Arthritis & Rheumatism, Volume 60,
October 2009 Abstract Supplement

The 2009 ACR/ARHP Annual Scientific Meeting
Philadelphia October 16-21, 2009.

Risk of Atypical Subtrochanteric Fractures in Older Veterans at High Risk for Osteoporosis

Safford1,  Monika M., Richardson2,  Kelly, Sarrazin2,  Mary, Cram2,  Peter, Curtis1,  Jeffrey R., Delzell1,  Elizabeth, Barasch1,  Andrei

UAB, Birmingham, AL,
Iowa City VAMC, Iowa City, IA


Atypical subtrochanteric fractures may be a risk for longer-standing bisphosphonate (BP) users. We examined the risk of these atypical fractures associated with BP therapy among veterans at high risk of osteoporotic fracture.


National administrative VA data from 1999–2008 included inpatient and outpatient utilization files and pharmacy data. We included individuals presumed to have osteoporosis, i.e., admitted for a non-trauma fracture to a VA facility at age >=45. We excluded individuals with enrollment in the VA system for < 180 days prior to 1st fracture, metastatic cancer, and history of chemotherapy. The primary outcome was a subsequent "atypical" subtrochanteric fracture, defined by ICD-9 code "fracture of subtrochanteric section of femur, closed"[820.22] in the absence of trauma ICD-9 codes or open fractures. Cumulative BP use was the main exposure of interest, and only individuals who initiated BP use at least 365 days after first appearance in VA data were included to better define duration of use. We constructed Cox proportional hazards models to examine associations between BP and risk of subsequent atypical fracture or more typical hip fracture. Available covariates included demographics (age at 1st fracture, sex, and race), baseline comorbid medical conditions, medications associated with osteoporosis, fall risk, and distance to the nearest VA Medical Center. Covariates bivariately associated with p<.20 were included in multivariable models.


Of 80,684 eligible veterans, 95% were men, 3.6% used BPs before their 1st fracture, and another 6.4% had BPs initiated afterward. We detected 45 atypical subtrochanteric and 1,805 typical hip 2nd fractures over a mean follow-up of 4.2 years. BP use duration of individuals with atypical subtrochanteric, typical hip and those with no 2nd fractures are shown in the Table. Compared with nonusers, BP users had the following hazard ratios (HR) for subsequent atypical subtrochanteric fractures: <1 year of BP use, HR 1.4 (95% CI 0.18–10.46); 1–6 years, HR 0.47 (95% CI 0.06–3.57); >6 years, HR 4.5 (95% CI 0.58–34.24). In contrast, compared with nonusers, BP users had the following HR for subsequent typical hip fractures: <1 year of use, HR 1.27 (95% CI 0.98–1.66); 1–6 years, HR 0.80 (95% CI 0.65-.099); >6 years, HR 0.39 (95% CI 0.16–0.95).

 Type of 2nd fracture
  Atypical subtrochancteric (N=45)Typical hip (1,805)No 2nd fracture (70,305)
Total BPs up to time of 2nd fracture or end of observationNone41 (91%)1,618 (90%)64,083 (91%)
 <1 yr1 (2%)71 (4%)1,344 (2%)
 1–6 yrs1 (2%)85 (5%)2,811 (4%)
 >6 yrs2 (4%)31 (2%)2,067 (3%)


While BP use >6 years trended towards an association with increased risk for atypical subtrochanteric fractures, atypical fractures were rare, and long-term BP use was associated with a substantially decreased risk for hip fractures, which were 40 times more common than atypical fractures. These findings should be confirmed to provide information for long-term BP users and their physicians about overall risks and benefits.

To cite this abstract, please use the following information:
Safford, Monika M., Richardson, Kelly, Sarrazin, Mary, Cram, Peter, Curtis, Jeffrey R., Delzell, Elizabeth, et al; Risk of Atypical Subtrochanteric Fractures in Older Veterans at High Risk for Osteoporosis [abstract]. Arthritis Rheum 2009;60 Suppl 10 :882
DOI: 10.1002/art.25962

Abstract Supplement

Meeting Menu