2.1. Statins
The 3-hydroxy-3-methylglutarylo-CoA (HMG-CoA) reductase inhibitors (statins) are the most widely prescribed lipid-lowering agents, constituting the mainstay of treatment both in adults and children with hypercholesterolemia
[110,111][42][43]. Preclinical and clinical data suggest a potential beneficial effect on bone metabolism. In particular, statins may inhibit osteoclastic activity since HMG-CoA blockade reduces the production of downstream products in the mevalonate pathway, such as farnesyl pyrophosphate (FPP) and geranylgeranyl pyrophosphate (GGPP)
[112][44]. The same pathway is also shared by nitrogen-containing bisphosphonates, thus preventing the prenylation of guanosine triphosphate (GTP)-ases, such as Ras, Rho, and Rac, which are essential for the survival and function of osteoclasts
[113][45]. Another mechanism could be the inhibition of RANKL, which is essential for osteoclast differentiation by averting the production of reactive oxygen species
[114][46]. Statins may also increase 25-hydroxy-vitamin D concentrations
[52][20].
Statins also exert osteoanabolic properties, inhibiting osteoblast apoptosis and fostering osteoblast activity. This mechanism is mediated through increased expression of the
BMP-2 gene, which promotes osteoblast differentiation
[115][47]. The latter is also induced by the depletion of FPP and GGPP, as mentioned above
[116][48]. Statins may also promote embryonic stem cell differentiation towards the osteogenic lineage, through activation of increased mRNA expression of runt-related gene 2 (
Runx2), osterix (
OSX), and osteocalcin (
OCN), as osteogenic transcription factors
[117][49].
However, data in humans regarding the effect of statins on bone mass and, more importantly, fracture risk are not robust and consistent. A meta-analysis of randomized controlled trials (RCTs), published in 2016, including seven studies involving a total of 27,900 subjects, showed an increase in BMD by 0.03 g/cm
2 (95% CI 0.006–0.053; I
2 99.2%;
p < 0.001) with statins. Concerning the skeletal site, four studies assessed BMD in LS, one in the distal radius and two in any of multiple skeletal sites. Regarding fracture risk, no association with statin use was observed [pooled hazard ratio (HR) 1.00, 95% CI 0.87–1.15; I
2 0;
p = 0.396]. These findings remained consistent and significant in sensitivity analysis
[118][50].
These results were replicated by another meta-analysis published in 2017, including 33 studies (23 observational and ten RCTs) with 314,473 patients on statin therapy and 1,349,192 controls
[119][51]. In particular, statins increased LS BMD (standardized MD (SMD) 0.20, 95% CI 0.07–0.32;
p = 0.002; I
2 43%), as well as TH BMD (SMD 0.18, 95% CI 0.00–0.36;
p < 0.05; I
2 62%). In subgroup analyses, these associations remained significant only for data derived from cohort studies but not for RCTs. Notably, there was no gender difference regarding TH, but LS BMD increased only in males. Concerning FN BMD, no association with statin use was found. Regarding fracture risk, statins decreased the risk of overall (odds ratio (OR) 0.81, 95% CI 0.73–0.89;
p < 0.0001; I
2 87.5%) and hip fractures (OR 0.75, 95% CI 0.60–0.92;
p = 0.007; I
2 77.2%), however, with no effect on vertebral and upper extremity fractures (data from 16 cohort and case-control studies). The authors also assessed the effect of statins on markers of bone turnover, showing a positive effect on osteocalcin concentrations (SMD 0.21, 95% CI 0.00–0.42;
p = 0.04; I
2 = 0%), but no effect on bone-specific alkaline phosphatase (bALP) and serum C-terminal peptide of type I collagen (CTX) concentrations
[119][51].
Interestingly, a recent Mendelian randomization (MR) study showed that the effect of statins on BMD was dependent on the degree of their LDL-C-lowering action. MR explained this by utilizing 400 single nucleotide polymorphisms, which provided evidence for a causal effect of LDL-C on BMD
[120][52].
Another recent meta-analysis assessed the effect of statin use exclusively on fracture risk in older adults
[121][53]. The authors included 21 observational studies and two RCTs (
n= 1,783,123 participants). Data from the observational studies showed an overall decreased fracture risk with statin use [pooled relative risk (RR) 0.80, 95% CI 0.72–0.88; I
2 93.1%]. In subgroup analysis, this association was more evident in men (RR 0.75, 95% CI 0.59–0.95) than in women (RR 0.87, 95% CI 0.76–0.99) and only for hip (RR 0.73, 95% CI 0.64–0.82) and low extremity fractures (RR 0.69, 95% CI 0.54–0.88). In terms of statin type, only atorvastatin was associated with a reduction in fracture risk (RR 0.77, 95% CI 0.71–0.84) compared to other statins. Interestingly, this beneficial effect was shown only for a short duration of statin use (<1 year) (RR 0.66, 95% CI 0.47–0.93), but not for a higher duration (1–3 or >3 years). Of note, it must be emphasized that the evidence for an anti-fracture efficacy for statins was based only on data derived from observational studies. In the two RCTs, there was no evidence for reducing fracture risk with statin use (RR 1.00, 95% CI 0.87–1.15; I
2 0%)
[121][53].
2.4. Fibrates
Fibrates are proliferator-activated receptor (PPAR)-α agonists, mostly used in patients with hypertriglyceridemia. They are moderately effective agents in reducing plasma TG (by 50%) and, to a lesser extent, LDL-C (≤20%), as well as in increasing HDL-C concentrations (≥20%)
[127][59].
Concerning bone metabolism, preclinical data have demonstrated that fibrates and, in particular, fenofibrate promote
BMP-2 gene expression, thus, stimulating the osteoblast differentiation
[128][60]. Fenofibrate has been shown to maintain FN and whole-body BMD and bone architecture in ovariectomized rats, compared with pioglitazone
[129][61]. However, others have shown a detrimental effect on bone quality in mice with diabetes mellitus, through decreased collagen I and osteocalcin secretion, due to down-regulation of
Runx2 gene expression
[130][62].
The evidence for any clinical effect of fibrates on bone health is generally poor. In a case-control study, including 124,655 fracture cases and 373,962 age- and gender-matched controls, an increased risk for non-statin lipid-lowering agents (mainly cholestyramine and fibrates) was demonstrated. In contrast to statins, the use of these non-statin drugs was associated with an increased crude risk of vertebral (OR 2.25; 95% CI 1.22–4.16) and total fractures (OR 1.14, 95% CI 1.00–1.30). However, this association lost significance after adjustment for potential confounders
[131][63].
2.5. Omega-3 Fatty Acids
The omega-3 fatty acids (FA) and, in particular, docosahexaenoic (DHA) and eicosapentaenoic acid (EPA) are essential polyunsaturated FA, derived mainly from fish oil
[127][59]. They are moderately efficacious in lowering serum TG concentrations in a dose-dependent manner, with usual doses of 2–4 g/day, although their effect on other lipoproteins is trivial
[127][59]. A cardiovascular benefit has been shown in patients at very high CVD risk with high doses (2 g of EPA twice a day)
[127][59].
Preclinical data suggest a protective effect of omega-3 FA on bone metabolism since a high dietary intake increases the rate of bone formation
[132][64]. They also reduce osteoclastic activity and the ensuing bone resorption by 80%, as shown in rats fed with a purified diet rich in omega-3 FA
[133][65]. Furthermore, fat-1 transgenic mice, which can convert omega-6 to omega-3 FAs, demonstrate significant acceleration in callus formation and fracture healing compared with controls
[134][66].
Epidemiological data in humans regarding the effect of omega-3 FA on musculoske-letal outcomes have provided inconsistent results. A meta-analysis of observational stu-dies (including seven prospective and three case-control studies;
n = 292,657 participants) showed an inverse association between fish consumption and the risk of hip fractures (pooled effect size 0.88, 95% CI 0.79–0.98, for the highest compared with the lowest quartile)
[33][67]. However, in subgroup analysis, this association was evident only in case-control studies and in prospective studies with a sample size of ≥ 10,000 participants. Moreover, an inverse association between omega-3 FA intake and the risk of hip fracture was observed (pooled effect size: 0.89, 95% CI 0.80–0.99)
[33][67].
In a systematic review and meta-analysis of ten RCTs published in 2012, a favorable effect of omega-3 FA on BMD or bone turnover markers was demonstrated in four studies, but only when co-supplemented with calcium, whereas three studies showed no effect. No data on fractures were available
[135][68]. Another meta-analysis of 28 RCTs (23 studies on omega-3 FA; 0.4–5.8 g/day of EPA and/or DHA, 3.5–9.1 g/day of alpha-linoleic acid) showed no effect on LS (mean difference 0.03 g/cm
2, 95% CI from −0.02 to 0.07) or FN BMD (mean difference 0.04 g/cm
2, 95% CI from −0.00 to 0.07) (low or very low quality of evidence, respectively)
[136][69]. A high omega-3 dose induced a slight increase in osteocalcin concentrations, but no effect was observed on other bone formation or bone resorption markers
[136][69]. No data on fractures were available from both meta-analyses
[135,136][68][69].