Abstract

Objectives Evidence suggests that different indicators of socioeconomic position (SEP) contribute to oral cancer risk. Occupational status, as a measure of SEP, may be able to capture aspects of social hierarchy in societies in which employment is highly correlated with other social structures such as caste systems. Often in such societies, the life course of an individual is also influenced by this hierarchy. However, the influence of life course occupational status on the risk of oral cancer is not well understood. This study aims to identify the life course model that is best supported by the data using life course SEP-as represented by occupation-on oral cancer risk in a population in the South of India. Methods Data from the HeNCe Life study, Indian site were used. Incident oral cancer cases (N = 350) were recruited from two major referral hospitals in Kozhikode, Kerala, South India, from 2008 to 2012. Controls (N = 371), frequency-matched by age (5 years) and sex were recruited from the outpatient clinics at the same hospitals as the cases. Life grid-based structured interviews collected information on an array of exposures throughout the life course of the participant. Occupation was coded with the 1988 International Standard Classification of Occupations, transformed to the simplified European Socioeconomic Classification, and further dichotomized into advantageous and disadvantageous SEP at three different life periods (childhood, early adulthood and late adulthood). The analysis was conducted using the Bayesian relevant life course exposure model with a Dirichlet noninformative prior and a weakly informative Cauchy prior to the overall lifetime effect and confounders. Results Participants in disadvantaged SEP throughout their life had 3.6 times higher risk of oral cancer than those in advantaged SEP (OR = 3.6; 95% CrI = 1.6-7.2), after adjusting for potential confounders. While the crude and sex- and age-adjusted models showed a clear childhood sensitive period for this risk, the model further adjusted for behavioural factors could not distinguish the specific life course period best explained by data. Conclusion Occupation status alone could provide a similar overarching risk estimate for oral cancer to those obtained from more complex measures of SEP.