By 1967 a doctor had counted the heart attacks on the Indian railways, and they were at the wrong end of the country. The men of Punjab, who cooked in ghee and took their fat from the buffalo, were having a seventh as many as the men of Madras, who hardly ate fat at all. The railway was the nearest thing India had to a laboratory. The British had laid it down to move cotton and soldiers, and by the 1960s it ran from the Punjab plains to the Madras coast as one employer, on one set of pay scales, with its own hospitals keeping its own records. More than a million men shared a boss, a pension and a doctor, and ate about as differently as two halves of a country can. Sardari Lal Malhotra, a gut doctor who had spent years arguing that saliva protected the stomach from ulcers, went through two years of those records, 1963 and 1964, from twenty-four railway hospitals. Everything the 1960s believed said where the bodies should be. In 1961 Time put Ancel Keys on its cover, and the American Heart Association told Americans to cut their saturated fat and swap in polyunsaturated oil. By that reckoning Punjab was a graveyard waiting to be dug. A railwayman there ate around 75g of fat a day, most of it ghee and milk fat, got through 80g of sugar, and smoked about eight times as much as his opposite number in the south. The man in Madras lived on rice with a splash of groundnut or sesame oil, 7g of fat a day and a spoon or two of sugar, and nearly half of what little oil he used was polyunsaturated. He was already eating the plate the West would be handed in 1980, only stricter. Malhotra accepted 294 heart attacks. At the railway hospital in Perambur, Madras, the rate came out at 102.3 for every 100,000 men. At Ferozepur in Punjab it was 14.4. The blood should have explained it, because by the science of the day the blood was where fat did its damage. The sweepers, the poorest men on the line, had their cholesterol measured at both ends of the country, and it came back the same. Ten times the fat in Punjab, and not a flicker in the blood. Malhotra suspected the protection was in the ghee itself. Milk fat carries short-chain fats that seed oils do not, and he thought they changed how readily the blood clotted. He wrote it all up for the British Heart Journal in 1967, and the West carried on pouring. Two years before that, a Bombay family had put a cooking oil on sale named after the safflower and sold it to India as the oil for the heart. It went on to run free cholesterol screenings across Indian cities, hold conferences for cardiologists and put a dietitian on the end of a phone, and by the 1990s its television adverts were ambulance sirens and stretchers. On World Heart Day 2009 it took the music off six radio stations and played a heartbeat. South Asians now have their first heart attack at an average of 53, nearly six years before the rest of the world. India's own railway had run the free cholesterol screening years earlier, and found the same number at both ends of the country and seven times the heart attacks at the end that cooked in seed oil.