Bone fractures less likely in patients with hyperglycemia

Middle-aged patients with high two-hour glucose levels are at a lower risk for fractures.

Issue: April 2008

The risk for low-energy fractures is reduced as a result of impaired glucose tolerance, according to recent study results.

Given the link between type 1 diabetes and low BMD and increased fracture risk, researchers at Malmo University Hospital in Sweden aimed to determine the association between fracture risk and glucose tolerance in middle-aged adults by conducting the Malmo Preventive Project.

The researchers conducted physical analyses, questionnaires, laboratory analyses, fasting blood glucose tests and oral glucose tolerance tests. They measured fasting plasma insulin and identified fractures as low- or high-energy in 22,444 men aged about 44 years and 10,902 women aged about 50 years. Participants were then separated into quartiles based on their levels of fasting blood glucose. The researchers conducted follow-up to investigate incident fractures at about 19 years for men and about 15 years for women.

The researchers recorded fractures among 1,471 men, 1,246 of whom suffered low-energy fractures. Four hundred and thirty-eight men suffered multiple fractures (mean: 2.8 fractures, range 2-10), and 808 (65%) suffered only one.

Fractures occurred in 1,271 women, and 1,236 suffered the low-energy type. Three hundred fifty-six women experienced multiple fractures (mean: 2.5 fractures, range 2-20), and 880 (71%) suffered one. Despite smoking, which was more prevalent in those with multiple fractures, both groups were similar at baseline. Within the entire fracture population, 682 (55%) osteoporotic fractures were reported in men, and 939 (76%) were reported in women.

Fasting blood glucose greater than 6.1 mmol/L was reported in 879 (3.9%) of 22,100 men whose fasting blood glucose was measured and 365 (3.4%) of 10,730 women. Although BMI and body weight increased with each quartile of fasting blood glucose, mean age and body height were similar among quartiles in both sexes, according to the study results.

Two hundred and twenty-three (1.7%) of the 13,046 men who underwent OGTT had two-hour blood glucose levels greater than 10.0 mmol/L, along with 275 (4.7%) of 5,888 women tested. These participants were of a somewhat higher mean age than the entire cohort, according to the researchers.

The quartiles for two-hour glucose were similar to those for fasting blood glucose in that the highest mean body weight and BMI were seen in the fourth quartile of both sexes. The researchers also found that in the first quartile of two-hour glucose, especially in women, the proportion of participants with fractures was higher. In both sexes, the first quartiles had the highest amount of smokers.

Fracture risk evaluated

The researchers found that after adjusting for BMI, age and smoking, the low fracture risks experienced in the second to fourth quartile of fasting blood glucose in men were no longer significant.

A reduced risk for multiple fractures was linked to two-hour glucose levels greater than 4.2 mmol/L post-OGTT in men in the second, third and fourth quartiles (OR=0.49-0.66). After adjusting for age, smoking and BMI, the remaining reductions were present in the second and third quartiles (OR=0.57-0.71), according to the study results.

Although women in the fourth quartile experienced the greatest reduction in risk for multiple fractures, those in the second, third and fourth quartiles with two-hour glucose levels exceeding 5.4 mmol/L were also at a reduced risk. A decrease in the risk for osteoporotic fractures was also linked to the third and fourth quartiles (OR=0.53-0.77).

Unlike the cohort of men, risk for multiple fractures in the third and fourth quartiles and for osteoporotic fractures in the fourth quartile remained low, even after adjusting for BMI, age and smoking.

After measuring fasting plasma insulin, the researchers found no significant links between fracture risk and serum insulin. – by Stacey L. Adams

For more information:
Holmberg AH, Nilsson PM, Nilsson J-Å, Åkesson K. The association between hyperglycemia and fracture risk in middle age. A prospective, population-based study of 22,444 men and 10,902 women. J Clin Endocrinol Metab. 2008;93:815-822.

This is a fascinating article. The February issue of Calcified Tissue International (2008;82:87-91) has a lead article of which the opening sentence reads as follows: “There is increasing evidence that the risk of fractures is increased in older women with type 2 diabetes.” They list nine supporting references. I don’t doubt the data here in the Holmberg study; the data in this paper are good, but they vary from a lot of what is out there. One thing that typifies diabetes, compared with nondiabetes, is fragility. Fractures generally occur at a higher BMD in patients with diabetes. The researchers provide no bone density data for their population.

What the Calcified Tissue International paper shows is that the measurement of bone using ultrasound of the heel is a better determinant of fracture risk than a DEXA measure of bone density at the hip. The hypothesis in the Holmberg article is that patients with diabetes have peripheral neuropathy; this may increase their risk for falling. In this article, there is not much information about the patients and their diabetes and the complications of their diabetes; all we are given are their blood sugar values. The researchers include data on height and weight and so on, but there is no concept of the patients' complications of diabetes. It may be that they fall more than normal; that's a possibility. Unfortunately, we don't know why bone density is higher in patients with diabetes, but if BMD is higher in diabetes and a patient is not falling, then maybe their risk for fractures is indeed lower. If their BMD is higher and their risk for falling is high, then their risk for fracture increases. So the data from these two articles are not necessarily inconsistent; the researchers are simply looking at different aspects. It would be important to know what the complications in diabetes were in the Holmberg study.

There has been a lot of recent information indicating that thiazolidines are associated with an increased risk for fractures. So, the link between diabetes and fracture is important. This article says it’s not real, and I’m not sure that I fully accept that yet. But, the paper is very compelling, and if you have complications in diabetes and you fall more, then you may be at increased risk.

Practicing clinicians should do whatever they can, at all times, to make sure their patients' diabetes is as well controlled as possible, to check for complications of diabetes as often as practical and to take particular precaution in patients who have peripheral neuropathy associated with diabetes.

– Michael Kleerekoper, MD, MACE

Endocrine Today Editorial Board member