Diet Can Impact Microbiota Composition in Children With Autism Spectrum Disorder

Abstract

Diet is one of the most influential environmental factors in determining the composition of the gastrointestinal microbiota. Microbial dysbiosis in children with Autism Spectrum Disorder (ASD) and the impact of some bacterial taxa on symptoms of ASD has been recognized. Children with ASD are often described as picky eaters with low intake of fiber-rich foods, including fruits and vegetables. However, the impact of diet on the microbiota composition in children with ASD is largely unknown. Herein, fecal samples, 3 day food diaries and the Youth and Adolescence Food Frequency questionnaire (YAQ) were collected from children with ASD (ASD; n = 26) and unaffected controls (CONT; n = 32). Children's ASD symptoms were determined using the Pervasive Developmental Disorder Behavior Inventory Screening Version (PDDBI-SV). Differences in the microbiota composition at the phyla, order, family, and genus level between ASD and CONT were observed. Microbiota composition of children with ASD was investigated in relation to feeding behavior, nutrient and food group intake as well as dietary patterns derived from the YAQ. In children with ASD, two distinct dietary patterns (DP) were associated with unique microbial profiles. DP1, characterized by higher intakes of vegetables, legumes, nuts and seeds, fruit, refined carbohydrates, and starchy vegetables, but lower intakes of sweets, was associated with lower abundance of Enterobacteriaceae, Lactococcus, Roseburia, Leuconostoc, and Ruminococcus. DP2, characterized by low intakes of vegetables, legumes, nuts and seeds and starchy vegetables, was associated with higher Barnesiellaceae and Alistipes and lower Streptophyta, as well as higher levels of propionate, isobutyrate, valerate, and isovalerate. Peptostreptococcaceae and Faecalibacterium predicted social deficit scores in children with ASD as measured by the PDDBI-SV. Diet-associated microbial profiles were related to GI symptoms, but no significant interaction between nutrition and microbiota in predicting social deficit scores were observed. In conclusion, dietary patterns associated with fecal microbiota composition and VFA concentrations in children with ASD were identified. Future studies using a larger sample size and measuring other behaviors associated with ASD are needed to investigate whether dietary intake may be a modifiable moderator of ASD symptoms.

Keywords: autism; dietary patterns; feeding behavior; microbiota; nutrients.

Figure 1

Principal co-ordinate analysis based on unweighted UniFrac (A) and weighted UniFrac Distance (B) generated from fecal samples of children with ASD (ASD) and unaffected controls (CONT). PERMANOVA analysis indicated that overall bacterial communities differed between ASD and CONT group (p = 0.02) based on unweighted UniFrac, but not on weighted UniFrac distance.

Figure 2

(A) Relative mRNA abundance of butyrate-producing gene BCoAT and propionate-producing gene mmDA in feces of ASD and CONT children measured by qPCR. (B) Differences in VFA concentrations between CONT and ASD. ^*p ≤ 0.05; ^†p ≤ 0.1; Data is expressed as mean ± SD; mmDA, methylmalonyl-CoA decarboxylase; BCoAT, butyryl-CoA:acetate CoA acyltransferase; ASD, Autism Spectrum Disorder group; CONT, unaffected control group; VFA, volatile fatty acids; DMB, dry matter basis.

Figure 3

Correlation between intake of specific food group and abundance of bacteria in ASD. Spearman Correlation coefficient for selected food groups and nutrients and abundance of bacteria showing strongest correlation; (A) insoluble dietary fiber and Clostridiales; (B) Fried Food and Faecalibacterium; (C) Fruit and Faecalibacterium; all correlations are shown in Supplemental Table 4.