Urinary creatinine excretion is not stable: a new method for assessing urinary toxic substance concentrations.

Urinary concentrations of toxic substances require correction to adjust for the misleading effects of varying states of hydration. The most common method in current use involves calculation of substance-to-creatinine concentration ratios. For accuracy, this method assumes creatinine excretion rates to be stable despite varying rates of urinary flow. However, this underlying assumption has been challenged in recent studies. Our evaluation of separate data regarding individual voids confirmed that creatinine excretion rate depends significantly (p less than .0001) on urinary flow. We calculated a logarithmic regression model identical to one reported previously. The partial correlation coefficient for log flow was .21 after adjusting for inter-individual differences in creatinine excretion rates. We propose a simple method to correct creatinine concentrations in "spot" urine samples for the effects of varying hydration. The new method retains many benefits of the classical correction by substance-creatinine ratios.