ObjectiveCurrently, there are limited markers to predict the osteoporosis probability in patients with primary hyperparathyroidism. We studied the relationship between various parameters and results of DXAs at various skeletal sites. MethodsRetrospective review of data for 218 patients with primary hyperparathyroidism was performed. Age, BMI, bone mineral density, serum total calcium, ionized calcium, intact parathyroid hormone, albumin, alkaline phosphatase, phosphate, 25-hydroxy vitamin D, 1,25-dihydroxy vitamin D, 24-hour urinary calcium levels and parathyroid tumor weight were analyzed. Two different statistical models- linear regression and multivariate logistic regression were performed. ResultsAt the lumbar spine, with the linear model, BMI (P < .001), alkaline phosphatase (P < .001), and ionized calcium (P < .001) significantly correlated with T scores; whereas with the logistic model, BMI was the only variable predicting osteoporosis probability.At the femoral neck, BMI (P < .022), 25-hydroxy vitamin D (P < .001), 1,25-dihydroxy vitamin D (P < .034) correlated with T scores; whereas both BMI (P < .029) and age (P < .051) were the significant variables that predicted osteoporosis.At the total hip, BMI (P < .001) and age (P < .001) correlated with T scores; whereas with the logistic model, only BMI (P < .016) predicted osteoporosis. At the forearm, a model could not be generated due to limited number. ConclusionIn patients with primary hyperparathyroidism, BMI strongly correlated with T scores and probability of osteoporosis.