Objectives MRSA carriage prevalence is as high as 60% in livestock farmers in the Netherlands. Human carriage was associated with carriage in animals. However, animal-to-human transmission and especially airborne exposure are poorly studied. We determined MRSA air levels and a possible dose-response relationship with nasal MRSA carriage in farmers.

Methods Human MRSA carriage and stable MRSA air levels were assessed in three independent populations of 38 assumed frontrunner pig farms (A), 50 random pig farms (B), and 49 veal calf farms (C). Farmers were identified as human participants spending on average ≥20 hours/week working on the farm. Per farm 1–6 electrostatic dust collectors (EDCs) were placed. Nasal swabs were taken from farmers and analysed for presence of MRSA by culturing and PCR. EDCs were analysed by qPCR, which resulted in an equivalent of the number of colony forming units (CFUeq) per EDC.

Results Mean MRSA air levels were lowest on veal farms and highest on pig farms. Generally, the average MRSA air levels corroborated with the MRSA nasal carriage prevalence in the three populations. Mean MRSA/EDC/farm for population A was 225 CFUeq (95%CI: 170–281); for population B it was 385 CFUeq (95%CI: 239–530), and for population C it was 59 CFUeq (95%CI:26–93). Nasal MRSA carriage prevalences were 53% for population A, 72% for population B, and 31% for population C. Working hours ranged from 20–80 hours/week, with means of 48, 46, and 39 hours for population A, B, and C, respectively.

In a multivariate pooled analysis, a strong association between MRSA carriage and the mean MRSA concentration (expressed as log (MRSA)/EDCs/farm) was shown (RR = 1.37; p = 0.0008), independent of the association with working hours (RR = 1.01; p = 0.03) and adjusted for age, sex and smoking.

Conclusions Results suggest an important role of MRSA transmission through air, which has not been established earlier.