Toxoplasma gondii, a parasitic protozoan related to Plasmodium, infects about 30% of the human population worldwide. Latent toxoplasmosis, characterized by the life-long presence of cysts of the parasite in different host tissues, including the nervous system, and by the presence of anamnestic Toxoplasma IgG antibodies in the serum, was long considered asymptomatic. In the past 20 years, several effects of this form of parasitosis on the human organism were described in the literature. For example, latent toxoplasmosis increases the risk of schizophrenia [1] and Parkinson's disease [2] influences human personality and behavior [3, 4], impairs psychomotor performance, enhances the risk of suicide [5], of traffic accident [6–9] and increases probability of the birth of male offspring [10, 11]. Reportedly, the activity of the immune system is likely to play an important role in many of the observed effects of Toxoplasma infection. For example, the impairment of the immune system has been suggested to be at least partly responsible for the observed association between toxoplasmosis and schizophrenia [3]. Also, many of the observed behavioural effects of toxoplasmosis might be a result of the increased level of dopamine in the brain tissue in response to IL-2 produced by immune cells in the sites of local inflammation in the infected brain [12–14]. Similarly, the effect of latent toxoplasmosis on human reproduction, not only on the probability of the birth of male offspring, but also on the probability of the birth of a child with Down syndrome [15] and on the length of pregnancy [16], has been assumed to be a consequence of toxoplasmosis-associated immunosuppression. It is well known that most of the embryos, especially the more immunogenic male embryos and those with various chromosomal aberrations and physical malformations, are aborted in early phases of pregnancy[17, 18]. The immunosuppression hypothesis suggests that Toxoplasma relaxes the stringency of some mechanisms of quality control of early embryos to increase the probability of its transmission to the next generation through the congenitally infected offspring [10, 11].

Many reports are available concerning the effect of acute toxoplasmosis on the immunity of humans or mice [19–21]. However, the data showing similar effects in congenital toxoplasmosis are absent. The results obtained in infected laboratory female mice showed that mice in the early phase of latent infection exhibited temporarily increased production of IL-12 and decreased production of IL-10. In accordance with the immunosuppression hypothesis, the mice showed decreased production of IL-2 and nitric oxide and decreased synthesis of DNA in the mixed lymphocyte assay in the early and also in the late phases of latent toxoplasmosis [22]. It is difficult to study such effects in mouse models, as the duration of acute and post-acute stages of infection approaches the normal length of life in this species. On the other hand, there is a striking difference between the about one-month acute infection and life-long latent infection in humans. In the present cohort study, we searched for indices of immunomodulation in humans with latent toxoplasmosis by comparing the available clinical records, namely the flow cytometry and haematological data, in Toxoplasma-positive (infected) and Toxoplasma-negative (noninfected) outpatients undergoing routine immunological tests at the Institute of Clinical and Experimental Medicine in Prague.

The studied population consisted of 440 patients, 312 women and 128 men. The mean age of 238 Toxoplasma-negative women (40.22, SD = 17.21) was significantly lower than that of 74 (23.7%) Toxoplasma-positive women (49.11, SD = 13.72), P < 0.0001. Similarly, 114 Toxoplasma-negative men (36.04, SD = 18.28) were significantly younger than 14 (10.9%) Toxoplasma-positive men (50.14, SD = 19.32), P < 0.008. The logistic regression with the dependent variable toxoplasmosis and independent variables sex and age showed that men had a significantly lower probability of Toxoplasma infection than women (P = 0.009, OR = 0.43, CI₉₅ = 0.23-0.81) and the probability of being Toxoplasma positive increased with each year of age (P < 0.0001, OR = 1.035, CI₉₅ = 1.02-1.05). No effect of the diagnosis on probability of the Toxoplasma infection was observed when the binary variable diagnosis (immunodeficiency yes/no) was included into the model (P = 0.779, OR = 0.920, CI₉₅ = 0.51-1.65).

Although the length of infection in particular Toxoplasma-positive patients is unknown, it can be estimated from the level of Toxoplasma antibodies measured by CFT. Unlike the irregularly fluctuating concentrations of antibodies measured by ELISA, the CFT titres have been reported to decrease slowly and relatively regularly in years after the infection [23]. Results of the partial Kendall regression analysis with age as a covariate showed that leukocytes, CD19, CD16 + 56 and monocytes had a non-significant tendency to return to normal with decreasing CFT titres of Toxoplasma antibodies, for monocytes in women this tendency was significant (partial Kendall Tau = 0.170, P = 0.028), see Figure 1.

The prevalence of latent toxoplasmosis was significantly higher in female (23.7%) than in male (10.9%) patients. Patients with or without anamnestic titres of Toxoplasma antibodies, i.e. with or without latent toxoplasmosis, differed from each others in several important parameters, namely in the counts of leukocytes, CD19 cells (B-cells), CD16 + 36 cells (natural killers) and monocytes. The number of B cells was lower in both infected men and women; however, all the other immune cell counts were higher in infected women and lower in infected men. Such shifts in the counts of particular immune cells to the opposite directions are likely to be side effects of the infection rather than a result of the biological adaptation of the parasite aimed to increase the probability of survival in the host organism by modulating its immunoreactivity.

The opposite effects of toxoplasmosis in men and women were also observed in the personality profile [3, 24] and behaviour [25, 26]. The key to such opposite behavioral and immunological effects in infected men and women might be the variation in the effects of Toxoplasma infection on the concentration of free testosterone. Reportedly, infected men had increased while infected women decreased testosterone levels in comparison with controls [27, 28]. Testosterone is known to have specific effects on the behaviour as well as the immunity of animals, including humans. The increased level of testosterone is generally associated with the immunosuppression [29, 30] which could explain both the decreased number of leukocytes, B-cells, NK-cells and monocytes in men and the increased size of leukocyte, NK-cell and monocyte populations in women. It must be remind, however, that experiments performed with artificially infected laboratory mice demonstrated decreased level of testosterone in both males and females. Moreover, we did not found any significant differences in the concentration of testosterone between Toxoplasma-positive and Toxoplasma-negative solders (nonpublished results). It is critically needed to search for changes of cell counts in the Toxoplasma-infected mice and also to search for possible correlation between concentration of testosterone and blood cell count in our patients.

The trend towards a slow return of the immune cell counts to normal suggests that the observed differences are a carry-out effect of acute infection rather than a consequence of latent toxoplasmosis. It must be remind that this trend was significant only for monocytes in women. For leukocytes, CD19 and CD16 + 56 cells, and monocytes in men the cell counts returned to normal value in subjects with oldest infection, however, the trend was nonsignificant due to large variance in cell counts and low number of infected subjects. The correlation between cell counts and duration of the infection also indicates that toxoplasmosis induces changes in the immune system rather than that a change in the immune system increases the risk of Toxoplasma infection. The return of the immune cell counts to normal is in accordance with the diminishing effect of latent toxoplasmosis on the sex ratio in women and female mice. The increased probability of the birth of male offspring was only observed in women with high or moderate levels of Toxoplasma antibodies and in female mice 2-4 months after the infection. In women with low or moderate levels of Toxoplasma antibodies and in female mice more than 4 months after the infection, the probability of the birth of male offspring was even lower than in uninfected controls. It was suggested that the increased sex ratio in women and mice in early phases of latent toxoplasmosis was caused by immunosuppression that protects the more immunogenic males embryos against abortion. For example, in humans, the sex ratio decreases from about 2.6 in the first weeks of pregnancy to about 1.06 at the time of delivery [17, 18]. The diminishing effect of toxoplasmosis on immune cell counts is in accordance with the expected decrease of immunosuppression.

The most obvious problem of our study was related to low number of Toxoplasma-positive men in our population. The unbalanced design can result in false negative result - i.e. in missing some effects of toxoplasmosis. It must be stressed, however, that it cannot lead in false positive results, i.e. in detection of non-existent effects of toxoplasmosis.

The main limitation of the present study was the absence of the data on the immunoreactivity of immune cells subpopulations. We searched for indices of immunosuppression using only the clinical data that are routinely collected from patients with allergic diseases. Recent results obtained in infected laboratory female mice suggest that the concentrations of important interleukins are increased; however, the reactivity of immune cells measured by nitric oxide production and proliferation of stimulated spleen cells in the MLC assay are considerably decreased in comparison with uninfected controls [22]. Further studies are needed to search for indices of immunosuppression using more specific markers. The second potential limitation of the present study is the selected population group. The patients with immunodeficiences or with allergies are a rather specific cohort and their immunological data can hardly be generalized to the normal healthy population. Theoretically, subjects with toxoplasmosis-associated immunosuppression might be protected against severe allergies, thus escaping inclusion in our study. The prevalence of latent toxoplasmosis is significantly lower among our male than female patients, although in the Czech general population, the prevalence rates of toxoplasmosis are approximately the same in men and women [31]. Possibly, the Toxoplasma-positive men (with decreased counts of leukocytes, B-cells, NK cells and monocytes) are protected against allergies and therefore were not enrolled in our study, while the women (with increased counts of total leukocytes, NK-cells and monocytes) have higher probability of allergies and immunopathological diseases. Decreased frequency of Toxoplasma-positive subjects in patients with atopic disease has already reported in several studies [32–34]. It must be remind, however, the low frequency of Toxoplasma infected men as well as the statistically significant associations between the Toxoplasma infection and the cell counts were observed mostly in subjects with immunodeficiencies (D.89.9) not in patients with allergies (J.30.1-4) in our study.

We can conclude that the data from the outpatient immunology clinic suggest specific effects of latent toxoplasmosis on humans. The effects decreased with the length of infection and were opposite in direction in men and women, possibly as a result of the reported opposite effects of the infection on the level of testosterone in men and women. Given the high prevalence of latent toxoplasmosis, its possible immunosuppressive effects, although relatively weak in individual patients, might have a considerable impact on the health of the world population.