Dr Eisman received a grant from GE-Lunar. All other authors have no conflict of interest.
Incidence of Hip and Other Osteoporotic Fractures in Elderly Men and Women: Dubbo Osteoporosis Epidemiology Study†
Version of Record online: 5 JAN 2004
Copyright © 2004 ASBMR
Journal of Bone and Mineral Research
Volume 19, Issue 4, pages 532–536, April 2004
How to Cite
Chang, K. P., Center, J. R., Nguyen, T. V. and Eisman, J. A. (2004), Incidence of Hip and Other Osteoporotic Fractures in Elderly Men and Women: Dubbo Osteoporosis Epidemiology Study. J Bone Miner Res, 19: 532–536. doi: 10.1359/JBMR.040109
- Issue online: 2 DEC 2009
- Version of Record online: 5 JAN 2004
- Manuscript Accepted: 20 NOV 2003
- Manuscript Revised: 29 OCT 2003
- Manuscript Received: 29 APR 2003
- hip fracture;
In this prospective 12-year study in men and women 60 years of age and older, there was a 4–6% per year reduction in the incidence rate of overall osteoporotic fractures, but the study was unable to exclude any change in the hip fracture incidence rate. Approximately one-half of hip fractures occurred before 80 years in men and two-thirds before 85 years in women. The age distribution of hip fractures underlines the need for earlier intervention in osteoporosis.
Introduction: Although hip fracture is the major osteoporotic fracture in terms of health outcomes, quality of life, and costs, there is a paucity of long-term data on secular changes in men and women within a defined community. This long-term prospective population-based study over 12 years from 1989 to 2000 specifically examined the age distribution and secular changes in the incidence rates of hip and other osteoporotic fractures in men and women 60 years of age and older in a predominantly white population in Dubbo, Australia.
Materials and Methods: Hip and all other clinical fractures were ascertained by reviewing all radiography reports from the two area radiology services, ensuring complete ascertainment of all clinical osteoporotic fractures.
Results and Conclusion: Among the 1055 symptomatic atraumatic fractures (after excluding pathological fractures), there was a significant reduction in the overall fracture incidence rate in women (4% per year; p = 0.0003) and men (6% per year; p = 0.0004) over the 12 years. There were 229 hip fractures (175 in women and 54 in men) within 39,357 person-years of observation. The overall rate ± SE of hip fracture was 759 ± 57 per 100,000 person-years in women and 329 ± 45 per 100,000 person-years in men, with an exponential increase with age. With advancing age, the incidence rate of hip fractures in men approached that in women; the female:male ratio fell from 4.5 (95% CI: 1.3–15.7) to 1.5 (0.9–2.5) and 1.9 (1.2–2.8) in the 60–69, 70–79, and 80+ year age groups, respectively. In women, the absolute number of fractures and incidence rate continuously increased with age; however, in men, the absolute number of hip fractures peaked at 80–84 years of age and then decreased. Most importantly, despite the continuing increase with age, almost one-half (48%) of the hip fractures occurred before the age of 80 years in men, and 66% of hip fractures occurred before the age of 85 years in women. The overall hip fracture incidence is comparable with other white (except Sweden) and Asian groups as well as two other Australian studies. This study could not exclude a change in hip fracture incidence rate, even in those 80 years of age and over among whom the incidence of hip fractures was the highest. The incidence data highlight the fact that a large proportion of hip fractures occurs in those under 80 years of age, particularly in men. This age distribution underlines the need for earlier intervention in osteoporosis in women and particularly in men to achieve the most cost-effective outcomes.
OSTEOPOROSIS AND ITS consequence of low trauma fracture represent a major health burden in aging populations, because it results in increased morbidity, mortality, and high health care costs. Hip fracture is arguably the most costly fracture; thus, a reliable estimate of their present and likely future incidence, in both sexes, is important to help project the costs and resources needed to manage this problem. With the rapidly aging population, the absolute number of hip fractures is expected to increase significantly over the next few decades. Indeed, it has been estimated that the total number of hip fractures worldwide will increase from 1.26 million in 1990 to 2.6 million by the year 2025 and to 4.5 million by the year 2050.(1)
The projection could be compromised if the age-adjusted hip fracture incidence was increasing, as has been suggested for successive birth cohorts for both men and women.(1-6) Projections that fail to account for such changes in rates could over- or underestimate the future public health impact of hip fractures. Reliable data on the long-term incidence of hip fractures are therefore required to improve the worldwide epidemiology of hip fractures.
The aim of this study was to assess osteoporotic fracture, particularly hip fracture, rates in a stable Western urban population with long-term follow-up to allow estimation of any secular/cohort changes over three successive 4-year birth cohorts.
MATERIALS AND METHODS
Dubbo is a semi-urban city 400 km northwest of Sydney, Australia, with a population of ∼32,000. The population is representative of the Australian population, and over 98% are white. It is a relatively isolated city with its own health service, and it has a stable population, making it an ideal city for epidemiological studies.(7)
The Dubbo Osteoporosis Epidemiology Study (DOES) is an ongoing longitudinal population-based study of fracture incidence in men and women 60 years of age and older that commenced in 1989, and it has been described earlier.(8, 9) As part of this study, all radiography reports from two area radiology services that service the entire Dubbo population were reviewed, ensuring complete ascertainment of all clinical fractures in Dubbo. Hip fractures in both men and women 60 years of age and older were identified, and only minimal trauma fractures (fractures caused by falls from a standing height or less) were included. Persons with an underlying disorder, particularly cancer (three subjects) or Paget's disease (five subjects), which might result in a pathological hip fracture, were excluded. The circumstances of the fall and fracture were obtained from personal interview and review of the radiology request and report.
Fracture incidence rates and their associated SE were calculated as the ratio of the number of hip fracture cases that occurred over the 12 years of analysis over the entire Dubbo population at risk. The population at risk was obtained from population census of the Australian Bureau of Statistics.(10) Because of the prospective nature of the study, the denominator in the calculation of incidence rate was the number of person-years. It was assumed that the observed fracture incidence followed a Poisson distribution, in which the expected value is equal to the variance. Furthermore, to assess the secular change in fracture incidence over the study period, a Poisson regression model was fitted to the observed incidence rate. In this model, time was considered an independent variable, whereas age and sex were covariates. The Poisson regression analysis was performed with PROC GENMOD of the SAS v8 statistical package.(11)
Number of fractures and incidence
From April 1989 to December 2000, a total of 39,357 person-years (22,884 in women and 16,473 in men) were observed. During the period, 1055 individuals (786 women and 269 men) had sustained at least one symptomatic osteoporotic fracture. Of these, 229 were hip fractures (175 in women and 54 in men). Hip fractures represented 21.7% of total symptomatic fractures (22% in women and 20% in men).
The overall hip fracture incidence rate in women was 759 (95% CI: 647-871) per 100,000 person-years, which was significantly higher (p < 0.001) than in men, whose incidence rate was 329 (95% CI: 241-417) per 100,000 person-years. There was an exponential increase in hip fracture incidence in both sexes with increasing age, from 116 (women) and 0 (men) per 100,000 person-years in the 60-64 year age group to 2597 (women) and 1187 (men) per 100,000 person-years in the 85+ year age group (Fig. 1C).
Although the incidence rate of hip fractures increased with advancing age, the absolute number of hip fractures peaked in men at 80-84 years of age and then fell (Fig. 1A). Approximately one-half (48%) of hip fractures in men occurred before the age of 80 years. Although the number of hip fractures in women continuously increased with advancing age, 66% of hip fractures occurred before the age of 85 (Fig. 1B).
The overall female:male hip fracture incidence ratio was 2.3:1, but this ratio decreased with age such that the incidence of hip fractures in men approached that in women. The ratio fell from 4.5 (95% CI:1.3-15.7) in the 60-69 year age group to 1.5 (0.9-2.5) and 1.9 (1.2-2.8) in the 70-79 year and the 80+ year age groups, respectively.
The incidence of hip fractures in women and men in the Dubbo group was comparable with a range of white(12-14) and Asian populations.(2, 15) However, the age-specific incidence rate in Dubbo was statistically lower (p < 0.01) than that in Malmo, Sweden.(12) There was no significant difference in the hip fracture incidence in Dubbo and two other Australian populations in Geelong(16) and Canberra.(17) However, the incidence of hip fracture in women observed in Dubbo was significantly higher than that in Southern Tasmania (an Australian State),(18) but the difference was not significant in men (Fig. 2).
On average, the incidence rate of total osteoporotic fractures decreased by 4% per year in women (p = 0.0003) and by 6% per year in men (p = 0.0004; Fig. 3). For example, the osteoporotic fracture incidence in the 80-84 year old men decreased from 4251 to 3094 to 2473 per 100,000 person-years in the three successive 4-year periods from 1989 to 2000. Moreover, in the same three successive 4-year periods, a similar decrease from 6334 and 6985 to 4059 per 100,000 person-years was observed in the 80- to 84-year-old women. This decrease in total osteoporotic fracture incidence was not attributable to any one fracture type. There was no statistically significant change in the incidence rate of hip fracture in either women (p = 0.14) or men (p = 0.71; Fig. 4), although the number of hip fractures observed in this study was relatively low, and therefore the study would be underpowered to show such a change. Even in the oldest age groups i.e., the 80+ year group, where the hip fracture incidence was the highest, there was no significant change in hip fracture incidence rate in either women (p = 0.24) or men (p = 0.13).
In this relatively stable and well-characterized community population who had been followed-up for 12 years, it was found that one in four of all hip fractures occurred in men. Moreover, with increasing age, the incidence rate of hip fractures in men approaches that in women; after the age of 70 years, the female:male ratio was 1.9:1 or lower. In absolute terms, the incidence of hip fractures peaked in the 80-84 year age group for men before declining, whereas the incidence in women continuously increased with advancing age. Hence, the burden of hip fractures in men occurs earlier (80-84 years) than in women (85+ years) because of the shorter life expectancy in men. Indeed, almost one-half (48%) of the hip fractures occurred before the age of 80 years in men and 66% of hip fractures occurred before the age of 85 years in women. It may be intuitively clear that a large proportion of fractures in men is expected to occur before the age of 80; however, it is not consistent with current public health-based approaches that stress treating only the oldest old (age, 80+ years) because these are at the highest individual risk.
Several studies(15, 18, 19) have suggested a wide geographic variation in hip fracture incidence rates between, as well as within, countries, including in Australia, with the highest rate reported for the northern European countries and for the United States and the lowest rate reported in Africa and some Asian populations. In general, people who live in latitudes further from the equator seem to have a higher incidence of fracture. Results of this study confirm such a variation. Although the incidence rate of hip fracture in Dubbo was comparable with most white and some Asian populations,(2, 12-15) it was significantly lower than in Malmo (Sweden). Even within Australia, it seems that the incidence of hip fracture in Dubbo (and Geelong) was higher than in Tasmania.(16, 18) Apart from the effects of genetic and environmental factors, the difference could also be caused by the differential accuracy in the recording of fracture data. In some studies, the small number of fracture cases makes the incidence estimates unreliable.
In this population, there was a statistically significant reduction in the incidence of total symptomatic fractures in both men and women. There was no statistically significant change in the incidence rate of hip fracture in either men or women during the 12-year study period. However, it was estimated given the number of hip fracture cases that the power to detect a significant change in incidence rate was 43% for women and 7% for men. There is no empirical explanation for the overall reduction in the incidence of total osteoporotic fractures, although a number of suppositions related to weight change and dietary calcium intake could be proposed. In the larger subset of the Dubbo population from which longitudinal data are available, men and women in the latter part of the study (1997-2000) tended to gain more weight than in the earlier part (1989-1992). Furthermore, the average daily calcium intake in 1989 in Dubbo was 636 ± 338 (SD) mg/day in men and 642 ± 353 mg/day in women, and these intakes were significantly lower than the Australian(20) recommended daily calcium intake of 800 mg for men and 1000 mg for postmenopausal women. Although we have no data on this, the dietary calcium intake could have been improved because of widespread publicity on the need for higher calcium intake. The trend of increase in weight in both men and women in Dubbo and the possibility of improving calcium intake could be contributing factors to the overall downward trend of the total osteoporotic fracture incidence seen in this study.
During the time frame of the Dubbo study to the end of 2000, few (5% of all participants) were on some form of antifracture medication, including hormone replacement therapy, vitamin D compounds, selective estrogen receptor modulators, and bisphosphonates. While the use of antifracture medication could have reduced the overall fracture incidence, it seems unlikely that the prevalence of use of 5% in the population, even in those at high risk, would explain the 4-6% overall reduction in all symptomatic fractures.
The strengths of this study include the complete ascertainment of fractures in a stable and homogeneous population, long duration of follow-up, and the inclusion of both men and women. These positive features of the study allow subgroup differences to be reliably detected, which would not be possible with smaller and shorter duration studies. Nevertheless, because hip fracture is a relatively rare event, the actual number of fracture cases was modest, which limits the power to detect a small change in the incidence rate of hip fractures. These results from subjects of white background may not be generalizable to other populations.
In summary, this comprehensive study with full ascertainment of all types of osteoporotic fractures in men and women over 12 years in a white population showed comparable hip fracture incidences to other white (except Sweden) and Asian data and for two other regional centers within Australia. Of particular interest, the fracture incidence in men approached that in women with advancing age, almost one-half (48%) of hip fractures in men occurred before the eighth decade of life, and 66% of hip fractures in women occurred before the age of 85 years. There was a significant 4-6% per year decrease in overall osteoporotic fractures in both men and women. Although there was no apparent decrease in hip fracture incidence, the study had limited power to detect such a change. The incidence data highlight the fact that a large proportion of hip fractures occurs in those under 80 years of age, particularly in men. The age distribution of hip fracture incidence underlines the need for earlier intervention in osteoporosis in women, and particularly in men, to achieve the most cost-effective outcomes.
We gratefully acknowledge the expert assistance of Janet Watters and Donna Reeves in the interview, data collection, and measurement of bone densitometry, and the invaluable help of the Dubbo Base Hospital radiology staff, Dr R Slack-Smith, and Orana radiology. We thank Natasa Ivankovic for help in managing the database. This work was supported by the National Health and Medical Research Council of Australia.
- 11997 World-wide projections for hip fracture. Osteoporos Int 7:407–413., ,
- 22001 National Working Committee on Osteoporosis. Hip fracture incidence rates in Singapore 1991–1998. Osteoporos Int 12:311–318., , , ,
- 32002 Effect of birth cohort on risk of hip fracture: Age-specific incidence rates in the Framingham Study. Am J Public Health 92:858–862., , , ,
- 41992 Hip fractures in the elderly: A world-wide projection. Osteoporos Int 2:285–289., ,
- 51987 Secular trends in the incidence of hip fractures. Calcif Tissue Int 41:57–64., ,
- 61990 Trends in admissions for hip fracture in England and Wales, 1968–85. BMJ 300:1173–1174., ,
- 71990 The Dubbo study: An Australian prospective community study of the health of elderly. Aust NZ J Med 20:783–789., , , , , ,
- 81994 Symptomatic fracture incidence in elderly men and women: The Dubbo Osteoporosis Epidemiology Study (DOES). Osteoporos Int 4:277–282., , , , ,
- 91999 Mortality after all major types of osteoporotic fracture in men and women: An observational study. Lancet 353:878–882., , , ,
- 10Australian Bureau of Statistics 2003 Population by Age and Sex, New South Wales. Australian Bureau of Statistics, Canberra, Australia.
- 11SAS Institute Inc. 2000 SAS, version 8. SAS Institute, Inc., Cary, NC, USA.
- 121993 Incidence of hip fractures in Malmo, Sweden (1950–1991). Bone 14:S23–S29., , , , , ,
- 131999 Fracture incidence in Olmsted County, Minnesota: Comparison of urban with rural rates and changes in urban rates over time. Osteoporos Int 9:29–37., ,
- 141995 Hip fracture incidence among older people in Auckland: A population-based study. NZ Med J 108:426–428., , , , , ,
- 151999 International variation in the incidence of hip fractures: Cross-national project on osteoporosis for the World Health Organization Program for Research on Aging. Osteoporos Int 9:242–253., , , , , , , , , , ,
- 161999 Age- and gender-specific rate of fractures in Australia: A population-based study. Osteoporos Int 10:240–247., , , , , , ,
- 172001 Hip fractures in the Australian Capital Territory. Proceedings of the 2001 Annual Scientific Meeting of the Australian Society for Geriatric Medicine, Blue Mountains, NSW, Australia., , , ,
- 182001 A population based study of fracture incidence in Southern Tasmania: Lifetime Fracture risk and evidence for geographic variation within the same country. Osteoporos Int 12:124–130.,
- 191997 Regional variation in the incidence of hip fracture in New Zealand. NZ Med J 110:78–80., , , ,
- 20NHMRC 2003 NHMRC dietary guidelines for Australian adults. Available online at http://www.nhmrc.gov.au/publications/nhome.htm. Accessed August 23, 2003.