A new approach for the analysis of seroprevalence data: a mathematical analysis of the seroprevalence of Toxoplasma gondii infection in Hyogo prefecture (Japan) with an implication of periodic outbreaks among young children.

Abstract

We propose a new approach for the analysis of seroprecvalence data. The seroprevalence date on Toxoplasma gondii infection published in 1997 at Kobe University Hospital were analyzed, with a simple mathe- matical model, Y = exp(-λ t), where Y , λ and t represent the percentage of seronegative people, annual infection rate and age, respectively. After calculating the mean annual infection rate (MAIR) on the basis of the above data, we determined the relationship between MAIR and the year of birth of the female participants in the study. Our present study indicates that MAIR for women born between 1940 and 1960 decreased over the years, and that in- fection rates correlated with the year of birth. Moreover, assuming that the historically declining trends of MAIR mainly reflect infection rate changes in childhood, we created a simulation of MAIR for the age under 5 years. This simulation demonstrated that MAIR for the age under 5 years decreased from around 4% for females born in 1940, to about 1% for those born in 1960. For women born after 1960, it implied that infection rates might have been cyclically fluctuating from 0% to 1%, with an approximate 10-year interval. Our analyses imply periodic out- breaks of T. gondii infection among young children in Hyogo prefecture. In Japan, it is difficult to make a sophis- ticated statistical analysis of seroprevalence of T. gondii mainly due to the lack of available data. Despite simplic- ity of our new approach, we believe the approach will be useful to grasp the current and the past situations of T.