1989). The effect of lactation on magnesium absorption is not known. Despite the relatively large difference between recommended and actual magnesium intakes, the only evidence of maternal magnesium deficiency during lactation appears in two anecdotal reports of extraordinary cases: in a wet nurse whose milk output was estimated to be 1,700 ml/day for 3 months (Greenwald et al., 1963) and in a woman who secreted three times the normal level of magnesium in her milk (Kamble and Ookalkar, 1989). Women in Ghana who breastfed for up to 12 months post partum had levels of serum magnesium lower than those of nonlactating controls (Fenuku and Earl-Quarcoo, 1978), but in lactating U.S. women whose mean magnesium intake was 248 mg/day, neither plasma magnesium nor erythrocyte magnesium differed from values for nonlactating women (Moser et al., 1983). Magnesium intake from diet or supplements has not been associated with maternal plasma or erythrocyte magnesium (Moser et al., 1983), nor is it likely to influence magnesium concentrations in milk (see Chapter 6). The long-term impact of low magnesium intake on the mother's well-being has not been studied.
The vitamins most likely to occur in low levels (relative to the RDAs) in the diets of lactating women are B6, E, thiamin, and folate. Vitamin B6 levels in milk are strongly influenced by dietary intake: levels in the milk of women supplemented with 2.5 mg/day are twice as high as those of unsupplemented women (192 compared with 93 µg/liter) (Styslinger and Kirksey, 1985), but the RDA increment for lactation appears generous even if based on the vitamin B6 content of the milk of supplemented women. On the other hand, protein intakes by lactating women in the United States are high (average predicted intakes range from 111 to 166% of the RDA; Table 9-3); thus, the RDA increment allows for the increase in vitamin B6 requirement that accompanies increases in protein intake.
Low vitamin B6 intakes during lactation may adversely affect both the infant and the mother, although evidence of overt deficiencies is rare in the United States. Kirksey and Roepke (1981) reported three cases of breastfed infants with central nervous system disorders, which they attributed to vitamin B6 deficiency. The mothers of all three infants had been long-term (4- to 12-year) users of oral contraceptives prior to pregnancy at a time when the estrogen content of such preparations was much higher than it is currently; these mothers were considered to have inadequate vitamin B6 status. Levels of plasma pyridoxal 5-phosphate (PLP) tend to be lower in breastfed infants than in formula-fed infants, and there is a correlation between vitamin B6 levels in human milk and infant plasma PLP (Andon et al., 1989; McCoy et al., 1985). However, none of the infants in two studies of unsupplemented lactating women in the United States (N = 6 in Styslinger and Kirksey  and N =