The transport characteristics of the placenta, which favour higher phenylalanine concentrations in the fetus than in the mother, and regression data of head circumference at birth against phenylalanine concentration at conception in maternal phenylketonuria (PKU), suggest that treatment of maternal PKU should ideally aim to maintain plasma phenylalanine concentration within the normal range throughout pregnancy. A patient with classical PKU was treated from before conception by aiming to maintain plasma phenylalanine concentration within the range 50-150 mumol/l and tyrosine within the range 60-90 mumol/l. The diet was supplemented with phenylalanine-free amino acids (100-180 g/day) and tyrosine (0-5 g/day). Plasma amino acid concentrations were monitored weekly by amino acid analyser. Dietary phenylalanine intake ranged from 6 mg/kg/day at conception to 30 mg/kg/day at delivery. Normal weight gain and fetal growth were maintained throughout the pregnancy. A normal baby was born at term with a head circumference of 35.5 cm; at 1 year of age no abnormality is detectable. These results show that with careful monitoring and compliance it is possible, and may be advisable, to maintain plasma phenylalanine concentration within the normal range in the management of PKU pregnancy.

Acute metabolic decompensation in maple syrup urine disease (MSUD) during otherwise minor illnesses has generally been presumed to result from massive release of leucine from protein catabolism. A stable isotope method based on the continuous infusion of (2H5)phenylalanine was used to measure protein metabolism in vivo in two children with MSUD during acute illness and when well. Net protein catabolism was greater in the unwell state (0.51 and 0.40 gm/kg per 24 hours in each child, respectively) than in the basal state (0.34 and 0.32). This rate of release of leucine from protein is compatible only with a slow (several days) rather than a dramatic rise in plasma leucine levels during acute illness in MSUD. Poor oral intake leading to a relative increase in time spent in the fasting state appears to be a more important determinant of increasing leucine levels than the catabolic effect of infection in itself. These factors suggested that branched-chain amino acid restriction should be commenced at the start of minor illness in children with MSUD, and that intake of other nutrients should be maintained or increased throughout the illness. A regimen based on these concepts was used during nine episodes of minor illness in two children with MSUD. Plasma branched-chain amino acid levels remained acceptable (less than 700 mumol/L) throughout each of these episodes. Dietary supplementation of this type may reduce the risk of metabolic decompensation during acute illnesses in children with MSUD.

A preterm infant with classic phenylketonuria required rather less than 90 mg/kg of phenylalanine and between 270 and 290 mg/kg tyrosine daily to achieve a rate of weight gain of around 20 g/kg per day. Using Lofenalac as the low phenylalanine food, the intake of tyrosine, an essential amino acid for patients with phenylketonuria seemed to be limiting in respect of growth.

The outcome of 12 children with classical maple syrup urine disease is reviewed. All patients presented in the neonatal period at ages varying from 5 to 21 (median 8) days. The time taken to make the diagnosis ranged from 1 day to longer than 9 months (median 7 days). Each survived his initial illness but 3 died later after apparently mild infections. Three of the 12 patients had a spastic quadriplegia and 6 others abnormal neurological signs without clear cerebral palsy. The single most important factor determining the outcome appears to be the time taken to make the diagnosis after the first symptoms. Two patients were diagnosed within 24 hours of the first symptoms and one is of above average ability. The other is mildly retarded but control of the disease was poor in his first 4 years of life. Outcome is unpredictable if the delay is between 3 and 14 days. Two children are of normal ability but 6 others are retarded. A delay longer than 14 days is invariably associated with mental retardation and cerebral palsy. We conclude that early diagnosis is essential to improve the outcome of this condition.

A total of 795 382 infants born in north London was screened for phenylketonuria using the Guthrie test between October 1969 and December 1978. During this period it became recognised that phenylketonuria is not a single disease entity but one that encompasses a number of disorders of differing clinical and biochemical severity. The overall incidence of persistent hyperphenylalaninaemia was of the order of 7 per 100 000 births (or 1 in 15 000) and all the early treated patients made normal developmental progress. During the study there was an appreciable fall in the incidence of uncomplicated transient hyperphenylalaninaemia with or without tyrosinaemia. This reduction coincided with the change in infant feeding practice in the UK which led to lower intakes of protein and phenylalanine. It was concluded that any infant found to have a persistent blood phenylalanine concentration of 240 mumol/1 (4 mg/100 ml) or greater should be followed closely.

Cases of infants with life-threatening nutritional deficiencies due to certain strict vegetarian or macrobiotic diets have recently been reported.'-7It has been suggested that in some instances this should be regarded as a form of child abuse.'We have recently come across infants who have been severely ill as a result of similar nutritionally inadequate diets.During the past 12 months nine have presented to the Hospital for Sick Children, Great Ormond Street, and six to the Queen Elizabeth Hospital for Children, Hackney Road.In these cases the parents, not themselves on exotic diets, have used alternative "milks" for infant feeding in good faith on the basis of publications or on the recommendation of those providing professional advice.We wish to draw attention to the practice of lay, professional, or commercial recommendations of what may often be inappropriate infant feeds.Most mothers interested in vegetarian and other cult diets are keen to breast-feed their infants, and it is only when this is not possible that other feeds are used.Relatively minor problems may arise when infants are in- appropriately given supplements to their basic milk feeds or when they are on mixed diets.For example, the frequent use of Complan as an addition to infant feeds has resulted in at least two referrals to this hospital of infants with uraemia, later found to be due to excessive protein intake.The manu-

dietary management of glycogen the primary biochemical abnormalities.

Eighty-two cases of protracted diarrhoea in infancy presenting over a 6-year period have been analysed, with particular reference to diagnosis and management. The patients fell into 1 of 2 categories according to whether a specific diagnosis was established or not. A diagnosis (category 1) was established in 59 (72%), the commonest diagnoses being coeliac disease (33-2%), secondary disaccharide intolerance (12-2%), and cows' milk protein intolerance (12-3%). Other diagnoses included primary sucrase-isomaltase deficiency, Shwachman's syndrome, ulcerative colitis, ganglioneuroma, defective opsonization, staphylococcal pneumonia, and Hirschsprung's disease. Despite intensive investigation in diagnosis could not be established in 23 (28%) infants (category 2). Age of onset of symptoms in this group tended to be earlier than in category 1 patients, and 6 (7%) presented with diarrhoea dating from birth. Of particular interest in these 6 patients was the high incidence of associated extraintestinal anomalies, and of sibs who had died after protracted diarrhoea dating from birth. 4 of these 6 infants died, accounting for a mortality of 5% for the whole series. The remaining 17 (21%) patients in category 2 presented at a mean age of 4-9 weeks with a range of 1-18 weeks. All these 17 patients made an excellent response after institution of a chicken-based dietary formula, the details of which are presented. The pathophysiological mechanisms which may be operating in infants with protracted diarrhoea are discussed.

A diagnosis of Type I hyperprolinaemia was made in a 7-month-old infant who presented with hypocalcaemic convulsions and malabsorption. The plasma levels of proline were grossly raised and the urinary excretion of proline, hydroxyproline, and glycine was increased; neurological development was delayed and there were associated abnormalities of the electroencephalogram, renal tract, and bones. Restriction of dietary proline at the age of 9 months resulted in a prompt fall of plasma levels of proline to normal, and a low proline diet was continued until the age of 27 months when persistence of the biochemical defect was shown. During the period of dietary treatment, growth was satisfactory, mental development improved, and the electroencephalogram, and the renal, skeletal, and intestinal abnormalities disappeared. Proline should be regarded as a `semi-essential9 amino acid in the growing infant.

This study compares three feeding regimens for infants with phenylketonuria diagnosed by neonatal screening. Group 1 (five children) received Minafen (Cow & Gate) until they weighed twice their birthweights; Aminogran (Allen & Hanbury) was then added to the feeds in increasing amounts and replaced Minafen at between 8 and 10 months of age. Group 2 (five children) received Aminogran from the neonatal period. Group 3 (five children) received Minafen until they weighed twice their birthweights; Cymogran (Allen & Hanbury) was then added in increasing amounts and replaced Minafen at between 8 and 10 months of age. In all three groups growth was normal and control of phenylalanine levels satisfactory. During the first few months of life the Aminogran regimen proved more complicated and caused more practical difficulties than the regimens starting with Minafen. Later in the first year, when mixed feeding was introduced, and particularly when the bottle was exchanged for the cup, Aminogran had advantages over Cymogran because of its low calorie content, small bulk, and less unpleasant taste. At this age feeding problems were fewer and easier to manage with Aminogran than with Cymogran. A method of using Aminogran in the management of such problems is described. For these reasons, the regimen fed group 1, in which Minafen is used initially and then replaced by Aminogran, is preferred to the other two.

SummaryAn infant presented at three months with intolerance of glucose, galactose and fructose. It was thought that this was possibly secondary to staphylococcal infection. Recovery of monosaccharide intolerance coincided with recovery from infection. The infant was treated with a sugar‐free diet which is described in detail. It is important to detect and correct ketosis and/or hypoglycaemia at an early stage when such a diet is used.