Diabetes and fracture risk: A review

Several epidemiological studies have demonstrated an increased risk of osteoporosis and fracture in people with diabetes. Several studies (reviewed in Hofbauer et al, 2007) suggest that, in type 1 diabetes, bone mineral density (BMD) is reduced at the spine and this is associated with the presence of diabetic complications. Studies examining fracture risk estimate that people with type 2 diabetes have a 1.4- to 2.6-fold increased risk of fracture at the spine and hip compared with people without diabetes (Janghorbani et al, 2007). 

Studies in people with type 2 diabetes have been complicated by the well-known protective effect of obesity on increasing BMD. Thus a typical overweight person with type 2 diabetes will have a higher BMD than a normal-weight control. There are, however, suggestions that low BMD is more likely with increasing duration of diabetes and with poorer metabolic control determined by HbA_1c (Hofbauer et al, 2007).

The relationship between fracture risk and diabetes is also complex. The overall risk of fracture is increased in type 2 diabetes: the risk of hip fracture appears to increase approximately 1.7-fold, and for all fractures about 1.3-fold. People with a very recent diagnosis of diabetes appear to be protected against fracture, even after adjustment for body mass index (BMI; Leslie et al, 2007). It has been speculated that this is due to the high circulating levels of insulin in these insulin-resistant individuals: the insulin is still having an anabolic effect on bone cells despite no longer having a hypoglycaemic effect. However, in people with more than 5 years since diagnosis of diabetes, this protective effect is lost and fracture risk is elevated (reviewed in Hofbauer et al [2007] and Jangharbani et al [2007]). 

The importance of falls as a cause of fracture is illustrated in studies examining the impact of diabetes on risk of falls. These show a 2.8-fold and 1.7-fold increased risk of falls in insulin and non-insulin treated people with type 2 diabetes respectively (de Liefde et al, 2005).

Why is fracture risk increased in diabetes?
There are many ways in which diabetes can affect bone strength and the risk of fracture (Hofbauer et al, 2007) (Figure 1). Insulin is an anabolic hormone for bone, stimulating the formation of new bone. Insulin deficiency, as in type 1 diabetes, could therefore impair new bone formation. People with poorly controlled diabetes also tend to have low levels of insulin-like growth factor-1 (IGF1), which is a major stimulator of bone formation. 

It is also possible that in type 1 diabetes the underlying inflammation associated with pancreatic beta-cell destruction may have an adverse impact on bone since any cause of inflammation tends to be detrimental to bone health through increasing bone resorption.

Persistently high glucose levels probably also directly affect long-lived proteins such as collagen that are important for bone strength. This abnormal glycation will affect bone strength independently of bone mineralisation or density. 

Diabetic complications can also impact on bone both directly and indirectly. The integrity of the autonomic nerve supply to bone is essential for proper bone function. The neuropathy seen in an acute Charcot joint is associated with dramatic changes in bone density and a transient, very high risk of bone fracture. Peripheral neuropathy increases the risk of fracture through loss of protective sensation and by increasing the risk of falls through loss of joint position sense. Autonomic neuropathy and retinopathy also contribute to the risk of falls. Falls are also more likely in people with poorly controlled diabetes, and in those experiencing frequent hypoglycaemic episodes or on antihypertensive medication. 

Estimating the risk of fracture in an individual patient
Many factors influence the risk of fracture in people with diabetes. These have to be taken into account when advising an individual about the need for further tests or treatment. The most important factors to consider are:

• age (fracture risk rises steeply with age)
• gender (women have a higher risk than men)
• previous fractures (a previous fracture increases fracture risk two- to threefold)
• BMD. 

BMD is usually measured by a bone density scan, i.e. DXA (dual-energy X-ray absorptiometry) – a non-invasive technique that uses low doses of X-rays. Bone density is reported in terms of standard deviations below or above the young adult mean bone density (called the T-score) for each gender. A bone density T-score of <-1 is termed osteopenia, but this is common even in the healthy population. A T-score of <-2.5 is termed osteoporosis. In general, for each decrease of 1 standard deviation in BMD the risk of fracture doubles. 

Other risk factors include an early menopause, low BMI, a family history of fractures, medications such as glucocorticoids, and chronic inflammatory diseases such as rheumatoid arthritis. 

In people with diabetes, an additional factor to consider is the use of thiazolidinediones (Kahn et al, 2006). Both rosiglitazone and pioglitazone are associated with an increased risk of fracture in women but not men. The pattern of fractures was unusual in that there was no increase in the risk of spine fracture, but fracture of the upper and lower limbs was much more common. This pattern is quite different from that seen in patients taking other drugs that affect bone, such as glucocorticoids. The reasons for these adverse effects of glitazones, and why they only appear to occur in women, need further study, but glitazones do appear to reduce the ability to form new bone, increase the ability to resorb bone from the skeleton and decrease BMD (Grey et al, 2007). 

In an individual patient all these factors need to be taken into account when deciding whether treatment to reduce fracture risk is needed. Until recently this has been a difficult decision as the interaction of risk factors is complex. However, recent guidelines have attempted to quantify absolute risk in an individual more accurately by combining age, BMD and previous fractures (NICE, 2005).

Very recently, powerful equations such as FRAX (Kannis et al, 2007 [available online at http://www.shef.ac.uk/FRAX/. Accessed 17.06.08]) have been generated which estimate fracture risk, expressed as a percentage over 10 years, in an individual. It is likely that the FRAX algorithm will be used in a similar way to the Framingham equation when used to guide the use of cardiovascular protective medications. 

For people with diabetes, these equations should be broadly applicable, except that they are likely to underestimate the risk of fracture in people with type 1 diabetes and those with longstanding type 2 diabetes, and do not incorporate the additional risk associated with glitazones. It should also be borne in mind that the FRAX estimates do not incorporate the risk of falls, which may be an important factor in people with diabetes. 

Available treatments and when to use them
Most guidelines for the treatment of osteoporosis and prevention of fractures emphasise the use of medications that improve bone strength. This is largely because interventions that attempt to reduce the risk of falls have failed to show significant benefit in terms of reducing fractures. However, these studies were undertaken in the general population, and it is likely that in people with diabetes, who have a higher risk of falls, simple interventions to address risk factors for falls could be effective. 

There are a range of drug treatments for osteoporosis. The most commonly used are the bisphosphonates (e.g. alendronate, risedronate and ibandronate). These drugs are usually taken by mouth every week or month and bind almost exclusively to bone. They interfere with the ability of bone cells to resorb bone and thus preserve its architecture. They roughly halve the risk of fracture. 

The bisphosphonates are generally well tolerated. The main side-effect is upper gastrointestinal irritation, and patients must follow the instructions carefully to ensure that the tablet is taken properly. In those with upper gastrointestinal problems, e.g. delayed gastric emptying as a result of autonomic neuropathy, these drugs would be contraindicated. An alternative in this setting is to give a bisphosphonate by infusion (e.g. ibandronate or zoledronic acid). Annual infusion of bisphosphonate is becoming popular in patients in whom oral medication is contraindicated or not tolerated. 

The main alternative to the bisphosphonates is strontium ranelate. This drug is given as a daily sachet of powder and is of similar effectiveness to bisphosphonates. There is a small increased risk of deep vein thrombosis and some people experience an altered bowel habit, but the medication is generally well tolerated.

Hormone replacement therapy (HRT) was widely used in the past to protect bones, and estrogen does have a beneficial effect on bone and reduces the risk of fracture. The main problem is that HRT is also associated with an increased risk of breast cancer, and in the groups of women most likely to benefit it increases the risk of heart disease and stroke. HRT is therefore rarely used as a bone-protective treatment in the absence of other indications, such as troublesome post-menopausal flushing. 

Less commonly used medications for osteoporosis are raloxifene, which acts like oestrogen on bone but reduces the risk of breast cancer, and parathyroid hormone (PTH) analogues that stimulate the formation of new bone. The main problem with raloxifene is that it protects against spine fractures but not other types of fracture. The main problems with PTH analogues are that they have to be given as a daily subcutaneous injection, cannot be used indefinitely and are vastly more expensive that other treatments.

The role of calcium and vitamin D as a treatment for osteoporosis is controversial. In studies where calcium and vitamin D have been used in isolation, there is little evidence that their use reduces fracture risk outside of the care home setting. However, the use of calcium and vitamin D is encouraged when other treatments are given.

The decision to offer medication depends on the factors discussed above. They may be summarised as follows:

• In general, people with diabetes over the age of 75 years who have a low trauma fracture should be considered for treatment as they are at high risk of further fracture by virtue of their age and their previous fracture (NICE, 2005). 
• Younger people with fractures should be considered for treatment if they have osteoporosis on a bone density scan. 
• In the absence of fractures, individuals over the age of 65 years who have any of the risk factors described above should be considered for DXA screening. 

It is hoped that the use of more sophisticated tools such as FRAX will make it easier to screen women for fracture risk on the basis of their history; this might be used to decide who would benefit from a DXA examination. 

The increased risk of falls should not be forgotten as virtually all fractures are associated with a fall. Medications are a common correctable cause of falls and should be reviewed to ensure that they are not leading to postural hypotension or inappropriate sedation. Foot problems should be sought and treated. In patients with peripheral neuropathy, it is important to encourage the wearing of  appropriate footware, not just for ulcer prevention but also to improve stability. Vision should be assessed, and treatable causes of visual impairment addressed. 

Three case reports that illustrate the factors influencing the decision to give bone-active treatment are shown in Box 1 (on the following page).

Conclusion
We now know that some people with diabetes have an increased risk of fracture. Investigation and treatment should have a multidisciplinary focus with consideration of pharmacological therapy and an attempt to find and correct factors that predispose to falls.