UCSF

Introduction

Microbiome differences have been found in adults who smoke cigarettes compared to non-smoking adults, but the impact of thirdhand smoke (THS; post-combustion tobacco residue) on hospitalized infants' rapidly developing gut microbiomes is unexplored. Our aim was to explore gut microbiome differences in infants admitted to a neonatal ICU (NICU) with varying THS-related exposure.

Methods

Forty-three mother-infant dyads (household member[s] smoke cigarettes, n = 32; no household smoking, n = 11) consented to a carbon monoxide-breath sample, bedside furniture nicotine wipes, infant-urine samples (for cotinine [nicotine's primary metabolite] assays), and stool collection (for 16S rRNA V4 gene sequencing). Negative binomial regression modeled relative abundances of 8 bacterial genera with THS exposure-related variables (i.e., household cigarette use, surface nicotine, and infant urine cotinine), controlling for gestational age, postnatal age, antibiotic use, and breastmilk feeding. Microbiome-diversity outcomes were modeled similarly. Bayesian posterior probabilities (PP) ≥75.0% were considered meaningful.

Results

A majority of infants (78%) were born pre-term. Infants from non-smoking homes and/or with lower NICU-furniture surface nicotine had greater microbiome alpha-diversity compared to infants from smoking households (PP ≥ 75.0%). Associations (with PP ≥ 75.0%) of selected bacterial genera with urine cotinine, surface nicotine, and/or household cigarette use were evidenced for 7 (of 8) modeled genera. For example, lower Bifidobacterium relative abundance associated with greater furniture nicotine (IRR<0.01 [<0.01, 64.02]; PP = 87.1%), urine cotinine (IRR = 0.08 [<0.01,2.84]; PP = 86.9%), and household smoking (IRR<0.01 [<0.01, 7.38]; PP = 96.0%; FDR p < 0.05).

Conclusions

THS-related exposure was associated with microbiome differences in NICU-admitted infants. Additional research on effects of tobacco-related exposures on healthy infant gut-microbiome development is warranted.