Seven-Year Incidence of Hypertension in a Cohort of Middle-Aged African Americans and Whites
Abstract—Many studies have suggested that African Americans have a higher prevalence of hypertension than whites. The authors conducted a prospective study of hypertension incidence from 1987-1988 to 1994-1995 in 140 African American and 237 white adults aged 30 to 54 years at baseline. The study participants were screened for participation in the Trials of Hypertension Prevention, phase 1, an 18-month lifestyle modification intervention trial aimed at lowering blood pressure, at the Baltimore Clinical Center. Baseline age, blood pressure, body mass index, and heart rate were similar in the two groups. Compared with whites, however, African Americans had a lower percentage of men, college graduates, and households with an income ≥$40 000 per year. African Americans also had lower mean urinary sodium to creatinine ratio and potassium to creatinine ratio, but a similar sodium to potassium ratio. The incidence of hypertension (blood pressure ≥160/95 mm Hg and/or taking antihypertensive medication) over 7 years of follow-up was nearly identical: 25.7% in African Americans and 25.3% in whites. Baseline age, gender, blood pressure, and heart rate were all associated with the incidence of hypertension. Even after adjustment for these covariables, the risk of hypertension was not higher in African Americans compared with whites. These results indicate that middle-aged African Americans and whites have a similar risk of developing hypertension given the same age, initial blood pressure, and body mass index at baseline.
Hypertension is an important public health challenge in the United States, especially among African Americans. In general, African Americans are at higher risk of hypertension and its complications than are other ethnic groups.1 2 3 4 5 6 7
One of the first studies to compare BP between African Americans and whites in a community sample was conducted in Muscogee County, Georgia, by Comstock.8 In his study, Comstock found that African Americans had higher BP levels than whites at every age studied. Many subsequent investigations, including several national surveys of US residents conducted by the National Center for Health Statistics, confirmed that African Americans have higher BP levels as well as a higher prevalence of hypertension than whites.9 10 11 12 In most studies, the prevalence of hypertension was about 50% higher in African Americans than in whites.
Several longitudinal studies have also shown that African Americans have a higher incidence of hypertension than whites.1 2 3 4 5 However, this racial difference in the risk of hypertension may be modified by age. In the NHANES I Epidemiologic Follow-up Study, the incidence of hypertension in African Americans was over two times higher than in whites among those aged 25 to 34 years. In contrast, the incidence of hypertension did not differ between African Americans and whites who were 55 years of age or older.1 In other studies conducted in young adults, the incidence of hypertension in African Americans was an average of two times higher than in whites.3 4 5
The purpose of the present study was to compare the incidence of hypertension between African Americans and whites who had similar BPs at baseline and to identify risk factors for developing hypertension.
The study population consisted of 463 screenees (171 African Americans and 292 whites) from the Baltimore clinic center of the TOHP-1. The TOHP-1 was a national, multicenter, randomized, controlled trial designed to test the short-term feasibility and efficacy of three lifestyle (weight loss, sodium restriction, and stress management) and four nutritional supplement (calcium, magnesium, fish oil, and potassium) interventions aimed at lowering diastolic pressure in those whose BP was initially in the high normal range (80 to 89 mm Hg).13 14 Most of the study participants were identified through a work-site screening program conducted from September 1987 through October 1988. The participants primarily came from the Social Security Administration and the Health Care Financing Administration headquarters located in western Baltimore. Of the 463 screenees, 286 were randomly allocated to the trial. The major reasons for not randomizing those who were screened were presence of a mean diastolic BP ≤80 mm Hg or a decision by the participant. The principal eligibility criteria for study participation in TOHP-1 included age between 30 and 54 years and no evidence of medically diagnosed hypertension or other cardiovascular disease, extreme obesity (BMI ≥36.1 kg/m2), or heavy alcohol use (≥21 drinks per week).13 14
Baseline Data Collection
Baseline observations were obtained at three screening visits.13 All data were carefully collected by specially trained, experienced observers using standardized methods and a stringent level of quality control. All data collectors passed an initial series of certification examinations as well as periodic recertification evaluations. Demographic information, medical history, and health habit data were collected at screening visit 1. Physical activity was assessed with the Harvard Alumni Physical Activity Questionnaire at visit 2, which ascertains both work and leisure physical activity and the frequency of vigorous exercise resulting in perspiration each week.15 One 24-hour urine specimen was collected at screening visit 3 for estimation of sodium, potassium, and creatinine content.
Nine BP measurements, three at each of the three screening visits, were obtained at baseline. BP measurements were obtained using the Hawksley random-zero modification of the standard mercury sphygmomanometer.16 BP was measured after the individual had been seated quietly for 5 minutes. The subject was instructed not to eat or smoke for at least 30 minutes before their BP measurements. The first and fifth Korotkoff sounds were recorded as systolic and diastolic pressure, respectively. Body weight and height were measured at screening visit 1, and BMI (kg/m2) was calculated as an index of obesity.17
Follow-up Data Collection
The follow-up study was conducted between November 1994 and September 1995. Medical history, especially personal history of hypertension and use of antihypertensive medication, was evaluated by questionnaire. Three BP measurements were obtained using the same method as for the baseline visits. Body weight, height, and waist and hip girths were also measured at the follow-up visit.
Hypertension at follow-up was defined as diagnosis of hypertension by a physician during the follow-up period and initiation of antihypertensive therapy as assessed by questionnaire and/or2 systolic BP ≥160 mm Hg and/or3 diastolic BP ≥95 mm Hg at the follow-up examination. A mean BP ≥160/95 mm Hg was used as the cut point for definition of hypertension in the present study because the follow-up BP measurements were obtained at only one visit. Furthermore, the presence of a systolic BP <160 mm Hg was used as an eligibility criterion for inclusion in TOHP-1.
Incidence of hypertension was expressed as the proportion of hypertensives in all study participants seen at follow-up. The racial differences in the incidence of hypertension or its risk factors were examined using χ2 or Student’s t tests. Logistic regression analysis was used to compare the risk of hypertension between African Americans and whites after adjustment for other risk factors. Univariate and multivariate logistic regression analyses also were used to explore the relationship between baseline characteristics and risk of hypertension. Linear regression analysis was used to examine the relationship between these baseline characteristics and changes in BP level over follow-up. Study participants who were taking antihypertensive medications at the follow-up visit were excluded from the linear regression analysis because their BP level before initiation of antihypertensive medication was not available. OR or regression coefficients associated with a difference of 1 SD in continuous independent variables were reported. Assignment to sodium reduction or weight loss interventions in TOHP-1 was associated with the risk of hypertension during follow-up. Therefore, intervention assignment in TOHP-1 was included as a covariable in all multivariate analyses. To avoid possible bias due to regression toward the mean, the average of BP measured at baseline and at follow-up was used to predict change in BP.18 All analyses were performed using SAS software.19
Of the 463 African American and white TOHP-1 screenees, 377 (81.4%) participated in the follow-up study. Among the 86 nonrespondents, 1 was deceased, 3 were seriously ill, and the remaining 82 refused to participate. Nonrespondents and respondents had similar baseline characteristics including race, gender, socioeconomic status (education and employment), health habits (cigarette smoking, alcohol consumption, and physical activity), body weight, heart rate, and BP. However, nonrespondents were an average of 1.5 years younger than the respondents (P=.02).
Baseline Characteristics of African American and White Participants
Table 1⇓ shows baseline characteristics of the study participants by race. Because the proportion of male participants was significantly lower in African Americans than in whites, all analyses were adjusted for sex. African Americans were less educated and had lower household incomes than whites, although all study participants had a relatively high socioeconomic status. Compared with the whites, more African Americans were employed full-time. The proportion of study participants who were married, were current cigarette smokers, drank alcohol, or were allocated to sodium or weight reduction interventions in TOHP-1 was not statistically significantly different between African Americans and whites.
Average age, BP, and BMI were similar in the African Americans and whites because of the study’s eligibility criteria. African Americans had slightly lower physical activity measured by the number of times per week a person exercised enough to perspire, albeit this difference was not statistically significant. Ratios of urinary sodium to creatinine and potassium to creatinine were significantly lower in the African Americans compared with the whites, while the corresponding ratio of urinary sodium to potassium was not significantly different. The racial differences in baseline characteristics were similar in men and women.
Incidence of Hypertension in African American and White Participants
The average period of follow-up was 7.1 years (ranging from 6.3 to 7.9 years) and was similar in the African American and white participants. At the follow-up visit, 36 African Americans and 60 whites reported a history of hypertension since their screening examination in 1987-1988 and were taking antihypertensive medication or met the BP criteria for diagnosis of hypertension. The overall incidence of hypertension was 25.7% for African Americans and 25.3% for whites.
African Americans and whites had similar crude odds of developing hypertension (Table 2⇓). After adjustment for gender, the OR was reduced from 1.02 to 0.62, but this was still not significantly different from 1. After further adjustment for age, income, alcohol consumption, intervention assignment in TOHP-1, BMI, heart rate, and BP at baseline, the OR changed very little. The ORs adjusted for urinary sodium and potassium are not listed in Table 2⇓ because they were calculated using a smaller sample size. The OR adjusted for gender and ratio of urinary sodium to creatinine was 0.68 (95% CI, 0.37 to 1.25), and the OR adjusted for gender and ratio of urinary potassium to creatinine was 0.68 (95% CI, 0.37 to 1.27), respectively.
Likewise, changes in BP over the period of follow-up were not statistically significantly different between African Americans and whites (gender-adjusted systolic: 1.6 mm Hg in African Americans versus 1.9 mm Hg in whites, P=.8; gender-adjusted diastolic: −5.6 mm Hg in African Americans and −5.8 mm Hg in whites, P=.9). After further adjustment for other risk factors, these differences remained not significant.
Relationship of Baseline Characteristics to Risk of Hypertension
Univariate ORs of hypertension associated with baseline characteristics, overall and by race, are presented in Table 3⇓. Age was associated with an increased odds of developing hypertension in the total sample as well as in whites. Age was not significantly associated with the odds of hypertension in African Americans, and a test for interaction between age and race in the odds of hypertension was not significant (P=.11). Female gender was a significant predictor of hypertension in both races.
Baseline BP was strongly related to the odds of hypertension in all groups. For example, a 9.3 mm Hg higher systolic pressure at baseline was associated with a 2.98-fold greater odds of developing hypertension in African Americans and a 2.60-fold greater risk in whites. Likewise, a 4.2 mm Hg higher diastolic pressure was associated with 2.51 times higher odds of developing hypertension in African Americans and 2.96 times higher odds in whites. The magnitude of the elevated odds for hypertension related to baseline BP was similar for African Americans and whites (P values for test of interaction: 0.65 for systolic and 0.57 for diastolic pressure). After adjustment for gender, ORs of hypertension associated with baseline BP remained similar in African Americans compared with whites. Baseline heart rate was positively related to the odds of developing hypertension in whites and the total sample, while baseline BMI was related to the odds of developing hypertension in African Americans only (Table 3⇑). However, racial differences in the relationship of heart rate and BMI to the odds of developing hypertension did not achieve statistical significance.
The results from multivariate logistic regression models are presented in Table 4⇓. Analyses were conducted in the total sample because no significant interaction was detected between the effect of race and baseline characteristics on the odds of hypertension. In model 1, where baseline BP was not included, age and baseline heart rate were related to an increased odds of hypertension. Men had lower odds of developing hypertension than women. Baseline BP was introduced to model 2. In this model, age and heart rate were no longer significantly associated with odds of hypertension, but baseline systolic and diastolic pressures independently predicted odds of hypertension. Systolic pressure higher by 1 SD (9.3 mm Hg) was related to a 79% increase in the odds of hypertension, whereas diastolic pressure higher by 1 SD (4.2 mm Hg) was associated with a 2.19-fold increase in the odds of hypertension.
Relationship of Baseline Characteristics to Changes in BP
Over the 7 years of follow-up, systolic pressure increased by an average of 2.6 mm Hg, but diastolic pressure decreased by 5.1 mm Hg in the 110 African American and 183 white participants who were not taking antihypertensive medications. A 5.8-year older age was significantly associated with a greater increase in systolic pressure during follow-up in whites (2.7 mm Hg, P<.01) and in the combined group (1.7 mm Hg, P<.05). Women had a significantly greater increase in systolic pressure than men during follow-up in African Americans (5.1 mm Hg, P<.05), whites (6.4 mm Hg, P<.01), and the combined group (5.7 mm Hg, P<.001). Graduation from a 4-year college was inversely associated with an increase in systolic pressure in the African Americans (−5.4 mm Hg, P<.05) but not in the whites (0.27 mm Hg, P=.9). A test for interaction between race and education on systolic pressure was of borderline statistical significance (P=.06). Higher household income was inversely associated with an increase in systolic pressure over time in all groups, but this was significant only in the combined group (−3.5 mm Hg, P<.05). However, neither education nor income was significantly related to change in systolic pressure over time after adjustment for gender (data not shown).
A BMI higher by 1 SD (3.7 kg/m2) at baseline was associated with a greater increase in systolic pressure during follow-up in the African Americans (3.0 mm Hg, P<.01) and the combined sample (1.3 mm Hg, P<.05). An increase of 1 SD in BMI during follow-up (1.98 kg/m2) was positively related to change in both systolic and diastolic pressures among the African Americans (2.8 and 1.6 mm Hg, both P<.05), whites (2.7 and 2.1 mm Hg, both P<.001), and the combined sample (2.9 and 1.9 mm Hg, both P<.001). In regression models including both BMI at baseline and its change during follow-up, only change during follow-up was significantly associated with change in BP. Average BPs measured at baseline and at the follow-up visit were significantly related to change in BP during follow-up among all groups. For example, average systolic pressure higher by 1 SD (9.3 mm Hg) was associated with an increase in systolic pressure during follow-up of 5.4, 5.9, and 5.7 mm Hg for African Americans, whites, and the combined sample, respectively (all P<.001). Average diastolic pressure higher by 1 SD (5.0 mm Hg) was associated with an increase in diastolic pressure during follow-up of 4.3, 5.0, and 4.7 mm Hg for African Americans, whites, and the combined sample, respectively (all P<.001). The magnitude of this relationship did not differ significantly between African Americans and whites (P=.67 for systolic and P=.30 for diastolic).
In multivariate linear regression analysis, age, gender, ethnicity, intervention assignment in TOHP-1, baseline alcohol consumption, change in BMI during follow-up, and average BP measured at baseline and at the follow-up visit explained about 31% of the variation in systolic pressure change and 50% of the variation in diastolic pressure change during follow-up. In that model, on average, women had a 3.9 mm Hg greater increase in systolic pressure than men (P<.01) during follow-up. Changes in BMI over time were positively related to corresponding changes in systolic and diastolic pressures. For example, a 1.98 kg/m2 increase in BMI was related to a 2.3 mm Hg increase in systolic pressure (P<.001) and a 1.7 mm Hg increase in diastolic pressure (P<.001) during follow-up. A difference of 9.3 mm Hg in the average systolic pressure measured at baseline and at follow-up visit was related to an increase of 5.2 mm Hg in systolic pressure (P<.001) during the 7 years of follow-up, while a 5.0 mm Hg difference in average diastolic pressure was related to an increase of 4.7 mm Hg in diastolic pressure (P<.001).
In the present study, middle-aged African Americans did not have a higher incidence of hypertension over a 7-year follow-up period compared with whites with a similar age, initial BP, and BMI. Mean change in BP over time was also not significantly different in the two ethnic groups. Furthermore, the magnitude of increases in the odds of hypertension associated with baseline BP level was similar in African Americans compared with whites.
These findings are different from those of several previous studies that demonstrated an elevated risk of hypertension in African Americans compared with whites.8 9 10 11 12 20 21 22 23 However, in the NHANES I Epidemiologic Follow-up Study, the only longitudinal study of the age-specific incidence of hypertension by race, African Americans and whites had a similar risk of hypertension after age 55.1 In the present investigation, the study participants were middle-aged and had a relatively high normal diastolic pressure when they entered the study. The distributions of age, BMI, heart rate, and BP at baseline were almost identical between African American and white participants.
Some studies have suggested that African Americans develop hypertension at an earlier age than whites.24 25 26 Manatunga and colleagues followed 164 African American and 345 white children for 2 to 5.5 years.26 They found that the African American children had a significantly higher mean BP and rate of increase in BP over time. Earlier puberty and weight gain in African American children may partly explain the difference in BP.26 Given the higher overall risk of hypertension in African Americans compared with whites, the results from our study are consistent with other studies1 3 25 26 suggesting that the excess burden of hypertension in African Americans may be due primarily to a higher BP level and rise in BP with age in adolescence and young adulthood. Efforts to prevent high BP and its sequelae in African Americans need to target groups before they reach middle age.
In the present study, most of the participants were employed and had a relatively high socioeconomic status compared with general population. Lower socioeconomic status and a lower standard of health care have been proposed as possible explanations for the excess risk of hypertension in African Americans. The high socioeconomic status of the present study population may have also been partially responsible for the similar incidence of hypertension in the African American and white participants.
In the present study, women had a higher incidence of hypertension and greater increase in BP over time than men. These results may reflect the distribution of age and initial BP in this cohort. Several studies have demonstrated that men have a higher risk of hypertension in early adulthood. In contrast, women have a higher risk of hypertension in later life.1 27
In contrast to many previous studies,28 29 30 baseline BMI was not a strong predictor of the incidence of hypertension among whites in this study. However, longitudinal change in BMI was positively associated with an increase in BP over follow-up, independent of other important risk factors for hypertension. This finding suggests that gain in body weight over time may be a more important risk factor for high BP than a participant’s initial body weight at baseline.
Despite the narrow range of baseline BP in this cohort, both systolic and diastolic pressures at baseline were significantly related to subsequent risk of hypertension. In addition, the average BP over follow-up was a significant predictor of subsequent change in BP. The predictive value of baseline systolic or diastolic pressure on subsequent hypertension was not only independent of important risk factors—such as age, gender, race, alcohol consumption, body mass, heart rate, and intervention assignment in TOHP-1—but also of each other. The increased risk of hypertension in those with a higher initial BP warrants greater preventive effort in this group.
The limitations of the present study were that many of the study participants were recruited at work sites and thus had a relatively high socioeconomic status. Study entrance criteria restricted participants by age and initial BP, resulting in groups with similar characteristics. However, these similarities allowed for comparison of the incidence of hypertension in middle-aged African Americans and whites without confounding by these important factors.
In summary, the present study indicates that middle-aged African Americans and whites have a similar risk of developing hypertension given the same age, level of BP, and BMI at baseline.
Selected Abbreviations and Acronyms
|TOHP-1||=||Trials of Hypertension Prevention, phase 1|
This study was supported by an American Heart Association–Maryland Affiliate Grand-in-Aid (MDBG1195) and partially by Outpatient General Research Center grant 5M01RR00722 from the National Institutes of Health. Dr He was supported by training grant 5T32HL07024-21 from NIH National Heart, Lung, and Blood Institute. The authors would like to thank study coordinator Dolores Kaidy and research technicians Sharon Capelli, Shirley George, Charles Harris, Vicki Harris, Estelle Levitas, Selma Schlenhoff, and Bobbie Weiss at the Johns Hopkins Pro-Health Research Office for their excellent technical assistance during this project.
Reprint requests to Jiang He, MD, PhD, Department of Biostatistics and Epidemiology, Tulane University School of Public Health and Tropical Medicine, 1430 Tulane Ave SL18, New Orleans, LA 70112-2699.
- Received August 25, 1997.
- Revision received September 22, 1997.
- Accepted December 23, 1997.
Manolio TA, Burke GL, Savage PJ, Sidney S, Gardin JM, Oberman A. Exercise blood pressure response and 5-year risk of elevated blood pressure in a cohort of young adults: the CARDIA Study. Am J Hypertens. 1994;7:234–241.
Klag MJ, Thomas J, Mead LA, Thomas DJ, Levine DM, Pearson TA. Ethnicity and risk of hypertension in the Meharry-Hopkins Study. Circulation. 1996;93:623. Abstract.
Klag MJ, Whelton PK, Neaton JD, Randall BL, Ford CE, Brancati FL, Shulman NB, Stamler J. Race and incidence of ESRD: a prospective study. Circulation. 1994;89:8. Abstract.
National Center for Health Statistics, Gordon T. Blood pressure of adults by age and sex, United States, 1960-62. Vital Health Stat. 1964; series 11: No. 4. US Public Health Service publication 1000.
National Center for Health Statistics, Roberts J. Blood pressure of persons 18-74 years, United States, 1971-72. Vital Health Stat. 1975; series 11: No. 150. US Dept of Health, Education, and Welfare (HRA) publication 72-1632.
National Center for Health Statistics, Drizd T, Dannenberg AL, Engel A. Blood pressure levels in persons 18-74 years of age in 1976-80 and trends in blood pressure from 1960-1980 in the United States. Vital Health Stat. 1986; series 11: No. 234. US Dept of Health and Human Services (PHS) publication 86-1684.
Burt VL, Whelton PK, Roccella EJ, Brown C, Cutler JA, Higgins M, Horan MJ, Labarthe D. Prevalence of hypertension in the US adult population: results from the third National Health and Nutrition Examination Survey, 1988–1991. Hypertension. 1995;25:305–313.
Paffenbarger RS Jr, Wing AL, Hyde RT. Physical activity as an index of heart attack risk in college alumni. Am J Epidemiol. 1978;108:161–175.
SAS Institute Inc. SAS/STAT User’s Guide, Release 6.03. Cary, NC: SAS Institute Inc; 1991.
Stevens J, Gautam SP, Keil JE. Body mass index and fat patterning as correlates of lipids and hypertension in an elderly, biracial population. J Gerontol. 1993;48:249–254.
Hypertension Detection and Follow-up Program Cooperative Group. Race, education and prevalence of hypertension. Am J Epidemiol. 1977;106:351–361.
Hutchinson RG, Arnett DK, Watson RL, Barnes RA, Brown S, Davis CE, Tyroler HA. Gender differences in multiple risk factor clustering among blacks and whites: the ARIC Study. Circulation. 1992;88(suppl I):I-357. Abstract.
Eisner GM. Hypertension: racial differences. Am J Kidney Dis. 1990;16(suppl 1):35–40.
Voors AW, Foster TA, Frerichs RR, Webber LS, Berenson GS. Studies of blood pressure in children, aged 5–14 years, in a total biracial community: the Bogalusa Heart Study. Circulation. 1976;54:319–327.
Manatunga AK, Jones JJ, Pratt JH. Longitudinal assessment of blood pressures in black and white children. Hypertension. 1993;22:84–89.
Tyroler HA, Hyden S, Hames CG. Weight and hypertension: Evans County studies of blacks and whites. In: Paul O, ed. Epidemiology and Control of Hypertension. New York, NY: Stratton Intercontinental; 1975:177–204.