Maternal caffeine consumption and offspring DNA methylation Original paper

In this meta-analysis, there was little to no support for an effect of caffeine on DNA methylation.

This Study Summary was published on April 17, 2024.

Quick Summary

In this meta-analysis, there was little to no support for an effect of caffeine on DNA methylation.

What was studied?

The association between caffeine consumption during pregnancy and markers of offspring cord blood DNA methylation.

Who was studied?

A total of 3,725 mother-child pairs.

The mothers were between weeks 12 and 22 of gestation. The participants were from 6 different pregnancy and birth cohorts, including 2 UK cohorts, 1 Dutch cohort, 1 Norwegian cohort, 1 Spanish cohort, and 1 French cohort.

How was it studied?

In this meta-analysis, daily caffeine consumption was estimated from mother’s self reports of caffeine-containing beverages (e.g., coffee, tea, cola) on food frequency questionnaires.

The researchers adjusted their analyses for making multiple comparisons (Bonferroni correction) and adjusted their analyses for the potential confounders of offspring sex, gestational age at birth, maternal education, maternal age, maternal BMI, maternal smoking, and number of previous children.

What were the results?

Overall caffeine consumption was associated with one CpG site (cg19370043; the nearest gene was PRRX1), and cola consumption was associated with one CpG site (cg14591243; the nearest gene was STAG1). A CpG site is a particular sequence in the genome (cytosine nucleotide next to a guanosine nucleotide or C next to G) which can be methylated.

Anything else I need to know?

The researchers noted that the results from this study had a small effect size, and the findings were not consistent across different sources of caffeine. They speculated that if there was a true effect of caffeine on DNA methylation, it should have been present from analyses of coffee and tea as well as cola.

Smoking has been associated with 6,000 changes to DNA methylation at birth, in contrast with the 2 changes found in this study.

This sample included a somewhat diverse European population. However, the generalizability to other populations, especially outside of Europe, is reduced. One especially different aspect in this sample was the rate of smoking (15%, vs. 5% in the USA in 2021)ref=https://www.cdc.gov/nchs/products/databriefs/db458.htm.

This Study Summary was published on April 17, 2024.