点击显示 收起
Institute for Clinical Research and Health Policy Studies
Tufts-New England Medical Center
750 Washington Street
Box 63
Boston, MA 02111
E-mail: jlau1{at}tufts-nemc.org
Nutrition and Metabolic Disease Research Institute
Sanford Research–University of South Dakota
Department of Medicine
Sanford School of Medicine
University of South Dakota
Vermillion, SD
Cardiovascular Nutrition Laboratory
Jean Mayer USDA Human Nutrition Research Center on Aging
Tufts University
Boston, MA
Dear Sir:
We thank Vos et al for their comments on our article (1), but we disagree with their statement that our review showed "broad enthusiasm for fish oil and outright dismissal of ALA [-linolenic acid]." Our systematic review focused on the health effects of dietary n–3 fatty acids on clinical cardiovascular outcomes in humans and evaluated available evidence according to predefined questions. To minimize bias, conclusions were drawn based on the studies that met predefined criteria.
Vos et al did not disagree with the criteria we used to determine the quality of evidence. However, they seem to differ in what they would consider to be valid evidence by referencing several studies that did not meet our predetermined criteria. They refer to an acute, short-term experimental study in dogs to illustrate the potential beneficial antiarryhymic effects of -linolenic acid (ALA) (2). In the study by Billman et al (2), ALA was infused intravenously in an exercise-ischemia model of ventricular fibrillation. In fact, we reviewed this study in a separate article addressing the question of the effect of n–3 fatty acids on selected arrhythmia outcomes in animal models (3). Even if evidence from a dog study could be used to infer a benefit in humans, the experimental setting is highly unphysiological. Furthermore, whether such high plasma ALA concentrations are even achievable in a human consuming ALA is unknown (however, it is very unlikely).
Vos et al also cited the study by Freese (4), in which ALA supplementation reduced in vitro measures of platelet aggregation, as evidence that the reduction in platelet aggregation plausibly contributed to a reduction in the risk of cardiovascular events. In this study, very high intakes of both ALA (6 g/d) and eicosapentaenoic acid + docosahexaenoic acid (EPA + DHA; 5.2 g/d) were provided, neither of which has been shown at these doses to be cardioprotective. In addition, no effects were seen on plasma lipids for either ALA or EPA + DHA, and the minor effects on platelet aggregation do not explain the reduction in cardiac events that have been observed with 1 g EPA + DHA.
The Lyon Heart Study (5) was not designed to show and cannot be construed as showing that ALA was the agent responsible for the reduction in clinical events. Multiple variables were manipulated in that trial. Simply because serum concentrations of ALA were inversely associated with a reduced risk does not indicate that ALA that was responsible for the reduced risk. Association is not cause and effect, and a "true–true-and-unrelated" association is always possible.
Experimental studies conducted in animals and in vitro studies conducted in humans that focus on intermediate outcomes are important to uncover the mechanisms involved in the potential beneficial cardiovascular outcomes of n–3 fatty acid consumption in humans. The findings in the articles referred to by Vos et al support the hypothesis that ALA is cardioprotective and are reinforced by recent epidemiologic data (6, 7). However, these promising results must still be directly tested in human randomized controlled trials, rather than the benefits to humans assumed. Whether ALA has similar beneficial cardiovascular effects through diet or as supplements remains to be explored. Direct evidence does not support the view that ALA reduces the risk of cardiovascular events.
ACKNOWLEDGMENTS
JL and AHL have no conflict of interest to declare. WSH is a scientific advisor to OmegaMetrix LLC, Monsanto, CardioTabs, and TherRx.
REFERENCES