SUMMARY Adrenocortical function was studied in 52 newborn infants who had been divided into three groups: preterm well, preterm ill, and term ill. Basal plasma 17-hydroxyprogesterone concentrations were significantly increased in both groups of preterm infants. There was no significant difference in basal plasma cortisol concentrations, although they were highest in preterm ill infants. All infants responded to adrenocorticotrophic hormone (ACTH) stimulation (36 rig/kg intramuscularly) with a two to threefold increase in the concentration of both steroids. The peak plasma 17-hydroxyprogesterone response was significantly higher in preterm ill infants. A subgroup of five infants, who were highly stressed but had undetectable basal plasma cortisol concentrations, also showed an appropriate response to ACTH. The results provide useful reference data to assess adrenal function in the infant of a mother given glucocorticoids during pregnancy. There is also a change from the pattern of fetal adrenal steroidogenesis soon after birth, which may be affected by exogenous ACTH stimulation. Roughly 10% of stressed newborns failed to synthesise cortisol basally; temporary glucocorticoid replacement for such infants may be appropriate.Plasma concentrations of 17-hydroxyprogesterone are markedly raised in untreated infants with congenital adrenal hyperplasia due to 21-hydroxylase deficiency.

antiserum coupled to a magnetisable solid phase support. The sensitivity of the assay was 4-7 pg/tube, equivalent to 0 7 nmol/l. Intra-assay coefficient of variation did not exceed 5.5 % for low, medium, and high pools. The inter-assay coefficients of variation for low, medium, and high pools were 6.4 %, 5 1 %, and 41-% respectively. The cross re-

SUMMARY The adrenocortical response to stress was studied longitudinally in 10 ill preterm infants using measurements of cortisol and 170H-progesterone concentrations in filter paper blood spots. Mean cortisol and 170H-progesterone concentrations reached a peak of 2200 nmol/l and 65 nmolIl, respectively, between the third and fifth days of life. These concentrations far exceeded those observed in older children and adults subjected to stress as a result of surgery. Further pulses of endogenous cortisol production of 4000 nmolI or more occurred in association with clinical complications such as intraventricular haemorrhage. These results indicate that infants undergoing intensive care are unduly stressed. Consideration should be given to providing enough sedation and appropriate analgesia for ill preterm infants during painful procedures such as insertion of venous cannulae and arterial puncture.Increased plasma concentrations of 170H-progesterone (17P) and cortisol in ill preterm infants have previously been reported.' 2 These observations, which indicated a presumed stress response, were based only on single random measurements of plasma steroid concentrations. Little is known about the magnitude and duration of the adrenocortical response to stress in the newborn. A study of ill infants during the first three days after birth by Gutai et al showed that there was no significant increase in plasma cortisol concentrations. Most of the infants studied were not unduly stressed when assessed by clinical criteria.3 In a study of ill preterm infants with hyaline membrane disease,4 however, the peak plasma glucocorticoid response achieved was related to the severity of respiratory distress.Studies, in which serial measurements of plasma steroid concentrations in ill preterm infants have been made, have been limited by the difficulty of collecting venous samples from small infants. A steroid assay, which uses whole blood spotted on to filter gaper, has recently been developed however, and this technique, which requires only a small volume of blood, has been used to obtain longitudinal data on 17P and cortisol concentrations as an index of the adrenocortical response to stress in a group of ill preterm infants. Patients and methodsTen preterm patients were studied longitudinally for the first 13 days of life. All were very ill and required assisted ventilation for severe hyaline membrane disease. Table 1 (umbilical or radial) were inserted shortly after birth for measurements of blood gas and invasive blood pressure monitoring. When an arterial sample was collected additional drops were spotted on to a Guthne card sufficient to fill completely four 10 mm circles. The date and time of each sample collection was recorded for each infant. The filter paper card was left to dry at room temperature before being sent to the laboratory for analysis. In five of the infants blood spots were obtained at about six hourly intervals for three to five days.Blood spot 17P concentrations were determined by a specific radioimmunoassa...

Total plasma cortisol measurements may be misleading when there are variations in the plasma cortisol-binding protein capacity resulting from drugs, pregnancy or congenital alterations in cortisol-binding globulin (CBG). Salivary cortisol levels, which represent the free component of plasma cortisol, are less affected by alterations in protein binding and have been used in the investigation of hypothalamic-pituitary-adrenal disorders. This study compares these two indices of adrenal function in conditions of true hypercortisolism and spurious hypercortisolism (resulting from oral contraceptive medication or pregnancy). The circadian variation of cortisol in plasma and saliva was studied in six patients with unequivocal hypercortisolism and compared with normal volunteers. In the normal group, plasma and salivary cortisol levels taken at 0900 h were significantly higher than those taken at 2400 h. Patients with Cushing's syndrome failed to show a significant difference between plasma and salivary cortisol levels collected at 0900 and 2400 h. Five patients with pituitary-dependent Cushing's disease, one patient with an adrenal carcinoma causing Cushing's syndrome and seven normal subjects each received a dexamethasone suppression test using a continuous infusion of dexamethasone sodium phosphate at a rate of 1 mg/h. There was no significant difference in the half-life disappearance rate of endogenous cortisol in either plasma or saliva comparing grouped data from patients with pituitary-dependent Cushing's disease with that of normal subjects. Failure of suppression of both plasma and salivary cortisol levels was observed in the one patient with adrenal carcinoma during dexamethasone infusion.(ABSTRACT TRUNCATED AT 250 WORDS)