Background: Psychosocial stressors increase the risks for cardiovascular disease across diverse populations. However, neighborhood level resilience resources may protect against poor cardiovascular health (CVH). This study used data from three CVH cohorts to examine longitudinally the associations of a resilience resource, perceived neighborhood social cohesion (hereafter referred to as neighborhood social cohesion), with the American Heart Association's Life's Simple 7 (LS7), and whether psychosocial stressors modify observed relationships. Methods: We examined neighborhood social cohesion (measured in tertiles) and LS7 in the Jackson Heart Study, Multi-Ethnic Study of Atherosclerosis, and Mediators of Atherosclerosis in South Asians Living in America study. We used repeated-measures, modified Poisson regression models to estimate the relationship between neighborhood social cohesion and LS7 (primary analysis, n = 6,086) and four biological metrics (body mass index, blood pressure, cholesterol, blood glucose; secondary analysis, n = 7,291). We assessed effect measure modification by each psychosocial stressor (e.g., low educational attainment, discrimination). Results: In primary analyses, adjusted prevalence ratios (aPR) and 95% confidence intervals (CIs) for ideal/intermediate versus poor CVH among high or medium (versus low) neighborhood social cohesion were 1.01 (0.97–1.05) and 1.02 (0.98–1.06), respectively. The psychosocial stressors, low education and discrimination, functioned as effect modifiers. Secondary analyses showed similar findings. Also, in the secondary analyses, there was evidence for effect modification by income. Conclusion: We did not find much support for an association between neighborhood social cohesion and LS7, but did find evidence of effect modification. Some of the effect modification results operated in unexpected directions. Future studies should examine neighborhood social cohesion more comprehensively and assess for effect modification by psychosocial stressors.
Keywords: Neighborhood; Resilience; Psychosocial factors; Life's simple 7; Cardiovascular health
Supplementary Information The online version contains supplementary material available at https://doi.org/10.1186/s12889-022-14270-x.
Cardiovascular disease (CVD) is the leading cause of death globally and it is well-established that behavioral and physiological health factors and psychosocial stressors increase the risks for CVD across diverse populations (e.g., racially, ethnically, gender, geographically) [[
There are several direct and indirect pathways through which psychosocial stressors increase CVD risk. For instance, exposure to acute and chronic psychosocial stressors may lead to poor coping responses such as smoking, poor diet quality and physical inactivity [[
Resilience resources may protect against poor CVH outcomes via engagement in more positive health behaviors and reduction in the risks for poorer physiological functioning [[
Specifically, some of the gaps in resilience research relate to limited generalizability, analytical approach, time frame and the level of resilience resource examined. Much of the research is conducted in populations with diagnosed CVD or does not include racially, ethnically, socioeconomically or geographically diverse populations [[
Although some research examines neighborhood-level resilience (e.g., social cohesion), findings are mixed. Neighborhood social cohesion is a multicomponent concept that includes the extent to which neighborhoods [[
We used a harmonized dataset from three cohort studies in the United States – Jackson Heart Study (JHS), Multi-Ethnic Study of Atherosclerosis (MESA), and Mediators of Atherosclerosis in South Asians Living in America (MASALA). JHS is a study with 5,306 African American adults over the age of 21 and consists of three exams approximately every 4 years in addition to annual follow-up interviews every 12 months. MESA is a study among 6,814 racially/ethnically diverse adults over the age of 45 without a history of CVD at study enrollment. We used data from the first five exams which span 10 years. MASALA is a study of 906 South Asian adults over the age of 40 without CVD history at study enrollment. MASALA consists of two exams. Detailed description of the cohort study designs are detailed elsewhere [[
The exposure is based on participants' reports of neighborhood social cohesion at MESA/MASALA Exam 1 and JHS Third Annual Follow-up Interview (Cronbach's alpha = 0.73 in the harmonized dataset). This is a five-item measure assessing close-knit neighborhood, willingness to help, getting along, trust, and sharing the same values [[
The primary outcome is a binary variable of ideal and intermediate versus poor CVH defined by the composite LS7 metrics. The LS7 is based on self/proxy-reported and/or physical exams. A score of 0, 1, and 2 for poor, intermediate, and ideal, respectively, is assigned for each metric and summed across all of them (range: 0–14). Scores of 8–14 indicate ideal or intermediate CVH. Details on each LS7 metric are described elsewhere [[
Potential confounding variables include age, sex/gender, race, ethnicity, geographical region, nativity, marital status, self-rated health, health insurance type, self- and family-history of CVD and social support [[
Potential effect modifiers include the following psychosocial stressors - self-reported anger [[
Supplemental Fig. 1 shows the relationship between all variables included in the study in a causal directed acyclic graph. We selected variables as covariates if they were considered to be sources of confounding or selection bias or potential effect modifiers based on prior literature [[
We excluded participants who did not have available data for neighborhood social cohesion, potential confounders or sources of selection bias, potential effect modifiers, and outcomes at relevant time points. We used Chi-squared and Wilcoxon Mann-Whitney tests to compare Exam 1 characteristics between the included and excluded participants. Because the follow-up time between exams differed within and across cohorts, we constructed two equal bins of follow-up time that corresponded to 6-year intervals. If participants had multiple outcome assessments within an interval, we used the furthest observation in analysis. Outcomes assessed concurrently with the exposure were included. Hereafter, 'visit' refers to a given 6-year interval and 'exam' refers to the cohort exams. We treated death (9.3% and 7.7% in the primary and secondary analysis sample, respectively) as a censoring event and not an event that created undefined CVH outcomes [[
In primary analyses, we used repeated-measures, modified Poisson regression models to examine the overall relationship between neighborhood social cohesion and CVH. Unadjusted and adjusted models included neighborhood social cohesion, visit, and product terms between neighborhood social cohesion and visit as independent variables. Also, we included potential confounders and effect modifiers in the adjusted models to account for confounding and selection bias. The abovementioned primary analyses were repeated excluding the neighborhood social cohesion and visit product terms. To assess for effect measure modification by psychosocial stressors, we modified the adjusted models to include a product term between neighborhood social cohesion, visit and the psychosocial stressor of interest. When we assessed for effect measure modification by one psychosocial stressor measure at a time, we included the other psychosocial stressor measures in the outcome models to adjust for confounding and selection bias. Global chi-squared tests indicated whether at least one of the relevant product term coefficients was different from zero. Secondary analyses repeated the primary analyses but assessed for the two secondary CVH outcomes separately.
In our primary and secondary analyses, the continuous variable (age) was fitted using restricted quadratic splines with 4 knots at unequal intervals (i.e., 5th, 35th, 65th, 95th percentiles) [[
Interpretation of our findings aligns with recent literature on significance and hypothesis testing [[
The primary analysis included 6,086 participants (Supplemental Fig. 2). Table 1 highlights the characteristics among the included and excluded participants. The included participants had lower perceived neighborhood social cohesion than excluded participants. Specifically, for the included participants, 25.5% were in the high neighborhood social cohesion group, 40.3% in medium and 34.1% in low. The median (25th percentile-75th percentile) age was higher among included [61 (53–69)] than excluded participants [53 (44–62)]. Most included participants were female (51.7%), White, non-Hispanic (37.6%), born in the US (64.8%), and lived in the Midwest (38.2%). Compared to excluded participants, included participants were more likely to be married (64.2% versus 59.6%), report good health (91.1% versus 75.9%), have public or private health insurance (91.6% versus 87.6%) but less likely to be employed (50.7% versus 61.5%) and have a high school education or some college degree (44.0% versus 46.1%). Income levels were similar between the included and excluded participants. Further, the included participants, compared to the excluded, reported fewer depressive symptoms (87.8% versus 79.2%), included participants also were more likely to report that their neighborhood was safe (85.3% versus 59.9%).
Table 1 Characteristics of MASALA and MESA participants at the exam concurrent with neighborhood social cohesion assessment
Characteristics Included (n = 6,086) Excluded (n = 1,661) P-value* N % N % Low 2,078 34.1 382 23.0 < 0.01 Medium 2,455 40.3 502 30.2 High 1,553 25.5 777 46.8 61 (53–69) 53 (44–62) < 0.01 Female 3,145 51.7 1,066 64.2 < 0.01 Male 2,941 48.3 595 35.8 White non-Hispanic 2,288 37.6 0 0 < 0.01 Asian 1,081 17.8 15 0.9 African American 1,460 24.0 1,646 99.1 Hispanic 1,257 20.7 0 0 Other 2,145 35.2 15 0.9 < 0.01 U.S.-born 3,941 64.8 1,646 99.1 West 1,126 18.5 0 0 < 0.01 South 870 14.3 1,646 99.1 Midwest 2,325 38.2 15 0.9 Northeast 1,765 29.0 0 0 Never married, separated/divorced, widowed 2,178 35.8 671 40.4 < 0.01 Married 3,908 64.2 990 59.6 Not good 542 8.9 401 24.1 < 0.01 Good 5,544 91.1 1,260 75.9 None 509 8.4 206 12.4 < 0.01 Public or Private 5,577 91.6 1,455 87.6 No 2,676 44.0 692 41.7 0.09 Yes 3,410 56.0 969 58.3 Less than high school 949 15.6 156 9.4 < 0.01 High school or some college 2,675 44.0 766 46.1 College degree or more 2,462 40.5 739 44.5 Unemployed 3,002 49.3 639 38.5 < 0.01 Employed (Part/full-time) 3,084 50.7 1,022 61.5 $0-$19,999 1,299 21.3 364 21.9 0.88 $20,000-$49,999 2,181 35.8 593 35.7 $50,000+ 2,606 42.8 704 42.4 Low 2,410 39.6 499 30.0 < 0.01 Medium 2,053 33.7 433 26.1 High 1,623 26.7 729 43.9 No 5,344 87.8 1,315 79.2 < 0.01 Yes 742 12.2 346 20.8 Low 3,064 50.4 280 16.9 < 0.01 Medium 1,960 32.2 650 39.1 High 1,062 17.5 731 44.0 Low 2,431 39.9 365 22.0 < 0.01 Medium 2,076 34.1 500 30.1 High 1,579 25.9 796 47.9 Low 1,471 24.2 822 49.5 < 0.01 Medium 2,186 35.9 495 29.8 High 2,429 39.9 344 20.7 Safe 5,193 85.3 995 59.9 < 0.01 Not safe 893 14.7 666 40.1 Not high 2,322 38.2 425 25.6 < 0.01 High 3,764 61.9 1,236 74.4
Table 2 shows adjusted prevalence ratios (aPR) for ideal/intermediate versus poor CVH. Focusing on the overall findings across visits that were most compatible with the data, high and medium (versus low) neighborhood social cohesion was not associated with ideal/intermediate CVH (aPR: 1.01, 95% CI: 0.97–1.05, and aPR: 1.02, 95% CI: 0.98–1.06). The corresponding 95% CIs indicate that a weak positive association and a weak negative association was compatible with the data. Results by visit yielded similar findings. There was some evidence for effect measure modification by education and discrimination levels (Table 3). For instance, high (versus low) neighborhood social cohesion was negatively associated with ideal/intermediate (versus poor) CVH among participants with less than high school education (aPR: 0.77, 95% CI: 0.67–0.88). However, a positive association was most compatible with the data among those who had high school or some college (aPR: 1.05, 95% CI: 0.98–1.12) or college degree or more (aPR: 1.05, 95% CI: 1.00-1.10). Findings by visit were similar (Supplemental Table 6).
Table 2 Prevalence ratios (PR) for CVH outcomes
Outcome Neighborhood social cohesion and visit product term in outcome model High versus low neighborhood social cohesion PR (95% CI) Medium versus low neighborhood social cohesion PR (95% CI) Ideal or intermediate versus poor CVH Without product term 1.06 (1.01–1.10) 1.01 (0.97–1.05) 1.07 (1.03–1.11) 1.02 (0.98–1.06) With product term‡ Visit 1 1.06 (1.01–1.10) 1.01 (0.96–1.05) 1.08 (1.01–1.12) 1.02 (0.98–1.06) Visit 2 1.06 (0.99–1.12) 1.02 (0.96–1.08) 1.06 (1.03–1.12) 1.02 (0.96–1.07) Ideal or intermediate (but no poor) metrics versus 1 or more poor metrics Without product term 1.00 (0.93–1.08) 1.01 (0.95–1.09) 1.06 (1.00-1.13) 1.01 (0.95–1.07) With product term§ Visit 1 0.99 (0.92–1.07) 1.02 (0.95–1.09) 1.04 (0.98–1.11) 0.99 (0.93–1.05) Visit 2 1.03 (0.94–1.12) 1.00 (0.92–1.10) 1.09 (1.01–1.19) 1.03 (0.95–1.12) Lower cardiovascular risk (0–1 poor metrics) versus non-lower cardiovascular risk (2–4 poor metrics) Without product term 0.99 (0.97–1.02) 0.99 (0.97–1.02) 1.03 (1.00-1.05) 1.01 (0.98–1.03) With product term¶ Visit 1 0.99 (0.96–1.02) 0.99 (0.96–1.03) 1.03 (1.01–1.06) 1.01 (0.99–1.04) Visit 2 1.01 (0.97–1.04) 0.99 (0.96–1.03) 1.01 (0.98–1.04) 0.99 (0.96–1.02)
Note: Each modified Poisson regression model accounted for clustering within neighborhood (i.e., census tract at Exam 1) [[
Table 3 Assessment of effect measure modification of adjusted prevalence ratios
Psychosocial risk measure (Potential effect measure modifier) High versus low neighborhood social cohesion and ideal or intermediate versus poor CVH by level of psychosocial risk measure Medium versus low neighborhood social cohesion and ideal or intermediate versus poor CVH by level of psychosocial risk measure p† College degree or more 1.05 (1.00-1.10) 1.04 (0.99–1.09) < 0.01 High school or some college 1.05 (0.98–1.12) 1.03 (0.96–1.09) Less than high school 0.77 (0.67–0.88) 0.98 (0.89–1.08) Employed 1.02 (0.96–1.08) 1.02 (0.96–1.08) 0.81 Unemployed 1.00 (0.95–1.06) 1.02 (0.97–1.08) $50,000+ 1.03 (0.98–1.08) 1.01 (0.96–1.07) 0.33 $20,000-$49,999 0.99 (0.92–1.06) 1.01 (0.95–1.07) $0-$19,999 0.98 (0.88–1.09) 1.07 (0.99–1.15) Low 1.03 (0.97–1.09) 1.04 (0.98–1.10) 0.51 Medium 1.01 (0.96–1.07) 0.99 (0.94–1.05) High 0.98 (0.90–1.06) 1.03 (0.96–1.11) No 1.02 (0.97–1.06) 1.02 (0.98–1.06) 0.71 Yes 0.96 (0.84–1.10) 1.03 (0.93–1.14) Low 1.02 (0.96–1.07) 1.04 (0.99–1.08) 0.88 Medium 1.00 (0.93–1.07) 1.00 (0.93–1.07) High 1.03 (0.94–1.13) 1.01 (0.92–1.11) Low 0.96 (0.91–1.02) 0.99 (0.94–1.04) 0.12 Medium 1.02 (0.96–1.09) 1.02 (0.96–1.07) High 1.08 (1.01–1.15) 1.08 (1.01–1.15) Low 0.99 (0.89–1.09) 1.01 (0.93–1.09) 0.67 Medium 1.00 (0.93–1.08) 1.04 (0.97–1.11) High 1.03 (0.98–1.08) 1.02 (0.97–1.06) Safe 1.00 (0.96–1.04) 1.01 (0.97–1.06) 0.35 Not safe 1.09 (0.98–1.21) 1.05 (0.97–1.15)
Note: Each modified Poisson regression model accounted for clustering within neighborhood (i.e., census tract at Exam 1) [[
In total, 7,291 participants were included in the secondary analysis (Supplemental Fig. 3). Characteristics of these participants are shown in Supplemental Tables 1, and were similar to the primary analysis with slightly different distributions.
The aPRs (95% CIs) for ideal/intermediate (but no poor) metrics among high and medium (versus low) neighborhood social cohesion were 1.01 (0.95–1.09) and 1.01 (0.95–1.07), respectively, suggesting a null relationship based on the most compatible estimate (Table 2). Similar findings emerged when examining 0–1 versus 2–4 poor LS7 metrics (aPR: 0.99, 95% CI: 0.97–1.02, and aPR: 1.01, 95% CI: 0.98–1.03, for high and medium neighborhood social cohesion respectively). Findings by visit for both secondary outcomes were similar. When examining ideal/intermediate (but no poor) metrics (Supplemental Table 2), there was evidence for effect modification by psychosocial stressors such as lower education, lower income, and greater discrimination. Similarly, when examining 0–1 poor versus 2–4 poor metrics, there was evidence for effect modification by income (Supplemental Table 3). Findings by visit were similar (Supplemental Tables 7–8).
Again, focusing on the estimates most compatible with the data, inferences based on the primary and secondary analysis did not change when we solely accounted for outcomes correlated within individuals or models were specified with an exchangeable working correlation structure. Cohort-stratified findings for primary analyses showed that the aPRs for high (versus low) neighborhood social cohesion were consistently higher in MASALA compared to MESA based on the most compatible estimates (Supplemental Table 4). The aPRs for medium versus low neighborhood social cohesion were similar across cohorts, except for MASALA at visit 2. Cohort-stratified findings for the secondary analysis examining ideal/intermediate (but no poor metrics) showed similar aPRs for JHS and MASALA, while MESA had consistently null findings for high versus low neighborhood social cohesion. When comparing medium versus low cohesion, aPRs differed across all cohorts. For 0–1 poor metrics versus 2–4 poor metrics, aPRs for JHS were consistently lower than those in MESA and MASALA (Supplemental Table 5).
This study addresses some of the gaps in the literature by examining the role of a neighborhood-level resilience resource, perceived neighborhood social cohesion, on CVH outcomes over time and in the presence of adversities (i.e., psychosocial risks) among racially, ethnically and geographically diverse participants. We hypothesized that perceived neighborhood social cohesion (e.g., via interactions with neighborhood residents, social support, collective action, shared norms) would be a resilience resource that results in better CVH. However, analyses revealed little support for relationships between neighborhood social cohesion and LS7. In contrast, there was support for effect modification by some of the psychosocial stressors. We expected that individuals experiencing more adversity would benefit more from mobilizing resilience resources, but this direction was not supported across all psychosocial stressors that emerged as effect modifiers. For instance, higher levels of neighborhood social cohesion emerged as a protective factor for those with high school or more education but a risk factor for those with less than a high school education.
The current findings add to the growing body of literature finding mixed support for neighborhood social cohesion on CVH in both cross-sectional and longitudinal studies. In our study, neighborhood social cohesion was not associated with ideal/intermediate LS7 over time. Unger and colleagues did not find associations between higher neighborhood social cohesion and ideal CVH in their cross-sectional study with MESA participants [[
Although we did not find support for the role of neighborhood social cohesion as a resilience resource on LS7 over time, cross-sectional study findings suggest that there are relationships with LS7 components. Research with Asian American subgroups suggests that neighborhood social cohesion is associated with engaging in some CVH screening behaviors across genders, smoking among men, and with lower odds of hypertension and high BMI among women.[[
One reason for the equivocal findings may be that researchers conceptualize and measure neighborhood social cohesion in different ways [[
Although informative, this study is not without some limitations. The neighborhood social cohesion measure, while used extensively in prior studies, is limited in scope because it does not fully capture the social inclusion and social capital components of neighborhood social cohesion. While we assessed neighborhood social cohesion at the individual level and as tertiles, both of which are common, tertiles are sample dependent and may not truly reflect levels of high, medium or low in the general population [[
In conclusion, this study did not find much evidence of an overall relationship between neighborhood social cohesion and CVH. We also found mixed support for some psychosocial stressors as effect modifiers. Future studies should examine neighborhood social cohesion more comprehensively, assess it over longer periods of time and in socioeconomically diverse and nationally representative cohort studies. As well, future studies should assess for effect modification and include and examine composite LS7 over time, and only include prospective LS7 measures.
The authors thank the other investigators, the staff, and the participants of the JHS, MESA, and MASALA studies for their valuable contributions.
Akilah J. Dulin: Conceptualization, methodology, supervision, writing – original draft, Jee Won Park: Methodology, formal analysis, writing – original draft, Matthew M Scarpaci: Data curation, formal analysis, writing – review and editing, Laura A Dionne: Data curation, writing – review and editing, Belinda L. Needham: Supervision, investigation, writing – review and editing, Mario Sims: Supervision, investigation, writing – review and editing, Joseph L. Fava: Supervision, writing – review and editing, Charles B. Eaton: Supervision, writing – review and editing,, funding acquisition, project administration, writing – review and editing, Alka Kanaya: Investigation, writing – review and editing, Namratha R. Kandula: Investigation, writing – review and editing, Eric B. Loucks: Conceptualization, writing – review and editing, Chanelle J. Howe: conceptualization, methodology, writing – original draft, supervision, funding acquisition.
Research reported in this publication was supported by the National Heart, Lung, And Blood Institute of the National Institutes of Health under Award Number R01HL135200. 100% of the project costs are financed with Federal money. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
The Jackson Heart Study (JHS) is supported and conducted in collaboration with Jackson State University (HHSN268201800013I), Tougaloo College (HHSN268201800014I), the Mississippi State Department of Health (HHSN268201800015I) and the University of Mississippi Medical Center (HHSN268201800010I, HHSN268201800011I and HHSN268201800012I) contracts from the National Heart, Lung, and Blood Institute (NHLBI) and the National Institute on Minority Health and Health Disparities (NIMHD).
The Multi-Ethnic Study of Atherosclerosis (MESA) study was supported by contracts 75N92020D00001, HHSN268201500003I, N01-HC-95159, 75N92020D00005, N01-HC-95160, 75N92020D00002, N01-HC-95161, 75N92020D00003, N01-HC-95162, 75N92020D00006, N01-HC-95163, 75N92020D00004, N01-HC-95164, 75N92020D00007, N01-HC-95165, N01-HC-95166, N01-HC-95167, N01-HC-95168 and N01-HC-95169 from the National Heart, Lung, and Blood Institute, and by grants UL1-TR-000040, UL1-TR-001079, and UL1-TR-001420 from the National Center for Advancing Translational Sciences (NCATS). A full list of participating MESA investigators and institutions can be found at
The Mediators of Atherosclerosis in South Asians Living in America (MASALA) project described was supported by Grant Number R01HL093009 from the National Heart, Lung, And Blood Institute and the National Center for Research Resources and the National Center for Advancing Translational Sciences, National Institutes of Health, through UCSF-CTSI Grant Number UL1RR024131.
The views expressed in this manuscript are those of the authors and do not necessarily represent the views of the National Heart, Lung, and Blood Institute; the National Institutes of Health; or the U.S. Department of Health and Human Services.
Data may be obtained from a third party and are not publicly available.
The datasets supporting the conclusions of this article originate from three separate third party sources. Restrictions apply to the availability of the data, which were used under fully executed Data Materials Distribution Agreement and Data Use Agreements. Data from the Jackson Heart Study is available with a request to access at https://
The Institutional Review Boards (IRBs) at Jackson State University, Tougaloo College, and the University of Mississippi Medical Center in Jackson, Mississippi approved the Jackson Heart Study. The IRBs at Columbia University, Johns Hopkins University, Northwestern University, UCLA, University of Minnesota and Wake Forest University approved the Multi-Ethnic Study of Atherosclerosis. The IRBs at the University of California San Francisco and Northwestern University approved the Mediators of Atherosclerosis in South Asians Living in America study. All study participants provided written informed consent. Additionally, the Brown University IRB approved our secondary analysis study. All methods were performed in accordance with the relevant guidelines and regulations.
Not applicable.
The authors declare that they have no competing interests.
• aPRs
- adjusted prevalence ratios.
• BMI
- body mass index.
• CVH
- cardiovascular health.
• CVD
- cardiovascular disease.
• HRS
- Health and Retirement Study.
• IRB
- Institutional Review Board.
• JHS
- Jackson Heart Study.
• LS7
- Life's Simple 7.
• MASALA
- Mediators of Atherosclerosis in South Asians Living in America.
• MESA
- Multi-Ethnic study of Atherosclerosis.
• US
- United States.
Below is the link to the electronic supplementary material.
Graph: Supplementary Material 1
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
By Akilah J. Dulin; Jee Won Park; Matthew M. Scarpaci; Laura A. Dionne; Mario Sims; Belinda L. Needham; Joseph L. Fava; Charles B. Eaton; Alka M. Kanaya; Namratha R. Kandula; Eric B. Loucks and Chanelle J. Howe
Reported by Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author