Evolution of inequalities in cancer mortality in Hungary: a process interpreted in the context of epidemiological transition

Katalin Kovács, Demographic Research Institute, Budapest

In 2008 large educational disparities were detected in overall cancer mortality in Hungary, especially among men. This paper examines the onset of these inequalities by analysing time trends between 1971 and 2008 by detailed causes of cancer death and by educational groups. Results will be interpreted in the terms of the epidemiological transition theory. In concordance with findings and theories on risk factors of different types of cancer, the epidemiological theory is involved here in the sense than the processes of toning down of infections; the dietary transition, the tobacco epidemic, and improvement of medical interventions are considered. Cancer trends therefore were analysed in the following grouping: major risk factor is infection (example: stomach cancer), is tobacco smoking (example: lung cancer), is diet (example: colorectal cancer) or ineffective medical intervention (e.g. screening, example: cervical cancer). To examine time trends SMRs based on original data were fitted following the years of introduction of new ICD versions (1979, 1996) and new coding method (2005). Three major turning points were identified. During the seventies educational differences in most cancer mortality were negligible. From the beginning of the 1980s smoking related cancers started grow dynamically among the less educated, while rates among the more educated have been close to stable. This process induced 2-3 fold differences in all dominantly tobacco-related cancer rates among men. Diet-related cancers used to be higher among the more educated, which changes into the opposite at around 1990. Medically manageable or preventable cancers were also slightly more common among the more educated, but due to the decline of mortality from these causes among the more educated, this situation changes around 1995. Cancer of infectious origin, on the other hand, has being more common among less educated throughout the whole period, but rates of highly and less educated are strongly converge.

Presented in Poster Session 2