VII. How Toxoplasma slows a person down, and throws him under the wheels of cars
Along with the study on the effect of Toxoplasma on human personality and behavior, I tried to observe whether Toxoplasma has an effect on human performance, particularly reaction time (how quickly one can react to simple stimuli). I expected that since Toxoplasma lengthens reaction time in mice, thus increasing their risk of being captured by a cat, it could similarly influence a human.
The very first project didn’t work out, but I often joke about it. A high school peer of mine, who worked as a doctor in a Prague hospital, agreed to supply me with sera from patients who had suffered a car accident, as well as from patients hospitalized for different reasons. These sera I had examined for toxoplasmosis. Since blood testing is carried for almost all patients, I obtained samples from the blood routinely withdrawn – I needed at most half a milliliter per sample of blood for my studies. I predicted that among victims of various accidents there would be a higher percentage of Toxo positives, because these would be less coordinated, more likely to be run over by a car or to crash a car as a driver. As I mentioned, this project didn’t work out. The sera samples accumulated slowly, and it wasn’t always clear how the patient came to the accident – whether he wasn’t, perhaps, just the unfortunate passenger who had nothing to do with the crash. Mainly I wasn’t sure whether the control sample of people who did not suffer an accident came from the same region as the sample of people who did. For this reason, after about half a year I ended this project.
Why do I say the results were amusing? They’d likely appeal to any black humor enthusiast. I received the sera anonymously, marked by only number codes. But one time, when I arrived in the hospital for the samples, my doctor friend told me that I might know one of the patients and that maybe I should stop by him. Which I did only to find laying on one of the hospital beds my colleague Petr Kodym, who two days before had nearly cut his leg off with a chain saw he’d enlisted in an attempt to cut limbs up in a tree (this is a likely a new, hitherto little used method in pomology, which he evidently picked up in one of his frequent expeditions to the Balkans). For many years Kodym had been examining all the sera samples for toxoplasmosis, and both of us knew he, just as I, was Toxo positive. At the time I believe he was already head of the National Reference Laboratory for Toxoplasmosis, and he’d always been pretty skeptical of my studies. I think that in this particular case I showed him pretty clearly, that there was probably something to my theories.
This unsuccessful study was basically an analogy of the predation experiment, in which manipulation theory is proved on animals. The predation experiment observes a parasite distribution in the hosts which became the prey of predators (the final hosts of the given parasite), and this occurrence is compared with the parasite’s total prevalence in the host species in the studied locality. These studies were carried out, for example, on Frenkelia, a parasite related to Toxoplasma. Its intermediate hosts are rodents such as common voles (Microtus arvalis) or bank voles. The parasite forms typical round cysts in the rodent’s brain, and its final hosts are birds of prey. So one can collect prey (common and bank voles) from the nests of predatory birds, determine what percentage of prey individuals are infected by the parasite, and finally set out traps for voles in the same area to determine how many of them are infected (Fig. 21). By comparing the frequency of infection of the prey and that of the animals in the traps, one can calculate whether the parasite increases the probability of being captured by the predator.
The predation experiment was carried out by my former undergraduate student Jan Votýpka (28). The results of this very carefully conducted study showed that infection by Frenkelia significantly raises a rodent’s risk of capture by a buzzard. Votýpka and his colleagues supplemented the study with laboratory experiments – the predation experiment isn’t really an experiment, but an observational study (see Box 38 Experiments vs observational studies). In observational studies it’s usually impossible to distinguish whether infected rodents have an increased risk of capture by a predator, or whether they have a decreased risk of us capturing them in a trap (but sometimes it is possible, see Box 72 You can’t expect miracles, even from a prospective cohort study).
Fig. 21 A field predation experiment. The investigators placed wooden frames with nets over the nests of predatory birds (in this case, nests of the common buzzard, Buteo buteo). Parents bringing food to their nestlings have to put it on the net, where it was regularly collected by investigators and analyzed for the presence of parasites. In the same location, the prevalence of parasites in trapped rodents was analyzed. The prevalence of parasites in rodents caught by birds of prey was compared to the prevalence in rodents trapped by the investigators to determine whether infected individuals are more likely to be the prey of the definitive host – the bird of prey. I can assure my soft-hearted readers that the nestlings didn’t starve during the study – the investigators returned all analyzed rodents to the nests, so the birds were successfully raised. Photo by Petr Voříšek.
Votýpka and his coworkers infected 21 mice with the coccidian Sarcocystis dispersa, and when they stopped exhibiting signs of the disease’s acute phase, he shut them along with 21 uninfected mice into a room with lots of hiding place as well as a predator – a long-eared owl Asio otus (affectionately christened
Box 38 Experiments vs. observational studies
Empirical studies are primarily classified as experiments and observational studies. Both types of studies have their advantages and disadvantages. In an experiment, the researcher should have everything controlled; he himself decides which individuals are subject to the observed factor and which serve as the control (see also Box 15 Popular mistakes when making a control group). If he discovers that the individuals of the control group behave differently than do those of the experimental group, then he can conclude that this difference was caused by the observed factor. The disadvantage of experiments is the risk of experimental artifacts. In a laboratory (or in an experimental field), individuals are subjected to fairly different conditions than those freely in nature. For example, the natural infection method and dose of a parasite are often completely different from those a researcher will use in the lab. Therefore in the lab we can sometimes observe phenomena, which basically don’t exist in nature; and in the lab we might sometimes not observe phenomena, which do occur in nature. In the case of observational studies, the researcher should ideally not interfere in the course of events; so control over the happenings of the study, including, for example, which individuals will and won’t be infected, is entirely up to the study subjects (and, of course, chance). This is the fundamental problem in causality studies, which ascertain what is the cause and what the effect. If we find that men infected by Toxoplasma tend to disregard social norms, then without additional information we cannot determine whether the behavioral characteristic arose as a result of infection, or whether it was conversely the reason they got infected. A great advantage of observational over experimental studies is that they carry a much smaller risk of experimental artifacts, so the phenomena which we prove through them usually have biological significance – they really occur in nature. From a practical standpoint, there is one more difference between experiments and observation. Because a researcher has no control over the course of an observational study, he must generally use much more sophisticated statistical techniques during the evaluation of collected data.
Bubeek) (see Fig. 22). For the next nineteen days he followed how quickly the infected and uninfected mice disappeared from the room. Again it turned out that the infected mice were devoured by the predator first (28)
Fig. 22 A laboratory predation experiment. The long-eared owl (Asio otus), christened Bubeek, was placed in a room of 42 mice, half infected and half uninfected with Sarcocystis dispersa at the start of the experiment. Overall, Bubeek caught infected mice earlier than control mice. Photo Petr Voříšek.
Over the years I have conducted three experiments on humans, which I at first joking labeled as pseudopredation experiments. Rarely will a predator catch today’s human, whether Toxoplasma tries for it or not, but characteristics which should make prey more liable to get caught by a predator, should similarly predispose a person, for example, to get run over by a car. With my undergraduate student Jan Havlíček, for the first time I successfully finished this study, in which we followed the prevalence of toxoplasmosis in the victims of car accidents in a Prague hospital emergency room. It was a long-term study, in which we collected data for over three years. Fortunately, at the time we already had a control from the data of the aforementioned Petr Kodym, who discovered from epidemiological surveys, how Toxoplasma often occurs in healthy men and women in the center of Prague. We could therefore compare the prevalence of toxoplasmosis in various age groups of men and women in the normal population, with the prevalence of toxoplasmosis in the age groups of 146 people involved in a car accident, whether as the drivers who caused the accident or as the pedestrians who were hit by the car. The results were interesting. It turned out that Toxo positive people, both drivers and pedestrians, men and women, really are 2.65 times more likely than Toxo negative people to be involved in a car accident (29)
This study was immensely difficult to publish, most likely because it has quite a substantial practical impact. Even I, were I to receive a manuscript for review, especially one from a similarly exotic country like the Czech Republic (where is it? in former Yugoslavia?), I’d probably also view it with skepticism. According to the World Health Organization about 1.4 million people die as a result of traffic accidents. If we assume, for simplicity’s sake, that the occurrence of toxoplasmosis in the world is the same as in the Czech Republic – about 30% – we find that just in traffic accident hundreds of thousands of people die as a result of toxoplasmosis (see Box 50 How many road traffic victims does the “harmless” toxoplasmosis have on its conscience?). This would mean that latent toxoplasmosis, which most have assumed has no practical significance, is actually the second deadliest protozoan killer after malaria. We sent our article to several journals, at first to the most prestigious, and then to the less prestigious; and finally after two years we were able to publish it in the fifth or sixth journal. And this was still in those idyllic times, when the vast majority of editors forwarded almost all the manuscripts to be reviewed. Today editors likely have such an overabundance of manuscripts (perhaps from India, China and post-communist countries), that they send a large part of them back to the authors without review. It’s not unusual for a controversial manuscript to go through ten editorial boards, until it reaches its first reviewers (Box 39 How I outwitted unwelcoming editors and reviewers).
Box 39 How I outwitted unwelcoming editors and reviewers
From personal experience I can confirm that the game of ping-pong which a beginning author is forced to play with the editorial boards of scientific journals, whenever he tries to publish a manuscript that’s even a little controversial, is a quite frustrating event. A researcher sends his carefully pampered manuscript to a journal’s editorial board, which in a couple days sends him a formal email. The editor regrets to say that lately they have received such an overabundance of quality manuscripts that they can published only the very best of them, and the rest, like this certainly very valuable manuscript, must be rejected without review. In any case, the editor thanks the researcher for choosing to publish his article in specifically this journal, and looks forward to working with him in the future. Sometimes a member of the editorial board will write an additional brief rejection statement, concluding that the article likely wouldn’t have made it through thereviewing process anyway. About half the time the manuscript is forwarded to reviewers, and rejected based on two to three reports. Approximately every third review is valuable and the reviewer gives useful recommendations for modifying the manuscript. Unfortunately, in the case of controversial results, just one negative review out of three is enough, and the editor feels it’s safer to reject it. Then there is nothing to do but rearrange the manuscript to fit the format of another journal, send it to another editorial board, and hope that finally in some journal it will get three rational reviewers and an editor that’s in a good mood, and finally be accepted. Nevertheless, if an author doesn’t give in disgust, but perseveres, all quality (as well as most not quality) manuscripts will finally make it into some journal. There are a lot of scientific journals and the law of probability works reliably – sooner or later the manuscript must get a favorable editor and reviewers. So the main thing is to not get discouraged and find a way to make waiting for a fortuitous event more enjoyable. For myself I discovered a completely suitable method. I carefully follow how many editorial boards the manuscript went through before it was accepted for publication, and look forward to one day, based on this data, write out a study on the relationship of the number of citations, and thus the quality proven post-publication, and the number of boards which originally rejected the article. When today I get an editorial board’s rejection of a manuscript (and recently I got three such rejections within two days), I certainly won’t be too happy about it. But at the same time I’ll tell myself – hey, I got new data for my future study.From personal experience I can confirm that the game of ping-pong which a beginning author is forced to play with the editorial boards of scientific journals, whenever he tries to publish a manuscript that’s even a little controversial, is a quite frustrating event. A researcher sends his carefully pampered manuscript to a journal’s editorial board, which in a couple days sends him a formal email. The editor regrets to say that lately they have received such an overabundance of quality manuscripts that they can published only the very best of them, and the rest, like this certainly very valuable manuscript, must be rejected without review. In any case, the editor thanks the researcher for choosing to publish his article in specifically this journal, and looks forward to working with him in the future. Sometimes a member of the editorial board will write an additional brief rejection statement, concluding that the article likely wouldn’t have made it through thereviewing process anyway. About half the time the manuscript is forwarded to reviewers, and rejected based on two to three reports. Approximately every third review is valuable and the reviewer gives useful recommendations for modifying the manuscript. Unfortunately, in the case of controversial results, just one negative review out of three is enough, and the editor feels it’s safer to reject it. Then there is nothing to do but rearrange the manuscript to fit the format of another journal, send it to another editorial board, and hope that finally in some journal it will get three rational reviewers and an editor that’s in a good mood, and finally be accepted. Nevertheless, if an author doesn’t give in disgust, but perseveres, all quality (as well as most not quality) manuscripts will finally make it into some journal. There are a lot of scientific journals and the law of probability works reliably – sooner or later the manuscript must get a favorable editor and reviewers. So the main thing is to not get discouraged and find a way to make waiting for a fortuitous event more enjoyable. For myself I discovered a completely suitable method. I carefully follow how many editorial boards the manuscript went through before it was accepted for publication, and look forward to one day, based on this data, write out a study on the relationship of the number of citations, and thus the quality proven post-publication, and the number of boards which originally rejected the article. When today I get an editorial board’s rejection of a manuscript (and recently I got three such rejections within two days), I certainly won’t be too happy about it. But at the same time I’ll tell myself – hey, I got new data for my future study.
About four years after we did, a team of Turkish researchers obtained and published the same results (30). I was one of the reviewers of this manuscript and must admit that I had rather mixed feelings about their work. The usual prevalence of toxoplasmosis in Turkey is relatively high, but in this case the prevalence of toxoplasmosis in the controls turned
Fig. 23 Toxoplasma-infected blood donors (gray bars) have a longer reaction time than do uninfected blood donors (white bars). The increase in reaction time was most pronounced during the second minute of test, and accounted for 7% of the variability in reaction time.
out suspiciously low. I was glad that someone else had confirmed our results, but to this day I am not sure whether this study was completely sound. In 2009 another Turkish team published results of an analogous study, which once more confirmed that latent toxoplasmosis heightens the risk of traffic accidents (31). In this case the prevalence of toxoplasmosis in the control population fit the expectation. In the same year we also published further results of a broader pseudopredation study; but I’ll describe these results in the chapter dedicated to the protective role of the RhD antigen (32).
We observed the effect of toxoplasmosis on the behavior of infected persons not only using the pseudopredation tests, but also by measuring the reaction time of infected persons under laboratory conditions. We sat the test subject before a computer and let him complete two tests. In one we projected the image of a square white frame into the middle of the screen, and had a small white square appear in the frame in irregular 1 to 8 second intervals. The test subject was to click the mouse the instant the small white square appeared. The second performance test was a bit more complicated. At regular intervals a three digit number appeared in the middle of the screen and the person was to click the mouse whenever the current three digit number shared at least two digits with the previous. This was more difficult for the test subjects, for now they had to analyze information, not just react to a simple stimulus. This experiment was carried out by my colleague Havlíček on blood donors in a blood donation center near our department. While the blood donors were hooked up to a platelet separation machine, he convinced one after another to take our two performance tests after finishing their blood donation. In his undergraduate work, Jan Havlíček proved that Toxo positive persons really perform significantly worse in the first, simpler test than do Toxo negative persons. The differences between the two groups, however, were apparent primarily in the second minute of the three minute test. In the first minute the performance of both Toxo positive and Toxo negative persons was fairly high; in the second minute the Toxo positives got tired and their performance went down; and in the third minute the performance of the Toxo negatives also went down, so then the performance of the infected and uninfected pretty much evened out (Fig. 23).
Professor Andrew Smith of Cardiff University helped us out a lot in this study, providing us not only with the program for measuring reaction time, but even sending us the devices with which we carried out the measurements. Thanks to his technical support we were able to successfully finish this study. The second, more complicated test showed no difference between Toxo positive and Toxo negative persons. It might have been because fewer people took the test, or maybe because the effect of toxoplasmosis on processing information is more complicated and some of its manifestations in this test mutually canceled each other out (Box 40 What all influences the results of a performance test?). In any case it would probably be worthwhile to eventually go back to the results of this test, and once more try to analyze the data which Havlíček and, in the past few years, Martina Vitáková-Novotná have gathered.
Box 40 What all influences the results of a performance test ?
When testing simple reaction time, we want to measure how quickly the test subject is able to reaction to a simple stimulus. Unfortunately, the performance of the test subject is influenced not only by his actual reaction time, but also by a number of other factors. It’s clear that test subject’s performance is highly influenced by his motivation. When the person tries, he reaches a significantly better reaction time, than when his approach to the experiment is rather casual. Similarly, the test results are also influenced by how experienced the person is in such exercises, how little or well rested he is during the test and what instructions he received before the test itself. Therefore, the test must be carried out under the same conditions, the test subjects addressed under the same situations, in the same manner, and, if possible, by the same person. Even so it can easily happen that the same factor, such as Toxoplasma infection, can shorten the reaction time of one type of person but lengthen it in another. For example, men infected by Toxoplasma have, among other things, a heightened level of testosterone (33). Yet testosterone heightens the motivation to succeed as much as possible in performance tests, particularly when an attractive female student is giving the test. Toxoplasma infected men may have slower reactions, but in tests they may appear faster. By all accounts, toxoplasmosis directly or indirectly influences a number of psychological and physical characteristics in infected persons. The results of individual tests may depend on multiple characteristics, so the results, especially whether the infected subjects perform better or worse than the uninfected subjects, depend largely on which effects of toxoplasmosis are prevalent in the given population. A practical suggestion is to independently test various and, if possible, homogenous groups of people (such undergraduate students, blood donors, military drivers), and to analyze the results for each population separately.
We later tested the effect of toxoplasmosis both on undergraduate students and on soldiers. Again it turned out that toxoplasmosis significantly influences performance in tests of simple reaction time. The fact that we originally tested reaction time on blood donors later proved to be a crucial and immense advantage. In the first phase of the experiment, it was very helpful that we didn’t have to take blood from the blood donors to inspect for toxoplasmosis. We only had to persuade them to allow an examination for toxoplasmosis, and agree that scientific studies could use their results. These results included those from further examinations (in blood donors, routinely conducted), along with the results from reaction time tests. The main advantage of the blood donor subjects, which became apparent only much later, lay in that we had available the results of other hematological examinations, including data pertaining to their blood groups. That we even had data available about the presence or lack of Rh-factor in an examined person, finally led to perhaps the most surprising result of our study. But that will get its own chapter (Chapter XVI).