Childhood Family Living Arrangements and Blood Pressure in African American Men: The Howard University Family Study

Abstract

African American men have higher blood pressure levels and consequentially higher prevalence of hypertension compared to men from other ethnic groups in the United States. Socio-familial factors in childhood have been found to play an important role in hypertension, but few studies have examined this relationship among African American men. We investigated whether childhood family living arrangements are independently associated with mean blood pressure and hypertension in a cross-sectional sample of 515 unrelated African American male participants aged 20 years and older enrolled in the Howard University Family Study between 2001 and 2008. African American men who lived with both parents compared to the reference group of men who never lived with both parents during their lifetime had lower systolic blood pressure [−4.4 mmHg (95% Confidence Interval: −7.84, −0.96)], pulse pressure [−3.9 mmHg (95% Confidence Interval: −6.28, −1.51)] and mean arterial blood pressure [−2.0 mmHg (95% Confidence Interval: −4.44, 0.51)]. This protective effect was more pronounced among men who lived with both parents for 1 to 12 years of their lives; they had decreased systolic blood pressure [−6.5 mmHg, (95% Confidence Interval: −10.99, −1.95)], pulse pressure [−5.4 mmHg, (95% Confidence Interval: −8.48, −2.28)], mean arterial pressure [−3.3 mmHg, (95% Confidence Interval: −6.56, 0.00)], and a 46% decreased odds of developing hypertension (OR = 0.54; 95% Confidence Interval: 0.30, 0.99). No statistically significant associations were found for diastolic blood pressure. These results provide preliminary evidence that childhood family structure exerts a long-term influence on blood pressure among African American men.

Introduction

In the United States (US), over one-third of adults 20 years and older have elevated blood pressure (BP) high enough to be classified as hypertension, a major risk factor for stroke, myocardial infarction, heart and renal failure.¹ It has been consistently observed that African American men display higher BP and rates of hypertension compared to men from other US ethnic groups. For example, in the 1930s, African American men were reported to have a 7 mmHg and 4 mmHg higher mean systolic blood pressure (SBP) and diastolic blood pressure (DBP) respectively, relative to their male counterparts of European descent.² This differential in elevated blood pressure has not changed over time with recent statistics for 2007–2010 showing age-adjusted prevalence of hypertension at 40.5% for non-Hispanic black, 31.1% for non-Hispanic white and 28.6% for Mexican American men.^1,3

Socio-familial factors in childhood have been found to play an important role in the etiology of hypertension. Childhood socioeconomic position is a well-established determinant of BP in adulthood.⁴ Higher parental socioeconomic position during childhood, i.e., greater household income, parental education, and occupational status have been associated with lower DBP and SBP in European men and women independent of adult socioeconomic position and health behaviors.^5–9 To our knowledge only one study to date has shown that this association exists among African American men.¹⁰

The relationship between adult BP and other early-life socio-familial factors such as childhood family structure or living arrangements has been less studied. One study in Sweden suggests that there may be a relationship between childhood family living arrangements and adult BP. Swedish men and women who lived with their natural parents up to age 16 as compared with those who did not, have a lower prevalence of circulatory illness, a category that included cardiovascular ailments such as chest pains, heart attacks, heart weakness, dizziness and hypertension.¹¹ Based on the Swedish report and the low rate of African American children who grow-up in households with both parents,¹² we hypothesized that this early childhood disadvantage will have a negative influence on mean BP and risk of hypertension among African American men. Therefore, we examined whether childhood family living arrangements are independently associated with mean BP and hypertension in a sample of unrelated African American men in the Howard University Family Study.

Methods

Study Population

This cross-sectional study utilized 2001–2008 data from the Howard University Family Study (HUFS), a consecutive sample of families and unrelated African Americans from the Washington, DC metropolitan area. By not conditioning on any extreme phenotype, HUFS was purposely designed to shed light on the genetic and environmental basis of common complex traits including hypertension, obesity, diabetes and other health outcomes. Study participants were enrolled through multiple methods including door-to-door canvassing, advertisements in local print media, and at health fairs and other community gatherings.¹³ All study enrollees had to self-identify as African American, and were excluded from HUFS if they were under 14 years of age, pregnant at the time of enrollment and had acute febrile illness or an acute pain episode at the time of clinical examination. The present study capitalized on the socio-demographic information collected by interview during the clinical examinations of 515 unrelated African American male participants aged 20 years and older to examine the impact of childhood family living arrangements on BP and hypertension.

Measures

The main outcomes under study were mean SBP, mean DBP, mean pulse pressure (PP), mean arterial blood pressure (MAP) and hypertension. SBP and DBP were measured in mmHg using an oscillometric device (Omron HEM-431C) with a standard or large adult cuff (arm circumference 9–13 or 13–17 inches, respectively) in the seated position. Three BP readings were taken by trained observers with a 5-minute interval between each reading. Please access http://hyper.ahajournals.org for BP standardization conditions. Reported SBP and DBP are the average of the second and third readings. Pulse Pressure was defined as systolic minus diastolic blood pressure, and MAP was calculated as DBP + 1/3 (SBP – DBP). Hypertension was defined as an SBP ≥ 140 mmHg, and/or a DBP greater ≥ 90 mmHg, and/or currently on prescribed treatment with anti-hypertensive medication.^14,15

The socio-familial independent variables analyzed in this study were derived from responses to the following question - “How many years of your life have you spent living with (1) both parents together, (2) your mother only and (3) your father only?” Based on the numbers of years reported, a dichotomous variable, “none” vs. “any,” and an ordinal variable, characterized as “none,” “1–12 years,” “13–19 years” and “20+ years” were created for years lived with both parents.

Covariates utilized for statistical adjustment due to their known associations with SBP, DBP and hypertension included measures of demographic information, socioeconomic position, health behaviors, anthropometry, related cardiovascular medical conditions and family history, and blood micronutrient chemistries. Demographic information included age and a categorical variable of marital status, defined as “married or living with a partner,” “never married,” and “previously married.” Measures of socioeconomic position were education, categorized into “< high school (HS),” “HS,” and “>HS;” and current employment status dichotomized into “no” vs. “yes.” Health behaviors or lifestyle measures reported by the study participants included smoking status, which consisted of three categories, “non-smoker,” “smoker,” and “ex-smoker;” alcohol intake dichotomized into “none” vs. “any;” and physical activity specified as “none,” “moderate or vigorous activity” (for at least 10 minutes causing an increase in breathing or heart rate), and “other” (which included walking for at least 10 minutes and strength training).

Anthropometric indicators were ascertained via standard methods:¹³ weight was measured in light clothes on an electronic scale to the nearest 0.1 kg and height was measured with a stadiometer to the nearest 0.1 cm. Body mass index (BMI) was calculated as weight (kg) divided by the square of the height (m²).

Type 2 diabetes, a condition well-known to be associated with hypertension, was defined by American Diabetes Association (ADA) guidelines,¹⁶ fasting plasma glucose ≥ 126 mg/dl and/or currently on prescribed therapy for diabetes; the analysis sample was subsequently dichotomized by diabetes status as “no” and “yes.” Obesity was defined as BMI 30 kg/m², and was also categorized into “no” and “yes.” Family history of hypertension was specified as “no” and “yes” for any hypertension diagnosis in first-degree relatives. Current anti-hypertensive medication use was classified into “no,” “yes,” and missing. Finally, blood micronutrient chemistries obtained from overnight fasting samples were included, i.e., HDL, LDL and total cholesterol (mg/dl), triglycerides (mg/dl), and the sodium-potassium ratio defined as sodium (MEQ/L) divided by potassium (MEQ/L).

Statistical Analysis

Data management and all statistical analyses were conducted using SAS Version 9.2.¹⁷ General linear regression models (PROC GLM in SAS) describing differences in mean SBP, DBP, and MAP associated with men’s childhood family living arrangements were analyzed prior and subsequent to statistical adjustment for associated demographic, socioeconomic, lifestyle, anthropometric, medical history and micronutrient covariates. Multiple logistic regression models were used to assess the strength of the independent associations between the socio-familial measures and hypertension before and after covariate adjustment. Odds ratios (ORs) and 95% Confidence Intervals (CIs) are reported. Statistical significance was judged at p-values ≤ 0.05.

Results

African American men had a mean SBP of 133.0 mmHg, mean DBP of 84.0 mmHg, mean PP of 49.0 mmHg, and mean MAP of 100.0 mmHg; 54% were hypertensive (Table 1). 68% of the men lived with both parents at some point during their childhood. Their range in age was from 20 to 86 years, with a mean age of 48 years. The men were disadvantaged, as characterized by their high rates of unemployment (55%) and poor health behaviors, i.e., smoking (64%). Please see http://hyper.ahajournals.org for additional sample characteristics. Their prevalence of diabetes (12%) and obesity (33%) were similar to national estimates for African American men.¹⁸ Among the 32% of the men who never lived with both parents, 80% lived with their mothers alone, 7% lived with their fathers alone and 13% lived with other adults. Among the 152 men on antihypertensive medications, 38% were on combination therapy.

Table 1.

Selected characteristics of the African-American men in this study (N = 515): The Howard University Family Study.

Systolic BP, mean (SD), mmHg	133.0 (20.3)
Diastolic BP, mean (SD), mmHg	83.6 (13.8)
Pulse Pressure, mean (SD), mmHg	49.4 (13.4)
Mean Arterial Pressure, mean (SD), mmHg	100.1 (15.0)
Hypertension, No. (%)
No	239 (46.4)
Yes	276 (53.6)
Years Lived w/ Both Parents, No. (%)
None	166 (32.4)
1–12 Years	99 (19.3)
13–19 Years	161 (31.5)
20+ Years	86 (16.8)
Age, mean (SD), years	47.8 (11.9)

Table 2 presents the covariate-adjusted mean differences in SBP and DBP by childhood family living arrangements. Please see http://hyper.ahajournals.org for all age-adjusted tables. Men who lived with both of their parents at any point during their lifetime had an age-adjusted 4.9 mmHg decreased SBP compared to men who never lived with both of their parents (p-value = 0.008). After covariate-adjustment, the association remained. Furthermore, living with both parents for 1 to 12 years conferred an age-adjusted protective advantage of 7.6 mmHg decreased SBP compared to never living with both parents; (p-value = 0.002). After covariate-adjustment, men who lived with both parents for 1 to 12 years of their lives had a 6.5 mmHg lower SBP compared to men who never lived with both parents (95% CI: −10.99, −1.95). No significant differences in mean DBP were found between the socio-familial measures and mean DBP after age and covariate adjustment.

Table 2.

Covariate-adjusted differences in mean systolic and diastolic BP by childhood family living arrangements

Socio-familial Measures	SBP * Mean Difference (95%CI)	DBP † Mean Difference (95%CI)
Years Lived w/ Both Parents
Any vs. None	−4.40 (−7.84, −0.96)	−0.87 (−3.16, 1.43)
Years Lived w/ Both Parents
1–12 Years vs. None	−6.47 (−10.99, −1.95)	−1.68 (−4.70, 1.34)
13–19 Years vs. None	−3.36 (−7.38, 0.67)	−0.29 (−3.00, 2.43)
20+ Years vs. None	−3.70 (−8.55, 1.15)	−0.86 (−4.08, 2.35)

^*Mean differences in SBP (mmHg) adjusted for age, physical activity, family history of hypertension, anti-hypertensive medication use, sodium-potassium ratio, body mass index, and HDL cholesterol.

^†Mean differences in DBP (mmHg) adjusted for age, smoking, and anti-hypertensive medication use.

Men who lived with both of their parents during their lifetime had an age-adjusted 3.7 mmHg decreased PP compared to men who never lived with both of their parents (p-value = 0.003). Table 3 shows that after covariate-adjustment, this association remained. Furthermore, living with both parents for 1 to 12 years granted an age-adjusted protective advantage of 5.2 mmHg decreased PP compared to never living with both parents (p-value = 0.001). Living with both parents for 13–19 years also provided a benefit in PP. After covariate-adjustment, the associations between childhood family living arrangements and PP remained. No statistically significant associations were found between the dichotomous measure of living with both parents and MAP. There was however a statistically significant age-adjusted mean difference in MAP and the ordinal measure; men who lived with both parents for 1 to 12 years had a 4.1 mmHg lower MAP compared to men who never lived with both parents (p-value = 0.02). After covariate-adjustment, this association remained.

Table 3.

Covariate-adjusted differences in mean pulse pressure and arterial pressure by childhood family living arrangements

Socio-familial Measures	PP * Mean Difference (95%CI)	MAP † Mean Difference (95%CI)
Years Lived w/ Both Parents
Any vs. None	−3.89 (−6.28, −1.51)	−1.97 (−4.44, 0.51)
Years Lived w/ Both Parents
1–12 Years vs. None	−5.38 (−8.48, −2.28)	−3.28 (−6.56, 0.00)
13–19 Years vs. None	−3.12 (−5.93, −0.32)	−1.27 (−4.17, 1.64)
20+ Years vs. None	−3.32 (−6.65, 0.02)	−1.64 (−5.14, 1.86)

^*Mean differences in PP (mmHg) adjusted for age, employment, smoking, physical activity, family history of hypertension, anti-hypertensive medication use, sodium-potassium ratio, and body mass index.

^†Mean differences in MAP (mmHg) adjusted for age, anti-hypertensive medication use, sodium-potassium ratio, body mass index, and HDL cholesterol.

Additional analyses were performed among a subset of 271 men who reported no anti-hypertensive medication use, and analogous results were found. Among men who were not prescribed anti-hypertensive medications, men who lived with both parents for 1 to 12 years of their lives had, compared to men who never lived with both parents, 5.4 mmHg decreased SBP (95% CI: −11.20, 0.41), 5.2 mmHg decreased PP (95% CI: −8.60, −1.70), and 2.5 mmHg decreased MAP (95% CI: −7.02, 2.00) after adjusting for related covariates.

Table 4 presents the covariate-adjusted odds ratios for hypertension. No associations were found between hypertension and the dichotomous socio-familial measure. However significant covariate-adjusted associations were found between hypertension and the ordinal measure of living with both parents. Men who lived with both parents for 1 to 12 years of their lives had a 46% decreased odds of hypertension compared to men whom never lived with both parents in their lifetime, covariate-adjusted OR = 0.54 (95% CI: 0.30, 0.99).

Table 4.

Covariate-adjusted Odds Ratios (ORs) for hypertension by childhood family living arrangements

Socio-familial Measures	OR (95%CI) *
Years Lived w/ Both Parents
None	1 [Reference]
Any	0.82 (0.52, 1.29)
Years Lived w/ Both Parents
None	1 [Reference]
1–12 Years	0.54 (0.30, 0.99)
13–19 Years	0.85 (0.50, 1.45)
20+ Years	1.27 (0.66, 2.44)

^*Odds ratios for hypertension adjusted for age, education, marital status, employment, smoking, sodium-potassium ratio, HDL cholesterol, obesity, diabetes, and family history of hypertension.

Discussion

Summary of Findings

Our study found strong, protective effects of living with both parents for SBP, PP, MAP and hypertension among African American men. In particular, living with both parents between 1 and 12 years of one’s life conferred a significant advantage of decreased mean SBP, PP, and MAP and lowered risk for hypertension among men independent of adulthood measures of socioeconomic position, health behaviors, anthropometric measures, micronutrient blood chemistries, and medical history.

Analogous results have been found among European ancestral populations for a similar socio-familial factor, i.e., being born to unmarried parents. Swedish men who were born outside of marriage and who are unmarried at 55 to 80 years have a two-fold risk of mortality from ischemic heart disease compared to men who were born to married parents.¹⁹ This increased risk of being born to unmarried parents transmits across generations; sons and grandsons of Swedish men born outside of marriage also have increased risk of circulatory system diseases including essential hypertension.²⁰

Childhood family living arrangements have been found to be associated with other physical health outcomes in adulthood. Living with both parents at age 7 is associated with better self-rated health in Finnish men and women.²¹ Mortality rates are lowest among Finnish adolescents aged 17 to 19 years who live with married parents as compared to those teens who live with either cohabiting parents, single parents or no parents.²² Swedish men and women who lived with their natural parents up to age 16 have decreased mid-life mortality and reduced reports of physical and mental illness compared to those who never lived with both parents.^11,23

Middle-aged US men who lived at age 15 with both of their biological parents and had biological mothers who were not in the workforce, have the lowest risk for death compared to men with other familial living arrangements at age 15 years, including having biological parents with mothers who worked outside of the home, step-parents, single parents, no parents, and other family members.²⁴ Furthermore, racial differences in socio-familial conditions, i.e., lower levels of parental occupational status and decreased likelihood of living with both biological parents among African American men compared to white men, have been found to explain part of the black-white gap in overall mortality.²⁵

Explanatory Hypotheses

We propose three key explanations for our study results. First, living with both parents may be a proxy for childhood socioeconomic position. African American children who do not live with both parents are at a considerable socioeconomic disadvantage compared to those who live with both parents. Compared to African American children who reside with two parents, children who live with their mothers alone are approximately three times more likely to live in poverty, and those who live with neither parent or their father alone have an approximate two-fold risk of being poor.²⁶ Furthermore, children born to single mothers who do not reside with their child’s biological father have a four-fold risk of being born into poverty compared with children born to married, cohabiting parents.²⁷ Since the majority of the HUFS male participants who never lived with both parents resided with their mothers only, we can assume that these men were likely to experience the most adverse childhood socioeconomic conditions, established risk factors for elevated BP and hypertension.^5–9 Additionally, unfavorable childhood family living arrangements can operate directly or indirectly via low childhood socioeconomic position to decrease weight at birth and increase childhood BMI, recognized determinants of elevated BP and hypertension in adulthood.^28,29

Living with both parents during one’s childhood can also influence BP and hypertension via long-term unemployment stressors. Children who do not grow up with their natural parents compared to those who are raised by their two biological parents are significantly less likely to find and maintain steady employment as young adults.³⁰ The appreciably high unemployment rate among the HUFS male participants who never lived with both parents may implicate long-term, chronic unemployment ensuing from a disadvantageous childhood socio-familial environment. Although our statistical models included current employment status as a covariate, no data was available to adjust for duration, stability/security of employment and occupational status – established risk factors for BP and hypertension in men.^31–33 Longitudinal research will be necessary to unravel the various psychosocial and socioeconomic mechanisms through which adverse childhood family living arrangements elevate BP and risk for hypertension in African American men.

Third, the greatest protective effect on mean SBP, PP, MAP and hypertension was found for living between 1–12 years with both parents. This suggests that there may be a critical period and potential epigenetic mechanism through which this early life socio-familial factor operates to influence adult BP.³⁴ Animal models have shown that unfavorable early life experiences alter hypothalamic-pituitary-adrenal (HPA) axis stress reactivity throughout life via changes in glucocorticoid receptor expression by DNA methylation; this heightened stress reactivity elevates plasma cortisol concentrations, consequently raising BP throughout childhood and adulthood.^35–37 Research studies that explore DNA methylation patterns stratified by early childhood environment may determine whether epigenetic pathways can explain our study results.

Strengths and Limitations

To our knowledge this is the first study to document an association between childhood family living arrangements and SBP, PP, MAP and hypertension in a sample of African American men. Though not a large population-based study, the HUFS sample had enough statistical power to detect associations while adjusting for a number of BP covariates. We must note, however, some limitations with our study design and analyses. First, the cross-sectional nature of the HUFS data does not allow for any conclusion about causality to be drawn. The cross-sectional design also permits possible recall bias in measuring familial living arrangements decades prior, allowing for potential exposure misclassification in one’s childhood socio-familial environment. Nevertheless, there is no expectation that any bias in recall would systematically differ according to BP or hypertensive status, and we surmise that our results may have underestimated the true associations between childhood family living arrangements and adult BP and hypertension.

Second, the socio-familial exposure variables measure the number of years and do not measure the timing of those years that respondents spent living with both parents. This may not directly correspond to actual time across the life-course. However, data from the Fragile Families and Childhood Wellbeing Study show that African American parents tend to live together at the birth of their children and separate over time.²⁷ Therefore, the ordinal measure of residence with both parents is likely to parallel years across the life-course for African American men.

As previously mentioned, living with both parents during childhood may be a proxy for other important unmeasured factors which directly influence adult BP, such as childhood socioeconomic position or adiposity. Future research will need to ascertain whether childhood family living arrangements exert their effects on adult BP independent of these other early life socio-familial factors. Furthermore, investigations having more detailed measures of childhood family structure are needed. Grandparents are known to play an important role in providing care during early life. Since African Americans are most likely to have lived with at least one grandparent during childhood compared to all other racial/ethnic groups in the US,¹² additional study will be required to determine the extent to which residence with grandparents during childhood correlates with BP later in the life-course.

Misclassification of covariates due to the use of crude lifestyle measures within the analysis, i.e. physical activity and alcohol intake, may have also biased the results; nevertheless we anticipate any effects of confounder misclassification to be minimal as we do not expect it to be differentially distributed according to childhood family living arrangements. Other unmeasured confounders of hypertension in adulthood, including fruit and vegetable intake and adherence to anti-hypertensive therapy were not examined; however, analyses on a sub-sample of men who were not on anti-hypertensive medications revealed similar associations between the ordinal socio-familial measure and BP.

The small study sample size, the noticeably higher prevalence of hypertension among African American men in HUFS as compared to African American men nationwide,¹ and the two and one-half increased prevalence of smoking in HUFS relative to nationwide estimates for African-American men,¹⁸ may suggest possible selection bias within the sample for which the effects on our study results is unknown. However, the mean systolic and diastolic blood pressures among the HUFS men are remarkably similar to those found among black male participants in historical cohorts such as the Evans County Heart Study.³⁸ Moreover, studies of African American men in other urban cities reveal that men residing in these locations experience considerable health and socioeconomic disadvantages similar to HUFS male participants. In the Northern Manhattan Study three-quarters of African American men and women 40 years and older have hypertension, 60% have smoked during their lifetime and 23% have <HS education.³⁹ Therefore, our study findings are representative of and directly applicable to African-American men in urban metropolitan areas across the US.

Perspectives

These study findings have several implications for research and public health interventions designed to ameliorate hypertension among African American men. First, our results provide preliminary evidence that when describing its influence on blood pressure later in adulthood, the early-life socio-environment should include measures beyond childhood socioeconomic position, i.e., childhood family structure. Living with both parents in early life may identify a critical period in human development where a nurturing socio-familial environment can have profound, long-lasting influences on blood pressure through epigenetic processes and/or initiating cumulative sociocultural advantages across the life-course. Secondly, continued examination of critical periods across the life course during which socio-environmental stressors modify expression of genes that elevate BP is a worthwhile endeavor.⁴⁰ Forthcoming studies will assess differential patterns of DNA methylation by childhood family structure. Finally, replication of these findings within large-scale US population-based studies of ethnically-diverse men and women will be needed to confirm the long-term influence of childhood family living arrangements on later-life blood pressure and hypertension.

Novelty and Significance.

What is new?

• Study of the association between childhood family living arrangements and later blood pressure and hypertension among African American men.

What is relevant?

• Living with both parents during childhood protects against elevated systolic blood pressure, mean arterial pressure, pulse pressure and hypertension in African American men.

• Family structure may be another important early life socio-environmental factor in blood pressure, potentially operating through socioeconomic factors across the life-course and/or epigenetic mechanisms during critical periods of childhood development to increase blood pressure and risk for hypertension later in the life-course.

Summary

• The childhood socio-environment’s long-term influence on adult blood pressure and hypertension suggests that policies, programs and interventions designed to create and maintain stable families for African American children may lower risk for hypertension in later life.