Early-childhood dietary patterns associate with asthma and stool-plasma metabolomic signatures

Childhood asthma has been linked to individual foods, nutrients, diet-quality scores, and broad dietary patterns, but specific early-childhood food co-consumption patterns and their microbial/metabolic correlates remain unclear.

Objective

To identify data-driven early-childhood dietary patterns associated with asthma/wheeze, evaluate prospective associations with age-6 asthma/wheeze, assess external support in NHANES, and characterize associated gut microbiome and stool/plasma metabolomic profiles.

Methods

We analyzed age-3 food frequency questionnaire data from children in the Vitamin D Antenatal Asthma Reduction Trial. Dietary patterns were derived from log-transformed, energy-residualized, standardized food-frequency variables using principal component analysis. Associations with age-3 asthma/wheeze were tested using covariate-adjusted logistic regression. Prospective associations were evaluated using age-6 asthma/wheeze as the outcome. Leading PC food-cluster proxies were evaluated in NHANES 2021-2023 among children aged 2-3 years, with sensitivity analyses in ages 2-5 and 2-8 years. Selected PCs were tested for associations with gut microbiome, stool metabolome, and plasma metabolome features.

Results

PC1 contrasted a sweet snack/fried-food pattern with a fruit/vegetable-rich pattern, whereas PC3 captured a processed meat/fried-food axis. PC3 showed the strongest positive association with age-3 asthma/wheeze (odds ratio per 1-SD increase, 1.42; P = 0.00109). Age-3 dietary PCs were prospectively associated with age-6 asthma/wheeze, with the overall PC set improving model fit in permutation testing (likelihood-ratio statistic = 20.3; empirical P = 0.033) among 394 cases and 397 controls. In NHANES children aged 2-3 years, the PC3 food-cluster proxy was positively associated with current asthma (odds ratio, 1.59; 95% confidence interval, 0.95-2.67). PC3 was also linked to gut microbial and stool/plasma metabolomic variation, including steroid sulfate, vitamin E-related, nucleoside-related, and lipid-related metabolites.

Conclusions

Early-childhood asthma/wheeze-associated dietary signals were better represented as food co-consumption patterns than isolated single-food effects. Age-3 dietary patterns were associated with concurrent and prospective asthma/wheeze, showed directionally consistent NHANES support, and were linked to microbiome and metabolomic variation.