Table 1 Summary of key features of 45 reviewed studies
Author, yearPrimary research questions addressedStudy sampleNeighbourhood definition usedNeighbourhood features investigated*Depression measure investigatedStudy designKey analytical techniqueResults (support for neighbourhood effects on depression?)
Anehensel et al, 2007 48Are depressive symptoms in older individuals associated with low SES, high concentrations of ethnic minorities, low residential stability and low proportion of residents aged 65+ in urban neighbourhoods?3442 individuals aged >70 years living in urban areas in the USA, from AHEADCensus tractsSocioeconomic disadvantage, affluence, racial/ethnic composition (proportion of African-American residents, proportion of Hispanic residents), residential stability, proportion of persons >65 years (all from census)Eight-item version of CES-DCross-sectionalMultilevel analysisY and N: depressive symptoms are associated with residential stability (β (SE) 0.72 (0.27)) after controlling for individual-level characteristics, but not with any of the other neighbourhood characteristics
Aneshensel et al, 199618Are adolescents’ experiences of their neighbourhood as threatening or cohesive associated with their mental health?877 adolescents aged 12–17 from Los Angeles CountyCensus tracts, grouped using cluster analysis into eight clustersParticipant-reported subjective neighbourhood measures (ambient hazards, social cohesion), neighbourhood stabilityCDICross-sectionalSingle-level linear regressionY: Adolescents’ perceptions of ambient hazards (β (SE) 0.022 (0.008)) and negative social cohesion (β (SE) −0.122(0.032)) are both associated with depressive symptoms.
Berke et al, 200736Is there an association between neighbourhood walkability and depressive symptoms in older adults?740 adults aged 65 + from King County, WashingtonCircular buffers of radius 100, 500, and 1000 m around each subject’s homeNeighbourhood walkability, measured by the WBC. Greater walkability was assessed, on a scale of 0 to 100, according to the probability of meeting the CDC guidelines of 150+ minutes per week of physical activity versus noneCES-DCross-sectionalSingle-level logistic regressionY: for men, there was an association between neighbourhood walkability and depressive symptoms (OR for the interquartile range of walkability score = 0.31–0.33 for the buffer radii, p = 0.02) after adjustment for key individual-level factors. This association was not significant in women (p>0.68).
Caughy et al, 2003 39Is there an association between how well a parent knows their neighbours and their child’s internalising behaviours? Does this association differ by neighbourhood SES context?200 African-American families with a child aged 3–4.5 in BaltimoreCensus block groupsNeighbourhood poverty (from census), general sense of community, how well one knows one’s neighbours (participant reported)CBCL internalising problems scoreCross-sectionalSingle-level linear regressionY: Low sense of community was associated overall with higher levels of internalising problems (β(SE) 3.6(1.9)). In wealthy neighbourhoods, knowing neighbours decreased internalising problems; in poor neighbourhoods knowing neighbours was associated with increased internalising problems
Christie-Mizell et al, 2003 51Are the subjective appraisal of the neighbourhood and living in an urban versus rural area associated with maternal psychological distress? Does this differ by race?2204 women with at least one child from NLSYParticipant-defined neighbourhoods (for the subjective appraisal) and type of neighbourhood subjects lived in, as defined by the census(central city vs urban vs rural area)Living in a SMSA, perceived neighbourhood disorder (participant reported)Seven-item version of CES-D scoreLongitudinalSingle-level linear regression with change in CES-D as outcomeY: Across all racial groups, neighbourhood perceptions influence maternal distress (β (SE) 0.17 (0.02)). Also, objective neighbourhood location influences how mothers perceive their neighbourhoods
Cutrona et al, 2005 49Are women who reside in poor and/or dangerous neighbourhoods more likely to experience episodes of major depression than those in safe, affluent ones, after controlling for individual-level risk factors?720 women from FACHS (large-scale study of African-American families who live outside large metropolitan inner cities in the USA)Census block groupsEconomic disadvantage index (from census), neighbourhood-level social disorder (combination of community dilapidation and community deviance scales)UM-CIDICross-sectional and longitudinalMultilevel analysisY: neighbourhood disadvantage/social disorder was associated with recent onset of depression, after controlling for individual-level risk factors (OR 1.92, 1.04 to 3.52). However, neighbourhood disadvantage and disorder did not predict onset of depression at a later date
Cutrona et al, 2000 42What is the effect of neighbourhood context on psychological distress among African-American women? Does community context interact with individual-level risk factors in the prediction of distress?733 African-American women from the FACHS who were the primary caregiver for a 10–12-year-old childCensus neighbourhood block group areasNeighbourhood cohesion and disorder (participant reported), community economic disadvantage (from census)General distress and anxious arousal, two subscales from the mini-mood and anxiety symptom questionnaire (similar to UM-CIDI)Cross-sectionalMultilevel analysisY: neighbourhood disorder (β = 0.38, p<0.05), but not cohesion or economic disadvantage, was associated with level of distress, after controlling for individual-level characteristics. There were significant interactions between some neighbourhood- and individual-level characteristics in the prediction of distress
Dupéré and Perkins, 2007 52Is there variation at the block level in well-being and depression? Do community-level environmental stressors and social resources affect well-being and depression?412 residents from 50 neighbourhoods in a large city in the Mid-Atlantic region of the USACensus blocksNeighbourhood disorder, fear of crime, formal participation, informal ties with neighbours (participant-reported)The 6-item depression factor of the CES-D scaleCross-sectionalMultilevel analysisN: There was no significant variation at the block level for depression, although there was significant variation at the block level for well-being. The community-level stressors and resources had no impact on mental health over and above individual and block socioeconomic characteristics
Fitzpatrick et al, 2005 22Do bonding social ties of youth to their family, school, and community have an inverse relationship with depressive symptoms?1538 African-American middle and high school students from one school district in AlabamaParticipant-defined neighbourhoodsExposure to violence, human capital, social capital (participant reported)CES-DCross-sectionalSingle-level linear regressionY: adolescents exposed to threatening environments had higher rates of depression (β = 1.45, p<0.01); social capital had an inverse relationship with depression (β = −0.18, p<0.05).
Forehand and Jones, 2003 53What is the interactive influence of neighbourhood violence and co-parent conflict on child psychological adjustment?117 African-American children aged 8–14 from an inner-city area of New Orleans with single mothersParticipant-defined neighbourhoodsNeighbourhood risks (physical fighting, shootings or knifings, people being killed) (participant reported)CDILongitudinal and cross-sectionalSingle-level multiple regressionN: neighbourhood violence was not associated with child psychosocial adjustment ((β = −0.04, t = −0.45). However, girls living in homes with high levels of co-parent conflict were more vulnerable to the effect of neighbourhood violence ((β for interaction = 0.27, p<0.05).
Galea et al, 2005 27Are characteristics of the internal and external physical built environment related to depression?1355 residents of New York CityCommunity districts, New York CityCharacteristics of the internal built environment and the external built environment (from census, the New York City housing and vacancy survey, and the fiscal 2002 New York City mayor’s management report)National women’s study depression module, consistent with DSM-IV criteriaCross-sectionalMultilevel analysisY: characteristics of the built environment are associated with likelihood of depression: people living in poor quality built environments were 29–58% more likely to report past 6 month depression and 36–64% more likely to report lifetime depression.
Galea et al, 2007 28Is incident depression associated with urban neighbourhood poverty?1120 adult residents of New York CityCommunity districts, New York CityNeighbourhood SES (from census)Modified version of SCID, 3rd editionLongitudinalMultilevel analysisY: relative odds of incident depression were 2.19 (95% CI 1.04 to 4.59) for participants living in low versus high SES neighbourhoods
Gary et al, 2007 47Are perceptions of neighbourhood characteristics associated with mental health outcomes among African-American and white adults in Baltimore? Do these associations differ by race?1408 African-American and white adult residents of Baltimore, MarylandParticipant-defined (for items relating to perceptions), blocks (for resources)Perceptions of potential neighbourhood problems, availability of a community leader, community cohesion, resources (desirable and undesirable) within the community (participant reported)PHQ-9 and GHQCross-sectionalSingle-level linear and logistic regressionY: perception of severe community problems was associated with depression (OR 2.2 (White), 1.9 (African-American), p<0.05 (both)). Community cohesion was only associated with lower levels of depression in whites (OR 0.5, p<0.05).
Greiner et al, 2004 23What are the associations between level of community participation, self-reported community ratings (trust), and depressive symptoms?4601 subjects from the Kansas BRFSSCounties/participant definedOverall community ratings (neighbourhood-level trust) and social participation (participant reported)Optional depressive symptoms question from BRFSSCross-sectionalMultilevel analysis, single-level logistic regressionY and N: community rating was associated with depression (OR 0.65 (0.57 to 0.75)), but community involvement was not (OR 0.99 (0.71 to 1.36)), after adjustment
Gutman and Sameroff, 2004 54What are the ecological variables that influence depression in males and females from adolescence to young adulthood? Are there gender differences in these associations?372 youth from the first (when subjects were aged 11–15) and second (7–8 years later) waves of the Philadelphia family management study, from four inner-city areas of PhiladelphiaParticipant definedNeighbourhood cohesiveness, neighbourhood problems (participant-reported)Youth depressive symptoms: nine items assessing how often they felt certain symptoms in the past couple of monthsCross-sectional and prospectiveSingle-level linear regressionY: neighbourhood cohesiveness in adolescence was associated with higher levels of depressive symptoms in females in adolescence and early adulthood, and in males depressive symptoms in adolescence. Neighbourhood problems were only associated with increased levels of depressive symptoms in young female adults
Hadley-Ives et al, 2000 26Is the impact of the environment on mental health determined by perception of that environment in adolescents?792 subjects from the Youth Service Project, aged 13–18 in St Louis, MissouriParticipant definedNegative neighbourhood environment (extent to which drug dealing, shootings, murder, abandoned buildings, neighbours on welfare, homeless people on the street, and prostitution exist in neighbourhood), perception of violence (participant reported)DISC-RCross-sectionalSingle-level linear regressionY: perception of neighbourhood was associated with adolescent mental health (β = 0.04, p<0.001)
Henderson et al, 2005 40What is the relation between neighbourhood socioeconomic and ethnic characteristics and depressive symptoms in young adults, and do they modify the relation between individual SES and depression?3437 adults aged 18–30 from the CARDIA study at four US sites (Chicago Illinois, Birmingham Alabama, Minneapolis Missouri, Oakland, California)Census block groupSix area census variables reflecting wealth/income, education, and occupation, were summed to create a neighbourhood summary score; ethnic densityCES-DCross-sectionalMultilevel analysisN: Neither neighbourhood socioeconomic characteristics nor ethnic density were consistently associated with depression after controlling for individual-level characteristics
Hybels et al, 2006 16What is the association between neighbourhood context and level of depressive symptoms in older adults?2998 adults 65+ years old in North CarolinaCensus tractsNeighbourhood SES, racial/ethnic heterogeneity, residential stability, and neighbourhood age structure (from census)CES-DCross-sectionalMultilevel analysisN: none of the neighbourhood characteristics was significantly associated with depressive symptoms conditional on census tract random effects, both before and after adjustment for individual characteristics
Kubzansky et al, 2005 13What is the contribution of neighbourhood disadvantage, neighbourhood service environment, and individual characteristics to depression in older people?2109 non-institutionalised people 65 and older in New Haven, ConnecticutCensus tractNeighbourhood socioeconomic disadvantage and advantage, racial/ethnic heterogeneity, residential stability, age structure (from census), service density (services promoting social engagement, providing care, and undesirable amenities) (constructed from phonebook listings)CES-DCross-sectionalMultilevel analysisY: low neighbourhood SES (β = 6.51 (1.02, to 12.00)) and presence of more older people (β = −13.55 (−24.76 to −2.34)) were associated with depressive symptoms in older people after controlling for individual characteristics, but none of the other neighbourhood measures were
La Gory and Fitzpatrick, 1992 55What is the impact of environmental context (social networks and neighbourhood characteristics) on depressive symptoms? Are there joint effects of personal competence and the residential environment?725 adults aged 55+ from four metropolitan counties in AlabamaCensus tractsRacial congruence, age density (% of people aged 55+ in the census tract) (from census), neighbourhood resource accessibility (availability of automobile transport), perceived environment, social support (participant reported)CES-DCross-sectionalSingle-level linear regressionY: being environmentally dissatisfied, having limited social supports, and living in neighbourhoods with transportation problems are associated with increased levels of depressive symptoms. Significant interactions were found between environmental dissatisfaction and resource accessibility and functional health (p<0.01)
Latkin et al, 2003 24Do subjects from neighbourhoods with outward signs of disorder/that are decaying experience greater uncontrolled stress and symptoms of depression?818 participants in high drug use areas in Baltimore, MarylandParticipant-defined neighbourhoodSocial support, social integration, perception of neighbourhood characteristics (participant-reported)CES-DCross-sectional, follow-up with depressionSingle-level linear regressionY: strong, prospective association between negative perceived neighbourhood characteristics and subsequent depressive symptoms, after adjusting for baseline depression (β = 0.28, p<0.01).
Matheson et al, 2006 43Are depressive symptoms associated with neighbourhood ethnic diversity, dependency, residential instability and material deprivation? Does chronic stress explain gender differences in depression?56428 adults aged 18–74 living in census metropolitan areas in CanadaCanadian Census TractsResidential instability, material deprivation, dependency, ethnic diversity (from Canadian census)CIDI-SF MDCross-sectionalMultilevel analysisY: residential instability (OR 1.04, p<0.05) and material deprivation (OR 1.05, p<0.01) were associated with depression after controlling for individual-level characteristics. Chronic stress was not associated with gender differences in depression
Mulvaney and Kendrick, 2005 37What is the relationship between maternal depressive symptoms and individual- and neighbourhood-level measures of deprivation, social support, and stress in mothers living in deprived areas?846 mothers of young children living in deprived areas (Townsend deprivation scores >0) of Nottingham, UKEnumeration districtsSocial capital, stress, perceived social support, neighbourhood deprivation (participant-reported)CES-DCross-sectionalMultilevel analysis (random effects logistic regression)Y: Neighbourhood deprivation (OR for highest vs lowest fifth 2.4 (1.28 to 4.48)), lack of social support (OR 2.51 (1.75 to 3.61)), and self-reported stress (OR 10.42 (6.29 to 17.28)) were all associated with depressive symptoms in a model adjusting for all these characteristics plus social capital, receiving means-tested benefits, and having 3+ kids <5 years.
Natsuaki et al, 2007 44Does observed neighbourhood disorder prospectively influence African-American adolescents’ depressive symptoms? Does it interact with parents’ engagement in inductive reasoning?777 African-American children aged 9–12 at baseline, from Iowa and GeorgiaClusters of census block group areasInterviewers’ observed neighbourhood disorder at baselineDISC-IVProspectiveMultilevel analysisY: there is an interaction between baseline neighbourhood disorder, parents’ use of inductive reasoning, and depressive symptoms, such that parental use of inductive reasoning was a protective factor for depressive symptoms especially for youths living in highly disordered neighbourhoods (β (SE) −0.14 (0.07))
Ostir et al, 2003 17Is neighbourhood poverty associated with increased depression, and is increasing proportion of older Mexican-Americans associated with decreased depression?2710 non-institutionalised Mexican-Americans aged 65 years or older, from five southwest statesCensus tractPercentage of Mexican-Americans in census tract, neighbourhood SES (from census)CES-DCross-sectionalMultilevel analysis and single-level linear regressionY: after adjustment for individual characteristics, each 10% increase in neighbourhood poverty was associated with a 0.76 (95% CI 0.06 to 1.47) increase in CES-D score, while each 10% increase in Mexican-American population was associated with a 0.55 (95% CI 0.96 to 0.13) decrease
Reijneveld and Schene, 1998 30Is the distribution of mental disorders associated with neighbourhood SES? If this relationship exists after controlling for individual SES, is it due either to selective migration or causation?5121 residents of AmsterdamBoroughs, categorised by deprivation into three levelsArea deprivation, assessed through registered income, household income below minimum, and unemployment rateGHQCross-sectionalMultilevel analysisN: the prevalence of mental disorders is higher in deprived areas, but can be almost fully explained by the sex and SES of residents
Ross, 2000 11Is the impact of neighbourhood disadvantage on adult mental health mediated by disorder in the neighbourhood?2482 Illinois residents from the community, crime and health data setCensus tractRoss-Mirowsky perceived neighbourhood disorder scale (participant reported), objective neighbourhood disadvantage from censusCES-DCross-sectionalMultilevel analysisY: neighbourhood disadvantage affects adult depression (β = 0.228), although more than half of the contextual effects are really due to individual disadvantage (β = 0.081 when individual-level characteristics are added). The effect of neighbourhood poverty is mediated by perceived neighbourhood disorder
Ross et al, 2000 57What are the joint effects of neighbourhood stability and poverty on depression and what mechanisms connect objective neighbourhood characteristics to depression?2482 Illinois residents from the community, crime and health data setCensus tractObjective neighbourhood characteristics (neighbourhood stability, poverty, and their interaction) (from census), perceived neighbourhood disorder (Ross-Mirowsky scale), informal social ties with neighbours, fear, and sense of personal powerlessness (participant reported)Seven-item modification of CES-DCross-sectionalMultilevel analysisY and N: neighbourhood stability is associated with psychological well-being only in economically advantaged neighbourhoods; it has a slight negative effect in poor neighbourhoods
Schieman et al, 2004 45What is the association between perceived neighbourhood problems and mental health among older adults?1167 men and women aged 65+ in Washington DC and two adjoining countiesParticipants told to refer to “the area around where you live”Neighbourhood problems (a modified version of the Ross-Mirowsky perceived neighbourhood disorder scale) (participant reported)Seven items about depressive symptoms in the past weekCross-sectionalOrdinary least squares regression, men and women separatelyY: neighbourhood problems are associated positively with depression in men (β (SE) 0.095 (0.094)) and women (0.087 (0.082)). Received support buffers this association in women
Schulz et al, 2006 41Are household income and length of residence protective of mental health? Are these effects mediated through perceived financial stress, perceived discrimination, and perceived safety?679 African American women living in DetroitParticipant-defined neighbourhoodTwo measures of stressful neighbourhood conditions (concern about police responsiveness, safety stress scale), instrumental social support, emotional social support (participant-reported)CES-DCross-sectionalStructural equation modellingY: household income may be protective of mental health, both directly and indirectly (through reduced financial stress and increased social support). Length of residence is not associated with depressive symptoms
Silver et al, 2002 2Do neighbourhood structural characteristics affect prevalence of mental disorder, after controlling for individual SES?11686 residents from five areas of the USA, in the epidemiological catchment areaCensus tractnine census tract measures used to create two factors: neighbourhood disadvantage and neighbourhood residential mobilityDiagnostic interview schedule, DSM-III diagnosesCross-sectionalSingle-level logistic regressionY: depression was more prevalent in residentially mobile (OR 1.14 (1.03 to 1.27)) and disadvantaged neighbourhoods (OR 1.14 (1.01 to 1.31)), after controlling for individual risk factors
Simons et al, 2002 19What are the community-level correlates of childhood depressive symptoms in an African-American sample?867 African-American children aged 10–12 in Georgia and Iowa.Community groups, made up of census block group areas from cluster analysisCommunity cohesion, community ethnic identification, prevalence of discrimination and crime (participant-reported), and neighbourhood poverty (from census)DISC-IVCross-sectionalMultilevel analysisY: community ethnic identification (β = −0.392, p-value = 0.04) and prevalence of discrimination (β = 0.313, p-value = 0.04) were associated with child depressive symptoms, after controlling for individual- and community-level characteristics. The other community-level variables were not associated with depressive symptoms
Skapinakis et al, 2005 29Do regional mental health differences in Wales persist after taking into account individual characteristics and regional social deprivation?26710 people from the Welsh Health Survey (covering all of Wales)The 22 regional unitary authorities of WalesWelsh index of multiple deprivationSF-36Cross-sectionalMultilevel analysisY: there was a significant difference in psychiatric morbidity between regions (1.47% of variance), which was reduced but persisted after adjusting for individual-level characteristics (0.99%)
Steptoe et al, 2001 31Is neighbourhood stress associated with psychological distress? Is the association independent of neighbourhood SES, individual SES, and social capital differences?658 subjects (survey respondents) living in the London areaU.K. postal sectorsNeighbourhood problems, social cohesion, informal social control, neighbourhood SES (combined participant reports and census information)GHQCross-sectionalMultilevel analysis, single-level logistic regressionY: Highest quartile of neighbourhood problems had higher distress levels (OR = 2.65(1.47–4.47)), adjusted for social cohesion and control. Neither social cohesion nor social control was associated with depression.
Stevenson, 1998 56Do African American youth in self-reported unsafe neighbourhoods have higher levels of depression? Do teens with supportive families and neighbourhoods have better psychological outcomes than those with only one of these supports?160 low-income, inner-city African-American adolescents in a northeastern US cityParticipant-defined neighbourhoodsNeighbourhood social capital, neighbourhood risk, fear of calamity (a measure of negative urban life experiences) (participant reported)SMDICross-sectionalSingle-level linear regressionY: only neighbourhood social capital was significantly associated (t = −2.16, p<0.05) with depression in multiple regression analyses
Tweed et al, 1990 38What is the effect of exposure to racially dissonant residential environments on depressive psychopathology?3481 adults aged 18+ from the eastern third of the city of BaltimoreCensus tractsRacial congruence (% of the residential area population that is the same racial/ethnic group as individuals)Depressed mood (fulfilling criterion A of DSM-III), major depressive episode (DIS/DSM-III diagnosis)Cross-sectionalCalculation of z-scores, summary tests of significance comparing prevalence rates of depressionY: an inverse relationship exists between racial congruity and depressed mood
Wainwright and Surtees, 2004 32Is there area level variation in mental functional health after controlling for individual level SES? What’s the extent of the area level and individual level variation?20921 participants in the EPIC-Norfolk studyElectoral wardsOverall index of multiple deprivationSF-36Cross-sectionalMultilevel analysisY: area deprivation was associated with poor mental health, but the residual variation after adjusting for individual level risk factors was modest for men and non-existent for women
Wainwright and Surtees, 2004 33What is the relative strength of the association between individual and area-level demographic and socioeconomic factors and depression?19687 participants in the EPIC-Norfolk studyElectoral wardsOverall index of multiple deprivationDSM-IV criteriaCross-sectionalSingle-level logistic regression, multilevel analysisY: an association remained between area deprivation and current mood disorders, after adjusting for individual-level risk factors (OR for top vs bottom quartile of deprivation 1.29 (1.1 to 1.5)). Significant area-level residual variation remained
Weich et al, 2001 46Do individuals in regions with the highest income inequality have a higher prevalence of depression, after adjustment for individual income?5511 participants in BHPS (a representative sample of individuals in private households in England, Wales, and Scotland)Standard regionsGini coefficient (income inequality)GHQCross-sectionalSingle-level linear regressionN: no significant association between Gini coefficient and depression (0R 0.99 (0.87 to 1.13), although there was a significant interaction between individual income and Gini (p<0.01).
Weich et al, 2002 35Is depression most prevalent in areas characterised by derelict buildings and abundant graffiti, open public spaces and few buffers between public and private spaces?1887 people from two wards in London, UK“Housing areas,” 86 areas with homogenous housing type and formBuilt environment site survey checklist, an inventory for rating housing areas carried out by an urban design postgraduate who did not live in the areaCES-DCross-sectionalSingle-level logistic regression, linear regressionY: there was an association between depression and characteristics of the built environment, which remained after adjusting for individual SES and internal characteristics of dwellings (OR for properties with deck access 1.28 (1.03 to 1.58); OR for recent construction 1.43 (1.06 to 1.91)).
Weich et al, 2003 50Do people living in urban areas have higher rates of depression, after adjusting for personal SES? Do people living in disadvantaged neighbourhoods have higher rates of depression, after adjusting for personal SES?8978 respondents from the BHPS in Britain, Scotland, and WalesElectoral wards, grouped into 14 principal groups by demographic and socioeconomic composition, and householdsCarstairs index of socioeconomic deprivationGHQCross-sectionalMultilevel logistic and linear regressionN: there is little independent area-level variance in the prevalence of depression and anxiety, except amongst unemployed residents. There is no association between socioeconomic deprivation and depression, but there is household-level variation
Weich et al, 2005 34What is the variance in onset and maintenance of common mental disorders at individual, household and electoral ward levels? Will ward-level socioeconomic deprivation be associated with episode maintenance after controlling for individual and household factors?BHPS participants aged 16–74; 9518 at wave 1 and 7659 at wave 2Electoral wardsCarstairs index of socioeconomic deprivationGHQProspectiveMultilevel logistic regressionN: differences in rates of maintenance and onset of depression and change in score between waves across electoral wards are negligible (<1%) compared to those between households and individuals. Ward level socioeconomic deprivation does not influence the onset and maintenance of common mental disorders
Xue et al, 2005 20Is children’s mental health associated with neighbourhood structural characteristics (concentrated disadvantage, residential stability, immigrant concentration)? Do collective efficacy and organisational participation underlie these effects?2805 children from Chicago, IllinoisNeighbourhood clusters, made up of two or three census tractsNeighbourhood structural measures from census loaded into three factor scores (concentrated disadvantage, immigrant concentration, residential stability), informal social control, and social cohesion (participant reported)CBCL/4–18ProspectiveMultilevel analysisY: neighbourhood concentrated disadvantage was associated with the prevalence of children’s mental health problems, after controlling for individual characteristics (β = 0.088, p<0.01). This effect was accounted for by informal social control and social cohesion
Yen and Kaplan, 1999 21Will poverty area residence lead to increased levels of depressive symptoms?1737 participants in the Alameda County study who resided in Oakland County California in 1965 and who responded in 1974Poverty areas and non-poverty areasPoverty area (contiguous census tracts based on 1965 criteria)From response to 18 questions; similar to CES-DLongitudinalSingle-level logistic regressionY: living in a poverty area was associated with increased risk of depressive symptoms after adjustment for age and sex (OR 2.09 (1.49 to 2.99)); this became non-significant (OR 1.21 (0.76 to 1.93)) with additional confounders added to the model
Yen et al, 2006 25What are the associations between perceived neighbourhood problems and quality of life and depressive symptoms amongst adults with asthma?435 adults with asthma in northern CaliforniaParticipant-defined neighbourhoodsNeighbourhood problems (too much traffic, excessive noise, trash and litter, smells, smoke) (participant reported)CES-DCross-sectionalSingle-level linear regressionY: subjects in the top quartile of neighbourhood problems were more likely have depressive symptoms than the bottom quartile, after adjustment (OR 4.8 (2.4 to 9.5).
  • *It is difficult to distinguish between neighbourhood constructs and neighbourhood measures in some studies, so they have been combined for the purposes of this table. AHEAD, Study of Asset and Health Dynamics Among the Oldest Old; BHPS, British Household Panel Survey; BRFSS, Behavioral Risk Factors Surveillance System; CARDIA, Coronary Artery Risk Development in Young Adults; CBCL, Child Behavior Checklist; CES-D, Center for Epidemiologic Studies-Depression score; CDI, Children’s Depression Inventory; CIDI-SF MD, Composite Diagnostic Interview Schedule Short Form for major depression; DISC-IV, Diagnostic Interview Schedule for Children, Version 4; DISC-R, Diagnostic Interview Schedule for Children; DSM-IV, Diagnostic and Statistical Manual of Mental Disorders; FACHS, Family and Community Health Study; EPIC-Norfolk, European Prospective Investigation into Cancer and Nutrition in Norfolk, UK; GHQ, General Health Questionnaire; N, no; NLSY, National Longitudinal Survey of Youth; PHQ-9, Patient Health Questionnaire; SCID, Structured Clinical Interview for Diagnostic and Statistical Manual of Mental Disorders; SES, socioeconomic status; SF-36, Mental health index of the Short Form Health Survey 36; SMDI, Multiscore Depression Index, short form; SMSA, Standard metropolitan statistical area; UM-CIDI, University of Michigan Composite International Diagnostic Instrument; WBC, Walkable and Bikable Communities Project; Y, yes.