Objective The ermB gene is a key mediator of macrolide antibiotic resistance that may increase the risk of respiratory infections and affect clinical efficacy.This study aims to explore the effect modification of cold spells on the association between the antibiotic resistance gene ermB-originating from layer farms and carried by fine particulate matter (PM_2.5)-and the risk of hospitalization for lower respiratory infections (LRIs).

Methods Daily data on LRI-induced hospitalizations, concentrations of the antibiotic resistance gene ermB in PM_2.5, air pollutant levels, and meteorological variables were collected in Beijing, China, from 2016 to 2018.A cold spell was defined as a period with a daily mean temperature below the 5th percentile (-3.6℃) during the study period, lasting at least two consecutive days.Using a time-series model with an interaction term between ermB concentration and cold spells, quasi-Poisson regression was applied during the cold seasons (November to March of the following year) to estimate the effect modification of cold spells on the association between ermB in PM_2.5 and risk of LRI-induced hospitalizations, adjusting for daily mean temperature, relative humidity, air pollutants, and other variables.

Results During the cold seasons from 2016 to 2018, the daily M(P₂₅, P₇₅) value for LRIs in Beijing was 341(169, 389).The median concentration of ermB on cold-spell days was 0.005 0 copies/m³, which was significantly lower than that on non-cold-spell days (0.007 7 copies/m³; Z=4.97, P < 0.001).On non-cold-spell days, no significant association was observed between ermB concentration in PM_2.5 and the risk of LRI-induced hospitalizations.On cold-spell days, a single-day lag effect (lag1 d) showed that each interquartile range (0.003 2 copies/m³) increase in ermB concentration was associated with a 27% increase in the risk of LRI-induced hospitalizations (95% confidence intervalCI: 2%-57%).Cumulative exposure analysis over multiple days indicated that the hospitalization risk associated with ermB on cold-spell days increased progressively with longer accumulation periods, peaking at lag06 d, where the relative risk (RR) rose to 1.46(95%CI: 1.05-2.04).Both risk estimates differed significantly from those on non-cold-spell days (Z=2.19 and 2.04, respectively; P < 0.05).Subgroup analyses showed that on cold-spell days, ermB in PM_2.5 significantly increased hospitalization risk only among females and individuals aged ≥65 years.At lag1 d, RRs were 1.34(95%CI: 1.08-1.67) for females, 1.29(95%CI: 1.00-1.66) for those aged 65-74, and 1.27(95%CI: 1.03-1.55) for those aged ≥75.At lag06 d, RRs were 1.55(95%CI: 1.11-2.17) for females and 1.62(95%CI: 1.09-2.41) for those aged 65-74.

Conclusion Cold spells enhance the association between the antibiotic resistance gene ermB in PM_2.5 and the risk of LRI-induced hospitalizations, with a more pronounced effect among females and individuals aged ≥65 years.