The role of background diet on the effects of eicosapentaenoic acid and docosahexaenoic acid supplementation in healthy pre-menopausal women: a randomized, cross-over, controlled study

Abstract

Background: The links between dietary fat intake, polyunsaturated fatty acid intake and breast cancer risk remain equivocal, with some studies pointing to improvements in risk upon omega-3 supplementation. However, the background diet is poorly controlled in most studies, potentially confounding this link. Therefore, this study examined the hypothesis that in order to see the benefits of omega-3 fatty acid supplementation, the background diet must be low in fat.

Methods: Of the 56 healthy, pre-menopausal women randomized to one of two experimental arms, consisting of a two-treatment, randomized, cross-over design, 41 completed the 10 month intervention. The two diet phases (habitual and low-fat) were separated by a washout phase, each lasting 3 menstrual cycles. During each diet phase, women were supplemented with 1.2 g eicosapentaenoic acid + docosahexaenoic acid per day.

Results: Red blood cell fatty acid composition indicated that more eicosapentaenoic acid and docosahexaenoic acid was incorporated in the low-fat diet than the habitual diet, though both diet phases resulted in significant increases in the omega-3 to omega-6 ratio. In the context of omega-3 supplementation in breast cancer risk reduction, we also measured fatty acid incorporation into nipple aspirate fluid. Similar changes to red blood cells were noted in nipple aspirate fluid, with higher incorporation of eicosapentaenoic acid in the low-fat diet phase.

Conclusions: These data suggest that the total level of dietary fat has some direct impact on fatty acid partitioning in addition to the recognized importance of fatty acid ratios, and supports the hypothesis that dietary fat intake must be considered a confounder in supplementation trials. Additionally, we demonstrate that n3 supplementation both reaches and imparts improvements in lipid content and n3:n6 at the target breast tissue.

Trial registration: Trial was been retrospectively registered at clinicaltrials.gov (Reg NCT02816125 ).

Keywords: Estrogen; Estrogen metabolites; Nipple aspirate; Omega-3 fatty acids; Urinary biomarkers.

Figures

Fig. 1

Experimental design and participant flowchart. MC = menstrual cycle

Fig. 2

Omega-3 and omega-6 content in red blood cells. a Omega-3 (n = 41), b omega-6 (n = 8) and c n-3:n6 content in red blood cell membrane phospholipids. Data are expressed as a percentage of total fatty acids ± standard error. a = significantly different from baseline, b = significantly different from low fat diet phase by paired t-test (p < 0.05)

Fig. 3

Omega-3 and omega-6 content in nipple aspirate fluid. a Omega-3, b omega-6 and c n-3:n-6 (n = 15) content in nipple aspirate fluid. Data are expressed as a percentage of total fatty acids ± standard error. a = different from baseline, b = different from LFD + supplement by ANOVA and Tukey's post-hoc test (p < 0.05)

Fig. 4

Blood serum and NAF estrogen levels. 17-β-estradiol was measured by radioimmunoassay at baseline and 3 MC after each intervention or washout period. a Average E2 levels ± SE in serum (n = 41). b Average E2 levels ± SE in NAF expressed per μg protein (n = 15). c correlation between serum and NAF E2 concentrations. Points represent the serum E2 value plotted as a function of the corresponding NAF value. a = different from baseline by ANOVA and Tukey's post-hoc test (p < 0.05)

Fig. 5

Relative concentrations of urinary estrogen metabolites 2-OHE1 and 16α-EOE1. Complete morning voids were collected on the same day of the menstrual cycle each month. Portions were frozen at −80 °C until analysis. Freshly thawed samples (n = 34) were thawed and estimates of metabolites determined by ELISA as described in the Methods. Values are in arbitrary units (ratio) ± SE. a = significantly different from LFD + supplement (p < 0.05) by ANOVA followed by Tukey's post-hoc test