This work was presented in abstract form at the Seventeenth Annual Meeting of the American Society for Bone and Mineral Research, Baltimore, Maryland, U.S.A.
While noninvasive studies of bone mass and turnover in blacks and whites abound, histologic evaluations are very rare. We have performed a comparative bone histomorphometric study of iliac biopsies from 55 healthy, premenopausal women including 21 blacks (mean age 33.4 + 1.2 years) and 34 whites (mean age 32.5 + 0.8 years) of comparable age, weight, body composition, education, and lifestyle. Biochemical indices of mineral metabolism: parathyroid hormone, 25-hydroxyvitamin D, 1,25-dihydroxyvitamin D, serum ionized calcium, serum phosphorus, and urinary calcium/creatinine were measured in the fasting state. Blacks had lower 25-hydroxyvitamin D (31.5 ± 3.36 vs. 63.21 ± 3.79 nmol/l, p = 0.0001). Histomorphometric indices of bone volume, structure, and connectivity were not different between groups. The following indices of bone remodeling were also similar in both groups: eroded perimeter, osteoid width, mineralizing perimeter, tissue-based bone formation rate, osteoid maturation time, active formation period, and activation frequency. However, osteoid perimeter (black [B] = 15.85 ± 1.30 vs. white [W] = 9.49 ± 0.70%, p = 0.0002), osteoid area (B = 2.55 ± 0.32 vs. W = 1.39 ± 0.12%, p = 0.003), single-labeled perimeter (B = 5.46 ± 0.54 vs. W = 4.04 ± 0.33%, p = 0.03), mineralization lag time (B = 38.18 ± 4.04 vs. W = 21.83 ± 1.60 days, p < 0.009), and total formation period (B = 148.15 ± 19.70 vs. W = 84.04 ± 7.62 days, p = 0.0056) were higher in blacks than in whites. The quiescent perimeter (B = 76.91 ± 1.40 vs. W = 84.25 ± 0.91%, p = 0.0001), mineral apposition rate (B = 0.70 ± 0.02 vs. W = 0.75 ± 0.02 μm/day, p = 0.066), mineralizing osteoid perimeter (B = 0.49 ± 0.04 vs. W = 0.75 ± 0.04%, p = 0.0001) and adjusted apposition rate (B = 0.35 ± 0.04 vs. W = 0.58 ± 0.04 μm3/μm2/day, p = 0.0001) were all lower in blacks than in whites. These results indicate that there are no differences in bone volume, microstructure, or turnover between black and white premenopausal women. However, there are significant differences in the mechanism of bone formation between the two groups, with a lower rate of mineralized matrix apposition within each remodeling unit and a longer total formation period in blacks than in whites. The differences appear to be the result of more frequent and/or longer inactive periods in the life span of the bone formation units in blacks. These differences may allow a greater overall deposition of bone mineral in black women and therefore help explain a higher bone mass and perhaps better bone quality in black than white women.
A lower incidence of osteoporotic fractures in blacks than in whites has been reported.(1–5) This may be due in part to greater size and weight of the skeleton in blacks(6–9) and higher bone mineral density (BMD) in black adults and children.(10–18)
Histomorphometric investigations on this racial discrepancy are extremely rare. Two such studies have been performed to date and have yielded conflicting results.(19,20) Schnitzler et al.(19) studying South Africans found similar values of cancellous bone area in black and white women, but greater bone area in black than white men. In both genders, indices of bone structure were similar in the two races, except for trabecular width which was greater in blacks than whites. Osteoid and eroded perimeters were higher in blacks than whites, leading the authors to conclude that blacks had increased bone turnover. Based on this conclusion, the authors hypothesized that the more frequent cycles of renewal would ensure better bone quality and, hence, favorably influence fracture rate in blacks. While this report studied a large number of blacks and whites, it did not include a dynamic evaluation of bone turnover. Moreover, most of the bone specimens were derived from autopsied subjects on whom little background information was available. Finally, blacks and whites were not comparable with regard to socioeconomic background and lifestyle.
In contrast to this study, the report by Weinstein and Bell(20) included living American subjects prelabeled with tetracycline. Both groups were found to have similar bone volume and structure. However, blacks had lower bone turnover based on static and dynamic indices of bone remodeling. On the assumption that coupling between resorption and formation is maintained, the authors suggested that a lower rate of bone resorption would help maintain a higher peak bone mass in blacks. However, the number of subjects in each group was small, and data from men and women were combined in this study.
Because of the paucity and disparate nature of the available data, we have conducted a complete histomorphometric evaluation of iliac crest bone biopsies from healthy premenopausal black and white American women. This represents the first interracial biopsy study of blacks and whites based on two carefully selected, comparable groups, prelabeled with tetracycline, and the first comparative evaluation of trabecular connectivity in blacks and whites.
MATERIALS AND METHODS
This study was approved by the Institutional Review Boards of Helen Hayes Hospital, Columbia-Presbyterian, Mount Sinai, and St. Lukes-Roosevelt Medical Centers. All subjects gave informed consent.
Women were recruited from the New York City and Rockland County New York areas. Eligible volunteers were reached by advertisement through newspaper and radio announcements, television interviews, poster distribution on college and university campuses, and contact with women's organizations. The decision to assign recruits to one or the other ethnic group was based on self-identification and on the racial identity of three of the subject's four grandparents. This was determined by each woman's racial identification of her grandparents through a questionnaire. Initial screening procedures were performed by telephone to assure premenopausal status and exclude subjects with certain diseases. Formal screening procedures (history and physical examination) were then performed on eligible women. Pregnant women and subjects suffering from malignancy, eating disorders, endocrine, renal, hepatic, or other chronic diseases such as rheumatoid arthritis, were excluded. Subjects with a history of gastrointestinal surgery, alcoholism, or other substance abuse, use of medication such as oral contraceptives within 6 months of the study, steroid or anticonvulsant therapy, or other factors known to affect bone metabolism were similarly excluded. Women who smoked more than one and one half packs of cigarettes per day were excluded, as were women who consumed on average more than two alcoholic beverages per day. To minimize the effects of obesity on bone mass and metabolism, only women within 20% of ideal weight for height and average frame size (using a modification of the 1983 Metropolitan Life Insurance Tables guidelines) were recruited.(21) Screening laboratory evaluation included blood counts, routine biochemistries, urinalysis, thyroid function tests, estradiol, and follicle stimulating hormone (FSH) determinations.
Of a total of 239 subjects who responded to advertisements, 163 subjects were eliminated by telephone interview. The other seventy-six agreed to participate in the study and were eligible to undergo formal screening procedures. Seventy-one of the women screened were found suitable and were enrolled in a larger interracial study of bone metabolism. Transiliac bone biopsy was performed on 60 of the 71 consenting subjects. This group was comprised of 37 whites and 23 blacks.
The blood samples were obtained on the day of the biopsy in approximately half of the subjects recruited, and in the remainder, the samples were obtained at the beginning of the tetracycline labeling period, approximately 3 weeks prior to biopsy. In each subject, two serum samples were obtained at 30-minute intervals, and values for various indices were averaged. Fasting urine samples were also obtained. Serum was analyzed for parathyroid hormone (PTH) (1–84) by immunoradiometric assay (Allegro Intact PTH, Nichols Institute, San Juan Capistrano, CA, U.S.A.). 25-hydroxyvitamin D (25(OH)D) and 1,25-dihydroxyvitamin D (1,25(OH)2D) were analyzed by competitive protein binding and radioreceptor assays, as previously described.(22,23) Serum ionized calcium was measured by NOVA 8 Ionized Calcium Analyzer (Nova Biomedical Inc., Newton, MA, U.S.A.) and serum phosphorus was measured using an Automated Discrete Chemistry Analyzer (Cobas, MIra-S, Roche Diagnostic System, Montclair, NJ, U.S.A.). Urine calcium was analyzed by atomic absorption spectrophotometry. Urine creatinine was determined with the Beckman II creatinine analyzer, utilizing the Jaffe Rate technique.
Beginning approximately 1 month before the biopsy, women were prelabeled with tetracycline administered orally in two time-spaced cyclical doses of tetracycline hydrochloride (Sumycin) (250 mg four times daily) and demethylchlortetracycline (Declomycin) (150 mg four times daily) following a 3 days on, 14 days off, 3 days on, 5–7 days free schedule. Transiliac bone biopsy was performed according to standard technique,(24) and no significant postoperative complications were observed in any of the subjects.
Of the 60 biopsied subjects, one woman was later eliminated for nondisclosure of a past history of an eating disorder and drug use. Of the remaining 59 biopsy cores, four were technically unsuitable for analysis (two were severely fragmented and two were too narrow for quantification). A detailed histomorphometric analysis was performed on the remaining biopsies from 55 subjects comprising 21 blacks and 34 whites.
Methods of tissue processing, sectioning, and staining followed established procedures.(25,26) To ensure analysis under similar laboratory conditions and avoid observer bias, biopsies from both groups were, within a given time frame, evenly allocated for analysis, and the operator was blinded to the subject's racial identity.
Conventional indices of bone structure were evaluated using Goldner stained, 7-μm-thick sections. Prior to the measurements, the cancellous space was carefully demarcated by well-established criteria.(27) The cancellous bone area as a percentage of total cancellous tissue area, trabecular number, trabecular width, and trabecular separation were derived from automated measurement of bone area and perimeter (31× magnification) and defined as previously reported(28) (Optomax V AMS system, Optomax Inc., Hollis, NH, U.S.A.). Cortical width was measured semiautomatically, at 40× magnification, as previously described (Optomax VIDS IV).(29)
Trabecular microstructure was evaluated, by the strut analysis method(30) at 12.5× magnification, using an automated image capturing and processing system. Using Solochrome Cyanin R–stained sections, eight bit, gray scale images of the detailed trabecular structure were obtained with image-capturing software (Optimas, Bioscan Inc. Edmunds, WA, U.S.A.) and analyzed with a trabecular analysis system (TAS) program.(31) In a defined area of the cancellous space, approximately 1 mm away from the corticomedullary junction, the trabecular image was skeletonized, and the resulting strut structure analyzed. Two-dimensional indices of connectivity were defined as previously reported.(32,33) Node number and terminus number were expressed per square millimeter of cancellous tissue area, and the ratio of node number to terminus number calculated. The node to node strut length was expressed in millimeters per square millimeter of cancellous tissue area and terminus to terminus strut length as a percentage of the total strut length.(32,33)
Table Table 1. Characteristics of the Study Population
All perimeter indices, except dynamic indices of bone turnover, were measured on 7-μm-thick sections with the Solochrome Cyanin R stain used for osteoid parameters and the Goldner's trichrome stain for eroded perimeter. Twenty-micron unstained sections were used to measure all dynamic indices of bone remodeling. Perimeter indices were expressed as percentages of total trabecular perimeter. Osteoid perimeter, eroded perimeter, mineralizing perimeter, i.e., the extent of all double labels plus half the extent of single labels were measured by the point counting method at 125× magnification, and the proportion of mineralizing osteoid perimeter was calculated. The mineral apposition rate (MAR) was calculated from the semiautomated measurement of interlabel distance (Optomax VIDS V) and the interval in days between the two labels. The bone formation rate and adjusted apposition rate were calculated according to the formulas listed in the Appendix, and expressed in cubic micrometers per square micrometer per day.
The wall width of completed trabecular bone packets was measured semiautomatically (VIDS IV, Optomax), at 160× magnification, following the method of Kragstrup et al.(34) Osteoid width was measured semiautomatically (VIDS IV, Optomax) at four equidistant points on each seam, at 400× magnification. The osteoid area was measured by the point counting method at 125× magnification and expressed as a percentage of total trabecular bone area. Quiescent perimeter, mineralization lag time, and osteoid maturation time were calculated according to the formulas listed in the Appendix.
The active formation period of the bone structural unit (FP[A]) assumes that bone formation is continuous and therefore does not take into account the resting periods in the lifespan of osteoblasts (off times).(35) The total formation period of the bone structural unit (FP[T]) represents the total time spent by the osteoblasts in completing the bone structural unit (BSU) including off times. FP[A] and FP[T] were calculated using the standard formulas listed in the Appendix.(36) The resorption period, quiescent period, remodeling period, and activation frequency were calculated according to the formulas listed in the Appendix. Histomorphometric indices were designated in accordance with the nomenclature recommended by the American Society for Bone and Mineral Research.(36)
Table Table 2. Basal Biochemical Indices of Mineral Metabolism (n = 55)
Table Table 3. Histomorphometric Indices of Bone Structure in Black and White Premenopausal Women (n = 55)
All indices are reported as mean ± SEM. The significance of differences between the two groups was evaluated by unpaired, two-tailed Student's t-tests for normally distributed values and by the Wilcoxon rank sum test for abnormally distributed values, using SAS software (SAS Inc., Cary, NC, U.S.A.). The difference between percentages of smokers in each group was evaluated by a Chi-square test.
The general characteristics of the study population are shown in Table 1. There were no significant differences in age, weight, body mass index, parity, level of education, or tobacco use between black and white women. Calcium intake was similar in both groups and was much below the recommended daily allowance.
Mean basal values for the two groups are presented in Table 2. Serum levels of ionized calcium, phosphorus, and 1,25(OH)2D were similar in the two groups. The mean 25(OH)D level was significantly lower in the blacks than in the whites. Although PTH(1–84) was 13% higher and urinary calcium/creatinine was 50% lower in the blacks, differences were not significant.
Mean values of conventional indices of bone structure and indices of trabecular connectivity were similar in both groups (Table 3). This is illustrated qualitatively in Figure 1, which compares photomicrographs of iliac bone biopsies from two representative pairs of black and white women.
The eroded perimeter, activation frequency (Fig. 2A), mineralizing perimeter, bone formation rate (BFR) at the tissue level (Fig. 2B), as well as active formation, resorption, remodeling, and quiescent periods were similar in both groups (Table 4). In contrast, significant differences in some indices of bone formation were seen. Osteoid perimeter (Fig. 2C) and osteoid area were significantly higher in the blacks than in the whites, although the mean osteoid width was similar in both groups (Table 4). Furthermore, the single-labeled perimeter (Fig. 2D) was higher, while the quiescent perimeter, mineralizing osteoid perimeter (Fig. 2E), and adjusted apposition rate (Fig. 2G) were lower in the blacks than in the whites (Table 4). Osteoid maturation time was similar in both groups, but mineralization lag time was higher in blacks than whites (Table 4). Mineral apposition rate (Fig. 2F) was lower in blacks than in whites, the difference almost reaching statistical significance. Finally, the total formation period was higher in blacks than in whites (Fig. 2H).
Our data clearly show that in two comparable groups of healthy premenopausal black and white women of similar age, weight, body composition, lifestyle, and education level, bone volume and structure in the iliac crest bone biopsy are very similar. However, we found significant racial differences in the indices reflecting the cellular mechanism of bone formation.
Very few histological data comparing blacks and whites exist. Of two reports published to date, one compared black and white subjects from the U.S.A.,(20) while the other studied a South African population.(19) Although the first study(20) evaluated living individuals, the sample size, which included 12 blacks (6 men and 6 women) and 13 whites (7 men and 5 women), was small, necessitating the combined evaluation of subjects of both genders. While the South African study(19) involved a large number of subjects (346), most of the bone samples were obtained from cadavers. Although the circumstances of death were known, it is likely, as in all autopsy studies, that there was very little information on these subjects. In such cases of acute or violent death, a history of hidden factors such as alcohol or drug use which could affect bone metabolism is usually not available. Moreover, marked discrepancies in socioeconomic and lifestyle factors, which were acknowledged to exist between the two groups, may have biased the results. By contrast, the present study compared two groups of healthy premenopausal women who were carefully screened for all factors known to influence skeletal metabolism, and who were chosen from the same geographic region providing further similarities in lifestyle.
We found comparable values of cortical width in the two groups, a finding similar to that of the only other histomorphometric evaluation of cortical bone between racial groups.(20) Weinstein and Bell also reported no difference in cortical porosity between black and white men and women ranging in age from 19 to 46 years. Previous evaluations of cortical bone by noninvasive means have produced conflicting data. Using radiogrammetry, greater cortical width and cortical area and lesser medullary cavity expansion were reported in aging black than white women,(10) while lower cortical density and thinner cortices were found in Bantus than in white South Africans despite a lower fracture rate in the former.(37,38) Additionally, while BMD at the mid-distal radius was found to be higher in black than white women in both premenopausal(12) and in mixed pre- and postmenopausal groups by some,(39) no ethnic difference was found at these predominantly cortical sites by others, regardless of weight.(40) Moreover, bone mineral content measured at the distal radius was significantly higher in black than in white women in each decade, except in those ranging from 24 to 35 years, an age group similar to that of our study population.(13)
In agreement with our data, Schnitzler et al.(19) found a similar cancellous bone area in black and white women, although this index was higher in black than in white men. Also consistent with our findings, conventional indices of structure (trabecular number, width, and separation) in the South African study were similar in both races, except for trabecular width, which was higher in blacks than in whites in both genders. This finding is, however, difficult to reconcile with the fact that in women, cancellous bone area and trabecular number were similar in the two groups. Although Schnitzler et al.(19) studied women ranging from 21 to 83 years, our structural data in premenopausal women are comparable to those of their young black and white women aged 21–40 years, with the exception of the trabecular width, as noted above. Consistent with our data, Weinstein and Bell found similar indices of bone volume and structure in both races. However, it is difficult to compare their results with ours, because their population, although somewhat similar to ours in age, was a mixed sample of men and women.(20)
Our study is the first interracial assessment of trabecular connectivity. Using the best available histological techniques, the similarity of the data in these two comparable populations suggests that there are no differences in cancellous bone area nor in connectivity between premenopausal black and white women at the iliac crest. Noninvasive measurements of sites containing large amounts of cancellous bone, e.g., the spine have, in most studies, revealed higher BMD in blacks than in whites. However, discrepancies exist with some authors reporting higher BMD at the spine in black than in white premenopausal women whether obese or nonobese,(12) and others finding no ethnic differences at this site regardless of weight.(40) In mixed groups of pre- and postmenopausal women, 53% of the ethnic difference in spinal BMD was dependent on body weight.(39) Furthermore, there was substantial racial overlap in vertebral BMD, although values were generally higher in blacks than in whites.(13,41) Although there is a relationship between the cancellous bone area at the iliac crest and cancellous bone area at the spine,(42) the correlation between cancellous bone area at the iliac crest and BMD at the lumbar spine is modest.(43) Several conclusions can be drawn from our data. First, the reported racial difference in cancellous bone of the spine may not exist in the ilium. Second, the remarkable similarity between the two racial groups in indices of bone structure suggests that the reported racial differences in fracture incidence cannot be accounted for to a large extent by differences in the structure of cancellous bone at skeletal maturity.
Schnitzler et al. reported higher skeletal turnover in blacks, based on greater values of osteoid and eroded perimeters in black compared with white subjects. One weakness of this study, however, is that it did not include a dynamic evaluation. In particular, no data were available on activation frequency, which is necessary to validate any conclusion regarding bone turnover.(19) Weinstein and Bell found decreased bone turnover in blacks, based on lower values of mineralizing perimeter, labeled osteoid perimeter, and bone formation rate in black than in white subjects.(20) Unlike these authors, we found similar values in both groups for the major indices of bone turnover, most importantly activation frequency. Therefore, in this population of premenopausal women, we were not able to confirm Weinstein and Bell's finding of lower turnover in blacks which, together with their observation of higher PTH in blacks, led them to hypothesize that the black skeleton was resistant to the effects of PTH.(20,44) However, despite the fact that we did not see basal differences in activation frequency, we have obtained biochemical data strongly suggesting skeletal resistance to the resorptive action of PTH during acute elevations of the hormone,(45) in support of Bell's hypothesis.(44)
The most striking histological difference between the two groups was the higher osteoid perimeter and osteoid area in the blacks. This was consistent with the finding of Schnitzler et al., although our interpretation differs from these authors.(19) Since the activation frequency of remodeling units was the same in blacks and whites, it can be deduced that the greater extent of osteoid in the black subjects is a reflection of a longer life span of the osteoblasts(46) rather than of higher bone remodeling, as argued by Schnitzler. Although serum 25(OH)D levels were significantly lower in blacks than whites, there was no evidence of osteomalacia even when the strictest histomorphometric criteria were applied.(47–50) A more detailed analysis of the reason for the higher osteoid perimeter in blacks revealed that a substantially greater proportion of osteoid seams was unmineralized, and that the extent of single-labeled perimeter and the mineralization lag time were significantly higher in blacks. Moreover, the adjusted apposition rate, which is the best estimate of the effective mineralized matrix apposition rate by the bone remodeling unit,(47) was significantly lower in blacks. As a consequence of the lower adjusted apposition rate, the total bone formation period (FP[T]) was significantly longer in the blacks, indicating that bone formation is achieved at a slower pace in blacks than in whites. This may be the result of more inactive time (off time) in the osteoblast's life span in the blacks. As proposed by Frost, osteoblasts are not active throughout their entire lifespan, and there are off times or rest periods in the normal osteoblast's lifespan.(35) When the active formation period FP[A] is calculated in the blacks, 59.3 of a total of 148.2 days (40.0%) were spent actively forming bone. In contrast, of a total lifespan of 84.0 days in the whites, an active period of 55.4 days (65.9%) is found. This leaves 88.9 days in the blacks compared with 28.6 days in the whites spent in the inactive phase. With the osteoblasts having more off time, a greater proportion of remodeling units “escape” one of the two tetracycline labels, as suggested by the higher extent of single-labeled perimeter in the blacks than in the whites.(35)
Table Table 4. Histomorphometric Indices of Bone Remodeling in Black and White Premenopausal Women
As black subjects synthesize less vitamin D3, presumably due to greater skin pigmentation than whites,(51) better renal calcium handling, shown by lower urinary calcium in blacks,(44,45,52–54) and skeletal resistance to bone resorption by PTH(17,44,45) must ensure the maintenance of calcium homeostasis in these subjects. In addition, relatively higher PTH levels may promote optimal renal hydroxylation of 25(OH)D and, hence, maintain normal levels of 1,25(OH)2D and adequate intestinal calcium absorption. The relatively slower rate of bone formation, reflected in the reduced adjusted apposition rates may represent yet another mechanism of calcium conservation at the level of the mineralization front.
Thus, our data point to important differences in the cellular mechanism of bone formation between the two races. One can speculate that the longer bone formation period in blacks may result in an improvement in bone quality. The rationale for this, although indirect at present, derives from previous observations that bone that is formed rapidly is of poor quality. Examples of this include woven bone, pagetic bone, and the cancellous bone in the subset of fluoride-treated patients who gained bone mass very rapidly.(55–57) One can also argue that the more slowly the bone is formed, the more time there is for secondary mineralization and for higher ultimate mineral density, which is associated with reduced fragility.(58) Further investigation into the mechanisms underlying the reported differences in fracture rate between the two racial groups, such as mechanical tests of bone fragility, are clearly warranted.
The authors thank Diana Sherwood, Julie Horton, and Ollie Brown for invaluable assistance in recruitment and screening of research subjects, Bonnie Fuchs, Susan Gordon, David Healy, and Patricia Garrett for expert technical assistance, and Lester Ferguson, Stephanie Roberts, and Brian Yarborough for the preparation of illustrative material. This study was supported by National Institutes of Health grants AR41386 and AR 39191.
APPENDIX: FORMULAE FOR THE CALCULATION OF BONE REMODELING INDICES