Can New-type Urbanization Pave a Way against Depression? Evidence from a Quasi-Natural Experiment in China

Abstract

The escalating threat of depression demands urgent action from the research community. As a policy that prioritizes a people-centered approach, the New-type Urbanization Policy (NTU) holds promise for alleviating depression. However, whether and how NTU positively affects individual mental health remains underexplored. This study draws on three waves of data from the China Health and Retirement Longitudinal Survey (CHARLS) and employs the difference-in-differences (DID) method as a quasi-natural experiment to empirically analyze the effects of NTU on individual depression. The results indicate that NTU significantly reduces depression, with environmental pollution serving as a mediator in this relationship. Moreover, NTU’s impact on depression reduction is more pronounced in non-resource-based cities and those with lower population concentration. Additionally, the ecosocial theory emphasizes that health arises from the biological embodiment of structural exposures embedded in social and ecological environments. Based on this theory, it serves as a theoretical framework for analyzing NTU’s impact on depression. This study expands the existing research on pilot policies related to individual health and provides concrete policy recommendations for mitigating depression in the context of NTU implementation.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11524-025-01004-8.

Introduction

Depression is estimated to affect 300 million individuals worldwide, making it a major factor in the global mental health burden [1, 2]. In China alone, over 54 million individuals are affected [2]. In addition to its impact on individual health, depression also profoundly affects families and society [3, 4]. The rising prevalence of depression is even more concerning [5, 6], a trend that further exacerbates health, social, and economic burdens both globally and within China. Therefore, researching depression and its interventions is essential for effectively addressing this challenge [3].

China has undergone rapid urbanization since its reform and opening-up [7–11]. However, traditional urbanization often comes with significant resource consumption and environmental damage, leading to environmental degradation and mental health issues [9–11]. To address these issues, the Chinese government issued the National New-type Urbanization Plan (2014–2020) and, guided by this plan, launched the New-type Urbanization Policy (NTU). This policy reflects a strategic shift toward a people-centered model of urbanization that emphasizes sustainable and coordinated development across economic, social, and environmental dimensions [9, 10, 12]. It aims to comprehensively improve urban living quality through a set of integrated measures, including strengthening infrastructure construction, optimizing urban spatial layout and morphology, promoting green development and energy conservation, advancing urban–rural integration, and enhancing social security and public service provision [12–15]. To promote replicable and scalable models for nationwide implementation, NTU was advanced in three phases through the designation of pilot cities: the first phase was launched in March 2014, followed by the second and third in 2015 and 2016, respectively. Ecosocial theory, proposed by Nancy Krieger in 1994 [16], emphasizes that health arises from the biological embodiment of structural exposures embedded in social and ecological environments [17, 18]. Within this framework, NTU can be viewed as a structural intervention that reshapes these exposures, thereby helping reduce drivers of individual depression. Specifically, NTU’s emphasis on environmental protection plays a crucial role in mitigating ecological hazards such as pollution [13, 19], which constitute environmental exposures harmful to health [20]. Additionally, NTU promotes balanced urban–rural development, improves social security, and enhances public services [14, 15], which helps alleviate social deprivation and, in turn, enhances individuals’ life satisfaction. Moreover, NTU’s emphasis on public service facilities and infrastructure encourages individual participation in social activities [12, 15], which in turn strengthens social connections and helps buffer against psychosocial stressors caused by isolation.

Prior research has mainly examined the macro-level impacts of NTU, such as energy [9, 10, 21], carbon emissions [13, 14, 22, 23], environmental pollution [19, 23], resource pressure [24], and urban resilience [12]. However, limited research has examined NTU’s effects on individual physical and mental health. Additionally, there is a limited body of literature on urban pilot policies targeting individual depression, with only a few studies addressing pilot policies like “Broadband China” [25, 26], “Smart Cities” [4], and “Healthy Cities” [27]. Therefore, it is essential to analyze the relationship between NTU and individual depression. To this end, this study treats NTU as a quasi-natural experiment and employs the difference-in-differences (DID) method to investigate its relationship with individual depression. The empirical results indicate that the implementation of NTU helps alleviate depression, with environmental pollution serving as a mediating factor in this relationship. Additionally, the impact of NTU on individual depression varies by resource-based cities and population concentration.

This study makes the following contributions: from a theoretical perspective, it empirically investigates NTU’s impact on individual depression, filling the gap in the micro-level literature on NTU and expanding the research on pilot policies for alleviating depression. Additionally, by incorporating the ecosocial theory, this study provides a novel theoretical framework to analyze the influence of NTU and environmental pollution on individual depression. From a practical perspective, this study provides concrete policy recommendations for alleviating depression during the implementation of NTU.

Methodology

Data Sources

Data from various sources are used in this study. First, data on the NTU pilot areas are obtained from the National Development and Reform Commission. Second, CESD-10 and individual-level control variables are drawn from the China Health and Retirement Longitudinal Survey (CHARLS). CHARLS is a nationally representative microdata survey in China that began in 2011, and this study utilizes the three waves (2013, 2015, and 2018). Third, data on annual PM2.5 concentrations are obtained from NASA’s Socioeconomic Data and Applications Center (SEDAC) [20, 28, 29]. Finally, city-level control variables and the population concentration variable used in the heterogeneity analysis are sourced from the China City Statistical Yearbook.

In addition, this study deletes only rows containing missing values and balances the panel data to ensure that each individual has complete records for all three years, as detailed in Fig. 1. Ultimately, the empirical analysis is based on 18,066 individual-year observations from 6,022 participants surveyed in 2013, 2015, and 2018. Figure 2 illustrates the distribution of NTU implementation across cities, with additional details provided in Supplementary Table S1.

Fig. 1.

Flowchart of analytic sample selection

Fig. 2.

Distribution of NTU implementation

Definition and Measurement of Variables

Independent Variable

This study uses the dummy variable Treat_j × Post_t as the independent variable, where j denotes city and t denotes year. It takes a value of 1 if the j-th city implements NTU in the current and subsequent years, and 0 otherwise. To avoid overestimating the policy effect, a city is classified as a treated city only if NTU is implemented across the entire prefecture-level city; if not, it is placed in the control group [14].

Dependent Variable

In this study, the dependent variable is individual depression, measured by the Center for Epidemiologic Studies Depression Scale (CESD-10). The CESD-10 is recognized as a reliable and valid instrument for assessing depressive symptoms in older adults. It consists of 10 items, including 8 items related to negative emotions and 2 to positive emotions [4, 30, 31]. During the survey, participants recalled their experiences over the past week and selected one of four frequency categories for each item: rarely or none of the time (< 1 day), some or a little of the time (1–2 days), occasionally or a moderate amount of the time (3–4 days), and most or all of the time (5–7 days). All responses were coded on a 4-point scale ranging from 0 to 3. The total score was calculated by summing these 10 items, with positive emotion items reverse-coded. Therefore, scores range from 0 to 30, with higher scores indicating more severe depression among individuals.

Mediator

The mediating variable is environmental pollution, measured by annual PM2.5 concentrations. Annual PM2.5 concentrations, which refers to particulate matter with a diameter of 2.5 µm or less, is widely used to measure air pollution levels [32]. Additionally, urban air pollution remains a major environmental concern [19, 33] and is closely related to mental health issues [33–36]. Therefore, annual PM2.5 concentrations is used as the mediating variable to examine the mediating role of environmental pollution in the relationship between NTU and individual depression.

The data are derived from the Global Annual PM2.5 Grids, which are based on Aerosol Optical Depth (AOD). It is published by NASA’s SEDAC, which is operated by Columbia University’s Center for Integrated Earth System Information (CIESIN). We then use ArcGIS software to process the Global Annual PM2.5 Grids into annual PM2.5 concentrations for prefecture-level cities in China [20, 28].

Control Variables

Based on previous literature, this study classifies the control variables into two categories to mitigate potential confounding. The first category includes individual factors: 1) Mar represents respondents’ marital stability, with lower scores indicating greater stability [37]. 2) LS represents respondents’ level of life satisfaction, with scores classified into five levels, ranging from 1 (completely dissatisfied) to 5 (completely satisfied) [25, 38]. 3) Edu represents education level, where higher scores indicate greater educational attainment [25, 38]. 4) Res indicates whether the respondent resides in an urban area (1 = Yes, 0 = No) [4]. The second category includes city-level factors: 1) RGDP represents the regional gross domestic product, with its logarithmic value used in the analysis [25]. 2) Pop denotes the total population of the city [25]. 3) RA represents the per capita road area [4]. 4) PIS denotes the percentage of the primary industry in the regional gross domestic product [4]. 5) ES represents educational expenditure, with its logarithmic value used in the analysis. City-level covariates are included to mitigate confounding in the relationship between NTU and annual PM2.5 concentrations. In addition, all covariates are used to reduce potential bias in the associations between NTU and individual depression, and between annual PM2.5 concentrations and individual depression.

Model Specification

Following prior research [12, 26], this study adopts a multi-period DID approach, leveraging the staggered implementation of NTU across cities as a quasi-natural experiment to examine its causal impact on CESD. Formally, we specify:

$${\mathit{CESD}}_{\mathit{ijt}}={\beta }_0+{\beta }_1{\mathit{NTU}}_{\mathit{jt}}+\phi X_{\mathit{ijt}}+{\mu }_i+v_t+{\epsilon }_{\mathit{ijt}}$$ 1

where CESD_ijt represents individual i’s depression level in city j during year t; NTU_jt is a treatment indicator denoting whether the city j implemented NTU in year t; X_ijt represents a set of control variables; μ_i and v_t denote individual and time fixed effects; and ε_ijt is the error term. In model Eq. (1), β₁ reflects the total effect of NTU on CESD and should be significant if NTU influences CESD. To reduce selection bias arising from imbalanced individual characteristics, we apply propensity score matching (PSM) with a logit model based on individual-level covariates [39]. Each treated unit is matched with two untreated units with similar propensity scores using nearest-neighbor matching. We then perform DID regressions on the matched panel data to estimate the causal effect of NTU. Moreover, the traditional two-way fixed effects (TWFE) model may produce biased estimates in the presence of treatment timing heterogeneity, as units treated earlier can be mistakenly used as controls for later-treated units [40, 41]. To assess the robustness of the TWFE estimates, this study further employs the method proposed by Callaway and Sant’Anna (2021) as a complementary estimation approach.

To investigate how NTU impacts CESD, we introduce a mediation effect model [42]:

$${PM2.5}_{\mathit{jt}}={\alpha }_0+{\alpha }_1{\mathit{NTU}}_{\mathit{jt}}+\phi X_{\mathit{ijt}}+{\mu }_i+v_t+{\epsilon }_{\mathit{ijt}}$$ 2

$${\mathit{CESD}}_{\mathit{ijt}}={\gamma }_0+{\gamma }_1{\mathit{NTU}}_{\mathit{jt}}+{\gamma }_2{PM2.5}_{\mathit{jt}}+\phi X_{\mathit{ijt}}+{\mu }_i+v_t+{\epsilon }_{\mathit{ijt}}$$ 3

In models Eq. (2) and Eq. (3), α₁ reflects the effect of NTU on the mediator variable (annual PM2.5 concentrations), and γ₂ reflects the effect of annual PM2.5 concentrations on CESD. If α₁ and γ₂ are both significant, then NTU affects CESD through annual PM2.5 concentrations. Furthermore, if γ₁ is smaller than β₁ and significant, then a partial mediation effect exists; if γ₁ is not significant, then a complete mediation effect exists. To evaluate the significance of the indirect effect, we applied the Sobel test and a bootstrap method with 1,000 resamples. The Sobel test calculates a Z statistic by dividing the estimated indirect effect by its standard error, with significance determined by comparison to the standard normal distribution [43]. The bootstrap method constructs a 95% confidence interval through repeated resampling, and significance is determined if the interval excludes zero [44]. Moreover, to test for potential exposure-mediator interaction, we extend Eq. (3) by adding an interaction term between NTU and annual PM2.5 concentrations:

$$\begin{array}{cc}\hfill {\mathit{CESD}}_{\mathit{ijt}}& ={\theta }_0+{\theta }_1{\mathit{NTU}}_{\mathit{jt}}+{\theta }_2{PM2.5}_{\mathit{jt}}+{\theta }_3({\mathit{NTU}}_{\mathit{jt}}\times {PM2.5}_{\mathit{jt}})\hfill \\ \hfill & +\phi X_{\mathit{ijt}}+{\mu }_i+v_t+{\epsilon }_{\mathit{ijt}}\hfill \end{array}$$ 4

In model Eq. (4), if the interaction term is not significant, the above mediation effect model remains valid. Otherwise, other methods should be used [45].

Hence, by leveraging policy variation in a DID setting and combining it with mediation analysis, we examine NTU’s causal impact on CESD and explore a potential mechanism through annual PM2.5 concentrations.

Results

Benchmark Analysis

We examine NTU’s impact on individual depression, and the descriptive statistics of the core variables are presented in Table 1. The average CESD score was 7.735, with a minority of respondents residing in urban areas. To ensure the reliability of the model, variance inflation factor (VIF) values were examined prior to analysis, and all values were below 10, indicating no multicollinearity.

Table 1.

Descriptive statistics of variables

Variable name	Abbreviation	N	Mean	SD	Min	Max
CESD	CESD	22,049	7.735	6.074	0	30
Marital stability	Mar	22,688	1.450	1.196	1	6
Life satisfaction	LS	22,003	3.279	0.744	1	5
Education level	Edu	22,046	0.884	0.826	0	3
Residence	Res	22,025	0.151	0.358	0	1
Regional GDP	RGDP	22,574	16.88	0.882	14.73	19.61
Population	Pop	22,691	614.1	438.5	120.1	3404
Road area per capita	RA	21,912	12.61	5.385	2.166	37.03
Primary industry share of regional GDP	PIS	22,691	12.51	7.374	0.04	36.43
Educational expenditure	ES	22,691	13.46	0.698	11.93	16.14
PM2.5	PM2.5	22,691	46.93	17.38	14.23	103.6

Abbreviations listed in this table are used for variable names in all subsequent tables. Res is coded as 1 = urban and 0 = rural. GDP = gross domestic product

In Model (1) of Table 2, after controlling for relevant factors that may influence individual depression, along with individual and time fixed effects, NTU demonstrates a significant negative association with individual depression (β = −0.379, p < 0.01). This suggests that NTU helps reduce individual depression. In line with the ecosocial theory, several potential explanations exist for the negative correlation between NTU and individual depression. First, NTU improves the environment by promoting green development and increasing energy efficiency [10, 13], which helps reduce health problems resulting from environmental pollution and other harmful exposures [20]. Additionally, NTU enhances healthcare coverage [15], ensuring timely check-ups and treatment for individuals, both of which may help lower depression levels. Second, NTU promotes urban–rural integration, which helps reduce anxiety related to social deprivation [10, 14] while enhancing social security and public services to improve individuals’ life satisfaction [13]. These measures may help lower the prevalence of depression. Third, NTU’s enhancements in public services and infrastructure further reduce isolation [15]. This can strengthen social connections among individuals, thereby buffering against psychosocial stressors caused by isolation [38].

Table 2.

Results of regression analysis

	(1)	(2a)	(2b)	(3)	(4)	(5)	(6)
Variable	CESD	PM2.5	CESD	CESD	CESD	CESD	CESD
DID	−0.379***	−3.118***	−0.321**	0.107	−0.450**	−0.320**	−0.387***
	(0.129)	(0.181)	(0.130)	(0.204)	(0.187)	(0.130)	(0.129)
PM2.5			0.019**			0.019***
			(0.007)			(0.007)
DID × PM2.5						0.003
						(0.007)
Mar	0.317***	−0.063	0.318***	0.318***		0.318***	0.309***
	(0.080)	(0.094)	(0.080)	(0.080)		(0.080)	(0.081)
LS	−1.270***	0.020	−1.270***	−1.270***		−1.271***	−1.261***
	(0.069)	(0.075)	(0.069)	(0.069)		(0.069)	(0.070)
Edu	0.060	−0.227***	0.065	0.055		0.064	0.066
	(0.071)	(0.080)	(0.071)	(0.071)		(0.071)	(0.072)
Res	−0.297**	0.020	−0.298**	−0.292**		−0.297**	−0.325**
	(0.141)	(0.205)	(0.141)	(0.141)		(0.141)	(0.142)
RGDP	0.176	−0.998*	0.194	0.031		0.212	0.122
	(0.442)	(0.545)	(0.442)	(0.441)		(0.443)	(0.444)
Pop	0.001	−0.006***	0.001	0.000		0.001	0.001
	(0.001)	(0.001)	(0.001)	(0.001)		(0.001)	(0.001)
RA	0.003	0.213***	−0.001	0.012		−0.001	0.004
	(0.018)	(0.024)	(0.018)	(0.018)		(0.018)	(0.018)
PIS	0.001	0.456***	−0.008	−0.001		−0.007	−0.003
	(0.030)	(0.036)	(0.031)	(0.031)		(0.031)	(0.031)
ES	−0.049	−1.098***	−0.029	−0.089		−0.030	−0.080
	(0.372)	(0.398)	(0.372)	(0.374)		(0.372)	(0.377)
N	18,066	18,066	18,066	18,066	10,642	18,066	17,883
ID FE	YES	YES	YES	YES	—	YES	YES
Year FE	YES	YES	YES	YES	—	YES	YES

Standard errors are reported in parentheses. *p < 0.1, **p < 0.05, ***p < 0.01

Mediation Analysis

Building on the above benchmark analysis, it is evident that NTU helps mitigate environmental pollution, thereby reducing depression. Therefore, we examine whether environmental pollution (measured by annual PM2.5 concentrations) mediates the relationship between NTU and individual depression. Model (2a) in Table 2 demonstrates a significant negative relationship between NTU and annual PM2.5 concentrations (β = −3.118, p < 0.01). Model (2b) in Table 2 reveals that annual PM2.5 concentrations are significantly related to individual depression (β = 0.019, p < 0.05), suggesting that PM2.5 concentrations act as a mediator in this relationship. Furthermore, after controlling for annual PM2.5 concentrations, NTU remains significantly associated with individual depression (β = −0.321, p < 0.05; and |−0.321| <|−0.379|), indicating that annual PM2.5 concentrations serves as a partial mediator. In addition, the Sobel test yielded a significant result (Z = −2.520, p < 0.05), and the bootstrap confidence interval excluded 0, confirming the statistical significance of the indirect effect (β = −0.058, p < 0.05). Both methods consistently supported the mediating role of annual PM2.5 concentrations. Therefore, we conclude that the negative correlation between NTU and individual depression is partially mediated by environmental pollution. Moreover, prior studies have shown that NTU reduces environmental pollution [19], and reducing environmental pollution has been shown to alleviate individual depression [33, 34].

Robust Tests

To eliminate estimation bias due to unobserved variables, we conducted two placebo tests. First, we hypothesized that NTU was implemented one year earlier, with the results reported in Model (3) of Table 2. We found that the NTU coefficient was insignificant (β = 0.107, p > 0.1), which supports the DID estimates. Second, Fig. 3A shows the estimated coefficients for 500 random samples, which are predominantly centered around 0 and approximately normally distributed. Additionally, Fig. 3B displays the kernel density distribution of the t-values, which similarly cluster around 0. These results indicate the robustness of this study’s results. Additionally, to address potential bias arising from treatment timing heterogeneity, we use never-treated cities as the control group. Model (4) in Table 2 reports that the average treatment effect on the treated remains negative and significant (β = −0.450, p < 0.05). Although this approach results in a substantial reduction in the sample size, the finding further supports the robustness of the TWFE estimates. In addition, we apply the Bacon decomposition method (see the supplementary information) to examine the weights of treatment effect variation. The results show that over 92% of the TWFE estimator’s weight comes from comparisons with “treated groups vs. never-treated groups” (see Supplementary Table S2), suggesting that treatment effect heterogeneity is unlikely to bias the results. As shown in Model (5), the interaction term between NTU and annual PM2.5 concentrations is not significant (β = 0.003, p > 0.1), thereby supporting the assumption of no exposure-mediator interaction and the validity of the mediation framework.

Fig. 3.

Kernel density plot of (A) the estimated coefficients; and (B) the t-value

Endogeneity Analysis

Satisfying the parallel trends assumption is a prerequisite for applying the DID model. That is, prior to NTU implementation, the treatment and control groups should not differ significantly. We use an event study approach (see the supplementary information) to test this assumption. Figure 4 presents the estimated event-time coefficients, showing that the pre-treatment effects are not significant, thereby supporting the validity of the DID design. To address potential selection bias arising from imbalanced individual characteristics, we implement a PSM-DID approach. After matching, all individual-level covariates exhibit standardized bias below 20% (see Supplementary Table S3), indicating good balance. As shown in Model (6) of Table 2, the NTU coefficient remains significant (β = −0.387, p < 0.01), further confirming that NTU alleviates individual depression.

Fig. 4.

Parallel trend test

Heterogeneity Analysis

We examine the heterogeneous effects of NTU on individual depression along two dimensions: whether a city is resource-based and population concentration.

First, cities are classified as resource-based or non-resource-based according to the list issued by the State Council, comprising 33 and 61 cities, respectively. A resource-based city is one where core industries rely on extracting and processing local natural resources. Model (1a) and Model (1b) in Table 3 indicate that NTU significantly reduces individual depression in non-resource-based cities (β = −0.354, p < 0.05) but has no significant effect in resource-based cities (β = −0.418, p > 0.1). These findings indicate that NTU exerts a stronger impact in non-resource-based cities. One possible explanation is that resource-based cities face severe transformation challenges, as their economies are overly reliant on a single industry. This limitation hinders the flow of capital, talent, and technology, resulting in a lack of resources for implementing NTU [46, 47]. In contrast, non-resource-based cities have more diverse economic structures and stronger policy implementation capabilities, enabling them to more quickly realize the effects of NTU and more effectively reduce individual depression.

Table 3.

Heterogeneity analysis

	(1a)	(1b)	(2a)	(2b)
	Non-resource cities	Resource-based cities	Low population concentration	High population concentration
Variable	CESD	CESD	CESD	CESD
DID	−0.354**	−0.418	−0.494**	−0.152
	(0.151)	(0.261)	(0.223)	(0.177)
Mar	0.245**	0.437***	0.219**	0.414***
	(0.099)	(0.136)	(0.111)	(0.117)
LS	−1.136***	−1.473***	−1.333***	−1.193***
	(0.085)	(0.115)	(0.100)	(0.095)
Edu	−0.009	0.162	0.034	0.058
	(0.088)	(0.118)	(0.105)	(0.097)
Res	−0.058	−0.791***	−0.323	−0.236
	(0.167)	(0.259)	(0.199)	(0.203)
RGDP	0.023	0.566	0.974*	−0.613
	(0.579)	(0.747)	(0.530)	(1.092)
Pop	0.001	−0.002	0.003	0.002
	(0.002)	(0.004)	(0.004)	(0.002)
RA	−0.013	0.058	0.007	0.017
	(0.022)	(0.038)	(0.024)	(0.030)
PIS	0.022	0.000	0.054	−0.086*
	(0.039)	(0.053)	(0.040)	(0.048)
ES	−0.780	0.509	−0.331	0.194
	(0.577)	(0.508)	(0.468)	(0.604)
N	11,202	6,864	9,108	8,958
ID FE	YES	YES	YES	YES
Year FE	YES	YES	YES	YES

Standard errors are reported in parentheses. *p < 0.1, **p < 0.05, ***p < 0.01

Second, cities are classified into two groups based on the median population concentration: low and high [38]. Models (2a) and (2b) in Table 3 show that NTU significantly reduces depression in cities with low population concentration (β = −0.494, p < 0.05) but has no significant effect in cities with high population concentration (β = −0.152, p > 0.1). This supports the conclusion that the policy is more effective in cities with low population concentration. NTU alleviates depression by enhancing individuals’ social connections [38]. In cities with low population concentration, where social connections tend to be weaker, NTU is more effective at bridging this gap, potentially reducing depression. In contrast, in cities with high population concentration, social connections are already strong. As a result, NTU may encourage individuals to step beyond their social comfort zones, thus reducing the policy’s effectiveness in alleviating depression [48].

Discussion

With over 300 million people affected by depression globally, the number continues to rise, exerting a heavy burden on individuals, society, and the global economy [1–3]. The people-centered NTU has been gaining increasing attention. Given its people-centered approach, NTU has the potential to alleviate individual depression. Although prior studies have examined NTU’s impact on the ecological environment [19], and some studies suggest that a healthy ecological environment can help alleviate individual depression [33], no study has specifically explored the relationship between NTU and depression. Therefore, our research expands the existing literature on health policy.

Our study suggests that NTU plays a significant role in reducing individual depression. By incorporating the ecosocial theory, this study develops a holistic theoretical framework for understanding how NTU contributes to depression alleviation. First, NTU can help reduce diseases caused by long-term exposure to polluted environments, such as stroke and lung cancer, while also improving patients’ health through timely treatment, thereby alleviating depression related to health issues [20]. Previous literature also indicates a negative correlation between diseases and depression [31, 49]. Second, NTU alleviates the social deprivation caused by uneven urban–rural development, while also enhancing individual life satisfaction through a well-developed social security system and comprehensive public services, thereby reducing individual depression [10, 13, 14]. Prior research has confirmed that life satisfaction negatively correlates with depression [4]. Nonetheless, heterogeneity analysis indicates that NTU’s depression-alleviating effects are weaker in resource-based cities, mainly due to their dependence on a single industry and limited policy implementation capacity [47]. In implementing NTU, it is particularly important to promote urban–rural integration and to strengthen the development of social security systems and public services, as these factors can enhance individual life satisfaction and thereby help reduce depression. Moreover, the relatively limited policy implementation capacity in resource-based cities may constrain the mental health benefits of NTU, suggesting a need for stronger institutional support in these areas. Third, NTU’s improvements in public services and infrastructure promote social participation, which in turn strengthens social connections and alleviates depression associated with social isolation [15]. Previous studies have confirmed a negative correlation between social connections and depression [38]. While the development of public services and infrastructure can help alleviate depression, its effects are less pronounced in high-population-concentration areas than in low-population-concentration areas. Overly dense social connections may lead to discomfort for individuals. Therefore, when implementing NTU, while enhancing public services and infrastructure, it is also important to avoid the excessive concentration of public service facilities and infrastructure.

Our empirical results also suggest that environmental pollution mediates the negative correlation of NTU with individual depression. To maximize the mental health benefits of NTU, efforts to reduce environmental pollution such as promoting green development and improving energy efficiency may be particularly important. Taken together, this study provides specific policy guidance to mitigate depression during the implementation of NTU.

Nevertheless, our research is not without limitations. First, the CHARLS data primarily covered the middle-aged and elderly population, and may not have fully represented the entire social population. Second, the CHARLS data were based on personal interviews, which may not be as accurate as clinical results. Finally, although we examined the mediating role of environmental pollution, other potential mechanisms were not considered.

Conclusion

This study examined the association of NTU with individual depression and further analyzed the underlying mechanisms through which NTU affects depression, accounting for its heterogeneity. This study yields the following significant conclusion: first, the implementation of NTU significantly alleviated depression, and this result remained robust across multiple robustness tests. Second, NTU reduced individual depression by mitigating environmental pollution. Finally, the implementation of NTU exerted a stronger positive impact in non-resource-based cities and cities with lower population concentration. Furthermore, the main contributions are as follows: first, this study found that NTU could reduce individual depression levels, contributing to research in the field of health policy. Second, this study adopted the ecosocial theory, offering a novel theoretical perspective to enhance the understanding of NTU’s and environmental pollution’s effects on individual depression. Third, this study proposed policy recommendations for mitigating depression during NTU implementation. Future studies could incorporate additional data sources to encompass a broader age group, thereby enhancing the study’s representativeness and applicability. Additionally, further research should explore additional mediation pathways to gain a clearer insight into how NTU affects individual depression.

Supplementary Information

Below is the link to the electronic supplementary material.

Funding

This study was funded by the National Natural Science Foundation of China (72271126, 72372079), Qinglan Project of Jiangsu Province, and and the National Statistical Science Research Project (2024LY059).

Data Availability

All data, models, or code that support the findings of this study are available from the corresponding author upon reasonable request.