Arch Dis Child 2000;82:493–494 493 Breast milk docosahexaenoic acid (DHA) correlates Arch Dis Child: first published as 10.1136/adc.82.6.493 on 1 June 2000. Downloaded from with DHA status of malnourished infants Ella N Smit, Esther A Oelen, Ejaz Seerat, Frits A J Muskiet, E Rudy Boersma Abstract ished, mostly breast fed, 4–56 month old chil- Aim—To investigate whether low docosa- dren, living in and around Islamabad, North hexaenoic acid (22:6ù3; DHA) status of Pakistan.3 malnourished, mostly breast fed infants is We hypothesised that the low RBC DHA a result of low ù3 fatty acid intake via content in these infants was a result of low breast milk. ù3-LCPUFA intake via breast milk. In an Methods—Fatty acid composition of attempt to test this hypothesis we studied the breast milk of eight Pakistani mothers, FA composition of both the breast milk of eight and of the erythrocytes of their malnour- Pakistani mothers and that of the RBC of their ished children was analysed. malnourished children. Results—The milk of the Pakistani moth- ers contained low percentages of all ù3 Subjects and methods and most ù6 fatty acids, compared with The study population consisted of eight milk of Dutch mothers. Breast milk DHA mother–child pairs from a low socioeconomic was positively correlated with infant class. They were recruited from the Nutrition erythrocyte DHA and arachidonic acid Rehabilitation Center of the Pediatric Depart- (20:4ù6). ment, Federal Government Services Hospital, Conclusion—DHA status of these mal- Islamabad. The children were classified as nourished children is strongly dependent malnourished, defined as weight for age below on the ù3 fatty acid intake from breast the mean minus 2SD, according to WHO milk. Augmentation of the infants’ ù3 long growth charts. EDTA anticoagulated blood chain polyunsaturated fatty acid status, or (2.5 ml at most) of the children was taken in an the ù3 and ù6 fatty acid status in general, undefined metabolic state. Breast milk samples by supplementation is indicated in de- were collected by manual expression. Fatty acids (FA) were determined as their methyl Departments of prived circumstances where access to Obstetrics and fresh fish is diYcult. However, in terms of esters by capillary gas chromatography with Pediatrics, Perinatal prevention, maternal supplementation of flame ionisation detection. Nutrition and these long chain polyunsaturated fatty Milk FA composition of Pakistani mothers http://adc.bmj.com/ Development Unit, acids, preferably from early pregnancy was compared with previously collected ma- Groningen University onwards, may be a better option. ture milk FA data of 25 Dutch mothers. Hospital, Groningen, Between group diVerences were analysed using Netherlands (Arch Dis Child 2000;82:493–494) E N Smit the Mann–Whitney U test at p < 0.05. Corre- E A Oelen Keywords: fatty acid; docosahexaenoic acid; lations of the diVerent PUFA between breast E R Boersma malnourished; breast milk milk and infant RBC, and between the various PUFA in infant RBC were tested with the Department of Spearman test at p < 0.05. on September 24, 2021 by guest. Protected copyright. Pathology and Postnatal docosahexaenoic acid (22:6ù3; Laboratory Medicine, DHA) status has been found to be related to Results Groningen University visual acuity, neurodevelopment, and behav- Hospital The median (range) duration of lactation of the F A J Muskiet iour. DHA and other long chain polyunsatu- Pakistani mothers was 12.5 (4.5–21) months. rated fatty acids (LCPUFA) are widely consid- Table 1 shows the breast milk FA composition, Federal Government ered to be essential in the prenatal and early together with the data of the Dutch mothers on Services Hospital, postnatal periods, because the synthesis of postnatal day 89 (77–103). Milk of Pakistani Department of LCPUFA from their precursors, á-linolenic mothers contained significantly lower amounts Pediatrics, Nutrition Rehabilitation Center, acid (18:3ù3; ALA) and linoleic acid (18:2ù6; of all ù3FA. Among the ù6FA, LA, 18:3ù6, Islamabad, Pakistan LA), does not seem to cover the infants’ high 20:2ù6, 20:3ù6, and AA contents were signifi- E Seerat needs at this stage of development. Breast milk cantly lower. contains in general a suYcient quantity of Breast milk DHA and infant RBC DHA Correspondence to: LCPUFA, including DHA and arachidonic were strongly correlated (r = 0.8571, Dr E N Smit, c/o Prof. Dr E R Boersma, Department of acid (20:4ù6; AA) to serve these needs. p = 0.007). Breast milk DHA correlated posi- Obstetrics and However, the LCPUFA composition of human tively with infant RBC AA (r = 0.7857, Gynecology/Pediatrics, milk is to a great extent dependent on the p = 0.021), but negatively with 18:1ù9 Perinatal Nutrition and 1 Development Unit, maternal diet. Therefore in North Pakistan, (r = −0.8571, p = 0.007). Milk eicosapentae- University Hospital low breast milk ù3 (LC)PUFA levels might be noic acid (20:5ù3; EPA) correlated with infant Groningen, Dept CMC 5, expected, because of the predominant use of RBC DHA (r = 0.7186, p = 0.045). There k.Y.3179, PO Box 30.001, 9700 RB Groningen, ALA poor corn oil and ghee, the low intake of were no other significant correlations between Netherlands green leafy vegetables (a source of ALA), and milk PUFA and infant RBC PUFA. email: almost no consumption of (ù3LCPUFA rich) There was a positive correlation between [email protected] fish.2 In a previous study we found low infant RBC DHA and infant RBC AA Accepted 4 February 2000 erythrocyte (RBC) DHA levels in malnour- (r = 0.8571, p = 0.007). 494 Smit, Oelen, Seerat, Muskiet, Boersma Table 1 Fatty acids in mature human milk of Pakistani and other favourable eVects, but also as the mothers of malnourished children compared with Dutch major source of dietary ù3LCPUFA. Arch Dis Child: first published as 10.1136/adc.82.6.493 on 1 June 2000. Downloaded from controls Apart from the adequate supply of Pakistani (n = 8) Dutch (n = 25) LCPUFA, many other factors are important for cognitive function. Overall nutritional ù3 series 18:3ù3 0.34 (0.28–0.51)* 1.12 (0.64–2.19) status, iron, iodine, and zinc deficiency, all 20:5ù3 0.02 (0.01–0.09)§ 0.05 (0.00–0.29) seem to be related to mental development. 22:5ù3 0.05 (0.04–0.14)† 0.13 (0.09–0.22) Finally, poor social and family circumstances 22:6ù3 0.06 (0.04–0.14)† 0.14 (0.10–0.65) sumù3 0.53 (0.37–0.64)* 1.45 (0.90–2.54) and a less stimulating environment could also LCPUFAù3 0.13 (0.09–0.29)† 0.32 (0.23–1.17) aVect later achievements. These unfavourable ù6 series conditions do prevail in circumstances where 18:2ù6 8.73 (7.45–10.61)† 13.84 (6.44–27.69) 6 18:3ù6 0.05 (0.03–0.12)‡ 0.10 (0.06–0.19) malnutrition is often encountered. 20:2ù6 0.15 (0.14–0.17)* 0.25 (0.18–0.37) Nevertheless, the very low DHA status of 20:3ù6 0.21 (0.15–0.36)† 0.29 (0.18–0.43) these children is a matter of concern, because 20:4ù6 0.26 (0.20–0.44)§ 0.33 (0.24–0.43) 22:4ù6 0.06 (0.04–0.08) 0.06 (0.04–0.08) of its potentially adverse eVects on brain devel- 22:5ù6 0.02 (0.00–0.04) 0.02 (0.00–0.04) opment, including that of the retina.1 Regard- sumù6 9.35 (8.22–11.69)† 14.80 (7.40–28.97) LCPUFAù6 0.69 (0.57–0.96)† 0.96 (0.77–1.22) ing the poor ù3LCPUFA status of their moth- ers, it may be expected that this condition was Data represent median (range), and are expressed in mol% already present during gestation and immedi- (mol/100 mol). Duration of lactation: Pakistan (median 12.5; range 4.5–21 months); Netherlands (mean 89; range 77–103 ately after birth, exposing the vulnerable fetus days). and neonate to low concentrations of *p < 0.0001; †p < 0.005; ‡p < 0.001; §p < 0.05. ù3LCPUFA. LCPUFA, long chain polyunsaturated fatty acid (C > 20, at least three double bonds). Milk DHA was also positively associated with infant RBC AA. Moreover, infant RBC In summary, milk of Pakistani mothers con- DHA and AA were positively correlated. tained low percentages of all ù3 and most ù6 Others have suggested that ù6FA cannot accu- mulate normally in cell membranes when fatty acids, compared with milk of Dutch 7 mothers. Breast milk DHA and EPA were ù3FA supply is low. It is therefore tempting to positively correlated with infant RBC DHA, speculate that the low DHA intake by these while milk DHA was also associated with malnourished children may not only cause low infant RBC AA. RBC DHA concentrations, but may also have adverse eVects on AA status. Augmentation of the infants’ ù3LCPUFA Discussion status, or the ù3 and ù6FA status in general, by Milk of Pakistani mothers feeding malnour- supplementation is indicated in deprived areas ished children was found to contain low where access to fresh fish is diYcult. Fish oil percentages of all ù3PUFA and most ù6PUFA seems to be the most logical supplement. As (LA, 20:2ù6, 20:3ù6, and AA) compared to purified fish oil is rather expensive, the cheaper http://adc.bmj.com/ milk of Dutch mothers. As the milk FA cod liver oil could be a useful alternative. composition is largely dependent on the Moreover, cod liver oil contains vitamins A and maternal diet1 it is conceivable that the above D.