Fractures Among the Elderly: An Old Problem

Fractures among the elderly are a major health problem in developed countries. Numerous fracture risk factors have been identified; many exert their effects through bone mass, whereas others (such as risk factors for falls) act independently.1-9 Although effective treatments are now available for patients with low bone density, questions remain about how to maximize skeletal health earlier in life so that fewer people require treatment. Skeletal features other than bone density such as bone geometry and microarchitecture are also important, but little is known about how to maximize or preserve these features, and we are still learning which have the strongest influence on fracture risk. It is also becoming increasingly clear that nonskeletal risk factors such as falls must also be addressed. There are large differences among countries in rates of hip and other fractures.10-17 As noted by Johnell et al.,13 there exists “a larger difference in incidence between countries than between sexes, which suggests important genetic or environmental factors in the causation of hip fracture.” Large differences in risk imply the presence of important risk factors that could potentially be exploited to improve interventions. One country with low rates of hip fracture is Japan. Both men and women in Japan have hip fracture rates half that of U.S. Caucasians.10-12 Even so, osteoporosis and hip fractures are a serious problem in Japan because the life expectancy is among the longest in the world, and a large segment of the population is approaching old age, when fracture incidence becomes extremely high. Osteoporotic fractures are also a concern for other countries that are experiencing improvements in longevity; the number of hip fractures in Asia is projected to represent half of all cases in the world by the year 2050, almost double the number for North America and Europe combined.18 It is sometimes possible to identify environmental (nonhereditary) factors responsible for differences in disease patterns between migrants and their original country by taking advantage of the genetic similarities and cultural diversities of the two populations. Compared with many countries, Japan is relatively racially homogeneous because it has remained isolated both physically and politically throughout much of its history. Comparisons between Japanese in Hawaii and other populations, including Japanese in Japan and in California, have yielded important insights into cardiovascular disease, cancer, and now osteoporosis.19 As with people in Japan, the incidence of hip fracture among Japanese in Hawaii is about half that of Caucasians in the U.S.12 In contrast, the prevalence of vertebral fractures is similar among Japanese Americans and U.S. Caucasians, but higher in Japan.14 The divergent findings suggest that different risk factors probably influence hip and spine fractures to some extent. Cross-sectional analyses suggest that the higher prevalence of spine fracture appears to be at least partly a result of lower bone mineral density (BMD) in Japan.15, 20, 21 However, it remains uncertain why the hip fracture incidence in Japan is similar to Japanese in Hawaii, and lower than for Caucasians. The lower risk of hip fracture (compared with Caucasians) does not appear to be related to BMD. Numerous studies have demonstrated that BMD is similar or lower among Asians in comparison with Caucasians after adjusting for body size.22-30 While this is certainly true for both young (25–34 years) and early postmenopausal women, the majority of hip fractures occur after age 80, and there have been few if any comparisons between Japanese and Caucasians for this age group. There have also been few comparative studies that included native Japanese in Japan. Another candidate is bone geometry; the shorter hip axis length among Asian women is of sufficient magnitude to explain much of the lower hip fracture risk relative to Caucasians.5, 31 However, the difference in risk of hip fractures might also be explained by the lower incidence of falls among Japanese in both Hawaii and Japan, which is only half that of Caucasians, or by other factors.32-35 Prospective studies are needed that compare these and other risk factors between populations, to determine which factors truly account for the observed differences in fracture incidence. Perhaps more than any other country, the influence of changes in diet and other cultural factors on skeletal maturation since the turn of this century is strikingly apparent in Japan. Significant increases in height have occurred between generations in Japan. The temporal changes in culture and physique among Japanese create both opportunities and dilemmas. If differences in bone mass, bone geometry, and other variables can be demonstrated between generations, they may be associated with differences in fracture risk (secular trends), thereby providing insights into fracture causation. Furthermore, if differences in fracture risk factors between generations can be shown to result from changes in diet and other factors, this information may assist the design of interventions. One dilemma is that it may not be possible to sort out cause and effect because many variables are interrelated and because prospective data are not available for the majority of past experiences, such as dietary intake during childhood. Another dilemma is that the age and cohort effects cannot be distinguished from each other without good prospective data covering several decades of follow-up. Most epidemiologic studies in Japan have only recently been initiated. One exception is the Adult Health Study (AHS) cohort being studied by the Radiation Effects Research Foundation, consisting of survivors of the tragic atomic bombing of Hiroshima and Nagasaki as well as nonexposed controls. The large size and long follow-up make the AHS cohort a valuable resource. Although bone density measurements were not available until recently, a wealth of other variables was collected many years ago, forming the basis for the prospective study of risk factors for hip fracture reported in this issue of JBMR.36 As reported by others,6 the presence of vertebral fractures at baseline was an important predictor of hip fracture risk in the study of Fujiwara et al.,36 implying a generalized reduction in skeletal strength among fracture cases. A previous report based on serial radiographs acquired over several decades suggests that the incidence of vertebral fractures has declined among later birth cohorts within the AHS.37 Thus, the higher prevalence of spine fracture relative to Japanese Americans and U.S. Caucasians may disappear as later birth cohorts replace the older generations. In contrast to the decrease in vertebral fractures, there is tantalizing, but limited, evidence of an increase in trochanteric hip fracture incidence among women, but not men, in Japan.38 Certain risk factors including vertebral fractures, low BMD, and bone geometry reportedly differ between trochanteric and cervical hip fractures.39-41 However, prospective bone measurements are needed to determine whether such trends are related to changes in skeletal features (such as BMD, geometry, or microarchitecture) or nonskeletal characteristics. Age at menarche is one factor that may have influenced skeletal development long ago but appears to exert a long-term influence on fracture risk in the study by Fujiwara et al.36 Age at menarche of 15 years or older was associated with a doubling of hip fracture risk, and almost 60% of the AHS women fell into this category. As in many other countries, age at menarche has been gradually decreasing in Japan over the last 50 years, but has remained consistently 2 years later than for comparable birth cohorts of Japanese Americans in the U.S.42 One would therefore expect women in Japan to have a higher rate of hip fracture because of their later menarche compared with Japanese Americans, but the rates are similar in both populations. Age at menarche is easy to obtain, but it may not be the best (or the only) indicator of hormonal influences during development. One hypothesis is that a “longer period of prepubertal growth might result in greater bone size and cortical thickness,” but it is not clear what role age at menarche plays.22 The timing of menarche may simply be an indicator of any number of other factors that have more direct influences on skeletal maturation. In Japan, the average height at maturity increased approximately 10 cm for both men and women born after 1960, compared with those born before 1920.43 However, greater height is associated with increased hip axis length, and both height and bone geometry are stable once maturity is achieved.44 Thus, even if BMD is increased among later generations in Japan (which remains to be seen), a secular trend of increasing hip axis length, such as that reported for women in the United Kingdom45 may negate the potential benefit on hip fracture risk. Indeed, increased height itself is also associated with increased hip fracture risk, an ominous sign for the Japanese.1, 46 In contrast to the increase in height during the first half of this century, there was little if any change in weight for women, and only a minor increase of 2 or 3 kg for men over the same period.43 This implies that the Japanese physique has become less stocky over time, and that there is less muscle per unit of height. One might speculate that the ratio of muscle to fat mass may have declined during this same period as the country became more industrialized, which would further reduce the amount of muscle. Such changes may have implications for fracture risk, because a lower amount of muscle acting on longer bones may result in less strength and control, increasing the risk of falls and fractures. Increased calcium intake appears to result in greater BMD and reduced fracture risk.2 The benefit of increasing calcium consumption may be more pronounced among people with calcium-deficient diets, and calcium intake may have been lower in Asia 50 years ago. As with BMD, there are multiple determinants of skeletal height. For example, milk consumption during adolescence, birth weight, age at menarche, and race were all related to peak adult height among young women in Hawaii.47 To make matters even more complex, some factors appear to influence BMD and bone size at certain skeletal sites more than others.48 Fujiwara et al.36 found that increased milk intake among adults reduced fracture risk, but it remains uncertain if the primary effect on fracture risk occurred long ago during skeletal development, more recently by preserving bone mass, or both. The finding that women with five or more children had a higher risk of hip fracture also implies that nutritional and physiologic stresses after skeletal maturation may have important effects on fracture risk. It is easy to invoke the effects of changes in diet and other factors on skeletal development as a potential explanation for the Japanese trends in body size, but deciphering which nutrients are responsible, and the final effects on fracture risk, is another matter. It may be difficult or impossible to distinguish between hormonal and mechanical factors, or between physical activity and muscular loading effects, especially since such factors may influence each other. Among young Japanese women in Hawaii, increased milk consumption was associated with longer leg length, and to a lesser extent, greater sitting height (primarily a measure of spine length).47 After adjusting for milk intake, later age at menarche was also associated with longer leg length. Although age at menarche has declined in successive generations in both Japan and Hawaii, stature overall has increased, suggesting that dietary influences on height more than compensated for the changes in age at menarche, or that menarche is not the best indicator of hormonal effects. Nutritional factors other than calcium such as protein, vitamin D status, magnesium, zinc, and total caloric intake may also play important roles during both development and later life.49-51 What effect might these changes have on fracture incidence? There have been significant advances in our understanding of skeletal growth recently, but many questions remain. It may never be possible to reconstruct the life events that led to differences in fracture risk between populations, or between past and current generations. However, careful studies of the determinants of bone mass, structure, and geometry from conception through adulthood, combined with information about which risk factors best predict fractures among adults, may lead to effective interventions. Hopefully, we will learn enough to anticipate which factors are most important and plan appropriately to optimize the skeletal health of future generations.