One Health approach probes zoonotic non-typhoidal Salmonella infections in China: A systematic review and meta-analysis

Abstract

Background: Zoonotic infections, particularly those caused by non-typhoidal Salmonella (NTS), pose a significant disease burden. However, there is a notable lack of comprehensive and integrated studies employing the One Health approach to address Salmonella prevalence. In this study, we aimed to analyse NTS spatiotemporal prevalence, serovar distribution, and antimicrobial resistance (AMR) across China.

Methods: We conducted a systematic review and meta-analysis to understand the dynamics of NTS in a One Health context in China. We searched the CNKI, Wanfang, and PubMed databases for Chinese and English peer-reviewed articles published before 1 January 2022 dealing with Salmonella in the context of China. We examined the dynamic prevalence along the food chain, the risk of dominant serovars and the carriers' regional contribution by principal component analysis, and the AMR burden before and after the ban on using antimicrobials as feed additives across five decades. We used the inverse variance index as an indicator of the inconsistency across studies, and we adopted the restricted maximum likelihood model due to high heterogeneity for analysis with a 95% confidence level for the pooled prevalence estimate.

Results: Based on 562 retrieved high-quality studies during 1967-2021, representing 5 052 496 samples overall and 80 536 positive samples for NTS isolates, the overall average prevalence was 7.35% (95% confidence interval (CI) = 0.069-0.087), which was regionally relatively higher in northern China (8.19%; 95% CI = 0.078-0.117) than in southern China (6.94%; 95% CI = 0.067-0.088). Poultry was the primary vehicle for serovars Enteritidis and Indiana (especially in the north), while swine and ruminants for Typhimurium and Derby were the first to highlight the regional livestock contribution to serovar prevalence. The overall AMR rate was 73.63% (95% CI = 0.68-0.99), decreasing after the ban on excessive use of feed-based antibiotics in livestock since 2020, with a relatively low resistance towards front-line and last-resort drugs.

Conclusions: Our study emphasises the importance of adopting a One Health framework to better understand the zoonotic nature of human NTS and highlights the dominant serovars on food contamination and human infection. The similarity in AMR patterns between poultry and human isolates further emphasises the integrated approach for evaluating disease burden and implementing targeted interventions.

Figure 1

Summary of study processes and data. The black bars indicate the sample size on the left axis and the red lines indicate prevalence on the right axis. Prevalence was calculated per the formula prevalence (%) = N (positive sample)/(N (total sample size) × 100%). Panel A. Flow diagram of the literature search. Panel B. The publication year of selected studies. Panel C. Percentage of methods in articles used for bacterial isolate identification. Panel D. Typing method. Panel E. Method for Antimicrobial susceptibility test. AIS – automatic identification systems, PCR – polymerase chain reaction, PFGE – pulsed-field gel electrophoresis.

Figure 2

Spatiotemporal prevalence from different sources. Panel A. Prevalence of NTS in food animals (FA), food (F), and humans (H) across regions. Panel B. Prevalence of NTS in food animals (FA), food (F), and humans (H) across provinces. Panel C. Temporal changes in overall sample size and prevalence. C – China, N – northern China, NTS – non-typhoidal salmonella, S – southern China.

Figure 3

The dynamic pattern of the top fifteen serovars from 1953 to 2021. Total strands for the number of isolates serotyped were calculated per the formula proportion (%) = N (strains of serovars X)/(N(total of strains serotyped) × 100%). Panel A. Top fifteen serovars and proportion. Panel B. The proportion of the top fifteen serovars in poultry, swine, and ruminants, collectively referred to as food animals, food, and humans. Panel C. Dynamics of the top fifteen serovars from 1953 to 2021 from all origins. Panels D–I. Dynamics of the top fifteen serovars from 1953 to 2021 from different hosts. F – food, FA – food animals, H – human, P – poultry, R – ruminants, S – swine.

Figure 4

The contribution to the prevalence of the top fifteen serovars among various host groups in northern and southern China using PCA analysis. Numerics 1–15 stand for the top fifteen serovars in order: 1 – Typhimurium, 2 – Enteritidis, 3 – Derby, 4 – Agona, 5 – Indiana, 6 – Anatum, 7 – Senftenberg, 8 – Stanley, 9 – London, 10 – Thompson, 11 – Rissen, 12 – Newport, 13 – Choleraesuis, 14 – Corvallis, and 15 – Kentucky. F – food, H – human, N – northern China, P – poultry, S – southern China, S – swine.

Figure 5

Antimicrobial resistance rate in different hosts and periods. The dark grey wells with a cross indicate that data are not available in this point. Panel A. Antimicrobial resistance rate of eleven antimicrobial classes in poultry, swine, ruminant, food animal, food, humans, and all sources. Panels B–M. Dynamics of antimicrobial resistance rates from different origins. ALL – all sources, ARR – antimicrobial-resistant rate, F – food, FA – food animals, H – human, P – poultry, R – ruminants, S – swine.