Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2015 May 1.
Published in final edited form as: Health Place. 2014 Mar 4;27:120–126. doi: 10.1016/j.healthplace.2014.02.001

Socioeconomic deprivation, perceived neighborhood factors, and cortisol responses to induced stress among healthy adults

Wendy E Barrington 1,4, Mai Stafford 2, Mark Hamer 3, Shirley AA Beresford 1,4, Thomas Koepsell 1, Andrew Steptoe 3
PMCID: PMC4212220  NIHMSID: NIHMS572890  PMID: 24603009

Abstract

Associations between measures of neighborhood socioeconomic deprivation and health have been identified, yet work is needed to uncover explanatory mechanisms. One hypothesized pathway is through stress, yet the few studies that have evaluated associations between characteristics of deprived neighborhoods and biomarkers of stress are mixed. This study evaluated whether objectively measured neighborhood socioeconomic deprivation and individual perceived neighborhood characteristics (i.e. social control and fear of crime) impacted cortisol responses to an induced stressor among older healthy adults. Data from Heart Scan, a sub-study of the Whitehall II cohort, were used to generate multilevel piecewise growth-curve models of cortisol trajectories after a laboratory stressor accounting for neighborhood and demographic characteristics. Neighborhood socioeconomic deprivation was significantly associated with individual perceptions of social control and fear of crime in the neighborhood while an association with blunted cortisol reactivity was only evidence among women. Social control was significantly associated with greater cortisol reactivity and mediation between neighborhood deprivation and cortisol reactivity was suggested among women. These findings support a gender-dependent role of neighborhood in stress process models of health.

Keywords: neighborhood deprivation, stress, social control, cortisol

1. Introduction

Recent reviews of the neighborhood effects literature have concluded that neighborhood deprivation, typically measured by aggregated census information on income, employment and education, is modestly associated with chronic disease and mortality after adjustment for individual-level socioeconomic factors (Diez Roux and Mair, 2010, Meijer M et al., 2012). Neighborhood effects on mental health have also been demonstrated(Truong and Ma, 2006). Although associations between neighborhood deprivation and health outcomes may be small, addressing these exposures is of public health importance given the large absolute number of persons exposed to deprived neighborhoods and associated neighborhood factors as well as the socioeconomic and racial disparities in exposure to these neighborhoods (Diez Roux and Mair, 2010). Identifying potential mechanisms that explain how living in deprived neighborhoods influence more proximal health-related processes may contribute to our understanding of associations between neighborhoods and health and ultimately inform efforts to reduce health disparities. To this end, stress has emerged as one potential embodiment framework for neighborhood effects on health (Diez Roux and Mair, 2010, Brenner et al., 2013, Schulz et al., 2013).

Stress has been defined as a condition of living resulting in the activation of neurobiological systems that preserve viability through change or allostasis (Cohen et al., 2007, McEwen and Seeman, 1999). When external demands exceed the adaptive capacity of an individual, cognitive, behavioral, as well as physiological processes occur which may increase disease risk (Cohen et al., 2007, McEwen and Seeman, 1999). Stress may stem from the presence of socio-environmental demands such as social or physical factors associated with deprived neighborhoods or other elements of disadvantaged social status (Pearlin et al., 1981, Steptoe and Feldman, 2001) and these demands may be more frequent, severe, and sustained (Latkin and Curry, 2003). Studies among adults evaluating neighborhood influences on perceived stress after controlling for individual socioeconomic factors are inconclusive (Brenner et al., 2013, Everson-Rose et al., 2011). More evidence of the impact of neighborhood factors on biomarkers of stress among adults is emerging related to diurnal cortisol profile (Karb et al., 2012, Do et al., 2011, Roe et al., 2013). While no studies have evaluated neighborhood context with respect to cortisol reactivity among adults, a related and mixed literature among adolescents does exist (Brenner et al., 2012, Hackman et al., 2012, Kapuku et al., 2002, Rudolph et al., 2013). While it is has been suggested that alterations to biomarkers of stress, either diurnal cortisol profile or cortisol reactivity, may more accurately reflect the cumulative effects of chronic exposure to neighborhood stressors (Brenner et al., 2012), it is still possible that the severity and timing of the exposure varied across studies and contribute to the lack of consistent findings.

The hypothalamic-pituitary-adrenal (HPA) axis is one of the major systems that modulate physiologic responses to stress (Cohen et al., 2007). Cortisol, the main regulatory hormone of the HPA axis, exhibits a basal diurnal profile with superimposed fluctuations when cortisol temporarily increases in response to an encountered stressor. Cortisol is regulated by feedback loops that may be disrupted by the effects of previous stress exposure (Cohen S et al., 2012). A departure from normal HPA functioning could manifest as differences in total daily cortisol output (i.e. hypo- and hyper-cortisolism) (Miller et al., 2007, Fries et al., 2005) as well as differences in cortisol responses to stress (i.e. blunted and exaggerated responses) (Cohen S et al., 2012, Silverman MN and Sternberg EM, 2012) which may have health consequences (Phillips AC et al., 2013, Cohen S et al., 2012, Hamer et al., 2012).

Diurnal profile and acute stress responsivity measures of cortisol likely measure distinct functional aspects of the HPA system; separate evaluation of the impact of neighborhood deprivation on these patterns is needed (van Eck M et al., 1996). Studies linking neighborhood factors and diurnal cortisol are emerging; exposure to neighborhood stressors has been associated with blunting of the diurnal cortisol secretory profile among adults in two large diverse samples (Karb et al., 2012, Do et al., 2011). With respect to cortisol reactivity, one of three studies among predominantly African-American adolescents found that measures of neighborhood deprivation were associated with blunted cortisol responses to induced stress (Hackman et al., 2012, Kapuku et al., 2002, Brenner et al., 2012). Studies relating measures of neighborhood to cortisol stress reactivity in adults, however, are lacking. To address this gap in the literature, the primary objective of this study was to evaluate whether neighborhood socioeconomic deprivation was associated with the acute responsivity of cortisol among older white British adults.

Furthermore, individual perceptions of the neighborhood may also be important to consider when assessing associations between neighborhood deprivation and cortisol response as perception is integral to the appraisal of stress. Social control and perceived crime are individual-level neighborhood perceptions that are associated with deprivation and may mediate associations with health (Ross and Mirowsky, 2001, Feldman and Steptoe, 2004). These perceptions may provide a means of internalizing the effects of neighborhood deprivation: social control has been shown to attenuate the impact of neighborhood stressors on depression (Diez Roux and Mair, 2010) while fear of crime has been associated with poorer mental health, reduced physical functioning, and lower quality of life in the larger Whitehall II cohort (Stafford et al., 2007). These relationships suggest that individual perceptions of the neighborhood could mediate associations between neighborhood deprivation and cortisol responses to stress (Figure 1). Our secondary objective, therefore, was to determine whether individual-level social control or fear of crime explained any relationship between neighborhood socioeconomic deprivation and cortisol response.

Figure 1.

Figure 1

Open in a new tab

Conceptual model relating neighborhood deprivation, fear of crime, social control, and cortisol responses to stress.

2. Methods

Data were drawn from the Heart Scan Study, an ancillary investigation of socioeconomic and psychosocial factors, physiological stress responsivity, and sub-clinical cardiovascular outcomes within the Whitehall II cohort(Steptoe et al., 2010). Inclusion of men and women aged 53–76 years was determined by eligibility criteria: no history or clinical signs of coronary heart disease; no previous diagnosis or treatment for hypertension, inflammatory diseases, or allergies; and no intake of medication which might affect cortisol levels and sympathetic nervous system activity (i.e. hormone treatment and beta-blockers). The study group comprised 543 individuals of white European ancestry; our sample is limited to those living in England without missing data on exposures, covariates, or outcome for a total of 514 individuals. Individual-level age, gender, income, and employment as well as neighborhood-level socioeconomic deprivation were similar between the study sample and those dropped from the analysis (p>0.06 for all).

2.1 Cortisol stress responses

Psychophysiological testing for the Heart Scan Study occurred between 2006–2008 and methodology has been described elsewhere (Hamer et al., 2010). Briefly, saliva was collected for baseline assessment of cortisol after the participant rested for 30 minutes. Two 5-minute behavioral tasks (color/word interference and mirror-tracing) were then administered in random order. After completion of both tasks, a saliva sample was collected along with subjective stress ratings of each task. Participants then rested for 75 minutes and provided post-task cortisol measurements at 30 minutes, 55 minutes, and 85 minutes after baseline. Levels of cortisol (nmol/L) were assessed using a time resolved immunoassay with fluorescence detection where the intra and inter-assay coefficients of variation were <8%.

2.2 Perceived neighborhood factors

Social control

We used a modified scale of informal social control (Sampson et al., 1997) during the 2006–2008 survey administration to assess whether respondents perceived that their neighbors would take action to safeguard the welfare of the neighborhood. Participants responded whether they thought that their neighbors could be counted on to intervene in various ways if: 1) children were truant and loitering; 2) children were doing graffiti on a local building; 3) children were showing disrespect to an adult; 4) fighting broke out in front of your home; and 5) the school closest to home was threatened with budget cuts (Feldman and Steptoe, 2004). Responses ranged from “very unlikely” to “very likely” on a 5 point Likert scale. The social control scale ranged from 0 to 25 (Cronbach’s alpha=0.87).

Fear of crime

To assess the level of perceived danger due to violence in their neighborhood, participants indicated how worried they were regarding: 1) having their home broken into; 2) being mugged or robbed; 3) having their car broken into or stolen; and 4) being raped. Possible responses were “very worried”, “fairly worried”, “not very worried”, or “not worried at all”. This measure originated from the British Crime Survey (Kaewthummanukul and Brown, 2006) and was included during the 2002–2004 survey administration. We dropped the last scale item as the relationship between fear of crime and neighborhood context likely differs for men and women, especially with respect to the fear of sexual assault (Lane et al., 2009). The modified fear of crime score ranged from 0–9 (Cronbach’s alpha=0.78).

2.3 Neighborhood socioeconomic deprivation

Participants provided their home address as well as the number of years they lived at that address at the time of psychophysiological testing. Each postcode was matched to a lower layer super output area (LSOA), the smallest aggregation used to disseminate census information in England and Wales. There are 32,482 LSOAs in England with a population range of 1,000 to 3,000 and an average of 1,500 people (Department for Communities and Local Government) which is analogous to a US Census blockgroup (Kim D et al., 2010). The mean number of participants per LSOA was 5.1 with a range between 4 and 10.

Linked neighborhood socioeconomic measures were obtained from the Index of Multiple Deprivation (IMD) 2007 which provides deprivation information based on 38 indicators grouped in seven domains including income and employment (UK Statistics Authority). Percent income and employment deprivation within each LSOA are each based on six indicators capturing proportions of residents living in poverty and in receipt of unemployment-related government support, respectively. We linked IMD income and employment deprivation variables to participants via matching each person to the LSOA within which they resided. The percent income and employment deprivation variables were highly correlated (r=0.89): we therefore standardized and averaged both to create a composite continuous z score of socioeconomic deprivation (Nazmi et al., 2010).

2.4 Independent Variables and Covariates

Grade of employment and household income were used as individual-level socioeconomic measures. Employment categories corresponding to status included low (i.e. grades 7 to 1), intermediate (i.e. higher executive office and senior executive office), and high (i.e. administrative assistant, administrative officer, executive officer). Participant’s household income comprised three categories: less than 20,000 pounds; 20,000–39,999 pounds; greater than 40,000 pounds. Covariates used in all analyses included age, gender, number of years lived at provided address, time of day of stress testing (am or pm), and subjective stress rating of administered tasks (response range from 1 to 7 with higher ratings indicative of higher stress). Other psychosocial and behavioral variables such as depression, anxiety, smoking-status, and body mass index (BMI) were not included as confounding variables in these analyses as we conceptualized these variables as potential mediators linking associations between our neighborhood-related measures and cortisol response.

2.5 Statistical analysis

For descriptive purposes, differences in socio-demographic characteristics, length of neighborhood residence, and neighborhood perceptions were examined by tertile of neighborhood deprivation using chi-square test for trend. The data had a hierarchical structure with cortisol measurements over time (level 1) nested with individuals (level 2) nested within neighborhoods (level 3). We first plotted unadjusted mean cortisol values over time using a two-power fractional-polynomial curve. Visual inspection of the plotted data suggested two linear functions for the reactivity period (baseline to immediately post-task) and the recovery period (immediately post-task to 75 minutes post-task).

We used piecewise multilevel growth curves to describe cortisol trajectories with respect to time and to assess whether neighborhood deprivation and perceptions of the neighborhood were associated with cortisol trajectories while controlling for covariates. These methods have been employed in similar studies (Goldman-Mellor et al., 2012, Hackman et al., 2012) with coding of the time variable derived from Llabre and colleagues (Llabre et al., 2001). In brief, time is partitioned into “reactivity” and “recovery” vectors. The reactivity time vector corresponds to the total slope between baseline (time 0) and immediately post-task (time 1) while the recovery time vector corresponds to the total slope between immediately post-task (time 1) and 75 minutes post-task (time 4). The main effects of variables in the model specify the relationship of that variable with absolute baseline levels of cortisol while variable interactions with time (i.e. reactivity and recovery) specify that variable’s relationship with those pieces of the cortisol trajectory. We log transformed cortisol values to account for skewness and attempted to fit a quadratic term when modeling recovery, but this variable did not significantly improve model fit (p=0.11).

Model 1 included fixed effects for neighborhood deprivation and its interaction with the time vectors (i.e. reactivity and recovery) while accounting for time of laboratory visit, task stress rating, and individual-level demographic characteristics. This allowed us to test for differences corresponding to neighborhood deprivation in both baseline cortisol levels and slopes during the two time periods. Models 2 and 3 included fixed effects for fear of crime and social control, respectively. These models also specified interactions between these variables and the reactivity and recovery time vectors Model 4 included fixed effects for both neighborhood deprivation and social control to evaluate whether social control accounted for any association between neighborhood deprivation and cortisol response. Random effects were specified for the intercept as well as reactivity and recovery slopes for both level 2 (person) and level 3 (neighborhood). An unstructured working correlation structure was assumed. For illustrative purposes, we also plotted the adjusted predicted mean cortisol values (untransformed) over time by tertile of neighborhood deprivation (Model 1) and social control (Model 3). All analyses were conducted using XTMIXED in Stata SE version 12.1 (College Station, TX, USA).

3. Results

The mean baseline cortisol of the total group was 6.7 nmol/dL, the mean age was 62.9 years (SD= 5.7), slightly more than half of individuals were male (54.9%), 21.5% had an income < £20,000, and 19.8% were of lower employment grades. The mean social control score was 11.7 (SD=4.7) and mean fear of crime score was 2.7 (SD=1.6). Within represented LSOAs, income deprivation ranged between 1% and 46% of persons (Mean= 9.5%, SD=7.5) and employment deprivation ranged between 1% and 24% of persons (Mean=6.3%, SD=3.7).

Table 1 shows selected characteristics of Heart Scan participants by tertile of neighborhood socioeconomic deprivation score. Individuals living in more deprived areas were more likely to be female, of lower employment status, and have lower incomes (p<0.0001 for all). Individuals living in more deprived areas also reported greater fear of neighborhood crime as well as lower neighborhood social control (p<0.0001 for both). The subjective stress rating from laboratory tasks did not vary by level of neighborhood deprivation. The average number of years participants lived in their neighborhood was high (>20 years) and did not vary across levels of deprivation. Individual reports of fear of crime and social control varied across tertiles of neighborhood deprivation (p<0.0001 for both). Social control and fear of crime were not significantly correlated.

Table 1.

Selected characteristics of Heart Scan participants by tertile of neighborhood socioeconomic deprivation score (n=514).

Tertile of Neighborhood
Deprivation Scorea P for Trendb
1-Low (n=189) 2-Medium (n=176) 3-High (n=176)
Afternoon visit (%) 60.6 60.7 56.6 0.10
Mean (SD) baseline cortisol, nmol/L 6.7 (4.8) 6.3 (3.8) 7.1 (5.0) 0.99
Mean (SD) task stress rating 4.0 (1.5) 4.3 (1.4) 4.1 (1.3) 0.19
Mean (SD) age, years 62.9 (5.7) 63.0 (5.9) 62.6 (5.6) 0.48
Female (%) 37.2 46.4 52.4 <0.0001
Employment status (%) <0.0001
 Low (grades 7 to 1) 48.9 38.7 32.5
 Intermediate (e.g. higher executive office) 39.4 41.1 39.2
 High (e.g. executive office) 11.7 20.2 28.3
Household income (%) <0.0001
 <£20,000 12.2 22.0 28.9
 £20,000 to £39,999 43.9 39.3 35.6
 >£40,000 43.9 38.7 35.5
Mean (SD) neighborhood residence, years 23.1 (12.1) 21.8 (14.1) 24.5 (14.9) 0.19
Mean (SD) perceived neighborhood factors
 Fear of crimec 2.5 (1.7) 2.8 (1.6) 2.9 (1.6) <0.0001
IQR: 1–3 IQR: 2–3 IQR: 2–3
 Social controld 12.1 (4.7) 12.0 (4.6) 10.9 (4.7) <0.0001
IQR: 10–15 IQR: 10–15 IQR: 8–14
Open in a new tab
a

Neighborhood socioeconomic deprivation based on the IMD 2007 income and employment deprivation rates within a lower super output area (LSOA)

b

P values by non-parametric test for trend across ordered tertiles of neighborhood deprivation

c

Fear of crime score range: 0–9

d

Social control score range: 0–20

Table 2 displays multivariate results from piecewise growth curve models. In model 1, neighborhood deprivation bore little relation to baseline cortisol values or recovery slope, but was marginally associated with a significant decrement in reactivity. In model 2, we tested whether these cortisol profiles were related to fear of crime, but this was not the case. In model 3, social control was significantly positively associated with cortisol reactivity. For each unit increase of social control, there was a significantly steeper (i.e. positive) cortisol reactivity slope of 0.01 nmol/L per minute. Model 4 tested whether variations in social control were responsible for the association we observed between greater socioeconomic deprivation and reduced cortisol reactivity. Yet, social control did not appear to mediate the association between neighborhood deprivation and cortisol trajectories. Although including social control in the model of neighborhood deprivation and cortisol dramatically improved model fit (as demonstrated by the lower log-likelihood of model 4 vs. model 1), its presence in the model did not appreciably alter the relationship between neighborhood deprivation and cortisol trajectories as evidenced by the lack of change in regression coefficients. Higher age and lower individual-level SES (as measured by employment grade) was also associated with higher baseline cortisol levels in this model. For illustrative purposes, adjusted predicted cortisol trajectories over the laboratory testing session are presented by tertiles of socioeconomic deprivation (Figure 2) and social control (Figure 3).

Table 2.

Piecewise growth-curve models of log cortisol.

Model 1 Model 2 Model 3 Model 4
Esta SE pb Esta SE pb Esta SE pb Esta SE pb
Intercept (baseline) 1.65 0.23 1.60 0.24 1.67 0.24 1.68 0.24
Reactivity (slope, per minute) 0.07 0.02 <0.0001 0.10 0.03 0.001 −0.04 0.04 0.28 −0.03 0.04 0.38
Recovery (slope, per minute) −0.09 0.01 <0.0001 −0.09 0.01 <0.0001 −0.07 0.02 <0.0001 −0.06 0.02 <0.0001
Afternoon visit −0.21 0.03 <0.0001 −0.20 0.04 <0.0001 −0.21 0.04 <0.0001 −0.21 0.04 <0.0001
Task stress rating 0.01 0.01 0.28 0.02 0.01 0.16 0.01 0.01 0.36 0.01 0.01 0.39
Age (years) 0.01 0.003 0.03 0.01 0.003 0.03 0.01 0.003 0.03 0.01 0.003 0.04
Female −0.001 0.03 0.97 −0.003 0.04 0.93 −0.01 0.03 0.79 −0.002 0.03 0.95
Employment status d 0.03 0.12 0.005 0.03 0.01 0.01 0.03 0.01 0.02 0.03 0.01 0.007
Household income (£) −0.01 0.03 0.62 −0.02 0.03 0.55 −0.01 0.03 0.74 −0.01 0.03 0.62
Neighborhood tenure (years) 0.001 0.001 0.27 0.002 0.001 0.23 0.001 0.001 0.41 0.001 0.001 0.35
Neighborhood deprivation (NDEP)c −0.02 0.02 0.45 −0.02 0.02 0.48
NDEP x reactivity −0.03 0.02 0.08 −0.03 0.02 0.08
NDEP x recovery −0.004 0.01 0.60 −0.003 0.01 0.65
Fear of Crime (FOC) 0.01 0.01 0.54
FOC x reactivity −0.01 0.01 0.31
FOC x recovery 0.00 0.004 0.94
Social control 0.00 0.004 0.99 0.00 0.004 0.98
Social control x reactivity 0.01 0.003 0.003 0.01 0.003 0.006
Social control x recovery −0.002 0.001 0.11 −0.002 0.001 0.99
Random- effects parameters
Between neighborhood
 Variances
  Intercept 0.004 0.007 0.002 0.002
  Reactivity 0.02 0.03 0.02 0.02
  Recovery 0.002 0.002 0.002 0.002
 Covariances
  Intercept/reactivity −0.01 −0.01 −0.01 −0.01
  Intercept/recovery 0.0005 0.0009 −0.0002 −0.001
  Reactivity/recovery −0.0006 −0.001 0.0004 0.0001
Between Person
 Variances
  Intercept 0.16 0.16 0.16 0.16
  Reactivity 0.03 0.03 0.03 0.03
  Recovery 0.01 0.01 0.01 0.01
 Covariances
  Intercept/reactivity 0.004 0.009 0.002 0.001
  Intercept/recovery −0.03 −0.03 −0.02 −0.03
  Reactivity/recovery −0.002 −0.002 −0.001 −0.001
129 137 122 122
AIC 0.82 5.27 8.67 7.36
Open in a new tab
a

Beta coefficients calculated using multilevel piecewise growth curve modeling assuming exchangeable correlation

b

Wald test

c

Neighborhood socioeconomic deprivation based on the IMD 2007 income and employment deprivation rates within a lower layer super output area (LSOA)

d

Coded from low to high status

Figure 2.

Figure 2

Open in a new tab

Predicted cortisol reactivity to acute stressor over time among men and women by tertiles of neighborhood deprivation score.

Figure 3.

Figure 3

Open in a new tab

Predicted cortisol reactivity to acute stressor over time among men and women by tertiles of social control score.

As previous studies among adolescents have reported gender differences in associations between neighborhood and cortisol reactivity (Hackman et al., 2012), we explored differences in relationships between neighborhood-related variables and cortisol reactivity by gender. While the relationship between neighborhood deprivation and fear of crime did not vary by gender (p=0.66), there was a significant gender interaction in the relationship between neighborhood deprivation and social control (p<0.0001). Given the complexity of modeling an overall three-way gender x neighborhood x time interaction, we chose to stratify our models by gender and found that higher neighborhood deprivation (p=0.047) and lower social control (p=0.006) were both significantly associated with blunted cortisol among women only (see Figures 2 and 3). Furthermore, the magnitude of association between neighborhood deprivation and cortisol reactivity decreased and became marginally significant when social control was also included in the model. This suggests a partial mediating role of social control in the relationship between neighborhood deprivation and cortisol reactivity among women.

We also explored whether other potential mediators such as depression, anxiety, smoking-status or BMI could instead explain relationships between neighborhood socioeconomic deprivation and cortisol reactivity in women. Separate models for women were generated comprising variables in Model 4 with fixed effects for depression and its interactions with reactivity and recovery time vectors. Similarly, separate models were also generated to evaluate anxiety, smoking-status, and BMI. Independent associations between social control and cortisol reactivity remained in separate models where depression, anxiety, smoking-status, and BMI were added (p<0.02 for all) and partial mediation by social control in the relationship between neighborhood deprivation and cortisol reactivity was still suggested.

4. Discussion

Stress has been put forward as a mode of embodiment for neighborhood influences on health(Diez Roux and Mair, 2010, Truong and Ma, 2006). Neighborhood factors related to socioeconomic deprivation have been conceptualized as chronically experienced stressors (Pearlin et al., 1981, Steptoe and Feldman, 2001); cumulative stress exposure induces alterations to HPA functioning that have been linked to health outcomes (Cohen et al., 2007). The aim of this study was to explore associations between neighborhood deprivation and related individual-level perceptions of the neighborhood and cortisol stress reactivity to an induced stressor as a functional measure of the HPA system. Here, lower perceived social control in the neighborhood was significantly associated with blunted cortisol response to an induced stressor. Higher neighborhood socioeconomic deprivation exhibited a similar relationship that did not reach statistical significance. Exploratory analyses by gender, however, revealed that both neighborhood socioeconomic deprivation and social control were significantly associated with cortisol reactivity among women. Furthermore, a mediating role of social control in the relationship between neighborhood socioeconomic deprivation and cortisol reactivity was suggested. Perceiving less social control in one’s neighborhood may be indicative of perceiving more physical and/or social threats which, in turn, gives rise to psychological and physiologic stress responses which affect health (Hill et al., 2005). Specifically, these blunted cortisol responses might reflect exhaustion of allostatic regulatory mechanisms, including cortisol down-regulation or alterations in feedback sensitivity, associated with chronic stress exposure (Hellhammer and Wade, 1993, Phillips et al., 2013, Fries et al., 2005). Our results are consistent with some (Roy MP, 2004, Roy et al., 1998), but not all (Brenner et al., 2012) studies finding that individual-level social support is associated with greater cortisol reactivity. Another study among adolescents did find that measures of neighborhood socioeconomic deprivation was associated with blunted cortisol reactivity only among those reporting a high number of chronic stressors (Brenner et al., 2012). Differences by gender were not noted in these studies, however. Relatedly, the magnitude of associations between our exposure variables and cortisol trajectories were similar to associations seen between early-life stress and recurrent distress (as indicators of chronic stress) and cortisol trajectories within this same cohort (Goldman-Mellor et al., 2012).

Individual fear of crime did not impact cortisol responses in these data. It is possible that there was not a sufficient range in this exposure to detect associations as overall fear of crime was low in this cohort. Another study among African-American adolescents also found no association between fear of violence in the neighborhood and cortisol reactivity (Brenner et al., 2012). In one longitudinal study among adolescents, however, recent exposure to violence was associated with blunted reactivity in boys (Peckins et al., 2012). It is plausible that exposure to violence may be positively correlated with fear of crime and thus be similarly related to cortisol reactivity. Evaluating objective crime statistics for neighborhoods with respect to reactivity may therefore be informative.

While the present study is the first to evaluate neighborhood-related influences on cortisol reactivity among older white adults, three other studies have evaluated exposure to neighborhood deprivation and cortisol reactivity among predominantly African-American adolescents (Hackman et al., 2012, Brenner et al., 2012, Kapuku et al., 2002). In one study, concentrated neighborhood disadvantage was associated with greater cortisol reactivity among African-American boys, but not girls (Hackman et al., 2012). In 2 other studies among African-American boys, neighborhood deprivation was not associated cortisol reactivity (Brenner et al., 2012, Kapuku et al., 2002). Inconsistencies between the present study findings and those from studies of African-American youth could be attributed to differences in HPA responses based on age, sex, and race (Kudielka et al., 2004, Birditt et al., 2011, Wilcox et al., 2005). Additionally, although perceived social control in the neighborhood was measured at the individual-level, our findings may also complement other studies demonstrating associations between objective and perceived neighborhood factors and differences in diurnal cortisol pattern, another measure of HPA functioning (Do et al., 2011, Karb et al., 2012).

There were limitations and strengths to our study that should be acknowledged. First, perceptions of the neighborhood were measured by self-report and at the individual- rather than neighborhood-level. Neighborhood perceptions were also measured at a single-point in time which prevents clear conclusions about the directionality of the association we observed with social control. Cortisol reactivity was also measured during a single laboratory session, although strong reproducibility of these responses over two repeated stress sessions have been demonstrated (Hamer M et al., 2006). The Heart Scan Study comprised mostly white individuals with selection criteria aiming to minimize variation in physiologic stress responses due to health conditions such as coronary heart disease. It is possible that selection criteria may have controlled for downstream consequences of the exposure, thereby attenuating associations of interest. These findings may only generalize to white individuals who have lower exposures to neighborhood stressors. It is possible that ethnic or racial minority groups or those with greater exposure to neighborhood stressors may exhibit different reactivity patterns. There were also several assumptions to our model specifications which may be incorrect. First, we visualized cortisol response as a curve with one inflection point located at post-task. Due to the timing of cortisol collection, the measurement immediately post-task was designated as the maximum; it is possible that peaks in cortisol occurred somewhere in between immediately post-task and 20 minutes post-task. There were also several strengths to our study including the use of a large well-characterized sample to evaluate associations. We were also able to minimize extraneous variation in cortisol response by evaluating associations within a group of similar age and race and by using a controlled physiologic stress measure that accounted for differences in stress perception.

In conclusion, these findings indicate that neighborhood may play an important role in HPA functioning among adults, especially among women. Addressing individual perceptions of social control may facilitate mental health promotion efforts in neighborhoods. These findings may therefore provide weak support of stress process models linking neighborhood and health. Further investigation is warranted to verify the robustness of these findings.

  • Associations between neighborhood and stress reactivity were evaluated among healthy adults.

  • Socioeconomic deprivation was associated with blunted cortisol reactivity among women.

  • Social control was associated with higher cortisol reactivity among women.

  • Improving social control may aid community mental health promotion efforts among women.

  • These findings provide weak support of stress process models linking neighborhoods and health.

Acknowledgments

We thank Bev Murray for compiling postcodes for linkage with IMD data.

FUNDING

Dr. Barrington was supported by the National Cancer Institute Biobehavioral Cancer Prevention and Control Training (BCPT) Program (R25 CA092408) and a Supplement to Promote Diversity in Health-Related Research through the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) (R01 DK0790420142) at the University of Washington. The Heart Scan Study was supported by the British Heart Foundation.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. BIRDITT KS, CICHY KE, ALMEIDA D. Age differences in Exposure and Reactivity to Interpersonal Tensions among Black and White Individuals across Adulthood. Race Soc Probl. 2011;3:225–239. doi: 10.1007/s12552-011-9058-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. BRENNER AB, ZIMMERMAN MA, BAUERMEISTER JA, CALDWELL CH. The Physiological Expression of Living in Disadvantaged Neighborhoods for Youth. J Youth Adolesc. 2012 doi: 10.1007/s10964-012-9838-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. BRENNER AB, ZIMMERMAN MA, BAUERMEISTER JA, CALDWELL CH. Neighborhood Context and Perceptions of Stress Over Time: An Ecological Model of Neighborhood Stressors and Intrapersonal and Interpersonal Resources. Am J Community Psychol. 2013 doi: 10.1007/s10464-013-9571-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. COHEN S, JANICKI-DEVERTS D, DOYLE WJ, MILLER GE, FRANK E, RABIN BS, RB T. Chronic stress, glucocorticoid receptor resistance, inflammation, and disease risk. Proc Natl Acad Sci USA. 2012;109:5995–5999. doi: 10.1073/pnas.1118355109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. COHEN S, JANICKI-DEVERTS D, MILLER GE. Psychological stress and disease. JAMA. 2007;298:1685–1687. doi: 10.1001/jama.298.14.1685. [DOI] [PubMed] [Google Scholar]
  6. DEPARTMENT FOR COMMUNITIES AND LOCAL GOVERNMENT. [on 8/9/2011];Using the English indices of deprivation. 2007 Accessed at www.communities.gov.uk/documents/communities/doc/615986.doc. [Online]
  7. DIEZ ROUX A, MAIR C. Neighborhoods and health. Ann NY Acad Sci. 2010;1186:125–145. doi: 10.1111/j.1749-6632.2009.05333.x. [DOI] [PubMed] [Google Scholar]
  8. DO D, DIEZ ROUX A, HAJAT A, AUCHINCLOSS A, MERKIN S, RANJIT N, SHEA S, SEEMAN T. Circadian rhythm of cortisol and neighborhood characteristics in a population-based sample: the Multi-Ethnic Study of Atherosclerosis. Health & Place. 2011;17:625–632. doi: 10.1016/j.healthplace.2010.12.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. EVERSON-ROSE SA, SKARUPSKI KA, BARNES LL, BECK T, EVANS DA, MENDES DE LEON CF. Neighborhood socioeconomic conditions are associated with psychosocial functioning in older black and white adults. Health Place. 2011;17:793–800. doi: 10.1016/j.healthplace.2011.02.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. FELDMAN P, STEPTOE A. How neighborhoods and physical functioning are related: The roles of neighborhood socioeconomic status, perceived neighborhood strain, and individual health risk factors. Ann Behav Med. 2004;27:91–99. doi: 10.1207/s15324796abm2702_3. [DOI] [PubMed] [Google Scholar]
  11. FRIES E, HESSE J, HELLHAMMER J, HELLHAMMER D. A new view on hypocortisolism. Psychonueroendocrinology. 2005;30:1010–1016. doi: 10.1016/j.psyneuen.2005.04.006. [DOI] [PubMed] [Google Scholar]
  12. GOLDMAN-MELLOR S, HAMER M, STEPTOE A. Early-life stress and recurrent psychological distress over the lifecourse predict divergent cortisol reactivity patterns in adulthood. Psychoneuroendocrinology. 2012;37:1755–68. doi: 10.1016/j.psyneuen.2012.03.010. [DOI] [PubMed] [Google Scholar]
  13. HACKMAN DA, BETANCOURT LM, BRODSKY NL, HURT H, FARAH MJ. Neighborhood disadvantage and adolescent stress reactivity. Front Hum Neurosci. 2012;6:277. doi: 10.3389/fnhum.2012.00277. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. HAMER M, GIBSON EL, VUONONVIRTA R, WILLIAMS E, STEPTOE A. Inflammatory and hemostatic responses to repeated mental stress: individual stability and habituation over time. Brain Behav Immun. 2006;20:456–459. doi: 10.1016/j.bbi.2006.01.001. [DOI] [PubMed] [Google Scholar]
  15. HAMER M, ENDRIGHI R, VENURAJU SM, LAHIRI A, STEPTOE A. Cortisol responses to mental stress and the progression of coronary artery calcification in healthy men and women. PLoS One. 2012;7:e31356. doi: 10.1371/journal.pone.0031356. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. HAMER M, O’DONNEL A, LAHIRI A, STEPTOE A. Salivary cortisol responses to mental stress are associated with coronary artery calcification in healthy men and women. Eur Heart J. 2010;31:424–429. doi: 10.1093/eurheartj/ehp386. [DOI] [PubMed] [Google Scholar]
  17. HELLHAMMER D, WADE S. Endocrine correlates of stress vulnerability. Psychother Psychosom. 1993;60:8–17. doi: 10.1159/000288675. [DOI] [PubMed] [Google Scholar]
  18. HILL TD, ROSS CE, ANGEL RJ. Neighborhood disorder, psychophysiological distress, and health. Journal of Health and Social Behavior. 2005;46:170–186. doi: 10.1177/002214650504600204. [DOI] [PubMed] [Google Scholar]
  19. KAEWTHUMMANUKUL T, BROWN K. Determinants of employee participation in physical activity: critical review of the literature. AAOHNJ. 2006;54:249–261. doi: 10.1177/216507990605400602. [DOI] [PubMed] [Google Scholar]
  20. KAPUKU GL, TREIBER FA, DAVIS HC. Relationships among socioeconomic status, stress induced changes in cortisol, and blood pressure in African American males. Ann Behav Med. 2002;24:320–5. doi: 10.1207/S15324796ABM2404_08. [DOI] [PubMed] [Google Scholar]
  21. KARB RA, ELLIOTT MR, DOWD JB, MORENOFF JD. Neighborhood-level stressors, social support, and diurnal patterns of cortisol: the Chicago Community Adult Health Study. Soc Sci Med. 2012;75:1038–47. doi: 10.1016/j.socscimed.2012.03.031. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. KIM D, DIEZ ROUX AV, KIEFE CI, KAWACHI I, LIU K. Do neighborhood socioeconomic deprivation and low social cohesion predict coronary calcification?” the CARDIA study. Am J Epidemiol. 2010;172:288–298. doi: 10.1093/aje/kwq098. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. KUDIELKA BM, BUSKE-KIRSCHBAUM A, HELLHAMMER DH, KIRSCHBAUM C. HPA axis responses to laboratory psychosocial stress in healthy elderly adults, younger adults, and children: impact of age and gender. Psychoneuroendocrinology. 2004;29:83–98. doi: 10.1016/s0306-4530(02)00146-4. [DOI] [PubMed] [Google Scholar]
  24. LANE J, GOVER A, DAHOD S. Fear of violent crime among men and women on campus: the impact of perceived risk and fear of sexual assault. Violence Vict. 2009;24:172–192. doi: 10.1891/0886-6708.24.2.172. [DOI] [PubMed] [Google Scholar]
  25. LATKIN CA, CURRY AD. Stressful neighborhoods and depression: a prospective study of the impact of neighborhood disorder. J Health Soc Behav. 2003;44:34–44. [PubMed] [Google Scholar]
  26. LLABRE MM, SPITZER SB, SAAB PG, SCHNEIDERMAN N. Piecewise latent growth curve modeling of systolic blood pressure reactivity and recovery from the cold pressor test. Psychophysiology. 2001;38:951–60. doi: 10.1111/1469-8986.3860951. [DOI] [PubMed] [Google Scholar]
  27. MCEWEN BS, SEEMAN T. Protective and damaging effects of mediators of stress. Elaborating and testing the concepts of allostasis and allostatic load. Ann N Y Acad Sci. 1999;896:30–47. doi: 10.1111/j.1749-6632.1999.tb08103.x. [DOI] [PubMed] [Google Scholar]
  28. MEIJER M, ROHL J, BLOOMFIELD K, GRITTNER U. Do neighborhoods affect individual mortality? A systematic review and meta-analysis of multilevel studies. Soc Sci Med. 2012;74:1204–1212. doi: 10.1016/j.socscimed.2011.11.034. [DOI] [PubMed] [Google Scholar]
  29. MILLER GE, CHEN E, ZHOU E. If it goes up, must it come down? Chronic stress and the Hypothalamic-Pituitary-Adrenocortical axis in humans. Psychological Bulletin. 2007;133:25–45. doi: 10.1037/0033-2909.133.1.25. [DOI] [PubMed] [Google Scholar]
  30. NAZMI A, DIEZ ROUX A, RANJIT N, SEEMAN TE, JENNY NS. Cross-sectional and longitudinal associations of neighborhood characteristics with inflammatory markers: findings from the multi-ethnic study of atherosclerosis. Health Place. 2010;16:1104–12. doi: 10.1016/j.healthplace.2010.07.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. PEARLIN LI, LIEBERMAN MA, MENAGHAN EG, MULLAN JT. The stress process. J Health Soc Behav. 1981;22:337–56. [PubMed] [Google Scholar]
  32. PECKINS MK, DOCKRAY S, ECKENRODE JL, HEATON J, SUSMAN EJ. The longitudinal impact of exposure to violence on cortisol reactivity in adolescents. J Adolesc Health. 2012;51:366–72. doi: 10.1016/j.jadohealth.2012.01.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. PHILLIPS AC, GINTY AT, HUGHES BM. The other side of the coin: Blunted cardiovascular and cortisol reactivity are associated with negative health outcomes. International Journal of Psychophysiology. 2013;90(1):1–7. doi: 10.1016/j.ijpsycho.2013.02.002. [DOI] [PubMed] [Google Scholar]
  34. ROE JJ, THOMPSON CW, ASPINALL PA, BREWER MJ, DUFF EI, MILLER D, MITCHELL R, CLOW A. Green space and stress: evidence from cortisol measures in deprived urban communities. Int J Environ Res Public Health. 2013;10:4086–103. doi: 10.3390/ijerph10094086. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. ROSS CE, MIROWSKY J. Neighborhood disadvantage, disorder, and health. J Health Soc Behav. 2001;42:258–76. [PubMed] [Google Scholar]
  36. ROY MP. Patterns of cortisol reactivity to laboratory stress. Hormones and Behavior. 2004;46:618–627. doi: 10.1016/j.yhbeh.2004.06.015. [DOI] [PubMed] [Google Scholar]
  37. ROY MP, STEPTOE A, KIRSCHBAUM C. Life events and social support as moderators of individual differences in cardiovascular and cortisol reactivity. J Pers Soc Psychol. 1998;75:1273–81. doi: 10.1037//0022-3514.75.5.1273. [DOI] [PubMed] [Google Scholar]
  38. RUDOLPH K, WAND G, STUART E, GLASS T, MARQUES A, DUNCKO R, MERIKANGAS K. THE ASSOCIATION BETWEEN CORTISOL CHARACTERISTICS AND NEIGHBORHOOD DISADVANTAGE IN A US POPULATION-BASED SAMPLE OF ADOLESCENTS. American Journal of Epidemiology. 2013;177:S32–S32. doi: 10.1016/j.healthplace.2013.11.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. SAMPSON R, RAUDENBUSH S, EARLS F. Neighborhoods and violent crime: A multilevel study of collective efficacy. Science. 1997;277:918–924. doi: 10.1126/science.277.5328.918. [DOI] [PubMed] [Google Scholar]
  40. SCHULZ AJ, MENTZ G, LACHANCE L, ZENK SN, JOHNSON J, STOKES C, MANDELL R. Do observed or perceived characteristics of the neighborhood environment mediate associations between neighborhood poverty and cumulative biological risk? Health Place. 2013;24C:147–156. doi: 10.1016/j.healthplace.2013.09.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. SILVERMAN MN, STERNBERG EM. Glucocorticoid regulation of inflammation and its functional correlates: from HPA axis to glucocorticoid receptor dysfunction. Ann NY Acad Sci. 2012;126:55–63. doi: 10.1111/j.1749-6632.2012.06633.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. STAFFORD M, CHANDOLA T, MARMOT M. Association between fear of crime and mental health and physical functioning. American Journal of Public Health. 2007;97:2076–2081. doi: 10.2105/AJPH.2006.097154. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. STEPTOE A, FELDMAN P. Neighborhood problems as sources of chronic stress: development of a measure of neighborhood problems, and association with socioeconomic status and health. Ann Behav Med. 2001;23:177–185. doi: 10.1207/S15324796ABM2303_5. [DOI] [PubMed] [Google Scholar]
  44. STEPTOE A, HAMER M, O’DONNELL K, VENURAJU S, MARMOT MG, LAHIRI A. Socioeconomic Status and Subclinical Coronary Disease in the Whitehall II Epidemiological Study. PLoS One. 2010;5:e8874. doi: 10.1371/journal.pone.0008874. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. TRUONG KD, MA S. A systematic review of relations between neighborhoods and mental health. J Ment Health Policy Econ. 2006;9:137–54. [PubMed] [Google Scholar]
  46. UK STATISTICS AUTHORITY. [on 8/9/2011];Index of multiple deprivation 2007: Measure of multiple deprivation at small area level made up of seven domains. Accessed at http://data.gov.uk/dataset/index_of_multiple_deprivation_imd_2007. [Online]. Communities and Local Government (CLG): ID 2007 neighborhood statistics.
  47. VAN ECK M, NICOLSON NA, BERKHOF JS. Individual differences in cortisol response to a laboratory speech task and their relationship to responses to stressful daily events. Biological Psychology. 1996;43:69–84. doi: 10.1016/0301-0511(95)05159-7. [DOI] [PubMed] [Google Scholar]
  48. WILCOX S, BOPP M, WILSON DK, FULK LJ, HAND GA. Race differences in cardiovascular and cortisol responses to an interpersonal challenge in women who are family caregivers. Ethn Dis. 2005;15:17–24. [PubMed] [Google Scholar]

RESOURCES