Association Between Nonalcoholic Fatty Liver Disease and Reduced Bone Mineral Density in Children: A Meta-Analysis.

Abstract

Recent cross-sectional studies have examined the association between nonalcoholic fatty liver disease (NAFLD) and bone mineral density (BMD) in children or adolescents, but these have produced conflicting results. We performed a systematic review and meta-analysis of these published studies to quantify the magnitude of the association, if any, between NAFLD and BMD. We searched publication databases from January 2000 to September 2018, using predefined keywords to identify relevant observational studies conducted in children or adolescents in whom NAFLD was diagnosed either by imaging or by histology and BMD Z score was measured by dual-energy X-ray absorptiometry. Data from selected studies were extracted, and a meta-analysis was performed using random-effects modeling. A total of eight observational cross-sectional or case-control studies enrolling 632 children and adolescents (mean age 12.8 years), 357 of whom had NAFLD, were included in the final analysis. Meta-analysis showed significant differences in whole-body or lumbar BMD Z scores between children/adolescents with and without NAFLD (n = 6 studies; pooled weighted mean difference [WMD], -0.48; 95% confidence interval [CI], -0.74 to -0.21; I2 = 55.5%), as well as between those with biopsy-confirmed nonalcoholic steatohepatitis (NASH) and those with no-NASH (n = 4 studies; pooled WMD, -0.27; 95% CI, -0.40 to -0.13; I2 = 0%). The aforementioned WMDs in BMD Z scores were independent of common clinical risk factors, such as age, sex, race/ethnicity, and body mass index. Sensitivity analyses did not modify these findings. Funnel plot and Egger test did not reveal significant publication bias. Conclusion: This meta-analysis shows that the presence and severity of NAFLD are significantly associated with reduced whole-body BMD Z scores in children and adolescents; however, the observational design of the studies included does not allow for proving causality.