Resting Blood Pressure and Bone Mineral Content Are Inversely Related in Overweight and Obese Hispanic Women^*

Abstract

Background:

Previous studies have shown that hypertension is related to abnormalities of calcium metabolism such as increased calcium losses from kidney and secondary activation of parathyroid glands. In animal studies, high blood pressure (BP) has been shown to increase the risk of bone mineral loss; however whether hypertension is associated with reduced bone mineral content (BMC) in human beings is inconclusive. The relationship between BP and BMC has not been previously studied in Hispanic individuals.

Methods:

Total body BMC of 33 overweight and obese (mean BMI= 31.1 kg/m²) premenopausal Hispanic women 22 to 51 years of age from Los Angeles, CA, was measured using dual-energy x-ray absorptiometry. Seated systolic BP (SBP) and diastolic BP (DBP) were measured using a standard mercury sphygmomanometer.

Results:

Partial correlations revealed an inverse relationship among BMC and SBP (r = −0.61, P < .001), DBP (r = −0.52, P < .01), and hypertension (r = −0.69, P < .0001). In multiple linear regressions, SBP was negatively related (β = −0.31, P = .001) to BMC and explained 10% of the variance. The DBP did not make a significant contribution to the variance. When fat mass and fat-free mass were controlled for, hypertensive women (n = 9) had significantly lower BMC (2119 g v 2441 g; P < .0001) than normotensive women (n = 23).

Conclusions:

These results reveal that BMC is partially and inversely correlated with resting SBP and DBP in premenopausal Hispanic women; in addition hypertensive women have lower adjusted means of BMC than normotensive women. Sustained hypercalciuria and ensuing hyperparathyroidism as consequences of high BP may be the mechanisms that explain the pathophysiology of increased bone mineral loss in hypertension.

Hispanic, obesity, hypertension, osteoporosis, bone loss

Metabolic studies^1,2 in hypertensive rats show that hypercalciuria and ensuing hyperparathyroidism lead to reduced growth and detectable deficits in bone mineral content (BMC) later in life. In human beings, the observation that increased urinary calcium excretion may be associated with low bone mass was first documented in 1976 by Alhava et al³ in a study of adults with kidney stones. In 1980 McCarron et al⁴ published the first report of hypercalciuria in patients with essential hypertension. However the link between hypertension and bone loss in human beings was not reported until recently,^5–12 and the existing evidence is inconclusive.

Tsuda et al⁵ compared the bone mineral density (BMD) of 31 hypertensive Japanese women to that in 14 normotensive women and showed inverse relationships between lumbar spine BMD and systolic blood pressure (SBP); they concluded that increased urinary calcium may lead to reduced BMD in female hypertension. Similarly Wu et al,⁶ in a group of Taiwanese women, and Cappuccio et al,⁷ in a large cohort of British women, found BMD to be inversely related to SBP and diastolic blood pressure (DBP). In men, studies have found inverse relationships between trabecular (ultra-distal radius) BMC and DBP,⁸ between cortical (femoral neck) BMD and DBP,⁹ and between both trabecular (lumbar spine) and cortical bone (hip) and SBP and DBP.¹⁰ In contrast, in a study of bone mass and bone modeling markers in hypertensive postmenopausal women in Spain, Perez-Castrillon et al¹¹ found no relationship between SBP or DBP and lumbar spine bone mass. Similarly, using cross-sectional data from the First National Health and Nutrition Examination Survey (NHANES I), Mussolino et al¹² found no significant associations between BMD and hypertension in men or women of African American or white ethnicity.

An inverse association between stroke incidence and BMD¹³ as well as between cardiovascular mortality and bone mass¹⁴ has been reported. However it is evident that there is a lack of adequate studies relating hypertension with deficits in bone mass or osteoporosis, a clinically silent disease associated with pain, deformity, loss of independence, and mortality.¹⁵ Therefore in this article our aim is to relate BMC and BMD with SBP and DBP to understand the pathophysiology of hypertensive bone loss in a group of ethnic minority women. Because estrogen has antiatherosclerotic and antihypertensive properties,^16,17 we believe that it is also important to address the association between BP and bone mass in additional studies in premenopausal women, in whom estrogen deficiency has not yet become a confounder. To the best of our knowledge, the relationship between bone mass and BP has not been previously studied in premenopausal Hispanic women.

In this article we address the hypothesis that resting SBP and DBP will be independently and inversely related with total body BMC and BMD in a group of overweight and obese premenopausal Hispanic women living in the United States.

Methods

Subject Description

A total of 39 Hispanic women were recruited through the elementary school of a low-income Hispanic community in Los Angeles County, to participate in a pilot study assessing the feasibility of a family-based, health risk reduction program. Subjects underwent a screening procedure consisting of a telephone interview, health history questionnaire, and physical examination by a board-certified physician. Volunteers were invited to participate if they had no chronic systemic illness or physical disability.

Study protocols were approved by the University of Southern California Institutional Review Board. Participation required written informed consent. Consent forms and questionnaires were written or compiled in English and then translated into Spanish and back-translated into English by professional translators to ensure that the intended messages were conveyed. Participants were given the option of receiving materials in either Spanish or English. Of the participants, 89% chose the Spanish-language option.

One woman was excluded from the study because she indicated that she might be pregnant. Of the remaining 38 who were tested, five were postmenopausal. Because BP increases in a nonlinear fashion in women, with dramatic accelerations after menopause,¹⁸ these five women were excluded from the data analyses reported here. The current report is based on data collected from 33 premenopausal women between the ages of 22 and 51 (mean 36.5 years).

Evaluation of BMC and BMD

Whole-body dual-energy x-ray absorptiometry (Hologic QDR-1500, software, version 7.10, Hologic Inc., Waltham, Massachusetts) was performed to provide whole- body BMC (in grams) and BMD (in grams per square centimeter). The whole-body scan requires the subject to be placed supine with the arms and legs positioned according to the manufacturer’s specifications; the scans took 15 min. Quality control was performed daily using a phantom, and measurements were maintained within the manufacturer’s precision standards of ≤1.5%. Reproducibility of BMD values, assessed in 10 healthy volunteers, ranged from 0.8% to 2.0%.

Resting BP

The BP measurements were obtained in the morning from each seated subject using a standard mercury sphygmomanometer. The same technician made all measurements. The SBP was measured at the first appearance of a pulse sound (Korotkoff phase 1) and DBP at the disappearance of the pulse sound (Korotkoff phase 5). Both SBP and DBP were recorded to the nearest even digit. The average of three measurements was used for this analysis.

Hypertension was defined by the criteria of the Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure.¹⁹ The cut-off points for hypertension were SBP ≥140 mm Hg or DBP ≥90 mm Hg. No subjects were taking antihypertensive medication.

Anthropometric Measurements

Weight was measured in kilograms using a Healthometer calibrated scale (Continental Scale Corporation, Bridgeview, IL). Subjects were weighed in light clothing without shoes, and weight was recorded to the nearest 0.1 kg. A stadiometer was used to measure height. Subjects were measured barefoot or wearing thin socks. The measurement was recorded to the nearest 0.1 cm. Body mass index (BMI) was calculated as the ratio of body weight to height squared (kg/m²). Overweight was defined as BMI values between 25 and 29.9 kg/m². Obesity was defined²⁰ as BMI values ≥30 kg/m².

Waist circumference was measured at the smallest circumference of the torso, which is at the level of the natural waist.²¹ Hip circumference was measured at the level of maximal posterior extension of the buttocks.²¹ Subjects wore no clothing except underwear to ensure correct positioning of tape. Values were recorded to the nearest 0.1 cm. The waist-to-hip circumference ratio was calculated by dividing waist circumference by hip circumference. This ratio is an indicator of the pattern of subcutaneous distribution of adipose tissue.

Aerobic Capacity

Peak VO₂ was determined using a continuous, incremental protocol on a motorized treadmill. The initial speed and grade were 2.5 mph and 0% respectively, with increases of 0.5 mph and 2% every 2 min of exercise. The volume of expired air, volume of oxygen consumption, and volume of carbon dioxide production were determined by SensorMedics metabolic system (SensorMedics Corporation, Yorba Linda, CA). Subjects exercised to volitional fatigue, with 12-lead electrocardiographic monitoring heart rate taken at the end of each 1-min period during exercise and at peak VO₂ for determination of HR_max. Peak VO₂ was considered to be achieved if the test met two of the following criteria: 1) respiratory exchange ratio (RER) value was >1.05; 2) heart rate = ±10 beats/min of the age-predicted HR_max; or 3) a plateau in VO₂ with increasing workloads. Criteria for terminating the test before completion included indications of distress, arrhythmia, or S-T abnormalities. Because the third criterion (plateau or leveling-off in oxygen uptake) was not met in every subject, a “true” VO_2max was not achieved and the tests were therefore called “peak VO₂.”

Data Analysis

All analyses were performed using SPSS version 13.0 (SPSS Inc., Chicago, IL), with a type I error set at P < .05. As determined by Kolmogorov-Smirnov test of normality, both BMC and BMD were normally distributed and no transformations were necessary. Descriptive statistics, partial correlations, general linear models, and stepwise multiple linear regression models were used to estimate the effect of variables on the dependent variables, BMC and BMD. A power analysis software program (Statistical Solutions, nQuery Advisor, version 3, Saugus, MA) was used for the power analysis results presented in the discussion.

Results

Means and standard deviations of participant characteristics are shown in Table 1. The youngest subject was 22 years of age, and the oldest was 51 years. Mean age was 36.5 years. Mean BMI was 31.1 kg/m²with a median of 30.5 kg/m². Only five (15%) women had a BMI <25 kg/m²; 10 (30%) women were overweight (BMI ≥25 to 30 kg/m²), and another 18 (55%) were obese (BMI>30 kg/m²). Mean fat-free mass was 44.3 kg with a range of 30.9 kg to 59.0 kg. Mean fat mass was 28.5 kg with a range of 13.8 kg to 55.7 kg. Mean percent fat was 36.9%. Mean SBP was 121 mm Hg with a range of 99 to 163 mm Hg. Mean DBP was 77 mm Hg with a range of 55 to 102 mm Hg. Nine (27%) women were hypertensive and 24 (73%) were normotensive. Mean total body BMC was 2348.8 g. Mean total body BMD was 1.121 g/cm².

Table 1

Characteristics of the study population

Characteristic	Mean ± SD
N	33
Age (y)	36.5 ± 7.2
Weight (kg)	75.8 ± 14.7
Height (cm)	156.0 ± 6.9
BMI (kg/m2)	31.1 ± 5.4
Waist circumference (cm)	91.4 ± 12.0
Hip circumference (cm)	109.1 ± 13.8
Waist–hip ratio	0.84 ± 0.1
Fat-free mass (kg)	44.3 ± 6.7
Fat mass (kg)	28.5 ± 10.8
Percent fat (%)	36.9 ± 7.4
Peak VO2 (mL · kg−1 · min−1)	26.5 ± 7.5
Mean SBP (mm Hg)	120.7 ± 17.7
Mean DBP (mm Hg)	77.2 ± 12.2
Total body BMC (g)	2348.8 ± 385.9
Total body BMD (g/cm2)	1.121 ± 0.1

Data are mean ± SD except N, which is absolute number.

BMC = bone mineral content; BMD = bone mineral density; BMI = body mass index; DBP = diastolic blood pressure; SBP = systolic blood pressure.

Table 2 shows partial correlations (adjusting for age, weight, height, and peak VO₂) between the dependent variables (BMC and BMD) and mean SBP, mean DBP, and hypertension. Mean SBP and DBP were significantly and inversely correlated with BMC (r = −0.61, P < .001 and r = −0.52, P < .01, respectively). Hypertension was significantly and inversely correlated with BMC (r = −0.69, P < .0001). There were no significant associations between SBP, DBP, or hypertension and BMD. Table 2 shows correlations between BMC per kilogram of body weight (BMC/kg) and SBP, DBP, and hypertension (adjusting for age, height, and peak VO₂). These associations were similar to but weaker than those observed between BMC and SBP, DBP, and hypertension.

Table 2

Partial correlations with bone mineral content (BMC), BMC per kilogram of body weight (BMC/kg), and bone mineral density (BMD) after controlling for age, weight, height, and peak VO₂

Variable	BMC	BMC/kg*	BMD
Mean SBP	−0.61†	−0.46‡	−0.11
Mean DBP	−0.52§	−0.50§	−0.17
Hypertension	−0.69∥	−0.58§	−0.08

DBP = diastolic blood pressure; SBP = systolic blood pressure.

*

Controlling for age, height, and peak VO₂ only;

†

P < .001;

‡

P < .05;

§

P < .01;

∥

P < .0001.

Stepwise multiple linear regression analyses were used to examine the independent association of age, weight, height, peak VO₂, SBP, and DBP with total body BMC and BMD (Table 3). The SBP explained 10% of the variance in total body BMC only. Another 74% of the variance in BMC was explained by body weight. The DBP did not make a significant contribution to the variances in BMC or BMD. Body weight explained 59% of the variance in BMD.

Table 3

Analysis with multiple linear regression model for bone mineral content (BMC) and bone mineral density (BMD)

	BMC		BMD
Characteristic	β ± SE	r2	β ± SE	r2
Intercept	1331.35 ± 237.59		0.83 ± 0.05
Age (y)	NS		NS
Weight (kg)	25.35 ± 2.10	0.74*	0.004 ± 0.001	0.59*
Height (cm)	NS		NS
Peak VO2 (mL · kg−1 · min−1)	NS		NS
Mean SBP (mm Hg)	−7.36 ± 1.82	0.10*	NS
Mean DBP (mm Hg)	NS		NS
r2	0.84		0.59

β = multiple regression nonstandardized coefficient; NS = nonsignificant; SE = standard error; other abbreviations as in Table 1.

*

Significant at P < .0001.

Hypertensive women (n = 9) were compared with normotensive women (n = 23) using general linear models, and the results are presented in Table 4. When fat mass and fat-free mass were controlled for, hypertensive women had significantly lower BMC (2119 g v 2441 g; P < .0001) than normotensive women. The BMD was not significantly different between the two groups but was lower in hypertensive compared with normotensive women (1.108 g/cm²v 1.129 g/cm²).

Table 4

Bone mineral content (BMC) and bone mineral density (BMD) means of hypertensive versus normotensive women after controlling for fat mass and fat-free mass

Subjects	BMC (g)	BMD (g/cm2)
Hypertensive women (n = 9)	2119*	1.108
Normotensive women (n = 23)	2441	1.129

*

Significantly lower than in normotensive women (P < .0001).

Figure 1 shows a scatter plot of SBP versus BMC/kg, and Fig. 2 gives a scatter plot of DBP versus BMC/kg.

Scatter plot of systolic blood pressure (SBP) versus bone mineral content (BMC) per kilogram of body weight.

Figure 1.

Scatter plot of diastolic blood pressure (DBP) versus bone mineral content (BMC) per kilogram of body weight.

Figure 2.

Discussion

In this study we found SBP, DBP, and hypertension to be inversely correlated with BMC; this association was independent of age, weight, height, and peak VO₂ (Table 2). Results from multiple linear regression analysis revealed that SBP is significantly, independently, and inversely related to BMC and that it explains 10% of the variance (Table 3). Using general linear models, when BMC was compared in hypertensive women (n = 9) versus the normotensive women (n = 24) while controlling for fat mass and fat-free mass, hypertensive women had significantly lower BMC compared with normotensive women (2119 g v 2441 g) (Table 4). Although our analysis was restricted to a small population of Hispanic women and the findings may not be generalizeable to other ethnic groups or to men, they suggest that BP has an independent relationship to bone mass.

Previous studies by our group²² examining bone density in these women revealed that fat mass, fat-free mass, and aerobic capacity were the significant independent predictors of BMD, explaining 55%, 10%, and 8% of the variance, respectively. In this study, it is noteworthy that when fat mass and fat-free mass replaced total body weight in the partial correlations and the linear regression models, the results with regard to the inverse relationship between SBP and BMC did not materially change (data not shown). As far as what was observed for BMC versus BMD, most associations between SBP, DBP, hypertension, and BMC were quite strong, whereas the relationships between these variables and BMD were nonsignificant (Table 2). Similarly, 84% of the variance was explained for total body BMC compared with only 59% of the variance for total body BMD (Table 3). Witzke and Snow²³ have found a stronger BMC model compared with a BMD model for anthropometric measures, leg power, and leg strength in adolescent girls. Similarly, in 400 postmenopausal African American women, we²⁴ have also found a stronger BMC model compared with a BMD model for age, resting energy expenditure, and grip strength. This is plausible because BMC is influenced by the growth of the skeleton with a trajectory that appears to be established early in life, and skeletal content appears to track into adulthood more strongly than skeletal density.²⁵

Our results confirm previous findings^5–10 on the inverse relationship between BP, hypertension and bone mass. Sustained hypercalciuria as a result of high BP may be the underlying mechanism that explains the pathophysiology of hypertensive bone loss. The association between hypertension and hypercalciuria has been confirmed in several studies and has been reported consistently in case-control and cross-sectional investigations.^26–28 The cause of hypercalciuria in hypertension is unknown.¹¹ However, Cappuccio et al²⁸ suggested the “renal calcium leak hypothesis” and the “central blood volume hypothesis.” The renal calcium leak hypothesis is explained by alterations in renal calcium handling because of tubular disorder; and the central blood volume hypothesis states that hypercalciuria is caused by central volume expansion observed in hypertensive individuals.²⁸ Although studying the above mechanisms was not the objective of the current investigation, our findings seem to be in agreement with them. In this study we addressed the hypothesis that resting BP would be independently and inversely related to bone mass in a group of ethnic minority women, in the hope of being able to highlight the need for encouraging minority populations to benefit from healthy BP.

It is noteworthy that we faced several barriers with regard to recruitment of this population, which were similar to the challenges experienced by others who have included minority populations in their research. In a study of breast cancer survivors, Naranjo and Dirksen²⁹ encountered a high refusal rate and found that culture contributed to the challenge of recruiting Hispanic women. They found that the Hispanic value of “familialism,” which involves an individual’s strong identification with and attachment to her nuclear and extended families, was a prominent problem in recruitment because of family commitments that involved travel to spend time with or take care of family members. Gender role or “machismo” was also a factor because some husbands forbade their wives to participate in research studies. In this group of Hispanic women of low socioeconomic status, fear of loss of health benefits (ie, Medicaid), inability to afford child care or transportation, the need to work overtime or to hold additional jobs on weekends when data were being collected, as well as immigration status and fear of deportation could have been other possible factors that played a role in recruitment and retention. We believe that successful recruitment of Hispanic women for our future research studies requires a renewed and ongoing appreciation of cultural values and beliefs, building trust within the Hispanic community, and involving the community in the design, planning, and implementation of research studies.

Despite these challenges, with a sample size of 33 women and observed r values in the range of 0.52 to 0.69, which were much stronger than previous findings,^5,9 we had >99% power. Because of the challenges that we faced with this population, our time with each subject was also limited and we were unable to perform separate spine, hip, or forearm scans. Trabecular bone (lumbar spine, os calcis) has been shown to be more sensitive to metabolic changes³⁰ compared with cortical bone (femoral neck, distal radius). The association between hypertension and osteopenia has been shown to be site specific.^8–10 Furthermore there are differences in the timing of bone loss in healthy women, with trabecular bone diminishing with every decade of life but cortical bone levels being similar in the third, fourth, and fifth decades.³¹ Therefore, although it would seem more logical that trabecular bone would be more sensitive to metabolic changes in these Hispanic women, measurement of both cortical and trabecular bone is necessary in our future work. Nonetheless our current results from whole-body dual-energy x-ray absorptiometry provide valuable physiologic insight into an ethnic group not frequently studied. Results of the present study are the first to demonstrate in overweight and obese Hispanic women of predominantly Mexican descent that systolic BP and bone mass are inversely related and that, independent of fat mass and fat-free mass, hypertensive women have lower bone mineral content than normotensive women.

In conclusion, over the past decade, as our attention has focused on weight reduction and obesity prevention, the importance of maintaining a healthy BP seems to have become secondary. A healthy BP for the prevention of not only cardiovascular diseases but also of osteoporosis can be encouraged in Hispanic women. Our results need to be replicated in this population, and prospective studies are needed to determine whether hypertension is an independent risk factor for osteoporosis.

Author notes

*

This study was supported by the University of Southern California’s Norris Comprehensive Cancer Center.

© American Journal of Hypertension, Ltd. 2006