Fecal Short-Chain Fatty Acids in Pregnancy and Offspring Asthma and Allergic Outcomes

To the Editor:

Asthma is the most common chronic disease in children and comorbid environmental allergies have prognostic relevance¹. Short chain fatty acids (SCFA), including acetic acid, butyric acid and propionic acid, are metabolites produced by intestinal microbes through fermentation of dietary fiber². Immune-modulating properties of SCFA are numerous and include increasing T regulatory cell differentiation³. Fecal SCFA during infancy are associated with reduced risk of asthma with allergic sensitization later in life^4–6. In mouse models, SCFA supplementation during pregnancy leads to reduced allergic airway disease in offspring^7,8 and in 40 mother-child pairs, low maternal serum acetic acid was associated with increased cough or wheeze in infant offspring⁷. Other than this, the association of SCFA during pregnancy and offspring risk of asthma and/or allergy has not been reported.

We hypothesized that SCFA during pregnancy, measured via mass spectrometry, are associated with reduced asthma and allergy in offspring. Methods are detailed in the Online Repository. Subjects were 150 mother-child pairs who participated in the Vitamin D Antenatal Asthma Reduction Trial (VDAART), a randomized controlled trial of Vitamin D during pregnancy for prevention of asthma ( NCT00920621, Tables E1–3). The study protocol was approved by the institutional review boards at each participating institution and at Brigham and Women’s Hospital. All participants provided written informed consent.

Participating mothers provided stool samples during the 3^rd trimester of pregnancy in which SCFA were measured by liquid chromatography with tandem mass spectrometry. We examined both absolute concentrations of SCFA and relative concentrations of each SCFA as a proportion of all measured SCFA. In general, the ratio of fecal acetic acid:butyric acid:propionic acid is approximately 60:20:20. Deviations from this ratio reflect differences in bacterial composition and other exposures. For example, relative acetic acid has repeatedly been associated with breastfeeding and associated gut microbes⁹. Additionally, abundance of one SCFA could influence the functions of others via mechanisms such as competition at common G-protein coupled receptors² and effects on bioavailability. So, relative proportions may be more important in some contexts than absolute concentrations.

Clinical outcomes in offspring were ascertained at age 6 years. Asthma/wheeze was based on parental report of physician-diagnosed asthma, recurrent wheeze and asthma medication use. Allergic sensitization was defined by serum specific IgE ≥ 0.35 kU/L to at least one environmental or food allergen. Hay fever was defined by parental report of physician-diagnosed allergic rhinitis. Seventy (47% of) children had asthma/wheeze, 38 (48%) had allergic sensitization and 54 (40%) had hay fever. Seventeen (21%) had both asthma/wheeze and sensitization and 33 (24%) had both asthma/wheeze and hay fever.

In logistic regression analyses, neither absolute nor relative maternal fecal SCFA were associated with the individual outcomes of asthma/wheeze, sensitization or hay fever (Figure 1, Table E4). However, relative, but not absolute, acetic acid concentration was inversely associated with asthma/wheeze with sensitization (odds ratio (OR) for 10% increase in relative acetic acid concentration 0.46, 95% confidence interval (CI) 0.22–0.94, p=0.036) (Figure 2, Table E4). A similar non-significant inverse association was seen between maternal fecal relative acetic acid and asthma/wheeze with hay fever (OR 0.63, 95% CI 0.37–1.05, p=0.077) (Figure 2). Permutation testing yielded similar results (Table E5 in the Online Repository).

Figure 1.

Associations of maternal absolute and relative fecal SCFA and offspring clinical outcomes. Odds ratios are for crude logistic regression analyses. Absolute SCFA concentrations were log-transformed prior to analysis. For relative concentrations, odds ratios pertain to change in odds of the outcome associated with a 10% increase in SCFA relative concentrations.

Figure 2.

Box plots summarize maternal fecal relative acetic acid for subjects with asthma/wheeze with sensitization (n=17) vs without asthma/wheeze with sensitization (n=63); and for subjects with asthma/wheeze with hay fever (n=33) vs without asthma/wheeze with hay fever (n=102).

As SCFA are products of microbial fermentation of dietary fiber, we investigated the maternal diet and fecal microbiome, hypothesizing that these factors correlate with fecal SCFA and offspring outcomes. Maternal dietary fiber intake was estimated from food frequency questionnaires from the 3^rd trimester. In adjusted linear regression analyses, fiber intake was positively associated with total SCFA (beta for a 1-unit increase in log-transformed dietary fiber nutrient density=0.26, 95% CI 0.02–0.44, p=0.035) and with absolute concentrations of acetic acid (beta=0.30, 95% CI 0.07–0.53, p=0.011) and propionic acid (beta=0.26, 95% CI 0.01–0.51, p=0.045) (Table E6). Fiber intake was positively but non-significantly associated with relative acetic acid (beta=0.03, 95% CI −0.0004–0.07, p=0.052). However, fiber intake during pregnancy was not associated with offspring outcomes (Table E7 in the Online Repository), and this may be because dietary fiber contributes to offspring disease risk only in combination with relevant intestinal microbes.

Bacterial 16S rRNA sequencing was performed on the same stool samples used for SCFA profiling in 114 of the 150 participating mothers. Shannon index, an alpha diversity metric, was not associated with SCFA (Table E8). Bray-Curtis dissimilarity, a beta diversity measure of taxonomic composition, was associated with relative and absolute concentrations of all three SCFA (all p < 0.05, Table E8), providing evidence that specific taxa contribute to production of SCFA. However, Bray-Curtis dissimilarity was not associated with offspring outcomes (Table E9).

In negative binomial regression analyses, four taxa were associated with relative concentrations of all three SCFA (FDR < 0.20, Table E10). Of these, Eubacterium dolichum and Streptococcus anginosus demonstrated directionally consistent associations with offspring asthma/wheeze with sensitization (log2fold change 1.7 and −1.2; p=0.002 and 0.045, respectively, Table E11, Figure E1). These species belong to genera known to produce SCFA², though given the broad array of species that produce SCFA, likely represent a small subset of those with relevance to offspring health. From these analyses, we concluded that specific taxa may impact relative acetic acid in the gut during pregnancy, which may in turn modify risk of atopic asthma/wheeze in offspring.

In summary, we found that relative concentration of the SCFA acetic acid during pregnancy is associated with reduced asthma/wheeze with allergic sensitization in offspring. A similar non-significant association was seen between relative acetic acid during pregnancy and asthma/wheeze with hay fever in offspring. Though prior literature suggests that maternal fecal SCFA may influence child outcomes via mechanisms including epigenetic modification, it is possible that maternal fecal SCFA also correlate with child diet and microbiome composition and the latter may have more causal significance. If findings are replicated, this study supports the use of microbiome-targeted therapies, potentially in combination with dietary fiber, to increase SCFA production as therapeutic strategies to prevent asthma and atopy in offspring.

This analysis is subject to limitations. Food frequency questionnaire data are subject to presumably non-differential measurement bias, though the significant correlations we saw between fiber intake and fecal SCFA provide some degree of validation for both measures. Both dietary intake and relevant elements of the gut microenvironment may have varied over pregnancy, and we assessed only one time point in the third trimester. Our sample size was limited as data were not available for all VDAART participants, and though our findings are biologically plausible, they require replication in an independent sample.

Clinical Implications:

Mothers with higher fecal acetic acid, compared to other short-chain fatty acids, during pregnancy were less likely to have offspring with atopic asthma. If replicated, findings support use of microbiome-targeted treatments during pregnancy to increase fecal acetic acid.