TOTAL CHOLESTEROL, LOW-DENSITY LIPOPROTEIN CHOLESTEROL, OR HIGH DENSITY LIPOPROTEIN AND BONE HEALTH: SYSTEMATIC REVIEW AND MENDELIAN RANDOMIZATION

Abstract

BackgroundCardiovascular calcification has been associated with osteoporosis and observational studies have linked lipids with bone mineralization suggesting a common link. However, the evidence that lipids associate with bone health is conflicting and the causal nature of this association remains unclear. Accordingly, we sought to (1) summarize the epidemiologic evidence for the association between total cholesterol (TC), low density lipoprotein (LDL), high density lipoprotein (HDL) or triglycerides (TG) on bone mineral density (BMD), osteoporosis or osteoporotic fractures by a systematic review of the literature; and (2) to provide evidence for a causal association, by performing a Mendelian randomization (MR) study of lipid-associated single nucleotide polymorphisms (SNPs) and BMD.MethodsWe searched the Cochrane Library, Medline, and Embase from inception until December 15, 2014 for studies that assessed the association of lipids (LDL-C, T-C and HDL-C) on bone mineral density, osteoporosis and osteoporotic fractures. We included observational studies; cohort, case-control and cross-sectional studies which included adults (age > 20 years). For MR analyses, we used summary level statistics from the GEFOS consortium, a large-scale genomewide association of BMD, to evaluate the effect of lipid SNPs on BMD.ResultsWe included observational studies (n=20), comprising 43,432 participants (58% women; women mean age 59 years, men mean age 57 years). Study designs were cross sectional studies (n=14), case control studies (n=1), and cohort studies (n=5). Among eligible studies, 12 studies reported an association between increasing TC or LDL-C with reduced BMD whereas 8 studies reported no association or a protective relationship with higher TC or LDL-C. The evidence for other lipids and BMD were even less consistent. The quality of observational studies was highly variable. The marked clinical and methodological heterogeneity in reported studies precluded quantitative meta-analysis. In MR analyses, effect sizes for genetic risk scores for LDL-C, HDL-C and TG were small and did not reach statistical significance for an association with BMD (Table).Conclusion BackgroundCardiovascular calcification has been associated with osteoporosis and observational studies have linked lipids with bone mineralization suggesting a common link. However, the evidence that lipids associate with bone health is conflicting and the causal nature of this association remains unclear. Accordingly, we sought to (1) summarize the epidemiologic evidence for the association between total cholesterol (TC), low density lipoprotein (LDL), high density lipoprotein (HDL) or triglycerides (TG) on bone mineral density (BMD), osteoporosis or osteoporotic fractures by a systematic review of the literature; and (2) to provide evidence for a causal association, by performing a Mendelian randomization (MR) study of lipid-associated single nucleotide polymorphisms (SNPs) and BMD. Cardiovascular calcification has been associated with osteoporosis and observational studies have linked lipids with bone mineralization suggesting a common link. However, the evidence that lipids associate with bone health is conflicting and the causal nature of this association remains unclear. Accordingly, we sought to (1) summarize the epidemiologic evidence for the association between total cholesterol (TC), low density lipoprotein (LDL), high density lipoprotein (HDL) or triglycerides (TG) on bone mineral density (BMD), osteoporosis or osteoporotic fractures by a systematic review of the literature; and (2) to provide evidence for a causal association, by performing a Mendelian randomization (MR) study of lipid-associated single nucleotide polymorphisms (SNPs) and BMD. MethodsWe searched the Cochrane Library, Medline, and Embase from inception until December 15, 2014 for studies that assessed the association of lipids (LDL-C, T-C and HDL-C) on bone mineral density, osteoporosis and osteoporotic fractures. We included observational studies; cohort, case-control and cross-sectional studies which included adults (age > 20 years). For MR analyses, we used summary level statistics from the GEFOS consortium, a large-scale genomewide association of BMD, to evaluate the effect of lipid SNPs on BMD. We searched the Cochrane Library, Medline, and Embase from inception until December 15, 2014 for studies that assessed the association of lipids (LDL-C, T-C and HDL-C) on bone mineral density, osteoporosis and osteoporotic fractures. We included observational studies; cohort, case-control and cross-sectional studies which included adults (age > 20 years). For MR analyses, we used summary level statistics from the GEFOS consortium, a large-scale genomewide association of BMD, to evaluate the effect of lipid SNPs on BMD. ResultsWe included observational studies (n=20), comprising 43,432 participants (58% women; women mean age 59 years, men mean age 57 years). Study designs were cross sectional studies (n=14), case control studies (n=1), and cohort studies (n=5). Among eligible studies, 12 studies reported an association between increasing TC or LDL-C with reduced BMD whereas 8 studies reported no association or a protective relationship with higher TC or LDL-C. The evidence for other lipids and BMD were even less consistent. The quality of observational studies was highly variable. The marked clinical and methodological heterogeneity in reported studies precluded quantitative meta-analysis. In MR analyses, effect sizes for genetic risk scores for LDL-C, HDL-C and TG were small and did not reach statistical significance for an association with BMD (Table). We included observational studies (n=20), comprising 43,432 participants (58% women; women mean age 59 years, men mean age 57 years). Study designs were cross sectional studies (n=14), case control studies (n=1), and cohort studies (n=5). Among eligible studies, 12 studies reported an association between increasing TC or LDL-C with reduced BMD whereas 8 studies reported no association or a protective relationship with higher TC or LDL-C. The evidence for other lipids and BMD were even less consistent. The quality of observational studies was highly variable. The marked clinical and methodological heterogeneity in reported studies precluded quantitative meta-analysis. In MR analyses, effect sizes for genetic risk scores for LDL-C, HDL-C and TG were small and did not reach statistical significance for an association with BMD (Table). Conclusion