Commentary |
Commentary: Statins and fracture—why the confusion?
1 MRC Epidemiology Resource Centre, University of Southampton, Southampton, UK
2 Department of Rheumatology, Southampton General Hospital, Southampton, UK
* Corresponding author. Department of Rheumatology, Southampton General Hospital, Tremona Road, Southampton, SO16 6YD, UK. E-mail: cedwards{at}soton.ac.uk.
In 1999 a paper published in Science suggested that 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors (statins) increased bone formation in rodents by effects on a potent bone forming cytokine bone morphogenetic protein-2 (BMP-2).1 Subsequent studies have suggested that statins increase bone mineral density (BMD) in humans2 and have investigated whether statin use is associated with reduced fracture risk. The fracture risk studies include four case–control studies, five cohort studies in which fracture end points where documented, and the post hoc analysis of the effect of statins on fracture in two cardiovascular outcome studies. To date there are no prospective randomized trials specifically designed to address the effect of statins on fracture risk.
The results of these studies have been variable; some demonstrating a significant reduction in fracture risk while others found no significant reduction. A meta-analysis of observational studies has suggested an association between statin use and reduction in hip fractures.3 However, post hoc analyses of randomized controlled trials (RCTs) of statin use for cardiovascular disease have not supported a reduction in fracture risk. Can these disparate results be explained by the different epidemiological techniques used in each of these studies?
To help us answer this we are fortunate to have two observational studies published by two separate groups using the same General Practice Research Database (GPRD) dataset, with different results. Meier et al.4 described a significant reduction in hip fracture associated with statin use: OR 0.12 [95% confidence interval (95% CI) 0.04–0.4] and Van Staa et al.5 found no significant effect of statins on hip fracture risk: OR 0.59 (95% CI 0.31–1.13). The design of these studies varied in a number of important ways allowing us to examine the effects of trial design on results. These issues have been explored in detail by de Vries et al.6 and give a unique insight into how different results can be generated from the same dataset.6 They describe a number of variations in the approach of the two groups that may have contributed to the different results including selection criteria, case definition, age matching, and duration of statin usage.
Van Staa selected cases and controls from the entire GPRD dataset, whereas Meier selected only from cohorts of patients prescribed a statin or other lipid-lowering drug, or with a diagnosis of hyperlipidaemia and a random sample of controls. Although this did not appear to produce major differences in the patients' clinical characteristics, the selected recruitment method of Meier did lead to problems in age matching the cases and controls, leading to a potential residual confounding effect of age. This difference appeared to contribute substantially to the differences in results between the two studies. It is also possible that case selection that included individuals at very high risk of fracture as done by Van Staa et al. might attenuate any effect of statins on fracture risk.
Another important issue was the inclusion criteria relating to statin usage. De Vries demonstrated that taking 6 months, as opposed to 30 days, as the definition of statin exposure attenuated the reduction in fracture risk. It is vital when defining inclusion criteria to consider biological plausibility: RCTs of bisphosphonates suggest that 6 months is the earliest time at which fracture reduction can be demonstrated. It seems likely that statins would take at least as long to have an effect on bone.
There are also a number of potential reasons for the discrepancies seen between the results of RCTs and observational studies of statin use and fracture. The observational studies may be subject to a healthy user bias whereby patients who are in better health, and therefore at a lower risk of fracture, are more likely to request, be prescribed, and persist with statins. This bias may explain any observed fracture reduction in the observational studies. A similar bias has previously been demonstrated in the studies of hormone replacement therapy and cardiovascular disease.7 Additionally, they may be subject to residual confounding for factors such as socioeconomic status, diet, and physical activity. Alternatively, it is possible that the RCTs underestimated the benefits of statins on fracture. These studies were designed to assess cardiovascular outcomes and as such recruited patients based on cardiovascular risk profiles who are likely to be at a relatively low risk of osteoporotic fracture. Published RCTs of bisphosphonates, which also act on the mevalonate pathway, suggest that their benefit is greatest in those at high risk of fracture, defined by prevalent fractures or low BMD, but of only limited benefit in patients with normal BMD.8 Additionally, the LIPID study assessed the use of Pravastatin9 and unlike other statins there is little evidence, from in vitro studies, that pravastatin has any effect on bone, possibly owing to its high water solubility.
So what of statins and fracture? The answer remains elusive and can only be definitively answered with a formal RCT of statins with fracture as the primary outcome. Interestingly the majority of the analyses reported by de Vries did confirm a reduction in hip fracture.6 In the meantime, observational studies provide epidemiological signals suggesting a reduction in fracture risk, whereas the RCT data suggest that the effect in a population defined by cardiovascular risk is at best small. It is important, however, not to over interpret such post hoc analysis of RCTs, which although providing the highest scientific rigour are often limited by their lack of generalizability. We cannot, therefore, rule out a beneficial effect in patients at high risk of osteoporotic fracture.
The story of statins and bone/fracture provides researchers with an opportunity to reflect on the degree to which differences in study design can change the conclusions reached. It also illustrates how important an understanding of the underlying biological mechanisms is in establishing study protocols so that informed decisions can be made. This should encourage epidemiologists and laboratory scientists to collaborate more closely when determining the questions to be asked and how to ask them.
 |
References |
|---|
 Top References |
|---|
1 Mundy G, Garrett R, Harris S et al. Stimulation of bone formation in vitro and in rodents by statins. Science 1999;286:1946–49.
2 Edwards CJ, Hart DJ, Spector TD. Oral statins and increased bone-mineral density in postmenopausal women. Lancet 2000;355:2218–19.[CrossRef][Web of Science][Medline]
3 Bauer DC, Mundy GR, Jamal SA et al. Use of statins and fracture: results of 4 prospective studies and cumulative meta-analysis of observational studies and controlled trials. Arch Intern Med 2004;164:146–52.
4 Meier CR, Schlienger RG, Kraenzlin ME, Schlegel B, Jick H. HMG-CoA reductase inhibitors and the risk of fractures. JAMA 2000;283:3205–10.
5 van Staa TP, Wegman S, de VF, Leufkens B, Cooper C. Use of statins and risk of fractures. JAMA 2001;285:1850–55.
6 de Vries F, de Vries C, Cooper C, Leufkens B, van Staa T-P. Reanalysis of two studies with contrasting results on the association between statin use and fracture risk: the General Practice Research Database. Int J Epidemiol 2006;35:1301–08.
7 Pedersen AT, Ottesen B. Issues to debate on the Women's Health Initiative (WHI) study. Epidemiology or randomized clinical trials—time out for hormone replacement therapy studies? Hum Reprod 2003;18:2241–44.
8 Cummings SR, Black DM, Thompson DE et al. Effect of alendronate on risk of fracture in women with low bone density but without vertebral fractures: results from the Fracture Intervention Trial. JAMA 1998;280:2077–82.
9 Reid IR, Hague W, Emberson J et al. Effect of pravastatin on frequency of fracture in the LIPID study: secondary analysis of a randomised controlled trial. Long-term Intervention with Pravastatin in Ischaemic Disease. Lancet 2001;357:509–12.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||