Statins and Osteoporosis: A Clinical Review

Michael J. Gonyeau, Pharm.D.


Pharmacotherapy. 2005;25(2):228-243. 

In This Article

Surrogate Marker Trials

To determine the potential role of statins in the prevention or treatment of osteoporosis, their clinical efficacy ideally should be compared with the standard of care. Studies evaluating surrogate markers of osteoporosis, including BMD, serum osteocalcin, serum bone-specific alkaline phosphatase, urine deoxypyridinoline (DPD), and urine cross-linked N -telopeptides of type I collagen (CrossLaps; Nordic Bioscience Diagnostics, Norfolk, VA) were reviewed and are summarized in Table 1 .[26,40,41,42,43,44,45]

One group[26] conducted an observational case-control study of statin effects on BMD from participants of the Chingford study.[46] The primary end point was BMD in statin users versus that in controls. Statin use was identified in 41 patients at the time of dual x-ray absorptiometry, and patients were matched with 2-3 controls (100 subjects) based on age and date of examination. Bone mineral densities at the hip and spine were compared. The most commonly used statin was simvastatin (51%), followed by pravastatin (24%), atorvastatin (15%), and fluvastatin (10%), although no dosages were provided. The authors found statistically significant increases in BMD in statin users compared with controls (hip 0.76 vs 0.68 g/cm2, p=0.05; spine 0.99 vs 0.91 g/cm2, p=0.001); these increases remained significant after adjustment for confounding factors such as age, height, weight, use of HRT, and smoking status. Although the authors did adjust for potential confounders, this trial is limited by the fact that it was open label and evaluated a patient population already enrolled in an observational study. This increases the risk for selection and observer bias and, along with the small sample, limits our ability to extrapolate the results to the entire population.

In a retrospective case-control study of 69 patients with type 2 diabetes mellitus, 33 control subjects were matched to 36 patients taking statins (lovastatin 20 mg, 25%; pravastatin 10 mg, 39%; and simvastatin 10 mg, 36%) on the basis of age, sex, body weight, postmenopausal status, and fasting blood glucose levels.[40] No subjects were taking any drugs that would affect bone metabolism. The primary end point was the comparison of BMD measurements at baseline and at 15 months.

No significant differences were noted in baseline characteristics or BMD scores. After 15 months, decreases in BMD in the lumbar spine (-0.035 g/cm2, p<0.05), femoral neck (-0.009 g/cm2, p=NS), and total hip (-0.01 g/cm2, p=NS) were observed in the control group. Statin-treated patients experienced an increase in BMD in the femoral neck (+0.025 g/cm2, p<0.05 vs control) and total hip (+0.014 g/cm2, p<0.05 vs control), but a nonsignificant decrease in BMD in the lumbar spine. No significant differences were noted in BMD changes among the three statins prescribed. However, almost all of the statistically significant BMD increases in those who took statins occurred in male patients. A possible explanation for the sex difference was theorized, as osteoporosis in men is usually the result of decreased osteoblast function, whereas women's primary defect occurs with increased bone resorption due to estrogen loss during menopause. Although this study showed a benefit with statins not likely due to chance, the small and very specific population, the retrospective design, as well as the disparity of results in the female population limit the ability to extrapolate these data to a larger patient base.

In one of the first prospective studies on this topic, researchers evaluated, in an open-label fashion, 17 nonosteoporotic Chinese patients treated with simvastatin 20 mg/day orally for 4 weeks.[41] The primary end point was alteration in biochemical bone markers, including levels of serum osteocalcin, serum bone-specific alkaline phosphatase, DPD, and urine CrossLaps. After controlling for disease states and drugs that affect bone metabolism, hypercholesterolemic subjects had fasting blood and urine samples collected before and after 4 weeks of simvastatin therapy. All lipid parameters significantly decreased over the treatment period. The only bone metabolism marker to reach statistical significance was serum osteocalcin level, which nearly doubled (from 15.6 to 28.9 B5g/L, p=0.021) during the 4 weeks of statin administration.

Although exclusion criteria were explicitly stated for this trial, patient demographics were not reported. The population studied may also influence results, as Asians are less prone to developing osteoporosis than are Caucasians. Also, the duration of treatment may not have been long enough to see effects on other bone markers, although the researchers stated the time period was derived from animal analyses that showed increased bone growth with statins in 35 days.[33] This information cannot necessarily be correlated with human response. Although these results were statistically significant, increases in osteocalcin concentrations occurred in only 8 (47%) of the 17 subjects. No significant differences were found with regard to sex, age, or changes in lipid levels; however, noting the sex of each subject to assess any differences as seen in previous studies may add to our body of knowledge.

In contrast to the previous study, another group performed a prospective, single-center, randomized, open-label trial of 64 elderly, hyperlipidemic women with osteoporosis.[42] All participants were randomly assigned to 12 weeks of treatment with fluvastatin 40 mg plus vitamin C, or with vitamin C alone. The primary end point assessed was any change in serum and urine bone metabolic markers at baseline and the end of the treatment period. No differences were noted in baseline characteristics of the study subjects, and no statistically significant differences in any bone parameter were observed except for serum CrossLaps (3736 vs 4718 pmol/L, p<0.05; Table 1 ). As with the first study discussed in this section,[26] this trial[42] is limited by its small sample. The authors also state that the dosage they used may have been too low to show significant effects and that any statin produced specifically for effects on bone would have to be manipulated to increase the amount of drug available at the site of action.

The purpose of another prospective case-control study of 620 patients from the Kupio Osteoporosis Risk Factor and Prevention trial[43] was to evaluate if statin therapy could prevent early postmenopausal bone loss. Patients were included if baseline and follow-up dual x-ray absorptiometry measurements were available and information on individual characteristics and statin therapy could be obtained. The 118 patients taking statins (lovastatin 31%, simvastatin 42%, fluvastatin 10%, atorvastatin 9%, and pravastatin 8%; dosages not recorded) were compared with 360 women who did not report hyperlipidemia or statin use and 142 women who self-reported hyperlipidemia for which they took no specific treatment. No statistically significant differences were noted between those taking statins and those not taking statins in the annual percentage BMD change in the femoral neck (-0.47% vs -0.33%, p=0.628) or lumbar spine (-0.20% vs 0.47%, p=0.134). An interesting finding noted by the researchers is the potential for hyperlipidemia itself to be protective against osteoporosis and fractures, as the smallest annual bone loss and greatest gain of lumbar bone were seen in hyperlipidemic control subjects.

This study has several limitations. This is an observational subanalysis of a larger trial, which increases the risk for population and observer bias. The trial also relied on patient self-reporting for medical and drug histories, although attempts were made to verify drug usage. The p values of BMD changes at the spine and hip reported in the study seem to compare all those who took statins with those not taking statins, but only 45% of statin users were continuously taking statins during the study period, so this may not be an accurate determination of statin effects. Although the authors did account for many confounders in their analyses, it is curious that HRT did not affect BMD measurements. This may indicate that the study was not powered to detect differences in BMD measurements.

Another group added to the body of evidence on this topic by conducting a single-center, retrospective, observational study of 983 patients at a Veterans Affairs medical center.[44] The main objectives were to determine any statin effects on BMD at the hip and spine and subsequent development of osteoporosis. Diseases and drugs affecting bone metabolism were accounted for, although specific statins and dosages were not recorded. A protective effect of statins in the development of osteoporosis in men (odds ratio [OR] 0.79, 95% confidence interval [CI] 0.51-1.22 and in women (OR 0.36, 95% CI 0.12-1.07) was observed, but it was not statistically significant. The BMD scores were significantly increased in the spine for men taking statins versus men not taking statins (1.05 vs 0.97 g/cm2, p=0.004) after 4 years, but no other BMD parameters reached statistical significance for men or women after adjusting for potential confounders.

Because this was a retrospective trial, evaluation of nonpharmacologic agents could not be evaluated. Another potential drawback is the predominance of men in the study population (71%). Selection bias may also be a factor, as patient consideration was dependent on those who may have been referred for BMD testing based on suspicion of osteoporotic disease. An interesting note is that the researchers never did a combined analysis of both sexes to evaluate the overall effect of statins in this population. One important limitation of this study is the lack of adjustment for smoking status, a known risk factor for osteoporosis.

Other investigators performed a prospective, cross-sectional, case-control trial of 280 postmenopausal women: 140 of the women were treated with a statin (simvastatin median dose 20 mg [65%], atorvastatin median dose 20 mg [15%], lovastatin median dose 20 mg [8.6%], pravastatin median dose 40 mg [5.7%], fluvastatin median dose 40 mg [5%], and cerivastatin median dose 0.1 mg [0.7%]) for 2 or more years, and the other 140 women served as age- and sex-matched control subjects.[45] The primary end point of the trial was to evaluate the effect of statins on BMD and bone-turnover markers. No statistically significant differences were noted between the case patients and the control subjects with regard to BMD measure-ments of the hip, spine, and forearm (0.84 vs 0.85 g/cm2, p=0.71; 0.89 vs 0.9 g/cm2, p=0.34; 0.46 vs 0.45 g/cm2, p=0.62, respectively). However, a significant 11% decrease was observed in serum CrossLaps, a marker of bone resorption (0.44 vs 0.51 mg/L, p<0.01). In contradiction to previously mentioned studies, a statistically significant decrease in markers of bone formation in case patients versus control subjects was observed (serum osteocalcin -9%, 36 vs 39.5 mg/L, p=0.03; and bone alkaline phosphatase -14%, 98 vs 109 U/L, p<0.01) The authors concluded that statins do affect markers of bone metabolism, but they cite the mechanism to be primarily through antiresorptive properties rather than increased bone formation.

One possible reason for the discrepancy in results in this study when compared with previous evaluations is the control of thiazide diuretic use in this population. Thiazides have been shown to be protective against osteoporosis by increasing BMD.[47,48,49] Another important point that differentiates this trial from others is that the population studied did not have osteoporosis, which may minimize any effects of statins.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.