Abstract
OBJECTIVES: to extensively investigate the relationship between serum total alkaline phosphatase (ALP) and osteoporosis in a large-scale population. METHODS: people who took routine health examinations between 2019 and 2024 at a teaching hospital were included. The health examination records were reviewed, and data were extracted. The variables included anthropomorphic measurements, laboratory tests, and biochemical indices. Osteoporosis was defined by bone mineral density at the spine and hip. The relationship between ALP and osteoporosis was analyzed by using logistic regression, Receiver Operating Characteristic (ROC) analysis, and restricted cubic spline analysis. RESULTS: total ALP in serum was significantly higher in osteoporosis than in non-osteoporosis groups(P<0.001). The relationship between ALP and osteoporosis was persistent in multivariate analysis after adjusting for available confounders. There was a non-linear relationship between ALP and osteoporosis; when ALP was greater than 100 IU/L, the relationship became less pronounced. The relationships between ALP and osteoporosis were stronger in younger, female, and metabolically healthy populations with normal liver enzymes. The best cut-off value for ALP to predict osteoporosis was 72IU/L in ROC analysis. CONCLUSIONS: the serum total ALP in serum persistently and negatively relates to the risk of osteoporosis in a general adults population when ALPs are clinic normal. The 72IU/Lof total serum ALP may be considered as a tentative threshold for initiating further bone health counseling in health management for adults. Cohort studies on the application of ALP in the prediction of osteoporosis or osteoporotic fracture are warranted.