Recent studies have shown appreciable differences in the composition of term (T) and preterm (PT) milk. During the early weeks of lactation PT milk is much richer in protein and several minerals3-8 and may be closer attuned to the nutritional needs of the preterm baby. Lemons et al.8 calculated that 170-200 ml/kg/day of 'average' PT milk would supply all the nutrients except calcium, phosphorus, and possibly iron, needed by a preterm infant of 33 weeks' gestation to sustain the intrauterine growth rate. They concluded that the milk from the baby's mother was preferable to T milk or pooled mature breast milk, but that supplementation with calcium, phosphorus, and possibly other factors might be warranted. Gross et al.4 stated that 'PT milk appears to approximate more closely the nutritional needs of the preterm infant than does other breast milk', but later5 they reported wide variability in the composition of PT milk from different women and warned that the nutrient concentrations might sometimes be inadequate for the baby's needs. Hibberd et al.9 also found wide variation in the composition of PT milk. They were unconvinced of the nutritional superiority of PT milk and suggested that a higher nitrogen content resulted from 'leakage' of serum proteins into the alveoli of the immature mammary gland and was not therefore an adaptation to the needs of the preterm infant.The question of the adequacy of milk from surrogate mothers and indeed of the maternal milk as sources of vitamins for the rapidly growing preterm infant has so far received little attention. There is as yet no published report comparing the vitamin composition of T and PT milk and the present study was undertaken to provide data on the B vitamin content of such milk. Materials and methods Milk samples. Milk samples were obtained from the

A survey was undertaken to update and extend available information on the vitamin content of pasteurized milk as produced at processing dairies in mainland UK and to investigate regional, seasonal and breed effects. The concentration of total retinol in milk from non-Channel Island (NCI) breeds averaged 619/tg/lOOg in summer and 41-2/£g/100g in winter. Concentrations of /^-carotene were 31*5 and 105/tg/100 g in summer and winter respectively. Concentrations of retinol in milk from Channel Island (CI) breeds were similar, but concentrations of /^-carotene were on average 3 times higher. The concentration of vitamin D 3 in milk from NCI breeds was 0-033/ig/100 g in summer and 0026 /ig/100 g in winter. There was no marked seasonal variation in the mean concentration of total vitamin C (14-5 /*g/ml). Values for the concentration of B vitamins (/jg/ml) were: folic acid 0-060, vitamin B 12 0'0042, riboflavin 1-78, nicotinic acid 0-71, pantothenic acid 3-60, biotin 0-020, thiamin 0-46 and vitamin B 6 061. Seasonal variation in the concentration was most marked for folic acid (c.v. 17-4%) and to a lesser extent for vitamin B 12 (c.v. 10-3%). The only breed differences in the B vitamin content were for riboflavin and folic acid, the mean values obtained for milk from CI breeds being respectively 20 and 10 % higher than those from NCI breeds.Because of its high nutritive value and the quantities consumed, milk is one of the most important primary commodities in the British diet. In addition to protein, fat, carbohydrate and Ca, milk is also a significant source of many vitamins and trace elements.Although monthly averages of butterfat and solids-not-fat in raw milk are published by the Milk Marketing Board, no recent comprehensive data on the nutrient content of pasteurized bulk milk from the British herd are available. In the most recent edition of Composition of Foods (Paul & Southgate, 1978) all values, apart from riboflavin, vitamin C and most minerals (analysed in bottled milk by the * Present address: Academia Medyczna, Zaklad Bromatologii,

SummaryConcentrations were measured of different vitamins in milk as delivered to the home and after storage in the domestic refrigerator, during summer and winter periods. Vitamin C concentrations varied greatly, ranging from 0·1 to 18·4 εg/ml. Most of the lower vitamin C values were associated with known exposure of the milk to daylight and there was a clear tendency for the concentrations to fall with increased time of refrigerated storage especially when associated with prior exposure of the milk to daylight. Riboflavin values were mostly in the range 1·35–1·92εg/ml, but occasional lower values were associated with low content of vitamin C and in some instances with prolonged exposure to daylight on the doorstep. Concentrations of the other vitamins (εg/m1) were: folic acid 0·041–0·074, thiamin 0·38–0·48, vitamin B12 0·0032–0·0044 and vitamin B6 0·56–0·71. All fell within the normal ranges and were not significantly influenced by exposure of the milk to daylight or subsequent storage in the refrigerator.