A combined case-control and molecular source attribution study of human Campylobacter infections in Germany, 2011-2014

Abstract

Campylobacter infection is the most commonly notified bacterial enteritis in Germany. We performed a large combined case-control and source attribution study (Nov 2011-Feb 2014) to identify risk factors for sporadic intestinal Campylobacter infections and to determine the relative importance of various animal sources for human infections in Germany. We conducted multivariable logistic regression analysis to identify risk factors. Source attribution analysis was performed using the asymmetric island model based on MLST data of human and animal/food isolates. As animal sources we considered chicken, pig, pet dog or cat, cattle, and poultry other than chicken. Consumption of chicken meat and eating out were the most important risk factors for Campylobacter infections. Additional risk factors were preparation of poultry meat in the household; preparation of uncooked food and raw meat at the same time; contact with poultry animals; and the use of gastric acid inhibitors. The mean probability of human C. jejuni isolates to originate from chickens was highest (74%), whereas pigs were a negligible source for C. jejuni infections. Human C. coli isolates were likely to originate from chickens (56%) or from pigs (32%). Efforts need to be intensified along the food chain to reduce Campylobacter load, especially on chicken meat.

Figure 1

Minimal Spanning Tree generated from MLST comparisons of 816 C. jejuni and C. coli isolates from human study participants and from animals and food samples from the study region. Colouring according to isolate source: blue, isolates from patients; green, isolates from animals; red, isolates from food samples. Only abundant sequence types (STs) are labelled. A version of the figure with full labelling is available as Supplementary Figure S1.

Figure 2

Minimal Spanning Tree generated from MLST comparisons of 816 C. jejuni and C. coli isolates from human study participants and from animals and food samples from the study region. Colouring according to host species as specified in colour legend. Only abundant sequence types (STs) are labelled. A version of the figure with full labelling is available as Supplementary Figure S2.

Figure 3

Source probabilities for human isolates (n = 613) to originate from each of the five sources (chicken, pet, pig, cattle, poultry other than chicken) as determined by source attribution analysis. MLST data from animal and food isolates obtained in Germany and in neighbouring European countries was used for source attribution analysis (Table 1).