ecological regression

Latent toxoplasmosis is probably the most common protistan parasitic disease with many indirect negative impacts on human health. One of the important impacts is impaired psychomotor function leading to reduced driving efficiency in Toxoplasma - seropositive subjects. Numerous case-control studies have established a positive relation between the seroprevalence of Toxoplasma gondii (Nicolle et Manceaux, 1908) and probability of traffic accidents in study populations. The prevalence of toxoplasmosis varies between populations according to local geographical conditions, hygienic practices and kitchen habits. Similarly, we see a striking variation in the incidence of traffic accidents across countries. Hence, we compiled the largest ever data set on the seroprevalence of toxoplasmosis and tried to understand its role in traffic accident-related deaths and disabilities across 87 countries. Simple non - parametric analysis showed a positive and strong relation of T. gondii seroprevalence and traffic accident related disabilities. Further, we conducted multivariate analysis to control for confounding factors. After controlling for wealth, geographical latitude, health of population, length of roads and number of vehicles, the correlation disappeared. When the frequency of RhD negativity and its in - teraction with toxoplasmosis were included into the model, the effects of toxoplasmosis seemingly returned. However, the analysed data suffered from the problem of multicollinearity. When a proper method of analysis, ridge regression, was applied, the effects of toxoplasmosis prevalence and RhD negativity frequency disappeared again. The existence of a strong correlation between the prevalence of toxoplasmosis and health of population in particular countries, which was the probable cause of multicollinearity and possible reason for the negative result of the present study, suggests that ‘asymptomatic’ latent toxoplasmosis could have a large impact on public health