The improvement path of depression and anxiety among adult women in Shanxi Province, China: a fuzzy-set qualitative comparative analysis

Abstract

Background: Depression and anxiety (D&A) are currently recognized as complex and prevalent mental disorders that pose major threats to mental health. Women are more susceptible to D&A than men.

Methods: We collected data from female participants in Shanxi Province between November 2021 and March 2022 through on-site investigations and an online survey. The survey collected information on sociodemographic traits, lifestyle factors, and physical and mental health. The degree of D&A was evaluated using the Center for Epidemiological Studies Depression Scale (CESD-10) and the Generalized Anxiety Disorder Assessment Scale (GAD-7). We assessed the impact of these factors on D&A symptoms among women using regression and fuzzy-set qualitative comparative analysis (fsQCA).

Results: D&A symptoms had many common influencing factors. Regression analysis identified key protective factors against D&A, including better self-rated health (Depression: OR = 0.11, 95% CI = 0.03-0.47; Anxiety: OR = 0.11, 95% CI = 0.02-0.57) and the absence of recent illness (Depression: OR = 0.56, 95% CI = 0.38-0.83; Anxiety: OR = 0.49, 95% CI = 0.35-0.70). Age exhibited marginal protective effects for both conditions (OR = 0.99, 95% CI = 0.98-1.00). In contrast, occupational stress constituted a significant risk factor, substantially increasing the likelihood of depression (OR = 2.66, 95% CI = 1.43-4.96) and anxiety (OR = 2.99, 95% CI = 1.43-4.96). FsQCA analysis did not identify the conditions for ideal mental health (all consistency < 0.9). However, it did identify eight condition configurations predicting mental health (absence of depression symptoms), each achieving consistency ≥0.87. Additionally, two distinct configurations explained resilience to anxiety (consistency ≥0.80). All configurations met fsQCA's consistency requirements, with self-rated health (present in 10/10 pathways), social support (9/10), and marital status (9/10) playing important roles in most configurations.

Conclusion: Women's mental health faces significant challenges, with D&A being closely intertwined. FsQCA did not identify any specific condition for the absence of D&A symptoms. However, it revealed multiple pathways to mental well-being, highlighting the need for personalized, multifactorial interventions rather than a one-size-fits-all approach. Regression and fsQCA complement each other, offering unique strengths, and their combined insights should be widely applied to broader research and practice.

Keywords: adult women; anxiety; depression; fuzzy-set qualitative comparative analysis; regression.

Figure 1

Relationship between depression symptoms and anxiety symptoms. (a) Scatter density plot of depression scale scores and anxiety scale scores. (b) Histogram of depression scale scores. (c) Histogram of Anxiety Scale Scores.

Figure 2

Forest plot of logistic regression analysis identifying factors associated with depression symptoms. The plot presents odds ratios (ORs) with 95% confidence intervals (CIs) for each predictor variable. Additionally, the distribution of influencing factors is displayed, with categorical variables represented as frequencies (%) and continuous variables as means ± standard deviations (SD), providing a comprehensive comparison between the depression and non-depression groups. ^R denotes the Reference category, indicating its role as the control group. The coefficients (parameter estimates) of the remaining groups are analyzed in relation to this control group.

Figure 3

Forest plot of logistic regression analysis identifying factors associated with anxiety. The plot presents odds ratios (ORs) with 95% confidence intervals (CIs) for each predictor variable. Variables with ORs greater than 1 indicate an increased likelihood of anxiety, while those with ORs less than 1 suggest a protective effect. ^R denotes the Reference category, indicating its role as the control group. The coefficients (parameter estimates) of the remaining groups are analyzed in relation to this control group.