Low prevalence of resistance genes in sheltered homeless population in Marseille, France, 2014–2018

Tran Duc Anh Ly,1,2 Linda Hadjadj,2,3 Van Thuan Hoang,1–2,4 Meriem Louni,1,2 Thi Loi Dao,1–2,5 Sekene Badiaga,2,6 Herve Tissot-Dupont,2,3 Didier Raoult,2,3 Jean-Marc Rolain,2,3 Philippe Gautret1,21IRD, AP-HM, SSA, VITROME, Aix Marseille Univ., Marseille, France; 2IHU-Méditerranée Infection, Marseille, France; 3MEPHI, Aix Marseille Univ., Marseille, France; 4Family Medicine Department, Thai Binh University of Medicine and Pharmacy, Thành Phố Thái Bình, Vietnam; 5Pneumology Department, Thai Binh University of Medicine and Pharmacy, Thành Phố Thái Bình, Vietnam; 6Emergency Department, North Hospital, AP-HM, Marseille, FranceObjectives: The present study has explored the prevalence and potential factors contributing to the presence of nasal/pharyngeal resistant genes in homeless people.Methods: During the winters 2014–2018, we enrolled sheltered homeless adults and controls and collected nasal/pharyngeal samples. Sixteen antibiotic resistance genes (ARGs), including genes encoding for beta-lactamases and colistin-resistance genes, were searched by real-time polymerase chain reaction (qPCR) performed directly on respiratory samples and followed by conventional PCR and sequencing.Results: Over a 5-year period, using qPCR, we identified in homeless group (n=715) the presence of blaTEM (396/710, 54.7%), blaSHV (27/708, 3.6%), blaOXA-23 (1/708, 0.1%), while other genes including colistin-resistance genes (mcr-1 to mcr-5) were absent. We found a significantly higher proportion of ARG carriage among controls (74.1%) compared to homeless population (57.1%), p=0.038. Tobacco smoking (OR=4.72, p<0.0001) and respiratory clinical signs (OR=4.03, p=0.002) were most prevalent in homeless people, while vaccination against influenza (OR=0.31, p=0.016) was lower compared to controls. Among homeless people, type of housing (shelter A versus B, OR=1.59, p=0.006) and smoking tobacco (smoker versus non-smoker, OR=0.55, p=0.001) were independent factors associated with ARG carriage. By sequencing, we obtained a high diversity of blaTEM and blaSHV in both populations.Conclusion: The lower risk for ARGs in the homeless population could be explained by limited access to health care and subsequently reduced exposure to antibiotics.Keywords: antibiotic resistance gene, homeless, real-time polymerase chain reaction (qPCR), potential risk factor