Prevalence of vitamin D inadequacy in Scottish adults with non-vertebral fragility fractures

Abstract
It is well established that vitamin D levels are sub-optimal in the elderly and that adults with fragility fracture are more likely to have serum vitamin D levels either lower than those of control patients of similar age, or below the normal range. To investigate the prevalence of vitamin D inadequacy in an elderly population presenting to the South Glasgow Fracture Liaison Service with non-vertebral fragility fractures in order to assess the extent of the problem. The retrospective arm of this study used data from an established database to identify patients aged over 50 years admitted to South Glasgow University Hospitals over the previous 4 years with hip fracture. The prospective arm identified the first 50 patients aged over 50 presenting with a clinical non-vertebral fragility fracture with osteoporosis as measured by axial spine and/or hip DEXA (T-score < -2.5) after November 2004. In the retrospective arm, 626 patients were identified from the database: mean age 80.5 years; 94% were aged over 60 and 74% were aged over 75. Data analysis was limited to 548 patients aged over 60 years with vitamin D recordings and not receiving supplementation with calcium and vitamin D. The mean vitamin D level was 24.7 nmol/L (9.9 ng/ml) SD = 17, however, it is likely that the true mean is lower since in approximately 25% of cases vitamin D levels were reported as < 15 nmol/L (effectively unrecordable). These were transcribed as 15 nmol/L in order to permit a numerical value to be calculated. In the absence of an agreement on what should constitute a diagnostic serum level of vitamin D inadequacy, a number of thresholds were considered--97.8% had vitamin D levels below 70 nmol/L and 91.6% had vitamin D levels below 50 nmol/L. There were no significant differences by patient sex, age or season of presentation. The mean age of patients in the prospective arm was 65.8 years (range 50.6-83.8), 72% were aged over 60 and 16% were aged over 75. The mean vitamin D level was 44.1 nmol/L (18.4 ng/ml) SD = 25.3; 82% had vitamin D levels below 70 nmol/L and 72% had vitamin D levels below 50 nmol/L. Although numbers were too small to justify extensive subgroup analyses, the mean vitamin D level in the 13 patients with hip fracture (34.5 nmol/L) was lower than in the 37 with non-hip fractures (48.2 nmol/L). This study confirms almost universal vitamin D inadequacy among 548 elderly patients admitted to hospital with hip fracture, regardless of whether a threshold of 50 nmol/L or 70 nmol/L was used. However, among a prospective subset of 50 patients with clinical fragility fractures, especially those with non-hip fractures, the prevalence of inadequacy was substantially lower. It may be that vitamin D represents a correctable risk factor for fragility fracture in the elderly, possibly specifically for the hip.