Longitudinal Correlates of Change in Blood Pressure in Adolescent Girls
Abstract—The objective of this study was to assess the longitudinal changes in blood pressure in black and white adolescent girls and evaluate potential determinants of changes in blood pressure, including sexual maturation and body size. A total of 1213 black and 1166 white girls, ages 9 or 10 years at study entry, were followed up through age 14 with annual measurements of height, weight, skinfold thickness, stage of sexual maturation, systolic and diastolic blood pressures, and other cardiovascular risk factors. Average blood pressures in black girls were generally 1 to 2 mm Hg higher than in white girls of similar age over the course of the study. Age, race, stage of sexual maturation, height, and body mass index (kg/m2) were all significant univariate predictors of systolic and diastolic blood pressures in longitudinal regression analyses. Black girls had a significantly smaller increase in blood pressure for a given increase in body mass index compared with white girls. The predicted increases in blood pressure per unit increase in body mass index (mm Hg per kg/m2) were as follows: systolic, 0.65±0.04 in whites and 0.52±0.04 in blacks (P<.001); diastolic fourth Korotkoff phase, 0.31±0.04 in whites and 0.15±0.03 in blacks (P<.001); and diastolic fifth Korotkoff phase, 0.31±0.05 in whites and 0.16±0.04 in blacks (P<.001). Understanding of the determinants of the racial differences in blood pressure could provide the rationale for future interventions to reduce the excess cardiovascular mortality in black compared with white women.
Elevated blood pressure is more prevalent in black compared with white women at every age from 18 to 74 years.1 This finding is one factor that may significantly contribute to the increased risk of earlier cardiovascular disease morbidity and mortality observed in black women.2 However, the timing of the onset and course of racial differences in blood pressure is uncertain. The majority of investigations of blood pressure differences by race in childhood and adolescence have been cross-sectional in design. Some of these studies have found higher blood pressure in black children,3 4 others have found no difference,5 6 and still other studies have found higher blood pressure in white children.7 8 The evaluation of potentially important covariates of blood pressure, including age, height, weight, obesity, and sexual maturation, in these studies has been variable. The evaluation of obesity is particularly important in this regard because obesity has been shown to be a risk factor for blood pressure elevation, and some studies have suggested a differential effect of obesity on blood pressure between white and black adults.9 10 11
The NGHS is a prospective cohort study designed to evaluate the emergence of differences in obesity and cardiovascular risk factors during adolescence in black and white girls. The purpose of the present report is to describe the longitudinal changes in blood pressure in black and white girls from age 9 to age 14 years and to determine the presence and persistence of racial differences in blood pressure in the cohort over time. The determinants of longitudinal change in systolic and diastolic blood pressure over time and the presence of racial differences in the growth-related determinants of change were also evaluated.
The NGHS is a collaborative, longitudinal, observational cohort study with data collected at three clinical centers: The University of California at Berkeley, the University of Cincinnati-Children’s Hospital Medical Center, and Westat, Inc, with the Humana Group Health Plan, Washington, DC. The centers at Berkeley and Cincinnati recruited participants from public and parochial schools, and Westat recruited subjects from a health maintenance organization. The Maryland Medical Research Institute, Baltimore, Md, serves as the coordinating center for the study. This study was approved by the appropriate institutional review committees.
NGHS enrolled 1213 black and 1166 white girls who were age 9 or 10 years at the onset of the study. Race was determined by self-declaration by the girls and their parents or guardians. Girls were eligible if they declared themselves black or white and there was concordance between the racial declaration of the parents or guardians and the child. Informed consent for participation was obtained from all participants and their parents.
The subjects were examined once a year (within a 3-month window of the anniversary date). Data in this report include follow-up of the girls through age 14 years (through the fourth or fifth annual follow-up visit, depending on the age of the girl at entry). Age was defined as age at last birthday. Retention decreased slightly with time: at the fourth follow-up visit, 89.9% of black and 83.2% of white participants were seen at the annual visit for an overall participation rate of 86.6%.
The methods for the NGHS have been described previously in detail.12 Measurements of height, weight, skinfold thickness, stage of sexual maturation, and blood pressure were obtained by examiners who were trained and certified to follow a common protocol. Examiners were retrained yearly, and consistent methods of measurement were maintained throughout the study. Height was measured using custom-made stadiometers with the girls wearing no shoes. Weight was measured with calibrated Health-O-Meter electronic scales with girls wearing provided T-shirts or gowns. Skinfold measurements were obtained at the triceps, suprailiac, and subscapular sites with Holtain calipers. All anthropometric measurements were taken twice and repeated a third time if the first two measurements differed by more than 0.5 cm (height), 0.3 kg (weight), or 1.0 mm (skin folds). Degree of overweight and adiposity was assessed using BMI, calculated as (weight in kilograms)/(height in meters squared), and sum of the three skinfold thicknesses. The degree of adiposity was assessed using the ratio of truncal (subscapular and suprailiac) to total sum of skin folds. Blood pressure examiners received 16 hours of instruction, and their techniques were evaluated before certification for blood pressure measurement. Subjects were seated with feet resting on a flat surface and the arm at heart level. The appropriate size cuff was selected based on upper arm circumference. Blood pressure was measured three times with measurements at least 60 seconds apart using a standard mercury sphygmomanometer. The second and third measurements of blood pressure were used to calculate the mean systolic and diastolic K4 and K5 blood pressure.12 In the present analyses, measured values of diastolic K5 blood pressure <20 mm Hg were treated as missing data. The diastolic K5 blood pressure measurement was substituted for the diastolic K4 measurement when K4 could not be distinguished from K5. Sexual maturation staging was performed using criteria developed by Garn and Falkner based on two indices, pubic hair and areolar development. These stages were based on Tanner staging principles and modified for the study of obesity.12 Specific areolar and pubic hair staging plates were developed for evaluation of maturation. The maturation stages were defined as follows: prepubertal girls were premenarcheal and at stage 1 in both areolar and pubic hair development; pubertal girls were premenarcheal, but at stage 2 or greater in either areolar or pubic hair development; and postmenarcheal girls had reported beginning their menses.
To examine racial differences with age, descriptive statistics by race for anthropometry and blood pressure measures (means and standard deviations) were calculated by age at last birthday. Figures are presented showing mean blood pressure levels by age and race; by age, race, and maturation stage; and by race and BMI at a specific study visit (corresponding to ages 11 and 12 years) to display associations of these variables with blood pressure. Means are not presented for age-race-maturation stage or race-BMI combinations for which fewer than 20 girls were observed. Because the girls in the cohort were ages 9 or 10 at entry, and participants were seen annually, data from different study visits were pooled in calculating statistics by age, such as mean BMI at age 10.
Regression models were fitted to estimate the effects of age, race, height, BMI, and maturation stage on blood pressure. To account for the correlations among repeated observations from the same participant, the GEE method of Liang and Zeger13 was used to fit linear regression models. With this method, least-squares estimates of models were obtained, and residuals from the fitted model were used to adjust the standard errors of model coefficients to take account of within-individual correlations.
Increased sexual maturation is accompanied both by growth in height and in adiposity (estimated by BMI). Multivariable regression models were fitted incorporating terms for race, age, race×age interactions, and maturation stage. Terms for height and for BMI were added successively to these models, to see whether either or both accounted for the association of sexual maturation with increases in blood pressure. Tests were performed to determine whether the effects of height or BMI on blood pressure differed between blacks and whites. BMI was used as a measure of adiposity in these models for comparability with other studies. Analyses were also performed using sum of skin folds as a measure of adiposity rather than BMI; the results of those models were generally consistent with the analyses using BMI reported here.
Height and BMI were included in these models as continuous variables. Age in years at last birthday was included in these models by a set of indicator variables for ages 10 to 14 years, with age 9 years as the reference level. Maturation stage was included with indicator variables comparing pubertal and postmenarcheal girls to prepubertal girls. Race was included using an indicator variable with white girls used as the reference point. Tests of interaction between race and age and between race and maturation stage were performed using global tests for any difference between blacks and whites in the effect of age or maturation, before testing for black/white differences at particular ages or maturation stages. Otherwise, these exploratory analyses did not attempt to adjust values of probability to take into account the multiple tests required to arrive at the final model. Interactions were also tested between race and height and race and BMI. Values of P≤.05 were considered to indicate evidence of associations worth further investigation.
At the baseline visit (visit 1), NGHS enrolled 539 black and 616 white girls at age 9 years and 674 black and 550 white girls at age 10 years. Mean values and standard deviations for systolic, K4 diastolic, and K5 diastolic blood pressures by age and race are shown in Table 1⇓. Mean values and standard deviations for anthropometric variables by age and race are shown in Table 2⇓. The smaller numbers for age 9 years reflect that only 49% of the cohort enrolled in the study at age 9 years. Ages 10 years and older represent pooled data from two successive study visits.
The distribution of sexual maturation stage by age and race is shown in Table 3⇓. Sexual maturation begins earlier for black than for white girls: a larger proportion of black girls than white girls have begun puberty at age 9 years, and by age 12 years, over 60% of the black girls have reached menarche compared with only 39% of the white girls. By age 14 years, over 90% of the girls from both race groups have reached menarche.
In Fig 1A⇓ through 1C⇓, mean values by race of longitudinal blood pressures (systolic, diastolic K4, diastolic K5) are displayed from age 9 to 14 years. Systolic and diastolic blood pressures increased with increasing age for both black and white girls, and black girls had higher average blood pressures than the white girls of corresponding age over the course of the study, with the smallest differences at age 13 years. This variation by age in the black/white difference in blood pressure is statistically significant for systolic blood pressure (P=.02) and diastolic K4 blood pressure (P=.03), but borderline significant for diastolic K5 blood pressure (P=.06).
In Fig 2A⇓ through 2C⇓, mean systolic and diastolic blood pressure levels are displayed by age, sexual maturation stage, and race. Within maturation stage groups, black girls generally have higher blood pressure than white girls at each age, although there is variation in the magnitude of the black/white difference over ages 9 to 14 years. Because increased sexual maturation is generally accompanied by increases in height and in BMI, longitudinal regression models were used to investigate whether the effects of sexual maturation on blood pressure were accounted for by these factors.
Height was a highly significant predictor of blood pressure in all models, although adjustment for BMI and maturation stage reduced the predicted increase in blood pressure per centimeter increase in height from 0.28±0.02 mm Hg to 0.20±0.02 mm Hg for systolic and from 0.20±0.02 mm Hg to 0.15±0.02 for diastolic (K4 or K5). The effect of height on blood pressures did not differ by race (P>.3 for all comparisons).
After adjustment for age, race, and height, blood pressure remained related to sexual maturation stage. Mean blood pressure increased with increasing sexual maturation. After adjustment, it was estimated that the mean±SE blood pressure differences between pubertal and prepubertal girls were 1.2±0.3 mm Hg for systolic blood pressure (P<.001), 0.9±0.4 mm Hg for diastolic K4 blood pressure (P=.01), and 1.0±0.4 mm Hg for diastolic K5 blood pressure (P=.01). Similarly, after adjustment for age, race, and height, the estimated mean±SE blood pressure difference between postmenarcheal and pubertal girls was 1.6±0.3 mm Hg for systolic blood pressure (P<.001), 0.9±0.3 mm Hg for diastolic K4 blood pressure (P=.001), and 1.0±0.3 mm Hg for diastolic K5 blood pressure (P=.001). The relationship between maturation stage and diastolic (K4 or K5) blood pressure was similar in black and white girls, but there was some evidence (P=.05) for black/white differences in the effect of maturation stage on systolic blood pressure.
After the addition of BMI to the previous model, differences in blood pressure by maturation stage were no longer statistically significant, although trends remained for higher blood pressures with increasing sexual maturation. After age, race, height, BMI, and the age×race and race×BMI interactions were added to the GEE regression model, the estimated mean±SE blood pressure differences between pubertal and prepubertal girls were as follows: systolic, 0.3±0.3 mm Hg (P=.29); diastolic K4, 0.6±0.4 mm Hg (P=.10); and diastolic K5, 0.7±0.4 mm Hg (P=.10). The estimated differences (±SE) in blood pressure between postmenarcheal and pubertal girls were as follows: systolic, 0.5±0.3 mm Hg (P=.08); diastolic K4, 0.5±0.3 mm Hg (P=.09); and diastolic K5, 0.6±0.3 mm Hg (P=.06). After adjustment for both height and BMI, estimated differences in blood pressure between prepubertal and pubertal, and pubertal and postmenarcheal girls were similar in blacks and in whites.
The effect of BMI on blood pressure differed between blacks and whites (P=.02, P=.004, and P=.02 for tests for interaction for systolic, diastolic K4, and diastolic K5 blood pressures, respectively). Using regression models, the average (±SE) increase in blood pressure per unit increase in BMI was larger in whites than in blacks: 0.65±0.04 mm Hg in whites (P<.001) and 0.52±0.04 in blacks (P<.001) for systolic blood pressure. For diastolic K4 and diastolic K5 blood pressure, the increase per unit increase in BMI in whites was approximately double that for blacks, 0.31±0.04 mm Hg in whites (P<.001) and 0.15±0.03 in blacks (P<.001) for diastolic K4 blood pressure; and, 0.31±0.05 mm Hg in whites (P<.001) and 0.16±0.04 in blacks (P<.001) for diastolic K5 blood pressure.
To illustrate, the relationship between BMI and blood pressure is depicted in Fig 3A⇓ through 3C⇓ for black and white girls at study visit 3 (ages 11 to 12 years). Whereas black girls tend to have higher blood pressure than white girls at lower levels of BMI, there is a steeper increase in blood pressure with increasing BMI in white girls than in black girls. This results in 11- and 12-year-old black and white girls having similar blood pressures at higher levels of BMI. The relationship between blood pressure and BMI in black and white girls at other ages was similar to that observed at ages 11 to 12 years. Because the effect of BMI on blood pressure differed in black and white girls, as demonstrated by GEE modeling, the magnitude of black/white differences in blood pressure depended on BMI. The magnitude of these racial differences at each level of BMI also depended on age (each P<.05).
To evaluate the potential impact of regional differences in adiposity, the ratio of truncal to total sum of skin folds was included in the final model. This variable was not significantly associated with either systolic or diastolic blood pressure.
Essential hypertension is a prevalent condition affecting approximately 50 million adults in the United States.14 Whereas the prevalence seems to have decreased in the past decade, essential hypertension remains an important public health problem.1 Essential hypertension is a major cause of cardiovascular morbidity and mortality.15 It has been well established that hypertension is more common in black than in white adults, and this finding is especially true for women (eg, for women aged 18 to 74 years, using a cutpoint of 140/90 mm Hg, black women were found to have an age-adjusted prevalence of hypertension of 28.1% versus 16.7% for white women).1 This difference in blood pressure is reflected in the racial difference in mortality rates from hypertension-related disease, with a nearly 4-fold increase in death from cardiovascular disease in black compared with white women.16
The timing of the onset of racial differences in blood pressure remains controversial. Cross-sectional studies in children and adolescents have yielded conflicting results. It is well known that cross-sectional studies may not provide a valid assessment of longitudinal changes.17 Manatunga et al18 conducted a longitudinal study of 345 white children and 164 black children with mean age of 9 years at baseline. In their study, blood pressure was measured every 6 months for 2 to 5.5 years. For girls the mean systolic and diastolic blood pressures were higher by 2 mm Hg and 1.5 mm Hg, respectively, in blacks compared with whites. After accounting for age, weight, height, and BMI, the racial difference in systolic blood pressure remained significant, but there was no significant difference in diastolic blood pressure between black and white girls. The rate of increase in systolic and diastolic blood pressure with age was significantly greater for black girls compared with white girls.
The results of the present study are consistent with those of Manatunga et al.18 At enrollment, there were already differences between the black and white girls, with the black girls having higher blood pressure. Systolic and diastolic blood pressure progressively increased with increasing age for both black and white girls, and black girls consistently had higher blood pressure than white girls. However, Manatunga et al did not evaluate the role of sexual maturation on blood pressure change. Kozinetz19 20 has suggested from cross-sectional analyses using self-assessment of maturation that there is an association between stage of sexual maturation and blood pressure in girls, and that observed racial differences in blood pressure could be explained by racial differences in the level of pubertal development. However, in the present study in which maturation was assessed according to a standard protocol by trained observers, black/white differences in blood pressure are seen at all stages of sexual maturation. This result is consistent with the concept that additional factors other than sexual maturation are also important in the racial difference of blood pressure in adolescent girls.
Blood pressure increases with increasing height and BMI in children and adolescents.21 22 23 Therefore, it is important to determine the extent to which height and BMI influence the observed longitudinal racial differences in blood pressure. Height was a significant explanatory variable in the regression models for systolic as well as diastolic K4 and K5 blood pressures. There was no interaction between race and height, suggesting that the effect of linear growth on blood pressure is similar for white and black girls. Although black girls were initially taller on average, by age 14, average height was greater in white than in black girls.
BMI was also a significant independent variable in the regression models for systolic and diastolic blood pressures. This finding is consistent with previous cross-sectional epidemiologic studies22 23 and with studies demonstrating that weight loss lowers blood pressure in obese adolescents.24 When height and BMI are both included in the regression model, the effect of sexual maturation on blood pressure is greatly attenuated. This finding supports the hypothesis that the increases in height and BMI accompanying sexual maturation are major contributors to changes in systolic and diastolic blood pressure seen as girls go through the sexual maturation process.
In the present study, there was a significant interaction between the effects of race and BMI on blood pressure. BMI had a smaller impact on blood pressure in black than in white girls. However, average BMI increased at a greater rate with age in black than white girls, which may have contributed to the black/white differences in blood pressure seen in Fig 1⇑. It is of interest that the distribution of adiposity was not a significant correlate of blood pressure in this study.
There has been ongoing debate about the presence of racial differences in the relationship between BMI and blood pressure in adults. Some studies of adult populations have suggested that the association between BMI and blood pressure is weaker in blacks than whites,9 10 11 whereas others have found no racial differences in the relationship.25 26 27 28 In the present study, the racial difference in the BMI–blood pressure relationship results in similar blood pressures for black and white girls at higher levels of BMI, whereas black girls seem to have higher blood pressures than whites at lower levels of BMI. These differences are not accounted for by black/white differences in sexual maturation stage. The etiology of the racial difference in blood pressure at lower BMI remains unexplained. This difference could be due to factors not included in the present analysis, such as dietary intake of sodium and potassium, sensitivity to sodium intake, or physical activity.29 During ages 9 to 14 years, BMI increased with age at a higher average rate in black girls than in white girls. This greater rate of increase in BMI in black girls helps account for the maintenance of racial differences in blood pressures during ages 9 to 14 years, despite the finding of a lower increase in blood pressure per unit increase in BMI in black girls compared with white girls. The mechanism for a racial difference in the relationship between BMI and blood pressure remains unknown. Future research should focus on a better understanding of this relationship.
NGHS demonstrates significant racial differences in blood pressure in girls 9 to 14 years of age, the years of prepuberty and early puberty to late puberty. If the higher blood pressures observed in black girls compared with whites continue during late adolescence and young adulthood, and if BMI increases to a greater extent in young black women, then large differences in blood pressure can be expected to be seen between black and white women. Understanding the determinants of this racial difference in blood pressure could provide the rationale for future interventions to reduce the excess cardiovascular morbidity and mortality seen in black women.
Selected Abbreviations and Acronyms
|BMI||=||body mass index|
|GEE||=||generalized estimating equations|
|K4||=||fourth Korotkoff phase|
|K5||=||fifth Korotkoff phase|
|NGHS||=||National Heart, Lung, and Blood Institute Growth and Health Study|
This work was supported by National Heart, Lung and Blood Institute contracts HC 55023–26.
- Received February 24, 1997.
- Revision received May 5, 1997.
- Accepted August 13, 1997.
Burt VL, Cutler JA, Higgins M, Horan MJ, Labarthe D, Whelton P, Brown C, Rocella EJ. Trends in the prevalence, awareness treatment and control of hypertension in the adult US population: data from the Health Examination Surveys, 1960 to 1991. Hypertension. 1995;26:60–69.
National Center for Health Statistics. Vital Statistics of the United States, 1988. Vol 2, Part A, Mortality. Washington, DC: Public Health Service; 1991.
Voors AW, Foster TA, Frerichs RR, Webber LS, Berenson GS. Studies of blood pressures in children, ages 5–14 years, in a total biracial community: the Bogalusa Heart Study. Circulation. 1986;54:319–327.
Cornoni-Huntley J, Harlan WR, Leaverton PE. Blood pressure in adolescence: the United States Health Examination Survey. Hypertension. 1979;1:566–571.
Rowland M, Roberts J. Blood pressure levels and hypertension in persons aged 6–74 years: United States, 1976–80. Hyattsville, Md: National Center for Health Statistics. Advancedata. 1981;84:1–12.
Blair D, Habicht J-P, Sims EAH, Sulwester D, Abraham S. Evidence for an increased risk for hypertension with centrally located body fat and the effect of race and sex on this risk. Am J Epidemiol. 1984;119:526–540.
Tyroler HA, Heyden 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 Medical Book; 1975.
Liang KY, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika. 1986;73:13–22.
Fifth Report of the Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure. National High Blood Pressure Education Program. Washington, DC: NIH publication No. 93–1088; 1993.
Stamler J. Metabolic and nutritional factors in hypertension. Blood pressure and high blood pressure: aspects of risk. Hypertension. 1991;18(suppl I): I-95–I-107.
Gottdiener JS. Hypertensive heart disease in blacks. In: Saunders E, ed. Cardiovascular Disease in Blacks. Philadelphia, Pa: FA Davis Co; 1991.
Dawson-Saunders B, Trapp RG. Study designs in clinical research. In: Basic and Clinical Biostatistics. Norwalk, Conn: Appleton and Lange; 1990.
Manatunga AK, Jones JJ, Pratt H. Longitudinal assessment of blood pressures in black and white children. Hypertension. 1993;22:84–89.
Burns TL, Moll PP, Lauer RM. The relation between ponderosity and coronary risk factors in children and their relatives: the Muscatine Ponderosity Family Study. Am J Epidemiol. 1989;129:973–987.
Schieken RM, Mosteller M, Goble MM, Moskowitz WB, Hewitt JK, Eaves LJ, Nance WE. Multivariate genetic analysis of blood pressure and body size: the Medical College of Virginia Twin Study. Circulation. 1992;86:1780–1788.
Rocchini AP, Katch V, Anderson J, Hinderliter J, Becque D, Martin M, Marks C. Blood pressure in obese adolescents: effect of weight loss. Pediatrics. 1988;82:16–23.
Gillum RF. Pathophysiology of hypertension in blacks and whites: a review of the basis of racial blood pressure differences. Hypertension. 1979;1:468–475.
Stamler J, Berkson DM, Dyer A, Lepper MH, Lindberg HA, Paul O, McKean H, Rhomberg P, Schoenberger JA, Shekelle RB, Stamler R. Relationship of multiple variables to blood pressure: findings from four Chicago epidemiologic studies. In: Paul O, ed. Epidemiology and Control of Hypertension. New York, NY: Stratton Intercontinental Medical Book; 1975.