点击显示 收起
CMC-V, Room Y 1.165
Groningen University Hospital
PO Box 30.001
9700 RB Groningen
Netherlands
E-mail: f.a.j.muskiet{at}lc.umcg.nl
Department of Pathology and Laboratory Medicine
University Medical Center Groningen
Groningen
Netherlands
Faculty of Earth and Life Sciences
Vrije Universiteit Amsterdam
Amsterdam
Netherlands
Dear Sir:
Brenna et al (1) conducted a meta-analysis of human milk docosahexaenoic acid (DHA) and arachidonic acid (AA) contents worldwide. Their study consisted of 65 reports on 2474 women with mostly Western dietary habits. In agreement with at least 2 previous studies (2, 3), they concluded that milk DHA is more variable than is milk AA and that the ratio of DHA to AA varies widely. The calculated mean (±SD) values for DHA (0.32 ± 0.22% by wt) and AA (0.47 ± 0.13% by wt) were suggested to serve as a guide for infant feeding.
Although Brenna et al's conclusion on variation is consistent with one of our previous reports (2), we currently feel that the low variation in AA might in reality derive from a sampling bias. We found much higher milk AA (median: 0.70 mol%) and high DHA (0.75 mol%) contents in lactating women in Doromoni (Tanzania) (4). Their milk showed a clear correlation between DHA and AA. Both of these long-chain polyunsaturated (LCP) fatty acids could be traced to life-long consumption of DHA- and AA-rich fish from nearby freshwater, Lake Kitangiri. It seems that milk AA is notably dependent on long-term AA intake, because supplementation of lactating women in Jerusalem with 300 mg AA for 1 wk did not increase milk AA (5). The sizeable short- and long-term dietary influences on human milk fatty acid composition raise the question of whether the resulting variance should be taken as testimony of the wide variety of foods that are tolerated by humans and their offspring or whether it indicates a lack of evolutionary pressure because of a rather constant dietary composition in the past. We recommend that worldwide human milk fatty acid composition should not be used as guide for infant formula. Just as Western serum cholesterol concentrations and vitamin D status should not serve as targets for recommendations, it seems inappropriate to use calculated mean milk DHA and AA concentrations of mostly Western women as a basis for infant formulas. At least 3 arguments favor higher intakes of both DHA and AA by our ancient ancestors, who consumed diets that were much closer to the environment on which our genome has adapted during the past 2–3 million years of evolution. First, the sites at which their fossil remains have been discovered support the notion that evolution to Homo sapiens took place on an LCP n–3-rich diet from African ecosystems that were located in places where the land meets with water (6, 7). Food from these ecosystems is rich in iodine, vitamins A and D, and n–3-fatty acids of both vegetable and animal origin. Contrary to popular belief, our ancient ancestors did not need fishing gear to benefit from the abundance of LCP n–3 fatty acids, and probably of LCP n–6 fatty acids, in such ecosystems, where it is relatively easy to hunt and gather anything ranging from spawning catfish, shellfish, and crustaceans (lobster, crab, shrimp, etc) to eggs, birds, and reptiles, which all ultimately receive their LCP n–3 fatty acids from plankton via the local food chain. The resulting diet with a high content of iodine, vitamins A and D, and LCP fatty acids seems to have been somewhat abandoned since the Out-of-Africa Diaspora, because deficiencies of these nutrients are among the most widely encountered in the current world population (6, 8). Second, epidemiologic data have shown a negative association of fish oil with coronary artery disease (CAD) and of fish consumption with (postpartum) depression. Landmark trials with -linolenic acid (9), fish oil (10), and eicosapentaenoic acid (EPA) (11) in CAD and with EPA in depression and schizophrenia (12) support the causality of these relations. As acknowledged by Brenna et al (1), the intake of marine food in inland and developed countries is usually low. Many authoritative organizations have issued recommendations to the general public ranging from "choose fish as a food item more often" to "consume 3 servings of fish per week." One may wonder what the milk DHA and resulting DHA status would be of children born to parents reaching the necessary DHA status to decrease their risks of CAD and psychiatric disease. If their parents would benefit from a higher than current Western DHA status, which cannot be achieved by the mere consumption of -linolenic acid, it seems reasonable to assume that this DHA status is appropriate across the Homo sapiens life cycle and that our genes may have evolved on this high DHA status. Meanwhile, it has also become clear that LCP fatty acids are not only important structural components of membranes but, together with their eicosanoid metabolites, are firmly implicated in gene expression, eg, as modulators of nuclear transcription factors such as peroxisome-proliferator activated receptors, sterol regulatory element-binding protein, and nuclear transcription factor B. Finally, many randomized controlled trials that used formulas with and without LCP fatty acids, and measured outcome variables such as retinal function, visual acuity, behavior, and cognitive and motor developments, have shown beneficial effects of LCP, notably of LCP n–3 fatty acids, in both preterm and term infants. Preterm infants benefit most, but many of the effects are transient. The effects are especially the result of DHA, but AA might be important for preserving the balance between n–3 and n–6 fatty acids. The present consensus from human and animal studies is that LCP fatty acid supplements have no effect on growth, neonatal brain DHA is positively related to cognitive and behavioral performance, the differences are difficult to detect with currently available tools, and that the encountered differences may nevertheless be relevant (13). All of these studies have been performed with DHA and AA intakes in the current Western human milk range, which might explain the modest effects.
In conclusion, we agree that DHA seems to be the most variable fatty acid in human milk worldwide. The variance in AA, however, may be underestimated because of sampling bias. Higher AA concentrations occur in women who consume diets similar to our ancient diet, which is part of the environment on which our genes evolved. Current Western human milk DHA contents do not comply with the recommendations of authoritative organizations, which advise increases in fish intake, and should consequently not serve as a guide for infant feeding. It would be of interest to see whether infant formulas with LCP fatty acid contents consistent with those of non-Western, traditionally eating populations would produce more pronounced effects using the many endpoints that have been studied with limited success until now.
ACKNOWLEDGMENTS
None of the authors declared a conflict of interest.
REFERENCES