Diet and Obesity: Effects on Birth Outcomes Via DNA Methylation

• In Event: Metabolic State and Nutrition during Pregnancy: Mechanisms and Moderators of Effects on Offspring Health and Development

Abstract

Introduction: Nutrient intake during pregnancy alters the DNA methylation of metabolism-related genes in the placenta and the offspring. In the adult, dysregulated transcription of LEP, LEPR, ABCG1 and PHOSPHO1 genes are known to cause chronic disease. No study has tested the individual effects of each essential nutrient on the methylation pattern of these genes in the human placenta. A quantile regression method may account for the biological interdependence of DNA methylation within a gene by detecting patterns of association across the methylation distribution.

Hypotheses: 1) Average nutrient intake during pregnancy is associated with DNA methylation of LEP, LEPR, ABCG1, and PHOSPHO1 in the placenta. 2) DNA methylation of LEP, LEPR, ABCG1, and PHOSPHO1 in the placenta is a mediator of nutrient intake’s effects on birth weight.

Study population: Healthy, pregnant individuals, ages 20 to 45, who consented to providing placental samples.

Methods: This was a cross-sectional analysis of nutrient intake during pregnancy and birth outcomes. Dietary and supplemental intakes were collected up to three times across 2nd and 3rd trimesters using a web-based automated self-administered 24-hour dietary recall (ASA24). DNA methylation of placental samples were analyzed using the Illumina Infinium 450K BeadChip. Average fractional DNA methylation (%5mC) of all available BeadChip probes associated with LEP, LEPR, ABCG1 and PHOSPHO1 were averaged for each gene. Quantile regression analysis was used to measure the effect of nutrition on %5mC and birthweight at the 5th, 10th, 25th, 50th, 75th, 90th and 95th quantiles.

Results: Ninety patients with dietary and placental methylation data were included in this analysis. Nutrient intakes were below the estimated average requirements for calcium (16.7% of participants), magnesium (53.3%) and potassium (98.9%), and above the tolerable upper intake level for sodium (84.4%), vitamin B6 (86.7%), folate (93.3%), iron (51.1%), zinc (21.1%). Median macronutrient densities were within population normal. Eighteen participants were categorized as having had pre-pregnancy obesity (18.6%).

Both calcium (β=- 0.0005, p=0.008) and maternal pre-pregnancy maternal obesity (β =2.1315, p=0.005) were independently associated with LEPR %5mC at the 95th quantile, in a multivariable quantile regression model controlling for maternal age, nutritional supplemental assistance, gestational age at birth, and baby sex. Of the B vitamins, only riboflavin (β=-0.8584, p<0.001, at the 95th quantile of LEPR %5mC) was associated with %5mC. Fat density was inversely associated with both LEPR %5mC (β= -13.748, p<0.001) and ABCG1 %5mC (β=-10.598, p<0.001) at the 5th quantiles. All models reported are fully adjusted.

Both calcium (β=0.00061, p<0.0001) and fat density (-2.5805, p=0.041) were associated with birthweight, but only in at the 95th quantile. Neither LEPR nor ABCG1 %5mC explained these associations.

Associations between micronutrient intake, fat density, obesity and placental methylation were observed, but primarily at the tails of methylation distribution. Transcriptional activity would need to be measured in order to confirm the impact of these effects. The population’s high intakes of many B vitamins may have reduced sensitivity in detecting the effect of these critical methyl donors and cofactors.

Authors

• Angelie Singh, Ben Gurion University of the Negev
  Presenting Author

• Catherine Monk, Columbia University Irving Medical Center

• Michael Friger, Ben Gurion University of the Negev