A Randomized Trial of Maternal Docosahexaenoic Acid Supplementation to Reduce Inflammation In Extremely Preterm Infants

Abstract

Maternal supplementation with 1000 mg/day docosahexaenoic acid (DHA) provides third trimester DHA accretion levels in breast milk for the preterm infant. We hypothesized that DHA supplementation to mothers providing breastmilk for extremely preterm infants would decrease inflammatory markers, in the infant. Mother/infant dyads (n=27) were enrolled at birth and mothers were assigned to receive 200 or 1000 mg/day of DHA. Milk and plasma samples were analyzed for fatty acids and inflammatory markers. Decreases in inflammation were observed in both maternal and infant plasma and correlated with RBC DHA levels. The fact that maternal DHA supplementation decreases infant markers of inflammation implies that DHA, delivered through breastmilk, has the potential to decrease inflammation in the infant.

Introduction

Low circulating concentrations of docosahexaenoic acid (DHA) have been associated with poor development in preterm infants (1). Unfortunately, intravenous sources of DHA are not widely used in the United States to meet third trimester targets of intrauterine accretion (55mg/kg/day). In addition, key nutrients such as DHA in the maternal diet can be limiting (2) and thus human milk DHA content is variable (3–5). Mother’s milk can however reach intrauterine DHA accretion targets with maternal dietary supplementation of 1000mg DHA per day (6, 7). The effects of maternal DHA supplementation on reducing inflammation in the infant. Proof of this concept has been demonstrated in our murine model that resulted in increased DHA concentrations, decreased inflammation, and improved lung development in the offspring (8); and in preterm infants correlating low DHA levels with increased risk of chronic lung disease. (9). DHA is the fatty acid responsible for optimum neurological development, and supplementation with single source DHA is likely to provide additional benefits over fish oil which is largely comprised of eicosapentaenoic acid (EPA) (10). For this study, our hypothesis was that 1000 mg (intervention) vs 200 mg (current clinical recommendations and control group) of DHA per day using an algal source, would result in improved maternal breast milk concentrations and thus enhance infant DHA intake. Furthermore, we hypothesized that improved human milk DHA intake by the preterm infants would result in decreased inflammation.

Methods

Study Design.

The present study was designed as a prospective, randomized trial to give either 200 mg (low) or 1000 mg (high) DHA (Expecta®) to mothers delivering and providing breast milk for their extremely preterm infants (ClinicalTrials.gov Identifier: , Institutional Review Board number 2012–0329, and the Food and Drug Administration number108241). All studies were conducted at University of Cincinnati Hospital (Cincinnati, Ohio). Mothers were recruited in the postpartum unit after delivery of an infant <28 weeks gestational age, who intended to provide breast milk for their infant, and whose infants did not have any congenital abnormalities that would interfere with the outcome parameters. Mother/infant dyads were randomized by the Investigational Pharmacy staff. The randomization scheme was generated by the Web site Randomization.com (http://www.randomization.com⟩using the method of randomly permuted blocks. The first randomization list was created using six blocks of four patients each, followed by a list containing five blocks of two patients each. Using computer generated codes and corresponding labeled bottles, supplements were either mailed or picked up at the Pharmacy. The investigators and mothers were blinded to the groups. DHA supplementation was started on the day of birth. The study was powered to detect differences in inflammatory markers in infant samples at 4 weeks of age from mothers supplemented with 200 mg or 1000 mg (alpha =0.05 and power = 80%).

Fatty acid measurement.

Maternal and infant blood or breast milk was collected to examine fatty acid levels at birth and at 2 and 4 weeks of age. Plasma and red blood cells (RBCs) were separated by centrifugation at the time of collection and stored at −80°C until analysis. Fatty acids were analyzed by gas chromatography with flame ionized detection as previously described (11).

Inflammatory markers.

Aliquots of maternal and infant plasma were used for cytokine measurements at baseline (prior to 7 days after delivery) and at 4 weeks. The inflammatory markers were measured using the Meso Scale Discovery multiplex platform. Samples were analyzed for TNFα, IL-6, IL-8, IL-1β, IL-10, INFγ, and IL12p70 in a single analysis. sRAGE was also measured using the same technique using a single analysis platform.

Statistics.

Data were analyzed by Univariate and Multivariate Linear Regression with treatment and time as variables and GA as a co-variate. Repeated measures analysis could not be performed due to the number of missing variables. Correlations were assessed by Spearman’s Correlation.

Results

Demographics.

One hundred thirty-nine mothers were assessed and twenty-seven mothers met the criteria and were enrolled. Eligibility included <28 weeks gestation and no genetic anomalies. Sixteen mother/infant dyads were enrolled in the 200 mg per day arm and 11 mothers and 14 infants in the 1000 mg/day arm. A total of 6 infants expired (2 after completing the study), 4 were withdrawn due to transfer, and 2 withdrew during the study. Eight infants completed the 200 mg/day arm (5 withdrawn or dropped, 3 deceased) and 10 completed the 1000 mg/day arm of the study (1 withdrawn, 3 deceased). Demographic data are indicated in Supplemental Table 1. There were no significant differences between 200 mg/day and 1000 mg/day DHA supplemented groups in sex, race, gestational age at birth, maternal body mass index (BMI), mode of delivery, or infant morbidities (Supplemental Table 1).

Effects of DHA supplementation.

DHA levels were measured in maternal breast milk and in both maternal and infant RBCs (Figure 1A–C). Greater breast milk DHA concentrations were observed at 4 weeks of 1000 mg/day compared to 200 mg/day DHA supplementation. Maternal RBC DHA concentrations were higher in the 1000 mg/day DHA supplemented group than the 200 mg/day group at 4 weeks (Figure 1B). Infant RBCs DHA concentrations were not different as a result of enhanced maternal supplementation. However, there was an effect of treatment on ARA/DHA ratio indicating either lower ARA or higher DHA levels (Figure 1 C and D).

Figure 1.

DHA measurements. DHA concentrations were measured in breast milk and maternal and infant RBCs as described in Methods. White bars represent low DHA supplementation and gray bars represent high DHA supplementation. Data were analyzed by Univariate General Linear Regression with treatment and time as variables. Data from each time point were analyzed by multiple t-tests, significance was determined when * p<0.05.

Maternal inflammatory markers.

Cytokine levels and sRAGE concentrations were measured in maternal plasma samples obtained at 0 and 4 weeks after birth (Table 1). Multivariate Regression indicated no differences in the levels. Between subject analyses identified an effect of time on IL-6, an effect of treatment in IL-8, and an interaction between time and treatment in IFNγ levels.

Table 1.

Inflammatory markers in infant and maternal plasma.

Maternal				Infant
	Low DHA (200mg/day)	High DHA (1000 mg/day)	between subject effects	Low DHA (200mg/day)	High DHA (1000 mg/day)	between subject effects
TNFα  0 weeks  2 weeks  4 weeks	3.66 (9.63, 2.50) 3.13 (7.04, 2.67)	3.44 (4.68, 2.23) 3.11 (4.21,2.64)		10.72(14.3, 6.85) 13.82 (18.3, 7.32) 11.74 (23.2, 7.36)	7.38 (19.4, 4.14) 14.1 (28.42, 9.20) 8.88 (11.40, 7.37)	treatment, p=0.001 time, p=0.009
IL-6  0 weeks  2 weeks  4 weeks	2.16 (15.6, 0.84) 1.26 (4.23, 0.62)	3.4 (14.13, 0.72) 0.61 (1.2, 0.45)	time, p=0.011	18.8 (89.1, 1.79) 15.2 (79.3,6.82) 20.4 (40.4, 1.12)	7.91 (49.8, 2.80) 9.66 (87.5, 2.19) 3.43 (65.81, 0.94)
IL-8  0 weeks  2 weeks  4 weeks	8.55 (10.54, 5.12) 6.80 (11.58, 2.36)	3.52 (6.88, 2.67) 5.00 (5.67, 3.60)	treatment, p=0.001	120 (1210, 14.4) 108 (182, 20) 285 (484, 61)	115 (589, 25) 72 (140, 12.7) 45 (190, 29.2)
IL1β  0 weeks  2 weeks  4 weeks	0.50 (1.83, 0.16) 0.26 (0.44,0.18)	0.30 (1.43,0.11) 0.30 (1.54, 0.25)		9.18 (45.5, 0.94) 7.03 (12.6, 4.35) 10.4 (16.6, 0.66)	3.70 (12.6, 0.35) 3.82 (8.75, 1.90) 1.63 (7.87, 0.41)
IL-10  0 weeks  2 weeks  4 weeks	1.40 (2.12, 0.64) 1.00(2.01, 0.61)	1.04 (3.34, 0.93) 1.39 (1.8, 0.52)		2.02 (6.19, 1.34) 6.01 (11.5, 2.18) 3.50 (14.7, 2.53)	2.92 (11.8, 1.84) 4.88 (20.6, 2.10) 2.94 (9.55, 1.23)
IFNγ  0 weeks  2 weeks  4 weeks	0.82 (2.47, 0.50) 0.81 (1.5, 0.57)	0.59 (3.98, 0.03) 0.94 (2.04, 0.45)	treatment*time, p=0.041	0.41 (2.60, 0.18) 1.95 (3.33, 0.04) 1.01 (1.23, 0.65)	0.29 (1.99, 0.07) 0.29 (0.60, 0.10) 0.87 (1.34, 0.34)
IL-12p70  0 weeks  2 weeks  4 weeks	0.50 (0.70, 0.09) 0.40 (0.58, 0.26)	0.36 (0.71, 0.09) 0.42 (0.67, 0.33)		0.23 (1.05, 0.06) 0.67 (1.39, 0.19) 0.60 (0.76, 0.30)	0.36 (0.66, 0.09) 0.76 (2.71, 0.26) 0.87 (1.34, 0.34)	time, p=0.034
sRAGE  0 weeks  2 weeks  4 weeks	526 (2520, 263) 486 (1340, 161)	491 (1270, 195) 580 (1170, 411)		746 (2201, 428) 1220 (2080, 122) 900 (1650, 218)	5336 (7270, 416) 1107(2320, 214) 1127 (2050, 256)	treatment, p=0.019 time, p=0.000 treatment*time, p=0.001

Data are presented as medians and (range).

Infant inflammatory markers.

Cytokine levels and sRAGE concentrations were measured in infant plasma samples obtained at 0, 2, or 4 weeks after birth (Table 1). Multivariate Regression indicated effects of time, treatment, and an interaction between time and treatment in the model. Between subject analyses indicated an effect of time and treatment on TNFα levels and an effect of time on IL12p70. sRAGE levels in the 1000 mg/day DHA supplemented group were disproportionately high at birth however, they returned to levels matching the 200 mg/day DHA group at 4 weeks, indicating statistically, an effect of time, effect of treatment, and an interaction between time and treatment. Overall, our data suggest a decrease in inflammatory markers in the infants whose mothers received 1000 mg/day DHA supplement compared to infants whose mothers received 200 mg/day DHA supplement.

Correlations.

Correlations were identified between maternal breast milk DHA levels and DHA levels detected in maternal RBCs (r²=0.55, p=0.004, Supplemental Figure 1A). Inverse correlations were observed between maternal plasma IL-6 (r²=0.523, p=0.005) or IL-8 (r²=0.635, p=0.006) levels and RBC DHA content (Supplemental Figure 1B and 1C). Infant plasma levels of IL-12 were negatively correlated with infant RBC DHA concentrations (r²=0.635, p=0.006, Supplemental Figure 1D). These data suggest a correlation between higher circulating DHA levels and lower concentrations of specific inflammatory markers.

Discussion

Maternal dietary supplementation has been shown to influence breast milk composition of select nutrients, including DHA (6). Supplementation with pre-formed DHA is important as studies with the precursor, linolenic acid, do not correlate with increased breastmilk DHA levels (12). Furthermore, preterm infants likely have limited lipid absorption and synthetic capacity. DHA has been shown to have anti-inflammatory properties (13). Consequently, supplementing preterm infants with milk that contains higher concentrations of DHA could attenuate inflammation and thus improve development.

While the mechanisms are not completely understood, the anti-inflammatory effect of DHA likely includes changes in receptor-mediated signaling, changes in membrane fluidity, and/or enhancement of the production of anti-inflammatory lipid mediators due to the availability of DHA as substrate (14, 15). More specifically, modulation of immune cell activity by long chain fatty acids such as DHA can affect macrophage adhesion and decrease phagocytosis (16). Premature infants, who miss DHA accretion during the third trimester of development have lower overall intake of DHA and this deficit is likely to contribute to their vulnerability to inflammation and inflammation-mediated cell injury after birth.

This study tested the hypothesis that maternal dietary DHA supplementation, while providing breast milk for their infant, would provide anti-inflammatory effects to the infant. We developed a Consort Framework for a randomized control trial of 1000 mg/day versus 200 mg/day DHA (control group-current standard of care). No statistical differences were found in demographics or postnatal morbidities between the two groups (Supplemental Table 1).

As we have previously reported, supplementation with 1000 mg/day peaked maternal breast milk DHA levels within four weeks (Figure 1A) (6). Furthermore, increased breast milk DHA concentrations correlated with increases in maternal RBC DHA levels (Figure 1B). No differences were observed in infant RBC DHA levels (Figure 1C). However, as an alternative measurement, the relative ratio of ARA to DHA were calculated and this ratio was significantly lower in the infants receiving milk from high DHA supplemented mothers indicating that relative amounts of DHA were greater in the infant plasma (Figure 1D). Collectively, our data support several other studies that maternal DHA supplementation provides additional DHA to the mother, and by means of breast milk to the infant.

Our findings suggest that maternal DHA supplementation of 1000 mg/day suppresses the expression of inflammatory cytokines both in the mother and in the infant (Table 1). Furthermore, the expressions of specific cytokines are inversely related to RBC DHA content (Supplemental Figure 1). We propose that the attenuated inflammation in the 1000 mg/day DHA supplemented infants provides a more suitable environment to promote growth. Poor growth has been associated with other chronic inflammatory diseases in children (17) and elevation in inflammatory cytokines such as IL-6 (18). DHA supplementation in animal models has been shown to increase bone growth (19, 20) and may offer a mechanism by which linear growth in infants may be improved.

The primary limitations of our findings are the small sample size and the complex morbidities found in such immature infants. Our sample size prevented the use of Repeated Measures Analyses which would have given the study more power to detect differences. In addition, the small size also prevented the ability to statistically control for any of the infant morbidities defined in Supplemental Table 1, however the low and high supplementation groups had similar diagnoses. In conclusion, maternal supplementation with a higher dose of DHA can impact maternal milk DHA concentrations and may influence infant cytokine levels. While not conclusive, our data provides information that the clinical team may use to discuss maternal diet strategies.

Summary.

What is known?

• Preterm infants experience inflammatory responses to clinical interventions

• Unchecked inflammation can lead to further tissue injury

• DHA has anti-inflammatory properties

• DHA is essential to neurological development

What is new?

• Maternal DHA supplementation provides anti-inflammatory effects for the mother

• Maternal DHA supplementation provides anti-inflammatory effects to preterm infants through breast milk

Abbreviations:

DHA

docosahexaenoic acid

EPA

eicosapentaenoic acid

ARA

arachidonic acid

GA

gestational age

TNFα

tumor necrosis factor alpha

IL-6

interleukin-6

IL-8

interleukin-8

IL-1β

interleukin-1 beta

IL-10

interleukin-10

IL-12

interleukin-12

INFγ

interferon gamma

sRAGE

soluble receptor for advance glycation end products

ELISA

enzyme-linked immunosorbent assay

RBC

red blood cells

BMI

body mass index