Surrounding Residential Greenness does not Confound Associations between Pet Ownership and Endotoxins in Germany

Abstract

Background/Aim: Pet ownership is consistently associated with higher indoor endotoxin concentrations. We examined whether surrounding residential greenness may confound or modify this association in homes in Munich, Germany, because pet ownership might be related to neighbourhood greenness. Methods: Pet ownership data were collected during participant recruitment in the LISA cohort. Endotoxin concentrations were measured in settled house dust sampled from mothers’ mattresses and living room floors (N=1199). Residential greenness was defined as the mean Normalized Difference Vegetation Index in a 500m buffer around the home. Linear regression models assessed cross-sectional associations between log-transformed endotoxin concentrations and pet ownership, adjusted for known predictors of endotoxin concentrations. Confounding was assessed by additionally adjusting the models for residential greenness. Effect modification was assessed by including interaction terms between pet ownership and residential greenness, and stratifying the models by residential greenness tertiles. Results: In adjusted models, dog ownership (5.3%) was associated with higher endotoxin concentrations in mattresses (means ratio: 1.67 [95% confidence intervals: 1.17, 2.39]) and on floors (2.32 [1.60, 3.35]). Cat ownership (9.2%) was only associated with mattress concentrations (1.49 [1.14, 1.97]). No associations were found with the other pets considered (bunny, bird). Associations were robust to further adjustment for greenness. Associations were slightly stronger in the lower greenness tertile for endotoxin concentrations in mattresses (low: 2.10 [1.01, 4.35], medium: 1.55 [0.86, 2.77] and high: 1.43 [0.79, 2.59] for dogs, and low: 1.93 [1.13, 3.29], medium: 1.40 [0.92, 2.14] and high 1.48 [0.89, 2.44] for cats) but not on floors. Interaction terms were not significant. Conclusions: Residential greenness did not confound the associations between pet ownership and indoor endotoxin concentrations, and there was only limited evidence of effect modification. Studies should continue investigating how pets influence indoor endotoxin concentrations, either by their presence and/or by acting as transmission vectors of the outdoor environment.