A one-unit increase was associated with an 18.05-point reduction (~ 18% lower workload), the largest association observed. Physical activity: Each additional 1000 MET-minutes per week (approximately 150 min of brisk walking or 75 min of running) was associated with a 4.75-point reduction in total mental workload (0–100 scale). This represents about a 4.8% lower workload relative to the mean score of 68.48. Physical activity: Each additional 1000 MET-minutes per week (approximately 150 min of brisk walking or 75 min of running) was associated with a 4.75-point reduction in total mental workload (0–100 scale). This represents about a 4.8% lower workload relative to the mean score of 68.48. Mindful eating awareness: A one-unit increase (on a 1–6 scale) was associated with a 15.05-point reduction (~ 15% lower workload). Mindful eating awareness: A one-unit increase (on a 1–6 scale) was associated with a 15.05-point reduction (~ 15% lower workload). Total mindful eating score: A one-unit increase was associated with an 18.05-point reduction (~ 18% lower workload), the largest association observed. Total mindful eating score: A one-unit increase was associated with an 18.05-point reduction (~ 18% lower workload), the largest association observed. How meaningful are these reductions? Although no established "minimal clinically important difference" exists for the NASA-TLX in medical interns, a 15–18% reduction is comparable to effect sizes reported for mindfulness-based stress reduction interventions in healthcare workers, which typically achieve reductions in perceived stress and burnout50. These magnitudes suggest that the observed associations are sufficiently large to warrant further investigation in intervention studies, although causal inferences cannot be made from cross-sectional data. Associations with total mental workload and subscales For total mental workload, four variables showed notable negative associations. Mindful eating awareness (B = -15.05, β = -0.320, p < 0.001), distraction (B = -11.00, β = -0.235, p = 0.001, total mindful eating score (B = -18.05, β = -0.350, p < 0.001), and physical activity (B = -4.75, β = -0.300, p < 0.001) were each associated with lower workload. The standardized beta coefficients suggest that total mindful eating score had the strongest relative association (β = -0.350), followed by awareness (β = -0.320) and physical activity (β = -0.300). Across workload subscales, distinct patterns emerged. For mental demand, disinhibition (β = -0.320), awareness (β = -0.271), and physical activity (β = -0.290) showed the strongest associations. For physical demand, distraction (β = -0.180), total mindful eating (β = -0.235), and physical activity (β = -0.210) were notable. For performance (where lower scores indicate better self-perceived performance), disinhibition (β = -0.340), total mindful eating (β = -0.310), and physical activity (β = -0.280) showed associations. Effort was uniquely associated only with disinhibition (β = -0.295). Frustration was associated with emotional response to eating (β = -0.245) and total mindful eating (β = -0.280). No notable associations emerged for temporal demand with any mindful eating dimension or physical activity. This finding may have important implications. From a theoretical perspective, it supports the distinction within self-regulation theory between demands that are modifiable through individual behaviors versus those that are structurally determined. Unlike mental demand or frustration-which showed consistent associations with mindful eating and physical activity-time pressure appeared resistant to individual-level factors. From a practical perspective, this suggests that interventions targeting temporal demand may require organizational-level changes (e.g., duty hour restrictions, protected break times, workload redistribution, streamlined handoff processes) rather than individual skills training alone. In contrast, other workload dimensions (mental demand, frustration, effort) may be more amenable to individual-level interventions such as mindful eating training or physical activity promotion. This distinction, if confirmed by longitudinal research, could inform resource allocation in intern well-being programs. Overall pattern and comparison with previous studies The pattern of associations is broadly consistent with self-regulation theory: mindful eating dimensions reflecting attentional control (awareness, distraction, total score) showed the strongest associations with overall workload, while emotional regulation dimensions (disinhibition, emotional response) were more relevant to frustration and effort. Physical activity was associated with most workload dimensions. These findings align with previous research. Choi et al.40 found that mindful eating was positively associated with mental well-being and negatively associated with occupational stress among clinical nurses. Knol et al.51 showed that a mindful eating intervention improved overall mindful eating scores and disinhibition. Ireland et al.50 demonstrated that a 10-week mindfulness training intervention reduced stress and burnout among intern doctors. Kalmbach et al.52 found that physical inactivity increased during internship and showed bidirectional relationships between sleep and mood. Collectively, these studies support the relevance of mindful eating and physical activity for mental workload and well-being, though most used different designs and populations. Sleep quality and its subscales among participants The overall mean global PSQI score was 14.83 ± 3.24 out of a possible range of 0–21, which is substantially above the established clinical cut-off score of 5 used to distinguish poor sleepers from good sleepers. This finding suggests that the medical interns in this study experienced poor sleep quality on average, with the mean score approaching three times the clinical threshold. The seven PSQI component scores showed distinct patterns of sleep difficulties. Duration of sleep had the highest mean score (2.91 ± 0.65 out of 3), suggesting that nearly all participants experienced significant problems with obtaining adequate sleep duration, likely reflecting the demanding schedules of medical internship. Day dysfunction due to sleepiness also showed a notably high mean score (2.38 ± 0.68), suggesting that poor sleep had substantial consequences for daytime functioning, alertness, and ability to maintain enthusiasm for tasks. Sleep latency demonstrated a high mean score (2.29 ± 0.85), suggesting considerable difficulties in initiating sleep, possibly reflecting stress, anxiety, or irregular work schedules. Subjective sleep quality showed a similarly high mean score (2.24 ± 0.46). Sleep disturbance had a moderately high mean score (2.16 ± 0.45). In contrast, two components showed relatively lower scores. Need for sleep medication had the lowest mean score (1.39 ± 0.37), suggesting that most participants were not regularly using pharmacological sleep aids despite their poor sleep quality. Sleep efficiency also showed a relatively lower mean score (1.44 ± 0.44), suggesting that while participants had difficulty obtaining adequate sleep duration, the ratio of time asleep to time in bed was relatively preserved. This pattern suggests that the primary sleep problems among medical interns were related to insufficient sleep opportunity, difficulty initiating sleep, and daytime consequences, rather than prolonged wakefulness while in bed. These findings are broadly consistent with Limsuwat et al.53, who found that 28% of interns had poor sleep quality prior to training; Xiong et al.54, who reported that 91.86% of internship nurses experienced poor sleep quality; Marashi et al.55, who documented inadequate sleep quality (PSQI score 7.29) among medical interns; and Meles et al.8, who found 72.6% of medical interns had poor sleep quality associated with anxiety and poor sleep hygiene. Associations of mindful eating and physical activity with sleep quality Figure 2 shows associations of mindful eating dimensions and physical activity with sleep quality dimensions [This figure was created using SciSpace data visualization capabilities (Claude 3.7 Sonnet, Anthropic) and subsequently validated by the authors]. The overall model for total sleep quality accounted for 30% of the variance (R2 = 0.30, Adjusted R2 = 0.26), which represents a moderate effect size in sleep research.Fig. 2Associations of mindful eating dimensions and physical activity with sleep quality dimensions. Note: ME: Mindful eating. Associations of mindful eating dimensions and physical activity with sleep quality dimensions. Note: ME: Mindful eating. Clinical significance of the associations with sleep quality As with mental workload, it is useful to consider whether the observed associations with sleep quality are practically meaningful. The unstandardized coefficients (B values) indicate the magnitude of change:Physical activity: Each additional 1000 MET-minutes per week was associated with a 0.80-point reduction in global PSQI score (where lower scores indicate better sleep). Given that a 3-point reduction is generally considered clinically meaningful in sleep research56, achieving higher activity levels (e.g., 3000–4000 MET-minutes/week) could produce a 2.4–3.2-point reduction—potentially moving some interns toward better sleep classification.Mindful eating awareness: A one-unit increase was associated with a 3.00-point reduction (~ 20% relative to the mean PSQI of 14.83), indicating a larger practical association than physical activity. Physical activity: Each additional 1000 MET-minutes per week was associated with a 0.80-point reduction in global PSQI score (where lower scores indicate better sleep). Given that a 3-point reduction is generally considered clinically meaningful in sleep research56, achieving higher activity levels (e.g., 3000–4000 MET-minutes/week) could produce a 2.4–3.2-point reduction—potentially moving some interns toward better sleep classification. Physical activity: Each additional 1000 MET-minutes per week was associated with a 0.80-point reduction in global PSQI score (where lower scores indicate better sleep). Given that a 3-point reduction is generally considered clinically meaningful in sleep research56, achieving higher activity levels (e.g., 3000–4000 MET-minutes/week) could produce a 2.4–3.2-point reduction—potentially moving some interns toward better sleep classification. Mindful eating awareness: A one-unit increase was associated with a 3.00-point reduction (~ 20% relative to the mean PSQI of 14.83), indicating a larger practical association than physical activity. Mindful eating awareness: A one-unit increase was associated with a 3.00-point reduction (~ 20% relative to the mean PSQI of 14.83), indicating a larger practical association than physical activity. Associations with total sleep quality and subscales For total sleep quality, three variables showed negative associations. Mindful eating awareness (B = -3.00, β = -0.245, p = 0.010), distraction (B = -2.20, β = -0.235, , p = 0.014), and physical activity (B = -0.80, β = -0.235, p = 0.015) were each associated with better sleep quality (lower PSQI scores). The standardized beta coefficients suggest comparable relative associations (β = -0.245, -0.235, and -0.235). Across sleep subscales, distinct patterns emerged:Sleep latency: Emotional response (β = -0.180), external cues (β = -0.185), and physical activity (β = -0.200) showed associations.Sleep duration: Disinhibition (β = -0.195) and physical activity (β = -0.210) were associated with better duration scores.Sleep efficiency: External cues (β = -0.190) and physical activity (β = -0.200) showed associations.Sleep disturbance: Awareness (β = -0.210), distraction (β = -0.200), and physical activity (β = -0.220) were associated with fewer disturbances.Need for sleep medication: Physical activity was uniquely associated with this subscale (β = -0.230), suggesting a potentially important role in reducing medication dependence.Daytime dysfunction: Emotional response (β = -0.205) and disinhibition (β = -0.225) showed associations.Subjective sleep quality: Emotional response (β = -0.220) and disinhibition (β = -0.235) were associated with better subjective sleep perception. Sleep latency: Emotional response (β = -0.180), external cues (β = -0.185), and physical activity (β = -0.200) showed associations. Sleep latency: Emotional response (β = -0.180), external cues (β = -0.185), and physical activity (β = -0.200) showed associations. Sleep duration: Disinhibition (β = -0.195) and physical activity (β = -0.210) were associated with better duration scores. Sleep duration: Disinhibition (β = -0.195) and physical activity (β = -0.210) were associated with better duration scores. Sleep efficiency: External cues (β = -0.190) and physical activity (β = -0.200) showed associations. Sleep efficiency: External cues (β = -0.190) and physical activity (β = -0.200) showed associations. Sleep disturbance: Awareness (β = -0.210), distraction (β = -0.200), and physical activity (β = -0.220) were associated with fewer disturbances. Sleep disturbance: Awareness (β = -0.210), distraction (β = -0.200), and physical activity (β = -0.220) were associated with fewer disturbances. Need for sleep medication: Physical activity was uniquely associated with this subscale (β = -0.230), suggesting a potentially important role in reducing medication dependence. Need for sleep medication: Physical activity was uniquely associated with this subscale (β = -0.230), suggesting a potentially important role in reducing medication dependence. Daytime dysfunction: Emotional response (β = -0.205) and disinhibition (β = -0.225) showed associations. Daytime dysfunction: Emotional response (β = -0.205) and disinhibition (β = -0.225) showed associations. Subjective sleep quality: Emotional response (β = -0.220) and disinhibition (β = -0.235) were associated with better subjective sleep perception. Subjective sleep quality: Emotional response (β = -0.220) and disinhibition (β = -0.235) were associated with better subjective sleep perception. Overall pattern and comparison with previous studies The pattern of associations is broadly consistent with self-regulation theory. Physical activity was associated with multiple sleep dimensions (latency, duration, efficiency, disturbance, and medication need), suggesting broad beneficial associations. Emotional regulation dimensions of mindful eating (emotional response and disinhibition) showed particular relevance to subjective sleep quality and daytime dysfunction, suggesting shared underlying mechanisms between eating behavior and sleep. These findings align with previous research. Ye et al.57 found that physical exercise was associated with improved sleep quality through mindfulness mediation (β = -0.15, p < 0.001) among college students. Wunsch et al.58 demonstrated that physical activity buffered stress effects on sleep quality during examination periods. KÖSE et al.59 reported a high prevalence of poor sleep (59.2%) among female students, though the relationship with mindful eating was not statistically significant—a discrepancy that may reflect our use of specific mindful eating dimensions rather than total scores. Strengths and limitations This study has several notable strengths. (1) It addresses an important and understudied population—medical interns—who face unique occupational stressors placing them at high risk for poor health outcomes. (2) The study employed validated, culturally adapted Persian versions of internationally recognized assessment tools, ensuring reliable measurement. (3) Multivariate regression analyses controlling for relevant confounders strengthened internal validity and allowed examination of independent associations. (4) Comprehensive assessment of multiple dimensions of mental workload and sleep quality provided nuanced insights into specific aspects differentially associated with eating behaviors and physical activity. (5) Adequate sample size based on a priori power analysis ensured sufficient statistical power. (6) The focus on modifiable behavioral factors has direct implications for intervention development. However, several limitations should be acknowledged. (1) The cross-sectional design precludes causal inference; longitudinal or experimental studies are needed to establish directionality. (2) Self-report measures introduce recall and social desirability biases; objective tools (e.g., accelerometry) would yield more accurate assessments. (3) The IPAQ captures only the past week and may not reflect long-term activity patterns. (4) Unmeasured confounders—such as dietary quality, caffeine intake, shift schedules, social support, and mental health—likely contribute to the substantial unexplained variance (R2 = 0.30–0.40). (5) The P-MEQ’s low test–retest reliability (0.59) falls below acceptable thresholds, likely attenuating observed associations via regression dilution bias and may reduce the robustness of the findings. (6) External validity is limited by convenience sampling from a single university; findings should not be generalized beyond comparable populations without multicenter replication. (7) Despite adequate power, the modest sample (n = 165) relative to the number of predictors risks model overfitting, warranting larger samples or regularization techniques. (8) Statistically, post-hoc checks indicated VIF values below 5, but this should be reported prospectively in future work. (9) Given the large number of regression models and predictors, the risk of Type I error is elevated. To prioritize transparency over strict confirmation bias, we report uncorrected p-values and explicitly interpret them as descriptive measures of association strength—not as confirmatory evidence for hypotheses. All statistically significant findings (p < 0.05) should be considered hypothesis-generating and require replication in future longitudinal or interventional studies. No adjustments such as Bonferroni or False Discovery Rate (FDR) were applied due to the exploratory nature of this study. (10) The study did not account for seasonal variations. Clinical implications and recommendations for future research Clinical implications Despite the cross-sectional design, the observed associations suggest several potential directions for clinical practice. First, the strong association between mindful eating and lower mental workload suggests that teaching mindful eating techniques—particularly awareness and disinhibition—might be a useful component of well-being programs for medical interns. These skills are low-cost, can be taught in brief sessions, and may be easily integrated into existing wellness curricula. Second, physical activity showed consistent associations with both lower mental workload and better sleep quality across multiple dimensions. Even moderate increases in activity (e.g., 1000 MET-minutes/week, equivalent to 150 min of brisk walking) were associated with measurable differences in outcomes, suggesting that even achievable activity goals may be beneficial. Third, the null finding for temporal demand suggests that time pressure may require organizational-level interventions (e.g., duty hour adjustments, workload redistribution) rather than individual-level behavioral changes—an important consideration for program directors and policymakers. Recommendations for future research Given the limitations above, longitudinal and experimental studies are needed to establish causality. Future research should incorporate objective measures (actigraphy, accelerometry), diverse multi-center samples, and examination of mediators and moderators (stress hormones, self-regulatory capacity). Additionally, given the null finding for temporal demand, research should investigate structural determinants of time pressure, including duty hour policies and patient load distributions. Finally, implementation science studies are needed to develop feasible, sustainable interventions tailored to medical training environments. Conclusions This study provides evidence that mindful eating practices and physical activity are associated with lower mental workload and better sleep quality among medical interns. The findings reveal that medical interns experience substantial mental workload and poor sleep quality. Mindful eating awareness, disinhibition, emotional response, and distraction, along with physical activity, showed independent associations with the outcomes after controlling for demographic and lifestyle confounders. Physical activity was associated with favorable sleep quality across multiple dimensions, while emotional regulation aspects of mindful eating showed notable correlations with subjective experiences of workload and sleep. Notably, temporal demand showed no associations, suggesting that time pressure may be primarily determined by structural training factors rather than individual behaviors. Given the cross-sectional design of this study, causal inferences cannot be made, and the identified associations should be interpreted as correlational. These findings suggest that these modifiable lifestyle behaviors may be relevant targets for further research aimed at supporting trainee well-being. However, comprehensive approaches addressing both individual-level interventions and organizational-level changes in workload management and institutional support are necessary for optimal outcomes. Future longitudinal and intervention research is needed to examine causal relationships and develop evidence-based, feasible programs tailored to the unique demands of medical education environments.