Epidemiology of Fractures – Effects of Age, Gender, Race and Geography
Osteoporosis has long been considered synonymous with vertebral fractures in postmenopausal women. Only in the past few decades has it been shown that men account for about 20% of all hip fractures and, more recently still, that vertebral fractures may be as common in men as in women. The lifetime risk of any fracture of the hip, spine, or distal forearm in men has been estimated at 13%, compared to 40% in women, and is similar to the lifetime risk of prostate cancer (Melton et ai, 1992). Even though it is recognized that diaphyseal fractures of arms, legs, hands, and feet are more common among men, there has been little interest in these injuries which are generally attributed to severe trauma. However, these same fractures in elderly women are due in part to low bone mineral density (BMD) levels (Seeley et al., 1991).
There is a similar relationship between bone density and fractures in men (Nguyen et ai, 1996), and a panel of experts judged that osteoporosis might account for 60-85% of hip fractures in men, depending on age, along with 70-90% of vertebral fractures, 40-45% of distal forearm fractures, and 15-45% of fractures at other skeletal sites (Melton et al., 1997).
These osteoporotic fractures in men account for annual expenditures of $2.7 billion, or one-fifth of the total cost of osteoporotic fractures in the United States each year (Ray et al., 1997). Interest has also grown with the advent of potent drugs that can be used to treat osteoporosis in men and with the realization that there is a large population of men at risk. For example, data from the Third National Health and Nutrition Examination Survey (NHANES III) indicate that 7, 5, and 3%, respectively, of white, African-American, and Hispanic men currently have osteoporosis of the hip (Looker et al., 1997). This post provides an overview of fracture epidemiology in men and provides a basis for more detailed exploration of these issues in the subsequent posts.
2. Effects of Age
Overall fracture incidence in the community is bimodal. Among adolescents and young adults, fractures are more common among males than females and usually result from significant trauma (Melton, 1995). Fractures of the shafts of long bones typify this pattern of occurrence.
Distal forearm fractures are more frequent among males than females below age 35 years, but over this age they become more frequent in women (Owen et al., 1982), as do the other fractures that have been associated with osteoporosis. These traditionally include fractures of the hip and spine (Melton, 1995) but, as mentioned earlier, essentially all fractures in elderly women are associated with low bone density. Men also lose bone with aging. Moreover, most limb fractures are the result of falls, and the risk of falling increases with aging in men as well as women (Winner et al., 1989). Consequently, there are age-related increases not only for hip and spine fractures in men but also for fractures of the proximal humerus (Baron et al., 1996; Bengner et al., 1988; Donaldson et al., 1990; Knowelden etal., 1964; Kristiansen etal., 1987; Rose etal., 1982), pelvis (Baron etal., 1996; Knowelden et al., 1964; Liithje et al., 1995; Melton et al., 1981; Ragnarsson and Jacobsson, 1992), patella (Baron etal., 1996) and, in some studies, the ankle (Daly etal., 1987).
3. Effects of Gender
Fractures are actually more frequent among men than women at most skeletal sites (Donaldson et al., 1990). Fractures of the hands and feet, for example, are nearly three times more common among men (Garraway et al., 1979); over a third of these fractures are incurred during recreational and sporting activities, whereas another fifth are related to crush injuries, often occupational. Only for fractures of the proximal humerus, distal forearm, pelvis, and proximal femur are rates greater among women (Table I). For these four sites combined, the incidence was over 60% greater among women, and, because there are many more elderly women than men, the female excess in the actual number of cases is even higher. The greater risk of these fractures in women has been attributed both to lower average bone mass compared to men as well as a greater risk of falling. Data from NHANES III show that total hip BMD is 12-13% greater in white, African-American, and Hispanic men compared to women of the same ethnicity (Looker et al., 1995), whereas the risk of falling at any age above 65 years is 10-70% greater among women (Winner etal., 1989). This may not be true in all populations, however. Among the Maori in New Zealand, men and women have similar hip fracture incidence rates (Stott and Gray, 1980), whereas hip fracture rates are higher among Bantu men in South Africa (Solomon, 1968) and Chinese and Malay men in Singapore (Wong, 1964). Most studies show that prevalence rates for vertebral fractures are similar in women and men, but prevalence rates were higher in men in two surveys in rural North Dakota, where farming was the primary occupation (Bernstein et al., 1966). When all fractures are considered, however, the actual number in men exceeds that in women by 10-20% (Donaldson et al., 1990; Garraway et al., 1979).
|TABLE Average Annual Incidence (per 100,000 population) of Fractures at Different Sites among Men and Women in Leicestershire, England
4. Effects of Race
Within each gender, fracture rates are usually highest for whites and lower for other ethnic groups. Rates are high among persons of northern European extraction whether they live in North America, Scandinavia, New Zealand, or South Africa (Melton, 1991) . Conversely, hip fracture incidence is very low among the Maori people in New Zealand (Stott and Gray, 1980) and the Bantu in South Africa (Solomon, 1968). The lower incidence among African-Americans (Farmer et al., 1984; Griffin et al., 1992; Jacobsen et al., 1990a; Rodriguez et al., 1989; Silverman and Madison, 1988), has been explained on the basis of their greater bone mass. In NHANES III, for example, total hip BMD was 10% greater in African-Americans compared to whites of each sex (Looker et al.,
1995). The South African Bantu, on the other hand, have the lowest hip fracture incidence of any population, yet have metacarpal bone density lower than Johannesburg whites (Solomon, 1979). Similar observations have been made for women in Gambia (Aspray et al., 1996). Likewise, the incidence of hip fractures among men and women of Asian ancestry is about half that of their white counterparts (Ho et al., 1993; Lau et al., 1990; Lauderdale et al., 1997; Ross et al., 1991; Silverman and Madison, 1988) even though their bone mass is somewhat lower. However, racial comparisons of bone density are confounded by differences in bone size, and, when these are taken into account, there is little difference between white and Asian women (Bhudhikanok et al., 1996; Cundy et al., 1995; Ross et al., 1996; Russell-Aulet et al., 1993). Alternatively, the lower risk of hip fracture in nonwhites could be due to a lower risk of falling (Lipsitz et al., 1994; Nevitt et al., 1989; Tinetti et al., 1988) or to other risk factors like decreased hip axis length or femoral neck angle (Cummings et al., 1994; Nakamura et al., 1994; Villa et al., 1995).
The prevalence of vertebral fractures among Asians is about as high as in whites (Lau etal., 1996; Ross et al., 1995), despite their lower hip fracture rates. For example, hip fracture incidence rates are lower among Japanese women than among Japanese-Americans, whose rates in turn are lower than those of whites (Ross et al., 1991). Vertebral fracture prevalence, on the other hand, was almost twice as high among women in Hiroshima compared to Hawaiian women of Japanese descent but was also 20-80% greater than rates for white women (Ross et al., 1995). Few data are available for other ethnic groups. Hospital discharge rates for vertebral fractures in the United States are about four times greater for elderly whites than for African- American men and women (Jacobsen et al., 1992). Likewise, the prevalence of vertebral fractures is lower among Mexican-American as compared to non-Hispanic white women (Bauer and Deyo, 1987). Forearm fractures are less frequent in African-American (Anonymous, 1996; Baron et al., 1994; Griffin et al., 1992) and Japanese populations (Hagino et al., 1989), but there is still a substantial female excess. In Africa and Southeast Asia, however, distal forearm fractures are even less common and rates for women are little more than those for men (Adebajo et al., 1991; Wong, 1965). Lower rates for African-Americans than whites have also been reported for fractures of the proximal forearm, humerous, ribs, pelvis, patella, ankle, hands, and feet (Anonymous, 1996; Griffin et al., 1992).
5. Effects of Geography
Fracture rates at different sites tend to correlate within a population (Table II). Thus, forearm fracture rates in the United Kingdom are around 30% lower than those in the United States, as are hip fracture rates (Melton, 1995). More difficult to explain is the variation in fracture incidence within populations. For example, hip fracture rates vary more than sevenfold from one country to another within Europe (Elffors et al., 1994; Johnell et al., 1992), and comparable variation has been observed for vertebral fractures (Johnell et al., 1997; O’Neill et al., 1996). Although the latter data mostly reflect the experience of Mexican-Americans, hip fracture incidence is comparable in Madrid, Seville, Barcelona, and Salamanca (Diez et al., 1989; Elffors et al., 1994; Ferrandez et al., 1992) even though the American “Hispanic” and Spanish populations are not genetically comparable (Hanis et al., 1991).
Likewise, hip fracture incidence (Ross et al., 1991) and bone mass (Sugimoto et al., 1992) differ among Asian populations. An important role for environmental factors is also suggested by the marked variation in fracture incidence seen even within specific countries. Although the higher incidence of fractures in urban as opposed to rural districts has been explained on the basis of lower bone mass among urban residents (Gărdsell et al., 1991), studies in America indicate that the problem is more complex. In over 2000 countries nationwide, hip fracture rates in white women were higher in the South than the North (Jacobsen et al., 1990b), whereas the incidence of distal forearm and proximal humerus fractures was higher in the East and lower in the Western United States (Karagas et al., 1996). More detailed studies are needed to identify the environmental factors responsible for such regional differences.
|TABLE II Age-Adjusted° Incidence (per 100,000 per year) of Distal Forearm Fractures Compared to Hip Fractures in Different Populations of Persons 35 Years of Age or Older
6. Secular Trends
Osteoporosis and its attendant fractures impose a formidable burden on the medical system now increases in hip fracture incidence are occurring in many regions of the world (Melton et al., 1987). In Rochester, hip fracture incidence rates increased dramatically among women between 1928 and 1950 only to fall slowly thereafter, whereas rates in men rose steadily until 1980 but have since declined (Melton et al., 1996). Overall age- and sex-adjusted hip fracture incidence rates fell by almost 8% between 1963 to 1972 and 1983 to 1992 (Melton et al., 1998a), and rates now appear to be stabilizing for women in Sweden (Naessen et al., 1989), Great Britain (Spector et al., 1990) and Australia (Lau, 1993). However, a sharp increase in hip fracture incidence in some parts of the Far East in recent decades is bringing rates there closer to those in Northern Europe (Lau et al., 1990; Ling et al., 1996). Rochester rates for Colies’ fracture have also been relatively stable (Melton et al., 1998b), but the incidence of distal forearm fractures, ankle fractures, proximal humerus fractures, proximal tibia fractures, and possibly vertebral fractures appears to be increasing in other areas (Obrant et al., 1989). Detailed studies have found no evidence of an increase in vertebral fracture incidence in more recent years (Cooper et al., 1992a; Hansen et al., 1992), but the incidence of tibia and ankle fractures was found to have increased by 72% and 272%, respectively, in Rochester between 1969 to 1971 and 1989 to 1991 (Melton et al., 1999).
7. Public Health Implications
With continued aging of the population, the annual number of fractures is expected to rise dramatically in coming decades. In the United States, for example, the number of individuals aged 65 years and over is expected to rise from 32 million to 69 million between 1990 and 2050, and the number aged 85 years and over will grow from 3 million to 15 million. As a consequence, the number of hip fractures and their associated costs could triple by 2040 (Schneider and Guralnik, 1990), and other estimates are comparable (Cummings etal., 1990). Likewise, hip fractures might increase in the United Kingdom from 46,000 in 1985 to 60,000 in 2016 (Anonymous, 1989) and in Australia from 10,150 in 1986 to 18,550 in 2011 (Lord and Sinnett, 1986). Health authorities in Finland anticipate a 38% increase in the number of hip fractures between 1983 and 2010 with a 71% increase in resulting hospital bed-days (Simonen, 1988).
On a worldwide basis, the 323 million individuals who are 65 years of age and over will grow to an estimated 1.555 billion by the year 2050 with the greatest increase, from 190 million in 1990 to 1.271 billion in 2050, in Asia, Latin America, the Middle East and Africa (Cooper et al., 1992b). Because hip fracture incidence rates rise exponentially with aging, these demographic trends could cause the number of hip fractures worldwide to increase from an estimated 1.7 million in 1990 to a projected 6.3 million in 2050 (Cooper et al., 1992b). Any rise in incidence rates, over and above that due to population aging, will increase future fractures still further. Indeed, taking this into account, the number of hip fractures worldwide in 2050 could be as high as 21.3 million (Gullberg et al., 1997). It is clear, therefore, that large numbers of individuals will experience the pain, expense, disability, and decreased quality of life caused by osteoporotic fractures (Melton, 1995). If the enormous costs associated with osteoporotic fractures are to be reduced, increased attention must be given to the design and implementation of effective control programs. The posts that follow provide more insight into the pathophysiology of bone loss and fractures in men, and they provide essential clinical guidance with respect to osteoporosis prevention, diagnosis and treatment.