In 1984, an estimated 238,000 hip fractures occurred in the United States.1 By 1993, more than 307,000 hospitalizations for hip fractures were recorded in the United States, and more than 90% occurred in people older than 50 years.2,3 It has been suggested that the incidence of hip fractures increases exponentially with age-approximately doubling every decade after age 50. Given the advancing age of the US population, the overall incidence will continue to increase sharply. According to the projected figures of Cummings et al,1 the incidence of such fractures could increase to 347,000 by the year 2020 and 512,000 by the year 2040.
Several previous studies assessed the incidence and relative risks of a second fracture involving the contralateral proximal femur. In 1982, Boston4 examined a large series of patients presenting with a femoral neck fracture. Fifty-three (10.6%) of 500 patients sustained fractures of both hips; of these 53 patients, 83% demonstrated symmetric fracture patterns.
Dretakis et al5 identified a 7.5% incidence of bilateral, noncontemporary fractures of the proximal femur; symmetrical fracture patterns were noted in 70%. In a population-based study, Melton et al6 estimated that the risk of recurrence of a hip fracture, whether ipsilateral or contralateral, was 29% by 20 years, or 1.6 times greater than expected. These findings and the work of others have generated further interest in identifying an underlying etiology for this trend.
All fractures about the hip that presented to our institution over a 7-year period were reviewed. This study identified all patients who sustained fractures of both hips, with at least one of the fractures occurring in the study period. The relative incidence of bilateral fractures, the time interval between fractures, and whether the second fracture would present as a similar pattern as the first were determined.
Materials and Methods
Computerized medical records at Dartmouth-Hitchcock Medical Center in Lebanon, New Hampshire, were used to identify the patient population. From January 1989 through December 1995, all patients admitted with the diagnosis of a fracture about the hip were selected. Patient records were reviewed for information regarding the mechanism of injury, nature of the fracture, age of the patient at the time of the initial and subsequent fracture, and type of operation performed for each fracture. Patients younger than 50 years, those who sustained a pathologic fracture, or a fracture secondary to a high-energy injury were excluded.
Standard anteroposterior pelvis and lateral radiographs were available for all patients. If, at the time of fracture presentation, the patient was noted to have evidence of a prior hip fracture involving the contralateral side, old radiographs and records were retrieved for review. Fracture categories included femoral neck, intertrochanteric, subtrochanteric, and unspecified trochanteric fractures.
Patients included in the study were drawn from the wide geographic area served by our institution. The hospitals in the surrounding communities provided overlapping coverage in the management of proximal femoral fractures. Therefore, hip fractures sustained in our geographic service area were not exclusively treated at our tertiary care center.
During the study period, 602 patients with fractures involving the proximal femur were admitted to our facility. Fracture incidence based on location included 313 closed femoral neck, 233 closed intertrochanteric, 50 closed subtrochanteric fractures, 4 unspecified closed trochanteric, 1 open intertrochanteric, and 1 open subtrochanteric fractures. Five hundred forty-five fractures met our criteria. These included 294 closed femoral neck, 215 closed intertrochanteric, 34 closed subtrochanteric, and 2 unspecified closed trochanteric fractures.
Hospital records of all admissions were reviewed. Patients were treated with nonoperative and operative means. The majority of patients underwent operative intervention including multiple pinning, mono- and bipolar hemiarthroplasty, dynamic hip screws, blade plate fixation, total hip arthroplasty (THA), and intramedullary nail fixation.
Of the 545 patients, 79 (14.5%) were treated with multiple pinning, 12 (2.2%) with a monopolar endoprosthesis, 182 (33.4%) with a bipolar endoprosthesis, 231 (42.4%) with dynamic hip screws, 4 (0.7%) with a blade plate, 4 (0.7%) with a THA, and 19 (3.5%) with an intramedullary rod of the femur. Fourteen (2.6%) patients were treated nonoperatively, primarily because of associated medical conditions that precluded operative intervention.
Fracture admissions identified 29 patients who sustained fractures of both hips, representing an overall incidence of 5.5%. By inclusion criteria, at least one of the fractures occurred during the study period. Of the 29 patients with bilateral fractures, adequate records and radiographs were available for 28 patients (5 [18%] men and 23 [82%] women); these patients comprised the study group. All fractures presented as closed injuries and were sustained after minor falls, except in one patient who sustained fractures of both right and left hips during separate cross-country ski accidents. Twenty-eight (50%) fractures involved the femoral neck; 24 (43%) were intertrochanteric and 4 (7%) were subtrochanteric fractures.
Eighteen (64%) of 28 patients sustained bilateral fractures with similar patterns. Of these 18 patients, there were 10 bilateral femoral neck and 8 bilateral intertrochanteric fractures. No cases of bilateral subtrochanteric fractures were identified.
Ten (36%) patients presented with dissimilar fracture patterns. Two patients sustained an intertrochanteric fracture followed by a subtrochanteric fracture on the contralateral side. Eight suffered one femoral neck fracture and one trochanteric fracture. Five of these patients initially sustained a femoral neck fracture and then an intertrochanteric fracture on the contralateral side; the remaining 3 patients initially suffered a trochanteric fracture, followed by a femoral neck fracture. One patient initially sustained a left subtrochanteric fracture that was treated with an intramedullary femoral nail. She subsequently sustained a femoral neck fracture above the intramedullary nail, requiring nail removal and conversion to a bipolar endoprosthesis.
In 15 patients, an initial fracture on the right was followed by a fracture on the left. Twelve patients initially presented with a left-sided fracture, followed by a fracture on the right. One patient presented with simultaneous bilateral femoral neck fractures. This patient was a nursing home resident whose history at the time of presentation was unclear. We were unable to precisely ascertain the timing of his fractures; however, it was believed that one fracture may have preceded the other by several days, although both fractures presented simultaneously.
Mean age at the time of the initial fracture was 77.5 years (range: 55.1-91.3 years). Mean age at the time of the subsequent fracture was 79 years (range: 66.1-91.6 years). Eleven (39%) patients experienced a second fracture within 1 year and 21 (75%) patients within 3 years. The average time elapsed between fractures was approximately 2.4 years (range: 0-10.99 years). The average time elapsed between fractures of similar patterns was approximately 1.7 years (range: 0-7 years), whereas the average time between fractures of dissimilar patterns was approximately 3.5 years (range: 0.07-10.99 years). For femoral neck fractures, the mean age at the time of fracture was 82 years. The mean age for all intertrochanteric fractures was 77.4 years.
The incidence of fractures about the hip among the elderly poses a serious concern to society. In the face of the advancing age of the at-risk population, an increasing number of patients will require care for proximal femoral fractures. Although relative risk factors for fractures about the hip have been identified, little information is available regarding the likelihood that a patient who has sustained a unilateral hip fracture will subsequently present with a fracture of the contralateral hip. This review identified the relative incidence of bilateral fractures of the hip, the time interval between fractures, and whether the second fracture would present with a similar pattern as the first.
Hip fractures are a significant source of morbidity and mortality, and account for a significant portion of health-care expenditures in the United States and other countries. Over 300,000 hip fractures occurred in 1993 in the United States. One of every 3 women and 1 of every 6 men aged older than 90 years will sustain a fracture about the hip.2 The vast majority of these fractures are treated via surgical means.
However, following surgical stabilization, it is estimated that 20% become functionally nonambulatory, 10% to 15% are rendered household ambulators, and the remaining 50% to 65% of patients will regain their previous level of ambulation.2 During the first 12 months following fracture and surgery, a relative increase in mortality rates exists for these patients. One-year mortality rates are estimated to range from 14% to 36%.
In 1984, approximately $7.2 billion was spent in the United States alone on the care of hip fractures. According to figures estimated by Cummings et al,1 the projected annual cost for hip fracture care, using conservative inflation rates of 3% and 5%, would increase to $31 and $62 billion, respectively, by the year 2020.
Much work has been done to determine risk factors that may predispose a patient to a hip fracture.7 Relative risk factors include physical deconditioning; alterations in neuromuscular function; previous hip fracture; altered mental status such as senile dementia; impaired vision; urban dwelling; excessive and prolonged alcohol consumption; the use of psychotropic medications; nutritional factors such as excess caffeine intake, copper, and albumin deficiencies; and premorbid anatomy. Additionally, predictors for the pattern of a fracture about the hip have been identified to include age, mechanism of injury, and premorbid anatomy. However, controversy still exists regarding the relative impact of these individual factors.
Dretakis et al8 investigated bone quality, as determined by dual photon absorptiometry and calcaneus broadband ultrasound attenuation, in patients with single and bilateral hip fractures. No significant difference in the bone quality was found. They did note that falls were significantly more common in patients with bilateral fractures.
Ferris et al9 studied the morphology of the femoral neck in an attempt to relate femoral neck length to fracture pattern. Their conclusions suggested a positive correlation between trochanteric fracture patterns and a short femoral neck (4.5±0.5 cm); longer necks (5.4±0.5 cm) were associated with subcapital fractures or osteoarthritis. They postulated that differences in premorbid anatomy could lead to differences in biomechanical loading and subsequent responses to injury.
Boston4 examined the records of 500 patients with femoral neck fractures. He found a 10.6% incidence of bilateral hip fractures. Of these patients, 83% demonstrated similar fracture patterns bilaterally. Thirty-eight percent fractured the contralateral hip within 1 year; 70% sustained the fracture within 3 years. Boston4 identified those patients with a fracture about the hip as a high-risk group and recommended prophylactic measures to reduce the risk of a contralateral fracture.
Similarly, Schroder et al10 found 256 second hip fractures among 3898 patients aged older than 40 years. Ninety-two percent of the second fractures involved the contralateral side, occurring at a mean of 3.3 years after the initial fracture. They noted the risk of the first fracture to be 1.6 per 1000 men/year and 3.6 per 1000 women/year. The risk for the second fracture was greatly increased: 15 per 1000 men/year and 22 per 1000 women/year.
Huang et al11 reviewed 1644 patients who sustained a proximal femoral fracture. Twenty-eight patients were readmitted with a second hip fracture, 26 involving the contralateral side. They found significant fracture symmetry between the initial and subsequent fracture in 69% (18/26) of the patients. They concluded this symmetrical fracture pattern was due to senile osteoporosis.
Chiu et al12 studied 35 patients with sequential fractures of both hips. Evidence of osteomalacia, concomitant neurologic disease, and institutionalization were common factors in this group. Twelve contralateral fractures occurred within 1 year, but specific fracture patterns were not addressed.
Other studies have suggested a relative increase in the ratio of intertrochanteric-to-femoral neck fractures with advancing age. Lawton et al13 studied 128 patients with fractures about the hip. They found the patients with trochanteric fractures to have more advanced physiologic age. In addition, this group of patients demonstrated significant increase in length of hospitalization and need for assistance in postoperative mobilization. They recommended maintenance of mobility in the elderly at-risk population to preserve bone strength and aid in the recovery after surgery.
Wolinsky and Fitzgerald14 found 27 cases of subsequent hip fractures in a series of 368 initial fractures. This incidence equated to a rate of 1 for every 33.8 person-years. Poor perceived health status and dizziness were identified as independent risk factors for subsequent hip fracture.
Finsen and Benum15 compared 575 hip fracture patients with fracture prevalence among the general population. The probability of later hip fracture increased with the number of previously sustained fractures in women younger than 70 years and men younger than 80 years.
In this study, 29 patients who had sustained fractures of both hips were identified. The overall incidence of bilateral hip fractures was 5.5%, which is comparable to previously reported incidence rates (e.g., 10.6% by Boston,4 7.3% by Wolinsky and Fitzgerald,14 6.6% by Schroder et al,10 and 1.6% by Huang et al11). Our figures may reflect several factors: the wide geographic area served by our hospital; the fact that our institution is a tertiary care facility, with surrounding hospitals that also treat fractures about the hip; and the lack of longitudinal care at our center for all of the patients included in this study. As reported by Boston4 and Huang et al,11 a significant similarity between fracture patterns in a majority of these patients was noted. Eighteen (64%) of 28 patients with bilateral fractures demonstrated similar fracture patterns on both sides (P=>066; 90% confidence interval). These findings suggest a subsequent fracture of the contralateral hip is likely to be of similar fracture pattern as the initial fracture.
Specific analysis of the interval between fractures and the fracture pattern has not been performed previously. In this study, the average interval between fractures of similar nature, when compared to fractures of dissimilar nature, approached a statistically significant difference, 1.7 years versus 3.5 years. This difference would argue against femoral neck length as a predictor of subsequent fracture pattern. Rather, it suggests the effects of physiologic age and the accordant changes in bone architecture play a larger role in predicting subsequent fracture pattern. This argument is supported by the work of Brocklehurst et al,16 who noted a relative increase in the ratio of trochanteric to femoral neck fractures with advancing physiologic age.
This review suggests that at least 1 in 20 patients who sustain a fracture about the hip can expect to suffer a fracture involving the contralateral side. Prophylactic measures are beneficial in reducing the risk of initial, but especially subsequent, contralateral hip fractures in the elderly. Cummings17 divided risk factors for hip fracture into treatable and untreatable factors. Treatable factors can be addressed via physical therapy for muscle strengthening and gait training, the use of external ambulatory aides, minimizing the use of psychotropic medications, improving visual acuity, removing potential obstacles within the home, and initiation of medical therapy for the treatment of underlying osteoporosis.
Although the orthopedic surgeons treating fractures about the hip usually do not have contact with patients prior to fracture, it is incumbent on the treating physician and involved medical personnel to initiate measures that can help reduce the risk of a subsequent, contralateral hip fracture.
- Cummings SR, Rubin SM, Black D. The future of hip fractures in the United States. Numbers, costs and potential effects of postmenopausal
- Zuckerman JD. Hip fracture.N Engl J Med.1996; 334:1519-1525.
- Kyle RF, Cabanela ME, Russell TA, et al. Fractures of the proximal part of the femur.Instr Course Lect. 1995; 44:227-253.
- Boston DA. Bilateral fractures of the femoral neck.Injury. 1982; 14:207-210.
- Dretakis E, Kritsikis N, Economou K, Christodoulou N. Bilateral non-contemporary fractures of the proximal femur.Acta Orthop Scand. 1981; 52:227-229.
- Melton LJ 3rd, Ilstrup DM, Beckenbaugh RD, Riggs BL. Hip fracture recurrence. A population-based study.Clin Orthop. 1982; 167:131-138.
- Eiskjaer S, Ostgard SE. Risk factors influencing mortality after bipolar hemiarthroplasty in the treatment of fractures of the femoral neck.Clin Orthop. 1991; 270:295-300.
- Dretakis KE, Dretakis EK, Papakitsou EF, Psarakis S, Steriopoulos K. Possible predisposing factors for the second hip fracture.Calcif Tissue Int. 1998; 62:366-369.
- Ferris BD, Kennedy C, Bhamra M, Muirhead-Allwood W. Morphology of the femur in proximal femoral fractures.J Bone Joint Surg Br. 1989; 71:475-477.
- Schroder HM, Petersen KK, Erlandsen M. Occurrence and incidence of the second hip fracture.Clin Orthop. 1993; 289:166-169.
- Huang KC, Ku MC, Lee TS. Second hip fracture [Chinese].Zhongua Yi Xue Za Zhi (Taipei). 1992; 50:149-152.
- Chiu KY, Pun WK, Luk KD, Chow SP. Sequential fractures of both hips in elderly patients-a prospective study.J Trauma. 1992; 32:584-587.
- Lawton JO, Baker MR, Dickson RA. Femoral neck fractures-two populations.Lancet. 1983; 2:70-72.
- Wolinsky FD, Fitzgerald JF. Subsequent hip fracture among older adults.Am J Public Health. 1994; 84:1316-1318.
- Finsen V, Benum P. Past fractures indicate increased risk of hip fracture.Acta Orthop Scand. 1986; 57:337-339.
- Brocklehurst JC, Exton-Smith AN, Lempert-Barber SM, Hunt LP, Palmer MK. Fractures of the femur in old age: a two-centre study of associated clinical factors and the cause of the fall.Age Ageing. 1978; 7:2-15.
- Cummings SR. Treatable and untreatable risk factors for hip fracture.Bone. 1996; 18(suppl):S165-S167.