Saliva analyses show potential in identifying predictive factors for infant food reactions

Key takeaways:

• Multi-omics involve the genome, microbiome and epigenome.
• Maternal age and Asian race were associated with food reactions.
• Messenger RNA and microRNA were elevated in infants with food reactions.

LOUISVILLE, Ky. — Multi-omic analyses of saliva may improve understanding of what causes immune system-mediated food reactions, according to a presenter at the American College of Allergy, Asthma & Immunology Annual Scientific Meeting.

“What are the risk factors that are associated with these food reactions, and can we potentially identify who these patients are before they develop these food reactions?” Shane Stone, DO, internist and pediatrician in the department of pediatrics, Penn State College of Medicine, said during his presentation.

Stone noted that food reactions have been associated with medical and demographic traits as well as environmental, immunological and biological factors.

“There certainly are epigenetic regulators such as microRNA and different microbial factors that may impact the development of food reactions,” Stone said. “Several studies have shown that the microbiome is able to influence different inflammatory modulators that may be involved in food reactions.”

However, Stone continued, not many studies have examined how these various factors interact or how they could be used to predict potential for food reactions.

“For our study, we aimed to identify multi-omic factors in infant saliva to better understand how the host-microbial interactions can contribute to the origins of food reactions,” he said.

Stone and his colleagues defined multi-omics as an analysis of various “-omes,” including the genome, microbiome and epigenome. Specifically, they examined infant saliva because of the diverse biogeographical continuum of the oral mucosa, which includes microbiota, messenger RNA (mRNA) and microRNA.

The prospective longitudinal study dichotomized 119 term infants (girls, n = 71; white, n = 93) into those with food reactions (n = 27) and those with no food reactions (n = 92) at 12 months. Standardized surveys were used to assess medical, demographic and environmental traits.

The researchers defined food reactions as any problem caused by food, such as an allergic reaction, sensitivity or intolerance on the Infant Feeding Practice-II survey confirmed by medical record review.

Next, the researchers collected two saliva samples from the infants at 6 months. One sample was used for enzyme-linked immunoassay testing to measure cytokine levels, and the other was used for RNA sequencing to measure microRNA and microbial RNA.

“We also performed a logistical regression that helped us to assess whether or not these ‘omic’ factors could identify which infants were at risk of developing food reactions relative to the other risk factors,” Stone said.

Medical and demographic factors associated with food reactions included maternal history of food allergies, maternal age and Asian race. However, there were no significant differences between the groups in terms of environmental factors.

“We assessed for different things like tobacco exposure, prematurity and antibiotic exposure,” Stone said, “but we did not find any significant difference here.”

Stone further said that the researchers were able to identify several factors that were elevated in infants who developed food reactions compared with those who did not. These included the STAT6 mRNA, the Let-7-e-5p and miR-151a-5p microRNA, and the Cyanobacteria phylum.

“We also noticed a slight elevation in the IL-8 to IL-6 ratio,” Stone said, adding that previous literature has indicated associations between these factors, pro-inflammatory states and allergic inflammation.

Unfortunately, Stone continued, none of these omic factors were more predictive for the development of food reactions compared with a simple model using demographic traits and environmental exposures.

“We suspected this is likely in part related to our somewhat smaller sample size,” Stone said.

Although there was no enhanced predictive value in using multi-omics compared with simple models, the researchers concluded, further investigation could enable clinicians to prospectively identify which infants are at risk and provide them with appropriate referral and anticipatory guidance.

“Our hope is that with future studies, we will be able to further identify which factors may be predictive and see if there is any underlying multi-omic network that could better explain the pathogens that set up these food reactions,” Stone said.