Why Do Older Women with Diabetes Have an Increased Fracture Risk?

Dorothy A. Nelson and Scott J. Jacober

Wayne State University School of Medicine Detroit, Michigan 48201

Address correspondence and requests for reprints to: Dorothy A. Nelson, Ph.D., Department of Internal Medicine, Wayne State University, Hutzel Hospital, 1 Lobby Center, 4707 St. Antoine, Detroit, Michigan 48201. E-mail: dnelson{at}intmed.wayne.edu


    Introduction
 Top
 Introduction
 References
 
By definition, osteoporosis is characterized by low bone mass, architectural deterioration, increased fragility, and an increased fracture risk (1). A multitude of studies has demonstrated the relationship between low bone mass and increased fracture risk, mainly in white populations. Conversely, populations with high bone mass generally have a lower fracture risk. However, an apparent paradox emerges from the study by Schwartz et al. (2), in this issue of the journal. They report that a population with relatively high bone mass—people with type 2 diabetes—have an increased fracture risk compared with people without diabetes (2). This is perhaps more surprising because of the association between type 2 diabetes and obesity, and the presumed protective effect of obesity against osteoporosis and fractures. So why do these older women with diabetes have more fractures? Can these fractures be attributed to osteoporosis?

Although type 1 diabetes seems to be associated with a low bone mass, most studies have indicated a relatively normal bone mass in type 2 diabetes. However, the importance of bone quality and the likelihood of injury (e.g. from falls) in determining fracture risk (3) should not be underestimated. The consensus definition of osteoporosis given above differs from earlier concepts in that it includes the deterioration of bone architecture, and bone fragility, in addition to low bone mass. When assessing whether a fracture in a person with normal bone density is osteoporotic, one must consider these other elements of the disease.

In diabetes, there are many factors that could predispose patients to injury sufficient to cause bone fractures. These include altered proprioception, balance, and gait due to neuropathy and visual impairment from diabetic retinopathy and cataracts. If these do increase the frequency of falls (and there is some evidence that people with diabetes have frequent falls; Ref. 4), then fractures at certain skeletal sites are predictable.

Little is known about bone quality in people with diabetes, although abnormal vascular function could theoretically have an adverse effect on bone remodeling and, therefore, on bone quality. The combination of poor bone quality and frequent, injurious falls would be expected to increase fracture risk independently of bone mass.

Schwartz et al. (2) were able to take into account more than two dozen covariates that might be relevant to the analysis of fracture risk in their cohort, who were participating in the Study of Osteoporotic Fractures (SOF). This is a large cohort of 9654 white women over the age of 65, followed for an average of 9.4 yr. The main outcome variables in SOF are osteoporosis and fractures, although the study is so large and comprehensive that many other diseases and variables can be meaningfully assessed. For example, SOF has data on covariates that are relevant to the diabetes study, such as factors that could be indicators of diabetic retinopathy and neuropathy. These include vibratory perception, pressure sensitivity, grip strength, hip extensor strength, static balance, and gait speed for diabetic neuropathy; and corrected visual acuity, depth perception, and contrast sensitivity (2).

As one might expect, the women with diabetes had poorer scores on measures of balance and gait, peripheral neuropathy, and vision compared with the nondiabetic subjects. Despite this wealth of information, because SOF was not designed to study diabetes, there are several potentially relevant variables that are not available in the database. For example, while data are available on lower extremity vibratory perception and pressure sensitivity of the great toe, the authors acknowledge that they do not have data on more sensitive measures of peripheral neuropathy such as nerve conduction velocity.

There are only a few other studies of bone mass and fractures in diabetic subjects that approach the magnitude of SOF. The Rotterdam Study, with 3450 elderly women, of whom 335 had type 2 diabetes, found a higher bone mass among the women with diabetes compared with those without (5). In contrast with the SOF study, the women with diabetes in the Rotterdam Study reported fewer nonvertebral fractures, but this was for the 5 yr preceding their baseline visit (odds ratio, 0.63; 95% confidence interval, 0.44–0.90) (6). The Nord-Trondelag Health Survey of 35,444 people 50 yr and older determined hip fracture incidence over 9 yr of follow-up (7). They found that among women ages 50–74 with type 2 diabetes of long duration, the relative risk of hip fracture was 1.8 (95% confidence interval, 1.1–2.9), a rate that is similar to that reported for SOF. Schwartz et al. (2) also reported that the duration of diabetes was associated with a higher risk of hip fracture and that there was a tendency for any fracture risk to be higher among insulin-treated diabetics, who may have more severe disease.

It is instructive to consider what types of fractures are associated with diabetes in the SOF study, and whether these are typical of osteoporosis. After adjusting for age, body mass index, and bone density, Schwartz et al. (2) found that the risk of hip and proximal humerus fractures were increased in noninsulin users, whereas there was an increased risk of foot fractures in insulin users. Certainly, hip fractures, and probably humerus fractures, are characteristic of osteoporosis, but would foot fractures fall into this category? An earlier paper from the SOF group identified which fractures were associated with low bone mass in their study: the hip, humerus, and foot were all included in the list, whereas fractures of the ankle, elbow, finger, and face were not (8). In the diabetes subgroup reported by Schwartz et al. (2), the increased rate of hip, humerus, and foot fractures do not seem to be explained by low bone mass and, by convention, would not be considered "osteoporotic." However, there is ample evidence in the osteoporosis literature that a variety of risk factors, which are independent of bone mass, contribute to fractures.

It is well known that falling is an independent risk factor for hip fractures, and poor depth perception has been reported as an independent risk factor for foot fracture (9). The latter provides a logical link between diabetic retinopathy, poor depth perception, and foot fractures in people with diabetes. The characteristics discussed above could theoretically contribute to an increased rate of falls, but there are few data on falls in diabetic people. There is one study of 116 diabetic African-Americans compared with 522 nondiabetic subjects (4), in which the diabetic subjects reported significantly more injurious falls than the nondiabetic subjects (12% vs. 6%, P < 0.025).

There are many features of diabetes, not included in the SOF database, that could contribute to an increased frequency of falls. These include the frequency of nocturia, visual field deficits from cataracts, and loss of peripheral vision and night blindness resulting from laser therapy of diabetic retinopathy. Variations in glycemic control may also be relevant to falls because acute fluctuations in glucose levels can transiently affect visual and nerve function, as well as cognition.

The role of bone fragility in fracture risk among people with diabetes is less easy to document because there are few available noninvasive methods of measuring bone quality. Schwartz et al. consider the possibility that diabetes is associated with a decrease in bone strength that is not apparent from conventional bone density measurements (2). The conventional bone mass measurement methods [i.e. dual-energy x-ray absorptiometry (DXA)] typically concentrate on the spine, hip, forearm, and sometimes heel, but not the humerus or foot bones that are apparently prone to fracture in people with diabetes. Furthermore, an estimate of mass per projected area is provided by these measurements, but not information about bone structure or architecture.

Using x-ray morphometry, Cundy et al. (10) found that cortical bone mass in the hands and feet was reduced in patients with diabetic neuropathy. This may suggest a preferential loss of cortical bone in diabetes that cannot be captured by integral (cortical plus trabecular) bone mass measurements. It is also consistent with the types of fractures found in the Schwarz study: hip, humerus, and foot, which are sites with a relatively high proportion of cortical bone. The distal radius, which has a higher trabecular bone component, does not seem to be a typical fracture site in diabetes, despite the implication of frequent falls in the etiology of these fractures. Cross-sectional geometry of the hip or humerus, based on computed tomography methods, or as derived from DXA data (11), might provide important information on the etiology of hip and arm fractures in diabetics that is not reflected in a conventional DXA measurement.

One possible source of poor bone quality is the abnormal vascular function associated with diabetes. For example, it has been suggested that the rate of bone turnover may be higher in people with type 2 diabetes (12), and that it may be affected by the microvascular complications associated with diabetes (13, 14). With data from SOF, Vogt et al. (13) evaluated the hypothesis that reduced blood flow to the lower extremities may affect bone remodeling. The ankle/arm index (the ratio of the posterior tibial and brachial systolic blood pressures) was used as a measure of blood flow to the legs. This index was positively correlated with bone density at several sites, such that a decrease of 2 SD in the index was associated with a decrease of 3.7% in hip bone density. Whereas bone density in the study by Schwarz et al. (2) was not lower at baseline in women with diabetes, a subsequent study of this cohort reported at this year’s annual meeting of the American Society for Bone and Mineral Research showed that women with diabetes lost hip bone density at a faster rate than women without diabetes (15). It is possible that a normal or high bone mass in people with diabetes is associated with a higher body mass index, or obesity, but that the comorbidites of diabetes lead to accelerated bone loss, a deterioration in bone quality, and frequent falls. This chain of events would, logically, lead to the increased fracture risk reported by Schwartz et al. (2). If this scenario holds true, such fractures would be considered osteoporotic based on the consensus definition.

In summary, the apparent paradox of a normal (or high) bone mass and increased fracture risk is less troublesome when one remembers that osteoporotic fractures result not just from a low bone mass but from decreased bone quality and an increase in injurious falls. An analogous case is found in some of the rheumatic diseases, in which initially normal bone mass may be adversely affected over time by treatment with glucocorticoids, leading to increased fracture risk. Furthermore, arthropathies interfere with normal gait and balance, possibly leading to more falls. In the diabetic patient, neuropathy may blunt the body’s usual defenses against injury. In both cases, the prevalence of overweight or obesity, and a high prevalence of these diseases in nonwhites, might lead to complacency by the health care providers with regard to fracture risk in these populations. For patients with diabetes, the article by Schwartz et al. (2) should remind the practitioner to consider the prevention of hip, foot, and humeral fractures. In general, the article should remind us that a normal bone mass, as conventionally measured by DXA and related modalities, can be offset by bone fragility at certain fracture sites and the likelihood of injury to these sites.

Received November 16, 2000.

Accepted November 16, 2000.


    References
 Top
 Introduction
 References
 

  1. WHO Study Group. 1994 Assessment of fracture risk and its application to screening for postmenopausal osteoporosis. WHO Technical Report Series 843. Geneva: WHO.
  2. Schwartz AV, Sellmeyer DE, Ensrud KE, et al. 2001 Older women with diabetes have an increased risk of fracture: a prospective study. J Clin Endocrinol Metab. 86:32–38.[Abstract/Free Full Text]
  3. Heaney RP. 1996 Pathophysiology of osteoporosis. Am J Med Sci. 312:251–256.[CrossRef][Medline]
  4. Miller DK, Lui LY, Perry HM, Kaiser FE, Morley JE. 1999 Reported and measured physical functioning in older inner-city diabetic African Americans. J Gerontol Series A, Biol Sci Med Sci. 54:M230–M236.
  5. Stolk RP, Van Daele PL, Pols HA, et al. 1996 Hyperinsulinemia and bone material density in an elderly population: The Rotterdam Study. Bone. 18:545–549.[CrossRef][Medline]
  6. Van Daele PL, Stolk RP, Burger H, et al. 1995 Bone density in non-insulin-dependent diabetes mellitus. The Rotterdam Study. Ann Intern Med. 122:409–414.[Abstract/Free Full Text]
  7. Forsen L, Meyer HE, Midthjell K, Edna TH. 1999 Diabetes mellitus and the incidence of hip fracture: results from the Nord-Trondelag Health Survey. Diabetologia. 42:920–925.[CrossRef][Medline]
  8. Seeley DG, Browner WS, Nevitt MC, Genant HK, Scott JC, Cummings SR. 1991 Which fractures are associated with low appendicular bone mass in elderly women? The Study of Osteoporotic Fractures Research Group. Ann Intern Med. 114:837–842.
  9. Seeley DG, Kelsey J, Jergas M, Nevitt MC. 1996 Predictors of ankle and foot fractures in older women. The Study of Osteoporotic Fractures Research Group. J Bone Miner Res. 11:1347–1355.[Medline]
  10. Cundy TF, Edmonds ME, Watkins PJ. 1985 Osteopenia and metatarsal fractures in diabetic neuropathy. Diabet Med. 2:461–464.[Medline]
  11. Beck TJ, Ruff CB, Warden KE, Scott WW, Rao RU. 1990 Predicting femoral neck strength from bone mineral data: a structural approach. Invest Radiol. 25:6–18.[Medline]
  12. Hampson G, Evans C, Petitt RJ, et al. 1998 Bone mineral density, collagen type 1 alpha 1 genotypes and bone turnover in premenopausal women with diabetes mellitus. Diabetologia. 41:1314–1320.[CrossRef][Medline]
  13. Vogt MT, Cauley JA, Kuller LH, Nevitt MC. 1997 Bone mineral density and blood flow to the lower extremities: The Study of Osteoporotic Fractures. J Bone Miner Res. 12:283–289.[Medline]
  14. Piepkorn B, Kann P, Forst T, Andreas J, Pfutzner A, Beyer J. 1997 Bone mineral density and bone metabolism in diabetes mellitus. Horm Metab Res. 29:584–591.[Medline]
  15. Schwartz AV, Sellmeyer DE, Nevitt MC, et al. 2000 Older women with diabetes have a higher rate of bone loss at the hip. J Bone Miner Res. 15(Suppl 1):S188.