The Effect of Early and Late Feeding and Glucagon upon Blood Sugar and Serum Bilirubin Levels of Premature Babies.

Many problems relating to the premature baby have received intensive study during the past few decades. This has resulted in more efficient care of the premature with a subsequent decrease in the mortality rate. It is surprising, however, that there is still so much difference of opinion regarding the feeding of premature babies, not only with respect to the type of food but also to the timing of the first feed. Former authorities apparently favoured early feeding. For instance Goodhart (1913) wrote: 'these feeble infants must not be allowed to wait two or three days for regular feeding with the mother's milk; the loss of weight and possible rise of temperature which such waiting involves-may be the last straw for one of these infants, who is fighting a feeble struggle for existence'. Hess (1923) recommended that milk should be given to premature babies from 12 hours of age; he stated, 'the necessity of an early supply of food cannot be over-emphasized, as even the better developed infants do not withstand prolonged starvation'. During the past decade or so a number of writers have advised that feeds should be withheld from premature babies for periods of 12 to 96 hours depending upon the degree of immaturity at birth, mainly on the grounds that the premature is particularly liable to pneumonia due to the aspiration of vomit. Newborn premature infants frequently manifest oedema suggesting that they are overhydrated, and Smith (Smith, Yudkin, Young, Minkowski and Cushman, 1949; Hansen and Smith, 1953; Smith, 1957) has shown that relatively large amounts of water and solutes are excreted during the first few days of life, the rate of excretion being proportional to the degree of oedema. He recommended that feeding should be delayed for as long as four days. Although thirsting babies were unable to maintain homeostasis, clinically their condition remained satisfactory and any haemoconcentration was promptly reversed after one day's fluid intake. Recently Smith (1962) appears to have had second thoughts about the advisability of an initial starvation period. Gaisford and Schofield (1950) were also strong protagonists of a three-tofour-day starvation period for premature infants. They thought the time of the first feeding should be determined by the strength of the cry and the disappearance of oedema. Crosse (1957) wrote about the possible dangers of overfeeding during the first week of life and considered that the survival rate of premature infants had been improved by giving nothing by mouth for several days after birth. She recommended starvation periods varying from 12 hours to four days depending upon the birth weight. Ylppo (1954) challenged the concept that prematures should be starved initially, believing that fasting caused acidosis and disturbances of renal function. He compared the effects of early and late feeding regimes in two hospitals and found little difference in mortality rates. He considered that prematures should be given fluid and food as early as possible. There appear to have been very few controlled studies on premature infant feeding. Gleiss (1955) reported a mortality of 41% in 92 babies first fed at 36 hours of age, and a mortality of 28% in 102 infants first fed between 12 and 24 hours. These death rates are very high and it is not clear what kind of baby he was studying. Analysis of his figures of mortality show that they are not statistically significant (X2 = 2 99). Bauman (1960) studied 50 premature babies; 26 were given no fluids for 24 hours, 24 received glucose and saline by