Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2026 Jun 15;54(5):e70108. doi: 10.1002/jcop.70108

Pathways to Mental Health: Exploring the Relationships Among Neighborhood Hostility, Crime, Built Environment, and Perceived Safety

Paul Rodrigues 1,2,3, Carolyn Côté‐Lussier 1,2,4,✉
PMCID: PMC13266458  PMID: 42290300

ABSTRACT

Neighborhood crime‐related stressors and the built environment are important determinants of mental health, yet research on the pathways through which they affect individuals remains limited. This study draws on structural equation modeling and high‐quality measures to test the pathways linking a novel measure of neighborhood‐level hostility, crime, and built environment features to individuals' perceived safety and mental health. The results suggest that neighborhood crime and perceived safety are directly associated with men's, but not women's, mental health. Additionally, neighborhood hostility, tree canopy, and walkability are indirectly linked to men's mental health via their associations with crime and perceived safety. Findings show that neighborhood hostility and walkability were directly associated with perceived safety for both women and men, while tree canopy was indirectly linked to perceived safety through its relationship with neighborhood hostility. According to these results, equity‐oriented built environment interventions have the potential to reduce neighborhood‐level hostility and crime, and enhance population mental health.

Keywords: built environment, mental health, neighborhood crime, neighborhood hostility, structural equation modeling

1. Introduction

Mental health is a public health issue defined by the World Health Organization as “a state of well‐being that enables people to cope with the stresses of life, realize their abilities, learn well and work well, and contribute to their community” (World Health Organization 2022). However, mental health disorders, such as depression and anxiety, are among the leading causes of disability, affecting a high proportion of people worldwide (World Health Organization 2022). According to a meta‐analysis, 29.2% of individuals may experience a common mental disorder in their lifetime (Steel et al. 2014). Mental disorders represent a major economic burden for society, due to the direct (e.g., healthcare resources), indirect (e.g., lost productivity), and intangible (e.g., psychological pain) costs associated with them (Christensen et al. 2020). In Canada, the annual cost of mental health and mental illness is estimated at over $50 billion (Wilson 2017). Thus, identifying determinants of mental health is essential to guide preventive strategies and reduce the prevalence of mental health disorders.

Socioecological theories of public health identify neighborhood crime as an important environmental determinant of mental health (Baranyi et al. 2021; Lorenc et al. 2012; Ross and Mirowsky 2001). Such exposure could affect individuals both directly (e.g., personal victimization) and indirectly (Baranyi et al. 2021; Lorenc et al. 2012). Specifically, neighborhood crime and various related stressors, such as disorder, risk of victimization, and an inhospitable environment (i.e., unwelcoming and unsafe place), may lead to individual psychological distress (Shmool et al. 2015). See Lorenc et al. (2012) for a detailed conceptual diagram of pathways between neighborhood crime and related factors, and individual responses shaping mental health.

Despite these theoretical frameworks and hypothetical pathways, empirical studies remain focused on police‐recorded neighborhood‐level crime and individuals' perceptions of safety when investigating crime‐related stressors and mental health (Baranyi et al. 2021). Moreover, such analyses are limited in understanding the interrelationships between neighborhood‐level factors, but also with regard to identifying pathways through which they come to influence individuals' mental health. For instance, characteristics of the built environment, such as walkability, greenness, and street lighting, are likely to influence perceptions of the living environment, as well as neighborhood crime‐ and safety‐related stressors (Lorenc et al. 2012). Some studies have used structural equation modeling to estimate the direct and indirect effects of neighborhood crime and related stressors (e.g., perceived disorder) on mental health. However, these studies have primarily focused on individual‐level mediators, such as personal experiences of victimization and perceptions of insecurity and disorder (Curry et al. 2008; Turner et al. 2013). They are limited in assessing the effects of overall neighborhood hostility conceptualized as an aggregated, area‐level measure, and the multiple pathways through which features of the neighborhood environment come to shape mental health. Examining the multiple pathways linking neighborhood‐level crime‐related stressors to mental health allows for the identification of modifiable elements at the neighborhood level that could be targeted by future preventive strategies.

This study aims to assess associations between the built environment, neighborhood crime, and safety‐related stressors at both the neighborhood level (using an aggregated measure of neighborhood hostility) and at the individual level (through a measure of perceived safety) in order to explain individuals' mental health. The study draws on new measures of neighborhood stressors, derived from rich representative data, and a structural equation modeling approach to enhance understanding of the pathways through which hostile neighborhood environments come to influence mental health.

1.1. Defining Mental Health in Relation to the Neighborhood Environment

Studies on the association between neighborhood crime and mental health have primarily focused on mental disorder outcomes, such as depression, anxiety, and suicide. Specifically, a recent literature review showed that neighborhood crime is significantly associated with depression and psychological distress, but there is still limited research to establish an association between neighborhood crime and anxiety or psychosis (Baranyi et al. 2021). Issues related to mental health, such as suicide, have also been analyzed in previous studies, although these studies are relatively few. For example, one study found that several neighborhood characteristics, including police‐recorded crime levels and perceptions of neighborhoods, are likely to affect suicidal behaviors (Allen and Goldman‐Mellor 2018). Other research measured mental health through a general perception of health and found a significant association with neighborhood crime, as well as with crime‐ and safety‐related stressors (Putrik et al. 2019; Ross and Mirowsky 2001). Though this approach may provide an imprecise assessment of mental health and well‐being and mask variability in mental health outcomes.

Positive mental health is a holistic concept that describes mental health as a continuum and emphasizes emotional, psychological, and social well‐being (Keyes 2002). However, no studies examining the association between neighborhood crime and positive mental health have been identified. In addition, studies that consider several components of mental health simultaneously are scarce. Such a conceptualization may capture multiple dimensions of mental health in line with the definition provided by the World Health Organization (i.e., the presence or absence of mental disorders, such as depression, anxiety, and suicidal ideation, as well as measures of well‐being), and may provide a better sense of how neighborhood environments can impede on emotional, psychological, and social well‐being.

1.2. Pathways Linking Neighborhood Crime to Mental Health

Neighborhood crime may have direct impacts on mental health through the consequences of personal victimization, which can lead to both physical and psychological trauma. For example, using path analysis, Curry et al. (2008) found that neighborhood crime is associated with depression through experiences of violence in the neighborhood. However, neighborhood crime is also indirectly associated with mental health through three key processes.

First, at the individual level, neighborhood crime may exacerbate feelings of insecurity, fear of crime, stress, anger, and agitation, which could lead to increased psychological distress (Baranyi et al. 2021; Foster et al. 2010; Lorenc et al. 2012; Putrik et al. 2019; Ross and Mirowsky 2009). Several empirical studies reported a positive association between fear of crime and stress (Macassa et al. 2017), depression (Stafford et al. 2007), anxiety (Stafford et al. 2007), and psychological distress (Ross et al. 2000). Fear of crime, which is distinct from actual crime levels, may also affect individuals by contributing to sedentary lifestyles, which reduce physical and social activities, both of which are important determinants of mental health (Lorenc et al. 2012). In the present study, perceived safety is conceptualized as antecedent factors of mental health. This decision is guided by theoretical and empirical evidence suggesting that perceptions of insecurity may act as chronic stressors that shape individuals' psychological well‐being through behavioral (e.g., avoidance of outdoor activities) and social mechanisms (e.g., reduced social cohesion and trust) (Lorenc et al. 2012). While reverse or reciprocal relationships cannot be ruled out, the temporal and conceptual ordering adopted in this study is consistent with the dominant theoretical framework in environmental psychology and criminology literature.

Second, crime is also likely associated with neighborhood‐level hostile conditions that can negatively impact individuals' mental health. The broken windows theory suggests that neighborhood disorder signals to potential criminals a lack of concern in the community, which could encourage more serious crimes (Wilson and Kelling 1982). Despite the theoretical underpinnings linking crime and disorder, empirical studies have produced mixed results, as this association could be confounded by informal social control (Lanfear et al. 2020). It is therefore essential to view neighborhood disorder in a broader sense by considering its social (e.g., low social control and people littering) and structural (e.g., physical decay that can be explained by institutional failings, such as vacant buildings or graffiti) dimensions. In this perspective, disorder may be seen as a dimension of neighborhood hostility. It should be noted that neighborhood hostility is conceptualized here as a contextual construct reflecting characteristics of neighborhoods themselves rather than individual perceptions, and is therefore analytically distinct from individual‐level appraisals such as perceived safety or mental health outcomes. In addition, this concept is related to neighborhood criminality but reflects a broader contextual construct, as it captures the social and environmental conditions shaping residents' experiences beyond the occurrence of criminal acts. Using a qualitative approach, Shmool et al. (2015) observed that several factors contribute to creating a hostile environment that likely leads to the experience of psychosocial stress, unhealthy behaviors and maladaptive biological processes (i.e., allostatic load). These factors include, among others, neighborhood social disorder and other crime‐related stressors, such as exposure to neighborhood‐level victimization risk (e.g., the prevalence of observed and witnessed criminal incidents within the neighborhood environment) and living in an unwelcoming community. Neighborhood stressors related to crime may erode the social environment (e.g., leading to an actual or perceived low social control) by increasing fear and distrust among residents, which can lead to greater social disorder and a higher risk of victimization (Lanfear et al. 2020). Such conditions may, in turn, contribute to higher levels of crime in neighborhoods, creating a feedback loop. Furthermore, while crime and neighborhood hostility are expected to influence perceived safety, this relationship is more complex and may involve other determinants. The risk–fear paradox suggests that individuals' fear of crime or perceived insecurity does not necessarily correspond to actual crime levels, but is also shaped by broader environmental (e.g., neighborhood built environment), interpersonal, and individual factors (Farrall et al. 2007).

Third, the neighborhood built environment is associated with neighborhood‐level crime and hostility, but also with individual‐level negative perceptions of the local environment, and social and physical activities that are detrimental for mental health. Features related to a walkable environment, greenness, and nighttime lighting are important built environment features related to a host of outcomes, including neighborhood crime‐related stressors. Walkability refers to the environment's ability to promote active transportation, such as walking (Frank et al. 2021; Lefebvre‐Ropars and Morency 2018). This concept is described through several dimensions, the most commonly analyzed of which are street connectivity, residential density, destination density, and transit stop density (Foster et al. 2010; Foster et al. 2015; Frank et al. 2021; Lefebvre‐Ropars and Morency 2018; Nichani et al. 2021; Rodrigues et al. 2022). Walkability may contribute to better mental health by increasing physical activity, but also by fostering social interactions among residents (Zhang et al. 2019). It is also associated with positive neighborhoods perceptions, which help reduce fear of crime (Foster et al. 2010). Conversely, walkability is also associated with neighborhood crime and crime‐related stressors, such as disorder, as a walkable environment facilitates movement for potential offenders and creates more opportunities for crime (Foster et al. 2014). The associations between walkability, neighborhood crime‐related stressors, and mental health are therefore complex.

Indeed, due to the built environment's footprint and the higher density of destinations, walkability is negatively correlated with the level of vegetation (Doiron et al. 2020; Shuvo et al. 2021). However, vegetation supports mental health by alleviating stress and mental fatigue (Kondo et al. 2018). This environmental characteristic is also linked to a positive perception of the surroundings and a greater sense of safety, which can contribute to better mental health (Kondo et al. 2018). Finally, neighborhoods with high levels of vegetation may also contribute to lower levels of crime and hostility as they encourage outdoor activities and enhance neighborhood surveillance (Bogar and Beyer 2015; Shepley et al. 2019).

Lastly, nighttime lighting is an additional environmental feature that influences crime‐related stressors and mental health. Artificial outdoor lighting affects the circadian rhythm and can disrupt individuals' biological and behavioral rhythms, negatively impacting sleep quality, mood, and mental health (Paksarian et al. 2020). Despite this potential detrimental effect on mental health, nighttime lighting is recognized for enhancing the sense of safety (Côté‐Lussier et al. 2020; Haans and de Kort 2012; Welsh et al. 2022) and reducing fear of crime (Welsh et al. 2022), both protective factors for mental health (Lorenc et al. 2012). Nighttime lighting could also help reduce crime and related stressors by encouraging a diverse use of public spaces (e.g., families, young adults, and elderly), thereby strengthening social ties and natural surveillance (Kaplan and Chalfin 2022; Welsh et al. 2022). Furthermore, nighttime lighting is correlated with other features of the built environment. This characteristic is positively associated with walkability, as it may promote neighborhood walkability by enhancing pedestrian security (Fonseca et al. 2022). Nighttime lighting is also likely to be negatively correlated with the level of greenness, as walkable neighborhoods—where street lighting is more prominent—often exhibit less vegetation (Shuvo et al. 2021).

1.3. Neighborhood Crime, Mental Health, and Gender

In spite of some of these broader patterns, it is well established that the relationship between the neighborhood environment and mental health may vary by gender. In general, the association between neighborhood features and health is likely to be stronger for women than for men, as women may have greater exposure to their local environment (Stafford et al. 2005). Furthermore, women may be more sensitive to adverse factors, such as neighborhood crime and exposure to a hostile environment, because of a perceived higher risk of the negative impacts of victimization and less perceived control over victimization, which could lead to poorer mental health outcomes (Clark et al. 2008; Santiago et al. 2011). These findings may help explain why, in urban settings, women tend to report greater fear of crime (Clark et al. 2008). Moreover, associations between features of the built environment and mental health may vary by gender, although these differences are less clear. For example, improving street lighting is known to enhance perceptions of safety, particularly for women (Kaplan and Chalfin 2022). However, the effects of greenness and walkability are more nuanced. For example, one study found a protective effect whereby greenness reduces sedentary behavior—a determinant of mental health—but only for women (Nichani et al. 2021). In contrast, another study found that greening is associated with increased outdoor time for men, which leads to a lower fear of crime, while for women, it is associated with greater fear due to a heightened male presence in public spaces (Kondo et al. 2021). Regarding walkability, some studies suggest stronger effects of walkability features on sedentary activities for women (Foster et al. 2015), while others found stronger effects for men (Nichani et al. 2021).

Gender differences in the expression of mental health problems are also noted, with women demonstrating more internalizing disorders (e.g., anxiety and depression) and men exhibiting more externalizing disorders (e.g., substance abuse, antisocial behaviors, and death by suicide) (Rosenfield and Mouzon 2013). In this respect, associations between hostile neighborhood environments and crime can also be understood through a lens which considers crime as an epidemiologic phenomenon, with differing mental health consequences according to gender (Skinner and Farrington 2024).

1.4. Objectives

The main objective of this study is to analyze the associations between individuals' mental health and crime‐related stressors (i.e., crime, neighborhood‐level perceived hostility, perceived safety and victimization, and features of the built environment) at the neighborhood level using structural equation modeling to simultaneously estimate direct and indirect effects. A secondary objective is to evaluate gender‐based differences in these associations. In line with the extant literature, individual‐level perceived safety is considered as a potential mediator linking the neighborhood environment to mental health (Lorenc et al. 2012). In this study, mental health is considered as a complex concept reflecting the absence of disorders, but also emotional and social well‐being (e.g., reported positive mental health). The hypothesized associations are summarized in Figure 1. In line with the second objective, the associations between neighborhood crime‐related stressors and mental health are expected to be stronger for women than for men. Additionally, it is expected that the strength and direction of the associations between features of the built environment and individuals' perceived safety and mental health will vary by gender.

FIGURE 1.

FIGURE 1

Open in a new tab

Hypothesized pathways linking built environment characteristics, crime, crime‐related stressors, and mental health. Note: Solid lines refer to neighborhood‐level variables; dashed lines refer to individual‐level variables.

2. Materials and Methods

2.1. CCHS Data

This study is based on data from the Canada Community Health Survey (CCHS, 2015–2016), a cross‐sectional survey of Canadians over the age of 12 conducted by Statistics Canada (Statistics Canada 2019). The CCHS employed a multistage sampling strategy that provides a representative sample of the target population. The information gathered by the CCHS covers a wide range of topics related to health, lifestyle, and living conditions. In the current study, only participants who reported a postal code located on the island of Montreal were included (n = 1281). The Island of Montréal consists of the City of Montréal, the second‐largest city in Canada, and 14 other municipalities, primarily suburban in nature. It had a population of 1,942,040 in 2016 (Statistics Canada 2017).

2.1.1. Mental Health

As described previously, mental health is a multidimensional concept that encompasses not only the presence or absence of specific disorders but also reflects a continuum of well‐being (World Health Organization 2004). To capture this complexity, mental health was operationalized as a latent factor reflecting a general underlying dimension rather than distinct subdomains, such as distress or well‐being, which are considered closely related manifestations of the same construct. It was measured using five variables: perceived mental health, anxiety, depression, suicidal ideation, and positive mental health.

Perceived mental health was measured using a single question: “In general, would you say your mental health is…?” The response scale ranged from 1 (Excellent) to 5 (Poor). This variable was reverse‐coded so that participants reporting good mental health received the highest scores.

Anxiety was operationalized using a dichotomous variable that distinguished participants who answered “no” from those who responded “yes” to the following question: “Do you have an anxiety disorder such as a phobia, obsessive‐compulsive disorder or a panic disorder?”

To assess depression, participants were asked the following question: “Do you have a mood disorder such as depression, bipolar disorder, mania or dysthymia?” A dichotomous variable was created by differentiating participants who reported having a mood disorder from those who did not.

Suicidal ideation was measured using a single question: “Have you ever seriously contemplated suicide?” The response options were yes and no, which were used to create a dichotomous variable.

Positive mental health was measured using the Mental Health Continuum – Short Form (MHC‐SF), which consists of 14 items (Lamers et al. 2011). This instrument assesses three dimensions of positive mental health: emotional (three items—e.g., “In the past month, how often did you feel happy?”), psychological (five items—“In the past month, how often do you feel that you had experiences that challenge you to grow and become a better person?”), and social well‐being (six items—e.g., “In the past month, how often do you feel that you had something important to contribute to society”). For each question, the response scale ranges from 0 (Never) to 5 (Every day). The response options were reverse‐coded and then summed to obtain a total score ranging from 0 to 70, with a higher score indicating greater well‐being.

A Confirmatory Factor Analysis (CFA) was conducted separately for women and men to derive the measures of mental health. This approach allows us to consider gender‐based differences in the expression of mental health, as women are more likely to experience internalizing disorders, such as anxiety, while men are more often affected by externalizing disorders, such as substance use (Kessler et al. 2005). Good fit indices were observed in the model for women (χ 2 = 295.801, degrees of freedom (df) = 10; p < 0.001; Comparative Fit Index (CFI) = 0.99; Root Mean Square Error of Approximation (RMSEA) = 0.00; Standardized Root Mean Square Residual (SRMR) = 0.038) and the model for men (χ 2 = 442.406, df = 10; p < 0.001; CFI = 0.99; RMSEA = 0.01; SRMR = 0.042). As a rule of thumb, a non‐significant Chi‐square statistic (p > 0.05), an RMSEA below 0.05, a SRMR below 0.08, and a CFI greater than 0.95 indicate a good fit (Hu and Bentler 1998).

2.1.2. Perceived Safety

Data from the CCHS was also used to measure perceived safety, based on a single question: “The crime rate in my neighborhood makes it unsafe to go on walks at night.” The response scale ranges from 1 (Strongly agree) to 4 (Strongly disagree). Higher scores reflect more perceived safety. A dichotomous variable was created to differentiate between participants who disagreed or strongly disagreed (1) and those who agreed or strongly agreed (0). This dichotomization was used to facilitate interpretation and to distinguish between respondents perceiving their neighborhood as safe vs. unsafe, consistent with prior studies using this item (Forsyth et al. 2015).

2.2. Neighborhood‐Level Variables

2.2.1. Definition of Neighborhoods

Egocentric neighborhoods were used to assess exposure to crime and crime‐related stressors. According to this approach, neighborhoods are defined as the set of territories located within a specific radius of an individual's place of residence (Rodrigues et al. 2022; Rodrigues et al. 2024). They are implemented separately for each individual by creating a buffer around the individual's place of residence. Egocentric neighborhoods are preferred over administrative units (e.g., census tracts), as they more accurately capture individuals' actual exposure to their neighborhoods and activity spaces (Rodrigues et al. 2022; Rodrigues et al. 2024).

To operationalize egocentric neighborhoods, participants were first located using the centroid of their postal code of residence. In Canada, postal codes provide a good approximation of exact addresses in urban areas (Bow et al. 2004), as is the case for the island of Montreal. Polygon‐based network buffers around the participants' postal codes were then created to define egocentric neighborhoods. This buffering method provides a better estimate of the spaces used by individuals compared to traditional methods, such as circular buffers based on Euclidean distance (Oliver et al. 2007). In this study, polygon‐based network buffers were created by calculating all possible routes within a distance of 1000 m (i.e., a 15‐min walk) from the participants' places of residence and connecting the resulting endpoints. This analysis was performed using ArcGIS Pro 2.7.0 (Network Analyst extension).

2.2.2. Neighborhood Crime

Data from the Service de police de la Ville de Montréal (SPVM) for the years 2014–2017 were used to estimate neighborhood crime density. In this database, the date and location of criminal incidents at the nearest intersection, along with the type of crime, were available. Only violent crime events (i.e., homicide and robbery), excluding incidents of intimate partner violence and sexual assault which are not made public for confidentiality reasons, were available to estimate neighborhood crime density. These crime types may directly impact individuals and are less likely to be underreported than minor crimes (Baumer 2002). To assess violent crime density, Kernel Density Estimation (KDE) was performed in ArcGIS Pro 2.7.0. This method estimates the probability density of events on a continuous surface, reflecting the intensity of criminal incidents (Hart and Zandbergen 2014). In this study, the KDE of violent crime was computed on a 30 m × 30 m cell grid, using a bandwidth of 1122 m determined by Silverman's rule of thumb (Silverman 2018). The average was then calculated for each egocentric neighborhood.

2.2.3. Neighborhood Hostility

Neighborhood hostility was derived from measures estimated in a previous study using a regression kriging approach to estimate the probability of exposure to specific crime and safety‐related stressors (Rodrigues et al. 2026). These estimates were based on survey data from the 2014 General Social Survey capturing multiple crime‐ and safety‐related stressors, combined with sociodemographic variables drawn from the 2016 Canadian Census at the dissemination area (DA) level, to produce area‐level estimates of neighborhood hostility. Values are expressed in percentage points, with higher scores indicating a greater likelihood of exposure to these neighborhood stressors. Accordingly, a one‐unit increase corresponds to a one percentage‐point increase in the estimated probability of neighborhood residents reporting the corresponding condition at the neighborhood level. These measures have been established as being robust and as demonstrating strong convergent validity (correlations with crime densities from police data) and predictive validity (the ability to predict individual‐level perceived safety) (Rodrigues et al. 2026).

In the present study, a composite measure was estimated to reflect the probabilities of perceived social disorder (e.g., noisy neighbors and vandalism to property or vehicles), of experiences of victimization, and of the inverse of a perception of living in a welcoming community. Neighborhood hostility was derived for each egocentric neighborhood using the same method described above. CFA was then performed to estimate the measure of neighborhood hostility, and showed good fit indices (χ 2 = 2540.570, df = 3; p < 0.001; CFI = 0.99; RMSEA = 0.00; SRMR = 0.010).

2.2.4. Built Environment

Four variables were used to capture features of walkability: density of intersections, residential density, density of destinations, and density of public transport stops. First, data from Addresses Québec (2016) were employed to measure the density of intersections, which was calculated by the number of intersections with three or more road segments (excluding highways). Second, the net residential density (i.e., the number of dwellings per hectare of residential area) was calculated using data from the property assessment roll (parcel‐level land use) provided by the Communauté Métropolitaine de Montréal (2016). Third, Enhanced Points of Interest from DMTI (2016) were used to estimate the density of destinations. This database provides the locations of business establishments along with their types (Standard Industrial Classification, SIC, codes). Based on these SIC codes, only daily services were retained (e.g., retail stores, restaurants, community organizations, daycare centers, and health and social services) to calculate the number of services (i.e., the density of destinations). Fourth, the number of public transport stops was obtained using the locations of metro stations and bus stops provided by the Société de Transport de Montréal (STM). Finally, to estimate measures of walkability, CFA was performed. The model showed good fit indices (χ 2 = 238.96, df = 5; p < 0.001; CFI = 0.93; RMSEA = 0.05; SRMR = 0.049).

Greenness was assessed by calculating the percentage of tree canopy cover. The spatial coverage of tree canopy for the island of Montréal was obtained from the Communauté métropolitaine de Montréal for 2015. This coverage was estimated by combining the normalized difference vegetation index (NDVI) from 1 m resolution Landsat images with height information derived from LiDAR data. The tree canopy corresponds to tiles with an NDVI greater than 0.3 and a height exceeding 3 m. Data and detailed methodology are publicly available online (Communauté Métropolitaine de Montréal 2017). The percentage of tree canopy in each egocentric neighborhood was calculated using the following formula: (area of tree canopy coverage/total area) × 100.

Nighttime lighting was operationalized using a measure derived from cloud‐free nighttime images of the island of Montréal captured by the International Space Station (ISS). This variable was obtained from previous analyses that estimated three measures of nighttime lighting corresponding to the red, green, and blue color channels of the ISS images (Côté‐Lussier et al. 2020). In this study, only the blue nighttime lighting measurement was used, as it was identified as the most valid predictor of health outcomes.

2.2.5. Covariates

Four individual‐level variables were used as covariates: gender (woman or man), age, income ($19,999 and below; $20,000–$39,999; $40,000–$59,999; $60,000–$99,999; and $100,000 and over), and visible minority status (white or non‐white). Statistics Canada defines visible minority status as “… whether a person is a visible minority or not, as defined by the Employment Equity Act. The Employment Equity Act defines visible minorities as persons, other than Aboriginal peoples, who are non‐Caucasian in race or non‐white in color” (Statistics Canada 2018).

2.3. Statistical Analysis

Neighborhood‐level data were assigned to the CCHS data using participants' residential postal code. Associations between the built neighborhood environment (2016), neighborhood crime (2014–2017), neighborhood hostility (2014), and individual‐level perceived safety and mental health (2015–2016) were estimated using structural equation models in R software (version 4.2.3, package Lavaan). To account for gender differences, and consistent with the study's theoretical approach, a gender‐stratified modeling approach was used to investigate how mental health may differ meaningfully between men and women, rather than assuming a single homogeneous structural model across genders. Although multigroup SEM is also recommended for testing gender differences and moderation effects (Marsh et al. 2004), such approaches aim primarily to test invariance in measurement and structural models (Heub et al. 2024) which was not this study's primary objective.

With respect to model estimation, the diagonally weighted least squares method was employed for these analyses due to the inclusion of several categorical variables. All models used sampling weights and adjusted for the effects of individual‐level covariates on mental health and perceived safety. Sampling weights were incorporated within the DWLS estimation framework, in line with established recommendations for handling survey weights in SEM. Missing data for covariates (N = 3–244) were handled using multiple imputation methods (20 iterations) with the Amelia package. To assess model fit, we used a Chi‐square test of fit, the RMSEA, the CFI, and the SRMR.

3. Results

3.1. Descriptive Analyses

Table 1 presents the characteristics of the sample for women and men separately. Most participants reported having good mental health (77.45% for men and 66.86% for women) and no issues with anxiety (94.86% for men and 93.27% for women), depression (96.10% for men and 93.29% for women), and suicidal ideation (84.83% for men and 80.06% for women). Furthermore, the majority of both men and women perceived their neighborhood as safe (80.45% and 69.69%, respectively).

TABLE 1.

Descriptive analysis.

Men (n = 599) Women (n = 682)
Mean (SD) or %a Mean (SD) or %a
Mental health
Positive mental health 42.94 (18.21) 45.02 (19.76)
 NA 96 148
No anxiety (%)
 No 94.86 93.27
 Yes 5.14 6.42
No depression (%)
 No 96.10 93.29
 Yes 3.81 6.29
No suicidal ideation (%)
 No 84.83 80.06
 Yes 10.98 14.15
Perceived mental health (%)
 Good 77.45 66.86
 Poor 21.27 30.71
Perceived safety (%)
Low 15.69 21.71
High 80.45 69.69
Density of crime 16.70 (19.49) 15.97 (15.00)
Neighborhood hostility
Prob. perceived social disorder 14.77 (4.79) 14.30 (5.00)
Prob. perceived living in an unwelcoming community 9.56 (3.56) 9.87 (3.52)
Prob. violent victimization 3.03 (0.87) 2.97 (0.86)
Percentage of tree canopy 25.67 (8.57) 24.60 (7.95)
Nighttime lighting 22,277.16 (5791.09) 23,112.50 (5695.59)
Walkability
Density of intersections 61.44 (25.05) 61.29 (24.21)
Density of destinations 219.64 (356.20) 200.33 (265.65)
Residential density 9.57 (6.84) 9.74 (6.00)
Density of public transport stop 56.31 (29.70) 58.52 (27.65)
Control
Age 42.94 (18.21) 45.02 (19.76)
Income (%)
$19,999 and below 9.86 15.12
$20,000–$39,999 17.43 27.29
$40,000–$59,999 16.71 19.35
$60,000–$99,999 22.92 21.40
$100,000 and over 33.07 16.84
Visible minority status (%)
White 64.12 62.41
Non‐white 29.45 30.89
Open in a new tab
a

The sum of certain percentages does not equal 100% due to missing data.

3.2. Structural Equation Models

3.2.1. Model for Men

Figure 2 summarizes the results of the structural equation model for men, indicating an acceptable fit (χ 2 = 347.967; df = 129; p < 0.001; CFI = 0.95; RMSEA = 0.05; SRMR = 0.076). The results suggest significant direct paths by which men's perceived safety (β = 0.29; SE = 0.15; p < 0.05) is associated with greater mental health, and by which neighborhood police‐recorded crime (β = −0.42; SE = 0.14; p < 0.05) is associated with poorer mental health. No significant direct paths were observed between other neighborhood‐level factors and mental health. However, there are indirect paths from several neighborhood‐level factors to mental health via police‐recorded crime and perceived safety. First, we observed a significant negative indirect association, whereby neighborhood hostility was associated with poorer mental health through lower perceived safety (indirect effect: β = −0.09; SE = 0.04; p < 0.05). Second, tree canopy (indirect effect: β = 0.06; SE = 0.03; p < 0.05) was indirectly associated with better mental health, while walkability (indirect effect: β = −0.29; SE = 0.10; p < 0.05) was indirectly associated with poorer mental health due to the association with police‐recorded crime. Walkability was also linked to better mental health through its positive association with perceived safety (indirect effect: β = 0.08; SE = 0.04; p < 0.05). Neighborhood‐level factors and individual‐level covariates accounted for 23.3% of the variance in perceived safety, while neighborhood‐level factors, men's perceived safety, and individual‐level covariates explained 13.2% of the variance in men's mental health.

FIGURE 2.

FIGURE 2

Open in a new tab

Structural equation model for men. N = 599; χ 2 = 347.967; df = 129; p < 0.001; CFI = 0.95; RMSEA = 0.05; SRMR = 0.076. *** p < 0.001; ** p < 0.01; * p < 0.05. Note: Solid lines refer to neighborhood‐level variables; dashed lines refer to individual‐level variables. All coefficients are standardized.

3.2.2. Model for Women

The model for women, displayed in Figure 3, provided an acceptable fit (χ 2 = 314.718; df = 129; p < 0.001; CFI = 0.96; RMSEA = 0.05; SRMR = 0.080). No direct and indirect associations between neighborhood‐level factors and perceived safety, on the one hand, and women's mental health, on the other, were observed. However, the results suggest that several neighborhood‐level factors were linked to their perceived safety. There were direct paths demonstrating that neighborhood hostility was associated with lower perceived safety (β = −0.32; SE = 0.09; p < 0.05), while neighborhood walkability was associated with higher perceived safety (β = 0.34; SE = 0.12; p < 0.05). In addition, tree canopy (indirect effect: β = 0.11; SE = 0.04; p < 0.05) and walkability (indirect effect: β = −0.21; SE = 0.07; p < 0.05) were indirectly associated with perceived safety because they are associated with neighborhood hostility. Neighborhood‐level factors and individual characteristics explained 23.3% of the variance in perceived safety, while neighborhood‐level factors, women's perceived safety, and individual characteristics explained 16.8% of the variance in women's mental health.

FIGURE 3.

FIGURE 3

Open in a new tab

Structural equation model for women. N = 682; χ 2 = 314.718; df = 129; p < 0.001; CFI = 0.96; RMSEA = 0.05; SRMR = 0.080. *** p < 0.001; ** p < 0.01; * p < 0.05. Note: Solid lines refer to neighborhood‐level variables; dashed lines refer to individual‐level variables. All coefficients are standardized.

4. Discussion

This study aimed to assess the relationships between neighborhood crime and mental health using (i) a broader conceptualization of crime and (ii) structural equation modeling to capture its complex associations with positive mental health. This approach provides a better understanding of how neighborhood crime‐related stressors—particularly community‐level perceptions (i.e., neighborhood‐level perceived hostility) which are linked to police‐recorded crime and the built environment—may be related to individuals' mental health by identifying their direct and indirect associations. To our knowledge, no other study has simultaneously analyzed as many neighborhood crime‐related stressors while considering their interrelationships and potential mediating effects. In this respect, the study makes important advances in identifying how social and physical neighborhood dynamics come to shape individuals' mental health.

Although several factors examined in this study are associated with mental health, the R 2 values were relatively modest for both women (17%) and men (13%), which is expected given the multifactorial nature of mental health and the range of determinants not captured in the present models, with individual factors likely playing a particularly important role. Moreover, consistent with previous studies (Clark et al. 2008), our results showed that women reported lower feelings of safety than men. However, associations between neighborhood characteristics, perceived safety, and mental health were observed only for men, even though the model for women showed a higher R 2. This suggests that women's mental health could be more explained by individual characteristics. Evidence also suggests that the relative importance of neighborhood features varies by gender and socioeconomic position, suggesting different mechanisms linking the neighborhood environment to health outcomes for women depending on their socioeconomic position (Burke et al. 2009). Moreover, previous research has also observed that neighborhood characteristics and perceived safety are more likely to be associated with men's mental health, indicating that men who perceive their neighborhoods as unsafe tend to report higher levels of distress than women (Robinette et al. 2025). This gender difference may also be explained by the fear of crime paradox, which suggests that women consistently report higher levels of perceived unsafety and fear of crime, despite being less frequently physically victimized than men in their neighborhoods (Macassa et al. 2017). From this perspective, men may be more frequently exposed to non‐intimate violence (e.g., neighborhood assaults), whereas women are more likely to experience sexual or intimate partner violence. In this context, it is important to note that the measure of crime used in this study excludes certain types of crime, particularly sexual assault and intimate partner violence, which disproportionately affect women. The associations between police‐recorded crime, perceived safety, and mental health may therefore be underestimated for women. Furthermore, although we observed gender differences in the relationships between neighborhood characteristics, perceived safety, and mental health, these results should be interpreted with caution, as these differences were not statistically tested. Future research could formally test gender invariance using multi‐group SEM approaches, which was beyond the scope of the present study.

Indeed, our results showed that crime density and perceived safety were the only factors directly associated with men's mental health. Both have been posited as key determinants of mental health (Lorenc et al. 2012; Ross et al. 2000; Stafford et al. 2007). They are likely to influence mental health by increasing risk of victimization and levels of anxiety and avoidance behaviors (e.g., limiting outdoor activities) and by impacting social life, and in turn the social environment (e.g., reduced social interaction in the community) (Lorenc et al. 2012). In line with the fear of crime paradox, men may be more affected by these factors, as higher crime density can increase the veritable risk of victimization, while perceiving the neighborhood as unsafe may reflect prior experiences of victimization (Lorenc et al. 2012; Macassa et al. 2017). Although the present study did not measure individual victimization, future studies should assess these pathways to better understand the impact of neighborhood crime on mental health.

While neighborhood‐level features and perceived safety were not associated with women's mental health, there were some consistent patterns in terms of the predictors of perceived safety for both men and women. Higher levels of neighborhood hostility were directly associated with lower perceived safety for both women and men. Factors related to neighborhood hostility measured in this study were neighborhood perceptions of social disorder, perceptions of living in an unwelcoming community, and greater experiences of victimization. The results are consistent with the literature which demonstrates that indicators of neighborhood hostility generate feelings of insecurity, fear, and mistrust between residents (Ross and Mirowsky 2001, 2009; Ross et al. 2000). Our results additionally demonstrate how higher levels of neighborhood hostility indirectly impact men's mental health by reducing their perceived safety. Previous studies, which focused exclusively on disorder, have also observed this pathway (Ross and Mirowsky 2001). However, the present study makes a novel contribution by demonstrating that neighborhood hostility also represents exposure to a perceived unwelcoming neighborhood and exposure to victimization, which remains uncaptured by classic measures of neighborhood disorder and crime.

In the present study, there were no direct associations between the built environment and mental health. However, the results indicated that neighborhood walkability was positively associated with neighborhood hostility, crime and perceived safety, while greenness was negatively associated with neighborhood hostility and crime. As such, greenness is indirectly associated with individuals' perceived insecurity, while neighborhood walkability is both directly and indirectly associated with this outcome, leading to an indirect association with men's mental health. Our results align with previous studies suggesting that neighborhood walkability may contribute to crime and disorder by increasing opportunities for criminal activity (Foster et al. 2014). Despite this deleterious effect, walkability has also been identified as reducing fear of crime (Foster et al. 2010) and promoting social and physical activities, which could contribute to better mental health (Zhang et al. 2019). It is possible that greater walkability is associated with more police presence, and in turn more detection and reporting of crime. Indeed, in other research, antisocial behaviors and crime are found to be spatially co‐located with prosocial behaviors (Weisburd et al. 2025). These pathways have not been tested in the current study but should be explored to better understand the associations between walkability, crime, and mental health.

In addition, the potential role of urbanicity should be considered when interpreting these findings. Although urbanicity was not directly measured in this study, neighborhood walkability partly captures characteristics typically associated with urban environments, such as higher residential density and greater land‐use mix. Urbanicity is often identified as a risk factor for both crime and mental health outcomes (Freedman and Woods 2013; Sampson and Groves 1989; Allen and Goldman‐Mellor 2018), while also being associated with higher levels of walkability (King and Clarke 2015). This pattern may also be influenced by neighborhood sociodemographic characteristics, which were not explicitly modeled in this study but are central to theories such as social disorganization, suggesting that structural factors (e.g., socioeconomic disadvantage, residential instability, and ethnocultural diversity) may shape both crime (Sampson and Groves 1989; Baranyi et al. 2021) and mental health outcomes (Ross and Mirowsky 2001). These factors were not included in the present models, in part for the sake of parsimony and to focus on novel indicators of neighborhood crime and hostility, but could be conceptualized as upstream determinants shaping neighborhood conditions. In this context, urbanicity and sociodemographic composition may contribute to the observed complex associations whereby walkability is linked to both potentially beneficial (e.g., increased activity and social interaction) and detrimental (e.g., greater exposure to crime or disorder) processes. Future research should explicitly account for urbanicity and sociodemographic factors, as they are likely to shape and interact with neighborhood conditions and, in turn, influence mental health outcomes.

In contrast, our results demonstrate that greenness is associated with a reduction in neighborhood crime and hostility. These findings align with studies suggesting that greenness contributes to a safer environment (Bogar and Beyer 2015; Shepley et al. 2019). Finally, we found a negative correlation between greenness and walkability, suggesting that walkable neighborhoods are more likely to lack vegetation. Several studies have reported similar results across various geographical contexts (Doiron et al. 2020; Shuvo et al. 2021), including in Montréal (Doiron et al. 2020). These findings highlight the importance of adopting an equity deserving approach that targets the modification of built environment features (Heinze et al. 2018; South et al. 2018), such as greenness, in highly walkable areas. Such an approach could be particularly helpful in addressing the gamut of crime‐related stressors and fostering feelings of safety, which could promote better mental health for individuals. Furthermore, given the potential influence of neighborhood‐level social and physical dynamics and individuals' perceived safety, considering their combined effect could provide a more accurate estimation of associations between environmental factors and mental health outcomes. For instance, future research could integrate the concept of green walkability, which integrates measures of walkability and greenness (Roscoe et al. 2022), to further study associations between the built environment, behaviors, and mental health.

4.1. Strengths and Limitations

This study drew on representative survey data and a number of robust high‐quality measures of neighborhood characteristics, enhancing the strength of the findings. Moreover, mental health was operationalized using five variables, addressing both the absence of mental disorders and a continuum of well‐being, in accordance with the World Health Organization's definition (World Health Organization 2004). This approach contributes to a more precise conceptualization and operationalization of mental health compared to previous studies on neighborhood crime and safety, which primarily consider specific mental health disorders or perceptions of health (Allen and Goldman‐Mellor 2018; Baranyi et al. 2021; Putrik et al. 2019; Ross and Mirowsky 2001). Lastly, the neighborhood measures capture ambient neighborhood crime‐related stressors, which overcome limitations of previous research relying on individual‐level perceptions or experiences (e.g., perceived social disorder and personal victimization) (Curry et al. 2008; Turner et al. 2013).

This study also has some limitations. First, the CCHS is a cross‐sectional survey, which limits causal inference. Consequently, the SEM approach used in this study estimates statistical mediation and associations among the variables, rather than temporal mediation. This limitation may increase the risk of reverse causality and same‐source bias when estimating the association between perceived safety and mental health, as individuals with poor mental health tend to have a more negative perception of their neighborhoods and report lower feelings of safety (Lorenc et al. 2012). Related to this issue is the idea that feelings of safety may also be considered as a manifestation of poor mental health, in particular for men whose feelings of safety are found to be less influenced by spatial features and more influenced by individual features (Keel et al. 2025). Future studies should use longitudinal data and perhaps other indicators of mental health (e.g., substance use) to address this limitation.

Second, cross‐sectional studies are particularly subject to the risk of residential self‐selection bias. This issue arises when residential choice leads to a non‐random distribution of individuals across neighborhoods, potentially creating a spurious or inflated association between neighborhood characteristics and outcomes (Diez Roux and Mair 2010; Boone‐Heinonen et al. 2011). For example, in this study, individuals with poorer mental health are more likely to reside in higher‐crime and hostile neighborhoods due to socioeconomic constraints, reduced mobility, or psychological factors, such as low self‐efficacy. To overcome this limitation, studies relying on a longitudinal design are considered a strong approach to capture changes in neighborhood characteristics and individuals' outcomes (Boone‐Heinonen et al. 2011). In the case of a cross‐sectional design, measuring residential preferences through a questionnaire can help address residential self‐selection bias by enabling adjustment for residential sorting in statistical models (Zang et al. 2019). However, this information was not available in the CCHS data used in this study.

Third, this study measured mental health outcomes using a multidimensional approach which aligns with leading definitions of this concept and not limiting it to specific mental health disorders or a well‐being measure (World Health Organization 2004). However, the five variables derived from CCHS data may introduce conceptual and temporal heterogeneity by reflecting various time frames of mental health. For example, positive mental health reflects past‐month well‐being, while suicidal ideation is lifetime‐base. In addition, lifetime suicidal ideation may not be fully temporally aligned with current neighborhood conditions, which could affect the interpretation of associations with contextual factors. Despite this limitation, good fit statistics suggest that the measures accurately reflect the underlying dimension of mental health. This specification was retained to capture a broader underlying dimension of mental health.

Lastly, while this study focused on the mediating processes related to neighborhood hostility and individuals' perceived safety, several other factors could explain the link between neighborhood characteristics and mental health. For instance, neighborhood factors, including crime, may influence individuals' behaviors, such as social and physical activities, which can, in turn, affect mental health (Lorenc et al. 2012). Failing to consider these factors may partly explain why we did not observe some associations between neighborhood characteristics and mental health, particularly for women. Therefore, future studies should additionally assess the mediating effects of individuals' physical activity, sedentary lifestyles, and social behaviors.

5. Conclusion

Neighborhood crime is associated with higher levels of hostility in neighborhoods, which in turn is associated with lower perceived safety and, ultimately, poorer mental health. Neighborhood crime‐related stressors are particularly detrimental to men, who are more likely to experience victimization in their neighborhoods (Macassa et al. 2017). For women, neighborhood crime‐related stressors are detrimental for feelings of safety, which may come to shape health‐related behaviors (e.g., using complex avoidance strategies) (Jacobsen 2025). This study highlighted the key role of neighborhood hostility and individuals' perceived safety, which have been identified as mediators in the relationship between neighborhood crime and mental health. Therefore, building and supporting welcoming neighborhoods is necessary for promoting equity in mental health outcomes. Our results also demonstrated how the built environment—specifically greenness and features related to walkability—shapes neighborhood crime and related stressors. Thus, future studies on mental health should systematically consider the role of crime‐ and safety‐related stressors, as well as characteristics of the built environment.

This study also has implications for intervention. For example, interventions such as community‐engaged greening, which are designed to improve the built environment, promote social connectedness, and are associated with reductions in crime (Heinze et al. 2018), are also associated with better mental health (South et al. 2018). Neighborhoods that lack vegetation and that are characterized by high levels of crime could be targeted by such interventions with the aim of enhancing the neighborhood environment, and in turn the mental health of local populations.

Ethics Statement

Ethical approval was obtained from the Ethics Board of the Institut National de la Recherche Scientifique and from the Ethics Board of the Quebec Inter‐University Centre for Social Statistics. Data from the General Social Survey and from the Canadian Community Health Survey used in the current study were anonymous, and no participants were contacted.

Conflicts of Interest

The authors declare no conflicts of interest.

Acknowledgments

This work was supported by a grant from the Canadian Institutes of Health Research (CIHR) awarded to Carolyn Côté‐Lussier and Mark Daniel (Co‐PIs) (grant number 202010PJT‐453186‐PH1‐CFJA‐230902).

Data Availability Statement

Canadian Community Health Survey data are provided by the Québec Inter‐University Centre for Social Statistics and are not publicly available. All other data used in the current study are available from the corresponding author on reasonable request.

References

  1. Allen, K. , and Goldman‐Mellor S.. 2018. “Neighborhood Characteristics and Adolescent Suicidal Behavior: Evidence From a Population‐Based Study.” Suicide and Life‐Threatening Behavior 48, no. 6: 677–689. 10.1111/sltb.12391. [DOI] [PubMed] [Google Scholar]
  2. Baranyi, G. , Di Marco M. H., Russ T. C., Dibben C., and Pearce J.. 2021. “The Impact of Neighbourhood Crime on Mental Health: A Systematic Review and Meta‐Analysis.” Social Science & Medicine 282: 114106. 10.1016/j.socscimed.2021.114106. [DOI] [PubMed] [Google Scholar]
  3. Baumer, E. P. 2002. “Neighborhood Disadvantage and Police Notification by Victims of Violence.” Criminology 40, no. 3: 579–616. 10.1111/j.1745-9125.2002.tb00967.x. [DOI] [Google Scholar]
  4. Bogar, S. , and Beyer K. M.. 2016. “Green Space, Violence, and Crime: A Systematic Review.” Trauma, Violence & Abuse 17, no. 2: 160–171. 10.1177/1524838015576412. [DOI] [PubMed] [Google Scholar]
  5. Boone‐Heinonen, J. , Gordon‐Larsen P., Guilkey D. K., D. R. Jacobs, Jr. , and Popkin B. M.. 2011. “Environment and Physical Activity Dynamics: The Role of Residential Self‐Selection.” Psychology of Sport and Exercise 12, no. 1: 54–60. 10.1016/j.psychsport.2009.09.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bow, C. J. D. , Waters N. M., Faris P. D., et al. 2004. “Accuracy of City Postal Code Coordinates as a Proxy for Location of Residence.” International Journal of Health Geographics 3, no. 1: 5. 10.1186/1476-072X-3-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Burke, J. , O'campo P., Salmon C., and Walker R.. 2009. “Pathways Connecting Neighborhood Influences and Mental Well‐Being: Socioeconomic Position and Gender Differences.” Social Science & Medicine (1982) 68, no. 7: 1294–1304. 10.1016/j.socscimed.2009.01.015. [DOI] [PubMed] [Google Scholar]
  8. Christensen, M. K. , Lim C. C. W., Saha S., et al. 2020. “The Cost of Mental Disorders: A Systematic Review.” Epidemiology and Psychiatric Sciences 29: e161. 10.1017/s204579602000075x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Clark, C. , Ryan L., Kawachi I., Canner M. J., Berkman L., and Wright R. J.. 2008. “Witnessing Community Violence in Residential Neighborhoods: A Mental Health Hazard for Urban Women.” Journal of Urban Health 85, no. 1: 22–38. 10.1007/s11524-007-9229-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Communauté Métropolitaine de Montréal . 2017. Indice Canopée Métropolitain 2015. https://observatoire.cmm.qc.ca/produits/donnees-georeferencees/.
  11. Côté‐Lussier, C. , Knudby A., and Barnett T. A.. 2020. “A Novel Low‐Cost Method for Assessing Intra‐Urban Variation in Night Time Light and Applications to Public Health.” Social Science & Medicine (1982) 248: 112820. 10.1016/j.socscimed.2020.112820. [DOI] [PubMed] [Google Scholar]
  12. Curry, A. , Latkin C., and Davey‐Rothwell M.. 2008. “Pathways to Depression: The Impact of Neighborhood Violent Crime on Inner‐City Residents in Baltimore, Maryland, USA.” Social Science & Medicine 67, no. 1: 23–30. 10.1016/j.socscimed.2008.03.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Diez Roux, A. V. , and Mair C.. 2010. “Neighborhoods and Health.” Annals of the New York Academy of Sciences 1186: 125–145. 10.1111/j.1749-6632.2009.05333.x. [DOI] [PubMed] [Google Scholar]
  14. Doiron, D. , Setton E. M., Shairsingh K., et al. 2020. “Healthy Built Environment: Spatial Patterns and Relationships of Multiple Exposures and Deprivation in Toronto, Montreal and Vancouver.” Environment International 143: 106003. 10.1016/j.envint.2020.106003. [DOI] [PubMed] [Google Scholar]
  15. Farrall, S. , Gray E., and J. Jackson. 2007. Theorising the Fear of Crime: The Cultural and Social Significance of Insecurities About Crime. Experience & Expression in the Fear of Crime, Working Paper no. 5. 10.2139/ssrn.1012393. [DOI]
  16. Fonseca, F. , Ribeiro P. J. G., Conticelli E., et al. 2022. “Built Environment Attributes and Their Influence on Walkability.” International Journal of Sustainable Transportation 16, no. 7: 660–679. 10.1080/15568318.2021.1914793. [DOI] [Google Scholar]
  17. Forsyth, A. , Wall M., Choo T., Larson N., Van Riper D., and Neumark‐Sztainer D.. 2015. “Perceived and Police‐Reported Neighborhood Crime: Linkages to Adolescent Activity Behaviors and Weight Status.” Journal of Adolescent Health 57, no. 2: 222–228. 10.1016/j.jadohealth.2015.05.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Foster, S. , Giles‐Corti B., and Knuiman M.. 2010. “Neighbourhood Design and Fear of Crime: A Social‐Ecological Examination of the Correlates of Residents' Fear in New Suburban Housing Developments.” Health & Place 16, no. 6: 1156–1165. 10.1016/j.healthplace.2010.07.007. [DOI] [PubMed] [Google Scholar]
  19. Foster, S. , Knuiman M., Villanueva K., Wood L., Christian H., and Giles‐Corti B.. 2014. “Does Walkable Neighbourhood Design Influence the Association Between Objective Crime and Walking?” International Journal of Behavioral Nutrition and Physical Activity 11, no. 1: 100. 10.1186/s12966-014-0100-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Foster, S. , Pereira G., Christian H., Knuiman M., Bull F., and Giles‐Corti B.. 2015. “Neighborhood Correlates of Sitting Time for Australian Adults in New Suburbs: Results From RESIDE.” Environment and Behavior 47, no. 8: 902–922. 10.1177/0013916514533188. [DOI] [Google Scholar]
  21. Frank, L. D. , Appleyard B. S., Ulmer J. M., Chapman J. E., and Fox E. H.. 2021. “Comparing Walkability Methods: Creation of Street Smart Walk Score and Efficacy of a Code‐Based 3D Walkability Index.” Journal of Transport & Health 21: 101005. 10.1016/j.jth.2020.101005. [DOI] [Google Scholar]
  22. Freedman, D. , and Woods G. W.. 2013. “Neighborhood Effects, Mental Illness and Criminal Behavior: A Review.” Journal of Politics and Law 6, no. 3: 1–16. 10.5539/jpl.v6n3p1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Haans, A. , and de Kort Y. A. W.. 2012. “Light Distribution in Dynamic Street Lighting: Two Experimental Studies on Its Effects on Perceived Safety, Prospect, Concealment, and Escape.” Journal of Environmental Psychology 32, no. 4: 342–352. 10.1016/j.jenvp.2012.05.006. [DOI] [Google Scholar]
  24. Hart, T. C. , and Zandbergen P. A.. 2014. “Kernel Density Estimation and Hotspot Mapping: Examining the Influence of Interpolation Method, Grid Cell Size, and Bandwidth on Crime Forecasting.” Policing‐an International Journal of Police Strategies & Management 37: 305–323. [Google Scholar]
  25. Heinze, J. E. , Krusky‐Morey A., Vagi K. J., et al. 2018. “Busy Streets Theory: The Effects of Community‐Engaged Greening on Violence.” American Journal of Community Psychology 62, no. 1–2: 101–109. 10.1002/ajcp.12270. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Heub, A. , Schaller A., and Lange M.. 2024. “Gender‐Specific Perception of Job Stressors and Resources: A Structural Equation Model‐Based Secondary Analysis.” Frontiers in Public Health 12: 1463868. 10.3389/fpubh.2024.1463868. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Hu, L. , and Bentler P. M.. 1998. “Fit Indices in Covariance Structure Modeling: Sensitivity to Underparameterized Model Misspecification.” Psychological Methods 3, no. 4: 424–453. 10.1037/1082-989X.3.4.424. [DOI] [Google Scholar]
  28. Jacobsen, S. K. 2025. “Gendered Responses to Fear of Victimization? A Comparative Study of Students' Precautionary and Avoidance Strategies in Suburban and Urban Contexts.” Violence Against Women 31, no. 6–7: 1383–1403. 10.1177/10778012241228284. [DOI] [PubMed] [Google Scholar]
  29. Kaplan, J. , and Chalfin A.. 2022. “Ambient Lighting, Use of Outdoor Spaces and Perceptions of Public Safety: Evidence From a Survey Experiment.” Security Journal 35, no. 3: 694–724. 10.1057/s41284-021-00296-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Keel, C. , Wickes R., Lee M., and Jackson J.. 2025. “Vulnerability and Place: A Test of the Psychology of Perceived Vulnerability for Women and Men.” Current Issues in Criminal Justice, ahead of print, April 28. 10.1080/10345329.2025.2489229. [DOI] [Google Scholar]
  31. Kessler, R. C. , Berglund P., Demler O., Jin R., Merikangas K. R., and Walters E. E.. 2005. “Lifetime Prevalence and Age‐of‐Onset Distributions of DSM‐IV Disorders in the National Comorbidity Survey Replication.” Archives of General Psychiatry 62, no. 6: 593–602. 10.1001/archpsyc.62.6.593. [DOI] [PubMed] [Google Scholar]
  32. Keyes, C. L. M. 2002. “The Mental Health Continuum: From Languishing to Flourishing in Life.” Journal of Health and Social Behavior 43, no. 2: 207–222. 10.2307/3090197. [DOI] [PubMed] [Google Scholar]
  33. King, K. E. , and Clarke P. J.. 2015. “A Disadvantaged Advantage in Walkability: Findings From Socioeconomic and Geographical Analysis of National Built Environment Data in the United States.” American Journal of Epidemiology 181, no. 1: 17–25. 10.1093/aje/kwu310. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Kondo, M. , Fluehr J., McKeon T., and Branas C.. 2018. “Urban Green Space and Its Impact on Human Health.” International Journal of Environmental Research and Public Health 15, no. 3: 445. 10.3390/ijerph15030445. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Kondo, M. C. , Clougherty J. E., Hohl B. C., and Branas C. C.. 2021. “Gender Differences in Impacts of Place‐Based Neighborhood Greening Interventions on Fear of Violence Based on a Cluster‐Randomized Controlled Trial.” Journal of Urban Health 98, no. 6: 812–821. 10.1007/s11524-021-00580-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Lamers, S. M. A. , Westerhof G. J., Bohlmeijer E. T., ten Klooster P. M., and Keyes C. L. M.. 2011. “Evaluating the Psychometric Properties of the Mental Health Continuum‐Short Form (MHC‐SF).” Journal of Clinical Psychology 67, no. 1: 99–110. 10.1002/jclp.20741. [DOI] [PubMed] [Google Scholar]
  37. Lanfear, C. C. , Matsueda R. L., and Beach L. R.. 2020. “Broken Windows, Informal Social Control, and Crime: Assessing Causality in Empirical Studies.” Annual Review of Criminology 3: 97–120. 10.1146/annurev-criminol-011419-041541. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Lefebvre‐Ropars, G. , and Morency C.. 2018. “Walkability: Which Measure to Choose, Where to Measure It, and How?” Transportation Research Record: Journal of the Transportation Research Board 2672, no. 35: 139–150. 10.1177/0361198118787095. [DOI] [Google Scholar]
  39. Lorenc, T. , Clayton S., Neary D., et al. 2012. “Crime, Fear of Crime, Environment, and Mental Health and Wellbeing: Mapping Review of Theories and Causal Pathways.” Health & Place 18, no. 4: 757–765. 10.1016/j.healthplace.2012.04.001. [DOI] [PubMed] [Google Scholar]
  40. Macassa, G. , Winersjö R., Wijk K., McGrath C., Ahmadi N., and Soares J.. 2017. “Fear of Crime and Its Relationship to Self‐Reported Health and Stress Among Men.” Journal of Public Health Research 6, no. 3: 1010. 10.4081/jphr.2017.1010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Marsh, H. W. , Wen Z., and Hau K. T.. 2004. “Structural Equation Models of Latent Interactions: Evaluation of Alternative Estimation Strategies and Indicator Construction.” Psychological Methods 9, no. 3: 275–300. 10.1037/1082-989X.9.3.275. [DOI] [PubMed] [Google Scholar]
  42. Nichani, V. , Turley L., Vena J. E., and McCormack G. R.. 2021. “Associations Between Neighbourhood Built Characteristics and Sedentary Behaviours Among Canadian Men and Women: Findings From Alberta's Tomorrow Project.” Preventive Medicine 150: 106663. 10.1016/j.ypmed.2021.106663. [DOI] [PubMed] [Google Scholar]
  43. Oliver, L. N. , Schuurman N., and Hall A. W.. 2007. “Comparing Circular and Network Buffers to Examine the Influence of Land Use on Walking for Leisure and Errands.” International Journal of Health Geographics 6, no. 1: 41. 10.1186/1476-072X-6-41. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Paksarian, D. , Rudolph K. E., Stapp E. K., et al. 2020. “Association of Outdoor Artificial Light at Night With Mental Disorders and Sleep Patterns Among US Adolescents.” JAMA Psychiatry 77, no. 12: 1266–1275. 10.1001/jamapsychiatry.2020.1935. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Putrik, P. , van Amelsvoort L., Mujakovic S., et al. 2019. “Assessing the Role of Criminality in Neighbourhood Safety Feelings and Self‐Reported Health: Results From a Cross‐Sectional Study in a Dutch Municipality.” BMC Public Health 19, no. 1: 920. 10.1186/s12889-019-7197-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Robinette, J. W. , Piazza J. R., and Stawski R. S.. 2025. “Community Crime and Safety: An Investigation of Gender Differences in the Daily Stress Process.” Journal of Community Psychology 53, no. 2: e23158. 10.1002/jcop.23158. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Rodrigues, P. , Côté‐Lussier C., Daniel M., et al. 2026. “Novel Canadian Micro‐Neighborhood Crime and Safety Related Stressors: Implications for Individual Safety and Mental Health.” GeoJournal 91, no. 1: 6. 10.1007/s10708-025-11557-5. [DOI] [Google Scholar]
  48. Rodrigues, P. , Hébert M., and Philibert M.. 2022. “Associations Between Neighborhood Characteristics and Dating Violence: Does Spatial Scale Matter?” International Journal of Health Geographics 21, no. 1: 6. 10.1186/s12942-022-00306-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Rodrigues, P. , Philibert M., and Hébert M.. 2024. “Assessing the Association Between Neighbourhood Sociodemographic Characteristics and Adolescent Dating Violence.” Journal of Child and Family Studies 33: 2057–2074. 10.1007/s10826-024-02869-0. [DOI] [Google Scholar]
  50. Roscoe, C. , Sheridan C., Geneshka M., et al. 2022. “Green Walkability and Physical Activity in UK Biobank: A Cross‐Sectional Analysis of Adults in Greater London.” International Journal of Environmental Research and Public Health 19, no. 7: 4247. 10.3390/ijerph19074247. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Rosenfield, S. , and Mouzon D.. 2013. “Gender and Mental Health.” In Handbook of the Sociology of Mental Health, Edited by Aneshensel C. S., Phelan J. C., and Bierman A., 277–296. Springer Netherlands. 10.1007/978-94-007-4276-5_14. [DOI] [Google Scholar]
  52. Ross, C. E. , and Mirowsky J.. 2001. “Neighborhood Disadvantage, Disorder, and Health.” Journal of Health and Social Behavior 42, no. 3: 258–276. [PubMed] [Google Scholar]
  53. Ross, C. E. , and Mirowsky J.. 2009. “Neighborhood Disorder, Subjective Alienation, and Distress.” Journal of Health and Social Behavior 50, no. 1: 49–64. 10.1177/002214650905000104. [DOI] [PubMed] [Google Scholar]
  54. Ross, C. E. , Reynolds J. R., and Geis K. J.. 2000. “The Contingent Meaning of Neighborhood Stability for Residents' Psychological Well‐Being.” American Sociological Review 65, no. 4: 581–597. 10.1177/000312240006500406. [DOI] [Google Scholar]
  55. Sampson, R. J. , and Groves W. B.. 1989. “Community Structure and Crime: Testing Social‐Disorganization Theory.” American Journal of Sociology 94, no. 4: 774–802. http://www.jstor.org/stable/2780858. [Google Scholar]
  56. Santiago, C. D. , Wadsworth M. E., and Stump J.. 2011. “Socioeconomic Status, Neighborhood Disadvantage, and Poverty‐Related Stress: Prospective Effects on Psychological Syndromes Among Diverse Low‐Income Families.” Journal of Economic Psychology 32, no. 2: 218–230. 10.1016/j.joep.2009.10.008. [DOI] [Google Scholar]
  57. Shepley, M. , Sachs N., Sadatsafavi H., Fournier C., and Peditto K.. 2019. “The Impact of Green Space on Violent Crime in Urban Environments: An Evidence Synthesis.” International Journal of Environmental Research and Public Health 16, no. 24: 5119. 10.3390/ijerph16245119. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Shmool, J. L. C. , Yonas M. A., Newman O. D., et al. 2015. “Identifying Perceived Neighborhood Stressors Across Diverse Communities in New York City.” American Journal of Community Psychology 56, no. 1: 145–155. 10.1007/s10464-015-9736-9. [DOI] [PubMed] [Google Scholar]
  59. Shuvo, F. K. , Mazumdar S., and Labib S. M.. 2021. “Walkability and Greenness Do Not Walk Together: Investigating Associations Between Greenness and Walkability in a Large Metropolitan City Context.” International Journal of Environmental Research and Public Health 18, no. 9: 4429. 10.3390/ijerph18094429. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Silverman, B. W. 2018. Density Estimation for Statistics and Data Analysis. Routledge. [Google Scholar]
  61. Skinner, G. C. , and Farrington D. P.. 2024. “Explaining and Reducing the Poor Health of Individuals Who Engage in Criminal Behavior.” In Handbook on Contemporary Issues in Health, Crime, and Punishment, 78–99. Routledge. [Google Scholar]
  62. South, E. C. , Hohl B. C., Kondo M. C., MacDonald J. M., and Branas C. C.. 2018. “Effect of Greening Vacant Land on Mental Health of Community‐Dwelling Adults: A Cluster Randomized Trial.” JAMA Network Open 1, no. 3: e180298. 10.1001/jamanetworkopen.2018.0298. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Stafford, M. , Chandola T., and Marmot M.. 2007. “Association Between Fear of Crime and Mental Health and Physical Functioning.” American Journal of Public Health 97, no. 11: 2076–2081. 10.2105/ajph.2006.097154. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Stafford, M. , Cummins S., Macintyre S., Ellaway A., and Marmot M.. 2005. “Gender Differences in the Associations Between Health and Neighbourhood Environment.” Social Science & Medicine 60, no. 8: 1681–1692. 10.1016/j.socscimed.2004.08.028. [DOI] [PubMed] [Google Scholar]
  65. Statistics Canada . 2017. Census Profile, 2016 Census Statistics Canada. https://www12.statcan.gc.ca/census-recensement/2016/dp-pd/prof/index.cfm?Lang=E.
  66. Statistics Canada . 2018. Dictionary, Census of Population, 2016: Census of Population, 2016. Statistics Canada. http://publications.gc.ca/collections/collection_2018/statcan/98-301-x2016001-eng.pdf.
  67. Statistics Canada . 2019. Canadian Community Health Survey: Public Use Microdata File, 2015/2016 https://www150.statcan.gc.ca/n1/en/catalogue/82M0013X2019001.
  68. Steel, Z. , Marnane C., Iranpour C., et al. 2014. “The Global Prevalence of Common Mental Disorders: A Systematic Review and Meta‐Analysis 1980–2013.” International Journal of Epidemiology 43, no. 2: 476–493. 10.1093/ije/dyu038. [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. Turner, H. A. , Shattuck A., Hamby S., and Finkelhor D.. 2013. “Community Disorder, Victimization Exposure, and Mental Health in a National Sample of Youth.” Journal of Health and Social Behavior 54, no. 2: 258–275. 10.1177/0022146513479384. [DOI] [PubMed] [Google Scholar]
  70. Weisburd, D. , Kuen K., Uding C. V., and Ready J.. 2025. “Antisocial and Prosocial Activities at Crime Hot Spots: Rethinking Conventional Paradigms.” Urban Studies, ahead of print, March 17. 10.1177/00420980251319208. [DOI] [Google Scholar]
  71. Welsh, B. C. , Farrington D. P., and Douglas S.. 2022. “The Impact and Policy Relevance of Street Lighting for Crime Prevention: A Systematic Review Based on a Half‐Century of Evaluation Research.” Criminology & Public Policy 21, no. 3: 739–765. 10.1111/1745-9133.12585. [DOI] [Google Scholar]
  72. Wilson, J. Q. , and G. L. Kelling. 1982. Broken Windows. The Atlantic. https://www.theatlantic.com/magazine/archive/1982/03/broken-windows/304465/.
  73. Wilson, M. 2017. Strengthening the Case for Investing in Canada's Mental Health System: Economic Considerations. Mental Health Commission of Canada. [Google Scholar]
  74. World Health Organization . 2004. Promoting Mental Health: Concepts, Emerging Evidence, Practice: Summary Report. World Health Organization. [Google Scholar]
  75. World Health Organization . 2022. World Mental Health Report: Transforming Mental Health for All. World Health Organization. [Google Scholar]
  76. Zang, P. , Lu Y., Ma J., Xie B., Wang R., and Liu Y.. 2019. “Disentangling Residential Self‐Selection From Impacts of Built Environment Characteristics on Travel Behaviors for Older Adults.” Social Science & Medicine (1982) 238: 112515. 10.1016/j.socscimed.2019.112515. [DOI] [PubMed] [Google Scholar]
  77. Zhang, C. J. P. , Barnett A., Sit C. H. P., et al. 2019. “To What Extent Does Physical Activity Explain the Associations Between Neighborhood Environment and Depressive Symptoms in Older Adults Living in an Asian Metropolis?” Mental Health and Physical Activity 16: 96–104. 10.1016/j.mhpa.2018.11.005. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

Canadian Community Health Survey data are provided by the Québec Inter‐University Centre for Social Statistics and are not publicly available. All other data used in the current study are available from the corresponding author on reasonable request.

Articles from Journal of Community Psychology are provided here courtesy of Wiley

RESOURCES