Osteoporosis is a common skeletal disease in the elderly, and osteoporosis-related fractures are one of the most serious medical problems that pose a considerable socioeconomic burden1,2,3). Of them, hip fractures are a major health concern and are more serious than other osteoporotic fractures in elderly in terms of decreased mobility4), diminished quality of life5), and excess mortality4,6). In addition, medical expenses for hip fractures are the highest compared to those of other osteoporotic fractures7).

Second hip fractures can occur after prior hip fractures. The estimated incidence of a second hip fracture among survivors is 2-20%, depending on the length of follow-up8,9). A number of studies have reported that sustaining an osteoporotic fracture increases the risk of subsequent fracture10,11). Several studies have reported that patients who had a second hip fracture are associated with higher mortality, more comorbidities, and decreased independence than those of patients with a unilateral hip fracture12,13).

Although several factors such as female sex, living institution, osteoporosis, low vision, dementia, dizziness, cardiac disease, and respiratory disease have been suggested as risk factors after hip fracture13), the overall risk of a second hip fracture is at least two-fold higher compared to the risk of the first14,15). The incidence, risk factors, and short-term mortality of a second hip fracture have been reported, but mid- to long-term mortality rates after a second hip fracture are seldom reported.

The purpose of this study was to determine the incidence, morbidity, and mid-term mortality of a second hip fracture and evaluate the cause of death after a second hip fracture.

The design and protocol of this study were approved by the institutional review board at each hospital, and informed consent was waived. Participants were drawn from a cohort that has been monitored since 2002 to determine the incidences of hip fractures and mortality and morbidity rates after hip fracture5,16). This larger cohort was recruited from Jeju Island, the largest Korean island, which is located southwest of the Korean peninsula. In 2011, Jeju Island had a population of 576,156 including 74,985 men and 90,481 women aged 50 years or more, and 12.1% were aged 65 years or more (9.3% of men and 14.9% of women). Eight hospitals (one university hospital and seven general hospitals) are on the island with orthopedic departments and emergency admission facilities. The medical records and radiography of the eight hospitals between January 2002 and December 2011 were reviewed to identify patients with hip fractures.

Patients who met the diagnostic criteria for femoral neck and intertrochanteric fractures in the International Classification of Diseases 10th revision (ICD-10; S720 and S721) and were older than 50 years at the time of the fracture were recruited. Exclusion criteria were nonresidents of Jeju Island, pathological bone fractures (e.g., metastasis), isolated fractures of the greater or lesser trochanter, and fractures of the subtrochanteric region due to a high energy injury16).

Hospital data evaluations were performed 6 and 12 months after discharge from the hospital, and every 6 months thereafter. We determined the diagnosis at admission, mechanism of injury, and procedures performed during hospitalization from the medical records.

A total of 2,055 hip fractures were enrolled. Of them, 98 patients (13 men and 85 women) experienced a second hip fracture between 2002 and 2011. There were femoral neck fractures in 45 hips (45.9%) and intertrochanteric fractures in 53 hips (54.1%). Sixty-four hip fractures occurred in the same diagnostic area as the first hip fracture. The mean age of the patients at the time of the second hip fracture was 80.8 years (range, 54-94 years). All patients were followed for a mean of 5 years (range, 2-9 years) (Table 1).

The incidence of a second hip fracture was 4.7% (98/2,055 hips). Second hip fractures occurred in 75 (76.5%) hips within 3 years and tended to decrease (Fig. 1).

Forty-one patients (four men and 37 women) died during the follow-up period. The cumulative mortality rate after fracture was 8.2% at 3 months (eight hips), 10.5% at 6 months (10 hips), 15.2% at 1 year (15 hips), 23.5% at 2 years (23 hips), and 41.8% at 5 years (41 hips) (Fig. 2).

Univariate analysis for risk factors for a second hip fracture showed that sex (P=0.103), age (P=0.976), type of fracture repair (P=0.521), pre-injury activity (P=0.123), and medical comorbidities using the modified Charlson's Comorbidity Index (P=0.324) were not associated with risk factor and surgical treatment (P=0.000) was only significant (Table 2). However, this finding might be a confounding factor for a second hip fracture because patients who undergo conservative treatment after a hip fracture have high mortality and little chance to survive over the long-term. The causes of death were pulmonary complications in 15 cases, congestive heart disease complications in 10, liver disease in four, cancer in three, and unknown in nine (Table 1).

Of the 57 patients that remained alive at the final follow-up, 21 of 45 patients (46.6%) who had pre-injury outdoor activity levels (Grade I and II) achieved outdoor activity at a mean of 5 years after the second hip fracture (range, 2-9 years) (P<0.001).

Although the incidence of hip fracture in several advanced countries has tended to decrease, hip fracture incidence in South Korea has been increasing steeply during the last decades16,19,20,21). In addition, the incidence of a first hip fracture in the Jeju cohort (0.2%; 304/165,466 in 2011) was much lower than the incidence of a second hip fracture (4.8%). Although mortality rates after a first hip fracture have been reported, few data are available on mid- to long-term mortality and morbidities after a second hip fracture. In this prospective cohort observational study, we found that the 5-year cumulative mortality rate after a second hip fracture was 41.8%.

The mid-term cumulative crude mortality rate after a second hip fracture in this cohort study was similar to our previous study after a first hip fracture in the same cohort5). Lee et al.5) performed a prospective cohort study to evaluate morbidity and mortality in 790 patients older than 50 years who sustained a femoral neck or intertrochanteric fracture from 2002 to 2006. They reported that accumulated mortality after the first hip fracture was 16.7% at 1 year and 45.8% at 5 years. In our study, the cumulative mortality rate continued to increase throughout the study period, and the mean crude mortality rates in patients older than 50 years at 1 and 5 years after the second hip fracture were 15.2% and 41.8%, respectively (Table 3). When we compared those previous studies regarding mortality after a second hip fracture8,9), our findings regarding crude mortality rates concurred with values reported previously. Ryg et al.9) performed a nationwide population-based cohort study on 169,145 cases with a first hip fracture and reported that patients with a second hip fracture had substantial mortality rates of 27% and 64% in men and 21% and 58% in women after 1 year and 5 years, respectively. Their data were similar to mortality rates published previously and observed in the cohort after a first hip fracture (32.9% vs. 63.4% in men and 23.9% vs. 56.2% in women)9,22).

Surgical treatment was the only risk factor associated with a second hip fracture in our study. However, patients who undergo conservative treatment after a hip fracture have high mortality and little chance to survive over the long-term5). Patients who undergo surgical treatment survive but have more chance to suffer a second hip fracture23). Therefore, surgical treatment might be a confounding factor for a second hip fracture. Liu et al.24) performed a systemic review and metanalysis of risk factors for a second hip fracture using 13 case-control studies. They suggested that possible risk factors for a second hip fracture were female sex, living institution, osteoporosis, low vision, dementia, dizziness, cardiac disease, and respiratory disease. However, in that study, medical comorbidities were not an associated risk factor for a second hip fracture.

Activity levels after the second hip fracture revealed increased mobility limitations and disability in the cohort at the latest follow-up. Patients who survived the second hip fracture had poor functional outcomes. Twenty-four of 45 patients (53.3%) who had pre-injury outdoor activity levels (Grade I and II) did not achieve independent outdoor activity at a mean of 5 years after the second hip fracture. This finding agrees with a previous study regarding mobility after a first hip fracture. Lee et al.5) reported that 67.9% (199/293) of patients capable of independent outdoor activity before fracture retained this ability at a mean of 6 years after the first hip fracture.

This study had several limitations. First, the Jeju cohort population comprised only 1.1% of the total registered population of South Korea. Therefore, this study was not large enough to generalize as nationwide data. However, we evaluated the epidemiological data prospectively during a mean follow-up of 5 years. Second, sex-specific mortality after the second hip fracture could not be evaluated because the proportion of men was too small. Third, walking ability as morbidity was evaluated over an extended period (range, 2-9 years). However, several studies have reported that functional recovery can be largely achieved during the first 4-6 months after a hip fracture25,26,27,28). Therefore, the activity evaluation bias in this study may have been minimized because all patients were followed for a minimum 2 years postoperatively. Fourth, in univariate analysis, surgical treatment (P=0.000) is only significant factor for second hip fracture. However, this finding is difficult to accept as an important risk factor. Mostly the reason of conservative treatment in patients with hip fracture was poor general condition and had high rate of early mortality. Therefore, second hip fractures occur in survivors from hip fracture. In addition, patients with surgical treatment might be susceptible to second hip fracture because of more severe fracture. However, severity of fracture in this study did not evaluate. These confounding factors might be influenced on this finding. Finally, antiosteoporosis medications are generally recommended to prevent second hip fractures29,30). The evaluations of type and duration of antiosteoporosis medications are helpful to understand this study. However, we did not evaluate postoperative antiosteoporotic medications in this study group.

The incidence of a second hip fracture was 4.8%. The cumulative mortality rate after the second hip fracture at the 5-year follow-up was 41.8%. Therefore, our results demonstrate that a secondary fracture prevention program is necessary to prevent second hip fractures in elderly patients.