10
LOOK WHAT THE CAT DRAGGED IN
I cannot make you understand. I cannot make anyone understand what is happening inside me. I cannot even explain it to myself.
—FRANZ KAFKA, The Metamorphosis
And in whatever houses a cat has died by a natural death, all those who dwell in this house shave their eyebrow.
—HERODOTUS
TO THE EXTENT that we manage the animal life in our houses, we tend to try to get rid of it, as with the case of German cockroaches. But there is one exception, one very important exception—our pets. Our pets are good. They keep us happy and healthy. In exchange, we feed them. We pet them. We walk them more than we walk our own human children. In a biological world full of ambiguities, our pets are unambiguous, unambiguously good. Or at least they seem to be until we begin to consider the species that ride into our houses with our pets. When we do, everything suddenly (once more) gets complex.
When most people think of pets, they are reminded of their own domestic animals. Maybe their first pet or a pet alongside which they weathered some storm. But, as an ecologist, when I think about pets, I am reminded of the first job I ever had in science, studying beetles. I was an eighteen-year-old undergraduate student. I applied for an internship to watch monkeys. I didn’t get it, so I applied for a second internship, to watch beetles. I got it. So it was that I began to help a graduate student at the University of Kansas, Jim Danoff-Burg, with his thesis.1 Jim was studying a group of beetles that live with ant species of the genus Liometopum; these ants give off an odor of citrus, apricots, and slightly sweet blue cheese when alarmed (which they always are when ant biologists are poking at them). These ants build large underground nests in the desert and can be found by turning stones or searching at the base of juniper or pinyon bushes. You can find them at night without a flashlight, by smell, if you don’t mind also finding rattlesnakes.
The beetles that live with Liometopum ants are, for all practical purposes, their pets. The beetles have evolved the ability to solicit food and shelter from the ants. Ants produce specialized compounds that they use to appease their colony mates; for example, these compounds can help calm the colony after some danger has passed. The beetles produce chemicals that mimic those produced by the ants themselves, chemicals that appease the ants, much as humans are appeased (and pleased) when petting a dog. The beetles also rub on the ants, like a cat rubbing your leg or a dog pushing against you to be petted. In doing so, they pick up the odor of the ants on their body and begin to smell like ants. Smelling like an ant is key; it keeps the ants from eating the beetles. Ants kill and eat nearly anything that moves and doesn’t smell like a close relative (distant relatives, on the other hand, inasmuch as they tend to be from neighboring colonies, are eaten without qualms). Having calmed the ants and made themselves invisible, the beetles walk around eating bits and pieces of the ants’ unattended food. Some species of these beetles even convince their host ants to feed them. They sit in front of the ants with their front “paws” in the air and beg.
At least part of the effect that the beetles have on the ants may be negative in that they extract a tithe of food. Though, like dogs or cats in early human societies, this food may be the leftovers the ants don’t want. The beetles may also feed on pests and pathogens that live in the ants’ garbage pile, which is beneficial to the ants. When working on these ants, Jim and I decided to test whether the beetles were, on average, costly or beneficial to the ants.2 We tried an experiment in which we put ants into film canisters with and without beetles and then tallied how long the ants in each treatment survived. The difficulty with the experiment was that we had to do it while driving around in Jim’s car (as we traveled from place to place looking for more sites with ants and beetles). The ants with the beetles seemed to survive longer than those without. We hypothesized that perhaps the beetles calmed the ants and prevented them from wasting energy on wild panic. That the ants might panic was understandable. After all, they were riding in a film canister through the desert in an old Toyota Tercel, bathed in the odors of their desperation and our peanut butter. The experiment suggested the potential benefit of the beetles to the ants, at least under some conditions.
The experiment with the ants and the beetles wasn’t easy, but it was nonetheless easier than the same experiment with humans and pets. No one would give you permission (anymore) to put a human and a dog in a giant jar and wait to see whether humans with dogs in their jar live longer than those without dogs. Actually calculating whether our dogs and cats (or for that matter indoor pigs, ferrets, and even companion turkeys) benefit us in terms of health and well-being is difficult. Dogs with special roles, such as performing services for people with disabilities or detecting cancer, have obvious direct benefits for humans. But what about the average dog or cat, the household pet? A small number of studies have found that having a dog, and to a lesser extent a cat, can reduce stress, anxiety, and feelings of loneliness, an effect similar to the one we thought the beetles had on the ants. It is an effect that is the basis of the increasing numbers of emotional support animals, be they dogs, cats, or even pigs and turkeys. In one study dog owners were even found to be more likely to recover after heart attacks than were dogless people. Cat owners, on the other hand, were less likely to recover than were catless people.3 But these types of studies are few, are correlative, and tend to consider relatively small numbers of people. Also, they do not include the other effects dogs and cats have on our lives. They do not account for the possibility that dogs and cats, just like house flies or German cockroaches, bring species into our lives, species that can make us sick and maybe even species that can make us well.
ONE OF THE SPECIES that cats bring in is the parasite Toxoplasma gondii.4 Toxoplasma gondii is emblematic both of the ways in which species ride into our lives on our pets and of the difficulty of understanding whether our pets are good or bad for us. The story of Toxoplasma gondii, as it concerns us here, begins in the 1980s. A group of researchers in Glasgow was studying house mice infected with Toxoplasma gondii. They noticed that the infected mice seemed hyperactive when compared to uninfected mice. They wondered whether it was because of the parasite, so they gave all their mice hamster wheels to run on. The student in the group, J. Hay, counted how many times each mouse went around the wheel. During the first three days, the uninfected mice did more than 2,000 rotations on the wheel. A lot! They were hardly calm. But the infected mice did twice that many turns. What was more, as the days went on, the difference grew even greater. By the twenty-second day of the experiment, the infected mice were doing 13,000 turns on the wheel compared to just 4,000 turns of the uninfected mice. Here was rodential pathos of an extreme sort. Something interesting must be happening, the researchers suggested, in the brains of the infected mice. But then they went one step further. They hypothesized that perhaps the hyperactivity of the infected mice was adaptive for the survival of the parasite; perhaps the parasite caused the mice to be hyperactive so that they were more likely to be eaten by a cat. The parasite Toxoplasma gondii can carry out the final stage of its life cycle only in cats.5 But that was as far as the group went. They published this work and offered their hypothesis for other scholars to investigate. This was a strange enough possibility on its own. Then, ten years later, things got far stranger thanks to Jaroslav Flegr.
Flegr was born and works in Prague. In Prague, he took the necessary steps to advance his career in evolutionary biology, did some fine work, obtained a PhD, and even landed that rarest of prizes, a faculty job at Charles University. It was at Charles University that Flegr started to study parasites. Initially, he studied the parasite Trichomonas vaginalis, the cause of trichomoniasis. Then, beginning in 1992, Flegr became fascinated by Toxoplasma gondii. He began to read the work Hay had done with the hyperactive mice on wheels. When he did, he was convinced that the parasite was indeed manipulating the brains of house mice for its own ends, as Hay had hypothesized. This was happening, he thought, in houses around the world, houses in which mice ran out from beneath the stove and were pounced upon by cats, all to the advantage of the parasite. It is hard to say why Flegr was so readily convinced that Hay was right; it is even harder to say why his next thought was to wonder whether he himself, like the hyperactive mice, was also infected.
Flegr started to list for himself the ways in which his own behavior was unusual. In some real way he felt like the infected mice. It wasn’t as if he was running faster than other people on a treadmill. But he did do things that, if he were a mouse, seemed more likely to get him killed and, were he living in the wild, maybe eaten by a large cat. Maybe, in addition to making mice more active, the parasite also made them less risk averse, and maybe the parasite was having the same effect in him. Once, in Kurdistan, he had found himself in a situation in which bullets were flying around him, and yet, he was not concerned about dying. At home in Prague, he was not afraid of traffic. He darted between cars, amid the orchestra of screeching brakes and honking horns, much like the infected mice dart out into the open. Nor, during the communist times, had he worried about announcing controversial ideas publicly, even with ample evidence that those who did were imprisoned or worse. How to explain it all? He must be, he began to think, infected, transformed, a man who, like Gregor in Kafka’s Metamorphosis, was acting out a drama beyond his control.
Not long after Flegr started to have these ideas did he decide to have himself tested for exposure to Toxoplasma gondii. When he did, he found out that, yes, his blood contained antibodies to the parasite. He had been infected. He began to wonder which actions were his own and which were instead the impulsive gestures of the parasite. Having this very idea itself—that the parasite might be manipulating him—was reckless in a way that was emblematic of the effect the parasite might have. It was the kind of idea that was likely to marginalize him from his international colleagues. To be frank, the whole thing just seemed absurd. But he was in Prague, where wild ideas have long had a home.
By the time Flegr became interested in Toxoplasma gondii, scientists had begun to learn a little more about the parasite. Toxoplasma gondii, as Hay and colleagues noted, infects house mice (Mus musculus). But it also infects other household rodents, including both Norway rats (Rattus norvegicus) and black rats (Rattus rattus).6 It can also infect geckos as well as pigs, sheep, and goats. The parasite gets into these animals when they inadvertently ingest soil or water in which its oocysts (akin to egg cases, from the ancient Greek oon for “egg” and kyst for “bag” or “bladder”) are present. In the host, the beginnings of a Greek drama, or a drama described in Greek terms anyway, unfold. The stomach enzymes break down the hard wall of the oocyst, which releases the sporozoite form of the parasite (where sporo is from the Greek for “seed” and zoite the Greek for “animal”) into the animal’s intestines. Once there, the sporozoites invade epithelial cells inside which they turn into tachyzoites (tachy is from the Greek for “swift”) that divide swiftly until the cells they are in die and rupture, whereupon the tachyzoites spread through the bloodstream and colonize cells in other tissues of the body. Eventually, the host immune system catches up with the parasites, and when it does, the parasites take on a new form, the bradyzoite (brady is from the Greek for “slow”), a form that hides within the cells of the host—in the brain, muscles, and other tissues—waiting, slowly, patiently, for the host to be eaten.
The parasite waits because in order to complete its life cycle it must be in the intestines of a cat. Toxoplasma gondii is a protist.7 Like many protists, it requires very specific conditions in which to have sex and produce oocysts. It cannot have sex or produce oocysts in the soil or in rodents, geckos, or, for that matter, pigs or cows (in which it sometimes finds itself). It is picky. It finds love and fruition in the lining of felid intestinal epithelia, and only there (and they say online dating is tough). It doesn’t seem to matter what kind of cat, but it must be some kind of cat. The parasite has been caught having sex in seventeen different species of felids. In this way, the life cycle of Toxoplasma gondii is incredibly dependent on a set of relatively unusual occurrences happening in a particular sequence. This dependence is a very important feature, a defining feature, of the parasite’s life.
Once a male and a female of Toxoplasma gondii meet in a cat’s intestine and mate, they finally produce more oocysts. The oocysts then ride the fecal highway, down the cat’s gut and out into the environment. A single, small hunk of cat poop can contain twenty million oocysts. The oocysts are as persistent as seeds. They can wait for months, or even up to a year, unseen, for a mouse or other animal to ingest them. There are about a billion cats on Earth, so if only one in ten cats is parasitized and sheds Toxoplasma gondii, there may be as many as three hundred trillion Toxoplasma gondii oocysts waiting to be ingested. Conservatively speaking, there are more than 760 times as many oocysts of Toxoplasma gondii as there are stars in the Milky Way, a wriggling galaxy of parasites.8
Where and when mice, rats, and cats are abundant, such as they were around the grain silos of ancient Mesopotamia, the odds that the parasite could complete its life cycle may be high. Nonetheless, any individual lineage of the parasite able to increase its odds of success by making its intermediate host (the mouse or rat, for instance) more likely to be eaten by a cat would have an advantage. All of this had indeed been known or inferred, to some degree, by Hay, whose initial intuition about what was going on would, in subsequent years of research, prove to be right: the parasite was manipulating the mouse.
Humans, as Flegr knew when he first thought he was infected, are often exposed to Toxoplasma gondii in homes via cat feces. As I’ve noted, in nature, the oocysts of the Toxoplasma parasite make their way into the feces of cats and then into the soil or water, setting the stage for the cycle to begin anew. But in houses, oocysts end up in the litter box instead, sometimes in extraordinary abundances.9 If a pregnant woman inadvertently ingests these oocysts, they break open in her stomach and the parasites divide asexually in the cells lining her intestines before spilling into her bloodstream and invading other tissues. Unfortunately, the parasite makes no distinction between the bloodstream of the mother and that of the fetus, and it travels through to the body of the fetus. Fetuses do not yet have their own immune systems. A fetus borrows antibodies from its mother, but not immune cells such as inflammatory T cells. This is a problem because Toxoplasma gondii is ordinarily kept somewhat in check by inflammatory T cells. As a result, Toxoplasma gondii can proliferate unchecked in a fetus during pregnancy, which can lead to mental retardation, deafness, seizures, and retinal damage in the fetus. (Older infections pose little risk to fetuses because parasites from older infections are likely to be established inside cells in the mother’s muscle or brain, not moving through the blood.) These consequences are not common, but neither are they rare.10 For years and years of study of Toxoplasma gondii, this was the end of the story: the parasite lives out a wild cycle in mice and rats and cats and, rather accidentally, poses a risk to pregnant women via cat litter.
But what Flegr also knew was that the form of the parasite to which pregnant women and other humans are exposed is the same form of the parasite to which mice and rats are exposed. At least in theory, if this parasite established itself in the cells of the brain, humans could suffer the same effects of the parasite that mice suffered. Once in the brain, the parasite could, at least theoretically, manipulate human behavior. But it seemed implausible. Whereas mice and rats are relatively small brained and so potentially can be manipulated by a tiny protist, we humans have big brains. The expansion of our frontal lobe, and the conscious thought it allows, is what makes us human and gave us the ability to invent fire, cheese curds, and computers. We have and express complex thoughts and make decisions to act. We are not simply at the mercy of our biochemistry. We are just too smart and too conscious to be controlled by the desires of a microscopic beast. Or so nearly everyone but Flegr thought.
WITH A PARASITE such as Toxoplasma gondii, it is hard to know just how to study what effect it might have on humans. The problem is that the way we usually study the effects of a particular pathogen or treatment is to study what it does to mice or rats, using them as model organisms. So as to avoid experimenting on humans, we poke rodents instead. The taxonomic order to which rodents belong, Rodentia, is relatively closely related to our own order, the Primates. As a result, our cells, physiology, and even immune systems are very similar to those of rodents’, similar enough so that when some chemical has a particular effect on a mouse or a rat, it is very likely to have the same effect on us. Interestingly, whereas one can debate just how much dogs and cats benefit our health, there is no such debate about house mice, Norway rats, or, for that matter, fruit flies. These household animals, each of which has been inadvertently ferried around the world alongside humans, have become central to how we study our own human biology. We study them to understand ourselves; they are our mirrors. But the trouble in the context of Toxoplasma gondii was that we already knew that the parasite seemed to have some effect (whether it was adaptive or not) on mouse behavior. The mice were more active. It was just hard to imagine that the same would be true for us. So, what next? One could cure individuals who showed evidence of latent Toxoplasma gondii infections (which is to say, their immune systems showed evidence of having been exposed to the parasite), but the problem was that no one knew how to kill the slow-growing form of Toxoplasma gondii present inside host cells (the bradyzoite) or even how to distinguish between people with living parasites inside their cells and people whose immune systems had killed off the parasite before it was ever able to establish (but who nonetheless bore the evidence of the fight). The other problem was that Flegr really didn’t have much money for any of these kinds of studies. He had his salary and he had his time. He decided to use the old approach of comparing people who were infected with people who were not. A correlation is not a causation, but it is nonetheless a starting point, a window—however muddy—into what has not been seen at all before.
The correlative study Flegr undertook was not easy, but it was cheap. He wanted to know the behaviors of large numbers of people, how they scored on personality profiles, how risk averse they were, and how likely they were to suffer problems associated with risky behavior (such as getting into car accidents). He went door to door, like a medieval salesman, hawking wild ideas and blood tests. He didn’t go out around Prague but instead kept things simpler. He just went through the halls of his own university department. As he would report in his paper, most of his participants were the faculty, staff, and students of the Faculty of Science at Charles University. He asked his coworkers, 195 men and 143 women in total, the 187 questions of the Cattell’s 16 Personality Factors (PF) survey, an assessment developed in the 1940s and used around the world to assess the magnitude of sixteen different personality factors, including warmth, liveliness, social boldness, and dominance. With the exception of Flegr and his collaborator (both of whom also participated in the study), none of the other participants knew whether they had been infected with Toxoplasma gondii before answering the questions. In addition to the personality test Flegr also gave each participant a skin test for Toxoplasma gondii. Each participant was given an injection of Toxoplasma gondii antigen, and if an immune reaction caused a small bump at the injection site after forty-eight hours, the participant was considered to have been at some point in their life infected with Toxoplasma gondii.11 This didn’t necessarily mean that the participant still had Toxoplasma gondii in their body, or even that the parasite had ever invaded their cells, just that at some point they had ingested the parasite in sufficient abundance so as to cause their immune system to try to fight it off. The work took fourteen months during 1992 and 1993. Flegr’s colleagues at Charles University thought he was quirky, but nonetheless, they consented to participate in the study (and, in doing so, disclosed many details about their lives).
When Flegr considered the data, he saw that men who had been exposed to Toxoplasma gondii—men like him—were different from those who were not. The men who were infected were more likely to be risk takers (to have a high “social boldness” factor on the test), and so to disregard rules and make rapid, and potentially dangerous, decisions. Overall, for both sexes, the personality types of those who were infected differed from those who weren’t. When Flegr looked more closely at the data, they seemed to explain key features of the human world to him. Inasmuch as the participants were his colleagues, the results explained his world. For example, twenty-nine of his fellow professors tested negative for Toxoplasma gondii. Those people tended to be leaders—people who made slow, considered decisions. Ten of the twenty-nine had been department heads, vice deans, or deans. Conversely, only one of the infected professors had ever had a leadership role (as a department head).12 Subsequent studies would show similar patterns. For instance, Flegr found that individuals infected by Toxoplasma gondii were two and a half times more likely to get into car accidents (a finding later repeated in two independent studies by Turkish research groups, a study in Mexico and a study in Russia).13
He was emboldened.14 He advocated more strongly for his ideas. There was something there, but he knew what people would say. They would argue that people who were exposed to Toxoplasma gondii were just different to start with, that risky people were just more likely to get the parasite, for example. He couldn’t rule this out, not in a formal way, but he also couldn’t imagine why being risky would make you more likely to be exposed to a parasite found in cat feces. The idea that risky people were more likely to have cats or to inadvertently ingest cat feces seemed like a stretch.15 Then again, so was his idea.
We don’t know when humans were first exposed to Toxoplasma gondii. In one possible scenario, our exposures to the parasite were relatively rare until the origin of agriculture. With the advent of farming, we began to store grains. Our grain stores fed large populations of grain-feeding insects and mice (Mus musculus). Grain was money, and mice were eating the money.16 As mice populations grew, so too did cat populations, and cats were domesticated so that agriculturalists could more persistently avail themselves of the cats’ services. Once cats were domesticated, our exposure to their feces increased. So too did the frequency with which we were exposed to Toxoplasma gondii.17 By 7500 BCE, a cat was buried in a shallow pit beside a human in Cyprus. The cat was not chopped into pieces. It was not cooked. It appears to have been curled neatly, as many cultures curl their human dead. Cats are not native to the island of Cyprus, so this cat (or its ancestors) must have come to the island with humans in a boat. The human beside the cat was buried with jewels and ornaments; he was powerful and wealthy. This burial implies our relationship with cats has long included an element of reverence, or at least appreciation.18 This cat, in particular, was probably already domesticated (though it is hard to know for sure on the basis of the bones alone).
Our first encounter with Toxoplasma gondii may have been in an early agricultural settlement such as one of those on Cyprus. The man buried beside his cat and the cat might both have been infected. The other possibility is that we began to be exposed to Toxoplasma gondii even earlier in human prehistory. As hunter-gatherers, we might have ingested it, accidentally, from the soil, much as would a mouse. Or we might have ingested it in uncooked meat (this is the other way we can be exposed to the parasite, by eating meat, such as that of pigs or sheep, in the cells of which the parasite dwells). Then, because our ancestors were not infrequently eaten by large cats, every so often we aided and abetted the parasite in its quest to get to its favored final destination. Our ancestors, particularly as young children, were eaten by cats more often than we might imagine. But even if this latter scenario is correct, it still seems likely that with the dawn of agriculture and the welcoming of cats into our homes, the frequency of our interactions with—and infections by—the parasite must have increased. Either way, if Flegr’s hypothesis was right, the parasite might have been affecting our behavior for a very long time. He was, in other words, considering not just how this parasite affects us now but also how it may have been affecting our ancestors for generations. One wonders, for example, whether Genghis Khan was infected by Toxoplasma gondii. Or maybe Columbus.
During the years in which Flegr considered the many ways Toxoplasma gondii might affect humans, and has affected humanity, a few other biologists quietly continued to unravel the effects the parasite had on rodents. One of them was Joanne Webster, a specialist on pathogens spread by nonhuman animals. (Webster calls herself a zoonotic epidemiologist.) Like Flegr, Webster had also decided to follow up on Hay’s experiments in Edinburgh. But unlike Flegr, Webster would do experiments. Hay had worked with house mice. Webster studied Norway rats, lab rats. In Norway rats, just as in house mice, Toxoplasma gondii divides asexually in the bloodstream and spreads through the body, inserting itself into the cells of muscles such as the heart as well as into the cells of the brain. When inside the cells of the brain, the parasite encysts and can stay encysted for years, for the lifetime of the host even. Webster was able to show in one careful experiment after another that when rats are infected with this parasite, the rats become more active, just as do house mice.19 Rats also become unafraid of the ordinarily terrifying smell of cat pee, a change that, just like the hyperactivity of mice, made the rats more likely to get eaten by cats.20 Nature can make ants love beetles; it can also make mice and rats walk right into the gaping mouths of their predators.
Slowly, Webster began to understand how the parasite causes all of these changes in the rats. It appears that once it arrives in the brain, the parasite produces the precursor to dopamine,21 which, together with other chemicals and mechanisms we have yet to understand, causes mice and rats to be more active, to be less afraid of cat pee, and to be more likely to be eaten by a cat. Because the species cats eat can live both indoors and outdoors, both indoor and outdoor cats are hosts to Toxoplasma gondii.22
Webster’s work helped to launch the study of the ways in which many parasites, not just Toxoplasma gondii, manipulate their host’s behavior. We now know such parasitic manipulation to be common. Fungi manipulate ant brains, wasps manipulate spiders, tapeworms manipulate isopods, and so on. But, apart from Flegr, no one who studied Toxoplasma gondii, including Webster, focused on the ways that it might influence humans.
Webster’s job offered her the scope to potentially study Toxoplasma and humans. One of Webster’s faculty appointments is in the School of Medicine at Imperial College, so she works every week among colleagues focused on human diseases. But the sort of correlative work Flegr was doing was unconvincing to her colleagues, as would be any similar work Webster might do. Webster didn’t necessarily have to convince her colleagues that the work was interesting or that the results were significant to continue, but it would help. Academia is built on a culture of respect gained and, just as readily, lost. If you lose the respect of your peers for your work, you lose their support, you lose their collaboration, you lose, too, potentially, any favors you might require of them in the future (and academics seem, nearly always, to be in need of favors). But just as problematic as the fact that the work wouldn’t convince Webster’s colleagues was that it wouldn’t fully convince her either. Her training was in experiments, in testing hypotheses in the lab, and there weren’t many aspects of the story of Toxoplasma and humans that could be studied experimentally. She couldn’t ethically give people Toxoplasma gondii, and no one knew how to get rid of it once it was established inside cells (so she couldn’t cure people and look at the effects). However, as Webster continued her work, she spotted a possibility. Along with many other phenomena, Flegr hypothesized that the parasite might influence not only behavior but also psychological health. More specifically, building on the work of Flegr, E. Fuller Torrey, a psychiatrist at the Stanley Medical Research Institute, and Robert Yolken, a professor of pediatrics at Johns Hopkins University Medical Center, suggested that Toxoplasma gondii might be partially or even wholly to blame for schizophrenia.23 Both schizophrenia and Toxoplasma gondii tended to be clumped within particular families, but not in a way that seemed to be exclusively genetic (rather that it was more related to the houses people lived in than the genes they had). In addition, the drug used to manage the symptoms of schizophrenia appears to sometimes rid patients of the Toxoplasma gondii hiding in their cells. Presented with these observations, Webster had an idea. She wondered whether the way in which the schizophrenia medications worked was by suppressing or even killing Toxoplasma gondii.
Webster did an experiment. It is what she does. She orally infected forty-nine rats with Toxoplasma gondii. She then pretended to infect an additional thirty-nine control rats by giving them an oral dose of what was effectively saltwater. Each group of rats, infected and control, was then further divided into four groups. One group received no additional treatment, one received valproic acid (a mood stabilizer), one received haloperidol (an antipsychotic), and the last group received pyrimethamine, a drug known to kill parasites, including, under some conditions, Toxoplasma gondii. Afterward, she put the rats, one by one, into a one meter by one meter square pen. In each corner of the pen she put fifteen drops of one of four odors. In one corner, she put wood chips soaked with fifteen drops of the rats’ own smell, rat urine. In another corner, she put wood chips soaked in a neutral smell, water. In the third corner, she put wood chips with rabbit urine, with the idea that rabbit urine should have no specific effect on the rats because rats have no reason to be scared of or attracted to rabbits. And in the final corner, Webster set down wood chips soaked with cat urine. Webster works at one of the most prestigious universities in the world. She has made major discoveries, and yet here she was, day after day, laying out the urine portfolio. After the pen was set up, an individual rat was released and Webster or someone from her team watched and noted the proportion of time the rat spent in each corner. This was done again and again and again for a total of 444 hours of observation of eighty-eight rats. When the data from these observations were tallied, they totaled 260,462 lines. Webster took the time to count. Those lines of data showed that the uninfected rats spent more of their time huddled near the familiar and “safe” smell of their own pee or that of other innocuous animals such as rabbits. The uninfected rats wisely avoided the cat pee areas. The infected, but unmedicated rats behaved differently. They entered the cat pee corner more often, and once there they tended to stay, as if completely unaware of the potential danger the pee might signal. Amazingly, the rats infected with Toxoplasma gondii but treated with either of the schizophrenia medicines or the antiparasite medicine acted more like uninfected mice. Compared to the rats infected with Toxoplasma gondii, but not treated with any medicine, they were less likely to enter the cat pee area and, if they did, they didn’t stay as long. They were, to use a loaded term, cured.24
Webster published her paper on schizophrenia, schizophrenia medicines, and Toxoplasma gondii in 2006. The work was compelling but still confined to mice, not humans. A human study needed to be done, but it couldn’t be just correlational (at least not for Webster). Nor could it be experimental. There was, though, a third option, a longitudinal study. Someone could track people through time to see whether individuals infected with Toxoplasma gondii were more likely, over years, to develop schizophrenia as compared to uninfected (but otherwise similar) individuals. This wasn’t Webster’s kind of study; it was not the sort of science she did, and yet, if someone were to find a way to do such work, it would be an elegant test, a test that would extend her work in a way that might finally get the attention of doctors. It was hard to imagine who might have the right sort of data. Such a data set would need to include not only health data from different time points but also blood samples from those time points. Among the only groups in the world with such data was the United States military.
The US military collects health data on all recruits. It also collects samples of their blood. Another epidemiologist, David Niebhur, at the Walter Reed Army Institute of Research, decided to study these data to see whether schizophrenia was, indeed, associated with Toxoplasma gondii infection. Niebhur went through the military database and found 180 service members who had been medically discharged from the army, navy, or air force between 1992 and 2001 because of a schizophrenia diagnosis. In the database, Niebhur and his colleagues then found another 3 soldiers without schizophrenia for each person diagnosed with schizophrenia. These control individuals matched the patients diagnosed with schizophrenia in terms of age, sex, race, and branch of military service. The researchers examined the blood serum samples collected by the military to see whether individuals who developed schizophrenia were more likely than controls to have been infected with Toxoplasma gondii prior to the onset of schizophrenia. They were. The soldiers who were discharged with schizophrenia were significantly more likely to have tested positive for exposure to Toxoplasma gondii than those solders who were not diagnosed with schizophrenia.25 Niebhur and his colleagues found that people exposed to Toxoplasma gondii have a 24 percent higher risk of developing schizophrenia at some point in their lives than people who have never been exposed. If you have been infected by Toxoplasma gondii, ever, your risk of schizophrenia is at least 24 percent greater than that of someone who has not been. Time has added nuance to the work led by Niebhur and his colleagues, as has replication. The number of papers published on the parasite has increased. To date, fifty-four studies have looked at links between schizophrenia and Toxoplasma gondii. All but five have found evidence of such links that suggest that infection by Toxoplasma gondii increases the risk of schizophrenia.26
Stepping back, it now seems Flegr was on the right path. Toxoplasma gondii appears to act in our brains much as it does in the brains of mice and rats. Nor are we the only primates affected. A recent study has shown that infection by Toxoplasma gondii appears to cause chimpanzees, our close relatives, to be attracted to the smell of cat pee, specifically leopard urine.27 Infected humans, or at least infected human men, are also more likely to regard the smell of cat urine as pleasant than are uninfected men.28
The proportion of people who have been infected by Toxoplasma gondii is huge. Some of these infections resulted from eating meat that was not fully cooked, meat in which the parasite lurks, wriggly, in muscle cells. But many are from our cats. Just how common are Toxoplasma gondii infections? In France, upward of 50 percent of all people show evidence of latent infections. This parasite could explain the behavior of much of a nation. It isn’t culture that makes the French enjoy red wine, meat, and cigarettes so much—it’s just that their parasites give no shit about risks. But lest the non-French get too haughty, I should note that the infection rate in other countries is also high. Forty percent of Germans have been infected. In the United States, upward of 20 percent of all adults have been infected by Toxoplasma gondii. Globally, more than two billion people have been infected at some point in their life.29
In and of itself, the story of Toxoplasma gondii is a big deal. Toxoplasma gondii may well be the most common parasite in humans, or at least the most common parasite with a big effect. Face mites, which we study in my lab, are more common than is Toxoplasma gondii (all adults we have ever sampled have face mites),30 but face mites don’t seem to cause any negative effects. Of parasites with negative effects Toxoplasma gondii, long neglected, appears to be king of commonness. But the story of mice, cats, Toxoplasma gondii, and the other parasites of cats is really a broader lesson about the complexities of opening our doors to domestic animals. Collectively, we seem to have eagerly decided that insects and microbes in our homes are bad, but our pets are good. Yet, when we let a cat in through our front door, the Toxoplasma hiding inside that cat comes in too. Toxoplasma gondii does not travel alone. Dozens of other species, all of them more poorly studied than it, also come inside with our feline friends. And lest cat people, be they ordinary women or men strangely drawn to the smell of cat pee, feel singled out, very similar things are happening with the other kinds of domestic animals we allow indoors.
Over the last twelve thousand years, we have welcomed in many domestic animals, be they cats, ferrets, dogs, guinea pigs, or comfort ducks. Each brought with it other species. Cats brought Toxoplasma gondii. Guinea pigs appear to have brought human fleas. But dogs, oh dogs, dogs are a veritable smorgasbord of worms, insects, bacteria, and more.
Seven years ago, I had the idea to have students in my lab compile a database of all the parasites associated with each kind of domestic animal. The idea was to make a complete list, pet by pet. Meredith Spence was the student charged with cataloging the species that live on domestic dogs. I imagined that after Meredith finished with dogs, someone else would do cats, another student would work on rabbits, and so on. We never got beyond dogs. Meredith spent a year compiling the list for dogs, then two years, then three. Eventually, she graduated from North Carolina State University with an undergraduate degree, went off to work in a veterinary clinic, came back to the university to start graduate school, and is now almost finished with her PhD. She continues to compile the list of species that live on and in dogs.31 The list is that long. It includes some species you might expect, fleas, of course, and the Bartonella parasites that ride in fleas.32 It also includes an entire circus of gorgon-headed worms, worms such as tapeworms of the genus Echinococcus.
Taxonomically, dogs are carnivores. They belong to the same order, Carnivora, as do cats. But, as we know, dogs are not final hosts for Toxoplasma gondii; to this particular parasite, something about a dog’s gut, though similar to the gut of a cat, is fundamentally less lovely. When it comes to parasites—and dogs host many—to each its own. Echinococcus tapeworms, for example, find the mood inside dog guts to be just right. Dogs are the “final” or “definitive” host of Echinococcus tapeworms, which is a boring scientific way of saying dog guts are “the place that the worms go to have sex, produce eggs, and die.”
The story of Echinococcus tapeworms is just beginning to be resolved. We are currently, in our understanding of Echinococcus, only as far along as we were with Toxoplasma gondii in 1980. Most tapeworm species have as their final hosts carnivores—whether those are dogs, cats, or sharks—but can be picky with regard to just which carnivore species. Adult Echinococcus tapeworms greatly prefer dogs. Dogs are the “final” or “definitive” host of Echinococcus tapeworms, which is a boring scientific way of saying dog guts are “the place that the worms go to have sex, produce eggs, and die.” One might imagine that because dogs are, like cats, carnivores, that Echinococcus tapeworms might be able to mate in cats. But they can’t (much as Toxoplasma gondii is unable to mate in dogs). To this particular parasite, something about a dog’s gut is just right.
Once two Echinococcus tapeworms have mated inside a dog, the resulting eggs leave the body of the dog in its feces. Having been deposited, they wait. Grazing animals often ingest a little dog feces, inadvertently, along with their grass. This is one of nature’s less known realities. The grazers are often goats or sheep. Though where goats and sheep are rare, a deer or even a wallaby will do. In the stomachs of the grazers, Echinococcus eggs hatch. The newborn larvae spread through the body of the animal and nestle there in cysts in organs or even in bones. When a grazing animal dies, the dogs are then exposed to these parasites if and when they eat any of the cysts. In the same way, humans can also be infected by the larval Echinococcus tapeworms when eating grazing animals such as sheep. The larval tapeworms can form cysts inside humans much as they might inside sheep, except that inside humans the cysts never stop growing; they can reach the size of basketballs. Eating an infected sheep is the more “glamorous” way to develop an Echinococcus cyst in your body. The less glamorous way is by accidentally ingesting a little dog feces that contain the eggs, which happens far more often than you might hope, such as when people let their dogs lick their faces. The world is vulgar and alive.
The story of the Echnicoccus parasite begs questions. Does this parasite manipulate the sheep that are infected or the humans who are infected so as to make them more attracted to dogs? Do dog people love dogs because they are possessed by the biochemistry of the worm? No one knows; stranger things, as should be clear by now, happen in the wilds of our daily lives.
Some parasites and pathogens of dogs, such as the rabies virus, are common in some regions (or times) but rare in most places, at least today. Echinococcus was one of the most common species in dogs on the list of dog parasites that Meredith Spence compiled and it was common in many regions, but it wasn’t as common as heartworms (Dirofilaria immitis). Meredith Spence now studies dog heartworms; it is the project to which her initial cataloging helped lead her. Heartworms are nematodes. They invade and live in the living hearts and pulmonary arteries of dogs, where they ultimately grow so dense as to clog the normal movement of blood. In the United States, up to 1 percent of dogs have been infected with heartworms. In some countries, more than half of dogs have been infected. The heartworm gets into dogs via mosquitoes. The worms ride inside mosquitoes. Then, in those moments when mosquitoes bite dogs, the worms quickly swim down and out of the proboscis of the mosquito into the wound left by its bite. From the wound, they crawl into subcutaneous tissue of the dog. From the subcutaneous tissue, the worms migrate through the muscle fibers before spilling into blood vessels en route to the heart. By the time they reach the heart, the worms have molted several times and are adults. Ever the romantics, these adult worms mate inside the heart. The evolution of dog heartworms has never really been studied in much detail, nor has that of the many other species of heartworm. Meredith is focused on the mosquito side of the story, so the evolutionary story of the heartworms won’t likely attract her attention anytime soon. Here, then, is a beautiful project for those who might be interested (my guess is that new species of heartworms are common, unnamed, and riding around your neighborhood right now in mosquitoes). Dog heartworms do not usually invade the hearts of humans. It happens rarely enough (hundreds, not thousands, of cases a year) that when it does, the doctors who have spotted the problem gather around and snap selfies with the afflicted patient. In only one case have heartworms been found mating in a human heart; most of them get stuck, during their corporeal peregrinations, in the pulmonary arteries, where, unable to move forward or back, they die. More rarely, a worm gets stuck and dies in the blood vessels of the eyes, brain, or testicles. Again, though, these cases are rare.33
What is not rare is the exposure of humans to heartworms. Many people show antibodies for dog heartworms, which is to say many (perhaps most) humans have been bitten at some point by a mosquito carrying dog heartworms. The worms tunneled into their (perhaps your) skin, but the human immune system killed the worms. When this happens, the person the worm has tried, unsuccessfully, to invade notices nothing different about their life. However, recent research suggests that even a single exposure to these worms can alter the immune health of people, favoring antibodies that predispose people to asthma relative to those who haven’t been exposed. In other words, it may be the case that a mosquito, in landing on you, gives you a worm that your immune system kills but that leaves a legacy, a kind of ghost, in the form of an increased predisposition to sneezing, coughing, and wheezing.34 The fact that we are often being bitten by mosquitoes carrying dog heartworms is largely a function of having let dogs into our lives. The worms are present in our environment because the dogs are present (they can also be present where coyotes or wolves are present, but few are the neighborhoods where wolves and coyotes outnumber dogs). You don’t even have to have a dog yourself to have the worms slide into your body. It is enough for there to be dogs present in your neighborhood. As many as twenty other parasites are common in dogs, connections to their wild wolf ancestry and to the world outside your home. What’s more, as Meredith’s cataloging has revealed, dozens of other parasites are at least occasionally found in domestic dogs.
I find the basic biology of Toxoplasma gondii, Echinococcus tapeworms, and heartworms to be wondrous and fascinating. But, like anyone else, I’d prefer to avoid being infected. Opening the door to a cat or a dog increases the risk of infection. Fortunately, the strongest negative effects of these infections—be they schizophrenia, a tapeworm, or dead heartworms in your testicles—are rare in most regions. Also, some of the risks posed by dogs and cats can be ameliorated by preventive measures dog and cat owners can take. For example, heartworm medicines reduce the abundance of heartworms in dog populations (though their use also increases the speed at which heartworms resistant to those same medicines evolve). Others, though, such as those posed by Toxoplasma gondii, can’t, at least not yet anyway.
I don’t propose to have the answer for how to balance the benefits of having a pet relative to the most common kinds of costs. The answer ultimately depends on where and how we live. In some regions, cats still help protect grain from mice and rats. In some regions, dogs still help shepherds work and protect sheep. But in a modern Western context, what these animals most often offer is companionship. Their value as companions increases in proportion to our need for company, in proportion even to our loneliness and despair. The more urban and isolated we become, the more likely they are to provide such benefits. Also, the more urban and isolated from nature we become, the more likely dogs and cats may provide another, new kind of benefit, that of connecting us to beneficial species.
We first considered the beneficial effect of pets on the bacterial life in houses when we surveyed forty homes across Raleigh and Durham, North Carolina. One of the questions we asked participants in that study was whether they had a dog. Almost 40 percent of the variation from one home to the next in terms of species of bacteria in the home resulted from the presence or absence of a dog.35 This was a huge effect. The effect of the dog was in part the result of a set of soil microbes that were more common in houses with dogs. We imagined that the dogs were just carrying these soil microbes in from the outside, but a recent study found soil microbes living in the fur of many different mammal species.36 It is possible that the normal fur microbiome for many mammals and the normal soil microbiome overlap. In addition to soil bacteria, dogs also left drool-associated bacteria around homes as well as a few fecal bacteria common in dogs, but not as common in humans (and so identifiable in the mix).
Once we had data from a thousand homes, we were able to consider whether cats had an effect on the bacteria in homes. They also do. For reasons we don’t totally understand, some species of bacteria, including some insect-associated bacteria, become rarer when a cat is in a home.37 Maybe the pesticides put on cats in the form of flea collars, drops, and powders kill the insects, which in turn kills their bacteria (though we’d expect this to be the case for dogs too). Maybe cats eat the insects (and kill their bacteria). Still, cats provided a way for hundreds of bacterial species to get into our houses. Most of these species, just as with dogs, appear to be associated with the bodies of cats—their skin, their fur, their feces, their saliva. What cats don’t seem to bring in with them are soil microbes. Perhaps it is because cats are smaller, perhaps it is because they clean their feet. We don’t know.
I suspect that at many points in our history the average bacterial species that a dog or cat brought in would have been, like the protists or worms, reasonably likely to have a negative effect on us if it had an effect at all. But our moment is an unusual one. Today, as outlined in the biodiversity hypothesis, in much of the world we are as likely to be sick from the bacteria we don’t have as from the bacteria or parasites we do. It is possible that, for children who fail to be exposed to enough of the right bacteria, exposure to what the dog or cat brings in offers some of the same benefits as sniffing the biodiverse dust from Amish houses. Recent studies suggest that having a dog does tend to lead to a reduced human risk of developing allergies, eczema, and dermatitis, particularly for children who are born into a household with those animals. The most comprehensive review of the literature found that children who live with pets tend to be less likely to suffer from atopic dermatitis.38 A similar study in Europe found the same result for allergies—having a pet reduced the owner’s risk of allergies, more in some regions than in others.39 Across studies, cats tend to show effects similar to those of dogs, but the effects tend to be weaker and less consistent.40
In parts of the world where we have become distant from wild biodiversity, dogs and cats may be good for our immune systems. The effect of dogs and cats on our immune systems could work in two ways. It may be that the bacterial species dogs and cats bring in compensate for the exposures we no longer receive. We live lives so disconnected from biological diversity that even being exposed to a little gunk from a dog’s foot can be an improvement. Alternatively, our children may inadvertently be benefiting from the fecal bacteria of dogs and cats in their own guts. Children with dogs in their homes tend to acquire dog gut bacteria by eating food from the ground with dog feces on it or by being “kissed” by a dog that has just kissed the backside of another dog.41 It is possible that what dogs (and maybe to a lesser extent cats) offer is not some generalized exposure to bacterial biodiversity but instead a chance to pick up our necessary gut bacteria when they have gone missing. It is now well documented that when specific gut bacteria are absent, their absence can cause a variety of health problems (from Crohn’s disease to inflammatory bowel disease and more). If this feces-eating hypothesis is right, we would expect C-section babies, who often fail to get all of the bacteria that they need,42 to get more benefit from dogs. They seem to. We would also expect that in houses where other sources of fecal microbes are more available, such as through interactions with dirty-fingered siblings, the effect of dogs would be less pronounced. This seems to be the case too. Dogs have less of an effect on the allergies and asthma of children with siblings. In general, I think the evidence so far supports the idea that dogs can benefit us by bringing in soil bacterial biodiversity and by offering up fecal microbes we have lost, but these services are only benefits because we now live in a world disconnected from wild nature, so diconnected that adding a little dog dirt and feces to our lives is a kind of solution. If we put these results back together with the stories of Echinococcus, and the heartworm, the results of having a dog seem likely to differ depending on just which species dogs bring in, whether the species are bacteria or worms, and if worms, which worms. Sometimes, when we are trying to figure out how to make simple decisions to better our lives, biodiversity is a real jerk in its complexity.
The truth is we don’t really know yet what the average consequences are of letting a dog or a cat, much less a ferret, tiny pig, or turtle, into our homes. And if we struggle to figure out whether dogs and cats make us healthy, you can get a sense of why it is also hard to figure out just which of the hundred thousand or so bacterial species from other sources sometimes found in our homes or on our bodies are the ones we want. But that hasn’t stopped people from trying. In fact, at one point in the 1960s it seemed as though doctors might soon plant gardens of bacteria on the bodies of babies across the United States, and perhaps in hospitals and houses too. Then they did.