Socio-economic position, area-level deprivation and gradients in cancer incidence: England and Wales, 1971-2016.

Abstract

Background: Social gradients for cancer mortality and survival have been reported but are less clear for cancer incidence where social factors external to health care systems are likely to be of more etiologic importance.

Methods: We examined social gradients in cancer incidence using data from the Office for National Statistics Longitudinal Study (ONS-LS), which selects an approximately 1.1% representative sample of the population of England and Wales. Data were analyzed for each successive ten-year census period from 1971-2011 with outcome data to 2016, the latest date available. Socioeconomic position of individuals was assessed using the National Statistics Socio-economic classification (NS-SEC). Areal level deprivation was measured using deciles of the Townsend Index. Cancer outcomes from the National Cancer Intelligence Network linked to the ONS-LS were examined for all cancers, and more common individual cancer sites. We used logistic regression to generate odds ratios to estimate the risk of a first incident cancer within each follow-up period.

Results: The 1971 ONS-LS census sample population initially comprised 257,803 individuals updated each census; and by 2016 137,755 incident cancer cases. Social gradients in cancer incidence were present for individual cancer sites of lung, stomach, and cervix for both individual and areal measures of socioeconomic standing with the least advantaged having higher incidence rates. Reverse gradients were present for prostate and breast cancers. The relationship of SES to increased cancer incidence for these common cancers is consistent with prior literature, but the striking gradients in these relationships reveal the strong association of SES factors with increasing social disadvantage for these cancers.

Conclusion: The findings demonstrate the importance of socioeconomic position in the incidence of some common cancers prior to diagnosis and treatment and reinforces the need for further research to address the contribution of upstream social determinants in the etiology of cancer.