Dr. Parker Replies

To the Editor: We agree that eicosapentaenoic acid efficacy for bipolar depression remains unproven. Of the four published randomized-controlled trials and one open-label trial of omega-3, two were reviewed in our article. Subsequently, a 12-week, double-blind, placebo-controlled study of 1 or 2 g/day of add-on ethyl-eicosapentaenoic acid for bipolar depression has been published (1) , reporting modest improvement in depression scores. As Drs. van Strater and Bouvy note, there have also been two negative clinical trials of 6 g/day of pure eicosapentaenoic acid for bipolar depression and for rapid cycling bipolar disorder—but published only as abstracts, reconciliation must be awaited.

We agree on the importance of identifying side effects, but since omega-3 supplementation has been researched for other conditions, much safety and side effect data are already available. Gastrointestinal complaints appear common, varying with quality and dose, but the American Heart Foundation (2) rates this side effect as very low for doses up to 1 g/day and moderate for high-dose (>3 g/day) supplementation.

Drs. van Strater and Bouvy raise questions regarding the possible effect of omega-3 on blood glucose levels in diabetic patients. However, a meta-analysis (3) of fish oil effects on glycemic control found no significant changes in HbA _1c percentages in diabetic patients. Fasting glucose levels were slightly increased in noninsulin-dependent diabetic patients and significantly lowered in insulin-dependent diabetic patients. While glucose level effects appear nonsignificant, caution with type II diabetes appears warranted, since the effect seems dose responsive (3) .

In relation to bleeding risk, omega-3 can decrease platelet aggregation, but there is little evidence that doses less than 3 g/day cause clinically significant bleeding (2) , even for individuals receiving aspirin or anticoagulants (4) . Extra care is suggested with supplementation exceeding 3 g/day for those with clotting disorders or receiving anticoagulants.

We note two other side-effects. Omega-3 can alter lipoprotein cholesterol levels, with high doses (4 g/day) decreasing serum triglyceride concentrations up to 30%, often with a corresponding increase in “good” high-density lipoprotein cholesterol (1% to 3%) and “bad” low-density lipoprotein cholesterol (5% to 10%) (2) . Second, omega-3 molecules are a potential target for free radical oxidative damage, albeit with health outcome effects unclear.

It is worth noting that the U.S. Food and Drug Administration has considered the potential effects of omega-3 on glycemic control in diabetes, on increased bleeding times, and on low-density lipoprotein cholesterol and determined that intakes of up to 3 g/day of marine omega-3 polyunsaturated fatty acids are “generally recognized as safe” (2) .