Prevalence of colistin-resistant Escherichia coli in foods and food-producing animals through the food chain: A worldwide systematic review and meta-analysis

Abstract

The purpose of this systematic review and meta-analysis was to summarize the available scientific evidence on the prevalence of colistin-resistant Escherichia coli strains isolated from foods and food-producing animals, the mobile colistin-resistant genes involved, and the impact of the associated variables. A systematic review was carried out in databases according to selection criteria and search strategies established a priori. Random-effect meta-analysis models were fitted to estimate the prevalence of colistin-resistant Escherichia coli and to identify the factors associated with the outcome. In general, 4.79% (95% CI: 3.98%-5.76%) of the food and food-producing animal samples harbored colistin-resistant Escherichia coli (total number of colistin-resistant Escherichia coli/total number of samples), while 5.70% (95% confidence interval: 4.97%-6.52%) of the E. coli strains isolated from food and food-producing animal samples harbored colistin resistance (total number of colistin-resistant Escherichia coli/total number of Escherichia coli isolated samples). The prevalence of colistin-resistant Escherichia coli increased over time (P < 0.001). On the other hand, 65.30% (95% confidence interval: 57.77%-72.14%) of colistin resistance was mediated by the mobile colistin resistance-1 gene. The mobile colistin resistance-1 gene prevalence did not show increases over time (P = 0.640). According to the findings, other allelic variants (mobile colistin resistance 2-10 genes) seem to have less impact on prevalence. A higher prevalence of colistin resistance was estimated in developing countries (P < 0.001), especially in samples (feces and intestinal content, meat, and viscera) derived from poultry and pigs (P < 0.001). The mobile colistin resistance-1 gene showed a global distribution with a high prevalence in most of the regions analyzed (>50%). The prevalence of colistin-resistant Escherichia coli and the mobile colistin resistance-1 gene has a strong impact on the entire food chain. The high prevalence estimated in the retail market represents a potential risk for consumers' health. There is an urgent need to implement based-evidence risk management measures under the "One Health" approach to guarantee public health, food safety, and a sustainable future.

Keywords: Colistin resistance; Escherichia coli; Foods and food-producing animals; Meta-analysis; Prevalence; mcr gene.

Fig. 1

Flow diagram (according to PRISMA guides) of selected studies included in the meta‐analysis. Abbreviation: n: Number of reports included and analyzed.

Fig. 2

Bar graph: Subgroup analysis comparing the prevalence of colistin-resistant E. coli (a) and mcr-1 gene (b) in food and food‐producing animals across geographic regions. World map graph: Geographical distribution (by country) of scientific papers included. The intensity of the blue color is proportional to the number of studies detected in each country. In countries without color, no studies were detected or included in this meta-analysis. References: The horizontal bars extend to the pooled prevalence estimate in each geographic region. The horizontal lines indicate confidence intervals. Abbreviations: *Number of reports included and analyzed, **Pooled prevalence estimate, ***Confidence interval.

Fig. 3

Subgroup analysis comparing the prevalence of colistin-resistant E. coli (a) and mcr-1 gene (b) in food and food‐producing animals considering the site of sampling. References: The horizontal bars extend to the pooled prevalence estimate in each geographic region. The horizontal lines indicate confidence intervals. Abbreviations: *Number of reports included and analyzed, **Pooled prevalence estimate, ***Confidence interval.

Fig. 4

Subgroup analysis comparing the prevalence of colistin-resistant E. coli (a) and mcr-1 gene (b) in food and food‐producing animals according to the species/plant-based product analyzed. References: The horizontal bars extend to the pooled prevalence estimate in each geographic region. The horizontal lines indicate confidence intervals. Abbreviation: *Number of reports included and analyzed, **Pooled prevalence estimate, ***Confidence interval.

Fig. 5

Subgroup analysis comparing the prevalence of colistin-resistant E. coli (a) and mcr-1 gene (b) in food and food‐producing animals according to type of sample analyzed. References: The horizontal bars extend to the pooled prevalence estimate in each geographic region. The horizontal lines indicate confidence intervals. Abbreviation: *Number of reports included and analyzed, **Pooled prevalence estimate, ***Confidence interval.