The China Labor-force Dynamics Survey (CLDS) is a nationally representative longitudinal study conducted biennially by the Social Science Survey Center at Sun Yat-sen University. Initiated in 2012, the survey employs a multi-stage, stratified probability sampling method to track labor force dynamics in 29 provincial - level administrative regions in China [37]. The 2018 wave (the fourth and most recent available dataset at the time of analysis) collected comprehensive data on working-age populations, including measures of education, employment status, occupational mobility, workplace protections, and subjective wellbeing.

From the original 2018 CLDS cohort of 16,537 respondents, we applied the following inclusion criteria: (i) aged 15–64, or 65 and older but still employed; (ii) with complete data for all study variables. After applying these criteria and conducting data quality checks, our analytical sample comprised 10,065 participants. The data screening process is shown in Figure 1.

Depression was assessed using the 20-item Center for Epidemiological Studies Depression Scale (CES-D), a validated self-report instrument designed to measure depressive symptomatology in the general population [38]. Respondents rated the frequency with which they experienced symptoms during the preceding week on a 4-point scale ranging from 0 (rarely or none of the time) to 3 (most or all of the time), yielding a total score range of 0–60, with higher scores indicating greater symptom severity. Based on a meta-analysis by Gemma Vilagut et al. [39], a cut-off score of 20 was deemed more appropriate for identifying cases of “possible depression or depressed”. Consistent with these recommendations, we established a cut-off score of 20 to optimize case identification. This threshold yielded two categories: no depression symptoms (CES-D ≤ 19) and possible depression or depressed (CES-D ≥ 20). The continuous depression score served as our primary dependent variable in regression analyses.

The independent variable, subjective social status, was assessed using the validated MacArthur Scale of Subjective Social Status [40]. This instrument employs a graphical representation of a 10-rung ladder, where respondents are instructed: “Imagine this ladder represents where people stand in society. At the top (rung 10) are those who are best off. At the bottom (rung 1) are those who are worst off. Please indicate where you would place yourself on this ladder.” In accordance with the CLDS protocol, responses were coded as a continuous variable ranging from 1 (lowest status) to 10 (highest status). The MacArthur Scale of SSS demonstrated strong psychometric properties in a study of 300 middle-aged and older married U.S. couples, with clear evidence of convergent and discriminant validity [41]. Moreover, the scale has now been widely applied in research involving the Chinese population [42, 43].

As a mediator variable, sense of social equity was assessed using a single-item: “How do you think of the current overall social equity?”. Responses were recorded on a 5-point scale ranging from 1 (totally unfair) to 5 (totally fair).

Existing research has demonstrated that occupation is closely associated with SSS in British civil servants and American urban populations. These studies also revealed a stronger association between SSS, education, and income [44]. Based on this evidence, our analysis adjusted for several key covariates, including employment category, annual personal income, education level, self-rated health, and time spent on housework per day. Additional demographic covariates included age, gender, marital status, and urban/rural residence.

Marital status, education level, residence, employment category, annual personal income, and self-rated health were all based on a questionnaire. Marital status was categorized into six groups: single, first-married, remarried, divorced, widowed, and cohabiting. Education level was classified across six levels, ranging from illiteracy to bachelor’s degree or above. Residence was classified as either urban or rural. The employment category was divided into six groups: “Production personnel in agriculture, forestry, animal husbandry, fishing and water resources,” “Heads of party organs, state organs, mass organizations and social organizations, enterprises and institutions,” “Professionals and technical personnel,” “Clerical and related personnel,” Social production service and life service personnel,” and “Manufacturing machine-related personnel.” Annual personal income was divided into five groups: less than 10,000 Yuan, 10,000–30,000 Yuan, 30,000–50,000 Yuan, 50,000–100,000 Yuan, and 100,000 Yuan or more. Self-rated health was assessed through the question: “How would you describe your current health status?”, with five response options from 1 (excellent) to 5 (poor). The age classification was based on the research of Lin Z et al. [45], and was divided into three groups: 15–47 years old, 48–63 years old, and ≥64 years old. Housework time per day was divided into four groups: less than 1 h, 1–2 h, 2–3 h, and 3 h or more. Detailed codes for all categorical variables are provided in Supplementary File 1.

We employed a comprehensive statistical approach to examine the relationship between SSS, sense of social equity, and depression. Continuous variables were summarized using means and standard deviations, while categorical variables were presented as frequencies and percentages. Prior to analysis, we conducted normality tests on all continuous variables, applying appropriate transformations when necessary to meet normality assumptions. The core analysis utilized a linear regression model to assess the association between SSS and depression. To investigate potential gender differences in this relationship, we included an interaction term between gender and SSS in the model. All regression analyses were performed using Stata, version 15.0. For the mediation analysis, we employed a Structural Equation Model (SEM) with a bootstrap method using 1000 resamples [46] to test whether sense of social equity mediated the relationship between SSS and depression. The Structural Equation Model was conducted using Mplus Version 7.0. All models were adjusted for relevant covariates, and all statistical tests were two-sided. For all effect estimates, either α = 0.05 or 95% confidence intervals are reported.

The analytic sample comprised 10,065 participants. Raw depression scores were skewed, with a median of 4 (interquartile range: 0–12); 12.37% of participants were classified as belonging to the “possible depression or depressed” category. Because depression followed a normal distribution after logarithmic transformation, our study used the logarithmically transformed depression as the dependent variable for analysis. The basic characteristics of the participants are presented in Table 1. The majority of participants were between 15 and 63 years old. Over half were men, and the majority were married (5 out of 6). Nearly three-quarters lived in rural areas, and more than one-third had a middle school education. Social production service and life service personnel formed the largest occupational group. The majority of participants had an annual personal income of 100,000 Yuan or less, and nearly five-sixths participants rated their health as good or better. In total, 40% of the participants spent 2 h or more on housework every day. One-way ANOVA indicated that depression varied significantly across groups defined by age, gender, marital status, residence, education, employment category, annual personal income, self-rated health, and housework time per day.

The mean SSS score was 4.47 ± 1.73, and the average sense of social equity score was 3.28 ± 0.88. Pearson correlation analysis revealed significant correlations between depression and both SSS (r = −0.19, P < 0.05) and sense of social equity (r = −0.22, P < 0.05).

As shown in Table 2, Model 1 examined the association between SSS and depression, with all covariates controlled for. Model 1 indicated a significant correlation between SSS and depression (β = −0.08, P < 0.05). Building on Model 1, Model 2 further explored the interaction between SSS and gender. It revealed no significant interaction between subjective social status and gender (β = −0.01, P > 0.05), with the findings visualized in Figure 2.

With all covariates controlled, we examined the mediating role of sense of social equity in the association between SSS and depression using SEM. The SEM analysis revealed significant associations between SSS, sense of social equity and depression. First, we observed that higher SSS was directly associated with both a greater sense of social equity (β = 0.13, 95% CI: 0.12 ∼ 0.14) and lower depression (β = −0.06, 95% CI: -0.08 ∼ −0.05) (Figure 3). This indicates that a one-unit increase in subjective social status corresponded to a 0.13-standard-deviation increase in sense of social equity and corresponded to a 0.06-standard-deviation decrease in depression score. Meanwhile, the mediation analysis demonstrated that the sense of social equity was also directly and negatively associated with depression (β = − 0.19, 95% CI: -0.22 ∼ −0.16). Second, we found a statistically significant indirect effect of SSS on depression through sense of social equity (β = − 0.03, 95% CI: −0.03 ∼ −0.02), which accounted for approximately 33% of the total effect (β = −0.09, 95% CI: −0.10 ∼ −0.07). These results suggest that sense of social equity partially mediates the relationship between SSS and depression.

This article has several limitations: first, as a cross-sectional study, it only identifies the associations between SSS, sense of social equity and depression, without verifying causal relationships. Second, SSS and sense of social equity were measured via a single item, which may not capture their multi-dimensional nature. Future research should adopt multi-dimensional measures and longitudinal data to explore causality. Third, due to data constraints, workplace factors such as discrimination and regional economic disparities were not included. Future studies should address these to better inform employee mental health interventions. Fourth, 6,472 observations were excluded due to age criteria or missing key variables. A comparative analysis revealed no significant differences in SSS or depression between the included and excluded samples. However, demographic characteristics such as age, marital status, and education differed significantly, which may still limit the generalizability of the findings.

This study elucidates the significant association between SSS and depression among working-age populations, while identifying sense of social equity as a critical mediating mechanism. Theoretically, our findings establish SSS as an independent psychosocial predictor of depression in the labor-force. Considering the sense of social equity as a mediator clarifies the mechanism linking SSS to depression, enriching our grasp of psychosocial pathways. Practically, our findings underscore that interventions targeting SSS and sense of social equity could mitigate depression across genders, offering a gender-neutral approach to mental health support in the workforce.

First, social comparison environments - particularly workplaces–should be optimized by reducing status-anxiety triggers; this can be done by avoiding public hierarchical evaluations and narrowing visible gaps in income or opportunities. This reduces psychological stress from “relative disadvantages”. Second, sense of social equity should be strengthened. Improving workplace fairness (e.g., transparent promotion systems, reduced discrimination) helps those with low SSS feel less injustice, easing life dissatisfaction and psychological distress, and breaking the “low SSS → perceived injustice → depression” chain.