Abstract
Purpose
We examined if receipt of public assistance during childhood lowered risk for hypertension by mid-life in a cohort of African Americans in the southeastern United States.
Methods
We utilized multiple logistic regression models to assess the relationship between receipt of public assistance during childhood and adult hypertension among 405 male and 737 female adult participants enrolled between 1988 and 2001 in the Pitt County Study, a community-based prospective cohort study of African Americans in North Carolina. Statistical analyses were adjusted for child and adult socio-demographic measures as well as adult psychosocial and lifestyle factors.
Results
Women who grew up in economically disadvantaged families and who received public assistance during childhood had a 65% decreased odds of hypertension by mid-life compared to women similarly disadvantaged in childhood but who did not receive public assistance, OR = 0.35; 95%CI: (0.14, 0.86). No association was observed for African American men.
Conclusions
Receipt of anti-poverty federal assistance during childhood was associated with reduced risk for hypertension by mid-life among African American women. It is possible that social expenditures on public assistance programs for families in need could produce long-term health benefits for children.
Keywords: African Americans, hypertension, public assistance, poverty, public policy
INTRODUCTION
In addition to rising income inequality and decreasing median household income, the United States (US) has also seen in recent decades an increase in the number of families participating in public assistance programs, including the Supplemental Nutrition Program (SNAP), Temporary Assistance for Needy Families (TANF), General Assistance, Housing Assistance, Medicaid and Supplemental Security Income (SSI) (1, 2). In 2012, for example, approximately 52.2 million Americans received means-tested federal assistance each month (2). Children have the most need for a social safety net, given that the average monthly participation in governmental assistance programs was 39% among children less than 18 years old in 2012 (2).
Only 61% of families in poverty, as defined by household income below standard thresholds adjusted for inflation (1), received federal assistance in 2012. This was the case despite clear evidence of improved psychosocial, cognitive and health outcomes among socioeconomically disadvantaged children who participate in public assistance programs such as the Food Stamp program (SNAP) and the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) (3–5). In addition, results from an economic study indicate that childhood participation in public assistance programs could provide a long-lasting positive impact on cardiovascular health, specifically lowering the risk of developing metabolic syndrome in adulthood (6). This finding underscores the importance of continued research into the potential contribution of federal anti-poverty programs to foster better health across the life course.
African Americans have the lowest median household income, the highest prevalence of unemployment and childhood poverty (1, 7), and the highest prevalence of hypertension relative to other US racial/ethnic groups: 44.9% for men and 46.1% for women (8). These multiple disadvantages faced by African Americans underscore the need for studies that shed light on social policies that not only diminish the severity of socioeconomic distress in this population, but which may also lower long-term risk for chronic diseases, such as high blood pressure. In the current study, we investigated whether receipt of public assistance during childhood, in the face of need, was associated with decreased risk for adult hypertension in a Southern African American cohort.
MATERIAL AND METHODS
Sample
The Pitt County Study (PCS) is a community-based, probability sample of African Americans in eastern North Carolina who were initially interviewed in 1988 and re-interviewed in 2001. The original study objective was to examine the contribution of social, economic, psychological, and behavioral risk factors for hypertension and related disorders in working and middle-class African Americans aged 25 to 50 years in 1988 (9–12). At baseline, neighborhoods with high concentrations of African Americans, especially those with middle class African Americans, were oversampled to insure a socioeconomically diverse study population (13). The research protocol was approved by the Human Subjects Institutional Review Board of the University of Michigan.
The 1988 baseline sample included 1773 African Americans (661 women and 1112 men). In 2001, information on life-course social and economic resources from early childhood through 2001 was obtained via interview for 1221 (428 men and 793 women). Further details on the 1988 baseline sample and household interview, and the 2001 follow-up interview, can be found elsewhere (12, 13). Analyses for the current study were based on 405 men and 737 women who had complete exposure and outcome data.
Measures
Blood Pressure
Hypertension, defined as a systolic blood pressure ≥ 140 mmHg, and/or a diastolic blood pressure greater ≥ 90 mmHg, and/or currently taking anti-hypertensive medication in 1988 was the study outcome. Blood pressure (BP) in 1988 was measured by a standard mercury sphygmomanometer on the study participant’s right arm; three BP readings were taken. For analytical purposes, BP was defined as the average of the second and third readings. The fifth-phase Korotkoff sound indicated diastolic blood pressure.
Childhood Socio-demographic Measures
In 2001, a computerized Event History Calendar (EHC) was administered to the PCS study participants to measure the childhood social environment. The EHC maximizes recall accuracy of distant autobiographical information by using more easily remembered landmark events (e.g., name and residence of the primary parental figure at a given age, year started school, age at first paid job, etc.) to ease recollection of more easily forgotten experiences (14, 15). The EHC assessed receipt of childhood public assistance, “yes” vs. “no,” and childhood household deprivation, defined by any occurrence of food insecurity, lack of heat, electricity, or plumbing from birth to 13 years of age. The primary exposure variable was a 3-category measure combining information on childhood public assistance and childhood household deprivation: (1) receipt of public assistance (2) no receipt of public assistance/household deprivation and (3) no receipt of public assistance/no household deprivation.
Other childhood socio-familial variables were included as covariates to estimate the independent influence of childhood receipt of public assistance on hypertension in early-middle adulthood. Residence from birth through age 18 in a two-adult household, (yes vs. no) characterized family structure composition. Stability of parental guardianship was determined by the number of guardianship changes from birth to 18 years (none vs. any) and was dichotomized into “yes” vs. “no.” Two measures were used to define childhood socioeconomic position from birth to 13 years of age: parental occupation, denoted by the highest ranking job held by a parental guardian of either sex, with jobs categorized as “skilled,” “farmer,” and “unskilled or housewife;” and parental education, defined by the highest level of education of a parent/guardian of either sex, with education categorized as “less than high school (<HS),” high school or greater (HS+), and “missing” (12).
Adult Measures
Covariates utilized for statistical adjustment included demographic information, socioeconomic position, health behaviors, anthropometry, and psychosocial factors measured in 1988. Demographics included age and marital status, specified as married “yes” vs. “no.” Measures of socioeconomic position were education (< HS, HS, and >HS), currently unemployed (yes vs. no), and occupation (laborers, operatives/blue collar and craft/white collar) (12).
Health behaviors included smoking status, dichotomized as “current smoker/non-smoker,” alcohol consumption dichotomized as “current drinker/non-drinker,” and physical activity, denoted by a response (yes vs. no) to the question, “do you exercise at least 3 times per week sufficient to breathe hard and sweat?” Anthropometric indices were the ratio of waist to hip circumference (WHR), and body mass index (BMI) calculated as weight (kg) divided by the square of the height (m2) (13).
Psychosocial covariates included John Henryism, defined as a strong behavioral predisposition to engage in high-effort coping with difficult social and economic stressors. Measured via the 12-item John Henryism Active Coping Scale, total scores were divided at the median (high vs. low) for analytical purposes (16). Instrumental social support was measured by a 6-item scale, and emotional support was measured by a 3-item scale; both were dichotomized (always/most of the time vs. sometimes/never) (13). Perceived Stress, a correlate of BP in the PCS, was measured by an 8-item scale (17), with scores dichotomized at the median (low vs. high) for analytic purposes.
Analysis
Analyses were weighted to account for the oversampling of middle-class households at baseline and nonresponse to both the 1988 and 2001 surveys, and thus provided more conservative estimates of standard errors (18). The characteristics of study participants were described using means for continuous variables and proportions for categorical variables. To estimate associations between adult hypertension and the public assistance and household deprivation measure, multiple logistic regression models were utilized. Because of a statistically significant sex interaction term (p-value < 0.0001), sex-stratified results are presented. Logistic regression was performed to add early-life and more proximal risk covariates individually, or as a group, to assess their contributions to the odds for hypertension. Odds ratios (ORs) and 95% Confidence Intervals (CIs) are reported. Statistical significance was set at α = .05 level, two-tailed tests. All analyses were performed using SAS software (Version 9.4; SAS Institute, Inc., Cary, NC).
RESULTS
Differences in selected characteristics between those lost to follow-up and those who completed the early childhood assessment in 2001 for the current analyses are summarized in Table 1. Among men, those lost to follow-up were more disadvantaged, in that they had a higher prevalence of unemployment, smoking, high stress and lower levels of instrumental support than men who remained in the final analytic sample. Among women, with the exception of age, there were no differences between those lost to follow-up and those in the analytic sample. There was no difference in the prevalence of hypertension between non-responders and the final study participants.
Table 1.
Selected Characteristics of Study Participants Lost to Follow-Up and in the Final Analytic Sample: The Pitt County, North Carolina Study
| Men | Women | |||||
|---|---|---|---|---|---|---|
| Non- Responders (N = 256) |
Analytic Sample (N = 405) |
P-valuea | Non- Responders (N = 376) |
Analytic Sample (N = 736) |
P-valuea | |
| Age, mean (SE) | 35.75 (0.46) | 36.16 (0.35) | 0.4753 | 35.14 (0.35) | 36.21 (0.27) | 0.0178 |
| Married,% | 48.44 | 62.62 | 0.0003 | 38.83 | 41.85 | 0.3327 |
| Unemployed,% | 11.72 | 6.17 | 0.0119 | 20.21 | 17.12 | 0.2057 |
| Education,% | 0.6072 | 0.9806 | ||||
| <HS | 32.03 | 31.60 | 30.05 | 30.57 | ||
| HS | 41.02 | 38.02 | 40.43 | 40.35 | ||
| >HS | 26.95 | 30.37 | 29.52 | 29.08 | ||
| Occupation,% | 0.1462 | 0.2769 | ||||
| Laborer | 33.47 | 26.43 | 36.59 | 36.86 | ||
| Operatives | 31.47 | 33.42 | 30.73 | 34.71 | ||
| Craft/White Collar | 35.06 | 40.15 | 32.68 | 28.43 | ||
| BMI (kg/m2), mean (SE) | 25.78 (0.25) | 26.45 (0.22) | 0.0524 | 28.79 (0.36) | 29.48 (0.25) | 0.1103 |
| WHR, Mean (SE) | 0.88 (0.004) | 0.89 (0.003) | 0.6562 | 0.86 (0.004) | 0.85 (0.003) | 0.3736 |
| Smoker,% | 51.97 | 44.20 | 0.0518 | 33.24 | 34.15 | 0.7629 |
| Physically Active,% | 53.33 | 58.02 | 0.2369 | 32.80 | 36.34 | 0.2431 |
| Drinker,% | 60.55 | 62.47 | 0.6204 | 42.55 | 38.45 | 0.1862 |
| High John Henryism,% | 46.09 | 48.64 | 0.5228 | 46.81 | 48.10 | 0.6838 |
|
High Instrumental Support,% |
25.00 | 33.83 | 0.0163 | 28.46 | 30.57 | 0.4663 |
|
High Emotional Support,% |
35.69 | 34.57 | 0.7693 | 39.36 | 41.85 | 0.4252 |
| High Stress,% | 24.71 | 17.12 | 0.0179 | 28.19 | 29.76 | 0.5875 |
| Hypertension,% | 30.08 | 32.35 | 0.5408 | 24.33 | 25.14 | 0.7695 |
Note: BMI = body mass index, WHR = waist-hip ratio, SE = standard error.
Chi-square Test (%), T-test (mean) of difference.
As shown in Table 2, men who received public assistance were least likely to live as a child within an intact family and two-adult household, and were most likely to have parents who attained less than a high school education compared with those who did not receive public assistance during childhood. In adulthood, they were most likely to attain low levels of education, and have the highest mean BMI. There was no significant difference in mean childhood household deprivation score between men who received public assistance during childhood and those who did not receive public assistance but who, nevertheless, reported household material deprivation.
Table 2.
Selected Characteristics by Childhood Public Assistance of African American Men: The Pitt County, North Carolina Study
| Men (N = 405) |
||||
|---|---|---|---|---|
| Public Assistance (N = 51) |
No Public Asst/ HH Deprivation (N = 233) |
No Public Asst/ No HH Deprivation (N = 121) |
P-valueh | |
| Childhooda | ||||
| HH Deprivation, mean (SE) | 1.36 (0.15) | 1.47 (0.05) | -- | 0.4776 |
| Intact Family,c % | 65.27 | 88.45 | 82.81 | 0.0018 |
| Two Adult Household,c % | 36.84 | 70.53 | 54.70 | 0.0001 |
| Parent/Guardian Education, b % | 0.0117 | |||
| Missing | 27.92 | 35.97 | 18.00 | |
| <HS | 50.90 | 42.33 | 43.99 | |
| HS+ | 21.18 | 21.70 | 38.01 | |
|
Parent/Guardian Occupation, b,d % |
0.0002 | |||
| Skilled | 12.71 | 12.03 | 29.23 | |
| Farmer | 32.13 | 40.70 | 15.26 | |
| Unskilled/Housewife | 55.16 | 47.27 | 55.50 | |
| Adulthood | ||||
| Age, mean (SE) | 31.66 (0.94) | 36.37 (0.52) | 33.45 (0.69) | <0.0001 |
| Married,% | 68.24 | 61.20 | 56.50 | 0.4719 |
| Unemployed,% | 10.17 | 5.37 | 4.44 | 0.4169 |
| Education,% | 0.0048 | |||
| <HS | 44.89 | 33.03 | 24.20 | |
| HS | 27.86 | 44.39 | 33.45 | |
| >HS | 27.25 | 22.59 | 42.35 | |
| Occupation, e % | 0.8703 | |||
| Laborer | 29.02 | 28.35 | 22.88 | |
| Operatives | 30.56 | 34.07 | 33.92 | |
| Craft/White Collar | 40.42 | 37.57 | 43.20 | |
| BMI (kg/m2), mean (SE) | 27.49 (0.75) | 26.38 (0.31) | 25.55 (0.41) | 0.0546 |
| WHR, Mean (SE) | 0.89 (0.01) | 0.89 (0.00) | 0.90 (0.01) | 0.6505 |
| Smoker, % | 46.61 | 42.79 | 52.48 | 0.3455 |
| Physical Activity, % | 72.86 | 60.05 | 59.01 | 0.2943 |
| Drinker,% | 67.10 | 60.61 | 70.87 | 0.2530 |
| High Instrumental Support,f% | 19.00 | 37.29 | 33.36 | 0.0750 |
| High Emotional Support,f% | 20.70 | 32.33 | 37.32 | 0.1797 |
| High John Henryism,g% | 54.89 | 44.16 | 49.69 | 0.4130 |
| High Stress,g% | 31.93 | 40.30 | 44.08 | 0.4562 |
| Hypertension,% | 29.46 | 33.62 | 28.26 | 0.6411 |
Note: Data were weighted for oversampling and nonresponse.
Measured in 2001 via Event History Calendar.
Birth to 13 years.
Birth to 18 years.
US Census Occupations, (see Ref 7 for full citation).
Hollingshead Scale (see Ref 7 for full citation).
Support always/most of the time.
Score above the median
Rao-Scott chi-squared test for comparison (%), T-test (mean) of difference.
Women who received public assistance also experienced the highest levels of disadvantage in the other childhood socio-familial factors compared with those who did receive public assistance during childhood (Table 3). There was also no significant difference in mean household deprivation score between women who received public assistance during childhood and those who did not receive public assistance but nevertheless reported household material deprivation.
Table 3.
Selected Characteristics by Childhood Public Assistance of African American Women: The Pitt County, North Carolina Study
| Women (N = 737) |
||||
|---|---|---|---|---|
| Public Assistance (N = 73) |
No Public Asst/ HH Deprivation (N = 430) |
N o Public Asst/ No HH Deprivation (N = 234) |
P-valueh | |
| Childhooda | ||||
| HH Deprivation, Mean (SE) | 1.27 (0.13) | 1.45 (0.03) | -- | 0.1843 |
| Intact Family,c % | 72.51 | 83.17 | 85.69 | 0.1093 |
| Two Adult Household,c % | 44.50 | 68.06 | 56.78 | 0.0010 |
|
Parent/Guardian Education, b,d % |
0.0025 | |||
| Missing | 18.29 | 25.62 | 19.80 | |
| <HS | 61.60 | 58.41 | 48.81 | |
| HS+ | 20.11 | 15.97 | 31.39 | |
|
Parent/Guardian Occupation, b,d % |
<0.0001 | |||
| Skilled | 28.57 | 10.74 | 25.02 | |
| Farmer | 13.71 | 45.06 | 20.06 | |
| Unskilled/Housewife | 57.71 | 44.19 | 54.91 | |
| Adulthood | ||||
| Age, Mean (SE) | 34.25 (0.98) | 36.71 (0.42) | 32.99 (0.45) | <0.0001 |
| Married,% | 36.45 | 47.00 | 36.03 | 0.0395 |
| Unemployed,% | 15.36 | 14.81 | 16.92 | 0.8271 |
| Education,% | 0.0045 | |||
| <HS | 27.55 | 32.85 | 17.39 | |
| HS | 39.44 | 40.14 | 44.05 | |
| >HS | 33.01 | 27.00 | 38.56 | |
| Occupation, e % | 0.3557 | |||
| Laborer | 39.91 | 34.00 | 29.93 | |
| Operatives | 30.81 | 37.26 | 33.01 | |
| Craft/White Collar | 29.28 | 28.74 | 37.05 | |
| BMI (kg/m2), Mean (SE) | 28.14 (0.82) | 30.07 (0.41) | 28.65 (0.48) | 0.0247 |
| WHR, Mean (SE) | 0.86 (0.01) | 0.85 (0.00) | 0.85 (0.01) | 0.4386 |
| Smoker,% | 25.21 | 32.13 | 36.24 | 0.3213 |
| Physical Activity,% | 31.85 | 39.87 | 29.11 | 0.0471 |
| Drinker,% | 32.58 | 32.56 | 43.83 | 0.0378 |
| High Instrumental Support,f% | 36.63 | 28.85 | 32.62 | 0.4544 |
| High Emotional Support,f% | 49.36 | 38.42 | 40.61 | 0.3434 |
| High John Henryism,g% | 46.33 | 47.65 | 47.66 | 0.9847 |
| High Stress,g% | 48.39 | 52.67 | 54.46 | 0.7562 |
| Hypertension,% | 11.12 | 29.10 | 19.78 | 0.0019 |
Note: Data were weighted for oversampling and nonresponse.
Measured in 2001 via Event History Calendar.
Birth to 13 years.
Birth to 18 years.
US Census Occupations, (see Ref 7 for full citation).
Hollingshead Scale (see Ref 7 for full citation).
Support always/most of the time.
Score above the median
Rao-Scott chi-squared test for comparison (%), T-test (mean) of difference.
Odds ratios (ORs) for associations between childhood public assistance and adult hypertension are summarized in Table 4 for men and Table 5 for women. No significant association was found between childhood public assistance and adult hypertension among men; unadjusted OR = 0.82, 95%CI: (0.39, 1.75) and covariate-adjusted (model 5) ORadj = 0.87, 95%CI: (0.37, 2.04). No significant association was found for men who grew up in families that neither received public assistance, nor experienced household deprivation, compared to men from socioeconomically deprived households that did not receive public assistance; unadjusted OR = 0.78, 95%CI: (0.45, 1.35) and fully-adjusted ORadj = 1.03, 95%CI: (0.54, 1.98).
Table 4.
Childhood Public Assistance and Adult Hypertension among African American Men: The Pitt County, North Carolina, Study
| Childhood Public Assistance |
Model 1a | Model 2b | Model 3c | Model 4d | Model 5e | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| OR | 95% CI | OR | 95% CI | OR | 95% CI | OR | 95% CI | OR | 95% CI | |||
| Public Assistance | 1.15 | 0.54, 2.48 | 1.01 | 0.43, 2.38 | 0.97 | 0.41, 2.29 | 0.98 | 0.41, 2.38 | 0.87 | 0.37, 2.04 | ||
| No Public Assistance/No Household Deprivation |
0.95 | 0.53, 1.70 | 0.96 | 0.52, 1.76 | 0.96 | 0.52, 1.77 | 0.93 | 0.49, 1.74 | 1.03 | 0.54, 1.98 | ||
| No Public Assistance/ Household Deprivation |
1.00 | Referent | 1.00 | Referent | 1.00 | Referent | 1.00 | Referent | 1.00 | Referent | ||
Abbreviations: OR = Odds Ratio, CI = Confidence Interval
Note: Data weighted for oversampling and non-response.
Age-Adjusted
Additionally adjusted for childhood socio-familial factors: household deprivation score, family structure, parental guardian stability, education and occupation
Additionally adjusted for socioeconomic factors in 1988: marital status, employment, education and occupation
Additionally adjusted for psychosocial factors in 1988: John Henryism, instrumental and emotional support and perceived stress
Additionally adjusted for lifestyle factors in 1988: body mass index, waist-hip ratio, smoking, strenuous exercise, and alcohol use
Table 5.
Childhood Public Assistance and Adult Hypertension among African American Women: The Pitt County, North Carolina, Study
| Childhood Public Assistance |
Model 1a | Model 2b | Model 3c | Model 4d | Model 5e | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| OR | 95% CI | OR | 95% CI | OR | 95% CI | OR | 95% CI | OR | 95% CI | |||
| Public Assistance | 0.35 | 0.16, 0.78 | 0.35 | 0.16, 0.78 | 0.36 | 0.16, 0.83 | 0.37 | 0.16, 0.84 | 0.35 | 0.14, 0.86 | ||
| No Public Assistance/No Household Deprivation |
0.83 | 0.52, 1.32 | 0.88 | 0.54, 1.42 | 0.90 | 0.55, 1.46 | 0.92 | 0.56, 1.51 | 0.86 | 0.52, 1.42 | ||
| No Public Assistance/ Household Deprivation |
1.00 | Referent | 1.00 | Referent | 1.00 | Referent | 1.00 | Referent | 1.00 | Referent | ||
Abbreviations: OR = Odds Ratio, CI = Confidence Interval
Note: Data weighted for oversampling and non-response.
Age-Adjusted
Additionally adjusted for childhood socio-familial factors: household deprivation score, family structure, parental guardian stability, education and occupation
Additionally adjusted for socioeconomic factors in 1988: marital status, employment, education and occupation
Additionally adjusted for psychosocial factors in 1988: John Henryism, instrumental and emotional support and perceived stress
Additionally adjusted for lifestyle factors in 1988: body mass index, waist-hip ratio, smoking, strenuous exercise, and alcohol use
A statistically significant inverse association was found between childhood public assistance and hypertension for African American women (Table 5). Women who resided from birth to 13 years of age in families that received public assistance had 69% decreased odds of hypertension compared to those from similarly disadvantaged families that did not receive public assistance; unadjusted OR = 0.31, 95% CI: (0.14, 0.68). The relationship between childhood public assistance and hypertension in 1988 persisted in all covariate-adjusted models, with the fully-adjusted model yielding a 65% decreased odds of hypertension for women, ORadj = 0.35, 95%CI: (0.14, 0.86). Women who grew up in families that neither received public assistance, nor experienced household deprivation, had 40% decreased odds of hypertension compared to women from socioeconomically deprived households who did not receive public assistance; unadjusted OR = 0.60, 95%CI: (0.39, 0.93). No significant association was detected after full covariate adjustment, ORadj = 0.86, 95% CI: (0.52, 1.42). In the appendix, we provide tables that include more detailed statistical models.
DISCUSSION
African American women who received needed public assistance in childhood had a 65% decreased odds of hypertension compared to those with equal need, but who did not receive public assistance. This association was independent of socio-familial factors in childhood and socioeconomic, demographic, lifestyle, and psychosocial characteristics in adulthood. We found no statistically significant association between public assistance, household deprivation and risk for hypertension among African American men before or after statistical adjustment. The observed sex difference, however, is consistent with previous research reporting the relationship between childhood socioeconomic position and adult blood pressure to hold only for women (19, 20).
That childhood public assistance was associated with lower odds for hypertension relative to women who reported no childhood material deprivation is an interesting finding. Since selection bias due to loss to follow-up has been demonstrated to be minimal among PCS women, our results suggest that the added socioeconomic benefit provided by public assistance may have negated some of the adverse impact of childhood household material deprivation on hypertension by early-middle adulthood. Studies designed to replicate this finding in other community-based samples of African Americans are needed. However, our findings of a potential long-term cardiovascular health benefit of childhood participation in a federal anti-poverty program are consistent with the documented positive association between childhood socioeconomic position and adult cardiovascular health in populations in other settings (21–23).
Hypertension has its origins in early childhood (22, 24), and socioeconomic resources are known to influence blood pressure and prehypertension in children as young as 5 and 6 years of age (25). The potentiation of the inverse association between age and blood pressure, by an impoverished childhood, suggests a latent effect where early-life disadvantage places individuals on an accelerated trajectory to develop hypertension in early/middle adulthood (23, 26). Public assistance programs that reduce the multiple socioeconomic disadvantages poor families face could help decelerate otherwise fast, age-related hypertension trajectories among children who grow up in these households.
Sex differences in the associations between childhood socioeconomic indicators and elevated blood pressure in adulthood remain unexplained by known risk factors. Childhood socioeconomic disadvantage has been shown to more strongly influence modifiable factors for elevated blood pressure (e.g., lower physical activity and physical fitness (27, 28), increased weight gain (29), and body mass (30)) among girls than boys. Although childhood body mass was not measured in this study, the higher BMI observed at midlife for PCS women, relative to men, probably has origins in early-life socioeconomic disadvantage (31). Therefore, sex differences in our reported association between childhood receipt of public assistance and adult hypertension could be due, at least in part, to sex differences in the relationship between early-life socioeconomic position and body mass, given that the latter is a major risk factor for elevated blood pressure (25). The absence of an observed association, for men, between childhood public assistance and later-life hypertension could also be due to selection bias, given that loss to follow-up was inversely associated with socioeconomic disadvantage; however, weighting our analyses for non-response should have mitigated the impact of this source of bias.
To our knowledge, this is the first study to document a long-term, sex-specific cardiovascular benefit of childhood receipt of public assistance among socioeconomically disadvantaged African American women. The fact that our study findings are based on an observational study, rather than a randomized trial, precludes unqualified statements about causal relationships. However; our findings are consistent with a growing body of evidence attesting to substantive and long-term health benefits of social policies that mitigate the well-known adverse effects of early childhood material and social deprivation (32).
Our retrospective assessment of childhood receipt of public assistance did not collect information on the types of public assistance (e.g., food stamps, Medicaid, public housing, etc.) study participants received in childhood; hence, we are unable to identify whether some types of public assistance were more impactful than others. This is a potentially important question for future research.
Study limitations include potential recall bias due to our retrospective study design. However, in addition to our use of the Event History Calendar (14, 15) to aid recall accuracy, we believe that weighting the estimates to account for differential non-response, along with the low probability of systematic misclassification on the exposure by hypertensive status, minimizes the contribution of recall bias on our results. Moreover, validation studies have found decades-long recollection of discrete and objective early-life socioeconomic experiences to be reasonably accurate and reliable (33–35).
Receipt of public assistance during childhood may be correlated with, or alternatively, be a proxy for other early-life determinants of adult hypertension that we did not adjust for in our analyses, i.e., childhood neighborhood deprivation (37), childhood maltreatment (38, 39), infant birth weight, breastfeeding duration, childhood BMI, and environmental tobacco smoke exposure (25), among others, including number of children in the household. It is also the case that statistical adjustment for some or all of these unmeasured mediators could lead to over-adjustment bias (40). Finally, misclassification of covariates due to our reliance on somewhat crude lifestyle measures (e.g., physical activity and alcohol status) could have biased our results, but we believe that such bias is likely to be random in nature.
Our research findings suggest that investments in social policies that reduce early-life socioeconomic disadvantage provide a long-term cardiovascular benefit for African Americans, especially women, who grow up in disadvantaged households. Furthermore, these social investments in public assistance programs for families in need could also reduce the nation’s social and economic burden from hypertension, which by recent estimates, costs the nation approximately $46 billion per year (8).
Acknowledgments
This study was supported by the National Institute on Minority Health and Health Disparities (NIMHD/NIH) grant K22MD006133 to Dr. Barrington. The Pitt County Study was supported by the National Heart, Lung, and Blood Institute (NHLBI/NIH) grants R01HL033211 and R01HL065645 to Dr. James.
Appendix
Appendix Table 1.
Regression Coefficients for Public Assistance and Adult Hypertension for Men
| Crude | Model 1 | Model 2 | Model 3 | Model 4 | Model 5 | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| N = 405 | N = 405 | N = 403 | N = 401 | N = 401 | N = 401 | |||||||
| Beta | Se | Beta | Se | Beta | Se | Beta | Se | Beta | Se | Beta | Se | |
| Intercept | −0.68 | 0.16 | −3.26 | 0.78 | −2.79 | 0.88 | −2.31 | 1.07 | −2.12 | 1.10 | −4.88 | 1.55 |
| Public Assistance (PA) | −0.19 | 0.38 | 0.14 | 0.39 | 0.01 | 0.43 | −0.04 | 0.44 | −0.02 | 0.45 | −0.14 | 0.43 |
| No PA/No Deprivation | −0.25 | 0.28 | −0.06 | 0.30 | −0.04 | 0.31 | −0.04 | 0.31 | −0.08 | 0.32 | 0.03 | 0.33 |
| AGE | 0.07 | 0.02 | 0.07 | 0.02 | 0.06 | 0.02 | 0.06 | 0.02 | 0.05 | 0.02 | ||
| Intact Family | −0.37 | 0.39 | −0.40 | 0.39 | −0.45 | 0.40 | −0.57 | 0.40 | ||||
| Two Adult Household | −0.15 | 0.33 | −0.19 | 0.33 | −0.17 | 0.33 | −0.05 | 0.34 | ||||
| Parental Ed - Missing | 0.19 | 0.31 | 0.14 | 0.32 | 0.16 | 0.33 | 0.33 | 0.34 | ||||
| Parental Ed - >HS | 0.47 | 0.35 | 0.39 | 0.34 | 0.45 | 0.35 | 0.56 | 0.36 | ||||
| Parental Occ - Skilled | −0.79 | 0.42 | −0.72 | 0.43 | −0.78 | 0.44 | −0.80 | 0.44 | ||||
| Parental Occ - Farmer | −0.04 | 0.32 | −0.04 | 0.33 | −0.11 | 0.34 | −0.03 | 0.36 | ||||
| Married | 0.12 | 0.27 | 0.07 | 0.27 | −0.04 | 0.30 | ||||||
| Unemployed | 0.30 | 0.52 | 0.39 | 0.54 | 0.59 | 0.52 | ||||||
| Education HS | 0.04 | 0.32 | −0.01 | 0.32 | −0.03 | 0.34 | ||||||
| Education > HS | −0.30 | 0.38 | −0.28 | 0.38 | −0.20 | 0.38 | ||||||
| Occ - Operatives | −0.37 | 0.29 | −0.38 | 0.29 | −0.27 | 0.30 | ||||||
| Occ - Craft/White Collar | −0.47 | 0.59 | −0.41 | 0.59 | −0.30 | 0.65 | ||||||
| High John Henryism | −0.13 | 0.27 | −0.14 | 0.28 | ||||||||
| Instrumental Support | 0.29 | 0.32 | 0.14 | 0.35 | ||||||||
| Emotional Support | 0.24 | 0.31 | 0.38 | 0.33 | ||||||||
| High Stress | −0.18 | 0.28 | −0.10 | 0.30 | ||||||||
| BMI | 0.10 | 0.04 | ||||||||||
| WHR_Q2 | −0.27 | 0.42 | ||||||||||
| WHR_Q3 | 0.89 | 0.38 | ||||||||||
| WHR_Q4 | 0.22 | 0.43 | ||||||||||
| Smoker | 0.05 | 0.30 | ||||||||||
| Physical Activity | −0.03 | 0.28 | ||||||||||
| Drinker | 0.01 | 0.30 | ||||||||||
Note: Data weighted for oversampling and non-response.
Appendix Table 2.
Regression Coefficients for Public Assistance and Adult Hypertension for Women
| Crude Model |
Model 1 | Model 2 | Model 3 | Model 4 | Model 5 | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| N = 737 | N = 737 | N = 737 | N = 737 | N = 725 | N = 725 | |||||||
| Beta | Se | Beta | Se | Beta | Se | Beta | Se | Beta | Se | Beta | Se | |
| Intercept | −0.89 | 0.12 | −4.14 | 0.59 | −4.01 | 0.63 | −4.34 | 0.73 | −4.60 | 0.79 | −8.44 | 1.08 |
| Public Assistance (PA) | −1.19 | 0.41 | −1.04 | 0.41 | −1.05 | 0.41 | −1.02 | 0.42 | −0.99 | 0.42 | −1.05 | 0.46 |
| No PA/No Deprivation | −0.51 | 0.22 | −0.19 | 0.24 | −0.13 | 0.25 | −0.11 | 0.25 | −0.09 | 0.25 | −0.16 | 0.26 |
| AGE | 0.09 | 0.01 | 0.08 | 0.02 | 0.09 | 0.02 | 0.09 | 0.02 | 0.09 | 0.02 | ||
| Intact Family | 0.20 | 0.29 | 0.15 | 0.29 | 0.17 | 0.30 | 0.17 | 0.31 | ||||
| Two Adult Household | −0.08 | 0.25 | −0.02 | 0.25 | −0.05 | 0.25 | 0.06 | 0.27 | ||||
| Parental Ed - Missing | −0.12 | 0.25 | −0.11 | 0.27 | −0.17 | 0.27 | −0.13 | 0.27 | ||||
| Parental Ed - >HS | −0.59 | 0.31 | −0.56 | 0.32 | −0.59 | 0.32 | −0.56 | 0.35 | ||||
| Parental Occ - Skilled | −0.19 | 0.33 | −0.18 | 0.33 | −0.12 | 0.34 | −0.24 | 0.37 | ||||
| Parental Occ - Farmer | −0.01 | 0.27 | −0.03 | 0.27 | −0.04 | 0.28 | −0.11 | 0.29 | ||||
| Married | 0.06 | 0.23 | 0.21 | 0.24 | 0.33 | 0.25 | ||||||
| Unemployed | 0.47 | 0.27 | 0.40 | 0.28 | 0.16 | 0.30 | ||||||
| Education HS | 0.12 | 0.26 | 0.14 | 0.27 | 0.32 | 0.28 | ||||||
| Education > HS | −0.05 | 0.31 | 0.00 | 0.31 | 0.12 | 0.33 | ||||||
| Occ - Operatives | 0.05 | 0.30 | 0.07 | 0.31 | 0.18 | 0.32 | ||||||
| Occ - Craft/White Collar | 0.20 | 0.25 | 0.21 | 0.26 | 0.14 | 0.27 | ||||||
| High John Henryism | 0.03 | 0.22 | 0.15 | 0.23 | ||||||||
| Instrumental Support | −0.82 | 0.29 | −0.74 | 0.32 | ||||||||
| Emotional Support | 0.21 | 0.25 | 0.16 | 0.29 | ||||||||
| High Stress | 0.30 | 0.23 | 0.28 | 0.24 | ||||||||
| BMI | 0.10 | 0.02 | ||||||||||
| WHR_Q2 | 0.37 | 0.36 | ||||||||||
| WHR_Q3 | 0.72 | 0.37 | ||||||||||
| WHR_Q4 | 1.14 | 0.36 | ||||||||||
| Smoker | 0.51 | 0.27 | ||||||||||
| Physical Activity | −0.42 | 0.25 | ||||||||||
| Drinker | 0.43 | 0.25 | ||||||||||
Note: Data weighted for oversampling and non-response.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
REFERENCES
- 1.DeNavas-Walt C, Proctor BD. Income and Poverty in the United States: 2014. Washington, DC: U.S. Census Bureau; 2015. [Google Scholar]
- 2.Irving SK, Loveless T. Dynamics of Economic Well-Being: Participation in Government Programs, 2009–2012: Who Gets Assistance. Washington, DC: U.S. Census Bureau; 2015. [Google Scholar]
- 3.Almond D, Hoynes HW, Schanzenbach DW. Inside The War On Poverty: The Impact Of Food Stamps On Birth Outcomes. Review of Economics and Statistics. 2011;93(2):387–403. [Google Scholar]
- 4.Lee BJ, Mackey-Bilaver L. Effects of WIC and Food Stamp Program participation on child outcomes. Children and Youth Services Review. 2007;29(4):501–517. [Google Scholar]
- 5.Jackson MI. Early childhood WIC participation, cognitive development and academic achievement. Social science & medicine (1982) 2015;126:145–153. doi: 10.1016/j.socscimed.2014.12.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Hoynes HW. Long run impacts of childhood access to the safety net. Cambridge, MA: National Bureau of Economic Research; 2012. [Google Scholar]
- 7.U.S. Bureau of Labor Statistics. Labor Force Characteristics by Race and Ethnicity, 2013. Washington, DC: BLS Reports; 2014. [Google Scholar]
- 8.Mozaffarian D, Benjamin EJ, Go AS, Arnett DK, Blaha MJ, Cushman M, et al. Heart disease and stroke statistics--2015 update: a report from the American Heart Association. Circulation. 2015;131(4):e29–e322. doi: 10.1161/CIR.0000000000000152. [DOI] [PubMed] [Google Scholar]
- 9.Roberts CB, Vines AI, Kaufman JS, James SA. Cross-sectional association between perceived discrimination and hypertension in African-American men and women: the Pitt County Study. American journal of epidemiology. 2008;167(5):624–632. doi: 10.1093/aje/kwm334. [DOI] [PubMed] [Google Scholar]
- 10.Curtis AB, Strogatz DS, James SA, Raghunathan TE. The contribution of baseline weight and weight gain to blood pressure change in African Americans: the Pitt County Study. Annals of epidemiology. 1998;8(8):497–503. doi: 10.1016/s1047-2797(98)00024-6. [DOI] [PubMed] [Google Scholar]
- 11.James SA, Jamjoum L, Raghunathan TE, Strogatz DS, Furth ED, Khazanie PG. Physical activity and NIDDM in African-Americans. The Pitt County Study. Diabetes care. 1998;21(4):555–562. doi: 10.2337/diacare.21.4.555. [DOI] [PubMed] [Google Scholar]
- 12.James SA, Van Hoewyk J, Belli RF, Strogatz DS, Williams DR, Raghunathan TE. Life-course socioeconomic position and hypertension in African American men: the Pitt County Study. American journal of public health. 2006;96(5):812–817. doi: 10.2105/AJPH.2005.076158. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Strogatz DS, James SA, Haines PS, Elmer PJ, Gerber AM, Browning SR, et al. Alcohol consumption and blood pressure in black adults: the Pitt County Study. American journal of epidemiology. 1991;133(5):442–450. doi: 10.1093/oxfordjournals.aje.a115911. [DOI] [PubMed] [Google Scholar]
- 14.Belli RF. The structure of autobiographical memory and the event history calendar: potential improvements in the quality of retrospective reports in surveys. Memory. 1998;6(4):383–406. doi: 10.1080/741942610. [DOI] [PubMed] [Google Scholar]
- 15.Belli RF, Shay WL, Stafford FP. Event history calendars and question list surveys: a direct comparison of interviewing methods. Public Opin Q. 2001;65(1):45–74. doi: 10.1086/320037. [DOI] [PubMed] [Google Scholar]
- 16.James SA, Keenan NL, Strogatz DS, Browning SR, Garrett JM. Socioeconomic status John Henryism blood pressure in black adults The Pitt County Study. American journal of epidemiology. 1992;135(1):59–67. doi: 10.1093/oxfordjournals.aje.a116202. [DOI] [PubMed] [Google Scholar]
- 17.Strogatz DS, Croft JB, James SA, Keenan NL, Browning SR, Garrett JM, et al. Social support, stress, and blood pressure in black adults. Epidemiology. 1997;8(5):482–487. doi: 10.1097/00001648-199709000-00002. [DOI] [PubMed] [Google Scholar]
- 18.Bergland PA, editor. Analysis of complex sample survey data using the SURVEYMEANS and SURVEYREG procedures and macro coding. Proceedings of the Twenty-Seventh Annual SAS Users Group International Conference. Cary, NC: SAS Institute Inc.; 2002. [Google Scholar]
- 19.Janicki-Deverts D, Cohen S, Matthews KA, Jacobs DR., Jr Sex differences in the association of childhood socioeconomic status with adult blood pressure change: the CARDIA study. Psychosomatic medicine. 2012;74(7):728–735. doi: 10.1097/PSY.0b013e31825e32e8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Schumann B, Kluttig A, Tiller D, Werdan K, Haerting J, Greiser KH. Association of childhood and adult socioeconomic indicators with cardiovascular risk factors and its modification by age: the CARLA Study 2002–2006. BMC public health. 2011;11:289. doi: 10.1186/1471-2458-11-289. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Power C, Atherton K, Strachan DP, Shepherd P, Fuller E, Davis A, et al. Life-course influences on health in British adults: Effects of socio-economic position in childhood and adulthood. International journal of epidemiology. 2007;36(3):532–539. doi: 10.1093/ije/dyl310. [DOI] [PubMed] [Google Scholar]
- 22.Kivimaki M, Smith GD, Elovainio M, Pulkki L, Keltikangas-Jarvinen L, Talttonen L, et al. Socioeconomic circumstances in childhood and blood pressure in adulthood: the cardiovascular risk in young Finns study. Annals of epidemiology. 2006;16(10):737–742. doi: 10.1016/j.annepidem.2006.01.004. [DOI] [PubMed] [Google Scholar]
- 23.Hardy R, Kuh D, Langenberg C, Wadsworth ME. Birthweight, childhood social class, and change in adult blood pressure in the 1946 British birth cohort. Lancet. 2003;362(9391):1178–1183. doi: 10.1016/S0140-6736(03)14539-4. [DOI] [PubMed] [Google Scholar]
- 24.Chen X, Wang Y. Tracking of blood pressure from childhood to adulthood: a systematic review and meta-regression analysis. Circulation. 2008;117(25):3171–3180. doi: 10.1161/CIRCULATIONAHA.107.730366. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.van den Berg G, van Eijsden M, Galindo-Garre F, Vrijkotte TG, Gemke RJ. Explaining socioeconomic inequalities in childhood blood pressure and prehypertension: the ABCD study. Hypertension. 2013;61(1):35–41. doi: 10.1161/HYPERTENSIONAHA.111.00106. [DOI] [PubMed] [Google Scholar]
- 26.Lawlor DA, Smith GD. Early life determinants of adult blood pressure. Curr Opin Nephrol Hypertens. 2005;14(3):259–264. doi: 10.1097/01.mnh.0000165893.13620.2b. [DOI] [PubMed] [Google Scholar]
- 27.Bohr AD, Brown DD, Laurson KR, Smith PJ, Bass RW. Relationship between socioeconomic status and physical fitness in junior high school students. The Journal of school health. 2013;83(8):542–547. doi: 10.1111/josh.12063. [DOI] [PubMed] [Google Scholar]
- 28.Fairclough SJ, Boddy LM, Hackett AF, Stratton G. Associations between children’s socioeconomic status, weight status, and sex, with screen-based sedentary behaviours and sport participation. International journal of pediatric obesity : IJPO : an official journal of the International Association for the Study of Obesity. 2009;4(4):299–305. doi: 10.3109/17477160902811215. [DOI] [PubMed] [Google Scholar]
- 29.Balistreri KS, Van Hook J. Trajectories of overweight among US school children: a focus on social and economic characteristics. Maternal and child health journal. 2011;15(5):610–619. doi: 10.1007/s10995-010-0622-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Pudrovska T, Reither EN, Logan ES, Sherman-Wilkins KJ. Gender and reinforcing associations between socioeconomic disadvantage and body mass over the life course. Journal of health and social behavior. 2014;55(3):283–301. doi: 10.1177/0022146514544525. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.James SA, Fowler-Brown A, Raghunathan TE, Van Hoewyk J. Life-course socioeconomic position and obesity in African American Women: the Pitt County Study. American journal of public health. 2006;96(3):554–560. doi: 10.2105/AJPH.2004.053447. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Schoeni RF, House JS, Kaplan GA, Pollack H, editors. Making Americans Healthier: Social and Economic Policy as Health Policy. New York: Russell Sage Foundation; 2008. [Google Scholar]
- 33.Yancura LA, Aldwin CM. Stability and change in retrospective reports of childhood experiences over a 5-year period: findings from the Davis Longitudinal Study. Psychology and aging. 2009;24(3):715–721. doi: 10.1037/a0016203. [DOI] [PubMed] [Google Scholar]
- 34.Krieger N, Okamoto A, Selby JV. Adult female twins’ recall of childhood social class and father’s education: a validation study for public health research. American journal of epidemiology. 1998;147(7):704–708. doi: 10.1093/oxfordjournals.aje.a009512. [DOI] [PubMed] [Google Scholar]
- 35.Berney LR, Blane DB. Collecting retrospective data: accuracy of recall after 50 years judged against historical records. Social science & medicine (1982) 1997;45(10):1519–1525. doi: 10.1016/s0277-9536(97)00088-9. [DOI] [PubMed] [Google Scholar]
- 36.Batty GD, Lawlor DA, Macintyre S, Clark H, Leon DA. Accuracy of adults’ recall of childhood social class: findings from the Aberdeen children of the 1950s study. Journal of epidemiology and community health. 2005;59(10):898–903. doi: 10.1136/jech.2004.030932. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Johnson RC, Schoeni RF. Early-life origins of adult disease: national longitudinal population-based study of the United States. American journal of public health. 2011;101(12):2317–2324. doi: 10.2105/AJPH.2011.300252. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Riley EH, Wright RJ, Jun HJ, Hibert EN, Rich-Edwards JW. Hypertension in adult survivors of child abuse: observations from the Nurses’ Health Study II. Journal of epidemiology and community health. 2010;64(5):413–418. doi: 10.1136/jech.2009.095109. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Springer KW, Sheridan J, Kuo D, Carnes M. Long-term physical and mental health consequences of childhood physical abuse: results from a large population-based sample of men and women. Child Abuse Negl. 2007;31(5):517–530. doi: 10.1016/j.chiabu.2007.01.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Schisterman EF, Cole SR, Platt RW. Overadjustment bias and unnecessary adjustment in epidemiologic studies. Epidemiology. 2009;20(4):488–495. doi: 10.1097/EDE.0b013e3181a819a1. [DOI] [PMC free article] [PubMed] [Google Scholar]