Individual and regional association between socioeconomic status and uncertainty stress, and life stress: a representative nationwide study of China

Life stress refers to the persistent daily worries in one’s life. Life stress could be related to a poor living situation, health conditions, interpersonal relationships and others [3, 6]. Most studies on mental stress and social disparities emphasize the deleterious impact of life stress. Life stressors are objective occurrences of external challenges to an individual’s coping reservoir. Uncertainty stress, on the other hand, damages mental wellbeing by challenging one’s capacity to predict and plan in such a way as to be able to act efficaciously. Compared to generic life stress, its coping requires more psychological resources because of the nature of its trigger. Uncertainty is directly related to important predictors of mental health such as self-efficacy and locus of control, which can be severely constrained when the origin of and solutions to the stress are ambiguous.

Although some scholars argue that stressors’ controllability and predictability (the lack of which leads to uncertainty stress) can sometimes be difficult to operationalize [7], a host of evidence now supports the assertion that uncertainty constitutes a powerful stressor [8]. For example, drawing upon the theories of control and defense mechanism, Mirowsky and Ross (1990) found depression to be associated with a feeling of not being in control of good and bad outcomes [9]. The stress-diathesis theory also recommends further classifications of generic stress because some stressors are desirable and controllable, while others may exert a negative or chronic influence and are harder to manage [10].

One’s social standing is a powerful determinant of the amount and quality of one’s social support, which mitigates the psychological impact of stress. Such social standing may comprise economic affluence, prestige, and ultimately the power to exercise the will [11]. Social exchange theory conceptualizes coping behaviors in response to uncertainty stress as structured by the uneven distribution of resources across social positions in a hierarchical system [12]. With fewer available material resources (money, etc.) and symbolic resources (education, prestige, etc.), a person of lower status is more likely be challenged to cope with stress. Those with fewer resources have fewer opportunities, less extensive social networks, less personal freedom, less healthy and safe work conditions, and less confidence in dealing with stress. [13]. Importantly, they have less perceived power to control their lives. The negative impact of a power differential has even been documented within non-human primates. Primates with lower power have demonstrated adverse adrenocortical, reproductive, immunological, and neurobiological functioning [14, 15]. It has been speculated that these same consequences may apply to humans in disadvantaged (less powerful) social positions [16].

In a vertically mobile social hierarchy, young adults tend to hold higher expectations for themselves regarding the development of their future career. As a result, they exert much pressure on themselves and face high expectations by the rest of society. Sorokin has argued that regardless of their objective economic standing, the upward mobile populace tend to have higher levels of stress [17]. The anticipation of any current and future threat of unknown intensity and duration constitutes a potent psychological stimulus [18]. Even when one’s socioeconomic standing has considerably improved, subjugation in a new symbolic and cultural order may still thwart self-esteem [19]. Being increasingly preoccupied with both their academic and professional development, young adults in universities experience an increasing level of uncertainty stress as college education has become a necessity for survival rather than a privilege. Students with lower SES and more challenging environments may have greater exposure to frequent and intense stress, but fewer means to manage stress [20]. This study will examine the association between SES and uncertainty stress along with life stress among college students. Given China’s regional differences in SES, region of residence might also be related to uncertainty stress and life stress. We will examine these associations at both the individual and regional level in this study.

This study reports individual data from students who completed the Global Health Professions Student Survey (GHPSS) on Tobacco Control in China GHPSS (Extended version). Compared to the original version, the extended version included additional health, mental stress, and behavioral items [21]. The survey was conducted between February and July 2013. A detailed description of the study methods can be found in Yang et al. [22]. Regional variables were retrieved from the National Bureau of Statistics database [23].

All data were entered into a database using Microsoft Excel. The data was then imported into SAS (9.3 version) for statistical analyses. Descriptive statistics were calculated to determine the prevalence of life stress and uncertainty stress. Both unadjusted and adjusted methods were considered in the data analyses, and utilized to assess associations between the dependent and independent variables. SAS survey logistic procedures were applied in the unadjusted analysis, using the university as the clustering unit, in order to account for a within-clustering correlation, attributable to the complex sample for unadjusted analysis. Associations were confirmed through application of a multilevel logistic regression model using the SAS GlIMMIX procedure [26]. We started with the Null Model, a two-level (individual and original regions) with random intercepts in building stress multilevel logistic regression models. The constant was the sole predictor in accounting for cross-regional variation in stress. To this base, we added demographic variables and different individual and regional SES variables as fixed main effects to form several multi-level models for evaluating the impact of stress. Only variables significant in the univariate analysis for the total sample were included in the final analysis. All regional and individual variables, with categories, are listed in Tables 2 and 3. The first category for each variable served as the referent in the logistic regression analysis. First, we constructed the first model (mother’s occupation model) which included variables relating to age and mother’s occupation. The second model (family income model), the third model (family location model), and fourth model (original region GDP model) included family income, family location, and original region GDP added to model 1 respectively. These models significantly improved the fit compared with the Null Model. Model fit was assessed using −2 Res Log Pseudo-Likelihood. We assessed the significance of the random parameter variance estimates using the Wald joint t test statistic.

All analyses were weighted. Weights included: (1) sampling weights, as the inverse of the probability of selection, calculated at university, and (2) post-stratification weights, calculated in relation to sex, based on estimated distributions of this characteristic from a national survey [27]. The final overall weights were computed as the product of the above two weights [26]. Unadjusted logistic regression analyses were weighted using the overall participant-level weights. The multilevel analysis was weighted using sampling in regional level, subject-level weights were used post-stratification weights, respectively [26].

The cross-sectional study design is an important limitation of this study; therefore, a causal link between SES and uncertainty stress along with life stress cannot be established with this work. On the other hand, we employed a large sample, and our findings met several criteria for inferring causality, including the strength of some associations, consistent multiple SES variables, regional SES variables being used, and theory supports and plausibility of effect. Future studies need to compile longitudinal data on uncertainty and other stresses. Second, mothers’ and fathers’ occupational group, family location, and regional GDP are only crude measures of SES, more and appropriate indicators will be needed. Third, this work only focused on college students. More research needs to be done on those who are not in college and are still facing significant amounts of uncertainty stress.