The Gender Differences of Mediating Effect of Body Mass Index Between Perceived Control and Depressive Symptoms in Korean Workers
Article information
Abstract
Objective
Depression impairs workplace productivity. Although obesity is associated with depression, findings differ across populations. Perceived control over stressors is an important psychological factor affecting both depression and weight regulation. This study investigated whether body mass index (BMI) mediates the relationship between perceived control and depressive symptoms, considering gender differences.
Methods
A total of 7,067 Korean employees (4,627 male and 2,440 female), aged 19–65, completed self-report measures: the Perceived Stress Scale (control subscale), the CES-D scale for depressive symptoms, and self-reported height and weight, which were used to calculate BMI. Gender-stratified mediation analyses were conducted using the PROCESS macro in SPSS.
Results
Among male, lower perceived control was significantly associated with higher BMI (B=-0.099, standard error [SE]=0.031, p= 0.001), and higher BMI predicted more severe depressive symptoms (B=0.105, SE=0.035, p=0.003). The indirect effect was significant (B=-0.010, 95% confidence interval [CI] [-0.021, -0.003]), indicating partial mediation. In female, perceived control was directly associated with depressive symptoms, but BMI did not mediate this relationship (indirect effect: B=-0.004, 95% CI [-0.017, 0.004]).
Conclusion
BMI partially mediated the association between perceived control and depression in male but not in female. Perceived control was negatively associated with both BMI and depressive symptoms in both genders; however, BMI contributed to depressive symptoms only in male. Gender-specific approaches may be warranted in workplace mental health interventions.
INTRODUCTION
Depressive disorders are a major public health concern, causing significant impairments in functioning, productivity, and quality of life [1]. In South Korea, workplace depression has gained attention for its association with reduced performance and increased suicide risk [2]. Among various psychosocial factors, perceived control—individuals’ belief in their ability to influence life events—has been consistently linked to mental health outcomes. Theoretical models, including Rotter’s locus of control [3], learned helplessness [4], and depressive realism [5], suggest that low perceived control fosters helplessness and increases vulnerability to depressive symptoms [6-8]. Recent meta-analyses confirm that individuals with lower perceived control or an external locus of control are at greater risk for depression and anxiety [2,9].
Beyond its direct psychological effects, low perceived control is linked to adverse health behaviors and physical health outcomes. Individuals with diminished control often exhibit poorer self-efficacy regarding health management, leading to maladaptive coping strategies such as emotional eating, reduced physical activity, and weight gain [10-12]. These behaviors can increase body mass index (BMI), a common proxy for obesity and physical health [13]. Prior studies suggest a potential indirect pathway from perceived control to depressive symptoms through increased BMI [14].
The association between obesity and depression is well documented, with evidence indicating a bidirectional relationship [15-18]. Depression can promote weight gain through altered appetite and inactivity, while obesity—particularly in sociocultural contexts that idealize thinness—may cause psychological distress. Recent studies suggest that the path from obesity to depression may be stronger than the reverse, driven by physiological factors such as hypothalamic–pituitary–adrenal (HPA) axis dysregulation and chronic inflammation, as well as psychosocial factors like reduced self-esteem or internalized weight stigma [19-21].
Gender plays a significant moderating role in these associations [22,23]. Female generally have a higher prevalence of both depression and obesity, and research has indicated that the association between the two may be stronger in female [24,25]. In societies emphasizing thinness, female may experience greater body dissatisfaction and psychological distress [26]. However, some studies report opposite trends. For example, in a longitudinal study, BMI gain was associated with stressful life events only in female, not male [27]. These inconsistencies may reflect cultural differences, sample characteristics, or gender-specific coping styles. Male often respond to stress with externalizing behaviors such as anger, drinking, or overeating, which can lead to weight gain [28-30], whereas female tend to use internalizing strategies like rumination or seeking emotional support [31,32]. While thinness is emphasized for female, male may also experience distress if weight gain is seen as a failure of self-control or a deviation from masculine norms, especially in cultures that value muscularity [33,34].
Although previous research has examined relationships between perceived control, BMI, and depression, most studies have addressed these as separate bivariate associations—either between control and depression or between BMI and depression. Few have tested a full mediation model assessing whether perceived control influences depressive symptoms indirectly through BMI, and fewer still have done so in large working populations with gender-stratified analyses. In occupational settings, low perceived control (e.g., low job autonomy) is a key psychosocial stressor linked to poor mental health outcomes via models such as the job demand–control model [35,36]. These effects may be partially mediated by physical health indicators such as BMI, and the pathways may differ by gender.
Understanding these mechanisms has practical implications for designing gender-sensitive workplace mental health interventions. If low perceived control leads to depression through weight-related mechanisms in male, integrated strategies combining psychological and physical health measures—such as stress management and lifestyle modification—may be more effective. For female, interventions that enhance perceived control and address psychosocial stressors may provide more direct benefits [37]. This study examined whether BMI mediates the relationship between perceived control and depressive symptoms in a large Korean working population, and whether this mediation pathway differs between male and female as outlined in the conceptual framework (Figure 1). The hypotheses were: 1) lower perceived control is associated with higher depressive symptoms; 2) this association is partially mediated by BMI, such that lower control is linked to higher BMI, which in turn is associated with increased depressive symptoms; and 3) the mediating effect of BMI may differ by gender, potentially being more pronounced in male than in female based on prior literature.
Conceptual framework. BMI, body mass index; PSS, Perceived Stress Scale.
METHODS
Study participants
This study analyzed data from employees aged 19 to 65 who underwent routine workplace mental health screening at the Workplace Mental Health Institute, Kangbuk Samsung Hospital in Seoul, Republic of Korea. The final analysis included 7,067 participants (4,626 male and 2,440 female) with complete sociodemographic and clinical data. All procedures were approved by the Institutional Review Board of Kangbuk Samsung Hospital (IRB No. KBSMC 2022-03-046). As anonymized data were obtained from routine mental health screenings, informed consent was waived in accordance with the Declaration of Helsinki.
Assessments
Perceived control was measured using the perceived control subscale of the 10-item Perceived Stress Scale (PSS-10). We summed the four PSS-10 items that assess one’s ability to handle stress (e.g., feeling confident about handling personal problems), consistent with the PSS-10’s established two-factor structure [9,38]. This validated subscale was chosen because it provides a brief measure of perceived control within a widely used stress instrument. Higher subscale scores indicate greater perceived control.
Depressive symptoms were measured using the Center for Epidemiologic Studies Depression Scale (CES-D), developed by the U.S. National Institute of Mental Health and validated for Korean populations [39]. The CES-D is a 20-item self-report questionnaire assessing the frequency of depressive symptoms over the past week. Each item is rated on a 4-point scale (0–3), producing a total score ranging from 0 to 60, with higher scores indicating greater severity. The total CES-D score was used as a continuous variable in the analyses.
BMI, the mediator in this study, was calculated from each participant’s self-reported height and weight using the standard formula (weight in kilograms divided by height in meters squared).
To improve model precision and control for potential confounding effects, several sociodemographic and occupational variables were included as covariates: age, educational attainment (high school or less, college graduate, graduate school or higher), marital status (married, single, other), years of employment, weekly working hours, and monthly income.
Statistical analysis
All statistical analyses were conducted using SPSS Statistics for Windows (version 25.0, IBM Corp.). Descriptive statistics (means, standard deviations, frequencies, and percentages) summarized the participants’ sociodemographic, occupational, and clinical characteristics. Gender differences in baseline characteristics were assessed using independent samples t-tests for continuous variables and chi-square tests for categorical variables.
To test the hypothesized gender-specific mediation effects, we used Model 4 of the PROCESS macro for SPSS developed by Hayes [40]. Analyses were stratified by gender, with separate models estimated for male and female participants to explicitly examine differential pathways.
The indirect effect of perceived control on depressive symptoms via BMI was evaluated using a bootstrapping procedure with 5,000 resamples. Statistical significance was determined when the 95% confidence interval (CI) excluded zero. The total effect (path c: X→Y), direct effect (path c': X→Y, controlling for the mediator), and component paths of the indirect effect (path a: X→M; path b: M→Y) were estimated and reported. A two-tailed p-value of <0.05 was considered statistically significant for all analyses.
RESULTS
Sociodemographic and clinical characteristics of the study participants
The sociodemographic, occupational, and clinical characteristics of the total sample (n=7,067) are shown in Table 1. The mean age was 37.58±9.00 years, with male participants (38.55±9.19 years) significantly older than female participants (35.74±8.32 years; p<0.001). Significant gender differences were observed across most sociodemographic variables, including educational attainment, marital status, and monthly income.
Baseline characteristics in participants
Baseline clinical characteristics of participants by gender
Significant gender differences were also observed in clinical characteristics. Male had a higher mean BMI (25.02±3.22 kg/m2) than female (21.04±2.62 kg/m2; p<0.001). Perceived control scores (PSS control) were also higher in male (4.69±1.56) than in female (4.36±1.59; p<0.001). Conversely, female had higher depressive symptom scores (CES-D) (17.52±10.37) than male (14.38±9.17; p<0.001). These baseline differences informed the stratification of mediation analyses by gender.
Results of mediation analysis by gender
The mediation analysis results on BMI’s role in the relationship between perceived control and depressive symptoms are presented in Table 2 (mediation analysis by gender), with the results for male and female shown side by side. Distinct patterns emerged between the two groups. See Figure 2 for a schematic summary of the mediation pathways by gender.
Results of the parallel mediation analysis
Comparison of the mediation pathways between perceived control (PSS), BMI, and depression by gender. A: Model in male. B: Model in female. Solid arrows indicate significant paths (p<0.05); dotted arrows indicate nonsignificant paths. BMI, body mass index; PSS, Perceived Stress Scale.
Mediation analysis in male employees
In the male employee sample (n=4,627), mediation analysis (Supplementary Table 1) showed that BMI significantly and partially mediated the relationship between perceived control and depressive symptoms. Lower perceived control was associated with higher BMI (path a: B=-0.099, SE=0.031, p=0.001), and higher BMI was associated with increased depressive symptoms (path b: B=0.105, SE=0.035, p=0.003). Bootstrap analysis confirmed the indirect effect of perceived control on depressive symptoms via BMI (B=-0.010, 95% CI [-0.021, -0.003]). The direct effect of perceived control on depressive symptoms remained significant after including BMI in the model (direct effect: B=-3.061, 95% CI [-3.206, -2.915]), indicating partial mediation.
Mediation analysis in female employees
In the sample of female employees (n=2,440), a different pattern emerged compared to males. As shown in Supplementary Table 2, BMI did not mediate the relationship between perceived control and depressive symptoms. The effect of perceived control on BMI (path a: B=-0.036, SE=0.033, p=0.277) and the effect of BMI on depressive symptoms (path b: B=0.106, SE=0.067, p=0.114) were both statistically non-significant. Bootstrap analysis also showed that the indirect effect of perceived control on depressive symptoms via BMI was not statistically significant (B=-0.004, 95% CI [-0.017, 0.004]). In contrast, the direct effect of perceived control on depressive symptoms remained significant after including BMI in the model (B=-3.560, 95% CI [-3.776, -3.344]).
DISCUSSION
This study, based on a large sample of Korean employees, examined the relationship between workplace psychosocial stressors and mental health, identifying a gender-specific mechanism. The key findings can be summarized as follows: among male employees, BMI significantly and partially mediated the association between lower perceived control and higher depressive symptoms. In female employees, no such mediating effect was observed; lower perceived control had a strong, direct effect on depressive symptoms, independent of BMI. These findings suggest that the pathways linking workplace stress to mental health differ by gender, highlighting the need for gender-specific intervention strategies.
Among male employees, the pathway from low perceived control to increased BMI and subsequently to elevated depressive symptoms may reflect a gendered stress response pattern characterized by somatization. This pathway is likely shaped by sociocultural expectations and occupational role demands placed on male in Korean workplaces. In this context, loss of perceived control may indicate not only reduced autonomy at work but also a perceived failure to fulfill the expected role of a successful provider, thereby intensifying psychological strain [41]. Chronic stress under such conditions can dysregulate the HPA axis, promote abdominal fat accumulation [42], and increase maladaptive health behaviors such as overeating and reduced physical activity [43]. These mechanisms may explain the significant association observed between low perceived control and higher BMI in male participants (path a). Elevated BMI may also constitute an additional psychological burden in male, especially in cultural contexts where masculinity is linked to physical strength and self-discipline [30]. Weight gain may be viewed as a failure in self-management, leading to shame, reduced self-esteem, and worsening mood [36,41,43]. Among working male, high job demands and limited time for physical activity may exacerbate these effects, contributing to the significant BMI → depression link (path b). Overall, the mediation pathway in male suggests that occupational stress may contribute to psychological distress through physiological and behavioral channels, reflecting a gender-specific pattern of embodied stress response.
Among female employees, the absence of a significant BMI mediation effect suggests that low perceived control influences depression primarily through direct psychological pathways rather than through weight gain [44]. In other words, female with low perceived control may experience increased depressive symptoms due to factors like stress-induced rumination, feelings of helplessness, or emotional exhaustion, without significant weight-related changes [32,41]. Indeed, previous studies imply that when weight does contribute to depression in female, it is often via subjective factors (e.g., body image dissatisfaction or internalized weight stigma) rather than BMI alone [36]. Such factors were not captured in our data, which could explain why BMI was not a significant mediator for female in our study [45]. Thus, for female, low perceived control appears to elevate depression risk mostly through immediate emotional/cognitive responses rather than via the indirect route of obesity [46].
The present results provide insights on the relationship between occupational stress, obesity, and depression. While Western studies have reported that stress is more strongly associated with BMI increases in female than in male [47-49], our findings indicate that underlying mechanisms may vary depending on the type of stressor (e.g., perceived control at work) and sociocultural context of the population studied (e.g., Korean employees). This underscores the need to consider contextual factors when examining stress-health relationships. Prior research among adolescents has shown a bidirectional association between obesity and depression in both sexes, with stronger effects in girls [50]. Our findings suggest that in working adults, psychosocial drivers may shift over the life course. Whereas body image and peer relations may be central stressors in adolescence, occupational roles, professional achievement, and perceived organizational justice may become more salient in adulthood, potentially leading to gender-divergent physical and psychological outcomes [51].
This study has several limitations. First, its cross-sectional design precludes causal inference; longitudinal data are required to clarify the temporal direction between perceived control, BMI, and depressive symptoms. Second, all measures—including perceived control, depressive symptoms, and BMI—were self-reported, which may introduce reporting bias; using objective or multi-method assessments in future studies would strengthen validity. Third, BMI is an imperfect indicator of adiposity and does not capture fat distribution or distinguish muscle from fat; more precise anthropometric indices (e.g., waist circumference or body fat percentage) should be considered, particularly given gender differences in body composition. Fourth, although we controlled for major sociodemographic and occupational variables, unmeasured confounders such as psychiatric history, physical illnesses, medication use, or specific occupational characteristics (e.g., job type, position, work autonomy) may have influenced the results. Finally, the relatively low R2 values observed in this study reflect the multifactorial nature of depressive symptoms, which are shaped by diverse psychosocial and biological determinants beyond the variables included in our simple mediation framework (PROCESS Model 4). Future research incorporating moderation or moderated mediation models could better capture these complex interactions and improve explanatory power.
This study examined gender-specific pathways linking perceived control, BMI, and depressive symptoms among Korean employees, revealing distinct mechanisms in male and female. In male, lower perceived control was associated with higher depressive symptoms through increased BMI, indicating a significant indirect effect. In female, BMI did not mediate this relationship; instead, perceived control had a strong, direct effect on depressive symptoms, independent of BMI. These findings suggest that workplace stress affects mental health through different mechanisms depending on gender. This study suggests that the pathways linking workplace stress to depressive symptoms differ by gender. In male, reduced perceived control contributed to depressive symptoms via increased BMI, while in female, perceived control directly influenced mental health, independent of BMI. It is noteworthy that while numerous studies have reported stronger links between obesity and depression in female, our findings demonstrate a BMI-mediated pathway exclusively in male. This discrepancy suggests that gender differences in stress-related depression may be contingent on the specific stress context and cultural setting. In summary, the findings of this study demonstrate that a universal approach to workplace mental health is likely to be inadequate. The conclusion drawn from this analysis is that any such interventions must be both gender-sensitive and pathway-specific. For male employees, organisational programmes that enhance job autonomy and incorporate weight management or other health-promotion components may help disrupt the stress–BMI–depression cycle. It is hypothesised that for female employees, the improvement of the psychosocial work environment, with an emphasis on the promotion of fairness, support, and empowerment, may offer a more direct protective measure against stress-related depression. These targeted strategies, informed by gender-specific findings, have the potential to enhance mental well-being in the workplace.
Supplementary Materials
The Supplement is available with this article at https://doi.org/10.30773/pi.2025.0292.
Results of the parallel mediation analysis: male
Results of the parallel mediation analysis: female
Notes
Availability of Data and Material
The data that would be necessary to interpret, replicate, and build upon the methods or findings reported in this article are available on request from the corresponding author S.C. The data are not publicly available because of ethical restrictions that protect patient privacy and consent.
Conflicts of Interest
The authors have no potential conflicts of interest to disclose.
Author Contributions
Conceptualization: Yoosuk An, Sra Jung, Jeong Hun Yang, Sung Joon Cho. Data curation: all authors. Formal analysis: Yoosuk An, Sra Jung, Jeong Hun Yang, Sung Joon Cho. Supervision: Sang-Won Jeon, Jeong Hun Yang, Sung Joon Cho. Validation: Sang-Won Jeon, Junhyung Kim, Eun Soo Kim, Jeong Hun Yang, Sung Joon Cho. Writing—original draft: Yoosuk An, Sra Jung. Writing—review & editing: all authors.
Funding Statement
None
Acknowledgments
None