Effect of prenatal EPA and DHA on maternal and umbilical cord blood cytokines

Inflammation has been associated with many complications of pregnancy, including perinatal depression, gestational diabetes, preterm delivery and preeclampsia [1, 2]. For example, a recent case-control study found that elevated concentrations of TNF- α in mid-pregnancy were significantly associated with diagnoses of preeclampsia later in pregnancy, [3] Similarly increases in circulating IL-1-β, IL-6, IL-18 and TNF have been demonstrated in women with GDM complicating pregnancy compared with controls [2]. Likewise, a recent case-control study of inflammatory markers among subjects with perinatal depression found that a diagnosis of major depressive disorder was significantly related to increased concentrations of TNFα [4].

The Western diet has been postulated to be one important cause of inflammation [5, 6]. Some investigators have related the increase in inflammatory disorders in the developed world to an imbalance in dietary intake in omega-6 to omega-3 fat compared to historical human diet [7]. Dietary supplementation with omega-3 fatty acids has therefore been proposed as a mechanism to potentially reduce inflammation [7].

Omega-3 fatty acids, both in food and through supplementation with fish or algal oil, have long been believed to limit or modulate the inflammatory response through modulation of cytokine production. Two early omega-3 fatty acid supplementation studies in healthy adults using greater than 2.4 g of EPA plus DHA daily reported significantly reduced blood cytokines TNFα, IL1β, and IL6 [7‐9]. However, a systematic review of randomized trials using DHA and EPA supplementation among healthy adults found no consistent effect of supplementation on blood cytokines [10]. The majority of studies investigating the effect of dietary supplementation on cytokines have measured cytokines after endotoxin stimulation of cells, whereas, more recent studies have measured circulating soluble cytokines using multiplex immunoassays [11]. It is not known whether fish oil supplementation during pregnancy modulates maternal cytokine production or whether supplementation alters cytokine concentrations in the umbilical cord blood.

This study is a secondary analysis of a double-blind randomized control trial carried out to test the hypothesis that EPA- or DHA-rich fish oil supplementation would reduce perinatal depressive symptoms. A secondary aim of the study was to determine the effect of prenatal EPA- and DHA- risk fish oil supplementation on maternal cytokine profiles as well as on cytokines in umbilical cord blood [12]. We hypothesized that prenatal EPA- and DHA-rich fish oil supplementation would reduce concentrations of the major pro-inflammatory cytokines IL-β, IL-6, and TNF-α in both maternal and umbilical cord plasma.

This study is a planned secondary analysis of plasma samples that were collected as part of a prospective, double-blind randomized controlled trial of fish oil supplementation for prevention of depressive symptoms among pregnant women [12, 13]. To be eligible to participate in the parent study, women had to be between 12 and 20 weeks gestation in a singleton pregnancy and to be at risk for depression based on an Edinburgh Postnatal Scale Score between 9 and 19 or a past history of depression. Potential subjects consuming more than two fish meals per week were not eligible to participate. Subjects were recruited from the antenatal clinics of the University of Michigan Medical Center (Ann Arbor, MI) and St. Joseph Mercy Health System (Ypsilanti, MI). The overall details of the trial design have been previously reported [12, 13].

The sample size for the study was determined based on the expected Beck Depression Inventory scores among women at risk for depression. The parent study was designed to detect a 50% reduction in the primary outcome for the trial, the Beck Depression Inventory score at 6 weeks postpartum [12, 13]. Maternal and umbilical cord cytokine levels were a pre-specified secondary outcome of the study [12]. We used a random number table maintained in the University of Michigan Investigational Drug Pharmacy for random allocation. Full details of the randomized, double-blinded study design as well as the primary and clinical outcomes have been previously reported [12, 13].

There were 126 subjects enrolled in the parent study between October 2008 and May 2011. They were randomly assigned to receive EPA rich fish oil (1060 mg EPA plus 274 mg DHA), DHA rich fish oil (900 mg DHA plus 180 mg EPA), or soy oil placebo and were followed longitudinally through pregnancy from their enrollment at 12–20 weeks’ gestation until 6 weeks postpartum. Maternal blood was drawn at enrollment and again at 34–36 weeks gestation. The minimum supplementation time was 14 weeks. Because of the double-dummy design of the trial, all subjects received some placebo capsules. Umbilical cord blood was obtained at delivery.

During the course of the study, eight subjects dropped out and 118 subjects completed the study. There were 126 maternal samples available for cytokine analysis at enrollment and 113 maternal samples available for analysis after supplementation (34–36 weeks). There were 102 cord blood samples available for analysis of cytokine profiles. The flow diagram of participant enrollment and sample availability is depicted Fig. 1.

We found that some maternal plasma cytokine concentrations may be significantly modulated by dietary supplementation during pregnancy. Our study found that supplementation with EPA-rich fish oil in particular resulted in reduced concentrations of IL6 and TNFα, inflammatory cytokines that have been associated with major depression in other studies [16]. These findings are similar to the results of a randomized control trial of supplementation with 2 g/day of omega-3 fatty acids among obese pregnant women. In that trial, prenatal supplementation was found to reduce inflammatory markers in maternal adipose tissue and in the placenta [17]. This finding is in contrast with previous observations among healthy non-pregnant adults in which supplementation with fish oil up to 3.5 g/day had no significant effect on cytokine profiles [10]. The reason for this finding is unknown, but may relate to the unique maternal immunology during pregnancy. Our findings were consistent with previous reports which suggest that EPA may be more effective than DHA in reducing inflammatory biomarkers [18].

By contrast, we found that fish oil supplementation in doses of approximately 1 g/day had no significant effect on umbilical cord blood cytokine concentrations. Our findings are not consistent with a systematic review of prior prenatal supplementation trials that found significant reduction of IL13 in infants of mothers who took prenatal fish oil supplementation [19]. This finding is also in contrast to our previously reported secondary analysis of the effect of fish oil supplementation on chemokine profiles among participants in this trial. That analysis found that the EPA- and DHA-rich fish oil had no significant effect on maternal chemokine concentrations, but significantly influenced cord blood chemokine ratios [20]. The reason for this discrepancy is unknown but may be related to the number of cord blood samples with cytokine concentrations that fell below detectable limits.

Based on our findings, we conclude that the fatty acid content of the prenatal maternal diet may modulate the maternal immune response, although prenatal supplementation did not significantly affect cord blood cytokine profiles in our study.

Strengths of our study included blinded assessment of outcomes, randomized design, and intent-to-treat analyses. The study also had several limitations. These included the differing times of day at which maternal blood samples were procured, a result of study procedures being timed with clinical care. However, all maternal samples were drawn after a four hour fast with only low-fat meals allowed on the day of the blood draw.

Another limitation of our study is that many cytokine concentrations were below the limits of detection for the assay used. These results were consistent with the results from other studies analyzed in the same laboratory and all of the study samples were analyzed on the same date. Likewise, the sample size of our study was based on the primary outcome of the Beck Depression Inventory score at 6 weeks’ postpartum. Nevertheless, our post-hoc power analysis suggests that this secondary study had reasonable power to detect a treatment effect of the fish oil supplementation on cytokine concentrations.

Additionally, our study was also limited by its double-dummy design which meant that all subjects in the intervention groups received soy oil placebo capsules along with the EPA- and DHA- rich fish oil capsules. Finally, it should be noted that the subjects were selected based on predisposition to depression and may have not been representative of the general population of pregnant women.

Our findings allow us to hypothesize that among women with perinatal depressive symptoms related to inflammation, prenatal EPA-rich fish oil supplementation may reduce plasma concentrations of inflammatory cytokines. Current research in depression is moving toward identification of a subset of depressed individuals in whom dysregulated cytokine production may play a role in pathogenesis. Future research should focus on the role of omega-3 fatty acids, in particular EPA, in modulation of the immune response in this population, and should work to elucidate the relationship of inflammatory cytokines with depressive symptomatology in these individuals.