Life Course Socioeconomic Status and Hypertension in African American Adults: The Jackson Heart Study

Abstract

BACKGROUND

Limited research has examined the association of life-course socioeconomic status (SES) with hypertension prevalence and incidence in a large cohort of African Americans.

METHODS

Among 4,761 participants from the Jackson Heart Study (JHS), we examined the association of SES indicators with prevalent and incident hypertension. We used multivariable Poisson regression to estimate prevalence ratios (PR, 95% confidence interval—CI) of baseline (2000–2004) hypertension by adult (education, income, occupation, wealth) and childhood (mother’s education) SES. Cox proportional hazards regression was used to estimate hazard ratios (HR, 95% CI) of incident hypertension by adult and childhood SES (2005–2013; 7.21 median years of follow-up). We also examined the association of childhood-to-adult SES mobility (parent-to-adult education) with prevalent and incident hypertension. Model 1 adjusted for age and sex. Model 2 added waist circumference, behaviors (smoking, alcohol, physical activity, diet), and diabetes prevalence.

RESULTS

High (vs. low) adult SES measures were associated with a lower prevalence of hypertension, with the exception of having a college degree and upper-middle income (PR: 1.04, 95% CI: 1.01, 1.07; PR: 1.05, 95% CI: 1.01, 1.09, respectively). Higher childhood SES was associated with a lower prevalence and risk of hypertension (PR: 0.83, 95%: CI 0.75, 0.91; HR: 0.76, 95% CI: 0.65, 0.89, respectively). Upward mobility and consistent high SES (vs. consistent low SES) from childhood to adulthood was associated with a greater prevalence, but lower incidence of hypertension.

CONCLUSION

Efforts to prevent hypertension among African Americans should consider childhood and current SES status.

Hypertension is a major risk factor for cardiovascular diseases (CVD) such as stroke, myocardial infarction, and coronary heart disease.¹ Among racial and ethnic groups in the United States, African Americans have the highest reported prevalence of hypertension.² Socioeconomic position has been referred to as a “fundamental cause” of disease.³ Research has found that objective measures of education, income, occupation, and subjective social standing are associated with hypertension^4,5; however, there is limited evidence of how socioeconomic status (SES) relates to hypertension in African Americans.

Life-course approaches have been used to examine adult and childhood SES and their impact on morbidities such as obesity,⁶ diabetes,^7,8 metabolic syndrome,⁹ subclinical disease,¹⁰ and heart disease.¹¹ A study by Matthews et al.¹² found an association between economic hardship and increased risk of hypertension among African Americans and Whites (n = 3,827). James et al.¹³ found a lower odds of hypertension among 379 African American men with high (vs. low) life-course socioeconomic position. Few studies have examined the association between adult and childhood SES and prevalent and incident hypertension in a large cohort of African Americans.

We used data from the Jackson Heart Study (JHS) to investigate the association of individual SES indicators (adult SES-education, income, occupation, and wealth and childhood SES-mother’s education) with prevalent and incident hypertension. We also examined whether socioeconomic mobility from childhood to adulthood was associated with prevalent and incident hypertension. Understanding the link between individual indicators of life-course SES and hypertension in African Americans is an important research priority that would potentially reveal mechanisms and improve intervention efforts to eliminate hypertension disparities.¹⁴ We hypothesized that hypertension prevalence and incidence would be inversely patterned by adult and childhood SES.

METHODS

The JHS is a longitudinal cohort study of CVD among African American men and women (N = 5,306) in 3 counties (Hinds, Madison, Rankin) of Jackson, MS. Participants (21–95 years old) were enrolled at baseline (2000–2004) and were followed through exam 2 (2005–2008) and exam 3 (2009–2013). Participants completed in-home interviews, questionnaires, and in-clinic examinations to obtain a wide range of demographic, psychosocial, anthropometric, and clinical data. A description of the recruitment and study methods has been reported elsewhere.^15,16 The study protocol was approved by the institutional review boards of the University of Mississippi Medical Center, Jackson State University, and Tougaloo College, and all participants provided informed consent.

Measurement of blood pressure

Trained staff conducted 2 blood pressure (BP) measurements in the right arm, and 1 minute elapsed between the 2 measurements using a random zero sphygmomanometer (Hawksley and Sons Ltd., Lancing, UK) at exams 1 and 2 and a semi-automatic oscillometric device (Omron HEM-907XL, Omron Healthcare Inc., Lake Forest, IL) at exam 3. The 2 clinic BP measurements were averaged for analysis. The random-zero BP measurements were calibrated to the semi-automated device using robust regression.¹⁷

Prevalent hypertension

Prevalent hypertension was defined as baseline systolic blood pressure (SBP) ≥ 140 mm Hg, or diastolic blood pressure (DBP) ≥ 90 mm Hg, or taking antihypertensive medications, or answering “yes” to “Have you ever been told by a physician that you had high blood pressure?”

Incident hypertension

Incident hypertension was defined as the first follow-up exam (2 or 3) at which a participant had SBP ≥ 140 mm Hg or DBP ≥ 90 mm Hg or self-reported antihypertensive medication use, or answering “yes” to “Have you ever been told by a physician that you had high blood pressure?”

Adult SES

Measures of adult SES included self-reported educational attainment, household income, occupational status, and wealth at exam 1. Educational attainment (years of schooling completed) was classified as less than a high school diploma (< HS), high school graduate/general equivalency diploma to some college/associate’s degree (HS4—C1-3), or college degree or more (C4+). Using census designations for family size and number of children under 18 years of age, household income was classified as poor (< poverty level), lower-middle (1–1.5 times the poverty level), upper-middle (>1.5 but <3.5 times the poverty level), and affluent (3.5+ times the poverty level) income. Occupation was categorized as construction/production, service, and management/professionals.

A total wealth score was based on the sum of 3 liquid assets: (i) owning a home, (ii) owning 1 or more cars, and (iii) participant’s net worth (including spouse, if applicable). For homeownership, participants received a value of 1 if they owned their home, and a score of 0 if they did not. For car ownership, each participant received a value of 1 if they owned 1 or more cars and a value of 0 if they did not own a car. For net worth, participants were asked to select the category, ranging from 0 to $200,000 or more, in which they would consider themselves if they needed emergency funds/liquid assets. If the participant had more than $200,000 in liquid assets, they were assigned a value of $250,000. A continuous asset measure was created by calculating the midpoint of each asset bracket, and then the summed score was dichotomized at the median; where below the median ($35,000) was coded 0 (low) and coded 1 (high) if above the median. A total of 3 points (from car, home, and assets) indicated “high wealth,” whereas 2 points indicated “moderate wealth,” and 1 or 0 points indicated “low wealth.”

Childhood SES

Childhood SES was measured using mother’s or most important female caretaker’s educational attainment. Participants were asked to describe their mother’s (or female caretaker’s) education up to when the participant was 16 years old: “What is the highest degree or years of school your mother (or female caretaker) completed, including trade or vocational school or college?” If the participant reported that their mother/female caretaker had less than 12 years of school, they were asked whether their mother/female caretaker completed a GED. The final classification was (i) less than high school (< HS), (ii) high school graduate or general equivalency diploma (HS4/GED), (iii) some college or trade school, and (iv) college degree or more (C4+).

Socioeconomic mobility

We defined childhood SES by mother’s education in the following categories: (i) low SES: < HS, (ii) lower-middle SES: HS4/GED, (iii) upper-middle SES: 1–3 years of college, and (iv) high SES: 4 or more years of college (C4+). Second, we defined adult SES by education: (i) low SES: < HS, (ii) lower-middle SES: HS4 to some college, (iii) upper-middle SES and high SES: C4+. We created two (2) categories where SES remained the same over time (consistent) [(i) low childhood SES to low adult SES, and (ii) high childhood SES to high adult SES]. We also created upward (low childhood SES to high adult SES) and downward (high childhood SES to low adult SES) mobility measures.

Covariates

Baseline covariates included demographics, health behaviors, diabetes, and waist circumference. Demographics included age (continuous) and sex (male/female). Health behaviors included alcohol intake in the past 12 months (yes/no), dietary intake [sodium (mg), % fat calories, total caloric intake (kcal)]. Other health behaviors were adapted from the American Heart Association Life’s Simple 7 (ideal vs. non-ideal). We adjusted for smoking (non-ideal: current smoker or had quit smoking less than 12 months; ideal: never smoked or had quit smoking more than a year before examination), and physical activity (non-ideal: fewer than 150 minutes of moderate physical activity and fewer than 75 minutes of vigorous physical activity per week; and ideal: more than 150 minutes of moderate physical activity and more than 75 minutes of vigorous physical activity per week). Waist circumference was measured in centimeters (cm). Diabetes prevalence was defined as having a fasting glucose ≥ 126 mg/dL, or HbA1c ≥ 6.5%, or use of antidiabetic medication.

Statistical analyses

Of the 5,306 participants in the JHS, 545 were excluded due to having CVD at exam 1, and 256 were excluded for missing covariates [education (n = 13), occupation (n = 54), wealth (n = 50), ideal smoking (n = 65), alcohol (n = 25), waist circumference (n = 3), diabetes (n = 46)]. Because of high missing for income (n = 820; 15.7%) and mother’s education (n = 1,132; 27.0%), we created missing dummy categories to retain missing values in the analysis. For these variables, we also performed multiple imputations using chained equations with 10 data sets to reduce bias and loss of statistical power. Imputed regression results were reported, as they were similar to the results with the missing dummy category.

Descriptive statistics were presented for select baseline covariates and SES measures by hypertension status. Associations of adult and childhood SES with baseline hypertension were examined using multivariable Poisson regression with robust standard errors to estimate prevalence ratios (PR, 95% confidence interval—CI) of hypertension. Model 1 adjusted for age and sex, and model 2 added smoking, physical activity, alcohol consumption, sodium, total energy, fat, waist circumference, and diabetes. An additional model adjusted for age, sex, and all other SES measures (Supplementary Table 1).

For incident analyses, those with hypertension at baseline were excluded (n = 2,538). New cases of hypertension included 642 participants at exam 2 and 334 participants at exam 3. Time-to-event variable(s) were generated using the midpoint of years between visit dates. Those without hypertension were censored at the date of their last exam. Cox proportional hazards regression was used to estimate the associations of exam 1 adult and childhood SES with incident (or first reported case of) hypertension to estimate hazard ratios (HR 95% CI) of hypertension risk. Model 1 adjusted for baseline age and sex, and model 2 added baseline health behaviors, dietary factors, waist circumference, and diabetes. An additional model adjusted for age, sex, and all SES measures (Supplementary Table 2). In both the prevalence and incidence regression analyses, each SES measure was classified in categories where the lowest category was the referent [e.g., education: < HS (referent), HS4-C1-3, C4+].

Poisson and Cox regression (described earlier) were used to estimate associations of SES mobility with prevalent and incident hypertension, respectively. We tested for effect modification by age and sex for all analyses. All tests were 2-tailed, and a P value < 0.05 was considered statistically significant. All analyses were conducted using Stata, version 14 (StataCorp, College Station, TX).

RESULTS

Hypertensive participants were more likely to be older, female, have less ideal physical activity, have more ideal smoking, have higher waist circumference, and lower sodium intake than normotensive participants (P < 0.05). Patently, adult SES measures were positively associated with hypertension prevalence; childhood SES was inversely associated with hypertension prevalence (Table 1).

Table 1.

Baseline characteristics by hypertension status: the Jackson Heart Study (2000–2004) (N = 4,761)

Characteristics	Normotensive (%)	Hypertensive (%)	P value
Age (mean, Std)	48.8 (12.3)	59.3 (11.1)	<0.001
Male	51.9	48.1	<0.001
Ideal physical activity	23.3	16.7	<0.001
Ideal smoking	84.6	87.5	0.005
Waist (mean, Std)	96.8 (15.9)	103.7 (16.0)	<0.001
Sodium (mean, Std)	4,564.6 (2,639.4)	3,830.1 (2,106.6)	<0.001
Adult SES
Education			<0.001
<High school	12.8	23.7
High school/GED-some college	42.2	38.7
College +	44.9	37.7
Income			<0.001
Poor	11.4	13.2
Lower-middle	17.8	22.1
Upper-middle	26.3	24.9
Affluent	29.6	24.6
Occupation			<0.001
Construction/production	45.1	54.9
Service/sales	43.2	56.8
Management/Prof	51.0	49.0
Total wealth			<0.001
Low	29.5	24.6
Moderate	47.3	43.9
High	23.3	31.5
Childhood SES			<0.001
<High school	36.9	46.7
High school/GED	18.0	13.2
Some college	12.5	6.6
College +	10.7	4.7

Abbreviations: GED, general equivalency diploma; SES, socioeconomic status; Waist–Waist circumference. Age is in years, waist circumference was measured in centimeters, and sodium was measured in milligrams based on the food frequency questionnaire. P value based on chi square and ANOVA tests.

The prevalence of hypertension was 53.3% (n = 2,538). Higher (vs. lower) education was associated with a greater prevalence of hypertension in model 2 (HS/GED-some college: PR: 1.11, 95% CI: 1.09, 1.14; College degree +: PR: 1.04, 95% CI: 1.01, 1.07) (Table 2). The trend was statistically significant in both models (P for trend < 0.001). Having a college degree was associated with 15% greater prevalence of hypertension (PR: 1.15, 95% CI: 1.03, 1.29) after adjusting for age, sex, and all SES measures (Supplementary Table 1). Affluent (vs. poor) income was significantly associated with a lower prevalence of hypertension in model 2 (PR: 0.91, 95% CI: 0.87, 0.96); however, upper-middle (vs. poor) income was associated with a greater prevalence in model 2 (PR: 1.05, 95% CI: 1.01, 1.09). Higher occupation and childhood SES were associated with a lower prevalence of hypertension in model 2. High (vs. low) total wealth was associated with a lower prevalence of hypertension in model 1 (PR: 0.96, 95% CI: 0.93, 0.98) and after adjustment for age, sex, and all SES measures (PR: 0.87, 95% CI: 0.79, 0.97; Supplementary Table 1).

Table 2.

Prevalence ratios (PR 95% CI) of hypertension by adult and childhood SES among African Americans in the Jackson Heart Study (2000–2004) (N = 4,759)

SES characteristics		PR 95% CI
Adult SES	Total (% prevalent HTN)	Model 1	Model 2
Education
Less than high school (ref)	23.7	1.0	1.0
HS grad-some college/technical	38.7	1.08 (1.05, 1.11)*	1.11 (1.09, 1.14)*
College degree or more	37.7	1.00 (0.97, 1.03)	1.04 (1.01, 1.07)*
P for trend		<0.001	<0.001
Income
Poor (ref)	13.2	1.0	1.0
Lower-middle	22.1	0.93 (0.90, 0.97)*	0.96 (0.93, 0.99)*
Upper-middle	24.9	1.01 (0.98, 1.05)	1.05 (1.01, 1.09)*
Affluent	24.6	0.87 (0.84, 0.91)*	0.91 (0.87, 0.96)*
P for trend		<0.001	<0.001
Occupation
Construction/production (ref)	37.3	1.0	1.0
Service/sales	28.8	0.94 (0.91, 0.96)*	0.93 (0.90, 0.95)*
Management/professional	33.9	0.89 (0.86, 0.92)*	0.91 (0.88, 0.94)*
P for trend		<0.001	<0.001
Total wealth
Low (ref)	24.6	1.0	1.0
Moderate	43.9	0.99 (0.97, 1.02)	1.01 (0.98, 1.04)
High	31.5	0.96 (0.93, 0.98)*	0.99 (0.96, 1.02)
P for trend		0.02	0.27
Childhood SES
Mother’s education
Less than high school (ref)	46.7	1.0	1.0
High school grad/GED	13.2	0.91 (0.86, 0.96)*	0.89 (0.85, 0.94)*
Some college	6.6	1.04 (0.96, 1.11)	1.01 (0.94, 1.09)
College degree or more	4.7	0.82 (0.74, 0.90)*	0.83 (0.75, 0.91)*
P for trend		<0.001	0.001

Abbreviations: CI, confidence interval; GED, general equivalency diploma; HS, high school; HTN, hypertension; ref, referent; SES, socioeconomic status. Total prevalent HTN % (hypertension case/total participants at risk). Model 1 adjusted for age and sex, model 2 adjusted for model 1 + physical activity, alcohol, smoking, sodium (mg), total energy (kcal), fat (%), and waist circumference (cm), and diabetes prevalence. Missing data from income (15.7%) and mother’s education (27.0%) were imputed with 10 data sets. *Bold = *P ≤ 0.05.

There were 976 (43.9%) new cases of hypertension after baseline (median follow-up years was 7.21). Incident rates of hypertension were generally higher for those with lower SES at exam 1; however, those with higher wealth and who worked in service/sales had a greater incidence rate of hypertension (Table 3). Marginally significant associations were found between having at least a high school diploma and having a higher risk of hypertension. Upper-middle income at exam 1 was associated with a 16% and 37% greater risk of hypertension in models 1 and 2 (HR: 1.16, 95% CI: 1.03, 1.30; HR: 1.37, 95% CI: 1.21, 1.57 respectively). After adjustment for age, sex, and all SES measures, the association was not significant. Occupation and wealth were not significantly associated with hypertension risk after full adjustment. Higher childhood SES was significantly associated with a lower risk of hypertension in models 1 and 2 (P < 0.05).

Table 3.

Hazard ratios (HR, 95% CI) of incident hypertension by adult and childhood SES among African Americans in the Jackson Heart Study (2005–2013) (N = 976)

SES characteristics		HR 95% CI
Adult SES	HTN rate/1,000 PY	Model 1	Model 2
Education
Less than high school (ref)	38.6 (19.6, 75.9)	1.0	1.0
High school grad-some college	24.5 (16.6, 36.1)	1.08 (0.99, 1.18)	1.08 (0.99, 1.18)
College degree or more	20.7 (13.8, 31.0)	0.95 (0.86, 1.04)	1.04 (0.94, 1.15)
P for trend		<0.001	0.18
Income
Poor (ref)	21.2 (11.6, 38.7)	1.0	1.0
Lower-middle	28.5 (16.5, 49.4)	1.01 (0.90, 1.14)	1.12 (0.99, 1.27)
Upper-middle	23.9 (14.5, 39.7)	1.16 (1.03, 1.30)*	1.37 (1.21, 1.57)*
Affluent	19.1 (11.1, 32.8)	0.93 (0.83, 1.05)	1.06 (0.94, 1.20)
P for trend		0.001	<0.001
Occupation
Construction/production (ref)	21.6 (14.5, 32.3)	1.0	1.0
Service/sales	30.1 (18.4, 49.2)	1.08 (0.99, 1.18)	1.04 (0.95, 1.13)
Management/professional	24.8 (16.0, 38.4)	0.96 (0.88, 1.05)	1.03 (0.94, 1.33)
P for trend		0.004	0.66
Total wealth
Low (ref)	21.7 (14.0, 33.6)	1.0	1.0
Moderate	22.3 (15.1, 33.1)	0.91 (0.85, 0.98)*	0.97 (0.90, 1.04)
High	30.2 (16.4, 55.7)	1.05 (0.96, 1.14)	1.09 (0.99, 1.19)
P for trend		<0.001	0.02
Childhood SES
Mother’s education
Less than high school (ref)	31.5 (19.9, 49.9)	1.0	1.0
High school grad/GED	27.1 (14.2, 51.5)	0.81 (0.72, 0.91)*	0.81 (0.71, 0.91)*
Some college	16.3 (8.0, 32.9)	0.77 (0.66, 0.89)*	0.76 (0.66, 0.88)*
College degree or more	18.7 (8.8, 39.9)	0.78 (0.67, 0.91)*	0.76 (0.65, 0.89)*
P for trend		<0.001	<0.001

Abbreviations: CI, confidence interval; GED, general equivalency diploma; HTN, hypertension; PY, person years; ref, referent; SES, socioeconomic status. PY were calculated as time to first hypertensive event; participants were censored after hypertensive event. Model 1 adjusted for age and sex. Model 2 adjusted for model 1, physical activity, alcohol, smoking, sodium (mg), total energy (kcal), fat (%), waist circumference (cm), and diabetes. Missing data from income (15.7%) and mother’s education (27.0%) were imputed using 10 data sets. *Bold = *P ≤ 0.05.

Table 4 presents the associations of childhood-to-adult SES mobility with prevalent and incident hypertension. Upward mobility and consistent high SES were associated with a greater hypertension prevalence in models 1 and 2. Upward mobility remained associated with hypertension prevalence when adjusting for age, sex, and all SES factors (Supplementary Table 3). All mobility categories were associated with a lower risk of hypertension compared to those with consistent low SES.

Table 4.

Prevalence ratios (PR 95% CI) of hypertension and Cox Proportional Hazard Ratios (HR 95% CI) of incident hypertension by socioeconomic mobility trajectories among African Americans in the Jackson Heart Study (2000–2013)

		PR 95% CI			HR 95% CI
Socioeconomic Mobility (N)	(%) Prevalent HTN	Model 1	Model 2	(%) Incident HTN	Model 1	Model 2
Consistent low SES (405) (ref)	67.8	1.0	1.0	62.7	1.0	1.0
Downward mobility (120)	26.7	0.84 (0.70, 1.02)	0.89 (0.74, 1.07)	53.9	0.88 (0.71, 1.08)	0.75 (0.60, 0.94)*
Upward mobility (2327)	53.4	1.25 (1.19, 1.31)*	1.27 (1.21, 1.33)*	51.5	0.80 (0.68, 0.94)*	0.79 (0.66, 0.94)*
Consistent high SES (526)	40.1	1.27 (1.18, 1.37)*	1.28 (1.18, 1.38)*	42.0	0.61 (0.50, 0.74)*	0.62 (0.50, 0.79)*
P for trend		<0.001	<0.001		<0.001	<0.001

Abbreviations: CI, confidence interval; HTN, hypertension; ref, referent. Socioeconomic mobility was based on mother’s education and adult education. Model 1 adjusted for age and sex. Model 2 adjusted for model 1, physical activity, alcohol, smoking, sodium (mg), total energy (kcal), fat (%), waist circumference (cm), and diabetes. Missing data from income (15.7%) and mother’s education (27.0%) were imputed with 10 data sets. *Bold = *P ≤ 0.05.

We performed effect modification by age and sex but did not find significant results, except for mother’s education by age (P values for interaction < 0.05), thus, we stratified by age (< 55 and ≥ 55). Participants younger than 55 (compared to those 55 and older) have a lower prevalence of hypertension for each level of mother’s education (Supplementary Table 3).

DISCUSSION

We examined the association between life-course SES with hypertension prevalence and incidence among African Americans. High adult income, occupation, and wealth were inversely associated with hypertension prevalence. Moderate wealth was inversely associated with incident hypertension in minimally adjusted models but not after full adjustment. Higher childhood SES was associated with reduced prevalent and incident hypertension. When compared to those with consistent low SES, upward mobility and consistent high SES from childhood to adulthood were associated with greater prevalent, but lower incident hypertension. We found partial support for our hypothesis.

Numerous studies have reported significant associations between low SES and hypertension prevalence,¹⁸ but the association between higher SES and hypertension in African Americans is not as commonly reported.^19,20 Our study found that affluent income, higher occupations, higher wealth, and higher mother’s education were protective of hypertension prevalence. Our results are comparable with a study of 379 African American men that reported a lower odds of hypertension with high (vs. low) life-course SES measured by an aggregate score of parent occupation and adult education, occupation, employment, and home ownership.¹³ Conversely, we found that participants with college degrees or more and upper-middle incomes had a higher prevalence of hypertension. Perhaps the higher prevalence of hypertension among college-educated and upper-middle income participants explains this finding, which may be indicative of a ceiling effect. It is plausible that these participants were predisposed to higher amounts of stress, as some may have experienced many challenges (e.g., first-generation college student and racial discrimination). According to James,²¹ frequently occurring psychosocial stressors explain why African Americans are more susceptible to hypertension, and higher SES, especially higher education, may not be as protective of hypertension because these stressors are unrelieved.

Moderate wealth was protective of incident hypertension after adjustment for age and sex but attenuated after full adjustment, which suggests mediation or moderation of behaviors and risk factors. High education and income were also inversely associated with hypertension incidence in minimally adjusted models but were not significant. On the other hand, upper-middle income was associated with a greater incidence of hypertension, after full adjustment. This group may experience stressors such as perceived discrimination and tokenism in respective environments (work, neighborhood, etc.), when compared to lower-income groups.²² In a previous study of JHS participants, high perceived everyday and lifetime discrimination were reported in over 30% of JHS participants with incomes of at least $50,000.²³ Evidence from Din-Dzietham et al.²⁴ suggests that race-based discrimination at work is associated with an increased likelihood of hypertension among African Americans (n = 356).

Higher childhood SES was protective of incident hypertension, with a significant inverse gradient across categories of mother’s education. Other studies have emphasized the lasting impact of low childhood SES on co-morbidities. For example, Roberts et al.²⁵ found that having low early-life SES was associated with increased risk of incident heart failure after adjustment for age and sex. Another study reported that low childhood SES was associated with poor cardiovascular health, substance abuse, and periodontal disease as adults.²⁶

Although upwardly mobile and consistent high SES participants had a higher prevalence of hypertension, normotensive participants who were upwardly mobile and had consistent high SES had a lower risk over time. The majority of participants in this study were upwardly mobile (n = 2,327), which suggests that they may have experienced economic hardships throughout the life course that affected their eventual risk of hypertension. Health behaviors learned from childhood in a low SES environment may have also continued into adulthood (i.e., avoiding medical examinations, poor dietary choices), which may contribute to a greater prevalence of hypertension. However, the associations of upward mobility (among lower childhood SES) with reduced risk of hypertension may suggest the adoption of better medication adherence and optimal health behaviors among this segment of the sample.

This study had a number of strengths, which includes the examination of multiple SES measures with prevalent and incident hypertension in a socioeconomically heterogeneous sample of African Americans. Over 35% of JHS participants had a college degree; most studies that use African Americans have a large proportion of poorly educated participants. Because the baseline prevalence of hypertension is over 60% in the JHS, we were most likely able to see variations in hypertension by SES. We examined wealth, which is rarely analyzed in studies that examine the social patterning of hypertension. Despite these strengths, this study has limitations. First, we were unable to account for stressors or changes in SES over time, and there may have been other covariates (e.g., family history of hypertension) that could have influenced the results, but were not adjusted for in the analysis. Second, mother’s education was the only measure used to define childhood SES, and it was measured retrospectively, which could induce measurement error. Third, in the mobility analyses, adult education was used to represent adult SES. Fourth, we were unable to estimate time to diagnosis of hypertension after exam 1. Fifth, there was a large amount of missing data for income and mother’s education and although we used multiple imputations to minimize bias and loss of power, it is possible that data were not missing at random. Finally, a single-site study limits generalizability to other African Americans in the United States.

Overall, efforts to eliminate hypertension disparities should focus on lower status African Americans and the mobility process they experience throughout the life course. In addition, prevention efforts should also target highly educated and upper-income persons where considerations of stressors and economic adversity during the life course could also play a role in hypertension disparities. More studies with African Americans in other regions of the United States are warranted to confirm these findings.

FUNDING

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 for Minority Health and Health Disparities (NIMHD). Dr. Sims was also supported by the grants P60MD002249 and U54MD008176 from the NIMHD; 15SFDRN26140001 and P50HL120163 from the American Heart Association and 1R01HL116446 from the NHLBI. Ms. Glover is supported by the Genetic Epidemiology of Heart, Lung, and Blood Traits Training Grant (GENHLB) T32 HL129982.

DISCLOSURE

No financial disclosures were reported by the authors of this article.

DISCLAIMER

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 US Department of Health and Human Services.