1) It could be a trivial type 2 error. In a good study, probability of such error (failure to detect existing effect) is about 20-25%. The authors probably did not perform power analysis, however, due to the usually low effect size of effects of toxoplasmosis on behavioral variables in "outbred" humans (toxoplasmosis usually explains less than 3% of total variability) I expect that the probability of false negative result of studies on similar number of Toxoplasma-infected subjects is higher than 50%. Whenever the negative results are being published (not just about toxoplasmosis), the power analysis should be rigorously done and its results should be presented in the same paper.
2) Most of studies showing the effect of toxoplasmosis on human behavior were performed on very young people, either Czech women in reproductive age, or the undergraduate students. Even here, the estimated frequency of false negative subjects (subjects that had been infected so long ago, that their concentrations of IgG decreased under the cut-off of particular serological test) among the controls is probably between 5-15%. In the present study the subjects are about 10 years older and the toxoplasmosis was diagnosed with just one serological test (IgG Immunosimplicity EIA kit), not two tests (IgG ELISA and CFR test). Therefore, the fraction of false negative subjects in the analysed population is probably higher. These false negative subjects are being infected for the longest period of time, therefore their behavior is changed the most. For analysis of such contaminated data, a special variant of permutation test [0] should be used.
3) The authors searched for the effect of toxoplasmosis on probability of having been diagnosed with schizophrenia and having been diagnosed with major depression. Only 24 schizophrenia patients were observed in the set and therefore difference between incidences 3.4% and 2.7% in Toxoplasma-infected and Toxoplasma-free subjects was not significant. The effects of toxoplasmosis on incidence of major depression has never been reported. Actually, our data even suggest that Toxoplasma-infected men have significantly lower, rather than higher incidence of major depression [1].
4) Similarly, the number of subjects who attempted suicide (16) and those with a non-suicidal injury (29) is too low to make even the large observed differences in incidences significant.
5) The authors searched for the effect of toxoplasmosis on human psychomotor performance. However, most of the published data suggest that the effect is most robust for simple reaction time test, while the results of more complex psychomotor performance tests are far less unambiguous. Authors did not test their probands using the simple reaction time test. Moreover, recent data showed that RhD-positive subjects are rather resistant to negative effects of toxoplasmosis. Therefore, this factor (as well as toxoplasmosis-sex interaction, see below) should be always included into the statistical models.
6) Authors searched for the effect of toxoplasmosis on criminal and driving conviction. I know just one paper showing that such effect of toxoplasmosis exists. The higher risk of traffic accidents which have been reported for Toxoplasma-infected (especially RhD-negative) subjects is most probably caused by impaired reaction times, not by their higher criminality.
7) Authors found no significant effect of toxoplasmosis on personality profile (Big Five traits) of humans. One study performed on very homogeneous sample of university students showed that two of five Big Five factors should differ between Toxoplasma-infected and Toxoplasma-free subjects (most of the previous studies have been done with Cattell's and Cloninger's questionnaiures. However, they did not show (or possibly even tested) the effect of toxoplasmosis-sex interaction on the personality. It is known that personality traits of men and women are often affected by toxoplasmosis in an opposite way [2]. Therefore, it is always necessary to search not only for main effect of toxoplasmosis and sex, but also for their interaction. Again, RhD negative subjects are mostly responsible for the observed effect of toxoplasmosis on human personality. It is not clear what fraction of the New Zealand cohort are RhD negative subjects.
8) Authors studied the effects of toxoplasmosis on human intelligence and other cognitive functions. However, such effects of toxoplasmosis have been reported to exist in just few studies and sometimes even the opposite effects of toxoplasmosis on the results of cognitive tests (caused, probably, by higher competiveness and cooperativeness of Toxoplasma-infected men and women, respectively) have been observed.
9) Authors did not perform (or at least did not report the result of) any aggregate test, such as exact binomial test. When I quickly checked the results presented in the tables 2 and 3 of the paper, I have found that the observed shifts of studied parameters were in the correct (the a priori expected) direction 17 times out of 22 comparisons. The binomial test provided the result 0.006. Therefore, also the present study confirms that toxoplasmosis has highly significant effect on human mental functions.

Jaroslav Flegr, Fac. Science, Charles Univ., Prague, Czech Republic


[0] Flegr J., Havl�cek J. 1999: Changes in personality profile of young women with latent toxoplasmosis. Folia Parasitologica, 46: 22-28.
[1] Flegr, J. 2015: Neurological and neuropsychiatric consequences of chronic Toxoplasma infection. Clinical Microbiology Reports, 2(4).
[2] Flegr, J. 2013: Influence of latent Toxoplasma infection on the human personality, physiology, and morphology: The Toxoplasma-human model in studying the manipulation hypothesis - pros and cons. Journal of Experimental Biology, 216, 127-133.