randomized designs; however, very few studies of the effects of maternal nutrition on milk volume meet this criterion. Thus, the subcommittee also reviewed observational studies in humans, which are useful in establishing associations between factors, and studies in animals, which can suggest hypotheses to be tested in humans. The following sections begin with discussions of data on animals and progress to studies in humans, reflecting the chain in which evidence is usually accumulated.
In examining the evidence relating energy balance to milk volume, the subcommittee addressed three major questions:
Is the volume of human milk affected if energy intake is curbed or supplemented during lactation?
Do maternal fat stores or weight relative to height affect the relationship between energy deficit and milk volume?
Are the mechanisms of energy utilization during lactation relevant to the volume of milk produced by lactating women?
Several investigators have developed animal models of malnutrition during lactation, primarily in rats. Studies by Warman and Rasmussen (1983), Young and Rasmussen (1985), and Kliewer and Rasmussen (1987) illustrate that milk yield is decreased by dietary restriction and that the decrease is more pronounced in rats restricted before and during lactation than it is in those restricted only during lactation. Milk yield was reduced only 12.5% in dams fed 75% of ad libitum intake by controls, suggesting that the underfed animals compensated for dietary restriction in some way. In contrast, there was a dramatic (52%) decrease in milk yield in rats restricted to 50% of ad libitum intake (Young and Rasmussen, 1985). These results suggest that there may be a threshold below which lactation can no longer be protected when food intake is restricted. Similar findings have been reported by Roberts et al. (1985) in studies of baboons. Among animals restricted to 80% of ad libitum intake, milk output was not significantly reduced, whereas milk output decreased 20% in those restricted to 60% of ad libitum intake. Reduced physical activity may protect milk output at moderate levels of energy restriction but not at high levels, when body stores were shown to be mobilized at a rapid rate in the baboons. However, this possible effect of physical activity has not been studied.
The relative energy costs of lactation are much lower for humans than for most other species, and it is not known whether there is an energy threshold for humans. Prentice and Prentice (1988) report that energy costs at peak milk output, as a function of maternal body weight, are 4- to 15-fold lower for humans than for either laboratory or domesticated animals. For example, the energy requirements of lactation in humans can be met by increasing energy intake by approximately 25%, whereas in rats, energy intake must increase by