91av

“TO INFINITY and beyond, to infinity and beyond, to infinity and beyond, to infinity and beyond,” chants the autistic child.

Across the city, a polar bear performs its own routine with the grace and precision of a prima ballerina. Several hundred times each day, it takes 12 steps north along the boundary of its enclosure. At the end, it scoops its head low, then turns and retraces its steps. Its paws touch the same spots of ground on every lap, and the circuit is always completed within a hair’s breadth of 14 seconds.

An estimated 80 million captive animals worldwide perform bizarre, repetitive rituals, known as stereotypies. Animal behaviour experts trying to work out why note that the movements often seem prompted by frustration-carnivores pace their boundaries as if to escape, voles jump against their cage roofs and pigs root fruitlessly at concrete floors. Although stereotypies can become so warped and intricate they bear no resemblance to the original acts, the consensus is that they help animals cope with a cramped or barren environment. Maybe a pacing polar bear can convince itself that the miles it covers each day are cross-country rather than back and forth along the fence. Or maybe such rituals-which some scientists think trigger opiate release-soothe animals.

But psychiatrists have long known that abnormal repetitive behaviours in humans are a symptom of something more worrying. They turn up in over 36 mental disorders including autism, schizophrenia, Tourette’s syndrome and obsessive-compulsive disorder, and brain scans of patients with these conditions often reveal abnormalities in the regions that control behaviour. Now two researchers have found disturbing parallels with animals. They suspect that conditions in some zoos, farms and laboratories are literally driving animals crazy. If they are correct this gives efforts to improve the welfare of captive animals a new urgency.

The first hints came around two decades ago from scientists searching for animal models for human mental disorders. Animals given amphetamines develop stereotypies just as humans do and the researchers found that they also fail tests similar to those that psychologists give to humans to measure their level of inappropriate repetitive behaviour. The tests require subjects to make a simple response and then change it. Human patients may be asked to guess the colour (red or green, say) of a card before it is turned over. Presented with the same choice over and over again, healthy people guess at random, but stereotypers tend to persevere with the choice they made first.

For lab animals the test takes a slightly different form. They are taught to press a coloured button to receive a food treat, or push through a certain doorway to find a hidden morsel. When the reward is removed, normal animals quickly learn to vary their response, but those on amphetamines keep on pushing the same button or going through the same doorway.

Of course, giving a rat a dose of speed is quite different to caging a bear and watching its behaviour go awry, so most experts in animal behaviour and welfare have ignored these findings and persisted with the view that stereotypical behaviour in captive animals is benign. Not Joe Garner from the University of California, Davis. “Stereotypies are seen as a way of coping, or a release of frustration, but that doesn’t explain why every movement is identical, or why improving the environment of older animals doesn’t stop some of them,” he says. Together with Georgia Mason from Oxford University, Garner set out to test the idea that caged animals showing stereotypical behaviour have a similar mental disorder to people who carry out inappropriate repetitive actions.

They looked at bank voles, which in captivity are notorious for repeatedly mouthing the bars of their cage and leaping against the walls and roof in stereotypical fashion. Garner rigged up a Y-shaped maze with a tempting treat of sugary water always in the end of the same arm. The voles followed their stomachs and easily learned to run down the correct arm to get the treat. The real test came when the cupboard was bare. With no food reward in the customary place, would the voles do the sensible thing and switch to running down either arm at random, or would they keep looking where the food had been before?

The outcome was similar to that from tests on amphetamine-dosed rats. Some voles persistently chose the arm that had initially held the sugar, long after the reward had been removed. Crucially, the worst offenders were those who showed the most stereotypical behaviour in their home cages. The researchers have now done similar studies with blue tits and parrots and got the same result: those animals with the most serious stereotypies found it hardest to shake their old habits.

In humans, such entrenched responses point to a failure of systems in the brain that control behaviour. We can all switch to autopilot-when we walk a familiar route, play a well-practised musical instrument or brush our teeth, for example. Responding in a set way to external cues saves having to figure things out consciously every time. If all’s well, we can adopt these habitual responses, or choose to operate on a more conscious, rational level. But people with diseases such as schizophrenia, Tourette’s syndrome, autism and obsessive-compulsive disorder often cannot suppress inappropriate habits. And scans reveal unusual levels of activity in regions of their brain called the frontal cortex and basal ganglia.

Habitual routines-like most behaviours-are controlled by the interplay between these two regions. Our frontal cortex weighs up the importance of information coming in from the five senses, in the context of memories and how we happen to be feeling, and then sends messages down loops of nerves to the basal ganglia. Depending on the information they receive, the basal ganglia either trigger body movement or inhibit it (see Diagram). Between the cortex and basal ganglia, the nerve loops operate on a stop-go system: one pathway within a given loop promotes movement, and the other opposes it. Whether a behaviour is performed or not depends on which pathway has a stronger influence on the basal ganglia’s output. And here’s where things can go wrong, because if there’s too much “go” in a loop, the person becomes overactive. If there’s not enough “stop”, the same behaviour gets repeated again and again-as when people with schizophrenia twitch, rock or chant.

Garner and Mason suspect that a faulty “stop” pathway is also at the root of the repetitive movement sequences performed over and over by stereotyping animals. They believe that when the cortex relays cues it has picked up from the environment, the automatic “go” messages flow out unchecked and a learned or instinctive sequence just keeps on firing. “It’s not that you’re activating something inappropriately, it’s that you’re failing to inhibit it,” says Garner. He notes that stereotypies are worse in barren environments. Without many external signals to cue the next behaviour, the animal continues with its routine. At the same time, such animals can be unusually fast to switch to new behaviours if the cues are there, suggesting that their “go” pathways are working normally, but are unchecked. When Garner’s stereotyping voles weren’t producing their rituals, they shuttled between behaviours at higher rates than normal.

Garner points out that it’s very stressful for animals to be trapped in an environment that is so different from the one for which they are adapted by evolution. Eventually, preventing them from following instincts such as burrowing or stalking prey may sabotage neurotransmitters in the brain-the same neurotransmitters that pass signals between the neurons of the behavioural control loops. “The behaviour is initially repeated for a good reason,” says Garner. “But if at the same time chronic stress is altering brain neurochemistry, it somehow turns whatever is being done into super-ritualised weirdness.”

He likens the motor systems of the brain to a vinyl record-the more you play one track, the deeper the groove becomes. “Somehow, perhaps via dopamine release, the coincident chronic stress enhances the deeper cutting of that groove,” he says. (Dopamine is the neurotransmitter that is overactive in schizophrenia.) This might explain why the stereotypies of young animals often stop when their surroundings are improved, but the same improvements have no effect on older animals. The behaviour takes time to become entrenched.

Michelle Turner of Durham University, who studies autism, agrees that animals might be prone to the same damage to their control system as humans, even though the cause is different. She notes that people with autism exhibit both true stereotypies and “variable repetitive behaviour”, which is more akin to that found in people with obsessive-compulsive disorder, where repetitive actions such as hand-washing don’t tend to be precisely the same every time. But Turner has found that only people with true stereotypies fail to choose randomly in tests. “Certain types of repetitive behaviour seem to [correlate] with performance at certain types of cognitive tasks,” she says. “I think there are enormous parallels with the animal work.”

But vet Daniel Mills of the University of Lincoln is more cautious, pointing out that not all stereotypies in animals become permanent. He has found that putting mirrors into horses’ stables to mimic social interaction almost instantly stops stereotypies that have been going on for up to 12 years. “Brain pathologies don’t disappear overnight,” he says, and since stereotypies can be stopped so quickly, he thinks there must be other reasons animals repeat certain behaviours. Perhaps they are just responding to a constantly present cue. “The animal is using normal processes and pathways in an attempt to cope. It might not be successful, but it is the best the animal can do,” he says.

Mason agrees that not all stereotypies are linked with brain damage-only the irreversible ones. And even Mills accepts that for some stereotyping animals, Garner and Mason’s theory is probably spot on. “Some stereotypies are very rigid,” he says. “Those animals have really blown a circuit, and are clearly being controlled at a very primitive level.” According to him, such wonky wiring is most likely to occur “when a mismatch between the environment in which the animal has evolved and the one in which it lives is extreme, or when it occurs during development when the nervous system is more plastic”.

But does this really matter to the animals? Being blissfully loopy might not be such a bad thing. Garner points out that the frustration expressed by human patients who feel trapped in a body on autopilot would suggest otherwise. “But maybe they are just embarrassed about behaving strangely in front of white-coated psychologists,” counters Mason. “Perhaps if left to their own devices, they wouldn’t feel so bad, and that may apply to animals too. We just don’t know yet.”

What is certain, however, is that many stereotyping animals injure themselves and show signs of poor health. There’s a simple solution to minimise such damage, say the scientists: enrich the environment. People with autism stereotype less when they have strong cues to direct their behaviour, and in animals the problem is less likely to emerge if they have more freedom to follow their instincts.

Enriching the environments of lab animals could do more than just improve their welfare-it could make research more reliable. If caged animals are going round the twist, it could invalidate the huge amount of research that is based on testing their behaviour. Ironically, one of the reasons lab cages are kept barren is to standardise conditions between labs and make results more consistent. But because some individuals seem more prone to stereotypy than others, Mason suspects that barren environments might be the unwitting cause of the notorious variation in behavioural test results. “If these animals are so messed up, perhaps there’s no point in doing science on them,” says Garner.