[Identification of risk factors for infection with Toxoplasma gondii in Serbia as a basis of a program for prevention of congenital toxoplasmosis].

BACKGROUND

Toxoplasmosis has long been known as a major cause of perinatal morbidity. Acute infection in pregnancy may lead to fetal infection and subsequent fetal loss or birth of a manifestly or latently infected infant. However, it is a preventable disease. In Europe, significant variations have been shown to occur not only between countries but also within a given country indicating local variations in the influence of epidemiological factors contributing to infection. Thus, many European countries have implemented prevention programs in measure with the respective estimated risk of congenital toxoplasmosis. Since in view of its cost, a general screening-in-pregnancy program is at present not an option in Serbia & Montenegro, insight into the risk factors of particular local significance may therefore improve the quality of and the compliance with the hygienic and dietetic advice given to pregnant women as a preventive measure, as well as identify the particular subpopulations at an increased risk of infection, who may then be selectively screened.

SUBJECTS AND METHODS

A retrospective study of risk factors for Toxoplasma gondii infection based on serological and epidemiological data (questionnaire) was performed in a series of 2936 women aged 15-49 years from throughout Serbia tested in our laboratory between 1988 and 1997. Inclusion criteria included availability of serological and epidemiological data (as specified below). Specific anti-Toxoplasma antibodies were detected by the reference Sabin-Feldman dye test as modified by Desmonts into the lysis test. The questionnaire included questions on age (stratified into five-year groups), degree of education (modalities: grade school, secondary or university level), and community of residence (urban/suburban), as well as on life-style habits pertaining to infection transmission risk factors: consumption of undercooked meat, exposure to soil, and exposure to cats (pet cat ownership). In addition, the year of entry into the study was taken as a variable per se.

STATISTICS

The difference in the rates of infection between groups was evaluated by the chi-square test (chi 2). Logistic regression (univariant and multivariant approaches) was used to analyze the association between Toxoplasma infection and the specific demographic and epidemiological factors.

RESULTS

The overall prevalence of infection was 69%, but with very wide variations both over time (decreasing from 86% in 1988 to 39% in 1997) and region (40-90%). The risk of infection increased from the south to the north (RR = 0.97, CI = 0.94-0.98). Within Belgrade, the risk was higher in urban than in suburban zones (RR = 0.79, CI = 0.64-0.93). The single infection transmission factor that was shown to be a predictor of infection in the whole study group was undercooked meat consumption (RR = 1.6, CI = 1.2-2.1), while exposure to soil was a predictor only in women aged 15-19 (RR = 10.3, CI = 2.7-38.6). Owing pet cats had no influence. While T. gondii infection was not associated with pathological pregnancies (p = 0.51) in the whole study population, it was significantly associated with pathological pregnancies in women exposed to both undercooked meat consumption (p = 0.009) and in those aged 15-19 in contact with soil (p = 0.022), as well as in women residing in highly urban communities (p = 0.048).

CONCLUSION

The dramatic decrease in the prevalence of T. gondii infection over the nineties resulted in a rising proportion of women exposed to infection in pregnancy and consequently in an increased risk of congenital toxoplasmosis. Since the financial status of the country's health system does not allow for a general screening-in-pregnancy program, we propose, based on the above data, health education of all pregnant women in combination with serological testing of those exposed to predictors of infection as an epidemiologically sound and financially sustainable alternative.