Background It is well documented that bone mineral density (BMD) declines with chronologic age. However, there are considerably fewer data regarding the relationship between BMD and “biologic age” or frailty. The Vulnerable Elders Survey (VES-13) questionnaire, which measures self-reported functional impairments, is widely used in geriatric medicine for assessment of frailty and evaluation of the health status and care needs of the elderly. In the current study, we examined whether the VES-13 questionnaire can be used to predict low calcaneal BMD and aid in identifying those elderly subjects who are at highest risk of fracture and would have the greatest benefit from an aggressive approach to the diagnosis and treatment of osteoporosis.
Methods Through community screening efforts, 207 community-dwelling elderly subjects (aged 65 to 95 years, mean 77 ± 7 years, 170 females, 37 males, 127 African Americans, 80 Caucasians) had measurement of calcaneal BMD using the portable PIXI densitometer and assessment of frailty using the VES-13 questionnaire. The total VES-13 score was calculated as described in the ACOVE (Assessing Care of Vulnerable Elders) project, in which the score, ranging from 0 to 10, is a composite of the points for age, general health, six functional activities, and five instrumental activities of daily living (IADL). Although the scoring system used in ACOVE was predictive of death and decline in other studies, it did not distinguish between subjects with a wide range of functional impairments. Furthermore, the total ACOVE score includes points for age, which is a strong independent predictor of BMD. Therefore, we developed a new scoring system where the new total score excluded age and was composed of the unweighted sum of the points for general health, functional activities, and IADL, with a minimum score of 12 and maximum of 60, with higher scores representing greater impairment. In addition, a functional subscore and IADL subscore were calculated by totaling the points for the six functional activities and five IADL questions, respectively. A physical subscore was then calculated as the sum of the functional and IADL subscores and reflected self-reported physical impairment without age or self-reported health status.
Results As expected, calcaneal BMD was significantly correlated with age (R2 = −.3193, p = < .0001), weight (R2 = .5559, p = < .0001), and height (R2 = .4204, p = < .0001) and was higher in men than in women (p = < .0001) and in African Americans than in Caucasians (p = < .01). BMD was also significantly correlated with VES-13 results using the total ACOVE score (R2 = −.2488, p = < .001), new total score (R2 = −.1824, p = < .01), IADL subscore (R2 = −.2398, p = < .001), and physical subscore (R2 = −.1881, p = < .01). In multivariate regression analysis with BMD as the outcome variable, VES-13 scores had a significant effect even when controlling for age, weight, sex, and race (new total score p = < .05, function subscore p = < .05, IADL subscore p = < .05, physical subscore p = < .01). In fact, BMD was best predicted by a model that included age, weight, race, sex, and the physical subscore (R2 = .4186, p = < .0001). According to the model, the difference in calcaneal BMD between subjects with the lowest and highest physical subscores was 0.102 g/cm2, corresponding to 1.275 BMD T-score units, and a doubling of the fracture risk.
Conclusions Frail elderly subjects have lower calcaneal BMD than expected for their age, sex, race, and weight. Assessment of frailty with an easily obtained, self-reported measure such as the VES-13 can identify community-dwelling elderly subjects who have higher fracture risk and greater potential benefit from an aggressive approach to the diagnosis and treatment of osteoporosis.