Systematic Review and Modelling of Age-Dependent Prevalence of Toxoplasma gondii in Livestock, Wildlife and Felids in Europe

Abstract

Toxoplasma gondii is a zoonotic parasite of importance to both human and animal health. The parasite has various transmission routes, and the meat of infected animals appears to be a major source of human infections in Europe. We aimed to estimate T. gondii prevalence in a selection of animal host species. A systematic literature review resulting in 226 eligible publications was carried out, and serological data were analyzed using an age-dependent Bayesian hierarchical model to obtain estimates for the regional T. gondii seroprevalence in livestock, wildlife, and felids. Prevalence estimates varied between species, regions, indoor/outdoor rearing, and types of detection methods applied. The lowest estimated seroprevalence was observed for indoor-kept lagomorphs at 4.8% (95% CI: 1.8-7.5%) and the highest for outdoor-kept sheep at 63.3% (95% CI: 53.0-79.3%). Overall, T. gondii seroprevalence estimates were highest within Eastern Europe, whilst being lowest in Northern Europe. Prevalence data based on direct detection methods were scarce and were not modelled but rather directly summarized by species. The outcomes of the meta-analysis can be used to extrapolate data to areas with a lack of data and provide valuable inputs for future source attribution approaches aiming to estimate the relative contribution of different sources of T. gondii human infection.

Keywords: Bayesian model; cat; cattle; goat; meta-analysis; pig; seroprevalence; sheep; systematic review; toxoplasmosis.

Figure 1

Categorization of Europe into regions. European countries are categorized into five European regions: Western (blue), Northern (white), Eastern (green), Southeastern (yellow), and Southwestern (red). Generated from: https://www.mapchart.net/index.html, accessed on 18 August 2022.

Figure 2

Direct detection estimates. Average weighted prevalence estimates with 95% CI per region based on direct detection and separated by sample types of blood, feces, muscle, and other types (muscle and organ tissues, milk, pleural fluids, cardiac fluids, and, in the case of felids, feces). Only animal species that had data available from at least one of the five regions are included.

Figure 3

Posterior probabilities of the model parameters. Values for the force of infection in regions, in species, by sample type and outdoor access are all exponentiated, which means that a value of one indicates the absence of an effect on the baseline probability (see Supplementary Data—Figure S4). To reconstruct the total force of infection, the exponentiated contributions must be multiplied. Gray area represents the uncertainty distribution. In all panels, the thin and thick black lines indicate 95% and 50% Bayesian credible intervals, respectively, with dots indicating the mean of exponentiated forces of infection dependent on region, species, sample type (other = pleural or cardiac fluids), and holding status as multiplicative corrections to the baseline (see Equation (6)).

Figure 4

SIS model fit for age-dependent seroprevalence of T. gondii in pigs by outdoor access in five regions of Europe. The lines indicate the fitted seroprevalence by age, based on detection in the blood (red), muscle fluid (green), or other matrices (blue). The grey dots represent seroprevalence data points at the best estimate of age in the data for the studied populations. The size of the dots indicates number of animals, in categories from 10,000 to 50,000. The dots are shifted horizontally along the grey line extending from the minimum to maximum possible age, to the best fitting age (red dots) (see Supplementary Files—Figure S15).