In the Wisconsin sample, Dr Christakis looked at those who had died before their 65th birthday. For the women, there was no significant relationship between their school’s sex ratio and their age of death. For the men, however, a significant relationship did emerge. A percentage-point increase in the male-to-female ratio of a man’s graduation class led to a percentage-point increase in his likelihood of dying before the age of 65. The Social Security data, moreover, suggest that a lack of women during men’s teenage years still haunts their health decades later.

The average white American male who was 65 in 1993 could expect to live another 15 years. Dr Christakis found, however, that those who had come of age around the most available women, however, had a life-expectancy three months longer than that of the least favoured. Three months may not seem a huge difference, but according to Dr Christakis it is comparable to the benefit an elderly person can expect from exercising or losing some surplus weight.