Warfarin: Blood-Thinner and Bone-Thinner?

Abstract & Commentary

By Allan J. Wilke, MD, Residency Program Director, Associate Professor of Family Medicine, University of Alabama at Birmingham School of Medicine—Huntsville Regional Medical Campus. Dr. Wilke reports no financial relationship to this field of study.

Synopsis: Long-term use of warfarin is associated with an increase in osteoporotic fractures.

Source: Gage BF, et al. Risk of osteoporotic fracture in elderly patients taking warfarin: results from the National Registry of Atrial Fibrillation 2. Arch Intern Med. 2006;166:241-246.

Bleeding is the most common adverse effect of warfarin use, but there are also reports linking it to osteoporosis and fractures in adults1 and children.2 Not all studies have confirmed this.3 The National Registry of Atrial Fibrillation II is a data set that includes Medicare Part A claims records, Medicare Part B records, and hospital chart abstracts from across the United States. The data were collected from 1995 to 1999. Gage and colleagues performed a retrospective cohort study to examine the relationship of warfarin (WARF) use to osteoporotic fractures (OF). They selected a stratified random sample of Medicare patients who were hospitalized with atrial fibrillation (AF). They looked at patients who had sustained vertebral, wrist, hip, and closed rib fractures, which are often related to osteoporosis. They identified patients likely to be taking WARF by searching for claims for prothrombin time (PT) charges. These patients were stratified into those with at least 1 year of use (long term) and those with 90 to 364 days of use (short term). They used multiple PT charges as a surrogate for continued WARF use. After applying reasonable inclusion and exclusion criteria, they identified 4652 patients with long-term WARF use, 1905 with short-term use, and 8007 with no use in the last 3 years. The patients were, on average, age 79 years, predominantly white, and more than half female. Thiazide diuretics, loop diuretics, and b-blockers were the most commonly prescribed additional medications. Of the total 14,564 subjects, 1005 (6.9%) had an OF. The 30-day mortality after suffering a fracture was high, ranging from 24.8% for wrist fractures to 39.0% for hip fractures. Gage et al performed a logistic regression on the 13,881 subjects who survived more than 30 days after the index hospitalization. The odds ratio (OR) of having an OF was significantly increased for long-term WARF use (OR, 1.25; 95% CI, 1.06-1.48). This did not hold true for short-term use (OR, 1.03; 95% CI, 0.82--1.29). Other independent risk factors for fractures were increasing age, high fall risk, hyperthyroidism, neuropsychiatric disease, and alcoholism. Protective factors included black race, male gender, and b-blocker use. When WARF use by gender was examined there was a statistically significant risk for men (OR, 1.63; 95% CI, 1.26-2.10), but not for women (OR, 1.05; 95% CI, 0.88-1.26).


Does long-term WARF use place your patient at risk for OF? This is a retrospective, observational, cohort study, so no firm conclusions can be made. We will have to wait for a prospective, randomized, control trial of sufficient size before we can claim anything more than an association. However, increasing age, high fall risk, hyperthyroidism, and alcoholism are acknowledged risk factors for OF; and black race and male gender ameliorate the risk.4-6 Previous studies (reviews published in Internal Medicine Alert7,8) have looked at the protective association of thiazide diuretic use and b-blocker use on OF. This study also supports the benefit of b-blocker use. The authors suggest that neuropsychiatric disease contributed to the increase in fractures through an increased risk of falls secondary to medication use. The basic science behind the risk of fracture from WARF use is that it blocks the y-carboxylation of glutamic acid. Besides residing on clotting factors II, VII, IX, and X, glutamic acid residues are also found on osteocalcin and other bone matrix proteins. Blocking y-carboxylation results in decreased bone formation and increased bone resorption.

This study does raise some interesting questions. Why was the risk significant in men, but not in women? The authors speculate that other factors (most likely pre-existing osteoporosis) had a greater influence on OF in women. Does WARF wreak its damage biochemically through y-carboxylation blockade, or do people taking WARF modify their diets in ways that are not bone-healthy? The authors wonder if by limiting their intake of leafy, green vegetables, a rich source of vitamin K and folic acid, they are controlling fluctuations in their PT/INR at the expense of becoming folate-deficient. A folate-rich diet helps prevent hyperhomocysteinemia, which is also associated with OF.

Assuming for the moment that WARF causes osteoporosis, how can a physician use the information from this study to best treat his or her patients? Are there alternatives to WARF for the treatment of AF? A Cochrane Review11 concluded that WARF is the best agent in patients with AF at average or greater risk of stroke, but that aspirin might be as useful in patients at low risk. Can the effects of WARF in osteoporosis be ameliorated? While there are no studies to address this issue, it would be prudent to monitor patients (especially males) taking WARF for osteoporosis with DXA scans and to promote general bone health measures, including calcium and vitamin C intake, smoking cessation, moderation in alcohol use, and physical exercise. Avoiding or discontinuing unnecessary psychotropic medication is always wise. Whether adding an antiresorptive agent (bisphosphonates, raloxifene, estrogen, or calcitonin) would be useful is a matter of speculation, but certainly worth considering.


1. Caraballo PJ, et al. Long-term use of oral anticoagulants and the risk of fracture. Arch Intern Med. 1999; 159:1750-1756.

2. Barnes C, et al. Reduced bone density in children on long-term warfarin. Pediatr Res. 2005;57:578-581.

3. Jamal SA, et al. Warfarin use and risk for osteoporosis in elderly women. Study of Osteoporotic Fractures Research Group. Ann Intern Med. 1998;128:829-832.

4. Cummings SR, Melton LJ. Epidemiology and outcomes of osteoporotic fractures. Lancet. 2002;359:1761-1767.

5. Heinemann DF. Osteoporosis: an overview of the National Osteoporosis Foundation clinical practice guide. Geriatrics. 2000;55:31-36.

6. Boelaert K, Franklyn JA. Thyroid hormone in health and disease. J Endocrinol. 2005;187:1-15.

7. Wilke AJ. Hip Pointers. Internal Medicine Alert. 2003; 25:145-146.

8. Wilke AJ. ß-Blockers Build Better Bones. Internal Medicine Alert. 2004;26:153-154.

9. Schoofs MW, et al. Thiazide diuretics and the risk for hip fracture. Ann Intern Med. 2003;139:476-482.

10. Schlienger RG, et al. Use of beta-blockers and risk of fractures. JAMA. 2004;292:1326-1332.

11. Segal JB, et al. Anticoagulants or antiplatelet therapy for non-rheumatic atrial fibrillation and flutter. Cochrane Database Syst Rev. 2001;(1):CD001938.