Much attention has focused recently on the fracture risk associated with the diabetes drug rosiglitazone (brand name Avandia). In the study known as A Diabetes Outcome Progression Trial (ADOPT), women with Type 2 diabetes who were randomly assigned to take rosiglitazone, rather than metformin or glyburide (two other diabetes drugs) had approximately twice the incidence of bone fracture. Based on warnings issued by their manufacturers, the Food and Drug Administration issued a Safety Alert in March 2007 stating that rosiglitazone and pioglitazone (Actos), a drug in same drug class as rosiglitazone, may be related to higher fracture rates in women. (Pioglitazone and rosiglitazone are in a class of drugs called the thiazolidinediones.)
While these findings are important, many studies showing a higher fracture risk in people with Type 2 diabetes were done before thiazolidinediones were available. Therefore, there must be other factors associated with Type 2 diabetes that increase the risk of fracture. Future research must determine the reasons for this increased fracture risk, as well as how to reduce it beyond avoiding medicines that may increase it even further.
Type 1 diabetes and bone mineral density
In one of the meta-analyses mentioned earlier, older age and longer duration of diabetes were found to correspond with lower BMD in the spine, though not in the hip, of people with Type 1 diabetes. Although no relationship was found between BMD and either body-mass index or HbA1c level (an indicator of blood glucose control), older age and longer diabetes duration often signify greater diabetic complications, which could be responsible for the result. Lower BMD in people with Type 1 diabetes does not completely explain the higher fracture risk that has been observed.
Type 2 diabetes and bone mineral density
Higher BMD in people with Type 2 diabetes, as stated earlier, does not protect against a higher fracture risk. One possible explanation for the higher BMD associated with Type 2 diabetes is simply that people with Type 2 diabetes tend to have a higher body weight and therefore a greater load on the skeleton. (In other words, carrying more body weight turns almost any activity into weight-bearing exercise, which is known to preserve bone density.) In one meta-analysis, a higher body-mass index in people with Type 2 diabetes was associated with higher BMD in both the hip and spine, while longer diabetes duration was associated with lower hip BMD. Again, longer diabetes duration may simply signify that a person has more diabetic complications. No relationship was found between BMD and either age or HbA1c level in people with Type 2 diabetes.
Few studies exist on BMD changes over time for either Type 1 or Type 2 diabetes. The available data demonstrate, however, that older white women with Type 2 diabetes have more rapid bone loss than those without diabetes. No such difference has been found in white men, black men, or black women. In one study, women with Type 2 diabetes showed greater loss of BMD for each year they took a thiazolidinedione drug, even when other factors that may affect bone loss were taken into consideration. No such effect was found in men. Another study of just men with Type 2 diabetes did find that those taking rosiglitazone lost more BMD than those not taking the drug, but this study did not take into account other factors that could affect bone loss, as the previous study did.
Unlike the study of rosiglitazone that led to new FDA warnings, both of these studies were observational, rather than being randomized clinical trials. Only a randomized trial (in which participants are randomly assigned to one of the treatments being compared) can eliminate the influence of factors that might make a person more likely to receive one treatment over another in a real-world situation. Since people who take thiazolidinedione drugs may tend to be different in unknown ways from those who do not, an observational study cannot determine the effect of these drugs on BMD with any statistical certainty.
