Parathyroid Hormone as a Mortality Predictor in Frail Aged Inpatients

Abstract

Background: Some prospective cohort studies have associated parathyroid hormone (PTH) levels with survival independently of renal function, calcaemic and vitamin D status in the elderly. Objective: In order to further evaluate the prognostic significance of subtle elevation of PTH and the involvement of vitamin D status in bedridden aged inpatients, the participants of a 6-month vitamin D supplementation trial were followed for 2 years. Methods: Eligible patients (n = 218) of 4 long-term care hospitals (1,215 beds) were randomized to receive 0, 400 or 1,200 IU/day cholecalciferol for 6 months. In addition to routine analyses, plasma 25-hydroxyvitamin D (25-OHD), PTH and ionized calcium (Ca²⁺) levels were measured. Functional capacity was evaluated by activities of daily living (ADL) hierarchy scale and cognition was assessed by cognitive performance scale (CPS). Body mass index and glomerular filtration rate (GFR) were calculated. Mortality data was collected from patient records. Results: The patients were aged (84.5 ± 7.5 years), vitamin D deficient (25-OHD = 23 ± 10 nmol/l) and frail (ADL 5.5, range 3–6; CPS 4.9, range 1–6). The PTH levels ranged from 12 to 268 ng/l, the cut points for IV quartile being 72 ng/l. The PTH levels in the IV quartile predicted a significant 1.58-fold over-mortality (95% CI 1.08–2.32, p = 0.020), resulting in a 9.1-month shortening (p = 0.019) in median life expectancy in patients with poorest renal function, but the prognostic significance of PTH was not explained by GFR. Furthermore, controlling for age, gender, body mass index, creatinine, 25-OHD, supplementation group, Ca²⁺ and albumin levels did not abolish this significance. Mortality was predicted neither by baseline 25-OHD nor vitamin D supplementation. Conclusions: Even a subtle elevation of PTH is a relatively independent predictor of increased 2-year mortality in severely frail older inpatients with poor overall survival prognosis. Further prospective studies on the involvement of 1,25 dihydroxyvitamin D status in the predictive value of PTH are needed.