Still, as the example of Lance Armstrong Human makes clear, sometimes exercise alone is not enough. When Evans began giving 516 to laboratory mice that regularly used an exercise wheel, he found that, after just four weeks on the drug, they had increased their endurance—how far they could run, and for how long—by as much as seventy-five per cent. Meanwhile, their waistlines (“the cross-sectional area,” in scientific parlance) and their body-fat percentage shrank; their insulin resistance came down; and their muscle-composition ratio shifted toward so-called slow-twitch fibres, which tire slowly and burn fat, and which predominate in long-distance runners. In human terms, this would be like a Fun-Run jogger waking up with the body of Mo Farah. Evans published his initial results in the journal Cell, in 2008. This year, he showed that, if his cookie-dough-scarfing mice were allowed to exercise, the ones that had been given 516 for eight weeks could run for nearly an hour and half longer than their drug-free peers. “We can replace training with a drug,” he said.

The drug works by mimicking the effect of endurance exercise on one particular gene: PPAR-delta. Like all genes, PPAR-delta issues instructions in the form of chemicals—protein-based signals that tell cells what to be, what to burn for fuel, which waste products to excrete, and so on. By binding itself to the receptor for this gene, 516 reconfigures it in a way that alters the messages the gene sends—boosting the signal to break down and burn fat and simultaneously suppressing instructions related to breaking down and burning sugar. Evans’s doped mice ran farther, in part because their muscles had been told to burn fat and save carbohydrates, which meant that they took longer to “hit the wall”—the painful sensation encountered when muscles exhaust their glucose store.