Subjects and methods

The study population consisted of eight mother–child pairs from a low socioeconomic class. They were recruited from the Nutrition Rehabilitation Center of the Pediatric Department, Federal Government Services Hospital, Islamabad. The children were classified as malnourished, defined as weight for age below the mean minus 2SD, according to WHO growth charts. EDTA anticoagulated blood (2.5 ml at most) of the children was taken in an undefined metabolic state. Breast milk samples were collected by manual expression. Fatty acids (FA) were determined as their methyl esters by capillary gas chromatography with flame ionisation detection.

Milk FA composition of Pakistani mothers was compared with previously collected mature milk FA data of 25 Dutch mothers. Between group differences were analysed using the Mann–Whitney U test at p < 0.05. Correlations of the different PUFA between breast milk and infant RBC, and between the various PUFA in infant RBC were tested with the Spearman test at p < 0.05.

Results

The median (range) duration of lactation of the Pakistani mothers was 12.5 (4.5–21) months. Table 1 shows the breast milk FA composition, together with the data of the Dutch mothers on postnatal day 89 (77–103). Milk of Pakistani mothers contained significantly lower amounts of all ω3FA. Among the ω6FA, LA, 18:3ω6, 20:2ω6, 20:3ω6, and AA contents were significantly lower.

Breast milk DHA and infant RBC DHA were strongly correlated (r = 0.8571, p = 0.007). Breast milk DHA correlated positively with infant RBC AA (r = 0.7857, p = 0.021), but negatively with 18:1ω9 (r = −0.8571, p = 0.007). Milk eicosapentaenoic acid (20:5ω3; EPA) correlated with infant RBC DHA (r = 0.7186, p = 0.045). There were no other significant correlations between milk PUFA and infant RBC PUFA.

There was a positive correlation between infant RBC DHA and infant RBC AA (r = 0.8571, p = 0.007).

In summary, milk of Pakistani mothers contained low percentages of all ω3 and most ω6 fatty acids, compared with milk of Dutch mothers. Breast milk DHA and EPA were positively correlated with infant RBC DHA, while milk DHA was also associated with infant RBC AA.

Discussion

Milk of Pakistani mothers feeding malnourished children was found to contain low percentages of all ω3PUFA and most ω6PUFA (LA, 20:2ω6, 20:3ω6, and AA) compared to milk of Dutch mothers. As the milk FA composition is largely dependent on the maternal diet1 it is conceivable that the above differences reflect nutritional differences. The low Pakistani milk ω3LCPUFA content is presumably caused by low fish consumption in North Pakistan.2 In this respect it is noteworthy that fish intake in the Netherlands is also considered to be low, which is reflected in relatively low plasma phospholipid DHA content compared to other countries.4 The milk ω3-LCPUFA contents were nevertheless two times lower in the Pakistani than in the Dutch samples.

A strong positive correlation between breast milk DHA concentrations and infant plasma and RBC phospholipid DHA in 12 week old exclusively breast fed children has been reported by Gibson et al.5 Similarly, we observed a strong correlation between breast milk DHA and EPA contents on the one hand and infant RBC DHA content on the other hand. This is of particular interest, as the children in the present study were much older (4.5–21 months) and weaned. It suggests that the DHA status of these malnourished children is still strongly dependent on the DHA and EPA intake from breast milk and probably much less on the ω3PUFA intake from weaning food. Prolonged breast feeding of these malnourished Pakistani children seems therefore not only important for its anti-infective properties and other favourable effects, but also as the major source of dietary ω3LCPUFA.

Apart from the adequate supply of LCPUFA, many other factors are important for cognitive function. Overall nutritional status, iron, iodine, and zinc deficiency, all seem to be related to mental development. Finally, poor social and family circumstances and a less stimulating environment could also affect later achievements. These unfavourable conditions do prevail in circumstances where malnutrition is often encountered.6

Nevertheless, the very low DHA status of these children is a matter of concern, because of its potentially adverse effects on brain development, including that of the retina.1 Regarding the poor ω3LCPUFA status of their mothers, it may be expected that this condition was already present during gestation and immediately after birth, exposing the vulnerable fetus and neonate to low concentrations of ω3LCPUFA.

Milk DHA was also positively associated with infant RBC AA. Moreover, infant RBC DHA and AA were positively correlated. Others have suggested that ω6FA cannot accumulate normally in cell membranes when ω3FA supply is low.7 It is therefore tempting to speculate that the low DHA intake by these malnourished children may not only cause low RBC DHA concentrations, but may also have adverse effects on AA status.

Augmentation of the infants' ω3LCPUFA status, or the ω3 and ω6FA status in general, by supplementation is indicated in deprived areas where access to fresh fish is difficult. Fish oil seems to be the most logical supplement. As purified fish oil is rather expensive, the cheaper cod liver oil could be a useful alternative. Moreover, cod liver oil contains vitamins A and D. Malnourished children often suffer from a deficiency of these essential nutrients as well.

However, future efforts could be better directed at the supplementation of these LCPUFA to Pakistani women from early pregnancy onwards in order to prevent any adverse affects on perinatal brain development of the child. By doing so both the low transplacental supply to the fetus and a low postnatal supply to the newborn via breast milk could be avoided. Consequently, the LCPUFA status of early brain development could be improved.