A comparison of two methods of milk sampling for calculating the fat intake of breast-fed babies

Fat intake at a breast feed was estimated by either sampling milk intermittently during the feed using a thin latex nipple shield ('shield method') or by expressing samples before and after the feed ('expression method'). These sampling techniques were compared within 30 mother/infant pairs using a randomized cross-over design. Mean fat intake estimated by the expression method was significantly greater than that estimated by the shield method (first breast P less than 0.0001, second breast P less than 0.002). This finding was partly explained by a significantly lower mean milk intake when the nipple shield was used, but the higher fat content of expressed milk samples compared to samples obtained during suckling also played a part. Use of the nipple shield confirmed that breast-milk fat content does not change in a predictable way during a feed. Nevertheless, for practical purposes, the advantages of studying changes in milk composition during the feed have to be set against the potential restrictive effect of the nipple shield system when it is used to sample intermittently during the feed.