Diurnal variation of estriol excretion in human pregnancy

Twenty-one normally pregnant patients had their urinary output fractionated into 4 hour aliquots for a total of 6 days. Each urinary specimen was analyzed for estriol. By using duplicate assays the statistical error was reduced to between 3 and 5 per cent. From these data it is not possible to predict the 24 hour urinary estriol excretion from a single 4 hour period specimen. Even the use of two 4 hour specimens in combination will not successfully predict the 24 hour excretion rate. While diurnal variation clearly exists, it is extremely complex and not simply predictable for a given patient. In these normally pregnant patients, bed rest had no significant influence on urinary estriol excretion rates.