Journal of Bone and Mineral Metabolism

, Volume 31, Issue 4, pp 367–380

Osteoporosis and treatments in Japan: management for preventing subsequent fractures

Authors

    • Japan Health Technology AssessmentEli Lilly Japan K.K.
  • Russel T. Burge
    • Global Health OutcomesEli Lilly and Company
  • Jennifer A. Flynn
    • Japan Health Technology AssessmentEli Lilly Japan K.K.
  • Shonda A. Foster
    • Global Health OutcomesEli Lilly and Company
  • Hideaki Sowa
    • Medical ScienceEli Lilly Japan K.K.
Review Article

DOI: 10.1007/s00774-013-0444-y

Cite this article as:
Nojiri, S., Burge, R.T., Flynn, J.A. et al. J Bone Miner Metab (2013) 31: 367. doi:10.1007/s00774-013-0444-y

Abstract

Prevalent fractures are major contributors to an increased risk of subsequent fractures, particularly in people with osteoporosis. While many studies have been conducted to assess the incidence of fracture in Japanese people with osteoporosis, far fewer have been conducted to assess the risk of subsequent fractures. This article reviews the morbidity, mortality, and risk of fracture in patients who are at high risk of subsequent fracture in Japan and the current treatment options available for these patients. Osteoporotic fractures in Japan are associated with high morbidity and mortality that result in significant financial and social costs. The rise in the proportion of elderly women in the Japanese population is contributing to a greater proportion of people with osteoporotic fractures and the high cost of osteoporosis. Although hip fractures have a significant effect on costs, a greater proportion of the Japanese population experience vertebral fractures. An increase in the incidence of vertebral fractures is concerning because preexisting vertebral fractures in older patients are associated with an increased risk of subsequent fractures. Hence, there is a clear rationale for pharmacological treatment of patients with prevalent vertebral fractures, or for those who are hospitalized or undergo surgery for osteoporotic fractures. Several pharmacological therapies are now available in Japan for the treatment of patients with osteoporosis. Understanding the consequences of subsequent fractures and the treatment options available for patients at high risk of subsequent fractures may contribute to clinical decision-making and improved outcomes for patients with osteoporosis.

Keywords

Drug therapyFracture riskJapanOsteoporosis

Introduction

Osteoporosis is a serious public health issue that is associated with high levels of morbidity and mortality worldwide [1]. Although most individuals with age-related bone loss are asymptomatic, progressive and unchecked bone loss is associated with increased bone fragility and a high risk of bone fracture. Since the late 1980s, several population-based and cohort studies have been conducted to assess the incidence and impact of osteoporotic fractures in Japan. Collectively, these studies have confirmed a significant increase in the incidence of vertebral and hip fractures over the past 20 years in elderly Japanese men and women [25]. Given Japan’s aging population [6], the social and financial impact of osteoporotic fractures is expected to be considerable, suggesting that fracture prevention is of vital importance to public health in Japan.

According to the World Health Organization (WHO), osteoporosis is diagnosed in adults with a femoral neck bone mineral density (BMD) that is ≤2.5 standard deviations (SD) below the young adult mean (YAM) [7]. In this definition, BMD is used as a proxy measure for bone strength and, therefore, does not take into account the role of bone quality in those individuals who do not meet the BMD criteria for osteoporosis, but who are at risk of subsequent fracture (i.e., adults with fragility fractures and a BMD that is 1.0–2.5 SD below YAM). Although low BMD is a major contributor to fracture risk, other clinical factors including, but not limited to, age, previous vertebral fracture, and multiple prevalent fractures, as well as a family history of fracture, exposure to corticosteroids, rheumatoid arthritis, and type 2 diabetes mellitus contribute to an increased risk of fracture, irrespective of BMD status [710]. As such, the Japanese Guidelines for the Prevention and Treatment of Osteoporosis [11] have taken a more practical approach in terms of early detection of impaired bone strength and have defined osteoporosis in persons with low BMD or in those who are at an increased risk of fracture. The Japanese Guidelines for the Prevention and Treatment of Osteoporosis [11] diagnostic criteria for osteoporosis are—(1) adults with fragility fractures and low BMD (i.e., nontraumatic bone fractures and a BMD <80 % of YAM) or (2) adults without fragility fractures and a BMD <70 % of YAM. Findings from a regional cohort study from Japan in 2005 showed the prevalence of osteoporosis with the Japanese diagnostic criteria (L2–L4) to be between 13.5 and 43.8 % for women >60 years old [12]. This study estimated that 800,000–2,600,000 Japanese men and 5,300,000–8,100,000 Japanese women aged ≥40 years would have osteoporosis at the lumbar spine, femoral neck, or total hip [12].

While many studies have been published on the incidence of fracture in Japanese individuals, few have focused on the risk of subsequent fractures in these patients and little is known about how patients at risk of subsequent fractures are treated. Findings from one population-based study showed that the incidence of a new vertebral fracture in Japanese men with an existing prevalent vertebral fracture was 31.5 per 1,000 person-years for those aged 60–69 years and 64.2 for those aged ≥80 years [13]. For women, the incidence was much higher: 59.7 per 1,000 person-years for those aged 60–69 years and 141 for those aged ≥80 years. During the past two years, a variety of new therapies with differing mechanisms of action (e.g., minodronate, bazedoxifene, teriparatide) have become available in Japan for the treatment of patients with osteoporosis. The objectives of this nonsystematic review are to describe the morbidity, mortality, and risk of fracture in patients who are at high risk of subsequent fracture and the current treatment options available for these patients. A systematic review of osteoporosis medication for preventing fractures in Japanese patients with an increased fracture risk has been published previously [14]. The aim of this review is to focus on the newer pharmacological therapies available for slowing bone loss or stimulating bone formation (e.g., bazedoxifene, minodronate, and teriparatide) that may help reduce the risk of subsequent fractures in Japan. Electronic searches of Medline (1990–2011, with no language restrictions, and using the search terms ‘Japan’ and ‘osteoporosis’ and ‘fracture’ and ‘epidemiology’ or ‘incidence’ or ‘prevalence’; and ‘Japan’ and ‘osteoporosis’ and the generic drug name of each osteoporosis drug available in Japan) and manual searches of the reference lists of retrieved references were conducted to identify primary published manuscripts from Japan.

Morbidity and mortality associated with fracture in Japan

In the past two decades there has been a dramatic increase in the proportion of the Japanese population >65 years, particularly elderly women. In 2011, people aged ≥65 years comprised 23 % of Japan’s total population, which has increased from 18 % in 2001 and 12.6 % in 1991 [1517]. Of the total population aged ≥90 years in 2011 (~1.5 million people), 77 % were women [18]. This dramatic increase in the elderly population in Japan is contributing to a parallel increase in the number of elderly people with osteoporotic fractures [2, 5] and a large increase in the financial and social costs of osteoporosis in Japan [19]. In 2010, fracture was estimated to be the fifth leading cause of the need for daily living support or nursing care in Japan [20]; overall, 67 % of the people in Japan who required daily living support or nursing care were women and 46.7 % were aged from 80–89 years. The annual incidence of osteoporotic fracture at any site has been estimated at between 499.9 and 608.2 per 100,000 person years [4, 21] and is estimated to be 2.8-fold higher in women than in men [21].

Two studies from Japan have shown the costs of osteoporosis to be high [22, 23]. In 1992, findings from a cost of illness study estimated that the total disease burden for osteoporosis and osteoporosis-related fractures in Japan was 1,497.3 billion yen (US dollars [USD] 12.1 billion, November 1992 estimate, 1 USD = 123.81 yen), with direct medical costs accounting for 1,310.2 billion yen (USD 10.6 billion, November 1992 estimate, 1 USD = 123.81 yen), nursing care and assistance accounting for 182.5 billion yen (USD 1.47 billion, November 1992 estimate, 1 USD = 123.81 yen), and lost production accounting for 4.7 billion yen (40 million, November 1992 estimate, 1 USD = 123.81 yen) [23]. More recently in 2002, findings from a cost-effectiveness study of risedronate treatment showed that the total medical costs from hospitals, clinics and long-term care facilities for Japanese women with osteoporosis who received care for fractures was 302.9 billion yen (USD 2,754 million, published value) [22].

In addition to the high financial cost, osteoporosis and osteoporotic fractures are associated with a significantly increased risk of death. Several studies have shown a 1.4- to 2.8-fold increased risk of death in women with osteoporosis, with or without fractures, and in elderly Japanese women with low BMD at the femoral neck that is independent of age or existing comorbidities [2426]. Findings from a prospective observational study of 1,429 postmenopausal women in Japan in from 1993 to 2010 showed that the severity of osteoporosis was associated with a 1.4-fold increase in risk of death [hazard ratio (HR) 1.390, 95 % CI 1.129–1.719] [24]. In this study, the 5-year survival of women with osteoporosis without major fractures was >90 % whereas the 5-year survival in women with osteoporosis and a major osteoporotic fracture was between 70 and 80 %.

Vertebral fractures

Several population-based and cohort studies have estimated the incidence of fracture at various sites in Japanese populations (Table 1). Two similar population-based studies from Sado City were conducted to compare the incidence of fracture at various fracture sites [4, 21]. Findings from these studies showed that of the fracture sites most commonly associated with osteoporosis (e.g., spine, hip, forearm, and proximal humerus), vertebral fractures appear to occur most often in the Japanese population [4, 21]. Between 2004 and 2006, the annual incidence of vertebral fracture in the older Japanese population per 100,000 persons was estimated to be between 232.8 and 282.3 persons [4, 21]. Findings from a cohort study showed that the incidence of vertebral fracture increased with increasing age and was 2-fold higher in women than in men after adjusting for age; however, the gender difference was not significant after adjustment for baseline BMD at the spine and femoral neck, and prevalent vertebral fracture [13]. There is some evidence to suggest that multiple vertebral fractures may occur more frequently among 65−75-year-old Japanese women than in non-native Japanese women or Caucasian women of the same age [27, 28]. This difference in prevalence of vertebral fractures may, in part, be because of the effect of differences in lifestyle and diet on BMD between Japanese women compared with non-native Japanese women [29].
Table 1

Estimated incidence of osteoporotic fracture in Japan

Fracture location

Population assessed

Study type Study dates

1-year incidence per 10,000 persons

References

Various

Sado city

Population-based 2004–2006

All, 60.8; VF, 28.2; hip, 17.7; forearm, 11.1; shoulder, 3.8a

[4]

Various

N = 70,011 (35,593 women), Sado City

Population-based 2004

All, 50.0; VF, 23.3; hip, 12.1; forearm, 10.9; shoulder, 3.7; females, 71.9; males, 25.7a

[21]

Various

N = 8,905 (6,841 women) nursing home residents, mean age 84.3 years, nursing homes in Niigata prefecture

Cohort 2007–2008

All, 347; hip, 137; other, 74; upper arm, 44; femur, 34; VF, 31; forearm, 16; lower leg, 11a

[51]

Various

N = 1,787 postmenopausal women, age 40–79 years, Nagano, Miyama, Taiji

Cohort 1993–2003

VF: Nagano cohort, 249; Miyama cohort, 2, 118 aHip: Nagano cohort, 39; Miyama cohort, 22a

[48]

Various

N = 773 postmenopausal women, age ≥69 years, Muramatsu

Cohort 2003–2009

All, 118; forearm, 45; upper arm, 19; femur, 16; lower leg, 7; VF, 33a

[112]

Hip and Vertebral

N = 2,356 (1,593 women) Hiroshima/Nagasaki atomic bomb survivors, mean age of women 65.4 ± 9.8 years

Cohort 1994–2000

VF: females 70+ years, 400; females 80+ years, 840a

[13]

Hip

N = 2,486,999 (1,278,804 women), Niigata prefecture

Population-based 1999

All subjects, 6.8; males, 3.3; males 65+ years, 16.9; females, 10.1; females 65+ years, 40.6a

[3]

Hip

N = 603,987 (208,582 women), age ≥35 years, Tottori prefecture

Population-based 2004–2006

All, 24.5; males, 9.9; females, 36.8 Residual lifetime risk of hip fracture in subjects aged 50 years: males 5.6 %; females 20.0 %a

[2]

Hip

N = 8,234 orthopedic institutions, 4,500 institutions selected for nationwide assessment

Nationwide study 2007

Males, 5.1; males 70+ years, 18.1; females, 18.1; females 70+ years, 39.7

[5]

Hip

Subjects with new fracture, age ≥50 years, Okinawa prefecture

Population-based 2004

Males, 11.5; females, 45.4a

[34]

Grp group, FN femoral neck, VF vertebral fracture

aIncidence values were adjusted from the original reported value to per 10,000 person-years and rounded to one decimal place for ease of comparison across studies

Multiple vertebral fractures are of great concern as they can result in chronic pain, a marked deterioration in the activities of daily living and quality of life, debilitating spinal deformity (kyphosis), and an increased risk of death (Table 2). Studies conducted in postmenopausal women in Japan have shown that quality of life is negatively correlated with increasing age, the number of vertebral fractures, and degree of spinal deformity, and positively correlated with isometric back extensor strength and spinal mobility [30, 31]. Moreover, older Japanese women with multiple vertebral deformities have significantly impaired activities in daily living, which appears to be independent of age, back pain, or the number of painful joints [32]. One observational study conducted from 1997 to 2007 in Japan in 629 elderly individuals with osteoporosis showed a marked difference in mortality between individuals with and without vertebral fracture [33]. The 10-year survival rate of individuals with vertebral fractures was 69 % compared with 86 % in individuals without vertebral fractures. The factors that were statistically significantly associated with an increased risk of death were advanced age, male gender, the presence of vertebral fractures, and the presence of three or more vertebral fractures. Compared with individuals with no vertebral fractures, individuals with at least one vertebral fracture were 1.7-fold more likely to die (HR 1.72, 95 % CI 1.12–2.65) and those with three or more vertebral fractures were 3.3-fold more likely to die (HR 3.28, 95 % CI 1.77–6.10).
Table 2

Morbidity and mortality of osteoporotic fracture in Japan

Fracture location

Population assessed

Study type Study dates

Outcomes and risk factors

References

Various

Subjects with osteoporotic fracture in Sado City

Prospective population-based 2004

Mean stay in hospital: VF (20.4 days) vs hip (30.5 days)

[21]

Discharge status (% of subjects)

VF: home (87 %), ward/hospital transfer (13 %)

Hip: home (55 %), nursing home (36 %), ward transfer (5 %), death (4 %)

Various

N = 122, postmenopausal women with fracture, age ≥45 years, Tottori prefecture

Prospective cohort 2004–2005

Change in mean EQ-5D (utility) after 1 years: hip −12.9; VF −7.8; wrist −5.8

[40]

NR

N = 1,232, postmenopausal women >50 years, mean age 63.9 ± 10.5 years, Nagano prefecture

Prospective cohort 1993–2007 Mean follow-up, 6.9 ± 3.6 years

Risk factors for all-cause mortality (Cox proportional hazards model; variables: age, low vitamin D, smoking, drinking, cardiovascular events, dementia, malignancy, BMD category)

[26]

Osteoporosis (HR 2.14, 95 % CI 1.22–3.75)

Age (HR 1.73, 95 % CI 1.51–1.98)

Low vitamin D (HR 2.17, 95 % CI 1.27–3.72)

Malignancy (HR 5.60, 95 % CI 3.36–9.31)

Vertebral

N = 584, women, mean age 64.4 ± 9.6 years, Oshima

Prospective cohort 1998/1999

Association with impaired function (difficulty with ≥3 ADLs; multiple logistic regression:

[32]

≥3 vertebral deformities adjusted for age, BMI, joint pain (OR 4.1, 95 % CI 1.8–9.5)

≥3 vertebral deformities adjusted for age, BMI, joint pain, back pain (OR 3.6, 95 % CI 1.5–8.7)

Vertebral

N = 174, postmenopausal women with osteoporosis, mean age 68 years, Hondo, Akita prefecture

Prospective cohort NR

Multiple vertebral fractures significantly correlated (Pearson’s correlation) with JOQOL scores: pain r = −0.217, ADL r = −0.255, posture and figure r = −0.265, recreation and social activity r = −0.195, falls and psychological factors r = −0.154

[30]

Vertebral

N = 157, postmenopausal women with osteoporosis, ≥60 years, Hondo, Akita prefecture

Prospective cohort NR

Multiple vertebral fractures significantly correlated (Pearson’s correlation) with JOQOL scores: pain r = −0.300, ADL r = −0.362, posture and figure r = −0.400, recreation and social activity r = −0.196, falls and psychological factors r = −0.179, general health r = −0.179

[31]

Vertebral

N = 629, subjects with osteoporosis, 66.6 % women, mean age 73 years, Miyagawa

Retrospective cohort 1997–2007 Follow-up, 10 years

10-year survival: with VF 69 %; without VF 86 %

[33]

Risk of mortality (Cox proportional hazards model; variables: VF, age, gender): VF (HR 1.72, 95 % CI 1.12–2.65); male (HR 1.82, 95 % CI 1.22–2.72); age (70s) (HR 0.18, 95 % CI 0.12–0.27)

Hip

N = 480, subjects hospitalized for hip fracture, 82.7 % women, mean age 82.3 ± 7.5 years

Retrospective cohort 1991–1996

1-year survival, 88.5 %

[45]

Significant prognostic factors (Cox proportional hazards models adjusted for age) for mortality following hip fracture: 1 comorbidity (HR 1.90, 95 % CI 1.50–2.42); 2 comorbidities (HR 3.65, 95 % CI 2.71–4.93)

Hip

N = 402,760, subjects (317,314 women), hospitalized for new hip fracture, ≥35 years

Prospective nationwide 2001–2008

Fall was major cause (77.7 % of subjects); 45 % of fractures were at the FN

[38]

Mean time from hospitalization to surgery 4.9–5.6 days

Mean time from surgery to discharge 40.7–53.4 days

Hip

N = 10,992, subjects with surgery for fracture, mean age 81.8 years

Retrospective cohort 1999–2001

Fall was major cause (76.1 % of subjects); 41 % of fractures were at the FN

[39]

1-year postoperative mortality rate, 10.1 %

Mean time from fracture to admission 3.1 ± 16.0 days; from admission to surgery 11.2 ± 34.0 days; from surgery to discharge 49.8 ± 42.0 days

Most subjects (92 %) had preoperative complications (e.g., hypertension, dementia, neuropathy, heart disease)

ADL activities of daily living, BMD bone mineral density, CI confidence interval, EQ-5D EuroQuol health-related quality of life instrument, FN femoral neck, HR hazard ratio, JOQOL Japanese osteoporosis quality of life questionnaire, NR not reported, OR odds ratio, VF vertebral fracture

Hip fractures

At least four population-based studies have been conducted in Japan to assess the incidence of osteoporotic fracture (Table 1). Of these, three studies have shown a significant increase in the incidence of hip fracture over the past two decades in older Japanese women (≥50 years) [2, 5, 34]. For example, findings from a nationwide study showed that the estimated annual incidence of hip fracture in women of any age per 10,000 persons was 9.20 in 1992 and 18.14 in 2007 [5]. In 2007, the annual incidence of hip fracture per 10,000 persons was estimated to be 39.71 in women ≥70 years and 313.58 in women ≥90 years [5]. In comparison, the annual incidence of hip fracture per 10,000 persons was 18.12 in men ≥70 years and 146.62 in men ≥90 years. The higher incidence of hip fracture in older Japanese populations is likely to reflect the time course of osteoporotic disease, with higher levels of cortical bone loss occurring later in the disease process [35].

In contrast to vertebral fracture, there appears to be a clear difference in the incidence of hip fracture between Asian and Caucasian populations [27]. Estimates of the relative probability of hip fracture averaged for age and gender have shown that the 10-year probability of hip fracture in Japan (0.39) is slightly higher than in China (0.29) or Korea (0.18), but is considerably lower than in the United States (0.78) and many European countries (0.57–1.24) [36]. Moreover, population-based studies that compare the incidence of fracture type suggest that the incidence of hip fracture is approximately 2-fold lower than vertebral fracture in Japanese people with osteoporosis [4, 21]. Given that there appears to be little difference in proximal femur BMD between Asian and Caucasian populations, other factors such as diet and lifestyle, body size, hip geometry, and differences in bone quality may contribute to the differences in the incidence of hip fracture between Japanese and Caucasian populations [27, 37].

Although fractures at any site are associated with significant morbidity and mortality, fractures at the hip have the greatest impact on public health [7]. Several population-based and cohort studies have been conducted to assess the morbidity and mortality of hip fracture in Japan (Table 2). Findings from these studies have shown that Japanese patients who are hospitalized for hip fracture (mean 30.5–49.8 days) [21, 38, 39] have longer hospital stays than patients hospitalized for vertebral fracture (mean 20.4 days) [21]. Fewer patients with hip fracture than vertebral fracture are discharged home (55 vs 87 %), with 36 % of patients with hip fracture being discharged to a nursing home [21]. Patients who have undergone surgery for hip fracture have a significant decrease in their independence in activities associated with daily living one year after discharge compared with before surgery and more severe loss of health-related quality of life compared with patients with wrist or vertebral fractures (Table 2). In one study, assessment of 10,992 hip fractures in patients who underwent surgery for hip fracture between 1999 and 2001 showed that 50.9 % of patients could catch public transport and visit their neighbors before surgery compared with 26.8 % after surgery [39]. Sequential analysis of the quality of life (EQ-5D) of postmenopausal women with hip fracture between 2004 and 2005 has shown that approximately 50 % of patients with hip fracture still report ‘some problem’ with their mobility, usual activities, and pain/discomfort, one year after fracture [40].

The mean length of hospital stay after hip surgery is considerably longer in Japan (30.5–49.8 days) [21, 38, 39] compared with the United States (6.2–10.5 days) [41, 42]. This difference in the length of stay is not due to slower recovery times, but is more likely to be a result of the reimbursement system in Japan, which covers a large part of the medical costs of hospitalization. Hence, patients who are hospitalized for hip fracture in Japan are more likely to receive rehabilitation and treatment for comorbidities before being discharged [41, 43]. Although patients with osteoporosis who are considered to be of low activity status before fracture are less likely to walk independently after fracture, patients considered to be of high activity status before fracture are more likely to walk independently if they receive appropriate treatment [44]. However, longer hospitalization, particularly for those who are bedridden, may place patients at an increased risk of hospital-acquired conditions and mortality. Between 1991 and 2001, the 1-year postoperative mortality rate for patients who had been hospitalized for hip fracture was estimated at approximately 10 % [39]. In this study, the 1-year survival rate for elderly patients who have undergone surgery for hip fracture was estimated to be between 91.5 and 94.9 % for patients in their 70s, 80.6 and 89.9 % for patients in their 80s, and between 81.8 and 83.3 % for patients in their 90s [39]. The factors found to be associated with an increased risk of death following hospitalization for hip fracture are male gender, trochanteric fracture, impaired ambulatory status before injury, dementia, diabetes mellitus, and a history of gastrectomy or colectomy [45]. Overall, patients with two comorbidities are estimated to have a 3.7-fold increased risk of death (HR 3.65, 95 % CI 2.71–4.93) compared with patients with no comorbidities [45].

Risk of subsequent fracture

The existence of a previous fracture confers an approximately 2-fold higher risk of subsequent fracture [8, 46] and is a major contributor to the risk of subsequent fractures in postmenopausal women in Japan (Table 3). Thus, given that the total estimated population aged >65 years was 29 million in 2010, and is projected to be 36 million in 2020 and 42 million by 2040 [18], the increased risk of a subsequent fracture is likely to affect a significant proportion of the older Japanese population.
Table 3

Risk of subsequent fracture in Japanese subjects with osteoporosis

Fracture location

Population (BMD, g/cm2)

Study type Study dates

Incidence of subsequent fracture

Outcomes

References

Vertebral

N = 712, 100 % women, mean age 64.1 ± 8.2 years, Japanese Population-based Osteoporosis (JPOS) Cohort Study [lumbar spine 0.81; total hip 0.75; FN 0.65]

Prospective population-based 1996 Follow-up 6 years

10.3 % of women with VF

Risk of subsequent VF in women with prevalent VF (Poisson regression analysis adjusted by propensity scorea):

[50]

all ages (RR 3.0, 95 % CI 1.8–4.9

50–59 years (RR 7.2, 95 % CI 1.0–9.6); 60–69 years (RR 3.2, 95 % CI 1.3–8.0); 70–79 years (RR 2.3, 95 % CI 1.3–4.1)

Vertebral and hip

N = 2,356, 67.6 % women, mean age 65 ± 9.8 years for women, Hiroshima / Nagasaki atomic bomb survivors [men: spine 0.98, FN 0.73; women: spine 0.82, FN 0.62]

Prospective population-based 1994–2000 Mean follow-up 4 years

Men 80+ years 64.2 VF per 1,000 person-years Women 80+ years 141 VF per 1,000 person-years

Risk of subsequent VF in subjects with prevalent VF (Cox regression models):

[13]

adjusted for age and vertebral BMD: men (RR 4.4, 95 % CI 1.5–13.5); women (RR 2.94, 95 % CI 2.0–4.3)

Adjusted for FN BMD: men (RR 4.5, 95 % CI 1.6–13.0); women (RR 3.2, 95 % CI 2.2–4.6)

Risk of subsequent hip fracture in women with prevalent VF (Cox regression models):

Adjusted for age and spine BMD (RR 5.16, 95 % CI 1.94–14.42)

Adjusted for age and FN BMD (RR 2.91, 95 % CI 1.10–7.90)

Hip

N = 384, subjects treated for fracture, 85.7 % women, mean age 85 years, Shimizu Hospital [NR]

Prospective cohort 2001–2007 Mean follow-up 3.0 ± 1.4 years

43 per 1,000 person-years

Comorbidities increased risk of subsequent fracture (Cox proportional hazards model including age, gender, respiratory disease, dementia): respiratory disease (HR 4.41, 95 % CI 2.33–8.34); dementia (HR 1.87, 95 % CI 1.02–3.41)

[52]

Other outcomes: median time between fractures, 21 months; incidence of fracture was greatest within the first year (40.8 % of subjects); mean length of rehabilitation, 71.3 ± 33.5 days

Hip

N = 714, subjects treated for fracture, 81.8 % women, mean age 80.7 years [NR]

Prospective cohort 1996–2001 Mean follow-up 2.4 ± 1.4 years

29 per 1,000 person-years

Comorbidities increased risk of subsequent fracture (logistic regression adjusted for age and gender): Parkinson disease (OR 3.27, 95 % CI 1.28–8.35); dementia (OR 3.07, 95 % CI 1.58–5.96)

[53]

Other outcomes: fall was the major reason for subsequent fracture (95.6 % of subjects); >70 % of subsequent fractures were the same type (r = 0.508, P < 0.001); mean length of hospital stay, 59.9 ± 24.4 days; most (71.1 %) subjects regained prefracture ambulatory status

Hip

N = 1,916, 89.7 % women, ≥35 years

Nationwide survey 2001–2004 Follow-up 4 years

NR

67.4 % had the same type of fracture (FN vs trochanteric; 2.3 % had the subsequent fracture on the same side (left vs right)

[38]

Hip

N = 2,663, women with a hip fracture, ≥65 years [60.1 ± 15.2 % YAM]

Prospective cohort 2006–2007 Follow-up 1 years

70 per 1,000 person-years, any fracture 34 per 1,000 person-years, hip fracture

Women ≥65 years with an initial hip fracture were 4-times more likely than the general population to sustain an additional hip fracture

[49]

BMD bone mineral density, CI confidence interval, FN femoral neck, NR not reported, OR odds ratio, RR relative risk, VF vertebral fracture

aPropensity scores were the probability estimated by logistic regression equation for prevalent vertebral deformities comprising age, weight, total hip BMD T-score, menopausal status, nonvertebral fracture history, bilateral oophorectomy, diabetes mellitus, hypertension, hyperlipidemia, and osteoporosis, and with lifestyle factors [physically active lifestyle, coffee consumption (≥1 cup/day)] as explanatory variables

The factors that have been found to be significantly associated with an increased risk of hip or vertebral fracture in Japanese populations are increasing age, older women, body weight/body mass index, previous fracture of any type, previous vertebral fracture, low BMD at the spine or hip, and regular alcohol intake (≥5 times per week) [13, 47, 48]. Men or women >50 years of age with a 1 SD decrease in BMD at the spine or femoral neck have a 1.58- to 1.78-fold higher risk of vertebral fracture and a 1.00- to 5.49-fold higher risk of hip fracture [13]. Recently, Hagino et al. [49] demonstrated that the incidence of all subsequent fractures among patients who sustained a first hip fracture was 70 per 1,000 person-years and that women >65 years who sustained a subsequent hip fracture had a 4-fold higher risk of a subsequent fracture than the general population. In addition, the risk of subsequent fracture is estimated to be approximately 3-fold higher in women and 4.5-fold higher in men with prevalent vertebral fractures [13, 50]. The incidence of a subsequent vertebral fracture in women in their 50s and 70s has been estimated at 36.2 and 88.0 per 1,000 person-years and in men in their 60s and 70s at 31.5 and 44.8 per 1,000 person-years [13]. Findings from population-based studies also suggest that the risk of subsequent fracture declines in women after the age of 70 years, so that middle-aged women (50–59 years) with previous fractures have a higher risk of subsequent fracture than older women (70–79 years), even after adjusting for additional risk factors known to be associated with fracture [13, 50]. Although there may be many reasons why older women are at a lower risk of subsequent fracture, competing risks such as increased mortality may be a contributing factor. In addition, reduced mobility in this older population may also contribute to a lower risk of fracture. For example, the risk of fracture incidence at any site in nursing home residents in Japan is estimated to be lower for immobile versus mobile residents (HR 0.34, 95 % CI 0.23–0.50) or for totally dependent versus partially mobile or independent residents (HR 0.49, 95 % CI 0.38–0.64) [51].

Although prevalent vertebral fracture is a major contributor to the risk of subsequent fracture in Japan, individuals with previous hip fracture are also at higher risk of a subsequent fracture. At least two studies conducted in Japan have estimated the annualized incidence of subsequent hip fracture in individuals with a previous hip fracture [52, 53]. The study conducted by Mitani et al. from 2001 to 2007 of 384 patients with hip fracture estimated the incidence of subsequent hip fracture to be 0.043 per person-years [52] and the study conducted by Yamanashi et al. [53] from 1996 to 2001 of 714 patients with hip fracture estimated the incidence of subsequent hip fracture to be 0.029 per person-years. The individuals in these studies were mostly women (>80 %) and had a mean age of >80 years. Many individuals (40.8–44 %) experienced a subsequent fracture within the first year of the original fracture [52, 53] and the median time between the first and second hip fractures in one study was 21 months [52]. The mean length of hospital stay or rehabilitation after subsequent fracture was 59.9–71.3 days and did not differ greatly compared with the first fracture [52, 53]; most (71.1 %) individuals regained their prefracture ambulatory status [53]. As for first fracture, the main reason for a subsequent hip fracture was a simple fall [53]. Both studies found the presence of comorbidity to be significantly associated with subsequent fracture. Respiratory disease, dementia, and Parkinson disease were significantly associated with subsequent fractures whereas neurological disease, diabetes mellitus, hypertension, cardiac disease, and rheumatoid arthritis were not (Table 3).

There is increasing recognition of the importance of prevalent vertebral fractures for predicting the risk of subsequent fractures. A recent analysis of the relative usefulness of the clinical risk factors for fracture suggests that, in the Japanese population at least, a model that combines age, BMD, previous clinical fracture, and the presence or absence of vertebral fracture can predict future fracture risk with greater simplicity and prognostic accuracy than the WHO fracture risk assessment tool (FRAX) [54]. This finding confirms that vertebral fracture status is an important indicator of bone strength, particularly in individuals with normal levels of BMD. Although BMD has been considered historically to be the predominant factor that determines bone quality and strength, bone architecture, turnover, mineralization, and microdamage also contribute in varying degrees to bone quality and strength. Indeed poor bone quality, characterized by low bone mineralization and detrimental changes to bone microarchitecture, has been observed in bone biopsies from patients with osteoporosis and hip fracture [55] and is thought to contribute to the increased risk of fracture in patients with type 2 diabetes mellitus, rheumatoid arthritis, or long-term corticosteroid use [9, 56].

Pharmacological treatment for prevention of subsequent fractures

The Japanese Guidelines for the Prevention and Treatment of Osteoporosis recommend that the main goals when managing patients with osteoporosis are to control the risk of fracture and to maintain or improve quality of life. To reduce the risk of fracture, individuals should maintain a balanced diet that is rich in calcium, vitamin D, vitamin K, and protein, and avoid smoking and excessive alcohol intake [11]. In addition, the use of hip protectors [57] and preventative lifestyle strategies that aim to maintain spinal mobility and general physical health are important for preventing subsequent fractures, and in reducing the significant morbidity and mortality associated with osteoporosis. However, the role of pharmacotherapy in the treatment of osteoporosis and in slowing bone loss or stimulating bone formation is increasingly recognized as of equal importance by Japanese orthopedists [58]. Despite this, some studies have shown that many patients who are hospitalized or undergo surgery for osteoporotic fractures in Japan are untreated before or after surgery and many patients who are identified as being at high risk of fracture are untreated [21, 59, 60]. One observational study of 422 patients who underwent surgery for osteoporotic fracture at 4 hospitals in Hokkaido Prefecture found that 87 % of patients were untreated before or after surgery and very few (4 %) commenced medication after surgery [59]. Of 723 Japanese individuals with rheumatoid arthritis who were assessed as being at high risk of fracture, only 63 % were found to be taking osteoporosis medication [60]. This underutilization of osteoporosis medication is not confined to Japan and is of concern, particularly for patients with fragility fractures who are very likely to benefit from treatment that can reduce the risk of subsequent fractures [6164].

Currently in Japan, there are two main pharmacological strategies for the treatment of patients with osteoporosis—(1) antiresorptive drugs, which inhibit osteoclastic activity to slow bone turnover, and maintain or increase bone mass and (2) anabolic drugs, which increase osteoblastic activity over osteoclastic activity to result in an overall stimulation of bone formation. Antiresorptive drugs currently approved and recommended by the Japanese Guidelines for the Prevention and Treatment of Osteoporosis for patients with osteoporosis include bisphosphonates (e.g., alendronate [6567], etidronate [68], risedronate [69, 70]), selective estrogen receptor modulators (e.g., raloxifene [71]), and active vitamin D3 derivatives (e.g., alfacalcidol [72, 73], eldecalcitol [74]). During the past two years a variety of new drugs have been approved for treatment of osteoporosis in Japan—minodronate [75, 76] (a bisphosphonate), bazedoxifene [77] (a selective estrogen receptor modulator), and teriparatide [78] (a parathyroid hormone analog). Currently, teriparatide is the only anabolic drug approved for treatment of osteoporosis in Japan.

Bisphosphonates are the most recognized and well-studied pharmacological treatment for osteoporosis in Japan [11, 58]. These drugs slow bone breakdown by inhibiting bone resorption via osteoclasts, which leads to a reduction in bone turnover, increased bone mass and mineralization, and improved trabecular microarchitecture [79]. Randomized and observational studies conducted in Japan for up to 7 years have shown that oral treatment with alendronate 5 mg/day, either as monotherapy or combined with alfacalcidol, results in statistically significant increases in BMD, a return to normal biomarker levels, and a marked reduction in the risk of vertebral and nonvertebral fractures in men and women at high risk of fracture [80, 81]. Other bisphosphonates that have been investigated in Japanese subjects include risedronate, etidronate, and minodronate [8285]. Minodronate, which is administered as a daily (1 mg/day) or monthly (50 mg/4 weeks) tablet is the most recent bisphosphonate to receive marketing approval for treatment of osteoporosis in Japan. Clinical studies conducted in postmenopausal women in Japan have shown that minodronate (1 mg/day) reduces the risk of vertebral fracture by 59 % (95 % CI, 36.6–73.3) and results in significant reductions in bone turnover markers compared with placebo after 2 years of treatment [85], and provides improvements in lumbar and hip BMD that are comparable to the changes seen with alendronate [83]. After 12 months of treatment, there was a 5.86 % (95 % CI, 5.22–6.50) and 6.29 % (95 % CI, 5.70–6.88) increase in lumbar spine BMD from baseline and a 3.47 % (95 % CI 2.83–4.11) and 3.27 % (95 % CI 2.60–3.95) increase in total hip BMD from baseline in the minodronate (1 mg/day)- and alendronate (5 mg/day)-treated subjects, respectively [83]. Although the acceptable safety profile of bisphosphonates is well established, concerns persist that prolonged therapy may lead to a failure to repair microdamage, resulting in accumulated damage and atypical bone fragility [8688]. As such, the American Society for Bone and Mineral Research has appointed a task force to investigate a possible association between atypical fractures of the femur and long-term use of bisphosphonates [89]. The task force has recommended that clinicians and patients should be made aware of the potential for an increased risk of atypical fractures and has called for increased surveillance and clinical data to help elucidate this issue further. However, in the absence of confirmatory evidence, the benefits of prolonged suppression of bone remodeling are considered to outweigh the risks of the potential for increased bone fragility with continued treatment [88].

Selective estrogen receptor modulators have estrogen antagonist or agonist activity, depending on the tissue site [90]. Raloxifene, a benzothiophene derivative, is a nonhormonal drug with estrogen agonist activity in bone and estrogen antagonist activity in breast tissue. Although raloxifene is an effective agent for improving bone density and the frequency of vertebral fracture [9194], it has not been shown to reduce the risk of nonvertebral fractures in patients versus placebo [95]. In a three-year clinical trial, the risk of fracture for treatment with raloxifene (60 mg/day) compared with placebo was 0.7 (95 % CI 0.5–0.8; 60 mg/mL) for new vertebral fractures and for treatment with raloxifene (60 mg/day or 120 mg/day) was 0.9 (95 % CI 0.8–1.1) for nonvertebral fractures [95]. Bazedoxifene is an indole-based selective estrogen receptor modulator that was approved for use in Japan in 2010. Clinical trials conducted in Caucasian populations have shown that bazedoxifene prevents bone loss and reduces the frequency of vertebral fractures in postmenopausal women [96, 97]. Although information in Japanese patients is limited, findings from a two-year randomized study in Japan have shown that bazedoxifene significantly improves lumbar spine BMD from baseline [20 mg/day, 2.43 % (95 % CI 1.75–3.12) and 40 mg/day, 2.74 % (95 % CI 2.04–3.45)] compared with placebo [−0.65 % (95 % CI −1.35 to 0.05 %), P < 0.001] and reduces bone turnover markers compared with placebo in postmenopausal women after one year of treatment [98].

Overall, selective estrogen receptor modulators are well tolerated. The side-effects most commonly associated with these therapies in the short term include vasodilatation, leg cramps, and fluid retention and, more rarely, venous thromboembolic events [97]. However, venous thromboembolic events may be of concern with raloxifene; findings from a three-year, placebo-controlled trial showed a 3.1-fold (95 % CI 1.5–6.2) increased risk of deep vein thrombosis and a 4.5-fold (94 % CI: 1.1–19.5) increased risk of pulmonary embolism with raloxifene (60 mg/day or 120 mg/day) compared with placebo [99]. Hence, more clinical trials need to be conducted to determine whether the favorable safety profile of selective estrogen receptor modulators, including bazedoxifene, is maintained after long-term treatment in Japan.

Teriparatide, which is a parathyroid hormone analog, reverses the structural damage caused by osteoporosis by affecting both bone formation and resorption during the course of therapy [100]. Treatment with teriparatide increases bone turnover and bone quality by stimulating bone formation earlier in the treatment regimen, before bone resorption occurs. In Japan, teriparatide is indicated for the treatment of patients with osteoporosis who are at a high risk of fracture (i.e., adults with low BMD, existing fracture, old age, or a family history of femoral neck or femur fracture) but should not be prescribed to patients with hypercalcemia or to patients who are at risk of osteosarcoma [78]. Findings from a randomized study conducted in a Caucasian population of postmenopausal women with previous fractures showed that once-daily injection of teriparatide (20 μg) for up to 2 years resulted in significant increases in bone density, a 65 % (95 % CI 45–78) lower risk of vertebral fractures, and a 53 % (95 % CI 45–75) lower risk of nonvertebral fractures defined by semiquantitative readings [101]. An 84 % (95 % CI 67–99) reduction in risk of vertebral fractures was determined when fractures were defined by quantitative morphometry [102]. A follow-up study of 1,262 women who had received 19 months of treatment with teriparatide also suggests that the reduction in risk of nonvertebral fragility fractures is sustained for up to 30 months after treatment discontinuation [103]. In addition, teriparatide has been found to be well tolerated and effective in increasing the levels of bone formation markers in institutionalized adults with severe developmental disabilities and osteoporosis for up to 2 years of therapy [104]. In Japan, findings from a 1-year randomized study have shown that teriparatide treatment results in significant improvements in markers of bone formation and in bone density at the spine, femoral neck, and hip [105]. The side-effects most commonly associated with teriparatide include nausea, headache, dizziness, muscle spasms, joint pain, and high levels of uric acid [78]. In addition, an increased incidence of osteosarcoma in rats after 2 years of exposure to teriparatide [106] has led to a ‘black box warning’ for osteosarcoma with use of teriparatide by the United States Food and Drug Administration [107]. As such, a 15-year safety surveillance program is being conducted to monitor the incidence of osteosarcoma in patients treated with teriparatide in the United States [108] and a company-sponsored safety monitoring program is being conducted worldwide [109]. To date, no causal association between teriparatide and osteosarcoma has been detected from these surveillance studies [108110].

Given the different mechanisms of action of bisphosphonates and teriparatide, it has been suggested that the prolonged suppression of bone remodeling that occurs with bisphosphonates may dampen the anabolic effects of teriparatide if treatment is commenced too early after switching from a bisphosphonate [111]. However, findings from a recent randomized study have shown that the changes in markers of bone formation and BMD with teriparatide are not greatly different between treatment-naive patients and patients with previous bisphosphonate therapy, irrespective of the length of washout (<2 months, 2–6 months, >6 months) [111].

Conclusion

Osteoporotic fracture in Japan is associated with high morbidity and mortality that results in significant financial and social costs to the individual, communities, and nation. Of the factors associated with high risk of fracture (low BMD, prevalent vertebral fracture, previous fracture of any type, age, etc), previous vertebral fractures impose the greatest risk of subsequent fractures. Clinicians should regard individuals with preexisting fractures as being at a significantly higher risk of subsequent fracture, irrespective of bone density. Hence there is a sound rationale for ensuring that patients in Japan who are diagnosed on the basis of prevalent vertebral fracture, or who are hospitalized or undergo surgery for osteoporotic fractures, receive the treatment necessary to reduce the risk of a subsequent fracture. Although nonpharmacological strategies are important for preventing the onset and progression of osteoporosis, pharmacological treatment can help to reduce the risk of fracture in patients who are at high risk of subsequent fractures. A variety of pharmacological therapies, with differing mechanisms of action, are now available for the prevention and treatment of patients with osteoporosis in Japan. Understanding the consequences of subsequent fractures and the treatment options available for patients at a high risk of subsequent fractures may better assist clinical decision-making and improve health outcomes f or the Japanese population.

Acknowledgments

The study was funded by Eli Lilly and Company, manufacturer of raloxifene and teriparatide. Medical writing services were provided by Serina Stretton, PhD, and Mark Woolley, PhD, of ProScribe Medical Communications and were funded by Eli Lilly Japan K. K. All authors are employees of Eli Lilly and Company or Eli Lilly Japan K. K and were involved in the development of the concept and study design (JAF, SF, SN, RB), publication collection (SN), and in the interpretation of the findings from the review (all authors), and in critical revision of the review and in the final approval for publication (all authors).

Conflicts of interest

Shuko Nojiri, MSc PhD, is an employee of Lilly Japan K.K. Russel T. Burge, MA PhD, is an employee of Eli Lilly and Company and an Eli Lilly shareholder. Jennifer A. Flynn, MSPH, is an employee of Lilly Japan K.K. and an Eli Lilly shareholder. Shonda A. Foster, PharmD MS, is an employee of Eli Lilly and Company and an Eli Lilly shareholder. Hideaki Sowa, MD PhD, is an employee of Lilly Japan K.K.

Copyright information

© The Japanese Society for Bone and Mineral Research and Springer Japan 2013