Some problems encountered in attempting to estimate catecholamine turnover using labelled tyrosine