The Role of Social Support in Moderating the Relationship between Race and Hypertension in a Low-Income, Urban, Racially Integrated Community

Abstract

In the US, African Americans have a higher prevalence of hypertension than Whites. Previous studies show that social support contributes to the racial differences in hypertension but are limited in accounting for the social and environmental effects of racial residential segregation. We examined whether the association between race and hypertension varies by the level of social support among African Americans and Whites living in similar social and environmental conditions, specifically an urban, low-income, racially integrated community. Using data from the Exploring Health Disparities in Integrated Communities-Southwest Baltimore (EHDIC-SWB) sample, we hypothesized that social support moderates the relationship between race and hypertension and the racial difference in hypertension is smaller as the level of social support increases. Hypertension was defined as having systolic blood pressure greater than 140 mmHg and/or diastolic blood pressure greater than 90 mmHg, or the participant reports of taking antihypertensive medication(s). The study only included participants that self-reported as “Black/African American” or “White.” Social support was measured as functional social support and marital status. After adjusting for demographics and health-related characteristics, we found no interaction between social support and race (DUFSS score, prevalence ratio 1.00; 95% confidence interval 0.99, 1.01; marital status, prevalence ratio 1.02; 95% confidence interval 0.86, 1.21); thus the hypothesis was not supported. A plausible explanation is that the buffering factor of social support cannot overcome the social and environmental conditions which the participants live in. Further, these findings emphasize social and environmental conditions of participants in EHDIC-SWB may equally impact race and hypertension.

Background

Hypertension is a major public health problem that disproportionately affects African Americans in the US. As of 2017, about 34% of adults in the US have hypertension [1]. When stratified by race and gender, 45.0% and 46.3% of Black males and females have hypertension, compared to 34.5% and 32.3% of White males and females [1]. African Americans are also more at risk of dying from stroke, heart disease, and other complications attributable to hypertension [1–3]. To gain a better understanding of racial differences in hypertension, there has been a burgeoning body of research that has moved beyond just understanding individual level factors to focusing on broader social factors particularly racial residential segregation that constrain individual choices and behaviors.

De jure segregation has been outlawed in the US, but the majority of Americans still live in racially segregated communities. Racial minorities are more likely to live in neighborhoods of lower socioeconomic status (SES) and have less accessible grocery stores, health facilities, and green spaces; accessibility to such resources influences health behaviors and risk of chronic diseases, particularly hypertension, than Whites [4–6]. Studies have found mixed results about the association between racial residential segregation and hypertension [7–10]. In Kershaw and colleagues’ study, Black-White differences in hypertension significantly varies by geographic US regions, with greater prevalence and racial differences in southern states and in cities like Baltimore, Maryland, Forsyth County, North Carolina, and New York, New York [7]. In a NYC-based sample, however, another study found that segregation was not associated with hypertension among US-born and younger foreign-born Blacks, but segregation may serve as a protective factor for older foreign-born Blacks [8]. Because of racial residential segregation, Americans are exposed to different social and environmental health risks.

Racial residential segregation also exacerbates poverty, a neighborhood- and individual-level risk factor of hypertension [9, 10]. Differences in prevalence of hypertension between Whites and Blacks varied by level of racial residential segregation and poverty and were positively associated with higher percentages of Blacks living in high-poverty neighborhoods [9]. In Usher and colleagues’ study, Blacks had higher odds of hypertension compared to Whites regardless of neighborhood and individual poverty statuses than Whites, but Whites in poor neighborhoods had higher odds of hypertension than Whites in nonpoor neighborhoods [10]. These studies underscore that the social and environmental context which African Americans and Whites live in must be considered in studying racial differences in hypertension.

National surveys like the National Health Interview Survey, National Health and Nutrition Examination Survey, the Medical Expenditure Panel Survey, and the Behavioral Risk Factor Surveillance System do not account for the social and environmental conditions that resulted from racial residential segregation [4]. As noted in LaVeist’s study [4], race and SES are confounded. Accounting for SES is not enough to capture the heterogeneity of exposure to various socio-environmental risks [4]. Not addressing the effects of racial residential segregation has the potential to lead to inaccurate conclusions about hypertension [4, 11, 12].

Some studies that have accounted for the lingering effects of racial residential segregation show that the racial differences in health outcomes including hypertension are minimized [4, 11, 12]. Thorpe and colleagues [11] studied the relationship between race and hypertension and accounted for the social and environmental conditions in which people live in by using the Exploring Health Disparities in Integrated Communities-Southwest Baltimore (EHDIC-SWB) sample. The EHDIC-SWB sample is from a low-income, racially-integrated community in Southwest Baltimore. This study also compared hypertension prevalence results reported in the 1999–2004 National Health and Nutrition Examination Survey (NHANES) and found that the racial difference in hypertension was narrower in EHDIC-SWB than in NHANES [11]. Even after controlling by design for social and environmental conditions in EHDIC-SWB, racial differences in hypertension persist. The remaining observed difference might be due to unmeasured psychosocial factors [11]. Psychosocial factors may shed light on how African Americans and Whites respond differently to stressors that may subsequently impact hypertension prevalence [11].

One understudied psychosocial variable that may contribute to the racial differences in hypertension is social support. According to Broadhead and colleagues, [13] social support includes “the social functions as a confidant relationship, positive affective expressions [affective support], affirmation or praise, assistance with material, cognitive, or task performance, or ‘tangible assistance’ [instrumental support].” A study suggests that African Americans living in some racially segregated contexts may experience protective factors because of reduced discrimination and increased social benefits [14]. Social support has also been found to buffer the relationship between depression and hypertension among a sample of African Americans [15]. However, these studies do not provide information on the racial differences in social support and hypertension because White individuals were not sampled for these studies. Other studies either focused on the relationship between social support and hypertension or race and social support, but few focused on the interrelationship between social support, race, and hypertension [14–20].

Bell and colleagues [16], using marital status as a proxy for social support, examined whether social support moderated the relationship between race and hypertension in the 2001–2006 NHANES dataset. This study found that racial differences in the prevalence of hypertension were significantly reduced among all participants when the interaction of social support was included [16]. When the results were stratified by race, Bell and colleagues found that among African Americans, levels of social support significantly affect the prevalence of hypertension, with higher levels of social support associated with lower prevalence of hypertension. However, there was no relationship between levels of social support and hypertension among Whites. Because this study used a national survey, conclusions about social support, race, and hypertension did not account for the social and environmental conditions that may result from racial residential segregation.

To compare the role of social support among Whites and African Americans, Strogatz and colleagues [20] studied a biracial sample in a rural community in North Carolina. After controlling for various factors, social support was more strongly associated with the prevalence of hypertension among the African American residents than with White residents. Although this study sampled Whites and African Americans living in the same community, the majority of the African Americans were of lower SES [20]. Because SES and race confound each other, this study’s conclusion about social support, race, and hypertension did not fully account for social and environmental conditions African Americans and Whites lived in.

The objective is to examine the role of social support in the relationship between race and hypertension among African Americans and Whites living in a racially integrated, low-income, urban community. Because of social support’s possible protective effect against hypertension, it is hypothesized that social support moderates the relationship between race and hypertension and the racial difference in hypertension would be smaller as the level of social support increases among participants in the EHDIC-SWB sample.

Methods

Study Design

EHDIC-SWB is a cross-sectional face-to-face survey of the adult population (age 18 and older) of two contiguous census tracts. In addition to being economically homogenous, the study site was also racially balanced and well-integrated, with almost equal proportions of African American and White residents. EHDIC-SWB defined an integrated community as one that had at least 35% of African American and at least 35% White residents, had a ratio of Black-to-White median income between 0.85 and 1.15, and had a ratio of Black-to-White high school graduation rates between 0.85 and 1.15 (for adults 25 and above) [4].

In the two census tracts, the racial distribution was 51% African American and 44% White, and the median income for the study area was $24,002, with no race difference. The census tracts were block listed to identify every occupied dwelling in the study area. During block listing, we identified 2618 structures. Of those, 1636 structures were determined to be occupied residential housing units (excluding commercial and vacant residential structures). After at least five attempts, contact was made with an eligible adult in 1244 occupied residential housing units. Of that number, 65.8% were enrolled in the study resulting in 1489 study participants (41.9% of the 3555 adults living in these two census tracts recorded in the 2000 Census). Because our survey had similar coverage across each census block group in the study area, the bias to geographic locale and its relationship with socioeconomic status should be minimal [21].

The EHDIC-SWB survey was administered in person by a trained interviewer and consisted of a structured questionnaire, which included demographic and socioeconomic information, self-reported health behaviors and chronic conditions, and three blood pressure (BP) measurements. The EHDIC-SWB study has been described in greater detail elsewhere [4]. The study was approved by the Committee on Human Research at the Johns Hopkins Bloomberg School of Public Health.

Participants

Participants who did not complete the questions on the social support scale (N = 2) and who did not identify as African American or White (N = 85) were excluded. This yields an analytic sample of 1406 participants.

Measures

Outcome

Hypertension was defined as systolic BP ≥ 140 mmHg and/or diastolic BP ≥ 90 mmHg or study participant reports of taking antihypertensive medication(s).

Main Independent Variable

Participants self-reported their race and ethnicity in the SWB Questionnaire. The options were (1) White, (2) Black or African American, (3) Hispanic or Latino, (4) Asian, (5) Native Hawaiian or other Pacific Islander, (6) American Indian or Alaskan Native, and (7) Biracial/Multiracial. The study only included participants that self-reported as “Black/African American” or “White.”

Moderators

There were two moderators for this study. Social support was measured using eight 5-category Likert-format items derived from the Duke-UNC Functional Social Support Questionnaire (DUFSS) as described in Table 1 [13]. Participants report how much they feel a statement of social support reflects their situation, with “1” meaning the statement occurs less than they would like and “5” meaning the statement occurs more than they would like. The eight-item instrument questions were in two content areas defined a priori as confidant (ability to share important information) support and affective (emotional) support [13].

Table 1.

Social support questions from the SWB questionnaire

1. I have people who care what happens to me

2. I get love and affection

3. I have chances to talk to someone about problems at work or with my housework

4. I get chances to talk to someone I trust about my personal and family problems

5. I get chances to talk about money matters

6. I get invitations to go out and do things with other people

7. I get useful advice about important things in life

8. I get help when I’m sick in bed

Derived from the Duke–UNC Functional Social Support Questionnaire

The eight-item DUFSS has been tested to be a valid and reliable instrument “in research applications that examine interactions between social support and other determinants of health” with test-test reliability of 0.55–0.77 and Cronbach’s alpha of 0.80–0.85 [22, 23]. DUFSS generates total scores of ranges 8 to 40 to represent overall social support, with higher scores representing more perceived social support [22, 23].

Marital status, the other moderator, was represented by one binary variable: currently or not.

Covariates

Covariates included demographics and health-related characteristics chosen based on previous literature to account for potential confounders of race and hypertension [16]. Demographics included age, gender, education attainment, annual income, and number of assets. Age was measured as a continuous variable, including participants ≥ 18 years old. Male sex was dichotomous (male or female). Three binary variables represented educational attainment: < 12 years of school or no diploma, high school graduate/GED, and more than high school. Annual income was measured as a categorical variable: less than $10,000, $10,000–$24,999, $25,000–$34,999, and $35,000–$54,999 or more than $50,000. Number of assets was measured as participants’ ownership of a home or other real estate, automobile, savings account, checking account, a business, mutual funds, stocks, or bonds. The number of assets was measured as a categorical variable: none, 1–2, greater than 2 assets.

Health-related characteristics include having a regular doctor, health insurance status, self-rated health, physical inactivity, weight status, smoking status, drinking status, and diabetes diagnosis. Having a regular doctor and health insurance status was measured as dichotomous. Five binary variables represented self-related health: excellent, very good, good, fair, or poor. For physical inactivity, respondents who were asked to rate how often they “exercise or participate in physical activity” for at least 1 h: none at all, less than once a month, once a month, once a week, 3 days a week, and more than three times a week. Responses for physical inactivity were measured as dichotomous. Weight status based on body mass index (BMI) was measured as categorical: < 18.5 as underweight, 18.5–24 as normal, 25–29.9 as overweight, and ≥ 30 as obese. Smoking status was measured as categorical (formerly, currently, never). Drinking status was dichotomous (regularly drink or not). Diabetes diagnosis was measured as dichotomous; diagnosis of diabetes was ascertained by respondents reporting whether a “doctor or other healthcare professional” informed them they had diabetes or not.

Perceived racial discrimination was measured using the Experiences of Discrimination Scale (EOD). Participants were asked, “Have you ever experienced discrimination, been prevented from doing something, or been hassled or made to feel inferior in any of the following situations because of your race or color?” for the following seven settings: (1) at school, (2) getting a job, (3) at work, (4) getting housing, (5) getting medical care, (6) from the police or in the courts, and (7) at a store, restaurant, or any other public place. The responses were dichotomized to having no experiences (none) and any experience across any setting (any). All dichotomous and binary variables were created with value 1 representing the attribute described in the name of the variable and zero otherwise.

Statistical Analysis

Means and proportional differences between African American and White participants for covariates were evaluated using Student’s t-tests for continuous variables and chi-square tests for categorical/dichotomous variables, respectively. Because the EHDIC-SWB sample had a high prevalence of hypertension (69.25%), modified Poisson regression models with robust standard errors were used to estimate prevalence ratios and corresponding 95% confidence intervals for the association between race and hypertension [24–26]. Model 1 was the unadjusted regression between race and hypertension. Model 2 adjusted for demographics, SES, health-related characteristics, and perceived racial discrimination. Model 3 is adjusted for the DUFSS score as measured by the x scale, and Model 4 is adjusted for marital status to determine which social support measure affects the race prevalence ratios on hypertension. The interaction of the DUFSS score and marital status was evaluated to determine whether social support modified race’s effect on hypertension, respectively. An interaction detected would mean that a racial difference in the prevalence of hypertension was observed within levels of social support. Interaction terms for the DUFSS score and marital status were then added to fully adjusted models; Model 5 is the fully adjusted model for with the DUFSS variable and the DUFSS interaction term, and Model 6 is the fully adjusted model with the marital status variable and marital status interaction term. All statistical procedures were conducted using Stata, version 15 (StataCorp LP, College Station, Texas).

Results

Table 2 displays the distribution of demographic and health related characteristics of the EHDIC-SWB sample by race. Compared to Whites, African Americans were more likely to be younger, less likely to be currently married, had about 0.7 assets, more likely to have high school or greater education, more likely to have health insurance, more likely to self-rate having excellent health, less likely to be physically inactive, more likely to ever drink, and more likely to have a higher DUFSS score. African Americans were also more likely to perceive any racial discrimination than Whites in EHDIC-SWB. The percentage of participants who had hypertension were similar for African Americans (70.8%) and Whites (67.2%).

Table 2.

Distribution of demographic variables and health-related characteristics of the EHDIC-SWB 2003 participants by race^a

Variable	Whites (N = 572)	African Americans (N = 831)
Age (years)*	43.9 ± 16.2	38.4 ± 13.3
Male (%)	43.1	45.7
Currently married (%)*	25.7	15.3
Income (%)
$0–$9999	22.2	25.3
$10,000–$24,999	40.3	41.5
$25,000–$34,999	15.5	12.7
$35,000–$54,999	11.7	11.3
$55,000+	10.3	9.1
Assets (numbers)*	1.1 ± 1.2	0.7 ± 1.1
Education level (%)*
Less than high school grad	47.5	35.5
High school graduate/GED	34.2	45.0
More than high school grad	18.3	19.5
Regular doctor (%)	61.6	61.2
Health insurance (%)*	59.8	65.1
Self-rated health (%)*
Excellent	10.1	17.1
Very good	17.6	23.6
Good	34.9	31.2
Fair	28.8	23.6
Poor	8.6	4.6
Physical inactivity (%)	44.7	42.5
Weight status (%)
Normalb	44.3	40.6
Overweight	25.8	27.4
Obese	29.9	32.0
Current smoker (%)	58.8	53.5
Ever drink (%)*	42.9	48.3
Diagnosis of diabetes (%)	11.8	9.3
Hypertension (%)	67.2	70.8
Perceived racial discrimination (%)*	38.3	59.7
Overall DUFSSc score*	30.8 ± 8.8	33.4 ± 7.5

The values which are represented with ± means ± SD. All binary variables are coded 1 and 0 represents the name of the variable

*p < 0.05

^aTests of significance are for differences in means for continuous variables (i.e., t-test) and differences in proportions (i.e., chi-square test) by experiences of racial discrimination

^b“Normal” weight comprised of both normal and underweight participants because of the small number of underweight participants (N = 32)

^cDuke-UNC Functional Social Support Study, score is on a scale of 8–40

Table 3 displays the association between race and hypertension. There is no association between race and hypertension in the unadjusted model (PR 1.05, 95% CI 0.98, 1.13). After controlling for demographics, SES, health-related characteristics, and perceived racial discrimination in Model 2, African Americans have a higher prevalence of hypertension than Whites (PR 1.11, 95% CI 1.03, 1.20). When the DUFSS score was added to Model 3, the prevalence ratio did not change but remained statistically significant (PR 1.11, 95% CI 1.03, 1.20). When marital status was added to Model 4, the prevalence ratio remained the statistically significant (PR 1.10, 95% CI 1.02, 1.19). In Model 5, there is no association between the DUFSS score and hypertension (PR 1.00, 95% CI 0.99, 1.01) and no interaction between DUFSS score and race (PR 1.00, 95% CI 0.99, 1.01). In Model 6, no association was observed between marital status and hypertension (PR 0.93, 95% CI 0.82, 1.06), and no interaction term was detected between marital status and race (PR 1.02, 95% CI 0.86, 1.21). There is no association between race and hypertension when the DUFSS score and its interaction term were included in Model 5 (PR 1.09, 95% CI 0.83, 1.43) and when marital status and its interaction term were included in Model 6 (PR 1.10, 95% CI 1.01, 1.19).

Table 3.

Prevalence ratios and 95% confidence intervals for the association between race and hypertension prevalence in the EHDIC-SWB study

Independent variable	Model 1 (N = 1406)	Model 2 (N = 1322)	Model 3 (N = 1322)	Model 4 (N = 1321)	Model 5 (N = 1,322)	Model 6 (N = 1,321)
Race
White	1.00	1.00	1.00	1.00	1.00	1.00
African American	1.05 (0.98–1.13)	1.11 (1.03–1.20)	1.11 (1.03–1.20)	1.10 (1.02–1.19)	1.09 (0.83–1.43)	1.01 (1.01–1.19)
DUFSS Score	-	-	-	-	1.00 (0.99–1.01)	-
DUFSS Score x Race	-	-	-	-	1.00 (0.99–1.01)	-
Marital Status	-	-	-	-	-	0.93 (0.82–1.06)
Marital Status x Race	-	-	-	-	-	1.02 (0.86–1.21)

White is the reference group. Model 1, unadjusted; Model 2, adjusted for demographics, SES, health-related characteristics, and perceived racial discrimination; Model 3, adjusted for demographics, SES, health-related characteristics, perceived racial discrimination, and DUFSS score; Model 4, adjusted for demographics, SES, health-related characteristics, perceived racial discrimination, and marital status. Adding DUFSS score or marital status into the model did not significantly change the relationship between race and hypertension; Model 5, fully adjusted model with DUFSS score and interaction term; Model 6, fully adjusted model with marital status and interaction term

Discussion

The objective of the study was to examine the role of social support in the relationship between race and hypertension in a sample of African American and White adults living in similar social and environmental conditions using data from the EHDIC-SWB study. No interaction between social support and race was found; thus, social support did not significantly moderate the association between race and hypertension in the EDHIC-SWB sample. Conversely, the hypothesis was not supported by these findings because African Americans were as likely to have hypertension as Whites regardless of level of social support.

From the EHDIC-SWB sample, the lack of social support’s protective influence on race and hypertension is inconsistent with prior work that have shown that social support protects African Americans from developing hypertension [14–20]. Previous studies that sampled predominantly Black or Black only participants showed social support as a buffer for African Americans [14, 15, 19]. Specifically, Coulon’s study demonstrated that social support buffered the effects of neighborhood disorder and poverty [14]. Coulon sampled African American communities in Southeastern region of the US, which experiences “high levels of chronic disease, poverty, and crime relative to state and national averages.” [14] These “at-risk” African American communities were predominantly Black and had variation of high, medium, and low SES [14]. At the lowest levels of income, greater perceived social support was protective against the negative link between income and diastolic blood pressure [14]. Heard and colleagues also sampled predominantly Black participants and found that social support mediated the relationship between depression and hypertension [15]. Heard and colleagues suggested frequent exposure to stress can improve the body’s adaptability to maintain normal blood pressure [15]. However, the EHDIC-SWB sample has a high prevalence of hypertension. Given the environment in which African Americans in EHIDC-SWB live in, there may be limits of the buffering effects of social support. Rather, macro-level, neighborhood factors might be involved such as neighborhood disorder, quality of housing, walkable spaces, and access to healthier food options.

This study also used different measures of social support compared to previous studies, which lead to different interpretations of the role of social support between race and hypertension. Coulon and colleagues used the Scale of Perceived Social Support (MSPSS), a 12-item, 7-point Likert scale with similar questions about social support from family, friends, and significant others [14]. Heard and colleagues measured social support as social support given versus received on a scale from the National Survey on Black Americans [15]. Bell and colleagues measured social support as emotional support, financial support, and marital status from NHANES [16]. Dressler and colleagues measured structural and functional support on the Problem-by-Source of Support Matrix [19]. Strogatz and colleagues only measured instrumental support and emotional support [20]. In this study, the DUFSS score nor marital status could not explain the high percentage of hypertensive participants in EHDIC-SWB sample, which may reflect other social and environmental conditions that are unique to the community.

One possible explanation for our findings is the historical and social contexts of social support in this low-income, urban racially integrated community. African Americans and Whites in EHDIC-SWB became neighbors in the 1960s when laws banned racial residential segregation [11]. African Americans moved into this community and worked with their White counterparts at a local manufacturing base [11]. As the manufacturing industry eroded, upper- and middle-class White neighbors moved to suburban areas [11]. Thus, low-income African American and White working families remained [11]. The history of the social and environmental conditions which they live in may mold how social support is perceived and structured in this community.

In EHDIC-SWB, White adults are exposed to the same social and environmental risks of hypertension that typically affects African Americans who live in low-income, segregated communities [27, 28] Hines and colleagues [28] found that in EHDIC-SWB, African Americans reported more discrimination, and discrimination and vigilant coping were associated with hypertension among African Americans. Vigilant coping, though operating on the individual-level, may reflect more mezzo- and macro-level factors associated with race and racism. This suggests that, though African Americans and Whites live in the same community, they have different lived experiences that can contribute to residual racial differences. Because social support did not moderate the weaker, yet persistent, association between race and hypertension among African Americans and Whites living in similar contexts, it is likely that individual-level factors like social support and marital status cannot overcome potential community- and structural-level factors that may interact with race to produce comparable prevalence of hypertension among African Americans to Whites living in the same social context.

This study has some limitations. This study was unable to compare results of analytic models from the EHDIC sample to national samples due to different measurements of social support [4, 11, 12, 16]. As a cross-sectional study, causal inferences cannot be made about the interrelationship between social support, race, and hypertension. Because this study samples a low-income, urban, racially integrated community, the findings are not generalizable to all communities in the US. Further studies should consider other integrated communities that are of higher income, non-urban, or with other racial minorities to examine the role of social support on race and hypertension. Including qualitative data, such as participant perceptions of their social support system, may shed light on the uniqueness of a racially integrated, low-income, urban community. Other psychosocial factors such as coping mechanisms and perceived stress should be considered as moderators of the relationship.

The strength of this study was having a sample of African Americans and Whites living in a racially integrated, low-income, urban community. This community setting allows us to account for the unmeasured heterogeneity of social and environmental factors from living in different segregated communities [4, 11, 12].

This study aimed to examine whether social support moderated the relationship between race and hypertension among African American and White adults living in similar social and environmental conditions. Social support did not significantly moderate this relationship in the EHDIC-SWB sample. Although social support was not related to race and hypertension, this study adds to the growing research in examining hypertension prevalence by accounting for the social and environmental conditions that are the lingering effects of segregation.