A comparative study of the determinants of bone strength and the propensity to falls in black and white South African women
Thesis (DMed)--Stellenbosch University, 2008.
The comparative study presented in this dissertation specifically aimed to assess fracture risk in black (Xhosa) and white South African women by evaluating known determinants of bone strength as well as the propensity to falls. We thus compared the prevalence of clinical (historic) risk factors for osteoporosis, measured and compared vertebral and femoral bone mineral density (BMD) employing dual energy X-ray absorptiometry (DEXA), ultrasound variables using the Sahara sonometer, serum parathyroid hormone (PTH) and 25-OH Vitamin D, mineral homeostasis and modern biochemical markers of bone turnover, bone geometry and the propensity to falls. Finally, we determined the prevalence of vertebral fractures in these black and white South African females. 1. Significant ethnic differences were noted in the presence and frequency of historical clinical and lifestyle risk factors for osteoporosis. Blacks were heavier and shorter, they consumed less calcium, were more inactive, preferred depot-medroxyprogesterone acetate as contraceptive agent and were of higher parity. Whites smoked more, preferred oral oestrogen containing contraceptive tablets and were more likely to have a positive family history of osteoporosis. Hormone therapy was used almost exclusively by postmenopausal whites. Inter-ethnic differences in weight, physical activity and high parity was most marked in the older subjects. 2. We found that peak spinal BMD was lower, but peak femoral BMD similar or higher (depending on the specific proximal femoral site measured) in black South-African females compared with whites. The lower peak spinal BMD was mainly attributed to lower BMD’s in the subgroup of black females with normal to low body weight, indicating that obesity either protected black females against a low spinal BMD or enhanced optimal attainment of bone mineral. An apparent slower rate of decline in both spinal- and femoral BMD with ageing was noted in the black females compared with whites in this cross-sectional study – an observation which will require confirmation in longitudinal, follow-up studies. This resulted in similar spinal BMD values in postmenopausal blacks and whites, but significantly higher femoral BMD measurements in blacks. The volumetric calculation of bone mineral apparent density (BMAD) at the lumbar spine and femoral neck yielded similar results to that of BMD. Spinal BMAD was similar in blacks and whites and femoral neck BMAD was consistently higher in all the menopausal subgroups studied. Weight significantly correlated with peak- and postmenopausal BMD at all sites in the black and white female cohorts. Greater and better maintained body weight may be partially responsible for slower rates of bone loss observed in black postmenopausal females. Most of the observed ethnic difference in BMD was, in fact, explained by differences in body weight between the two cohorts and not by ethnicity per se. 3. A low body weight and advanced age was identified as by far the most informative individual clinical risk factors for osteopenia in our black and white females, whereas physical inactivity was also identified as an important individual risk factor in blacks only. Risk assessment tools, developed and validated in Asian and European populations, demonstrated poor sensitivity for identification of South African women at increased risk of osteopenia. The osteoporosis risk assessment instrument (ORAI) showed the best results, with sensitivities to identify osteopenic whites at most skeletal sites approaching 80% (78% - 81%). The risk assessment tool scores appear to be inappropriate for our larger sized study cohort, especially our black subjects, thus resulting in incorrect risk stratification and poor test sensitivity. General discriminant analysis identified certain risk factor subsets for combined prediction of osteopenia in blacks and whites. These risk factor subsets were more sensitive to identify osteopenia in blacks at all skeletal sites, compared with the risk assessment tools described in the literature. 4. Higher ultrasonographically measured broadband ultrasound attenuation (BUA) and speed of sound (SOS) values were documented in our elderly blacks compared with whites, even after correction for differences in DEXA determined BMD at the spine and proximal femoral sites. BUA and SOS showed no decline with ageing in blacks, in contrast to an apparent significant deterioration in both parameters in ageing whites. If these quantitative ultrasound (QUS) parameters do measure qualitative properties of bone in our black population, independent of BMD as has been suggested in previous work in Caucasian populations, the higher values documented in elderly blacks imply better preservation of bone quality in ageing blacks compared with whites. The correlation between QUS calcaneal BMD and DEXA measured BMD at the hip and spine was modest at best. QUS calcaneal BMD was therefore unable to predict DEXA measured BMD at clinically important fracture sites in our study population. 5. Bone turnover, as assessed biochemically, was similar in the total pre- and postmenopausal black and white cohorts, but bone turnover rates appeared to differ with ageing between the two racial groups. A lower bone turnover rate was noted in blacks at the time of the menopausal transition and is consistent with the finding of a lower percentage bone loss at femoral sites at this time in blacks compared with whites. Bone turnover only increased in ageing postmenopausal blacks, and this could be ascribed, at least in part, to the observed negative calcium balance and the more pronounced secondary hyperparathyroidism noted in blacks. Deleterious effects of secondary hyperparathyroidism on bone mineral density at the proximal femoral sites were demonstrated in our postmenopausal blacks and contest the idea of an absolute skeletal resistance to the action of PTH in blacks. The increase in bone turnover and the presence of secondary hyperparathyroidism due to a negative calcium balance may thus potentially aggravate bone loss in ageing blacks, especially at proximal femoral sites. 6. Shorter, adult black women have a significantly shorter hip axis length (HAL) than whites. This geometric feature has been documented to protect against hip fracture. The approximately one standard deviation (SD) difference in HAL between our blacks and whites may therefore significantly contribute to the lower hip fracture rate previously reported in South African black females compared with whites. Average vertebral size was, however, smaller in black females and fail to explain the apparent lower vertebral fracture risk previously reported in this population. Racial differences in vertebral dimensions (height, width) and/or other qualitative bone properties as suggested by our QUS data may, however, account for different vertebral fracture rates in white and black women – that is, if such a difference in fact exists. 7. The number of women with a history of falls was similar in our black and white cohorts, and in both ethnic groups the risk of falling increased with age. There is a suggestion that the nature of falls in our black and white postmenopausal females may differ, but this will have to be confirmed in a larger study. Fallers in our postmenopausal study population were more likely to have osteoporosis than non-fallers. Postmenopausal blacks in our study demonstrated poorer outcomes regarding neuromuscular function, Vitamin D status and visual contrast testing and were shown to be more inactive with ageing compared with whites. An increased fall tendency amongst the black females could not however be documented in this small study. Quadriceps weakness and slower reaction time indicated an increased fall risk amongst whites, but were unable to distinguish black female fallers from non-fallers. 8. Vertebral fractures occurred in a similar percentage of postmenopausal blacks (11.5%) and whites (8.1%) in our study. Proximal femoral BMD best identified black and white vertebral fracture cases in this study. Quite a number of other risk factors i.e. physical inactivity, alcohol-intake, poorer physical performance test results and a longer HAL were more frequent in the white fracture cases and could therefore serve as markers of increased fracture risk, although not necessarily implicated in the pathophysiology of OP or falls. However, in blacks, only femoral BMD served as risk factor. Similar risk factors for blacks and whites cannot therefore be assumed and is deserving of further study. White fracture cases did not fall more despite lower 25-OH-Vitamin D, poorer physical performance and lower activity levels than non-fracture cases. Calcaneal ultrasonography and biochemical parameters of bone turnover were similar in fracture and non-fracture cases in both ethnic groups. Our study data on vertebral fractures in this cohort of urbanized blacks thus cautions against the belief that blacks are not at risk of sustaining vertebral compression fractures and emphasize the need for further studies to better define fracture prevalence in the different ethnic populations of South Africa. 9. In our study, hormone therapy in postmenopausal white women improved bone strength parameters and reduced fall risk. In hormone treated whites compared with non-hormone users, a higher BMD at the spine and proximal femur as determined by DEXA were documented and all QUS measurements were also significantly higher. The biochemically determined bone turnover rate, as reflected by serum osteocalcin levels, was lower in hormone users. Fall frequency was lower in the older hormone treated women (≥ 60yrs) and greater quadriceps strength and reduced lateral sway was noted. Only one patient amongst the hormone users (2%) had radiological evidence of vertebral fractures compared with four patients (6%) amongst the never-users. As hormone therapy was used almost exclusively by whites in this study population, the impact of hormone therapy on postmenopausal black study subjects could not be assessed.