Osteoporosis International

, Volume 16, Issue 8, pp 990–998

Hydroxymethylglutaryl-coenzyme A reductase inhibitors and osteoporosis: a meta-analysis

Authors

  • Christos Hatzigeorgiou
    • Tripler Army Medical Center
  • Jeffrey L. Jackson
    • Medicine-EDP
    • Uniformed Services University
Original Article

DOI: 10.1007/s00198-004-1793-0

Cite this article as:
Hatzigeorgiou, C. & Jackson, J.L. Osteoporos Int (2005) 16: 990. doi:10.1007/s00198-004-1793-0

Abstract

Studies determining the association between hydroxymethylglutaryl-coenzyme A (HMG-CoA) reductase inhibitors (statins) and bone metabolism are mixed. We conducted a systematic review to assess the potential impact of statins on fractures, bone mineral density and bone markers. We searched Medline, Embase, the Cochrane Library, and Federal Research in Progress (FEDRIP). Inclusion criteria consisted of human studies with measurable outcomes, which were rated as good or fair according to the United States Preventive Services Task Force (USPSTF) criteria. The effects of statins on bone mineral density (BMD), bone markers and fracture risk were independently extracted by two reviewers and were combined by use of a random-effects model. The 31 analyzed studies included 24 observational studies and seven randomized controlled trials. Overall, statin use was associated with fewer hip fractures (OR 0.60, 95% CI 0.45–0.78) and improved hip BMD (Z score 0.12, 95% CI 0.05–0.19), with a non-significant reduction in vertebral fractures and no effect on vertebral BMD. In subgroup analysis of studies that involved only women there was a reduction in hip fractures (OR 0.75, 95% CI 0.60–0.95) and improvement in hip BMD (Z score 0.11, 95% CI 0.04–0.18). Vertebral BMD was unchanged, and only one study reported on vertebral fractures, finding improvement. Statins had only small effects on bone markers, with a decrease in alkaline phosphatase [standardized mean difference (SMD) −0.18, 95% CI −0.34 to −0.01], an increase in NTX (SMD 0.39, 95% CI 0.07–0.71), with no effect on osteocalcin or CTX. The statistically significant improvement in hip fracture risk was seen only in case–control trials, not in either the eight prospective trials or the two randomized controlled trials (RCTs). Statins may have a beneficial impact on bone metabolism and fracture risk; randomized controlled trials are needed to explore this association.

Keywords

Bone mineral densityBone markerFractureHMG-CoA reductase inhibitorsMeta-analysisOsteoporosisStatin

Introduction

Hydroxymethylglutaryl-coenzyme A (HMG-CoA) reductase inhibitors (statins) are one of the most widely prescribed drugs, with greater than 3 million Americans taking one daily for the prevention of cardiovascular events. The beneficial effects of HMG-CoA reductase inhibitors in reducing cardiac events was initially believed to be secondary to low density lipoprotein (LDL) reduction, but increasing evidence suggests that the statins may have beneficial lipid-independent cardiovascular effects as well. They reduce markers of inflammation, including C-reactive protein [1, 2, 3, 4], increase hyperemia arterial flow dilatation [3], reduce TNF-alpha [3], and may exert an as-yet unexplained protection among patients with cytomegalovirus antibody positivity [2]. Statins also appear to lower markers of coagulation, systemic inflammation and soluble cell adhesion [5].

There is emerging evidence that HMG CoA reductase inhibitors may also have the potential to improve bone health. Mundy et al. first reported, in 1999, that statins stimulated bone formation in vitro [6]. Statins increase bone mineral protein 2 (BMP-2) mRNA and the production of BMP-2 by osteoblasts; BMP-2 enhances osteoblast differentiation [6]. This effect can be blocked by the addition of noggin, an inhibitor of BMP-2, or by the addition of mevalonate [7]. Statins inhibit the conversion of 3-hydroxy-3-methylglutaryl-CoA to mevalonate, a step in the production of cholesterol. In addition to effects on osteoblasts, mediated by BMP-2, there is also evidence of an inhibitory effect on osteoclasts [8]. Simvastatin produced a 15%–33% reduction in osteoclast number [6]. Others have found that mevastatin inhibits osteoclast differentiation [9]. In-vivo and in-vitro studies with human bone cultures and rodents have shown increases in bone growth ranging from 30% to 94% [8]. There have been a number of human studies investigating the effect of statins on bone markers, bone mineral density (BMD), and fractures, with mixed results, some studies showed benefit, others none. We therefore performed a systematic review of the literature, asking, “Is the use of an HMG-CoA reductase inhibitor associated with improvement in bone mineral density, biochemical markers of bone metabolism, or a reduction in fracture rates?”

Methods

We searched Medline (1966–July 2004) and Embase, using text, medical subject headings (MeSH) and key words (all languages): hydroxymethylglutaryl-CoA reductase inhibitors, osteoporosis, fractures, and bone mineral density. We searched the Cochrane registry for clinical trials and the Cochrane Database of Systematic Reviews (DARE) for systematic reviews. A search of the Federal Research in Progress (FEDRIP) database was conducted to identify unpublished literature. We also searched the references of review articles for additional articles missed by the computerized database search. Studies were screened for inclusion through review of the abstract or the published article if the abstract was unclear. We screened the articles based on the following criteria: human studies that were observational or randomized controlled trials with measurable outcomes reported. Inclusion was judged independently by two reviewers (C.H., J.L.J.), with no disagreement.

No authors were contacted for additional or unpublished data.

We assessed study quality with a component analysis instrument used by the USPSTF for rating internal validity [10]. Trials and studies were rated overall as good, fair, or poor; poorly rated trials were excluded. Two reviewers (C.H., J.L.J.) independently assessed study quality. Disagreements were arbitrated by consensus.

Abstracted data included setting, country of origin, HMG-CoA characteristics (type of statin, dose, and duration), demographics, number of participants enrolled, and outcomes. Outcomes were abstracted as continuous (hip and spine bone mineral density and bone markers) or dichotomous (hip, spine, and overall fracture risk). Abstracted bone markers included two formative bone markers (osteocalcin and bone alkaline phosphatase) and two resorptive bone markers (c-telopeptide of type I collagen (S-CTX) and n-telopeptide of type I collagen (N-CTX)). Some studies, while reporting extractable data, did not include an estimate of variance for the measurement. In those cases we computed the variance, using the methods of Follman [11].

For continuous outcomes (BMD and bone markers), we standardized differences between placebo and treatment groups by dividing by their standard deviation. This method of evaluating outcomes is known as standardized mean difference (SMD) and is equivalent to the BMD Z score. The difference between these standardized outcome scores were calculated for each study and analyzed. We analyzed the results within two groups of patients, among just women and among groups comprised of both men and women. We assessed for publication bias by using the methods of Begg [12]. Assessment of the impact of potential publication bias was done with the trim and fill method [13]. Heterogeneity was assessed visually with Galbraith plots [14] as well as Q statistics (chi-square) by the methods of Mantel and Haenzel [15]. Because of concern about heterogeneity of study design, we used a random-effects model for all analyses [16]. Assessment of the potential sources of heterogeneity included stratified analyses as well as meta-regression [17]. All analyses were done with STATA (STATA 8.0, College Station, Texas, USA).

Results

Our literature search produced 167 articles, and 25 met inclusion criteria for analysis. Excluded articles were reviews, editorials, or animal studies, or on an unrelated topic. Six additional articles were obtained from review of references from selected articles and pertinent reviews. A summary of the 31 selected articles is presented in Table 1. Eight of the selected 31 articles were case–control studies [18, 19, 20, 21, 22, 23, 24, 25], 11 were prospective cohorts [26, 27, 28, 29, 30, 31, 32, 33, 34], four were cross-sectional studies [35, 36, 37, 38], one was a retrospective cohort [39] and seven were randomized controlled trials [40, 41, 42, 43, 44, 45, 46]. Of the randomized controlled trials, two were secondary analyses of lipid-lowering trials, which assessed fracture outcomes, and five trials determined statin effects on bone metabolism, specifically bone markers and bone mineral density, with no data on fractures. Nineteen of the studies involved only women; 11 included cohorts comprised of both men and women, though some of those provided subgroup analysis of female subjects. All studies were rated as good or fair, and none was excluded for quality reasons.
Table 1

Characteristics of included studies

Study, country, year, [reference]

Design

Number

Percentage of women

Age (years)

Statin(s)

Duration

Outcome

Bauer, USA, 1999 [28, 31]

Prospective cohort (SOF trial)

8,412

100

>65

Multiple

4.2 years

BMD/fracture

Bauer, USA, 1999 [28, 31]

Prospective cohort (FIT Trial)

6,459

100

55–80

Multiple

3.8 year

BMD/fracture

Bauer, USA, 2003 [31]

Prospective cohort (HERS trial)

2,763

100

44–79

Multiple

4.5 years

BMD/fractures

Bauer, USA, 2003 [31]

Prospective cohort (Rotterdam Study)

4,878

100

>55

Multiple

5.3 years

Fractures

Bjarnason, Denmark, 2001 [43]

Randomized controlled trial

68

100

71.1

Fluvastatin

12 weeks

Bone markers

Barengolts E, USA, 2001 [25]

Case–control

436

0

69.9

Multiple

3 years

Hip fracture

Cauley, USA, 2000 [27]

Prospective cohort

6,442

100

63.7

Multiple

Unknown

BMD

Chan, USA, 2000 [24]

Case–control

3,675

100

76.7

Multiple

13 months

Fracture

Chung, Korea 2000 [18]

Case–control

69

58

53.6

Multiple

15 months

BMD

Edwards, UK, 2000 [19]

Case–control

141

100

66

Multiple

4 years

BMD

Funkhouser, USA, 2002 [36]

Cross-sectional

983

29

Men 65; women >49

Multiple

2+ years

BMD

Hsia, USA, 2002 [44]

Randomized controlled trial

24

100

56

Simvastatin

12 weeks

Bone markers

LaCroix, USA, 2000 [29]

Prospective cohort

93,723

100

50–79

Multiple

2–3 years

Fracture

LaCroix, USA, 2003 [33]

Prospective cohort

93,716

100

50–79

Multiple

4 years

Fractures/BMD

Lupattelli, Austria, 2004 [34]

Prospective cohort

40

100

61

Simvastatin

2 years

BMD

Meier, UK, 2000 [22]

Case–control

27,319

25

50–89

Multiple

Unknown

Fracture

Montagnani, Italy, 2003 [32]

Prospective cohort

30

100

61.2

Simvastatin

1 year

BMD/bone markers

Mostaza, Spain, 2001 [30]

Prospective cohort

36

100

52

Pravastatin

16 weeks

Bone markers

Pasco, Australia, 2002 [37]

Cross-sectional

1,375

100

70

Multiple

3 years

BMD

Pederson, Norway, 2000 [42]

Randomized controlled trial

4,444

19

Men 58; women 60

Simvastatin

5.4 years

Fracture

Ray, USA, 2002 [39]

Retrospective cohort

34,584

66

62

Multiple

10 years

Hip fractures

Reid, Australia, 2001 [41]

Randomized controlled trial

9014

17

62

Pravastatin

6 years

Fracture

Rejnmark, Denmark, 2002 [38]

Cross-sectional

280

100

62

Simvastatin

4 years

Bone markers

Rejnmark, Denmark, 2004 [23]

Case–control

39,934

70

78.3

Multiple

7 years

Hip fractures

Rejnmark, Denmark, 2004 [46]

Randomized controlled trial

82

100

63

Simvastatin

1 year

BMD, bone markers

Sirola, Finland, 2002 [26]

Prospective cohort

620

100

58.6

Multiple

4.4 years

BMD

Stein, USA, 2001 [45]

Randomized controlled trial

846

Men / Women

53

Simvastatin/atorvastatin

12 weeks

Bone markers

Van Staa, UK, 2001 [20]

Case–control

163,760

76

70

Multiple

Unknown

Fracture

Wada, Japan, 2000 [35]

Cross-sectional

440

51

61.7

Multiple

Unknown

BMD

Wang, USA, 2000 [21]

Case–control

6,110

84

>65

Fluvastatin

3 years

Fracture

Watanabe, Japan, 2001 [40]

Randomized controlled trial

25

100

73

Pravastatin/fluvastatin

6 months

BMD/bone markers

Included studies involved 510,646 individuals in 11 different countries with ages ranging from 53 years to 77 years. The specific statin and duration of use was variable; three trials involved fluvastatin, three pravastatin, one atorvastatin, six simvastatin and 18 reported results from cohorts that were exposed to a variety of statins (Table 1). Duration of use ranged from 12 weeks (for some studies of bone markers) to greater than 4 years, and typical lipid-lowering doses were utilized.

Overall effects (men and women)

A total of 29 studies reported data involving women or men and women (Table 2). Thirteen reported data on hip BMD, finding a significant improvement (Z score SMD 0.12, 95% CI 0.05–0.19), and the 15 studies reporting hip fracture rates found an overall reduction (OR 0.60, 95% CI 0.45–0.78). Spine BMD (n=16) did not improve (Z score −0.03, 95% CI −0.16 to 0.09) and there was no evidence of a reduction in spine fractures (n=5, OR 0.66, 95% CI 0.29–1.51).
Table 2

Meta-analysis results

Parameter

Both men and women

Women only

(n)

Effect (95% CI)

Heterogeneity (P)

Publication bias (P)

(n)

Effect (95% CI)

Heterogeneity (P)

Publication bias (P)

Hip

  Fracture (OR)

15

0.60 (0.45–0.78)

Q=41.2 (P<0.01)

0.43

9

0.75 (0.60–0.95)

Q=10.8(P=0.22)

0.60

  BMD (Z score change)

13

0.12 (0.05–0.19)

Q= 26.9 (P=0.01)

0.09

0.11(0.04–0.18)

Q=24.3(P=0.02)

0.25

Vertebra

  Fracture (OR)

5

0.66 (0.29–1.5)

Q=13.8 (P=0.08)

0.09

1

One study

NA

NA

  BMD (Z score change)

16

−0.03 (−0.16–0.09)

Q=75.6 (P<0.01)

0.32

15

0.01 (−0.07–0.10)

Q=23.7(P=0.05)

0.92

Formative bone markers

  Osteocalcin

No cohorts

Q=38.9 (P<0.01)

0.74

−0.09(−0.918 to 0.73)

Q=78.1(P<0.01)

0.85

  Alkaline phosphatase

14

−0.175 (−0.3 to −0.01)

6

−0.19(−0.38 to −0.004)

Q=32.3(P<0.01)

0.58

Resorptive bone markers

  S-CTX

11

0.19 (−0.06–0.44)

Q=101.9 (P<0.01)

0.28

11

0.26(−0.13–0.66)

Q=105(p<0.01)

0.03

  N-CTX

No cohorts

0.81

5

0.39 (0.07–0.71)

Q= 3.5 (p=0.48)

0.24

Thirteen studies provided data on alkaline phosphatase, finding a small, statistically significant decrease (SMD −0.18, 95% CI −0.34 to −0.01) and no effect on S-CTX (SMD 0.19, 95% CI −0.061 to 0.438, n=11). There were no studies of osteocalcin or N-CTX that included men.

Results among women

A total of 15 studies reported on cohorts comprised solely of women; an additional 18 studies reported subgroup analysis on the women members of the cohorts. Hip BMD improved (Z score 0.11, 95% CI 0.04–0.18, n=13), with a concomitant 25% reduction in hip fracture risk (OR 0.75, 95% CI 0.60–0.95, n=9). In 15 studies vertebral BMD remain unchanged (Z score 0.01, 95% CI −0.07 to 0.10). Only one study reported on vertebral fractures, finding a reduction (OR 0.42, 95% CI 0.24–0.75). Among markers of bone formation there was no effect on osteocalcin (SMD −0.09, 95% CI −0.92 to 0.97, n=4), while alkaline phosphatase decreased (SMD −0.19, 95% CI −0.38 to −0.004, n=9). Among markers of bone resorption, N-CTX increased (SMD 0.39, 95% CI 0.07 to 0.71, n=5), with no effect on S-CTX (SMD 0.26, 95% CI −0.13 to 0.65, n=11).

Randomized controlled trial data

Two randomized controlled trials reported fracture data. Both studies were secondary analyses of the effects of lipid reduction on cardiac outcomes. Both included mostly men in their cohorts, and neither found a reduction in hip (OR 0.87, 95% CI 0.48–1.6) or vertebral fracture (OR 0.75, 0.23–2.4) rates. There were numerous well-designed randomized controlled trials of the effect of statins on BMD and bone markers among postmenopausal women. Those trials found no effect on hip (Z score 0.46, 95% CI 0.09–1.0, n=8) or vertebral (Z score 0.03, 95% CI −0.26 to 0.33, n=5) bone mineral density. Those studies also found no significant effect on osteocalcin (SMD 0.10, 95% CI −1.3 to 1.5, n=4), alkaline phosphatase (SMD −0.21, 95% CI −0.50 to 0.08, n=9), or S-CTX (SMD 0.46, 95% CI −0.09 to 1.01, n=8), with an increase in N-CTX (SMD 0.51, 95% CI 0.06–0.95, n=4).

Sensitivity analysis

There was evidence of heterogeneity in 12 of the 14 analyses conducted (Table 2); only the hip fracture risk and N-CTX analysis had no evidence of heterogeneity. Assessment of potential sources of heterogeneity, by both stratified and meta-regression techniques, found that gender, age, year, country of publication, study type, or specific drug used failed to explain the variance. However, study design was important, with case–control trials consistently showing greater effect than prospective cohort or randomized controlled trial data among these outcomes (Figs. 1, 2, and 3).
Fig. 1

Effect of statins on hip fracture risk, stratified by study design.

Fig. 2

Effect of statins on hip bone mineral density, stratified by study design.

Fig. 3

Effect of statins on alkaline phosphatase, stratified by study design.

The five case–control studies of statin effects on hip fractures (Fig. 1) found a significant reduction in risk (OR 0.45, 95% CI 0.28–0.73), an effect not seen among either the studies involving prospective cohorts (0.73, 95% CI 0.53–1.02, n= 8) or randomized trials (OR 0.87, 95% CI 0.48–1.6, n=2). While the effect on hip BMD (Fig. 2) seen among the prospective cohorts was statistically significant (Z score 0.11, 95% CI 0.045–0.18, n=10), it was smaller than the effect reported from case–control trials (Z score 0.35, 95% CI 0.06–0.64, n=2).

While no study design found a reduction in spinal fracture rates, the effect reported from case–control trials (OR 0.52, 95% CI 0.04–5.9) was greater than that from prospective (0.55, 95% CI 0.18–1.74) or randomized controlled trial (RCT) (OR 0.753, 95% CI 0.24–2.42) cohorts. Similarly, case–control data reported significant differences for alkaline phosphatase (SMD −0.23, 95% CI −0.454 to −0.01, n=2), an effect not seen with prospective (SMD −0.16, 95% CI −0.5 to 0.21, n=2) or randomized controlled (SMD −0.12, 95% CI −0.42 to 0.06) trials. Case–control trials on osteocalcin found an effect (SMD −0.38, 95% CI −0.68 to −0.08), while RCTs did not (SMD 0.10, 95% CI −1.34 to 1.55). The one case–control trial reporting S-CTX found a reduction, an effect not seen with either prospective cohorts or RCTs. N-CTX data came exclusively from well-designed prospective cohorts or RCTs.

Publication bias was detected in one of the 15 analyses, spine fracture (P=0.03), but correction with the method of Duval and Tweedie [13] did not change the overall results (OR 0.73, 95% CI 0.36–1.5).

Discussion

When all studies are included in the analysis, our findings suggest possible benefit from the use of HMG-CoA reductase in treating osteoporosis. There was improvement in hip BMD, and reduction in hip fractures, both overall and among those studies reporting results from women alone. While there are potential biological explanations for these benefits, these results must be interpreted cautiously.

First, the finding of improved bone health was not consistent among the various measures of bone health reported. There was no improvement in vertebral BMD or reduction in vertebral fractures. Our results with bone markers do not support anti-resorptive effects or anabolic effects of statins on bone metabolism. An increase in formative markers and a decrease in resorptive markers are generally considered a pattern of bone-building potential [47, 48]. Our results were paradoxical, with a decrease in alkaline phosphatase, an increase in N-CTX and no effect on osteocalcin or S-CTX. These results would not be consistent with the posited beneficial effect of statins via an effect on BMP-2. It might be that statins have different effects on bone, ultimately not revealed by these bone markers.

Secondly, the effects differed markedly between the different trial designs. There were only two randomized controlled trials that reported fracture data, neither showing an improvement in hip fracture rates. When we stratified the results by study type, the significant benefit for hip fracture rates for statin use was collectively found in the synthesis of results from case–control trials; hip fracture rates were not improved in the data from prospective cohorts or from randomized controlled trials. Similarly, the beneficial effects in most of the data outcomes reported came from case–control data, with either less effect or no effect seen among the prospective cohorts and randomized controlled trials.

Of the various study designs available to assess the effect of therapy on outcomes, case–control trials are considered the weakest, having the greatest potential for confounding. This is particularly important in these trials because there is some evidence that these results might be biased. First, there is evidence that body weight was higher in statin users, an important risk factor for fractures. Some studies have shown that that other lipid-lowering drugs used by cohort participants were associated with reduced fracture risk [33, 39], though this was not seen in other studies [21, 22, 24, 28]. There are also data that indicate that other drugs likely to be used by statin users (e.g., beta blockers, nitrates, estrogens) are associated with beneficial effects on bone health. In addition, the metabolic abnormality for which statins are prescribed may itself be associated with difference in fracture risk or bone density. Given many potential confounders and the fact that the beneficial effects are largely reported from studies with weaker study design, there is clearly a need for properly conducted, adequately powered, randomized controlled trials to assess conclusively whether statins could potentially reduce fracture rates.

Another potential source of discrepancy among these study results is the widely varying pharmacodynamic elements of statins. The relationship between systemic circulation and statins’ effects on bone metabolism are possibly related to dosage, duration, and the specific statin used. The doses used in the animal studies showing benefit, involved statin doses ten-times higher than those used typically in human studies [6]. This is important, since the usual statin doses used to treat dyslipidemia target cholesterol synthesis in the liver. Less than 5% reaches systemic circulation, possibly producing minimal effect on bone metabolism [49]. Moreover, all statins may not have equal effects on bone health. Sugiyama et al. [50] have reported that pravastatin is ineffective in increasing BMP-2 or stimulating bone formation, possibly due to its molecular charge.

The most important limitation is the lack of RCTs dedicated to assessing the effects of HMG-CoA on BMD and fracture rates. While the randomized controlled data currently available from the many trials do not suggest a beneficial effect on spine or hip BMD, or any bone markers, these are only surrogate markers for the outcome on which clinicians hope to have an impact: fracture rates. Unfortunately, neither BMD nor bone markers are very accurate at predicting individuals at risk for fractures. The wealth of studies showing no effect on BMD or bone markers, unfortunately, are not enough to exclude the possibility of a beneficial effect on fracture risk [51]. The two randomized controlled trials in our analysis, reporting fracture rates, were secondary analyses of data of statin effects on cardiovascular outcomes. Those trials were predominantly on men, who have a lower rate of BMD change and fracture rates than women, rendering the study inadequately powered to detect potential differences in hip fracture rates. Moreover, many of the women were not postmenopausal, further reducing the likelihood of finding any benefit. Other important confounding factors, such as history of prior fractures, history of osteoporosis, use of drugs known to be associated with osteoporosis, or family history of osteoporosis, were not collected. Another important problem was the lack of objective assessment of fractures. In addition, one of the two trials [41] used pravastatin, a statin that appears to have little effect on bone in vitro.

Despite these limitations, these potential, interesting benefits of statins may be worth further examination. The benefit on hip BMD, excluding case–control data, suggests a potential improvement in Z score of 0.11. Epidemiological data suggest that such an improvement in BMD would result in a reduction in fracture risk of 0.90, less than seen in this analysis [52]. Animal studies with statins have revealed potential anti-resorptive and bone-building benefits, unlike most FDA-approved osteoporosis drugs, which target anti-resorptive mechanisms of bone metabolism. Parathyroid hormone is the only FDA-approved drug with anabolic effects. The potential use of HMG Co-A reductase inhibitors in osteoporosis is large, since osteoporosis is common, involving approximately 30 million people in the United States of America and affecting 30% of individuals greater than 65 years of age [53]. The complications and debilitating consequences of osteoporosis have enormous health and economic costs on the individual and society. The morbidity and mortality after hip fractures is troubling; approximately 50% of individuals are unable to ambulate independently, and mortality can range from 20%–40% in the first year [54, 55]. Although six FDA-approved drugs are available for prevention and treatment of osteoporosis, none works on both resorption and formation pathways. If the HMG-CoA reductase inhibitors were demonstrated to improve bone health, they would be particularly attractive for patients that also are at cardiovascular risk.

In conclusion, statins are commonly used drugs and their potential benefits in bone metabolism are intriguing. This drug class may provide an added benefit to the treatment of osteoporosis, a common disorder, implying a significant public health benefit, while significantly reducing cardiovascular events. Unfortunately, the evidence of benefit, at present, is limited to weak, case–controlled studies. Dedicated randomized controlled trials are needed to fully elucidate any possible benefit from statins on bone health.

Copyright information

© International Osteoporosis Foundation and National Osteoporosis Foundation 2005