Depression and anxiety symptoms and associated factors among community residents in Shenzhen, China: a cross-sectional study

Abstract

Background

Depressive disorders and anxiety disorders are the most common psychiatric disorders. This study aims to investigate and assess the symptoms of depression and anxiety in general community residents in Shenzhen, China, and to explore their associated factors.

Methods

A cross-sectional study was conducted with 1911 permanent community residents from three districts in Shenzhen as subjects. A self-designed structured questionnaire was used to collect basic demographic characteristics of the participants. Depression and anxiety symptoms were collected using Patient Health Questionnaire-9 (PHQ-9) and Generalized Anxiety Disorder-7 (GAD-7), respectively. Univariate and multivariate linear regression analyses were performed using SPSS to identify correlates of depression and anxiety symptoms.

Results

The detection rates of depression and anxiety symptoms among community residents in Shenzhen were 32.6% and 23.5%, respectively. Insomnia and neuroticism were the strongest predictors of depression and anxiety symptoms, and factors such as self-blame, growing up family environment stress, female gender, and alcohol consumption consistently predicted higher levels of depressive and anxiety symptoms. In addition, depressive symptoms among community residents in Shenzhen were associated with problem solving, life events, lie scale, and marital status. Negative emotions, withdrawal, individual exposure to abusive events, positive coping, negative coping, occupation, and frequency of alcohol consumption in the past 12 months, on the other hand, had an impact on anxiety symptoms.

Conclusions

The results of the study demonstrate the overall profile of depression and anxiety among community residents in Shenzhen and discuss the factors associated with their depression and anxiety symptoms. This may be instructive for providing accessible and targeted support and interventions for depression and anxiety symptoms among community residents in Shenzhen to improve their mental health and well-being in life, as well as for other cities undergoing similar rapid changes.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-025-23241-x.

Background

With the accelerated pace of life and increased social pressures, mental health issues are increasingly receiving widespread attention. The 2019 Global Burden of Disease Study confirms that mental disorders continue to rank in the top 10 of the global burden of disease [1]. One in eight people has a mental disorder, which means that a total of 970 million people worldwide have a mental disorder, with anxiety disorders and depressive disorders being the most common [2]. The 2019 China Mental Health Survey (CMHS) shows that in China, the lifetime prevalence of common mental disorders and psycho-behavioral problems is 16.57%, with a trend of increasing year by year, of which anxiety disorders have the Lifetime prevalence of anxiety disorders is the highest at 7.6%, followed by mood disorders at 7.4% [3]. The latest data from the 2021 CMHS shows that depressive disorders have the highest burden of disease among mental disorders, and that patients with depressive disorders suffer from significant impairment of social functioning, but the utilization rate of health services is very low, with the survey data showing that the lifetime prevalence of depressive disorders in adults is 6.8%, and only 0.5% of patients receive adequate treatment with a shortage of more than 20,000 physicians [4].

As a representative of China’s reform and opening-up frontier cities, Shenzhen’s rapid urbanization and economic development have brought about a great transformation of the social structure. This rapid development has not only changed the face of the city, but also profoundly affected the lifestyle and psychological state of its citizens. As a typical immigrant city, Shenzhen has a high proportion of non-indigenous population, especially migrant workers, and the problems they face, such as difficulties in adaptation and lack of social support, are more prominent [5–9]. Frequent population mobility, special population structure, fast pace of life, and high cost of living all have a potential impact on the mental health of the residents [10, 11], and the results of the Second Epidemiological Survey of Mental Diseases in Shenzhen in 2005 showed that the lifetime prevalence rate of mental diseases among adult residents in Shenzhen was 21.87%, which ranked the first in China [12].

The mental health status of community residents is directly related to the harmony and stability of society. Given the unique position of depressive and anxiety symptoms among the increasingly severe mental health issues, and considering Shenzhen’s distinctiveness within China, this study aims to explore the depressive and anxiety symptoms and their associated factors among community residents in Shenzhen. A population-based cross-sectional survey was conducted to assess the impact of various factors, including sociodemographic characteristics, insomnia, personality traits, coping traits, psychosocial stress, and lifetime stress events, on depressive and anxiety symptoms.

The selection of variables for this study integrated the biopsychosocial model with the stress-quality theoretical framework. The inclusion of socio-demographic variables was guided by the Social Determinants of Health theory [13], which emphasises the gradient effects of socio-economic status, educational level and material conditions on mental health. The inclusion of insomnia as a physiologically mediated variable was based on the mechanism of bidirectional sleep-mood regulation proposed by the hyperarousal theory [14]. The introduction of personality traits and coping styles (Lazarus Coping Theory, 1984) responds to the diathesis-stress model [15–17], and these individual difference variables are thought to modulate the process of transformation of stressors into psychological symptoms. Measures of stress-related variables (psychosocial stress, lifetime stressful events) were based on McEwen’s physiological model of chronic stress [18], which emphasises the cumulative effects of chronic stress load on the neuroendocrine system. This multisystemic measurement design aims to capture the biopsychosocial interaction mechanisms of depression and anxiety symptoms.

This study is intended to provide valuable insights for government departments and related organizations to develop targeted mental health interventions, as well as to serve as a reference for other cities undergoing similar rapid changes.

Methods

Study design and study population

This study used convenience sampling method to recruit permanent community residents who had lived in three districts of Shenzhen City for at least 6 months between January and March 2021. All subjects signed a written informed consent before participating in the study and met the following inclusion and exclusion criteria. Inclusion criteria included: (1) age distribution of 16–85 years old; (2) living in Shenzhen for at least 6 months; and (3) providing informed consent. The exclusion criteria included: (1) participants with serious diseases or mental disorders, etc., who could not complete the survey; (2) participants who refused to sign the informed consent form after repeated explanations by the researcher.

Based on the estimation of the mental health literacy level of 12% of the Chinese population in 2018, we used G*Power software to calculate the sample size to ensure statistical significance. The statistical validity of the study was defined as 0.95, the significance level was set at 0.05, and the minimum sample size was calculated to be 496 people. Considering invalid questionnaires, etc., a total of 1987 residents were finally invited to participate in this study and completed the basic demographic and mental health status questionnaires.76 subjects were excluded due to incomplete answers to the questionnaires, and finally 1911 subjects were included for data analysis with a compliance rate of 96.18%. Comprehensive mental health education was provided to all eligible participants, and those with abnormal results were counseled and recommended for referral to the nearest mental health facility for medical intervention.

Prior to the formal survey, the research team recruited investigators experienced in the field of mental health and provided them with a structured training program. The training covered the purpose of the study, community outreach strategies, appropriate questionnaire administration techniques, the importance of accurate assessments, and the study process. Investigators used multiple channels (posters, community health centers, and the media) to reach out to the selected communities and solicit the understanding, concern, and cooperation of community residents. The investigator then clarified the purpose, methodology and remuneration of the survey to the respondents with their consent. Given the sensitive nature of the inquiries involved in mental health surveys, we ensured that all participants completed the survey in a confidential one-on-one setting.

Measurements

A self-designed structured questionnaire was used to collect the basic demographic characteristics of the participants and to quantitatively assess their mental health-related status indicators, qualitative indicators and environmental stress indicators. Among the status indicators, the Patient Health Questionnaire-9 (PHQ-9) and Generalized Anxiety Disorder-7 (GAD-7) were used to assess symptoms of depression and anxiety, and the Insomnia Severity Index (ISI) was used to assess the subjects’ sleep disordered nature and symptoms. Qualitative indicators included personality factors, behavioral coping traits, which were collected using Eysenck Personality Questionnaire-RSC (EPQ-RSC) and Coping Style Questionnaire (CSQ), respectively. Environmental stress indicators including life events, measures of lifetime stress events were collected using the Psychosocial Stress Survey for Groups (PSSG) and the Self-Compiled Lifetime Stress Events Inventory. Participants were instructed to fill out the questionnaire by a trained investigator, and the completion time was approximately 30 min.

The PHQ-9 uses the diagnostic criteria for depressive disorders as the basis for its program, which focuses on the patient’s experiences in the past 2 weeks [19]. A 4-point (0–3) scale score is taken and summed, with a total score of 0–4 suggesting no depression; 5–9 may have mild depression; 10–14 may have moderate depression; 15–19 may have moderately severe depression; and 20–27 may have major depression [20]. In this study, we used a cutoff score of 5, with scores greater than or equal to 5 indicating the presence of depressive symptoms and scores less than 5 indicating the absence of depressive symptoms. The reliability of this scale in this study was good with a Cronbach' s alpha coefficient of 0.863.

The GAD-7 is used for screening, severity measurement and assessment of generalized anxiety disorder [21]. There are 7 entries to measure the degree of anxiety, which is taken as a summation of scores on a 4-point scale (0–3), with a total score of 0–4 suggesting no anxiety, 5–9 possibly mild anxiety; 10–13 possibly moderate anxiety; 14–18 possibly moderately severe anxiety; and 19–21 possibly severe anxiety [22]. In this study, a score of 5 was used as a critical value, with a score greater than or equal to 5 indicating the presence of anxiety symptoms and a score less than 5 indicating the absence of anxiety symptoms. The Cronbach' s alpha for this scale in this study was 0.896 with good reliability.

The ISI is a simple tool used to screen for insomnia and consists of seven entries, scored on a 5-point Likert scale, which are used to assess the nature and symptoms of the subject’s sleep disorder. Questions address the subjective evaluation of the subject’s sleep quality, including the severity of symptoms, the subject’s satisfaction with his or her sleep patterns, the impact of the degree of insomnia on daily functioning, the subject’s awareness of the impact of insomnia on him or her, and the level of frustration due to the sleep disorder. The total score of the 7-item scale reflects the results of the assessment. a score of 0–7 suggests no clinically significant insomnia; a score of 8–14 may be subclinical insomnia; a score of 15–21 may be moderate clinical insomnia; and a score of 22–28 may be severe clinical insomnia [23]. The scale had good reliability in this study with a Cronbach' s alpha coefficient of 0.842.

The EPQ-RSC is used to assess individual personality traits and is one of the most widely used questionnaires in the fields of medicine, justice, education and counseling. The questionnaire consists of 48 questions and measures four main dimensions: extraversion, neuroticism, psychoticism, and lie scale. A high score on Extraversion indicates that the individual is more outgoing, talkative, and enjoys social activities; a low score indicates that the individual is more introverted, quiet, and prefers to be alone. High scores on neuroticism imply emotional instability, anxiety-prone, and sensitivity; low scores indicate emotional stability, calmness, and less susceptibility to stress. High scores on psychoticism are associated with coldness and lack of empathy; low scores reflect more cooperative and pro-social behavior. Lie scale is used to detect whether subjects try to describe themselves in an idealized way, and high scores may indicate a tendency to give the “right” answer rather than an honest reflection of the self [24]. The Cronbach’s alpha for this scale in this study was 0.905, which is a good reliability.

The CSQ has 62 items, including six subscales: problem solving, self-blame, help-seeking, fantasy, withdrawal, and rationalization. Each item has two answers, and higher scores indicate a greater tendency to use a certain coping style [25]. The questionnaire has good reliability and validity with Cronbach' s alpha of 0.795.

The PSSG reflects a comprehensive subjective assessment of psychological stress over a certain period of time (5 years), and is suitable for group comparison studies, specifically to understand the differences in the total level of stress between groups through the controlled analysis of the total stress score, and to estimate the differences in the level of psychological stress among groups through the various stress factor scores, especially the negative factor scores [26]. The full questionnaire contains 44 entries, and the stress factor entries include life events, negative emotions, positive emotions, negative coping, and positive coping, and the total stress score consists of several of these factors. The Cronbach' s alpha for this scale in this study was 0.904, with good reliability.

The Self-Compiled Lifetime Stress Events Inventory was developed by psychiatrists based on life stress scales such as the PSSG in combination with clinical experience for assessing lifetime stress (Supplementary Table S1). It consists of 30 questions and two major categories of events: 1) Growing up with family environmental stress, such as unnatural death, injury, serious illness, mental abnormality, accidental injury, congenital disability, disappearance, delinquency, drug and alcohol abuse, major crises, unemployment of the primary caregiver, family indebtedness, parental divorce or extramarital affairs or conflicts, major conflicts and contradictions among other family members, loss of family property, family enrichment, family external conflicts, drastic changes in the reputation of the family, and migration of the family; 2) Individuals experience abusive events, such as being physically beaten or assaulted and injured, kidnapped and threatened with ransom, verbally abused and humiliated, sexually harassed, and raped. The Cronbach' s alpha for this scale in this study was 0.897 with good reliability.

Statistical analysis

Statistical analysis was performed using SPSS 22.0. First, a basic descriptive analysis was performed, using percentages and frequencies to describe variables. Second, the data were tested for normality using the Shapiro-Wilk test, and all continuous variables passed the test. The chi-square test was then performed on the categorical variables where appropriate. The t-test, analysis of variance (ANOVA) and Pearson’s correlation test were performed on continuous variables obeying normal distribution, and nonparametric tests such as Wilcoxon rank sum test were performed on non-normally distributed data. Dummy variables were created for categorical variables and the first group of each variable was set as the reference group for regression analysis. Univariate linear regression was used to identify factors associated with depressive symptoms and anxiety symptoms. Variables that were statistically significant in the univariate analysis were included in a multifactor stepwise linear regression model to explore factors associated with depressive and anxiety symptoms. The significance level was set at two-sided α = 0.05, with p < 0.05 considered statistically significant.

Results

Table 1 presents the basic demographic characteristics and mental health status of the participants in this study. Females were predominant (59.9%), and participants were predominantly young and middle-aged (86.7%), with the majority having college/undergraduate degrees and above (63.7%). The majority of the BMIs were in the normal range (59.9%). Occupation was dominated by commercial/service/logistic support workers (21.5%) and professional technician (13.8%). Married persons (61.3%) were the most prevalent, and the majority of participants did not report a history of chronic disease (85.1%). More than half of the participants had a monthly personal income of more than 5,000 yuan (53.7%) and a monthly household income of more than 10,000 yuan (61.7%). Regarding lifestyle, most did not drink alcohol (75.8%) and did not smoke (83.4%). Some participants (21.6%) drank alcohol occasionally in the past 12 months. Participants’ mental health was generally good, although some reported mild to moderate symptoms of depression (31.1%) and anxiety (22.1%), and very few reported moderate to severe and severe symptoms of depression (1.5%) and anxiety (1.4%).

Table 1.

Basic information of participants (n = 1911)

Variables	Category	N	%
Sex	Man	767	40.1
	Woman	1144	59.9
Age (years)	13–19	60	3.1
	20–34	905	47.4
	35–49	751	39.3
	50–64	168	8.8
	65–85	27	1.4
BMI	< 18.5	182	9.5
	18.5 ≤ BMI < 24.0	1144	59.9
	24.0 ≤ BMI < 28.0	476	24.9
	≥ 28.0	109	5.7
Education	Illiterate or minimally literate	7	0.4
	Primary school	33	1.7
	Junior high school	223	11.7
	High school/Vocational secondary school	430	22.5
	College	1143	59.8
	Master and above	75	3.9
Occupation	Civil servant and administrator	142	7.4
	Teacher	37	1.9
	Medical staff	136	7.1
	Professional technician	264	13.8
	Commercial/Service/Logistical support staff	411	21.5
	Farmer	32	1.7
	Laborer	114	6.0
	Military personnel	3	0.2
	Student	112	5.9
	Retirement	36	1.9
	Others	624	32.7
Marriage	Unmarried	650	34.0
	Married	1171	61.3
	Divorced	77	4.0
	Widowed	13	0.7
Suffering from chronic diseases	No	1626	85.1
	Yes	285	14.9
Monthly personal income (yuan)	0-1499	172	9.0
	1500–2999	115	6.0
	3000–4999	597	31.2
	5000–7999	663	34.7
	≥ 8000	364	19.0
Monthly household income (yuan)	0-2999	49	2.6
	3000–5999	199	10.4
	6000–9999	483	25.3
	10,000–20,000	688	36.0
	≥ 20,000	492	25.7
Drinking	Non-drinker	1449	75.8
	Ex-drinker	143	7.5
	Current drinker	319	16.7
Drinking frequency (last 12 months)	Never	1449	75.8
	Occasional (< 3 times per week)	412	21.6
	Regular (3–7 times per week)	27	1.4
	Daily (≥ 7 times per week)	23	1.2
Smoking	Non-smoker	1594	83.4
	Ex-smoker	73	3.8
	Current smoker	244	12.8
Depression symptoms	No depression	1290	67.5
	Mild depression	501	26.2
	Moderate depression	93	4.9
	Moderate to severe depression	22	1.2
	Major depression	5	0.3
Anxiety symptoms	No anxiety	1462	76.5
	Mild anxiety	367	19.2
	Moderate anxiety	56	2.9
	Moderate to severe anxiety	22	1.2
	Major anxiety	4	0.2

In this study, in order to identify potential correlates of depression and anxiety among Shenzhen residents, we analyzed the associations between multiple variables and depression (PHQ-9 score) and anxiety (GAD-7 score) by univariate linear regression. The results, as shown in Table 2, showed that the severity of depressive and anxiety symptoms among Shenzhen residents varied across gender, age, occupation, marital status, history of chronic illness, monthly personal income, alcohol consumption, insomnia symptoms, personality factors, coping traits, life events, and lifetime stressors. In addition, anxiety symptoms were associated with height and whether or not one smoked.

Table 2.

Significant factors of univariate linear regression results of depression and anxiety symptoms

Variables	Category (reference group)/ Dimension	PHQ-9 score					GAD-7 score
		B	95%CI		t	p	B	95%CI		t	p
			Lower	Upper				Lower	Upper
Sex	Male
	Female	0.413	0.076	0.750	2.405	0.016	0.458	0.159	0.757	3.003	0.003
Age (years)		-0.045	-0.060	-0.029	-5.618	< 0.001	-0.028	-0.041	-0.014	-3.904	< 0.001
Height		-0.009	-0.030	0.011	-0.889	0.374	-0.020	-0.038	-0.002	-2.132	0.033
Occupation	Civil servant and administrator
	Professional technician	0.762	0.012	1.511	1.993	0.046	0.977	0.313	1.641	2.886	0.004
	Student	1.475	0.564	2.385	3.177	0.002	1.521	0.715	2.328	3.701	< 0.001
Marriage	Unmarried
	Married	-1.221	-1.570	-0.871	-6.853	< 0.001	-0.664	-0.977	-0.351	-4.165	< 0.001
Suffering from chronic diseases	No
	Yes	1.063	0.601	1.524	4.513	< 0.001	0.814	0.403	1.225	3.887	< 0.001
Number of chronic diseases		0.860	0.517	1.203	4.915	< 0.001	0.781	0.476	1.086	5.028	< 0.001
Monthly personal income (yuan)	0-1499
	3000–4999	-0.699	-1.324	-0.073	-2.191	0.029	-0.726	-1.281	-0.171	-2.565	0.010
	5000–7999	-0.621	-1.239	-0.003	-1.970	0.049	-0.510	-1.059	0.039	-1.822	0.069
Drinking	Non-drinker
	Current drinker	0.934	0.489	1.379	4.115	< 0.001	0.455	0.059	0.852	2.251	0.024
Drinking frequency (last 12 months)	Never
	Occasional (< 3 times per week)	0.553	0.151	0.955	2.696	0.007	0.104	-0.254	0.463	0.572	0.568
	Regular (3–7 times per week)	0.998	-0.402	2.398	1.398	0.162	1.418	0.171	2.664	2.230	0.026
	Daily (≥ 7 times per week)	1.969	0.455	3.483	2.550	0.011	0.418	-0.931	1.766	0.607	0.544
Insomnia		0.526	0.495	0.557	32.851	< 0.001	0.457	0.429	0.485	31.760	< 0.001
EPQ-RSC	Extraversion	-0.070	-0.086	-0.053	-8.401	< 0.001	-0.046	-0.060	-0.031	-6.158	< 0.001
	Neuroticism	0.204	0.190	0.218	29.065	< 0.001	0.193	0.181	0.204	31.784	< 0.001
	Lie scale	-0.118	-0.133	-0.102	-14.700	< 0.001	-0.086	-0.100	-0.072	-11.879	< 0.001
CSQ	Problem solving	-0.060	-0.077	-0.044	-7.233	< 0.001	-0.036	-0.050	-0.021	-4.773	< 0.001
	Self-blame	0.162	0.147	0.177	21.423	< 0.001	0.145	0.132	0.158	21.559	< 0.001
	Help-seeking	-0.066	-0.082	-0.050	-7.966	< 0.001	-0.045	-0.059	-0.030	-6.015	< 0.001
	Fantasy	0.108	0.093	0.124	13.441	< 0.001	0.106	0.092	0.120	14.904	< 0.001
	Withdrawal	0.101	0.085	0.117	12.445	< 0.001	0.087	0.073	0.102	12.096	< 0.001
	Rationalization	0.081	0.064	0.097	9.787	< 0.001	0.072	0.058	0.087	9.914	< 0.001
PSSG	Life events	0.509	0.448	0.569	16.547	< 0.001	0.452	0.399	0.506	16.570	< 0.001
	Negative emotion	0.460	0.398	0.523	14.520	< 0.001	0.481	0.427	0.535	17.449	< 0.001
	Positive emotion	0.546	0.441	0.651	10.159	< 0.001	0.603	0.511	0.696	12.826	< 0.001
	Negative coping	0.112	0.007	0.218	2.086	0.037	0.125	0.031	0.219	2.612	< 0.001
	Positive coping	0.751	0.680	0.822	20.685	< 0.001	0.710	0.648	0.773	22.362	< 0.001
	Stress score	0.073	0.060	0.085	11.595	< 0.001	0.072	0.061	0.083	13.003	< 0.001
Lifetime Stress Events	Growing up family environment stress	0.443	0.376	0.511	12.891	< 0.001	0.394	0.334	0.454	12.891	< 0.001
	Individual exposure to abusive events	0.812	0.668	0.956	11.079	< 0.001	0.736	0.609	0.864	11.316	< 0.001

Only the variables that are significantly correlated in the univariate linear regression are listed in the table, and the categorical variables are listed together with the first group selected as the reference group

PHQ-9, Patient Health Questionnaire-9; GAD-7, Generalized Anxiety Disorder-7; EPQ-RSC, Eysenck Personality Questionnaire-RSC; CSQ, Coping Style Questionnaire; PSSG, Psychosocial Stress Survey for Groups

Bold values indicate statistically significant

Women were significantly higher than men on both depression and anxiety scores (PHQ-9: B = 0.413, p = 0.016; GAD-7: B = 0.458, p = 0.003). Age was negatively associated with depression and anxiety (PHQ-9: B = -0.045, p < 0.001; GAD-7: B = -0.028, p < 0.001). Married individuals had significantly lower depression and anxiety scores than unmarried individuals (PHQ-9: B = -1.221, p < 0.001; GAD-7: B = -0.664, p < 0.001). Individuals with chronic conditions had significantly higher depression and anxiety scores (PHQ-9: B = 1.063, p < 0.001; GAD-7: B = 0.814, p < 0.001), and with each additional chronic condition, depression and anxiety scores increased accordingly (PHQ-9: B = 0.860, p < 0.001; GAD-7: B = 0.781, p < 0.001). In addition, the student (PHQ-9: B = 1.475, p = 0.002; GAD-7: B = 1.521, p < 0.001) and professional technician (PHQ-9: B = 0.762, p = 0.046; GAD-7: B = 0.977, p = 0.004) groups were significantly higher than the other occupational categories in depression and anxiety scores. Psychometric indicators such as the insomnia; neuroticism, extraversion, and lie scale scores on the EPQ-RSC; problem-solving, self-blame, help-seeking, fantasizing, withdrawal, and rationalization dimensions on the CSQ; as well as life events, negative emotions, positive emotions, negative coping, positive coping, and total stress scores on the PSSG; and growing up with family environment stress and individuals experiencing abusive events on the lifetime stressors inventory were significantly correlated with the depression and anxiety scores were significantly correlated (p < 0.05). In particular, the ISI total score (PHQ-9: B = 0.526, t = 32.851; GAD-7: B = 0.457, t = 31.760) and neuroticism score (PHQ-9: B = 0.204, t = 29.065; GAD-7: B = 0.193, t = 31.784) showed the strongest associations.

Therefore, these variables were included in a multiple linear regression model to further explore the factors associated with depression and anxiety symptoms among Shenzhen residents. In this study, stepwise regression was employed to screen variables, aiming to identify core predictors from the initially included potential factors while preventing model redundancy. The method automatically selected variables through iterative inclusion and exclusion of non-significant terms, ultimately constructing a parsimonious yet effective explanatory model. The final models retained 10 key variables for depressive symptoms (e.g., neuroticism, and life events) and 13 variables for anxiety symptoms (e.g., negative coping strategies and occupational categories), with adjusted R² values of 0.501 and 0.511, respectively, demonstrating robust explanatory power despite variable reduction (as detailed in Table 3). Although this approach may exclude weakly associated variables, potential screening errors were mitigated through variance inflation factor (VIF < 5) assessment and bootstrap resampling for robustness verification.

Table 3.

Results of multivariate linear regression analysis of depression and anxiety symptoms

Variables	B (s.e.)	95% CI		β	t	p
		Lower	Upper
Depression model (Adjusted R 2  = 0.501, F = 192.530, p  < 0.001)
Insomnia	0.364 (0.016)	0.333	0.396	0.416	22.499	< 0.001
Neuroticism	0.079 (0.009)	0.062	0.096	0.214	9.156	< 0.001
Self-blame	0.040 (0.008)	0.024	0.055	0.107	5.012	< 0.001
Problem solving	-0.032 (0.006)	-0.045	-0.020	-0.088	-5.266	< 0.001
Life events	0.096 (0.028)	0.042	0.150	0.067	3.468	0.001
Lie scale	-0.018 (0.007)	-0.032	-0.005	-0.050	-2.699	0.007
Growing up family environment stress	0.087 (0.028)	0.032	0.141	0.055	3.094	0.002
Female	0.664 (0.131)	0.408	0.920	0.088	5.088	< 0.001
Married	-0.419 (0.128)	-0.669	-0.169	-0.055	-3.284	0.001
Current drinker	0.451 (0.17)	0.116	0.785	0.046	2.645	0.008
Anxiety model (Adjusted R2 = 0.511, F = 154.752, p < 0.001)
Neuroticism	0.097 (0.008)	0.083	0.112	0.298	12.828	< 0.001
Insomnia	0.299 (0.014)	0.271	0.327	0.384	21.119	< 0.001
Negative emotion	0.097 (0.025)	0.049	0.146	0.075	3.909	< 0.001
Growing up family environment stress	0.061 (0.026)	0.010	0.112	0.044	2.334	0.020
Self-blame	0.033 (0.008)	0.017	0.049	0.102	4.126	< 0.001
Withdrawal	-0.020 (0.007)	-0.034	-0.006	-0.061	-2.836	0.005
Individual exposure to abusive events	0.125 (0.055)	0.018	0.232	0.043	2.286	0.022
Positive coping	0.085 (0.036)	0.014	0.157	0.055	2.350	0.019
Negative coping	-0.088 (0.036)	-0.158	-0.018	-0.042	-2.468	0.014
Female	0.461 (0.112)	0.241	0.681	0.069	4.111	< 0.001
Professional technician	0.407 (0.153)	0.106	0.708	0.043	2.653	0.008
Ex-drinker	-0.534 (0.204)	-0.934	-0.134	-0.043	-2.619	0.009
Regular drinking	1.116 (0.451)	0.232	2.001	0.040	2.475	0.013

Only the variables retained in the model are listed in the table, with decreasing levels of significance from top to bottom. The reference group of categorical variables (the first group of categorical variables) is not presented in the table

As shown in Table 3, the depression multiple linear regression model had an F-value of 192.530 and an adjusted R² of 0.501, indicating that it explained 50.1% of the variance in PHQ-9 scores. Ten variables were significantly associated with PHQ-9 scores: insomnia, neuroticism, self-blame, problem solving, life events, lie scale, growing up family environment stress, gender, marital status, and whether or not alcohol was consumed. Table 3 shows that for each unit increase in ISI total score, neuroticism, self-blame, life events, and upbringing family environment stress, the severity of depressive symptoms increased by 0.364, 0.079, 0.040, 0.096, and 0.087 units, respectively. Problem solving, and lie scale decreased by 0.032 and 0.018 units of depressive symptom severity for each unit increase, respectively. Depressive symptom severity was 0.664 units higher for female participants than for male participants. Depressive symptom severity was 0.419 units lower in married participants than in unmarried participants. The severity of depressive symptoms was 0.451 units higher for alcohol drinkers than for non-drinkers.

As shown in Table 3, the multiple linear regression model for anxiety had an F-value of 154.752 and an adjusted R² of 0.511, indicating that it explained 51.1% of the variance in GAD-7 scores. Thirteen variables were significantly associated with GAD-7 scores: neuroticism, insomnia, negative affect, growing up family environment stress, self-blame, withdrawal, individual exposure to abusive events, positive coping, negative coping, gender, occupation, whether or not they drank alcohol, and frequency of alcohol use in the past 12 months. Table 3 shows that for each unit increase in neuroticism, insomnia, negative mood, growing up family environment stress, self-blame, individuals experiencing abusive events, and positive coping, the severity of anxiety symptoms increased by 0.097, 0.299, 0.097, 0.061, 0.033, 0.125, and 0.085 units, respectively. For each unit increase in withdrawal and negative coping, anxiety symptom severity decreased by 0.020 and 0.088 units, respectively. Anxiety symptom severity was 0.461 units higher for female participants than for male participants. The severity of anxiety symptoms was 0.407 units higher for professional technician than for civil servants/administrators. The severity of anxiety symptoms was 0.534 units lower for those who had quit drinking than for those who had not consumed alcohol. Regular alcohol drinkers in the past 12 months were 1.116 units higher than non-drinkers.

Discussion

This study investigated the symptoms of depression and anxiety in 1911 permanent community residents in Shenzhen and explored their associated factors. This study found that some of the Shenzhen community residents had mild to moderate symptoms of depression (31.1%) and anxiety (22.1%), and a very small proportion of the population reported moderate to severe and severe symptoms of depression (1.5%) and anxiety (1.4%), which was higher than the average prevalence rates of depression and anxiety in the national population as reported by the 2019 CMHS. Further analysis revealed that insomnia, neuroticism, self-blame, growing up with family environment stress, gender, and whether or not they drink alcohol are important influencing factors for depression and anxiety among Shenzhen community residents. In addition, depressive symptoms among Shenzhen community residents were also related to problem solving, life events, lie scale, and marital status. Negative emotions, withdrawal, individual exposure to abusive events, positive coping, negative coping, occupation, and frequency of alcohol consumption in the past 12 months, on the other hand, had an impact on their anxiety symptoms. Our findings demonstrate the overall profile of depression and anxiety among community residents in Shenzhen and shed light on the factors associated with their depression and anxiety symptoms, providing valuable insights for government departments and related organizations to develop targeted mental health interventions, as well as a reference for other cities undergoing similar rapid changes.

The results of this study show that the mental health status of Shenzhen community residents is not optimistic, with about 1/3 of Shenzhen community residents having symptoms of depression (32.6%) and anxiety (23.5%), which is higher than the average prevalence of depression and anxiety in the national population as reported by the 2019 CMHS. A study in 2023 showed that only about 1/5 of Shanghai residents suffered from symptoms of depression (18.76%) and anxiety (11.57%) [27]. A study in 2024 showed that only less than 1/5 of Beijing community residents suffered from symptoms of depression (13.79%) and anxiety (19.75%) [28]. Compared with other economically developed cities in China, the prevalence of depression and anxiety symptoms is higher in Shenzhen, suggesting that rapid urban change and development have had a greater impact on the psychology of Shenzhen residents, rather than just being caused by factors such as economic prosperity and cost of living. In 2018, the foreign floating population of Shenzhen was about 8.48 million, accounting for 65.1% of the total population of Shenzhen, and as one of the pilot cities for the development of a social psychological service system in China, the foreign population in labor-intensive industries has received great attention in the construction of the public service system [29]. A study in 2020 showed that the overall detection rates of depression and anxiety symptoms among migrant workers were 27.85% and 19.26%, respectively, which were lower than the results of community residents who were the focus of this study [30]. This is different from the conventional understanding that migrant populations are more prone to psychological problems due to insufficient social support such as education, employment, and medical care [31, 32], which may be related to Shenzhen’s unique environment, policies, and social welfare. We hypothesize that the high house price pressure faced by household residents (Shenzhen’s house price-to-income ratio is the highest in the country) and the existential anxiety of staying and developing in Shenzhen may have offset the social welfare advantages of household registration; and that Shenzhen’s more open-mindedness gives more support to migrant workers than other first-tier cities such as Beijing and Shanghai, resulting in a unique structure of psychological risk. Overall, the depression and anxiety situation of community residents in Shenzhen is not optimistic, and both migrant workers and community residents need more attention and support.

Further, we explored correlates of depression and anxiety symptoms to deepen understanding. Insomnia, neuroticism, coping styles predisposed to self-blame, growing up family environment stress, being female, and alcohol consumption consistently predicted higher levels of depression and anxiety symptoms. Insomnia is not only a common symptom of depression and anxiety, but may also be a potential risk factor for them. Shenzhen’s tech-industry ecosystem, where overtime is prevalent, may lead to greater work stress, poorer rest and more widespread circadian rhythm disruption, which can lead to sleep problems, suggesting that strict control of working hours should be incorporated into the city’s mental-health intervention system. Chronic insomnia can further exacerbate mental health problems by leading to mood instability, poor concentration, and impaired cognitive functioning. Therefore, early identification and intervention of insomnia is crucial for the prevention and treatment of depression and anxiety disorders, as demonstrated in many previous studies [33–36]. Individuals with high neuroticism scores tend to be more anxious and upset, more sensitive to negative events, and prone to negative emotions [37]. Neuroticism, as a stable personality trait, may increase an individual’s risk of developing depression and anxiety disorders by affecting his or her emotion regulation ability and coping strategies [38]. Therefore, psychological interventions for people with high neuroticism scores should pay special attention to the development of emotion regulation training and coping skills. Self-blame is a negative cognitive pattern that manifests itself as excessive guilt and self-criticism by individuals for their failures or mistakes [39]. This cognitive bias not only exacerbates depressive symptoms, but may also lead to a vicious cycle that plunges individuals into deeper emotional distress. In psychotherapy, cognitive-behavioral therapy has been shown to be effective in changing this negative cognitive pattern and helping individuals develop a more positive and healthy self-evaluation [40]. Adverse family environments such as strained parental relationships and domestic violence can have a profound impact on children’s psychological development and increase the risk of depression and anxiety disorders in adulthood [41]. Early interventions such as family education and support services can help improve the family environment and promote the development of children’s mental health [42]. Females were significantly higher than males in both depression and anxiety scores, a finding consistent with previous research [43]. Females may face more social role stress, hormonal fluctuations, and higher sensitivity, all of which increase their risk for depression and anxiety [44]. In addition, females are more inclined to seek help and support, which may result in them being more likely to report mental health problems [45]. Therefore, future research and interventions should pay special attention to the mental health needs of women. Alcohol consumption was significantly associated with symptoms of depression and anxiety, with drinkers more likely to be depressed relative to non-drinkers, while frequency of alcohol consumption in the past 12 months was significantly and positively associated with symptoms of anxiety. Drinking behavior may reflect individuals’ lack of effective coping strategies and social support networks, and those who tend to cope with stress by drinking may feel more helpless and anxious when facing life challenges [46]. The phenomenon that those who have quit drinking have significantly lower anxiety scores than nondrinkers may seem counterintuitive, but there are still studies that support this finding to some extent [47]. We hypothesize that abstainers tend to change other poor lifestyle habits during the process of abstinence, such as starting a regular routine, eating a balanced diet, and exercising moderately. These patterns of healthy behavior have a positive impact on mental health and can significantly reduce anxiety symptoms. At the same time, successful abstinence from alcohol is a significant personal achievement that enhances an individual’s self-confidence and self-efficacy. This positive mental state also helps to reduce anxiety and other negative emotions.

In addition, depressive symptoms were significantly associated with coping styles that tend to solve problems, life events, lie scale, and marital status. Good problem-solving skills help individuals to effectively cope with various challenges in their lives and reduce feelings of helplessness and hopelessness, thereby reducing the risk of depression [48]. On the contrary, lack of effective problem-solving skills may lead individuals to feel powerless in the face of difficulties, which in turn increases the likelihood of depression. Therefore, developing and improving individuals’ problem-solving skills should be one of the important components of mental health interventions. Major life events such as unemployment and divorce often trigger strong emotional reactions and increase the psychological burden on individuals. Prolonged exposure to high-stress life events may lead to the gradual accumulation of negative emotions in individuals, which eventually develops into depression [49]. Therefore, recognizing and coping with psychological shocks from life events in a timely manner is important for the prevention and treatment of mental health problems.

Lie scale was negatively associated with depressive symptoms, contrary to our expectation, probably because in the Chinese cultural context, where collectivist values emphasize harmony and face preservation, individuals may pay more attention to the feelings of others and social harmony, and tend to mask their negative emotions in order to avoid causing distress to others. This cultural trait may lead individuals high in lie scale to choose to internalize and deal with difficulties when facing them instead of expressing negative emotions openly, thus reducing negative evaluations and impacts on them from outside. Research suggests that in collectivist cultures, people are more inclined to use inhibitory strategies to cope with stress, which may be associated with lower levels of depression [50]. Individuals who are high in lie scale may have greater psychological resilience and more sophisticated coping strategies, and are better able to adjust their mindset in the face of life’s challenges [51]. They learn how to control their emotional responses in appropriate situations, which not only helps to maintain their personal mental health, but also makes them more adaptable in complex social environments. In the long term, this ability to regulate emotions may serve as an important barrier against depression [52]. Although high lie scale may help individuals avoid some negative social evaluations and conflicts in the short term, excessive suppression of true emotions in the long term may also lead to internalized stress accumulation and feelings of emotional detachment [53, 54].

Meanwhile, the unique phenomenon of a negative correlation between lie scale and depressive symptoms is inextricably linked to the unique socio-cultural characteristics of Shenzhen as a rapidly developing city of immigrants. Shenzhen, as an immigrant city rising rapidly after the reform and opening up, has a predominantly young foreign population with a stronger sense of adaptability and independence. In such environments, individuals are more likely to mask their emotions to maintain social relationships and work efficiency, and to avoid affecting their career development or social networks by revealing their emotions; Beijing and Shanghai, as traditional central cities with a higher proportion of local residents and more stable social relationships, are likely to be more dependent on established social support systems for the expression of their emotions, and have a lower need for concealment. Shenzhen’s entrepreneurial culture and fast-paced life reinforce pragmatic values, and emotion management is seen as part of professionalism. By masking negative emotions, individuals may be able to integrate more easily into high-pressure environments, reducing the frustration caused by emotional exposure and thus indirectly lowering the risk of depression; Beijing’s attributes as a political and cultural center and Shanghai’s international business traditions may place more emphasis on the emotional connection of interpersonal relationships, and emotional suppression may, on the contrary, exacerbate the feeling of loneliness. Shenzhen’s community structure is dominated by a “society of strangers,” with residents relying more on their professional circles than on traditional family or neighborhood relationships. The masked personality may help individuals maintain superficial stability in scenarios that lack deep emotional support and avoid isolation by exposing vulnerability. Therefore, it is equally important for mental health interventions to develop moderate emotional expression and to establish sincere interpersonal relationships.

Married individuals had significantly lower depression and anxiety scores than unmarried individuals. It has been shown that marital relationships not only provide emotional support, but may also bring financial stability and supportive social networks, which can help relieve stress and enhance mental health [55]. In contrast, those who are unmarried, divorced or widowed may lack such support systems and are more likely to experience mental health problems [56].

Anxiety symptoms were significantly associated with negative emotional experiences of life events and their positive and negative coping, coping styles that tend to be withdrawn, the individual’s exposure to abusive events, and occupation. Individuals experiencing more negative emotions (e.g., anger, sadness, or fear) have elevated levels of anxiety. Prolonged exposure to negative emotions not only reflects sensitivity to stressors and weaker coping skills, but may also create a vicious cycle that further exacerbates anxiety symptoms [40]. Positive coping was positively associated with anxiety, and negative coping and coping styles that tend to be withdrawn were negatively associated with anxiety, contrary to our general understanding. For example, a Shanghai-area study showed that active coping was protective against anxiety [27]. We hypothesize that while positive coping is generally a beneficial way of coping, in some cases, overuse of this style may instead lead to increased anxiety, such as an excessive desire for control over certain situations or events. Thus, we recommend emphasizing “strategic flexibility” rather than a single positive orientation in interventions. It should also be noted that although negative coping and withdrawal tendencies may temporarily reduce anxiety symptoms, they may not be the most effective coping strategies, and other more positive coping styles may be needed to manage anxiety symptoms in the long term. Individuals may choose to avoid situations to reduce anxiety because they feel powerless or unable to change them, but this approach typically accumulates negative emotions and anxiety and does not provide an effective solution. Individuals who have been abused often experience severe trauma that leads to chronic anxiety and fear [57]. These traumas not only affect the individual’s ability to regulate emotions, but can also lead to serious psychological problems such as post-traumatic stress disorder (PTSD). Therefore, psychological interventions for abused populations should pay special attention to trauma repair and the reconstruction of a sense of security in order to help them regain their psychological well-being. professional technician had significantly higher depression and anxiety scores than civil servants/administrators. For professional technician, their work often involves a high degree of specialized skills and continuous learning requirements, and they are more likely to be exposed to the pressure of a rapidly changing technological environment and technological updates, which increases their psychological burden, as in the case of IT industry practitioners in Shenzhen. Therefore, it is crucial to understand and support the mental health needs of this group, especially by providing career development guidance and psychological counseling services to help them better cope with challenges at work [58].

In the multiple linear regression we did not find a relationship between depression and anxiety symptoms and factors such as age, student population, and history of chronic illness, although they showed significant differences in one or more psychological indicators when using the univariate linear regression. Nonetheless, we should also note the influence of these variables as they have been found to be relevant psychological factors in other studies. For example, age was negatively correlated with depression and anxiety, i.e., symptoms of depression and anxiety tended to diminish with age, a finding reported in a larger number of studies [59, 60]. The student population had significantly higher depression and anxiety scores than civil servants/administrators, which may be related to academic stress and future uncertainty. Academic competition, examination pressure, and social adjustment may cause greater psychological burden on students [61]. Schools and educational institutions should enhance mental health support for students by providing psychological counseling and tutoring services to help them better cope with academic and life challenges. As the number of chronic diseases increased, depression and anxiety scores increased accordingly. Chronic diseases not only bring physical discomfort to patients, but also lead to long-term psychological stress and reduced quality of life. Effective chronic disease management and support services not only improve physical health, but also significantly enhance patients’ mental health [62].

Overall, this study provides an overall picture of depression and anxiety symptoms among community residents in Shenzhen, as well as the risk and protective factors associated with them. Depression and anxiety, as the two most common types of mental disorders, deserve special attention in Shenzhen as a representative of a rapidly developing Chinese city. In this study, although 1/3 of Shenzhen community residents suffered from depression and anxiety disorders, the proportion of severe cases was relatively small, suggesting that timely implementation of effective psychological interventions can have a positive impact, especially the need to focus on Shenzhen’s unique social environment. For example, regular psychological counseling and education for people prone to depression and anxiety, a pilot mental toughness training program in science and technology parks focusing on improving the emotional regulation skills of IT workers, as well as insomnia interventions for them and even for residents in the community; improvement of social welfare to reduce the occurrence and development of negative emotions in migrant workers and the community at large; and alleviation of the pressure of further education and employment to break the the stereotype that “successful people have to be strong”; encouraging positive communication, improving psychosocial skills, building more social support and interpersonal connections, and seeking professional help when necessary. It is worth noting that our regression models for depression and anxiety explained only 50.1% and 51.1% of the variance, respectively, which implies that there are still several correlates that are unexplained (e.g., social support, interpersonal relationships, etc.) and need to be further explored in the future. Many of the more refined potential predictors specific to Shenzhen will also be considered in subsequent studies: such as industry type segmentation (e.g., differences in stressors between IT and other industries), quantitative indicators of housing stress (e.g., mortgage-to-income ratios), and measures of policy perceptions (e.g., satisfaction with the Talent Settlement Policy), etc.

This study was conducted from January to March 2021. Due to the pandemic of novel coronavirus pneumonia (COVID-19) and government policy constraints, random sampling for door-to-door surveys and data collection was extremely challenging and impractical. Instead, a convenience sampling method was used whereby questionnaires were distributed and data collected by adequately trained enumerators. This method, while feasible in practical situations, may compromise the representativeness of the sample and limit the generalizability of the results. The data were considered to be authentic and reliable, reflecting the mental health status of the residents of the Shenzhen community studied.

Inevitably, this study has some limitations. First, convenience sampling may result in an unrepresentative sample, e.g., under-sampling of more transient or isolated groups. Second, this study utilized a cross-sectional design, which limited the identification of causal relationships. Future directions could include longitudinal studies or mixed-methods research to investigate causal mechanisms. Third, the use of self-rating scales to assess depression and anxiety symptoms inevitably introduces potential bias caused by subjective factors. Future studies may consider analyzing a combination of subjects’ self-reports and objective diagnoses by professional physicians. Fourth, this study includes only residents of permanent neighborhoods in Shenzhen, making it difficult to compare differences with other populations in Shenzhen or other urban populations, which limits our conclusions. Finally, depression and anxiety may lead to some risky behaviors such as suicide, and future studies should take into account both the factors influencing depression and anxiety symptoms and the risky behaviors that may result. Nevertheless, this study provides an overall picture of depression and anxiety symptoms among permanent community residents in Shenzhen. It also highlights factors that contribute to the reduction of depression and anxiety symptoms, which can help government departments and related organizations develop targeted mental health interventions, as well as provide valuable insights for other cities undergoing similar rapid changes.

Conclusions

In conclusion, the mental health status of Shenzhen community residents appears concerning. This study found that the detection rates for depression and anxiety symptoms were 32.6% and 23.5%, respectively. Insomnia, neuroticism, self-blame, growing up family environment stress, female gender, and alcohol consumption consistently predicted higher levels of depressive and anxiety symptoms. Our findings may be instructive for providing accessible and targeted support and interventions for depression and anxiety symptoms among community residents in Shenzhen to improve their mental health and well-being, as well as for other cities undergoing similar rapid changes.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

PHQ-9

Patient health questionnaire-9

GAD-7

Generalized anxiety disorder-7

CMHS

China mental health survey

ISI

Insomnia severity index

EPQ-RSC

Eysenck personality questionnaire-RSC

CSQ

Coping style questionnaire

PSSG

Psychosocial stress survey for groups

PTSD

Post-traumatic stress disorder

CI

Confidence interval

Author contributions

HH was responsible for the overall study design, facilitated the implementation of the study, analyzed the data, obtained the results, and wrote the article. KY and ZW assisted in facilitating the implementation of the study. WH and SJ were involved in the data collection. JW designed and implemented the overall study. All authors read and approved the final manuscript.

Funding

This research was funded by the Shenzhen Science and Technology Research and Development Fund for Sustainable Development Project (No. KCXFZ20201221173613036), Shenzhen Fund for Guangdong Provincial High-level Clinical Key Specialties (No. SZGSP013), and the Shenzhen Key Medical Discipline Construction Fund (No. SZXK072), and JW is the PI of the project.

Data availability

The datasets presented in this article are not readily due to privacy or ethical restrictions. Requests to access the datasets should be directed to the corresponding author.

Declarations

Ethics approval and consent to participate

The studies involving humans were approved by the Ethics Committee of Shenzhen Kangning Hospital (KN-2020-1209-1). The studies were conducted in accordance with ethical principles of the Declaration of Helsinki, as well as in compliance with local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

All authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest or non-financial interest in the subject matter or materials discussed in this manuscript.