Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 May 3.
Published in final edited form as: Am J Hypertens. 2011 Oct 6;25(2):140–144. doi: 10.1038/ajh.2011.177

Systolic Blood Pressure and Adiposity: Examination by Race and Gender in a Nationally Representative Sample of Young Adults

Beverly H Brummett 1, Michael B Babyak 1, IIene C Siegler 1, Richard Surwit 1, Anastasia Georgiades 1, Stephen H Boyle 1, Redford B Williams 1
PMCID: PMC3642773  NIHMSID: NIHMS458117  PMID: 21976277

Abstract

BACKGROUND

Adiposity, or more specifically, underlying body fat distribution, has been associated with systolic blood pressure (SBP), and it has been suggested that these associations vary between whites and blacks, as well as by gender.

METHODS

Here, we use data from the National Longitudinal Study of Adolescent Health (Add Health), a US study of over 15,000 participants (median age 29.0 years), to characterize the associations between measures of body fat distribution—waist circumference (WC) and WC adjusted for body mass index (BMI) (WC-bmi)—with SBP within white and black race and gender subgroups.

RESULTS

Our findings suggest that, at lower levels of WC-bmi, white women have significantly higher SBP as compared to black women, whereas black men have higher SBP than white men. Black women with WC-bmi >90 cm have higher SBP compared to white women with similar WC-bmi, whereas among black and white men the associations are essentially similar across the full range of WC-bmi.

CONCLUSIONS

The present results suggest that associations among anthropometric measures of adiposity and blood pressure are nonlinear, and importantly, vary for whites and blacks by gender. In black women, SBP increased more as WC increased from low- to mid-range levels, whereas it was only at higher WC levels that black men exhibited higher SBP than white men.

Keywords: blood pressure, body mass index, gender, hypertension, race, systolic blood pressure, waist circumference

It has been suggested that the association between anthropometric measures of adiposity (e.g., waist circumference (WC) and body mass index (BMI)) and high blood pressure vary by race.1 These anthropometric measures, however, only broadly capture more complex underlying patterns of body composition and adiposity (e.g., visceral vs. subcutaneous adipose tissue, upper body vs. abdominal fat), which also vary by sex and race.2 Despite their relative lack of specificity with respect to adiposity, techniques for assessing anthropometric measures such as BMI and WC are comparatively inexpensive and widely accessible. Thus, further elucidation of how these relatively simple measures are associated with disease remains important.

The present study describes the association between WC and systolic blood pressure (SBP), with a focus on how these relations may vary by race and sex, using a large nationally representative sample of young adults in the United States. We focus on SBP in the present study because, although associations have been shown to vary by age,3 many studies have shown SBP to be more important than diastolic blood pressure (DBP) with respect to health risk,4–9 and also possibly more responsive than DBP to changes in modifiable risk factors such as exercise, diet, and weight loss.10 WC has been shown to add to the prediction of visceral fat beyond measurements of BMI.11,12 Visceral adipose tissue is an independent predictor of the metabolic syndrome in both men and women,13 and has also been found to add to the prediction of hypertension.14,15 Following Tybor et al.,16 we use statistical partialling to remove the effect of BMI from the association between WC and SBP. WC partialled for BMI reflects WC independent of overall body size, and thus, by definition, must reflect body shape or body fat distribution. We also allow for the possibility that these associations may be nonlinear. Examination of these questions in this manner may provide more precise information about the pathophysiology underlying high blood pressure and also further guide risk stratification in clinical practice.

METHODS

Participants

The present study uses data from the Add Health study, a nationally representative sample of ~15,000 young adults that was designed to assess the effects of health-related behaviors during adolescence and into young adulthood. The study began with adolescents were in grades 7–12 in the United States during the 1994–1995 school year, and the cohort has been followed into young adulthood with four in-home interviews. The last interview was in 2008 and the sample was aged 24–32. Add Health participants have been assessed with regard to an extensive number of psychosocial variables, e.g., social, economic, and psychological well-being. In addition, specific contextual data on the family, neighborhood, and other social environments may be linked to health outcomes in young adulthood. Wave four interviews were expanded to collect biological data, e.g., blood pressure. The Add Health study was reviewed and approved by the institutional review board at the University of North Carolina–Chapel Hill and written consent was obtained at all waves from either the respondent or the parent if the respondent was under age 18. For detailed information regarding the study and its design, please see ref. 17. Data for the present study were collected at Wave IV, as SBP was not assessed at prior Waves.

Measures

Age was recorded in years; and gender 0 = men; 1 = women. Race was self-identified, and coded as white = 0, black = 1. education was coded as: 0 = some high school or less, or G.E.D.; 1 = graduated high school; 2 = some college or vocational-tech; 3 = bachelor’s degree; 4 = some graduate school or more.

For detailed information regarding assessment and reliability of SBP, see ref. 18. Trained and certified field interviewers measured respondents’ resting, seated systolic and DBPs (mm Hg) and pulse rate (beats/min). Following a 5-min seated rest three serial measurements were performed at 30-s interval using a factory calibrated, Microlife BP3MC1-PC-IB oscillometric blood pressure monitor (MicroLife USA, Dunedin, FL). SBP was constructed as the average of measures two and three. For cardiac medication status, participants were first asked, “Have you taken any medications in the last four weeks?” If the respondent answered yes, they were asked to collect their medications and the interviewer recorded a list of medications provided. We represented medication using a dichotomous yes/no variable that included all classes of medication prescribed that might lower blood pressure.

Height was measured to the nearest 0.5 cm. Weight was measured to the nearest 0.1 kg. BMI was calculated: BMI (kg/m2) = weight (kg)/height (m2). Height and weight were measured at Wave IV. WC was measured to the nearest 0.5 cm at the superior border of the iliac crest, including pregnant women.

Alcohol consumption was defined as: 0 = nondrinker; 1 = occasional drinker, drink 2 or fewer days of the week; 2 = light drinker, drink 5–7 days/week and drink 2 or fewer drinks (1 or fewer if female); 3 = moderate drinker, drink 5–7 days/week, 3 drinks for males, 2 drinks for females; 4 = heavy drinker, drink 5–7 days/week, >3 drinks for males, >2 for females. Smoking was represented by a dichotomous yes/no variable that indicated whether or not the participant was a daily smoker.

Statistical analyses

Sample characteristics were described using medians and interquartile ranges for continuous variables and frequency and percentage for categories. In a preliminary analysis, we examined mean SBP levels across race and gender using the general linear model. We then used ordinary least squares regression to examine associations between body size variables and SBP. In each model, we included age, level of education, smoking, alcohol consumption, and use of cardiac medication as adjustment variables. We estimated two regression models the first included WC but not BMI, the second included both WC and BMI. We modeled SBP, BMI, and WC as continuous variables, allowing the effect of BMI and WC variables to be nonlinear, using a three-knot restricted cubic splines,19 a flexible nonparameteric smoother. Models contained two three-way interaction terms, gender × race × waist and gender × race × BMI, as well as all corresponding lower-order terms. BMI and waist were allowed to be nonlinear within each interaction term, again using a three-knot spline. The primary effects of interest were the three-way interactions described above. Regression models were weighted using the grand sample weights, and within-school clustering was accounted for by including the school cluster codes in the model as covariates. The sample was limited to the 10,102 black and white study participants. We further excluded participants who had missing data on the BMI or waist variables, or who had no sample weight assigned. This resulted in a final sample size of 9,906 participants for the analysis. This small percentage (1.6%) of missing data generally does not require imputation.20

RESULTS

See Table 1 for presentation of sample characteristics. In an initial analysis, we examined the levels of SBP across race and gender, adjusting for age, level of education, cardiac medication use, smoking, and alcohol intake. The gender × race interaction was statistically significant (P < 0.001). Black women had a higher predicted mean SBP of 123.8 (95% confidence interval = 122.8, 124.9) mm Hg, compared to white women, 120.1 (95% confidence interval = 119.5, 120.8) mm Hg (Tukey’s adjusted P < 0.001). SBP also was slightly higher among black men, 130.9 (95% confidence interval = 129.9, 132.0) mm Hg compared to white men, 130.0 (95% confidence interval = 129.3, 130.6) mm Hg, though this difference was not statistically significant (Tukey adjusted P = 0.31).

Table 1.

Background characteristics of sample

White men Black men White women Black women Combined
N = 3,413 N = 1,144 N = 3,757 N = 1,592 N = 9,906
Age, years 29.1 (1.7) 29.0 (1.8) 28.8 (1.7) 28.9 (1.7) 28.9 (1.8)
Systolic blood pressure, mm Hg 129.8 (11.9) 130.1 (13.3) 119.3 (12.1) 122.1 (14.1) 124.6 (13.5)
Medication 4.2% (144) 3.7% (42) 3.6% (134) 5.5% (87) 4.1% (407)
Education
   <HS 7.4% (252) 8.7% (100) 5.3% (199) 6.7% (106) 6.6% (657)
   HS 17.6% (599) 18.7% (214) 13.2% (495) 11.6% (185) 15.1% (1,493)
   Some college 43.5% (1,483) 48.3% (552) 42.2% (1,586) 47.7% (760) 44.2% (4,381)
   College 21.2% (724) 14.3% (164) 23.2% (871) 17.8% (284) 20.6% (2,043)
   Postgraduate 10.4% (355) 10.0% (114) 16.1% (606) 16.1% (257) 13.5% (1,332)
Body mass index, kg/m2 28.6 (6.3) 29.0 (6.7) 28.2 (7.4) 31.5 (8.9) 28.9 (7.3)
Waist circumference, cm 99.2 (14.4) 97.8 (16.7) 95.7 (17.6) 100.6 (19.2) 97.9(16.9)
Alcohol use
   None 17.9% (610) 32.4% (371) 24.0% (900) 40.8% (650) 25.6% (2,531)
   Light 62.6% (2,137) 54.1% (619) 68.5% (2,574) 55.0% (875) 62.6% (6,205)
   Moderate 4.6% (156) 3.4% (39) 3.2% (121) 1.9% (30) 3.5% (346)
   Heavy 3.6% (124) 3.6% (41) 1.8% (69) 1.3% (20) 2.6% (254)
   Very heavy 11.3% (386) 6.5% (74) 2.5% (93) 1.1% (17) 5.8% (570)
Smoking 28.5% (972) 20.0% (229) 24.7% (926) 12.1% (192) 23.4% (2,319)
Open in a new tab

Values for continuous variables are mean (s.d.). Numbers after percents are frequencies.

HS, high school.

In both regression models, which included WC and then WC and BMI simultaneously, the race × gender × WC interaction term was statistically significant (P = 0.02, P = 0.03, respectively).

In addition, in the full model, the race × gender × BMI interaction also was statistically significant (P < 0.001). Among the adjustment covariables, cardiac medication, smoking, and heavier alcohol intake were associated with higher SBP. Age also was positively related to SBP, but was nonsignificant. See Supplementary Appendix online for the regression results for these latter variables. Regression weights from spline terms are not readily interpretable, particularly when involved in interactions. We therefore focus on interpreting the interaction graphically. The left panel of Figure 1 displays the association between SBP and WC without adjustment for BMI for each race and gender combination. When WC was relatively low (roughly <70 cm), Black and white women tended to have similar SBP. At the middle range of WC (~70–100 cm), however, the regression slope is somewhat steeper for black women than for white women, with black women exhibiting higher SBPs across this range, the races converging again at WC levels of about 140 cm. White and black men tended to have similar SBP when WC was under ~100 cm, but black men exceeded white men when WC exceeded ~100 cm. The right panel of Figure 1 displays the predicted values for WC after adjusting for BMI (WC-bmi). White women with WC-bmi lower than about 70 cm had higher SBP than black women in the same range, but their SBP slope was relatively flat across the entire range of WC-bmi. The slope for black women was curvilinear, with the difference between black and white women quickly becoming more pronounced for WC-bmi values greater than about 70 cm. Black men with WC-bmi lower than about 70 cm had higher SBP compared to White men in the same WC-bmi range, whereas the difference diminishes after WC-bmi approaches ~100 cm. For both unadjusted and adjusted analyses, the wide confidence limits for the very high levels of WC and WC-bmi reflect the relative sparseness of the data in that region, thus limiting inferences for those individuals with WC and WC-bmi in that region.

Figure 1.

Figure 1

Open in a new tab

Interaction of race, gender, and waist circumference predicting systolic blood pressure (SBP), with and without adjusting for body mass index (BMI). Shaded areas represent 95% confidence intervals for parameter.

DISCUSSION

The present findings, in a large sample of young adults, suggest that the relations between SBP and WC differ for whites and blacks, as well as by gender, and that these associations are not linear. In black women, SBP increased more as WC increased from low- to mid-range levels. In contrast, it was only at higher WC levels that black men exhibited higher SBP than white men. Adjustment for BMI tended to amplify the difference between black and white women, but diminish the race difference in males. The additional predictive ability of WC might reflect the involvement of fat depots more closely associated with WC than BMI, such as visceral fat.11 Studies of more direct measures of visceral fat will be needed to test this hypothesis. Very few studies appear to have examined both race and gender differences with respect to BMI-adjusted WC and blood pressure. Janssen et al.21 found that after adjustment for BMI, WC remained a significantly associated with hypertension in both men and women in a cohort of middle-aged adults (mean age ~43 years), but did not examine ethnic differences (nor is the ethnic composition of the sample reported). Of strictly methodological interest, Janssen et al. also showed that the use of artificially categorized body size variable led to potential distortions in the estimation of associations. Tybor et al.’s16 longitudinal study did examine race and gender differences in the association between BMI-adjusted WC and blood pressure and found that changes in BMI-adjusted WC predicted changes in SBP and DBP, but only in white women. In our cross-sectional design, the association between BMI-adjusted WC and SBP was actually weakest among white women. Among the critical differences between the present study and Tybor et al., is that as the present study used data from young adults around age 29, whereas Tybor et al. studied adolescents through the periods of about age 11–19 years. Thus, differences in findings across studies may reflect developmental differences in physiology. It also may be that Tybor et al.’s longitudinal design may be addressing a different question than the present study. Schaie22 for example, has long argued that cross-sectional data cannot be extrapolated to within-person change.

Strengths of the present study include the large, well-described group of participants taken from a nationally representative sample and highly reliable measures of body shape and blood pressure. The large sample also enabled us to examine relatively complex interactions with some confidence in the power and precision of the estimates. Consequently, our analytic approach allows for a richer, more nuanced description of the associations under study. There also are, however, several limitations to our approach. First, as with all observational studies, unmeasured confounding or explanatory variables may be producing spurious associations. Second, it is simply not practical to examine all possible interdependent effects. In other words, there may be other variables with which race, gender, BMI, and WC interact, such as age, income, education, alcohol intake, etc. To summarize, the mechanisms by which WC confers additional cardiovascular risk are complex and seem to differ between whites and blacks, as well as by race and gender.

Supplementary Material

Supplemental

Acknowledgments

This research uses data from Add Health, a program project directed by Kathleen Mullan Harris and designed by J. Richard Udry, Peter S. Bearman, and Kathleen Mullan Harris at the University of North Carolina at Chapel Hill, and funded by grant P01 HD31921 from the Eunice Kennedy Shriver National Institute of Child Health and Human Development, with cooperative funding from 23 other federal agencies and foundations. Special acknowledgment is due Ronald R. Rindfuss and Barbara Entwisle for assistance in the original design. Information on how to obtain the Add Health data files is available on the Add Health website (http://www.cpc.unc.edu/addhealth). This research was also supported by Grant Number P01 HL36587 from the National Heart Lung and Blood Institute.

Footnotes

Supplementary material is linked to the online version of the paper at http://www.nature.com/ajh

Disclosure: The authors declared no conflict of interest.

References

  • 1.Rosner B, Prineas R, Daniels SR, Loggie J. Blood pressure differences between blacks and whites in relation to body size among US children and adolescents. Am J Epidemiol. 2000;151:1007–1019. doi: 10.1093/oxfordjournals.aje.a010129. [DOI] [PubMed] [Google Scholar]
  • 2.Camhi SM, Bray GA, Bouchard C, Greenway FL, Johnson WD, Newton RL, Ravussin E, Ryan DH, Smith SR, Katzmarzyk PT. The relationship of waist circumference and BMI to visceral, subcutaneous, and total body fat: sex and race differences. Obesity (Silver Spring) 2011;19:402–408. doi: 10.1038/oby.2010.248. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Perry HM, Jr, Miller JP, Baty JD, Carmody SE, Sambhi MP. Pretreatment blood pressure as a predictor of 21-year mortality. Am J Hypertens. 2000;13:724–733. doi: 10.1016/s0895-7061(99)00214-9. [DOI] [PubMed] [Google Scholar]
  • 4.Beevers DG. Epidemiological, pathophysiological and clinical significance of systolic, diastolic and pulse pressure. J Hum Hypertens. 2004;18:531–533. doi: 10.1038/sj.jhh.1001702. [DOI] [PubMed] [Google Scholar]
  • 5.Strandberg TE, Pitkala K. What is the most important component of blood pressure: systolic, diastolic or pulse pressure? Curr Opin Nephrol Hypertens. 2003;12:293–297. doi: 10.1097/00041552-200305000-00011. [DOI] [PubMed] [Google Scholar]
  • 6.Izzo JL, Jr, Levy D, Black HR. Clinical Advisory Statement. Importance of systolic blood pressure in older Americans. Hypertension. 2000;35:1021–1024. doi: 10.1161/01.hyp.35.5.1021. [DOI] [PubMed] [Google Scholar]
  • 7.Leonetti G, Cuspidi C, Facchini M, Stramba-Badiale M. Is systolic pressure a better target for antihypertensive treatment than diastolic pressure? J Hypertens Suppl. 2000;18:S13–S20. [PubMed] [Google Scholar]
  • 8.Nielsen WB, Lindenstrøm E, Vestbo J, Jensen GB. Is diastolic hypertension an independent risk factor for stroke in the presence of normal systolic blood pressure in the middle-aged and elderly? Am J Hypertens. 1997;10:634–639. doi: 10.1016/s0895-7061(96)00505-5. [DOI] [PubMed] [Google Scholar]
  • 9.Staessen JA, Gasowski J, Wang JG, Thijs L, Den Hond E, Boissel JP, Coope J, Ekbom T, Gueyffier F, Liu L, Kerlikowske K, Pocock S, Fagard RH. Risks of untreated and treated isolated systolic hypertension in the elderly: meta-analysis of outcome trials. Lancet. 2000;355:865–872. doi: 10.1016/s0140-6736(99)07330-4. [DOI] [PubMed] [Google Scholar]
  • 10.Blumenthal JA, Babyak MA, Hinderliter A, Watkins LL, Craighead L, Lin PH, Caccia C, Johnson J, Waugh R, Sherwood A. Effects of the DASH diet alone and in combination with exercise and weight loss on blood pressure and cardiovascular biomarkers in men and women with high blood pressure: the ENCORE study. Arch Intern Med. 2010;170:126–135. doi: 10.1001/archinternmed.2009.470. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Janssen I, Heymsfield SB, Allison DB, Kotler DP, Ross R. Body mass index and waist circumference independently contribute to the prediction of nonabdominal, abdominal subcutaneous, and visceral fat. Am J Clin Nutr. 2002;75:683–688. doi: 10.1093/ajcn/75.4.683. [DOI] [PubMed] [Google Scholar]
  • 12.Lee K, Lee S, Kim YJ, Kim YJ. Waist circumference, dual-energy X-ray absortiometrically measured abdominal adiposity, and computed tomographically derived intra-abdominal fat area on detecting metabolic risk factors in obese women. Nutrition. 2008;24:625–631. doi: 10.1016/j.nut.2008.03.004. [DOI] [PubMed] [Google Scholar]
  • 13.Demerath EW, Reed D, Rogers N, Sun SS, Lee M, Choh AC, Couch W, Czerwinski SA, Chumlea WC, Siervogel RM, Towne B. Visceral adiposity and its anatomical distribution as predictors of the metabolic syndrome and cardiometabolic risk factor levels. Am J Clin Nutr. 2008;88:1263–1271. doi: 10.3945/ajcn.2008.26546. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Hayashi T, Boyko EJ, Leonetti DL, McNeely MJ, Newell-Morris L, Kahn SE, Fujimoto WY. Visceral adiposity is an independent predictor of incident hypertension in Japanese Americans. Ann Intern Med. 2004;140:992–1000. doi: 10.7326/0003-4819-140-12-200406150-00008. [DOI] [PubMed] [Google Scholar]
  • 15.Kotchen TA. Obesity-related hypertension: epidemiology, pathophysiology, and clinical management. Am J Hypertens. 2010;23:1170–1178. doi: 10.1038/ajh.2010.172. [DOI] [PubMed] [Google Scholar]
  • 16.Tybor DJ, Lichtenstein AH, Dallal GE, Daniels SR, Must A. Independent effects of age-related changes in waist circumference and BMI z scores in predicting cardiovascular disease risk factors in a prospective cohort of adolescent females. Am J Clin Nutr. 2011;93:392–401. doi: 10.3945/ajcn.110.001719. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Carolina Population Center, University of North Carolina. 2011 < http://www.cpc.unc. edu/projects/addhealth/design>.
  • 18.Nguyen QC, Tabor JW, Entzel PP, Lau Y, Suchindran C, Hussey JM, Halpern CT, Harris KM, Whitsel AE. Discordance in national estimates of hypertension among young adults. Epidemiology. 2011:22. doi: 10.1097/EDE.0b013e31821c79d2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Stone CJ, Koo CY, editors. Additive Splines in Statistics. Washington, DC: Proc Stat Comp Sect, Am Statist Assoc; 1986. [Google Scholar]
  • 20.Harrell FE. Regression Modeling Strategies: With Applications to Linear Models, Logistic Regression, and Survival Analysis. New York: Springer; 2001. [Google Scholar]
  • 21.Janssen I, Katzmarzyk PT, Ross R. Waist circumference and not body mass index explains obesity-related health risk. Am J Clin Nutr. 2004;79:379–384. doi: 10.1093/ajcn/79.3.379. [DOI] [PubMed] [Google Scholar]
  • 22.Schaie KW. A general model for the study of developmental problems. Psychol Bull. 1965;64:92–107. doi: 10.1037/h0022371. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental
NIHMS458117-supplement-Supplemental.doc (35KB, doc)

RESOURCES