Session V - Geriatrics

Saturday, October 23, 1999 Session V, Paper #38, 9:36 a.m.

Changes in Bone Mineral Density Six Years following a Hip Fracture: A Prospective Longitudinal Study

Douglas R. Dirschl, MD; Luis A. Piedrahita, MD; Richard C. Henderson, MD, University of North Carolina School of Medicine, Raleigh, NC

Purpose: One well-recognized risk factor for sustaining a hip fracture is a previous osteoporosis-related hip fracture. We and others have previously reported an accelerated rate of loss of bone mineral density (BMD) from the lumbar spine and contralateral femoral neck in the first year following a hip fracture; the observed rate of loss from the femoral neck was 4-5 times that seen in non-fractured postmenopausal women. The purpose of the present study is to extend the results of our previous prospective investigation to determine if an accelerated rate of loss of bone mineral continues for 6 years after a hip fracture.

Methods: Eighty-five elderly patients with osteoporotic fractures of the hip underwent an evaluation at the time of injury (baseline group). Sixteen patients (19%) died in the first year after injury and 14 patients declined follow-up; thus, 55 patients returned for repeat evaluation at one year following fracture (1-year subgroup). Between 1 and 6 years after fracture, an additional 24 patients died and 10 patients declined follow-up; thus, 21 patients returned for an evaluation an average of 6 years (range 67-86 months) after fracture (6-year subgroup). This subgroup of 21 patients, with complete data at the time of injury and at 1 and 6 years following injury, is the subject of this report. The baseline and follow-up evaluations were identical, consisted of clinical, nutritional, BMD, and serologic assessments, and have been reported previously. The BMD assessments were performed with a single DEXA unit over the six year study period. BMD was measured at the lumbar spine and contralateral hip and changes in BMD were expressed as a percentage of baseline BMD.

Results: There was a marked decreased in BMD in the 6-year subgroup in the first year following fracture, with the mean change in BMD being ­4.3+1.2% at the femoral neck and ­1.8+1.1% at the lumbar (mean + SEM). Between 1 and 6 years after fracture, however, there was a dramatic increase in the BMD at both the femoral neck and lumbar spine measurement sites. Relative to baseline, the mean changes were 7.7+4.3% at the femoral neck and 4.5+2.1% at the lumbar spine. In many cases, the loss of bone mineral which occurred in the first year after fracture was completely recouped in the subsequent five years; the absolute BMD six years following fracture was greater than baseline BMD in the femoral neck of 60% of patients (Figure 1)

Figure 1: Change from baseline in femoral neck BMD following a hip fracture

and in the lumbar spine of 48% of patients. Five of the 21 patients (24%) sustained a contralateral hip fracture in the 6 years following the index fracture. Lumbar spine BMD was lower at baseline (p=0.02), one year following fracture (p=0.007), and six years following fracture (p=0.003) in patients who sustained a second hip fracture than in those who did not. There was a general decrease in the functional activity level of patients in the six years following a hip fracture, but there were no statistically significant relationships between changes in BMD and the functional mobility of patients. The mean calcium intake in patients improved remarkably in the six years following fracture, but there was no correlation between daily calcium intake and changes in BMD.

Discussion: The patients in this report constitute only one-fourth of the 85 patients who completed the baseline evaluation for this prospective study; death (40 patients), relocation (12 patients), and unwillingness to complete the follow-up evaluations (12 patients) are responsible for the loss of patients. The 6-year subgroup of patients consists of younger and more active patients than the baseline group, but is probably representative of patients surviving six years after a hip fracture. The significant loss of BMD we and others previously observed in the first year following a hip fracture was at least partially recouped in 67% of patients, with 60% of patients having femoral neck BMD values greater at six years than at baseline. Although it is difficult to account for the unexpectedly large amount of BMD that was recovered at both measurement sites, the use of a single phantom and DEXA scanner throughout the study assure that the findings cannot be explained by technical error. Progression of osteoarthritis could perhaps falsely increase BMD measurements at the lumbar spine, but osteophytes would not be expected to increase BMD in the femoral neck. The decline in functional level of physical activity in these patients makes it unlikely that a progressive increase in weightbearing on the non-fractured hip could explain the findings in this study. Although calcium intake did increase in these patients following fracture, no correlation between this and changes in BMD could be demonstrated.

Conclusion: During the first year following a hip fracture, there is a rapid loss of bone mineral from the lumbar spine and contralateral femoral neck. Between one and six years following fracture, however, BMD is likely to increase, perhaps to levels greater than those at baseline. Although this investigation is quite small, the results of this study suggest that pharmacological intervention to prevent accelerated loss of BMD after a hip fracture may not need to be continued indefinitely. One or two years of pharmacologic therapy may be all that is necessary, at which time the period of accelerated loss of BMD following a hip fracture may be past. The findings of this study point to the importance of further larger studies to clarify the natural history of BMD following a hip fracture and the impact of pharmacological intervention on that natural history.