Hostname: page-component-78c5997874-4rdpn Total loading time: 0 Render date: 2024-11-14T23:21:59.780Z Has data issue: false hasContentIssue false

Not urbanization level but socioeconomic, physical and social neighbourhood characteristics are associated with presence and severity of depressive and anxiety disorders

Published online by Cambridge University Press:  15 March 2018

Ellen Generaal*
Affiliation:
Department of Psychiatry, Amsterdam Public Health Research Institute, VU University Medical Center, Amsterdam, the Netherlands
Erik J. Timmermans
Affiliation:
Department of Epidemiology and Biostatistics, Amsterdam Public Health Research Institute, VU University Medical Center, Amsterdam, the Netherlands
Jasper E. C. Dekkers
Affiliation:
Spatial Information Laboratory, Department of Spatial Economics, Faculty of Economics and Business Administration, VU University Amsterdam, Amsterdam, the Netherlands
Johannes H. Smit
Affiliation:
Department of Psychiatry, Amsterdam Public Health Research Institute, VU University Medical Center, Amsterdam, the Netherlands
Brenda W. J. H. Penninx
Affiliation:
Department of Psychiatry, Amsterdam Public Health Research Institute, VU University Medical Center, Amsterdam, the Netherlands
*
Author for correspondence: Ellen Generaal, E-mail: [email protected]
Rights & Permissions [Opens in a new window]

Abstract

Background

Which neighbourhood factors most consistently impact on depression and anxiety remains unclear. This study examines whether objectively obtained socioeconomic, physical and social aspects of the neighbourhood in which persons live are associated with the presence and severity of depressive and anxiety disorders.

Methods

Cross-sectional data are from the Netherlands Study of Depression and Anxiety including participants (n = 2980) with and without depressive and anxiety disorders in the past year (based on DSM-based psychiatric interviews). We also determined symptom severity of depression (Inventory of Depression Symptomatology), anxiety (Beck Anxiety Inventory) and fear (Fear Questionnaire). Neighbourhood characteristics comprised socioeconomic factors (socioeconomic status, home value, number of social security beneficiaries and percentage of immigrants), physical factors (air pollution, traffic noise and availability of green space and water) and social factors (social cohesion and safety). Multilevel regression analyses were performed with the municipality as the second level while adjusting for individual sociodemographic variables and household income.

Results

Not urbanization grade, but rather neighbourhood socioecononomic factors (low socioeconomic status, more social security beneficiaries and more immigrants), physical factors (high levels of traffic noise) and social factors (lower social cohesion and less safety) were associated with the presence of depressive and anxiety disorders. Most of these neighbourhood characteristics were also associated with increased depressive and anxiety symptoms severity.

Conclusion

These findings suggest that it is not population density in the neighbourhood, but rather the quality of socioeconomic, physical and social neighbourhood characteristics that is associated with the presence and severity of affective disorders.

Type
Original Articles
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution, and reproduction in any medium, provided the original work is properly cited.
Copyright
Copyright © Cambridge University Press 2018

Introduction

The role of neighbourhood characteristics in health research is receiving growing attention over the past 20 years (Arcaya et al. Reference Arcaya, Tucker-Seeley, Kim, Schnake-Mahl, So and Subramanian2016). Besides impacts on physical health (Diez Roux & Mair, Reference Diez Roux and Mair2010), an increasing amount of studies also focused on the potential psychological impacts of the living environment (Truong & Ma, Reference Truong and Ma2006; Peen et al. Reference Peen, Schoevers, Beekman and Dekker2010; Gong et al. Reference Gong, Palmer, Gallacher, Marsden and Fone2016). Urbanization and socioeconomic, physical and social neighbourhood characteristics might be associated with mental health through environmental stressors, such as overcrowding, violence and low social cohesion and social support (Evans, Reference Evans2003; Srivastava, Reference Srivastava2009; Blair et al. Reference Blair, Ross, Gariepy and Schmitz2014; van Deurzen et al. Reference van Deurzen, Rod, Christensen, Hansen, Lund and Dich2016). In addition, the neighbourhood might relate to mental health through their an unfavourable impact on biological stress regulation pathways, such as those involved in immune (Calderón-Garcidueñas et al. Reference Calderón-Garcidueñas, Solt, Henríquez-Roldán, Torres-Jardón, Nuse and Herritt2008; Chen et al. Reference Chen, Rey, Kwong, Copes, Tu and Villeneuve2017) or HPA-axis (Dulin-Keita et al. Reference Dulin-Keita, Casazza, Fernandez, Goran and Gower2012; Hackman et al. Reference Hackman, Betancourt, Brodsky, Hurt and Farah2012) regulation. However, which neighbourhood factors are most consistently associated with depressive and anxiety disorders remains rather unclear, mainly because previous studies are heterogeneous in their measurement of psychiatric outcomes, neighbourhood factors and confounders (Truong & Ma, Reference Truong and Ma2006; Diez Roux, Reference Diez Roux2007; Peen et al. Reference Peen, Schoevers, Beekman and Dekker2010; Gong et al. Reference Gong, Palmer, Gallacher, Marsden and Fone2016).

Studies found higher prevalences for clinical depressive and anxiety disorders in urban areas as compared to rural areas in several developed countries (Peen et al. Reference Peen, Schoevers, Beekman and Dekker2010; Zijlema et al. Reference Zijlema, Klijs, Stolk and Rosmalen2015). However, a recent study showed that increased urbanization was not significantly associated with DSM-based depressive and anxiety disorders, but only with self-reported indicators of mental health (de Vries et al. Reference de Vries, ten Have, van Dorsselaer, van Wezep, Hermans and de Graaf2016). Regarding socioeconomic neighbourhood characteristics, both person-level and neighbourhood-level income status has previously been associated with depressive symptoms, depressive disorders and major depressive episodes (Kim, Reference Kim2008; Annequin et al. Reference Annequin, Weill, Thomas and Chaix2015; Klijs et al. Reference Klijs, Kibele, Ellwardt, Zuidersma, Stolk and Wittek2016).

Previous studies have also covered the physical neighbourhood in relation to depression and anxiety outcomes (Miles et al. Reference Miles, Coutts and Mohamadi2012; de Vries et al. Reference de Vries, ten Have, van Dorsselaer, van Wezep, Hermans and de Graaf2016; Zijlema et al. Reference Zijlema, Cai, Doiron, Mbatchou, Fortier and Gulliver2016a, Reference Zijlema, Wolf, Emeny, Ladwig, Peters and Kongsgårdb). Some used subjective measures of ‘perception’ of the physical neighbourhood (Ruijsbroek et al. Reference Ruijsbroek, Mohnen, Droomers, Kruize, Gidlow and Gražulevičiene2017) whereas others used objective census data (Miles et al. Reference Miles, Coutts and Mohamadi2012; de Vries et al. Reference de Vries, ten Have, van Dorsselaer, van Wezep, Hermans and de Graaf2016), methods which have shown to yield different results (Ball et al. Reference Ball, Jeffery, Crawford, Roberts, Salmon and Timperio2008; Gebel et al. Reference Gebel, Bauman, Sugiyama and Owen2011; Linden & Vodermaier, Reference Linden and Vodermaier2012). One recent multi-cohort study found inconsistent evidence for an association of subjective reports of more green space and better green space quality with lower depressive symptoms among four European countries (Ruijsbroek et al. Reference Ruijsbroek, Mohnen, Droomers, Kruize, Gidlow and Gražulevičiene2017). Another study showed that census-data-based green space availability was associated with a lower prevalence of anxiety disorders, but not with depressive disorders (de Vries et al. Reference de Vries, ten Have, van Dorsselaer, van Wezep, Hermans and de Graaf2016). In that study, more water availability in the neighbourhood was associated with the presence of both anxiety and depressive disorders (de Vries et al. Reference de Vries, ten Have, van Dorsselaer, van Wezep, Hermans and de Graaf2016). Regarding other physical neighbourhood factors, air pollution has consistently been associated with physical health outcomes, such as pulmonary disease (Gehring et al. Reference Gehring, Gruzieva, Agius, Beelen, Custovic and Cyrys2013) and cardiovascular disease (Cesaroni et al. Reference Cesaroni, Forastiere, Stafoggia, Andersen, Badaloni and Beelen2014), but associations with affective disorders are less clear and absent in a recent European study (Zijlema et al. Reference Zijlema, Cai, Doiron, Mbatchou, Fortier and Gulliver2016Reference Zijlema, Wolf, Emeny, Ladwig, Peters and Kongsgårdb). Similarly, traffic noise has frequently been associated with physical outcomes such as high heart rate (Zijlema et al. Reference Zijlema, Cai, Doiron, Mbatchou, Fortier and Gulliver2016a), but inconclusive evidence was found for an association with depressive and anxiety symptoms, as indicated in a recent systematic review (Tzivian et al. Reference Tzivian, Winkler, Dlugaj, Schikowski, Vossoughi and Fuks2015).

The association between the social neighbourhood and depression and anxiety also remains to be elucidated, as most previous studies measured the subjective experience of social environmental factors, such as safety and poor social interactions (Mair et al. Reference Mair, Diez Roux and Galea2008; Ruijsbroek et al. Reference Ruijsbroek, Droomers, Hardyns, Groenewegen and Stronks2016). This might result in bias because poor mental health may change perception and influence self-report of the environment (Diez Roux & Mair, Reference Diez Roux and Mair2010). This underlines the need of objectively measured neighbourhood factors by using national registration systems. Besides this, most previous studies (Kim, Reference Kim2008; Mair et al. Reference Mair, Diez Roux and Galea2008; Diez Roux & Mair, Reference Diez Roux and Mair2010; Zijlema et al. Reference Zijlema, Klijs, Stolk and Rosmalen2015; de Vries et al. Reference de Vries, ten Have, van Dorsselaer, van Wezep, Hermans and de Graaf2016; Klijs et al. Reference Klijs, Kibele, Ellwardt, Zuidersma, Stolk and Wittek2016) were limited in examining only one aspect of the neighbourhood even though neighbourhood factors are highly intercorrelated. Large-scale multivariable studies are needed to test whether associations still hold when adjusting for other relevant neighbourhood factors such as urbanization grade and for individual indicators of socioeconomic status (Kim, Reference Kim2008). Strengths of the current study are (1) the inclusion of objective measures of a wide range of neighbourhood characteristics, (2) the assessment of depression and anxiety using semi-structured interviews according to DSM standards (Robins et al. Reference Robins, Wing, Wittchen, Helzer, Babor and Burke1988) and (3) the comparison with results from depression and anxiety symptom scales.

This large-scale cross-sectional study (N = 2980) examines to what extent (1) urbanization grade; (2) neighbourhood socioeconomic characteristics, i.e. socioeconomic status, home value, number of social security beneficiaries and percentage of immigrants; (3) neighbourhood physical characteristics, i.e. air pollution, traffic noise and availability of green space and water; and (4) neighbourhood social characteristics, i.e. social cohesion and safety, are (independently) associated with the presence and severity of depressive and/or anxiety disorders, while adjusting for individual socioeconomic variables.

Methods

Sample

The current cross-sectional study used data from the Netherlands Study of Depression and Anxiety (NESDA): an ongoing longitudinal cohort study in which 2981 participants (18–65 years at baseline) were followed to investigate the long-term course and consequences of depressive and anxiety disorders. To establish inclusion of patients at different developmental stages of their disorder, participants were recruited from the community (19%), from primary care (54%) and from specialized mental health care (27%). At baseline (2004–2007), persons with a range of psychopathology were included: from those without a depressive or anxiety disorder to those with a current, first or recurrent (in the past year) depressive or anxiety disorder and those with a remitted disorder (at baseline, a depressive and/or anxiety disorder was diagnosed in the past, but no diagnoses were present in the year before baseline). Exclusion criteria were not being fluent in Dutch and a primary clinically overt diagnosis of other psychiatric (e.g. diagnosis of psychotic, obsessive-compulsive, bipolar or severe addiction) disorder. Most NESDA participants were recruited from three different regions in the Netherlands, surrounding the cities of Amsterdam (north-west of the Netherlands), Leiden (mid-west) or Groningen (north-east). These cities have a variety of urbanization (cities with ±800.000, ±120.000 and ±200.000 habitants, respectively), and also persons from the larger rural areas around these cities were recruited. As compared with the whole Netherlands, NESDA included more persons from the extremely and strongly urbanized areas (65% v. 40%) and less from the hardly and not urbanized areas (16% v. 42%). Nonetheless, these more rural areas are still well represented in NESDA (±16% equals n = 481) and there is a rather large variation in urbanization grades within our study, which is comparable to the entire Netherlands. In NESDA, we included subjects from 666 different postal code areas (out of a total of ±4000 four-digit postal code areas in the Netherlands; CBS, 2006f) and from 200 different municipalities (out of a total of 458 municipalities in the Netherlands; CBS, 2005). On average, 25 participants live in one postal code area (range = 1–105), and 338 participants live in one municipality (range 1–787) within our study.

The research protocol was approved by the Ethical Committee of participating universities and written informed consent was obtained from all participants. Penninx et al. provide a detailed description of the NESDA study design and sampling procedures (Penninx et al. Reference Penninx, Beekman, Smit, Zitman, Nolen and Spinhoven2008).

Data for the current study were available of 2980 participants due to a failure to administer one postal code by the staff. Due to item missing values, the sample varied from n = 2966 [for a socioeconomic status score (SES)] to n = 2980 (for urbanization grade, home value and percentage of immigrants).

Presence and severity of depressive and anxiety disorders

The presence of current diagnoses of depressive disorders (major depressive disorder and dysthymia) and anxiety disorders (panic disorder, agoraphobia, generalized anxiety disorder and social phobia) were determined in the year prior to baseline interview using the semi-structured Composite International Diagnostic Interview (WHO version 2.1) according to DSM-IV criteria (Robins et al. Reference Robins, Wing, Wittchen, Helzer, Babor and Burke1988). Controls were defined as individuals without current diagnoses of depressive and anxiety disorders.

Severity of depressive and anxiety disorders were measured using the 28-item Inventory of Depression Symptomatology (IDS, range 0–84) (Rush et al. Reference Rush, Gullion, Basco, Jarrett and Trivedi1996), the 21-item Beck Anxiety Inventory (BAI, range 0–63) (Beck et al. Reference Beck, Epstein, Brown and Steer1988) and the 15-item Fear Questionnaire (FQ, range 0–120) (Marks & Mathews, Reference Marks and Mathews1979). The IDS measures the overall severity of depression in the past week, whereas the BAI assesses the overall severity of anxiety in the past week, e.g. feeling nervous, shaky or dizzy. The FQ reflects the level of avoidance (0 = ‘never avoid it’ to 8 = ‘always avoid it’) related to symptoms of social phobia, blood phobia and agoraphobia, e.g. fear of talking to people in authority, fear of injections and fear of large public spaces. Psychometric properties for the IDS (IDS, 2017), the BAI (Beck et al. Reference Beck, Epstein, Brown and Steer1988; Beck & Steer, Reference Beck and Steer1990) and the FQ (van Zuuren, Reference van Zuuren1988) have been well established.

Neighbourhood factors

Data of neighbourhood factors were retrieved from national registration organizations [e.g. Statistics Netherlands (CBS)]. Details have been described elsewhere (Timmermans et al., Reference Timmermans, Lakerveld, Beulens, Boomsma, Kram and Oosterman2018; CBS, 2006a, g). All neighbourhood factors were aggregated to mean values for each four-digit postal code area and subsequently linked to NESDA cohort data using respondents’ four-digit postal codes at baseline assessment. For all neighbourhood factors, data were used of the year 2006 (CBS, 2006a, c), with the exception of noise (2007) and air pollution (2009) as those were unavailable in 2006. Baseline assessment of NESDA took place between September 2004 and February 2007, and 89% of our sample were interviewed in 2005 or 2006.

Correlations of neighbourhood factors within each domain are moderate to high (see online Supplementary Table 1). The range in correlations were r =  −0.83 to −0.17 within the socioeconomic domain; r =  −0.47 to 0.66 within the physical domain, and r = 0.61 within the social domain, suggesting that all neighbourhood factors partly represent differential aspects of neighbourhood quality. In addition, correlations of neighbourhood factors across domains are generally somewhat lower than those within one domain (see online Supplementary Table 1).

Urbanization grade

Urbanization grade is the mean number of addresses per squared kilometre within a circle with a radius of 1 km (den Dulk et al. Reference den Dulk, van de Stadt and Vliegen1992). Data are provided by Statistics Netherlands and are defined in five categories: (1) extremely urbanized (⩾2500 addresses/km2), (2) strongly urbanized (1500–2500 addresses/km2), (3) moderately urbanized (1000–1500 addresses/km2), (4) hardly urbanized (500–1000 addresses/km2) and (5) not urbanized (<500 addresses/km2).

Socioeconomic neighbourhood factors

SESs were obtained from the Netherlands Institute of Social Research (SCP, 2017). The SES score is based on three dimension scores: education, income and position in the labour market. A principal component analyses defined the SES score based on mean income, percentage low incomes, percentage of low educated residents and percentage of unemployed residents. More details are available elsewhere (Knol, Reference Knol2012; SCP, 2017).

The average home value (in k€) was provided by Statistics Netherlands (CBS, 2017a). Residential housing and residential housing with (private) practices were included, whereas recreational housing and other non-residential housing such as garage boxes were excluded from the calculation of the score (CBS, 2006c; WOZ, 2017).

The number of social security beneficiaries per 1000 households was also provided by Statistics Netherlands (CBS, 2006c). Social security benefits in the Netherlands pertain to long-term unemployment, either due to disability or the inability to (find) work. We excluded benefits issued to residents who lived in (healthcare) institutions (CBS, 2006c, 2014).

The percentage of immigrants (from all continents) among all residents was provided by Statistics Netherlands (CBS, 2006c). Residents were considered as immigrants if they were born outside the Netherlands with one or both parents born outside the Netherlands (first generation) or if they were born inside the Netherlands and one or both parents were born abroad (second generation). This is the most widely accepted and valid assessment of ethnicity in the Netherlands (Stronks et al. Reference Stronks, Kulu-Glasgow and Agyemang2009). More details are described elsewhere (CBS, 2006c, d).

Physical neighbourhood factors

Daily mean noise of road- rail- and air traffic for several years were modelled by the Netherlands Environmental Assessment Agency by using the Empara Noisetool with a resolution of 25 × 25 m (PBL, 2017). This study uses noise data of 2007. Noise is measured in Level day–evening–night (Lden) and is expressed in decibels [dB (A)]. The measure Lden accounts for the fact that noise in the evening and the night are more annoying than during the day. The average noise levels during the day (7–19 h), the evening (19–23 h) and the night (23–7 h) were calculated first and the levels of noise in the evening and the night are increased with 5 and 10 dB (A), respectively. Subsequently, the daily mean noise was calculated by dividing the noise levels during day, evening and night by 3. The modelling of road-, rail- and air traffic noise accounts for several factors, such as sound barriers, rail type and flight paths. The noise data were linked to all addresses that were included in the Register of Addresses and Building (BAG-register, June 2015) of the Netherlands’ Cadastre, Land Registry, and Mapping Agency, by using GeoDMS software (Object Vision BV, Amsterdam, the Netherlands). More details on the assessment of traffic noise are available elsewhere (PBL, 2017). Traffic noise data for the Netherlands appears highly correlated over the years 2000–2008 (all Pearsons’ r = 0.98–0.99, with the exception of 2008 with 2000–2007: Pearsons’ r = 0.82), suggesting that noise exposure in the neighbourhood is relatively stable over time.

Annual average air pollution concentrations at the subjects’ home addresses were estimated by land-use regression models for the year 2009 by the Institute for Risk Assessment Sciences as part of the European Study of Cohorts for Air Pollution Effects (ESCAPE-project), as described elsewhere (Eeftens et al. Reference Eeftens, Beelen, de Hoogh, Bellander, Cesaroni and Cirach2012; Beelen et al. Reference Beelen, Hoek, Vienneau, Eeftens, Dimakopoulou and Pedeli2013). The current study uses data of the annual average concentrations of the mean blackness of PM2.5 filters, which is a proxy for elemental carbon (soot), the dominant light absorbing substance. Previous studies from Europe support the stability of spatial contrasts in black carbon levels over periods of 10 years and more (Eeftens et al. Reference Eeftens, Beelen, Fischer, Brunekreef, Meliefste and Hoek2011; Gulliver et al. Reference Gulliver, Morris, Lee, Vienneau, Briggs and Hansell2011; Cesaroni et al. Reference Cesaroni, Forastiere, Stafoggia, Andersen, Badaloni and Beelen2014).

The percentage of green space and blue space (water) of total land use was calculated using the Calculate Geometry-option in ArcGIS (ArcGIS, 2017) and the neighbourhood map from Statistics Netherlands (CBS, 2006g). Categories of green space included recreation (e.g. parks, gardens), agriculture and forest or nature (see CBS, 2006b). Recreational green space had a minimal size of either 0.1 hectare (ha) (garden complexes), 0.5 ha (sports fields) or 1 ha (e.g. parks or other recreational property) in order to be included (CBS, 2006b). For agriculture and forests/nature, this lower limit was 1 ha (CBS, 2006b). Blue space included inland water, sea and (large) lakes. Minimal size were either 0 ha for large water areas (sea, rivers or big lakes were included independent of size) or 0.5 to 1 ha for other types of inland water (e.g. canals or lakes) and recreational water (e.g. ponds). For an exact description of green and blue space, we refer to CBS Soil Statistics (CBS, 2006b).

Social neighbourhood factors

Social cohesion and safety in the neighbourhood have been measured as part of the assessment of neighbourhood livability by the Netherlands Ministry of the Interior and Kingdom Relations (Leidelmeijer & van Kamp, Reference Leidelmeijer and van Kamp2003; Leidelmeijer et al. Reference Leidelmeijer, Marlet, Van Iersel, Van Woerkens and van der Reijden2008; Livability scores, 2015). The social cohesion score (range, −50 to 50) is based on indicators such as a number of relocations and homogeneity of the family composition. The safety score (range, −50 to 50) is based on indicators such as annoyance (of drug abuse, loitering youth, neighbours, destruction, rubbish and daubing/graffiti) and number of crimes (Leefbaarometer, 2015).

Covariates

Basic covariates included the following person-level variables: age, sex, years of education and household income (net € per month). These covariates are in line with previous studies on neighbourhood factors and mental health (de Vries et al. Reference de Vries, ten Have, van Dorsselaer, van Wezep, Hermans and de Graaf2016; Zijlema et al. Reference Zijlema, Cai, Doiron, Mbatchou, Fortier and Gulliver2016a, Reference Zijlema, Wolf, Emeny, Ladwig, Peters and Kongsgårdb). The total income of the household was measured as a continuous variable (range, 1–24) with intervals of €200 for each value (i.e. 1 = ‘<€600’; 2 = ‘€600–800’; until 23 = ‘€4801–5000’; 24 = ‘>€5000’). In NESDA, one household comprises on average two persons, with a range from 1 to 10 persons (s.d. = 0.70). Missing values for baseline household income (n = 41) were imputed using NESDA follow-up data (first observation carried backward).

Statistical analyses

Sample baseline characteristics were compared between subjects with and without depressive and/or anxiety disorders using independent-samples t tests for continuous variables, χ2 tests for dichotomous and categorical variables, and Mann–Whitney U tests for non-normally distributed variables. Descriptive analyses were computed in SPSS V.20 for Windows (SPSSinc, Chicago, USA).

Multilevel logistic and linear regression analyses were conducted to examine the associations between each neighbourhood factor with (1) the presence of current depressive and/or anxiety disorders, and (2) the severity of depressive, anxiety and fear symptoms (separate models) for participants (level 1), nested within municipalities (level 2). We first conducted main analyses for the presence of depression and/or anxiety v. controls, followed by analyses separately for depression and anxiety to test for differences across type of disorder. To take intra-neighbourhood correlation into account (Merlo et al. Reference Merlo, Chaix, Ohlsson, Beckman, Johnell and Hjerpe2006), municipality code was included as a random intercept in the multilevel analyses, and this appeared of additive value (intraclass correlation coefficient = 0.15). Municipality codes/postal codes combinations were determined using CBS data of 2005 (CBS, 2005). Multilevel analyses were computed with R (version 3.3.2), using functions glmer and lmer from the ‘lme4’ package (version 1.1–12) (R software, 2017). Output of fixed parameter estimates were expressed as odds ratio for diagnoses of depression and anxiety, and as standardized regression coefficient (β) for severity of symptoms. Bootstrap confidence intervals were calculated (1000 bootstrap samples) using the ‘pbkrtest’ package (for lmer) and the Wald method of the confint.merMod function (for glmer). All multilevel analyses were adjusted for age, sex, years of education and household income. For all statistical tests, a probability level of ⩽5% was regarded as significant.

Separate multilevel regression analyses were conducted for each neighbourhood factor with the presence and severity of depression and anxiety. Subsequently, to test whether findings are confounded by other neighbourhood factors, we performed multivariable analyses including all neighbourhood factors. Variance inflation factors (Zuur et al. Reference Zuur, Ieno and Elphick2010) of multivariable models were below 4.3, which is considered acceptable for multicollinearity according to previous guidelines (Bowerman & O'Connell, Reference Bowerman and O'Connell1990; Myers, Reference Myers1990; Field, Reference Field2013).

To check whether our associations were affected by instability in home address of participants around the assessment period, we performed additional sensitivity analyses (1) excluding persons who lived at their home address for <2 years at baseline assessment (remaining n = 2485), and (2) excluding persons who moved between baseline and follow-up measurement in 2006–2009 (remaining n = 1986). Moreover, since the control group in our study (n = 1197) also includes individuals with remitted diagnoses of depressive and anxiety disorders, we performed additional analyses excluding persons from the control group who had experienced depressive and anxiety disorders earlier during their lifetime (remaining n = 2435 of which 652 controls).

Results

Baseline characteristics of the study population are shown in Table 1. In subjects with current diagnoses of depressive and/or anxiety disorders, median score for depressive symptoms was 28 (IQR = 19–37), median score for anxiety symptoms was 15 (IQR = 9.0–23) and median score for fear symptoms was 30 (IQR = 16–46). Compared with controls, subjects with depression and/or anxiety were significantly younger, had less years of education and a lower household income (all p < 0.05). Individuals with depression/anxiety seem to live in neighbourhoods with lower socioeconomic status, higher levels of air pollution and higher levels traffic noise, more water and lower social cohesion as compared with controls (See Table 1: p < 0.05).

Table 1. Baseline characteristicsa

a Values are illustrated as mean (standard deviation; s.d.), median (interquartile range; IQR) or percentages (%). Household income is presented in categories for illustrative purposes; a continuous variable was used in the multilevel regression analyses. N-values could vary from total N = 2980 due to missings on individual measures.

Neighbourhood characteristics and current diagnoses of depression and anxiety

Table 2 reports the associations of neighbourhood characteristics with current diagnoses of depressive disorders and anxiety disorders from multilevel logistic regression analyses (N = 2980). No significant associations were found for urbanization grade with depressive and anxiety disorders, neither in the main analyses nor in the analyses separately for depression and anxiety (p > 0.05). However, several socioeconomic, physical and social neighbourhood factors did appear associated with depressive and/or anxiety disorders (Table 2; p ⩽ 0.05). More specifically, neighbourhoods with low socioeconomic status (OR[95%CI] = 0.88[0.79–0.98]), a high number of social security beneficiaries [OR (95% CI) = 1.24 (1.11–1.39)], more immigrants (OR[95%CI] = 1.23(1.05, 1.44]), high levels of traffic noise [OR (95% CI) = 1.17 (1.03–1.32)], lower social cohesion [OR (95%CI) = 0.84 (0.76–0.93)] and less safety [OR (95% CI) = 0.84 (0.71–1.00)] were significantly associated with depressive disorders (Table 2, column 2; all p ⩽ 0.05). Of these factors, social security beneficiaries, traffic noise and lower social cohesion remained significantly associated with depression in multivariable analyses combining all neighbourhood factors (Table 2). In addition, higher home value and more water in the neighbourhood became significantly associated with depression in these multivariable analyses [OR (95% CI) = 1.29 (1.08–1.55) and OR (95%CI) = 1.13 (1.01–1.26)].

Table 2. Associations between neighbourhood characteristics and current diagnoses of depressive and anxiety disorders

Based on multilevel logistic regression analyses including municipality code as random intercept, adjusted for sex, age, education and household income. *p ⩽ 0.05; **p ⩽ 0.01. All continuous neighbourhood characteristics were standardized: OR are per 1 s.d. increase; socioeconomic status score: s.d. = 1.18, home value: s.d. = 70.5, social security beneficiaries: s.d. = 43.0, immigrants: s.d. = 15.0, road-, rail- and air-traffic noise: s.d. = 3.21, air pollution (PM2.5 absorbance): s.d. = 0.27, green space: s.d. = 22.6, water: s.d. = 6.21, safety: s.d. = 25.8, social cohesion: s.d. = 13.8.

a Urbanization grade was entered as a dichotomous variable (more or less than 1500 addresses/km2) in the multivariable model for ease of presentation.

Regarding anxiety, low neighbourhood socioeconomic status [OR (95% CI) = 0.88 (0.80–0.97)], more social security beneficiaries [OR (95% CI) = 1.21 (1.09–1.36)], more immigrants [OR (95%CI) = 1.16 (1.00–1.35)], high levels of air pollution [OR (95% CI) = 1.21 (1.04–1.41)], high levels of traffic noise [OR (95% CI) = 1.22 (1.09–1.38)], less green space [OR (95% CI) = 0.89 (0.79–1.00)], more water in the neighbourhood [OR (95% CI) = 1.12 (1.02–1.24)], lower social cohesion [OR (95% CI) = 0.86 (0.78–0.94)] and less safety [OR (95%CI) = 0.79 (0.67–0.93)] were significantly associated with anxiety disorders (Table 2, column 3; all p ⩽ 0.05). Of these factors, social security beneficiaries, traffic noise and water in the neighbourhood remained significantly associated with anxiety in multivariable analyses (Table 2). In addition, higher home value became significantly associated with anxiety in these multivariable analyses [OR (95% CI) = 1.37 (1.16–1.62)]. Thus, overall, not urbanization grade but rather socioeconomic, physical and social neighbourhood factors were associated with depression and anxiety, even after adjustment for covariates and other neighbourhood factors.

Additional sensitivity analyses only including subjects who lived at their home address for 2 years or longer (n = 2485) showed similar effect sizes for all neighbourhood factors that were statistically significant with the prevalence of depressive and/or anxiety disorder in Table 2 (ΔOR range, −6% to +3%). Also, analyses only including persons that did not report any relocations between baseline and follow-up measurement in 2006–2009 (n = 1986) resulted in comparable effect sizes (ΔOR range, −7% to + 6%). Overall, this suggests our findings seem rather stable and are not affected much by instability in home address around our assessment period. Additional sensitivity analyses including a more ‘clean’ control group without lifetime diagnoses of depression and anxiety (n = 2435 of which 652 controls) showed similar or even somewhat larger effect sizes (see online Supplementary Table 2). All earlier found associations with the prevalence of depressive and/or anxiety disorders remained statistically significant except the association for traffic noise [OR (95% CI) = 1.11 (0.96–1.27), p = 0.15] and water in the neighbourhood [OR (95% CI) = 1.07 (0.95–1.19), p = 0.26]. On the contrary, green space became significantly associated with lower prevalence of depression and/or anxiety disorders when using this ‘clean’ control group [OR (95% CI) = 0.84 (0.73–0.96), p = 0.01].

Previous research suggests multilevel data should include at least five cases per cluster (Clarke, Reference Clarke2008; McNeish & Stapleton, Reference McNeish and Stapleton2014). Thus, we performed sensitivity analyses only including subjects from municipalities with at least five persons (n = 2721). These sensitivity analyses performed for the variables that were significant in univariable analyses (Table 2, column 1) showed that odds ratios (OR) for all neighbourhood factors were similar (ΔOR ranging from −4% to 0%) in association with the prevalence of depressive and/or anxiety disorder. To be precise, effect sizes were as follows: socioeconomic status: OR (95% CI) = 0.88 (0.80–0.97); social security beneficiaries: OR (95% CI) = 1.20 (1.08–1.34); immigrants: OR (95% CI) = 1.13 (0.97–1.31); traffic noise: OR (95% CI) = 1.12 (0.99–1.27); water: OR (95% CI) = 1.08 (0.98, 1.19); social cohesion: OR (95% CI) = 0.85 (0.77–0.93); safety: OR (95%CI) = 0.81 (0.68–0.97). Thus, the exclusion of participants from municipalities with low sample size does not change our findings.

Neighbourhood characteristics and the severity of depressive, anxiety and fear symptoms

Multilevel linear regression analyses with symptom severity confirmed that not urbanization grade, but rather socioeconomic, physical and social neighbourhood characteristics were significantly associated with depressive, anxiety and fear symptoms, also after adjusting for covariates (Table 3). A higher number of social security beneficiaries and lower levels of safety in the neighbourhood were associated with increased levels of all types of symptoms. In addition, lower socioeconomic status, more immigrants and more traffic noise in the neighbourhood were associated with increased depressive and anxiety symptoms. Lower social cohesion was associated with increased depressive and fear symptoms. Multivariable analyses combining all neighbourhood factors showed that these effects remained significant for social security beneficiaries and traffic noise with depressive and anxiety symptoms (Table 3). Also, more water became significantly associated with increased depressive symptoms, and higher home value became significantly associated with increased depressive and fear symptoms in these multivariable analyses.

Table 3. Associations between neighbourhood characteristics and the severity of depressive, anxiety and fear symptoms

Based on multilevel linear regression analyses including municipality code as random intercept, adjusted for sex, age, education and household income. *p ⩽ 0.05; **p ⩽ 0.01. All symptom scales and continuous neighbourhood characteristics were standardized. Fixed parameter estimate (β) is per 1 s.d. increase; socioeconomic status score: s.d. = 1.18, home value: s.d. = 70.5, social security beneficiaries: s.d. = 43.0, immigrants: s.d. = 15.0, road-, rail- and air traffic noise: s.d. = 3.21, air pollution (PM2.5 absorbance): s.d. = 0.27, green space: s.d. = 22.6, water: s.d. = 6.21, safety: s.d. = 25.8, social cohesion: s.d. = 13.8.

a Urbanization grade was entered as a dichotomous variable (more or less than 1500 addresses/km2) in the multivariable model for ease of presentation.

Discussion

This cross-sectional study (N = 2980) showed that not urbanization grade, but rather socioeconomic, physical and social characteristics of the neighbourhood were associated with the presence and severity of depression and anxiety, independent of individual indicators of socioeconomic status.

These findings add to the body of evidence that characteristics of the neighbourhood seem to be associated with depression and anxiety (Evans, Reference Evans2003; Kim, Reference Kim2008; Mair et al. Reference Mair, Diez Roux and Galea2008; Diez Roux & Mair, Reference Diez Roux and Mair2010; Gariepy et al. Reference Gariepy, Kaufman, Blair, Kestens and Schmitz2014). Overall, our results were consistent across different assessment methods (DSM-based interviews v. symptom scales). Contrary to previous research (de Vries et al. Reference de Vries, ten Have, van Dorsselaer, van Wezep, Hermans and de Graaf2016) this might suggest that the neighbourhood context is relevant for both clinical diagnoses of depression and anxiety as well as more general reflections of negative mood. Our findings may implicate that affective disorders are not related to population density in the neighbourhood, but rather with the quality of the neighbourhood. This may be surprising in light of previous research which consistently showed urbanicity to be associated with increased prevalence of depressive and anxiety disorders (Peen et al. Reference Peen, Schoevers, Beekman and Dekker2010; Zijlema et al. Reference Zijlema, Klijs, Stolk and Rosmalen2015) and other types of psychiatric disorders, such as schizophrenia and psychosis (Krabbendam & van Os, Reference Krabbendam and van Os2005). On the other hand, our finding is in line with a literature review (Judd et al. Reference Judd, Jackson, Komiti, Murray, Hodgins and Fraser2002) showing that a variety of sociodemographic factors, such as unemployment, low socioeconomic status and poor social support, were stronger predictors of affective disorders than urbanization level, and that many previous studies failed to take these factors into account (Judd et al. Reference Judd, Jackson, Komiti, Murray, Hodgins and Fraser2002).

In our study, a high number of social security beneficiaries, high home values, high levels of road- rail- and air-traffic noise (measured as Lden in dB), more water in the neighbourhood and lower social cohesion were most consistently associated with depression and anxiety, independent of covariates and other neighbourhood factors. Social security beneficiaries are an indicator of long-term unemployment and socioeconomic deprivation of residents in the area, factors which have previously been associated with poor mental health and depression (Kim, Reference Kim2008; Annequin et al. Reference Annequin, Weill, Thomas and Chaix2015; Klijs et al. Reference Klijs, Kibele, Ellwardt, Zuidersma, Stolk and Wittek2016). Moreover, we showed traffic noise (Lden in dB) to be associated with depression and anxiety, independent of air pollution (measured as PM2.5 absorbance) and urbanization grade, whereas many previous studies failed to adjust for these characteristics (Tzivian et al. Reference Tzivian, Winkler, Dlugaj, Schikowski, Vossoughi and Fuks2015). Although causal processes remain speculative, it could be that traffic noise increases levels of stress (Arnsten & Goldman-Rakic, Reference Arnsten and Goldman-Rakic1998), subsequently affecting both physical and mental health (Recio et al. Reference Recio, Linares, Banegas and Díaz2016; Chen et al. Reference Chen, Rey, Kwong, Copes, Tu and Villeneuve2017). Our data did not allow one to distinguish between nocturnal and daily noise exposure, but likely these are highly correlated. High levels of traffic noise may possibly result in increased arousal (Griefahn et al. Reference Griefahn, Bröde, Marks and Basner2008) and concentration problems (Bodin et al. Reference Bodin, Björk, Ardö and Albin2015) but also in more sleep disturbances, which have consistently been associated with higher risks of depressive and anxiety disorders, also in the NESDA sample (van Mill et al. Reference van Mill, Hoogendijk, Vogelzangs, van Dyck and Penninx2010). According to the WHO guidelines, outdoor noise levels higher than 45 decibels could result in sleep disturbance effects (Berglund et al. Reference Berglund, Lindvall and Schwela2000). An alternative explanation is that traffic noise is a proxy of other forms of neighbourhood deprivation, such as annoyance (Ragettli et al. Reference Ragettli, Goudreau, Plante, Perron, Fournier and Smargiassi2015a) and overcrowding (O'Campo et al. Reference O'Campo, Salmon and Burke2009). Our finding that lower social cohesion is associated with an increased risk of depression is in line with previous research (Elliott et al. Reference Elliott, Gale, Parsons and Kuh2014), although most of these studies did not use objectively attained census data and did not account for other relevant neighbourhood correlates (Mair et al. Reference Mair, Diez Roux and Galea2008; Ruijsbroek et al. Reference Ruijsbroek, Droomers, Hardyns, Groenewegen and Stronks2016). The social neighbourhood could directly influence psychopathology, or indirectly through for example other person-level characteristics such as social relationships (Cohen & Wills, Reference Cohen and Wills1985; Helgeson, Reference Helgeson2003). Also, research suggests that social neighbourhood cohesion might mediate the relationship between income deprivation and mental health (Fone et al. Reference Fone, White, Farewell, Kelly, John and Lloyd2014).

Our findings that higher home values and more water in the area are associated with poor mental health are surprising in light of previous research showing beneficial health effects of high neighbourhood income (Klijs et al. Reference Klijs, Kibele, Ellwardt, Zuidersma, Stolk and Wittek2016) and living close to water (White et al. Reference White, Alcock, Wheeler and Depledge2013; Grellier et al. Reference Grellier, White, Albin, Bell, Elliott and Gascón2017). Regarding our result for water, one possible explanation is our assessment method of blue space, which is different from other previous studies that measured the percentage of blue space by calculating a ‘buffer’ around the centroid of the postal code area (e.g. de Vries et al. Reference de Vries, ten Have, van Dorsselaer, van Wezep, Hermans and de Graaf2016). Another explanation is that perhaps blue space is not a very good indicator of a ‘nature rich’ environment for the Dutch situation, because the percentage of water appeared to be highest in highly population areas. There are quite some cities with canals (e.g. Amsterdam) closely located to the Dutch coast, hence leading to positive effects of more water with higher risks of affective disorders. Consequently, the concept of ‘blue space’ in The Netherlands may not be exactly comparable to the concept in other countries, and therefore its findings should be interpreted carefully. Also, it is known that house prices in Dutch cities can be extremely high because of high demand and popularity. This might explain the association between high home values with higher risks of depression and anxiety (in multivariable analyses), and the positive correlations of high home values with air pollution and traffic noise (online Supplementary Table 1). High home values in the neighbourhood are likely to be a marker of higher socioeconomic status (r = 0.66; online Supplementary Table 1) and less social security beneficiaries (r = −0.64; online Supplementary Table 1), and this is found in the city center of Amsterdam. So, when interpreting these associations, it is important to realize that the Netherlands is amongst the most densely populated countries in the world. This may have introduced some specific characteristics. In other – less dense populated – areas, such associations could be different.

Although our study is cross-sectional of nature and therefore causality cannot be inferred, it would be adequate to provide some speculations regarding the pathways linking the neighbourhood to affective disorders. The social causation hypothesis asserts that poor neighbourhoods increase the risk of subsequent affective disorders (Mossakowski, Reference Mossakowski2014). The neighbourhood might be linked with mental health through behavioural and psychological stressors (e.g. social activity or overcrowding; Evans, Reference Evans2003) or biological regulation pathways (e.g. increased inflammation; Bird et al. Reference Bird, Seeman, Escarce, Basurto-Dávila, Finch and Dubowitz2010). Other proposed mediators between neighbourhoods and affective disorders could be unhealthy lifestyles (Saelens et al. Reference Saelens, Sallis, Black and Chen2003; Lorenc et al. Reference Lorenc, Clayton, Neary, Whitehead, Petticrew and Thomson2012), poor perception and satisfaction of the neighbourhood (Leslie & Cerin, Reference Leslie and Cerin2008), poor sense of control (Thoits, Reference Thoits2011), personality vulnerabilities such as neuroticism (Blair et al. Reference Blair, Ross, Gariepy and Schmitz2014) or the actual environmental exposure (i.e. some persons ‘use’ their neighbourhood more than others). However, a reverse link between environmental factors and affective disorders is also possible. The selection/drift hypothesis posits that poor mental health can inhibit socioeconomic attainment and lead people to drift into lower social class, resulting in (the maintenance of) less favourable environments (Pickett & Pearl, Reference Pickett and Pearl2001; Mossakowski, Reference Mossakowski2014). So, in sum, a bidirectional link is not unlikely. Deprived neighbourhoods may affect the resident's mental health and, reversely, the resident's health status may affect the state or selection of the neighbourhood. Future studies should shed light on potential causal and mediating pathways between neighbourhoods and affective disorders, and on how such a link could be intervened on in order to stimulate global mental health.

Strong aspects of this study are the large sample size, the objective measurement of neighbourhood factors using census data, the inclusion of both validated psychiatric interviews (Robins et al. Reference Robins, Wing, Wittchen, Helzer, Babor and Burke1988) and symptoms scales (van Zuuren, Reference van Zuuren1988; Beck & Steer, Reference Beck and Steer1990; Rush et al. Reference Rush, Gullion, Basco, Jarrett and Trivedi1996) for the assessment of depression and anxiety, and the adjustment for other neighbourhood-level and person-level confounders. Next to these important strengths, there are also several methodological considerations to our epidemiological study. First, our study is limited by the cross-sectional nature. However, we feel that generating a clear overview of the neighbourhood factors cross-sectionally associated with mental health is a most valuable first step in the neighbourhood and mental health research. It would be very useful to follow our study up with longitudinal approaches in which neighbourhood characteristics are for example linked with the development of mental health problems over time. Thus, future prospective studies and population-level experimental studies might shed more light on whether deprived neighbourhoods are contributors, correlates or consequences of affective disorders. These future studies are highly recommended to take a broad scope of the neighbourhood and select more than one factor and to place studies within a plausible theoretical framework (Schofield & Das-Munshi, Reference Schofield and Das-Munshi2017). Second, it could be that the use of more sophisticated Global Positioning System (GPS) tracking measures for the assessment of residential living space conceals somewhat different findings than the use of linkage through postal code areas, e.g. because environmental exposure at home and work could be combined (Ragettli et al. Reference Ragettli, Phuleria, Tsai, Schindler, de Nazelle and Ducret-Stich2015b). However, GPS methods may represent a step backward in terms of biases related to selective daily mobility and causal inferences (Chaix et al. Reference Chaix, Méline, Duncan, Merrien, Karusisi and Perchoux2013). Moreover, GPS methods are more costly, whereas the use of existing registration systems is easier, highly feasible and in line with various previous cohort studies (Mair et al. Reference Mair, Diez Roux and Galea2008; Zijlema et al. Reference Zijlema, Klijs, Stolk and Rosmalen2015; de Vries et al. Reference de Vries, ten Have, van Dorsselaer, van Wezep, Hermans and de Graaf2016; Klijs et al. Reference Klijs, Kibele, Ellwardt, Zuidersma, Stolk and Wittek2016; Chen et al. Reference Chen, Rey, Kwong, Copes, Tu and Villeneuve2017). Third, we aggregated neighbourhood data to average values for four-digit postal code areas as this was the most precise information available of our respondents, and this method is in accordance with many previous Dutch studies (Wittebrood, Reference Wittebrood2000; Reference Wittebrood2004; Van Wilsem, Reference Van Wilsem2003; Bernasco & Nieuwbeerta, Reference Bernasco and Nieuwbeerta2005; Nieuwbeerta et al. Reference Nieuwbeerta, McCall, Elffers and Wittebrood2008). Aggregation results in less variation in values for air pollution (as indicated by PM2.5 absorbance) and traffic noise (as indicated by Ldn in dB) that were estimated at a lower geographical level, i.e. residential addresses. However, our geodataset including all 4766 postal code areas of the Netherlands allowed us to compare distributions for air pollution and traffic noise data between four-digit and six-digit postal code areas and these appeared comparable [air pollution, mean (s.d.) = 1.18(0.22) v. 1.27(0.25); traffic noise, median (IQR) = 54 (51–56) v. 54 (51–59)]. In combination with our significant findings for noise with depression and anxiety and for pollution with anxiety, this seems to suggest that there is still enough variation in our variables to detect group differences. Fourth, as previously mentioned, the Netherlands is a very densely populated country (±483 residents/km2 in 2006; CBS, 2006e); hence, future research could test whether our results can be generalized to other countries.

Nonetheless, to our knowledge, this is the first large-scale study that assessed multiple objectively attained socioeconomic, physical and social neighbourhood factors in relation to both the presence and severity of depression and anxiety disorders, while adjusting for individual indicators of socioeconomic status.

It suggests that socioeconomic, physical and social neighbourhood characteristics are associated with affective disorders, implying the relevance of studying the psychological impact of neighbourhood quality.

Supplementary material

The supplementary material for this article can be found at https://doi.org/10.1017/S0033291718000612.

Acknowledgements

The infrastructure for the NESDA study (www.nesda.nl) is funded through the Geestkracht program of the Netherlands Organisation for Health Research and Development (Zon-Mw, grant number 10-000-1002) and financial contributions by participating universities and mental health care organizations (VU University Medical Center, GGZ inGeest, Leiden University Medical Center, Leiden University, GGZ Rivierduinen, University Medical Center Groningen, University of Groningen, Lentis, GGZ Friesland, GGZ Drenthe, Rob Giel Onderzoekscentrum). Geo-data were collected as part of the Geoscience and Health Cohort Consortium (GECCO), which was financially supported by the EMGO + Institute for Health and Care Research of VU University Amsterdam and VU University Medical Center Amsterdam, the Netherlands. We would like to thank the Institute for Risk Assessment Sciences of the Utrecht University, the coordinating centre of the European Study of Cohorts for Air Pollution Effects (ESCAPE), for providing data on air pollution. Furthermore, we would like to thank the Netherlands Environmental Assessment Agency for providing data on road-, rail- and air traffic noise. We gratefully acknowledge all organizations that collaborate with the six cohort studies in GECCO. We would also like to thank all respondents who participated in the six cohort studies that are involved in GECCO.

Declaration of interest

None.

Ethical standards

The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008.

References

Annequin, M, Weill, A, Thomas, F and Chaix, B (2015) Environmental and individual characteristics associated with depressive disorders and mental health care use. Annals of Epidemiology 25(8), 605612. doi: 10.1016/j.annepidem.2015.02.002.Google Scholar
Arcaya, MC, Tucker-Seeley, RD, Kim, R, Schnake-Mahl, A, So, M and Subramanian, SV (2016) Research on neighborhood effects on health in the United States: a systematic review of study characteristics. Social Science and Medicine 168, 1629. doi: 10.1016/j.socscimed.2016.08.047.Google Scholar
ArcGIS (2017) ArcGIS maps. Available at: https://www.arcgis.com/features/index.html (Accessed 16 February 2018).Google Scholar
Arnsten, AFT and Goldman-Rakic, PS (1998) Noise stress impairs prefrontal cortical cognitive function in monkeys: evidence for a hyperdopaminergic mechanism. Archives of General Psychiatry 55(4), 362368. doi: 10.1001/archpsyc.55.4.362.Google Scholar
Ball, K, Jeffery, RW, Crawford, DA, Roberts, RJ, Salmon, J and Timperio, AF (2008) Mismatch between perceived and objective measures of physical activity environments. Preventive Medicine 47(3), 294298. doi: 10.1016/j.ypmed.2008.05.001.Google Scholar
Beck, AT, Epstein, N, Brown, G and Steer, RA (1988) An inventory for measuring clinical anxiety: psychometric properties. Journal of Consulting and Clinical Psychology 56(6), 893897.Google Scholar
Beck, AT and Steer, RA (1990) Manual for the Beck Anxiety Inventory. San Antonio, TX: Psychological Corporation.Google Scholar
Beelen, R, Hoek, G, Vienneau, D, Eeftens, M, Dimakopoulou, K, Pedeli, X et al. (2013) Development of NO2 and NOx land use regression models for estimating air pollution exposure in 36 study areas in Europe – The ESCAPE project. Atmospheric Environment 72, 1023.Google Scholar
Berglund, B, Lindvall, T and Schwela, DH (2000) Guidelines for Community Noise. Geneva, Switzerland: World Health Organisation. whqlibdoc.who.int/hq/1999/a68672.pdf (Accessed 16 February 2018).Google Scholar
Bernasco, W and Nieuwbeerta, P (2005) How do residential burglars select target areas? A new approach to the analysis of criminal location choice. British Journal of Criminology 45, 296315. doi: 10.1093/bjc/azh070.Google Scholar
Bird, CE, Seeman, T, Escarce, JJ, Basurto-Dávila, R, Finch, BK, Dubowitz, T et al. (2010) Neighbourhood socioeconomic status and biological ‘wear and tear’ in a nationally representative sample of US adults. Journal of Epidemiology and Community Health 64(10), 860865. doi: 10.1136/jech.2008.084814.Google Scholar
Blair, A, Ross, NA, Gariepy, G and Schmitz, N (2014) How do neighborhoods affect depression outcomes? A realist review and a call for the examination of causal pathways. Social Psychiatry and Psychiatric Epidemiology 49(6), 873887. doi: 10.1007/s00127-013-0810-z.Google Scholar
Bodin, T, Björk, J, Ardö, J and Albin, M (2015) Annoyance, sleep and concentration problems due to combined traffic noise and the benefit of quiet side. International Journal of Environmental Research and Public Health 12(2), 16121628. doi: 10.3390/ijerph120201612.Google Scholar
Bowerman, BL and O'Connell, RT (1990) Linear Statistical Models: An Applied Approach, 2nd edn., Belmont, CA: Duxbury.Google Scholar
Calderón-Garcidueñas, L, Solt, AC, Henríquez-Roldán, C, Torres-Jardón, R, Nuse, B, Herritt, L et al. (2008) Long-term air pollution exposure is associated with neuroinflammation, an altered innate immune response, disruption of the blood-brain barrier, ultrafine particulate deposition, and accumulation of amyloid beta-42 and alpha-synuclein in children and young adults. Toxicologic Pathology 36(2), 289310. doi: 10.1177/0192623307313011.Google Scholar
CBS (2005) Statistics Netherlands. Municipality information 2006. Available at: https://www.cbs.nl/nl-nl/nieuws/2006/01/458-gemeenten-in-2006 (Accessed 16 February 2018).Google Scholar
CBS (2006 a) Statistics Netherlands / The Netherlands’ Cadastre, Land Registry, and Mapping Agency. Gebruik gegeneraliseerde geometrie Wijk- en buurtkaart 2006 [in Dutch]. Emmen: Statistics Netherlands / The Netherlands’ Cadastre, Land Registry, and Mapping Agency. Available at: https://www.cbs.nl/nl-nl/dossier/nederland-regionaal/geografische%20data/wijk-en-buurtkaart-2006 (Accessed 16 February 2018).Google Scholar
CBS (2006 b) Statistics Netherlands. Bestand Bodemgebruik [in English: Land use file] - Productbeschrijving. Available at: https://www.cbs.nl/NR/rdonlyres/5B353A4E-6B56-4756-A19C-4BF6AA520C65/0/BestandBodemgebruikProductbeschrijving.pdf (Accessed 16 February 2018).Google Scholar
CBS (2006 c) Statistics Netherlands. Buurtkaart met cijfers 2006; Gebruik gegeneraliseerde geometrie Wijk- en buurtkaart 2006 [in Dutch]. Voorburg/Heerlen: Statistics Netherlands. Available at: https://www.cbs.nl/-/media/imported/documents/2008/01/2006-b68-pub.pdf (Accessed 16 February 2018).Google Scholar
CBS (2006 d) Statistics Netherlands. Kerncijfers 2006. Available at: http://statline.cbs.nl/Statweb/publication/?DM=SLNL&PA=37296ned&D1=0-51&D2=56&VW=T (Accessed 16 February 2018).Google Scholar
CBS (2006 e) Statistics Netherlands. Kerncijfers 2006 (population density). Available at: http://statline.cbs.nl/Statweb/publication/?DM=SLNL&PA=37296ned&D1=0,68&D2=56&HDR=G1&STB=T&VW=T (Accessed 16 February 2018).Google Scholar
CBS (2006 f) Statistics Netherlands. Postal code information [in Dutch]. Available at: http://statline.cbs.nl/Statweb/publication/?DM=SLNL&PA=71273ned&D1=0-62,66-72&D2=0-70,72-8041504535&VW=T (Accessed 16 February 2018).Google Scholar
CBS (2006 g) Statistics Netherlands. Wijk en Buurtkaart 2006. Available at: https://www.cbs.nl/nl-nl/dossier/nederland-regionaal/geografischedata/wijk-en-buurtkaart-2006 (Accessed 16 February 2018).Google Scholar
CBS (2014) Statistics Netherlands. Bevolking en huishoudens; viercijferige postcode, 1 januari 2014 [In Dutch]. [In English: Population and households; 4-digit postal code, January 1, 2014]. Available at: https://www.cbs.nl/nl-nl/maatwerk/2014/52/bevolking-en-huishoudens-viercijferige-postcode-1-januari-2014 (Date last update: 16 December 2015, Date last accessed 16 February 2018).Google Scholar
CBS (2017 a) Statistics Netherlands. Available at: http://statline.cbs.nl/Statweb/ (Accessed 16 February 2018).Google Scholar
Cesaroni, G, Forastiere, F, Stafoggia, M, Andersen, ZJ, Badaloni, C, Beelen, R et al. (2014) Long term exposure to ambient air pollution and incidence of acute coronary events: prospective cohort study and meta-analysis in 11 european cohorts from the ESCAPE project. BMJ (Clinical Research Ed.) 348, f7412. doi: 10.1136/bmj.f7412.Google Scholar
Chaix, B, Méline, J, Duncan, S, Merrien, C, Karusisi, N, Perchoux, C et al. (2013) GPS tracking in neighborhood and health studies: a step forward for environmental exposure assessment, a step backward for causal inference? Health & Place 21, 4651. doi: 10.1016/j.healthplace.2013.01.003.Google Scholar
Chen, H, Rey, J, Kwong, C, Copes, R, Tu, K, Villeneuve, PJ et al. (2017) Living near major roads and the incidence of dementia, Parkinson's disease, and multiple sclerosis: a population-based cohort study. Lancet 389(10070), 718726. doi: 10.1016/S0140-6736(16)32399-6.Google Scholar
Clarke, P (2008) When can group level clustering be ignored? Multilevel models versus single-level models with sparse data. Journal of Epidemiology and Community Health 62(8), 752758. doi: 10.1136/jech.2007.060798.Google Scholar
Cohen, S and Wills, TA (1985) Stress, social support, and the buffering hypothesis. Psychological Bulletin 98(2), 310357. doi: 10.1037/0033-2909.98.2.310.Google Scholar
de Vries, S, ten Have, M, van Dorsselaer, S, van Wezep, M, Hermans, T and de Graaf, R (2016) Local availability of green and blue space and prevalence of common mental disorders in the Netherlands. British Journal of Psychiatry Open 2(6), 366372. doi: 10.1192/bjpo.bp.115.002469.Google Scholar
den Dulk, CJ, van de Stadt, H and Vliegen, JM (1992) A new measure for degree of urbanization: the address density of the surrounding area [article in Dutch]. Maandstatistiek van de bevolking (the Hague, Netherlands: 1982) 40(7), 1427.Google Scholar
Diez Roux, AV (2007) Neighborhoods and health: where are we and were do we go from here? Revue d’Épidémiologie et de Santé Publique 55(1), 1321. doi: 10.1016/j.respe.2006.12.003.Google Scholar
Diez Roux, AV and Mair, C (2010) Neighborhoods and health. Annals of the New York Academy of Sciences 1186(1), 125145. doi: 10.1111/j.1749-6632.2009.05333.x.Google Scholar
Dulin-Keita, A, Casazza, K, Fernandez, JR, Goran, MI and Gower, B (2012) Do neighbourhoods matter? Neighbourhood disorder and long-term trends in serum cortisol levels. Journal of Epidemiology and Community Health 66(1), 2429. doi: 10.1136/jech.2009.092676.Google Scholar
Eeftens, M, Beelen, R, de Hoogh, K, Bellander, T, Cesaroni, G, Cirach, M et al. (2012) Development of land Use regression models for PM 2.5, PM 2.5 absorbance, PM 10 and PM coarse in 20 European study areas; results of the ESCAPE project. Environmental Science and Technology 46(20), 1119511205. doi: 10.1021/es301948k.Google Scholar
Eeftens, M, Beelen, R, Fischer, P, Brunekreef, B, Meliefste, K and Hoek, G (2011) Stability of measured and modelled spatial contrasts in NO2 over time. Occupational and Environmental Medicine 68(10), 765770. doi: 10.1136/oem.2010.061135.Google Scholar
Elliott, J, Gale, CR, Parsons, S and Kuh, D (2014) Neighbourhood cohesion and mental wellbeing among older adults: a mixed methods approach. Social Science and Medicine 107, 4451. doi: 10.1016/j.socscimed.2014.02.027.Google Scholar
Evans, GW (2003) The built environment and mental health. Journal of Urban Health: Bulletin of the New York Academy of Medicine 80(4), 536555.Google Scholar
Field, A (2013) Discovering Statistics Using IBM SPSS Statistics, 4th edn., London, UK: Sage Publications LTD.Google Scholar
Fone, D, White, J, Farewell, D, Kelly, M, John, G, Lloyd, K et al. (2014) Effect of neighbourhood deprivation and social cohesion on mental health inequality: a multilevel population-based longitudinal study. Psychological Medicine 44(11), 24492460. doi: 10.1017/S0033291713003255.Google Scholar
Gariepy, G, Kaufman, JS, Blair, A, Kestens, Y and Schmitz, N (2014) Place and health in diabetes: the neighbourhood environment and risk of depression in adults with type 2 diabetes. Diabetic Medicine 32(7), 944950. doi: 10.1111/dme.12650.Google Scholar
Gebel, K, Bauman, AE, Sugiyama, T and Owen, N (2011) Mismatch between perceived and objectively assessed neighborhood walkability attributes: prospective relationships with walking and weight gain. Health and Place 17(2), 519524. doi: 10.1016/j.healthplace.2010.12.008.Google Scholar
Gehring, U, Gruzieva, O, Agius, RM, Beelen, R, Custovic, A, Cyrys, J et al. (2013) Air pollution exposure and lung function in children: the ESCAPE project. Environmental Health Perspectives 121(11–12), 13571364. doi: 10.1289/ehp.1306770.Google Scholar
Gong, Y, Palmer, S, Gallacher, J, Marsden, T and Fone, D (2016) A systematic review of the relationship between objective measurements of the urban environment and psychological distress. Environment International 96, 4857. doi: 10.1016/j.envint.2016.08.019.Google Scholar
Grellier, J, White, MP, Albin, M, Bell, S, Elliott, LR, Gascón, M et al. (2017) Bluehealth: a study programme protocol for mapping and quantifying the potential benefits to public health and well-being from Europe's blue spaces. BMJ Open 7(6), e016188. doi: 10.1136/bmjopen-2017-016188.Google Scholar
Griefahn, B, Bröde, P, Marks, A and Basner, M (2008) Autonomic arousals related to traffic noise during sleep. Sleep 31(4), 569577.Google Scholar
Gulliver, J, Morris, C, Lee, K, Vienneau, D, Briggs, D and Hansell, A (2011) Land use regression modeling to estimate historic (1962–1991) concentrations of black smoke and sulfur dioxide for Great Britain. Environmental Science and Technology 45(8), 35263532. doi: 10.1021/es103821y.Google Scholar
Hackman, DA, Betancourt, LM, Brodsky, NL, Hurt, H and Farah, MJ (2012) Neighborhood disadvantage and adolescent stress reactivity. Frontiers in Human Neuroscience 6, 277. doi: 10.3389/fnhum.2012.00277.Google Scholar
Helgeson, VS (2003) Social support and quality of life. Quality of Life Research 12(suppl. 1), 2531.Google Scholar
IDS (2017) The Inventory of depressive symptomatology. Psychometric properties. Available at: www.ids-qids.org (Accessed 16 February 2018).Google Scholar
Judd, FK, Jackson, HJ, Komiti, A, Murray, G, Hodgins, G and Fraser, C (2002) High prevalence disorders in urban and rural communities. Australian and New Zealand Journal of Psychiatry 36(1), 104113. doi: 10.1046/j.1440-1614.2002.00986.x.Google Scholar
Kim, D (2008) Blues from the neighborhood? Neighborhood characteristics and depression. Epidemiologic Reviews 30, 101117. doi: 10.1093/epirev/mxn009.Google Scholar
Klijs, B, Kibele, EUB, Ellwardt, L, Zuidersma, M, Stolk, RP, Wittek, RPM et al. (2016) Neighborhood income and major depressive disorder in a large Dutch population: results from the LifeLines Cohort study. BMC Public Health 16(1), 773. doi: 10.1186/s12889-016-3332-2.Google Scholar
Knol, F (2012) Statusontwikkeling van wijken in Nederland 1998–2010. The Hague, The Netherlands: Sociaal en Cultureel Planbureau. Available at: https://www.scp.nl/Publicaties/Alle_publicaties/Publicaties_2012/Statusontwikkeling_van_wijken_in_Nederland_1998_2010 (Accessed 16 February 2018).Google Scholar
Krabbendam, L and van Os, J (2005) Schizophrenia and urbanicity: a major environmental influence—conditional on genetic risk. Schizophrenia Bulletin 31(4), 795799. doi: 10.1093/schbul/sbi060.Google Scholar
Leefbaarometer (2015) Rapportage instrumentontwikkeling [In English: report on livability score]. Available at: https://www.rijksoverheid.nl/documenten/rapporten/2008/05/01/rapportage-instrumentontwikkeling (Accessed 16 February 2018).Google Scholar
Leidelmeijer, K, Marlet, G, Van Iersel, J, Van Woerkens, C and van der Reijden, H. (2008) De Leefbaarometer. Leefbaarheid in Nederlandse wijken en buurten; gemeten en vergeleken. Rapportage Instrumentontwikkeling [In Dutch]. [In English: Leefbaarometer (livability score). Liveability in Dutch residential districts and neighborhoods; measured and compared. Report on development of instrument]. Available at: https://www.leefbaarometer.nl/home.php (Accessed 16 February 2018).Google Scholar
Leidelmeijer, K and van Kamp, I (2003) Kwaliteit van de leefomgeving en leefbaarheid. Naar een begrippenkader en conceptuele inkadering [In Dutch]. [In English: Quality of the living environment and liveability. To a conceptual framework and conceptual framing]. Bilthoven: The National Institute of Public Health and the Environment in the Netherlands (RIVM) and RIGO Research and Consultancy BV (Amsterdam). Available at: https://www.rivm.nl/Documenten_en_publicaties/Wetenschappelijk/Rapporten/2004/mei/Kwaliteit_van_de_leefomgeving_en_leefbaarheid_Naar_een_begrippenkader_en_conceptuele_inkadering (Accessed 16 February 2018).Google Scholar
Leslie, E and Cerin, E (2008) Are perceptions of the local environment related to neighbourhood satisfaction and mental health in adults? Preventive Medicine 47(3), 273278. doi: 10.1016/j.ypmed.2008.01.014.Google Scholar
Linden, W and Vodermaier, A (2012) Mismatch of desired versus perceived social support and associated levels of anxiety and depression in newly diagnosed cancer patients. Supportive Care in Cancer 20(7), 14491456. doi: 10.1007/s00520-011-1228-3.Google Scholar
Livability scores (2015) The Netherlands Ministry of the Interior and Kingdom Relations. Data leefbaarheidscores 1998–2012, per postcode (pc4) [In Dutch]. [In English: Data liveability scores 1998–2012, per 4-digit postal code]. Available at: https://data.overheid.nl/data/dataset/scores-leefbaarometer-1998-2012 (Date last update: 30 March 2015, Date last accessed: 16 February 2018).Google Scholar
Lorenc, T, Clayton, S, Neary, D, Whitehead, M, Petticrew, M, Thomson, H et al. (2012) Crime, fear of crime, environment, and mental health and wellbeing: mapping review of theories and causal pathways. Health and Place 18(4), 757765. doi: 10.1016/j.healthplace.2012.04.001.Google Scholar
Mair, C, Diez Roux, AV and Galea, S (2008) Are neighborhood characteristics associated with depressive symptoms? A review of evidence. Journal of Epidemiology and Community Health 62(11), 940946. doi: 10.1136/jech.2007.066605.Google Scholar
Marks, IM and Mathews, AM (1979) Brief standard self-rating for phobic patients. Behaviour Research and Therapy 17(3), 263267. doi: 10.1016/0005-7967(79)90041-X.Google Scholar
McNeish, DM and Stapleton, LM (2014) The effect of small sample size on two level model estimates: a review and illustration. Educational Psychology Review 28(2), 295314. doi: 10.1007/s10648-014-9287-x.Google Scholar
Merlo, J, Chaix, B, Ohlsson, H, Beckman, A, Johnell, K, Hjerpe, P et al. (2006) A brief conceptual tutorial of multilevel analysis in social epidemiology: using measures of clustering in multilevel logistic regression to investigate contextual phenomena. Journal of Epidemiology and Community Health 60(4), 290297. doi: 10.1136/jech.2004.029454.Google Scholar
Miles, R, Coutts, C and Mohamadi, A (2012) Neighborhood urban form, social environment, and depression. Journal of Urban Health 89(1), 118. doi: 10.1007/s11524-011-9621-2.Google Scholar
Mossakowski, KN (2014) Social causation and social selection. In The Wiley Blackwell Encyclopedia of Health, Illness, Behavior, and Society. New York: John Wiley, pp. 21542160. doi: 10.1002/9781118410868.wbehibs262.Google Scholar
Myers, R (1990) Classical and Modern Regression with Applications, 2nd edn., Boston, MA: Duxbury.Google Scholar
Nieuwbeerta, P, McCall, PL, Elffers, H and Wittebrood, K (2008) Neighborhood characteristics and individual homicide risks. Effects of social cohesion, confidence in the police, and socioeconomic disadvantage. Homicide Studies 12(1), 90116.Google Scholar
O'Campo, P, Salmon, C and Burke, J (2009) Neighbourhoods and mental well-being: what are the pathways? Health & Place 15(1), 5668. doi: 10.1016/j.healthplace.2008.02.004.Google Scholar
PBL (2017) The Netherlands Environmental Assessment Agency. Geluidbelasting weg rail en luchtverkeer in Nederland 2000–2008 [In Dutch]. [In English: Road-, rail- and air-traffic noise in the Netherlands 2000–2008]. Available at: http://geoservice.pbl.nl/arcgis/rest/services/projecten/Geluidbelasting_weg_rail_en_luchtverkeer_in_Nederland_20002008/3MapServer (Accessed 16 February 2018).Google Scholar
Peen, J, Schoevers, RA, Beekman, AT and Dekker, J (2010) The current status of urban-rural differences in psychiatric disorders. Acta Psychiatrica Scandinavica 121, 8493. doi: 10.1111/j.1600-0447.2009.01438.x.Google Scholar
Penninx, BWJH, Beekman, ATF, Smit, JH, Zitman, FG, Nolen, WA, Spinhoven, P et al. (2008) The Netherlands study of depression and anxiety (NESDA): rationale, objectives and methods. International Journal of Methods in Psychiatric Research 17(3), 121140. doi: 10.1002/mpr.256.Google Scholar
Pickett, KE and Pearl, M (2001) Multilevel analyses of neighbourhood socioeconomic context and health outcomes: a critical review. Journal of Epidemiology and Community Health 55(2), 111122.Google Scholar
Ragettli, MS, Goudreau, S, Plante, C, Perron, S, Fournier, M and Smargiassi, A (2015 a) Annoyance from road traffic, trains, airplanes and from total environmental noise levels. International Journal of Environmental Research and Public Health 13(1), E90. doi: 10.3390/ijerph13010090.Google Scholar
Ragettli, MS, Phuleria, HC, Tsai, M-Y, Schindler, C, de Nazelle, A, Ducret-Stich, RE et al. (2015 b) The relevance of commuter and work/school exposure in an epidemiological study on traffic-related air pollution. Journal of Exposure Science and Environmental Epidemiology 25(5), 474481. doi: 10.1038/ jes.2014.83.Google Scholar
Recio, A, Linares, C, Banegas, JR and Díaz, J (2016) Road traffic noise effects on cardiovascular, respiratory, and metabolic health: an integrative model of biological mechanisms. Environmental Research 146, 359370. doi: 10.1016/j.envres.2015.12.036.Google Scholar
Robins, LN, Wing, J, Wittchen, HU, Helzer, JE, Babor, TF, Burke, J et al. (1988) The composite international diagnostic interview. An epidemiologic instrument suitable for use in conjunction with different diagnostic systems and in different cultures. Archives of General Psychiatry 45(12), 10691077.Google Scholar
R software (2017) Package ‘lme4’ in R. Available at: https://cran.r-project.org/web/packages/lme4/lme4.pdf (Accessed 16 February 2018).Google Scholar
Ruijsbroek, A, Droomers, M, Hardyns, W, Groenewegen, PP and Stronks, K (2016) The interplay between neighbourhood characteristics: the health impact of changes in social cohesion, disorder and unsafety feelings. Health & Place 39, 18. doi: 10.1016/j.healthplace.2016.02.001.Google Scholar
Ruijsbroek, A, Mohnen, SM, Droomers, M, Kruize, H, Gidlow, C, Gražulevičiene, R et al. (2017) Neighbourhood green space, social environment and mental health: an examination in four European cities. International Journal of Public Health 62(6), 657667. doi: 10.1007/s00038-017-0963-8.Google Scholar
Rush, AJ, Gullion, CM, Basco, MR, Jarrett, RB and Trivedi, MH (1996) The inventory of depressive symptomatology (IDS): psychometric properties. Psychological Medicine 26(3), 477486.Google Scholar
Saelens, BE, Sallis, JF, Black, JB and Chen, D (2003) Neighborhood-based differences in physical activity: an environment scale evaluation. American Journal of Public Health 93(9), 15521558. doi: 10.2105/AJPH.93.9.1552.Google Scholar
Schofield, P, Das-Munshi, J (2017) New directions in neighbourhood research—a commentary on Lovasi et al. (2016): cause and context: place-based approaches to investigate how environments affect mental health. Social Psychiatry and Psychiatric Epidemiology 52(2), 135137. doi: 10.1007/s00127-016-1332-2.Google Scholar
SCP (2017) The Netherlands Institute of Social Research. Sociaal en Cultureel Planbureau Statusscores [In Dutch]. Available at: http://www.scp.nl/Onderzoek/Lopend_onderzoek/A_Z_alle_lopende_onderzoeken/Statusscores (Accessed 16 February 2018).Google Scholar
Srivastava, K (2009) Urbanization and mental health. Indian Journal of Psychiatry 18(2), 7576. doi: 10.4103/0972-6748.64028.Google Scholar
Stronks, K, Kulu-Glasgow, I and Agyemang, C (2009) The utility of ‘country of birth’ for the classification of ethnic groups in health research: the Dutch experience. Ethnicity and Health 14(3), 255269. doi: 10.1080/13557850802509206.Google Scholar
Thoits, PA (2011) Mechanisms linking social ties and support to physical and mental health. Journal of Health and Social Behavior 52(2), 145161. doi: 10.1177/0022146510395592.Google Scholar
Timmermans, EJ, Lakerveld, J, Beulens, JWJ, Boomsma, DI, Kram, SE, Oosterman, M et al. (2018) Cohort Profile: The Geoscience and Health Cohort Consortium (GECCO) in the Netherlands. BMJ Open.Google Scholar
Truong, KD and Ma, S (2006) A systematic review of relations between neighborhoods and mental health. Journal of Mental Health Policy and Economics 9(3), 137154.Google Scholar
Tzivian, L, Winkler, A, Dlugaj, M, Schikowski, T, Vossoughi, M, Fuks, K et al. (2015) Effect of long-term outdoor air pollution and noise on cognitive and psychological functions in adults. International Journal of Hygiene and Environmental Health 218(1), 111. doi: 10.1016/j.ijheh.2014.08.002.Google Scholar
van Deurzen, I, Rod, NH, Christensen, U, Hansen, AM, Lund, R and Dich, N (2016) Neighborhood perceptions and allostatic load: evidence from Denmark. Health Place 40, 18. doi: 10.1016/j.healthplace.2016.04.010.Google Scholar
van Mill, JG, Hoogendijk, WJ, Vogelzangs, N, van Dyck, R and Penninx, BW (2010) Insomnia and sleep duration in a large cohort of patients with major depressive disorder and anxiety disorders. Journal of Clinical Psychiatry 71(3), 239246. doi: 10.4088/JCP.09m05218gry.Google Scholar
Van Wilsem, J (2003) Crime and Context: The Impact of Individual, Neighborhood, City and Country Characteristics on Victimization. Nijmegen: Katholieke Universiteit Nijmegen. http://repository.ubn.ru.nl/bitstream/handle/2066/63843/63843.pdf (Accessed 16 February 2018).Google Scholar
van Zuuren, FJ (1988) The fear questionnaire. Some data on validity, reliability and layout. The British Journal of Psychiatry 153, 659662.Google Scholar
White, MP, Alcock, I, Wheeler, BW and Depledge, MH (2013) Coastal proximity, health and well-being: results from a longitudinal panel survey. Health Place 23, 97103.Google Scholar
Wittebrood, K (2000) Buurten en geweldscriminaliteit: een multilevel analyse [In English: neighbourhoods and violent crime: a multilevel analysis]. Mens en Maatschappij 75, 92109.Google Scholar
Wittebrood, K (2004) Van delictmelding tot officiële aangifte: sprake van sociale ongelijkheid? [In English: from notifying to an official report: a matter of social inequality?]. Tijdschrift voor Criminologie 46, 5671.Google Scholar
WOZ (2017) Statistiek Waardering Onroerende Zaken [In Dutch]. Available at: https://www.cbs.nl/nl-nl/onze-diensten/methoden/onderzoeksomschrijvingen/korte-onderzoeksbeschrijvingen/waardering-onroerende-zaken--woz-- (Accessed 16 February 2018).Google Scholar
Zijlema, WL, Cai, Y, Doiron, D, Mbatchou, S, Fortier, I, Gulliver, J et al. (2016 a) Road traffic noise, blood pressure and heart rate: pooled analyses of harmonized data from 88336 participants. Environmental Research. Elsevier 151, 804813. doi: 10.1016/j.envres.2016.09.014.Google Scholar
Zijlema, WL, Klijs, B, Stolk, RP and Rosmalen, JGM (2015) (Un)healthy in the city: respiratory, cardiometabolic and mental health associated with urbanity. PLoS ONE 10(12), e0143910. doi: 10.1371/journal.pone.0143910.Google Scholar
Zijlema, WL, Wolf, K, Emeny, R, Ladwig, KH, Peters, A, Kongsgård, H et al. (2016 b) The association of air pollution and depressed mood in 70928 individuals from four European cohorts. International Journal of Hygiene and Environmental Health 219(2), 212219. doi: 10.1016/j.ijheh.2015.11.006.Google Scholar
Zuur, AF, Ieno, EN and Elphick, CS (2010) A protocol for data exploration to avoid common statistical problems. Methods in Ecology and Evolution 1(1), 314. doi: 10.1111/j.2041-210X.2009.00001.x.Google Scholar
Figure 0

Table 1. Baseline characteristicsa

Figure 1

Table 2. Associations between neighbourhood characteristics and current diagnoses of depressive and anxiety disorders

Figure 2

Table 3. Associations between neighbourhood characteristics and the severity of depressive, anxiety and fear symptoms

Supplementary material: File

Generaal et al. supplementary material

Generaal et al. supplementary material 1

Download Generaal et al. supplementary material(File)
File 35.1 KB