Renal handling of drugs in the healthy elderly. Creatinine clearance underestimates renal function and pharmacokinetics remain virtually unchanged.

OBJECTIVE: It is commonly assumed that renal function, and in parallel the excretion of drugs, is considerably reduced in the elderly. Endogenous creatinine clearance or indirect estimates of this parameter are generally recommended for adapting drug dosage. The present study evaluates the validity of both assumptions.
METHODS: We compared pharmacokinetics (and pharmacodynamics) of 50 mg atenolol, 800 mg piracetam and 25 mg hydrochlorothiazide plus 50 mg triamterene in ten healthy young [25 (2) years] and 11 healthy elderly subjects [68 (5) years]. Inulin (Cin) and para-aminohippurate [PAH (CPAH)] clearance (infusion clearance technique), endogenous (C(Cr)) and calculated (Cockroft-Gault) creatinine clearance, analysis of drugs and their metabolites (HPLC), were performed. Renal haemodynamics and the pharmacokinetics of beta-adrenergic blocking agent, diuretics and the nootropic agent piracetam, respectively, were measured on separate days.
RESULTS: Cin was significantly (P < 0.01) lower in the healthy elderly subjects [104 (12) vs 120 (14) ml x min(-2) x 1.73 m(-2) in the young], but remained within the normal range (> 90 ml x min(-2) x 1.73 m(-2)). In contrast, C(Cr) was even lower in healthy elderly subjects [95 (24) vs 121 (20) ml x min(-1) in the young], and the Cockroft-Gault clearance underestimated true glomerular filtration rate (GFR) even more seriously [74 (17) vs 122 (16) ml min(-1)]. For atenolol the mean area under the curve (AUC) was similar in both groups [3.16 (0.48) microg x h(-1) x ml(-1) in the elderly vs 3.01 (0.30) in the young], as was the mean maximal plasma concentration [0.42 (0.07) vs 0.44 (0.06) microg x ml(-1)], but the proportion of the drug excreted in urine was marginally (P < 0.025) lower in the elderly. Similar results were obtained for hydrochlorothiazide, whereas no marked differences between the groups were found for triamterene and its metabolite. Furthermore, the pharmacodynamic action of diuretics was not significantly altered in the elderly.
CONCLUSIONS: The true GFR of the healthy elderly remains within the normal range and is underestimated by creatinine clearance and more so by its surrogate (Cockroft-Gault clearance). In parallel, pharmacokinetics of renally excreted drugs are not affected in the healthy elderly to a clinically significant extent. For drugs with a narrow therapeutic window, indirect estimates of GFR appear to be an unreliable means for calculating correct dosage in the elderly.