The present study describes for the first time, a metabolic profile reflecting the
osteoporosis progression in 364 pre- and postmenopausal Chinese women using GC-MS. In order to accurately evaluate the dynamic changes of metabolites along with
estrogen deficiency and
osteoporosis progression, we divided these subjects into the following four groups: premenopausal women with normal bone mass density (BMD, group I), postmenopausal women with normal BMD (group II), postmenopausal women with
osteopenia (group III) and postmenopausal women with
osteoporosis (group IV), according to their menopause or low BMD status. Principal component analysis (PCA) and Partial least squares-discriminant analysis (PLS-DA) were used to evaluate the associations of metabolic changes with low BMD or
estrogen deficiency. Twelve metabolites identified by the PLS-DA model were found to be able to differentiate low BMD groups from normal BMD groups. Of the 12 metabolites, five
free fatty acids (LA,
oleic acid, AA and 11,14-eicosadienoic
acid) have the most potential to be used as
osteoporosis biomarkers due to their better correlations with BMD, and high sensitivity and specificity in distinguishing the low BMD groups from the normal BMD groups calculated by the receiver operating characteristic curve (ROC). The
lipid profile may be useful for
osteoporosis prediction and diagnosis.