Sixty infants with birth-weights less than 1500 g and who were less than 6 h old were randomly assigned to a group given phenobarbitone or a control group. Intravenous phenobarbitone was given in doses sufficient to achieve anticonvulsant serum levels within 12-18 h. Maintenance therapy was continued for one week. Periventricular/intraventricular haemorrhage (IVH) occurred in 13.3% (4/30) of the phenobarbitone group and in 46.7% (14/30) of the control group. The occurrence of risk factors related to IVH was similar in the two groups. Phenobarbitone may reduce the incidence of IVH in small preterm infants.

Human insulin derived from porcine insulin was given subcutaneously (s.c.), intramuscularly (i.m.), and intravenously (i.v.) to normal men. The dosage for all three routes was 0 . 075 IU/kg body weight. Diluting medium was administered by s.c. injection to obtain control values. Somatostatin (100 microgram/h) was given to inhibit pancreatic beta cell secretion. The plasma glucose responses to s.c. injection of this insulin into the anterior abdominal wall and to i.m. injection into the thigh were similar with respect to the extent, onset, and duration of effect. Plasma glucose fell from mean (+/- SE) pre-injection values of 4 . 3 +/- 0 . 15 and 4 . 4 +/- 0 . 27 mmol/l, to 3 . 06 +/- 0 . 25 and 2 . 98 +/- 0 . 16 mmol/l by 90 to 105 min for s.c. and i.m. studies, respectively, thereafter returning to mean basal level by 6 h after i.m. injection, but remaining about 0 . 5 mmol/l below basal level after s.c. injection. A much more sudden, but short-lived, hypoglycaemic response occurred after i.v. insulin, with plasma glucose failing from 4. 50 +/- 0 . 42 to 1 . 45 +/- 0 . 16 mmol/l by 25 min, returning to mean basal value after 3 1/2 h. The mean (+/- SE) peak insulin levels after s.c. and i.m. injection were 0 . 13 +/- 0 . 01 and 0 . 18 +/- 0 . 04 pmol/ml at 90 and 60 min, respectively. After i.v. injection the maximum plasma insulin concentration of 6 . 9 +/- 0 . 73 pmol/ml was seen at 2 min. No adverse side-effects were observed.

Clinic cuff blood-pressure measurements, obtained on at least three occasions, were compared with mean arterial pressures in 59 patients with borderline or essential hypertension who underwent direct ambulatory monitoring of blood pressure. In 22 patients (group I) mean cuff and ambulatory pressures were similar (+/- 10 mm Hg) while in 32 subjects (group II) cuff pressures were more than 10 mm Hg higher. Groups I and II could not be distinguished on the basis of clinical examination, indices of sympathetic nerve activity, or blood-pressure variability, or by the magnitude of pressure rise during physical or mental exercise. Group II had less cardiovascular target organ damage and better baroreflex sensitivity but there was considerable overlap. There was no reliable way of telling which subjects would have lower ambulatory than cuff pressures. 20 out of 59 subjects classified as hypertensive by cuff measurements had awake ambulatory pressures or less than 140/90 mm Hg.