Inferring disease mechanisms from epidemiological data in chronic kidney disease: calcium and phosphorus metabolism.

BACKGROUND/AIMS By applying numerical filtering to epidemiological data of 2,512 chronic kidney disease patients, we aimed to identify some of the underlying mechanisms of the calcium/phosphorus metabolism perturbations. METHODS The measured variables, serum calcitriol, calcidiol, total calcium ([Ca](s)) and phosphorus ([P](s)) and the urinary excretions of calcium and phosphorus, were paired in the same patients with the glomerular filtration rate (GFR) or the serum concentrations of parathormone (i[PTH](s)) (used as independent variables) numerically filtered with a moving average and partitioned into 15-25 frequency classes. All variables exhibited unimodal frequency distributions. RESULTS There was a steep fall of i[PTH](s), [P](s), and urinary excretion fractions of Ca and P up to a value of GFR in the range of 25-45 ml/min/1.73 m2. The increase in the phosphorus urinary excretion preceded the steep increase in i[PTH](s). Except [Ca](s), all factors exhibited their physiological correlation with i[PTH](s) when GFR was above 90 ml/min/1.73 m2 and reverted to a feedback correlation below 80 ml/min/1.73 m2. CONCLUSION The perturbation of mineral metabolism in chronic kidney disease results in the maintenance of a normal range of [Ca](s) and [P](s) acting as the controlled factors at the cost of large variations of i[PTH](s), and calcium and phosphate urinary excretions behaving as controlling factors.