Skip to main content
Fig. 1 | Parasites & Vectors

Fig. 1

From: Closing the gap on causal processes of infection risk from cross-sectional data: structural equation models to understand infection and co-infection

Fig. 1

Optimal SEMs showing host phenotype as a latent variable predicting disease co-infection in domestic cats: (a) all sites combined, (b) Australia, (c) USA, (c) Singapore. Boxes and ellipses represent measured and latent variables respectively. Coefficient of determination (r2) given where variables are predicted. P-values given for measured variables which are predicted, except for FcaGHV1 (abbreviated here to GHV for editorial purposes) which was set as the fixed variable (i.e., a direct response of host phenotype). Solid lines represent directional relationships, with non-standardised coefficients on each line. Dotted lines represent covariance relationships among pathogens (pathogen-facilitation), with their non-standardised coefficients. Line thicknesses are proportional to the strength of coefficient or covariance. In all cases, variables are binomial (0 or 1), with 1 equal to male, adult, or positive pathogen status. Since coefficients are non-standardised, care must be taken not to make direct comparisons of their effect sizes. Abbreviations: Bart: Bartonella spp;GHV: FcaGHV1, feline gammaherpesvirus 1; FeLV: feline leukaemia virus; FIV: feline immunodeficiency virus; Mhf: Mycoplasma haemofelis; Mhm: Candidatus Mycoplasma haemominutum

Back to article page