Nick Wilson, Christine Clement, Matt Boyd, Andrea Teng, Alistair Woodward, Tony Blakely

As relatively little is known about how socioeconomic position might have affected health prior to the Second World War, we aimed to study lifespan by occupational class in two cohorts in New Zealand. The first study included men on the electoral rolls in Dunedin in the period 1893 to 1902. The second study used an established cohort of male military personnel who were recruited for the First World War. Linear regression was used to estimate lifespan by occupational class. The first study of 259 men on the electoral rolls found no substantive lifespan differences between the high and low occupational class groups. But the second study of 2,406 military personnel found that men in the three highest occupational classes lived 3.5 years longer (95%CI: 0.3-6.8 years) than the three lowest classes (in the multivariable analysis adjusting for age in 1918 and rurality of occupation). We found no significant lifespan differences in one cohort, but a second cohort is the earliest demonstration to our knowledge of substantial differences in mortality by socioeconomic position in this country prior to the 1960s. Implications for public health: This study provides historical context to the long-term efforts to address health inequalities in society.