B40

There are some arguments on direct association between cancer and socioeconomic status (SES) because of its association with some lifestyle factors including smoking. In addition, there are many unresolved issues with regard to the rare information from undeveloped and developing country, where tobacco usage is increasing, role of other collateral exposure, and real impact of socioeconomic status in itself, which is important to formulate and carry out effective and strategic tobacco control activities.
 >In this reason, we investigated differences in cancer incidence and mortality between SES adjusting potential confounding factors with the data from the large cohort (Korean National Health Insurance Cooperation Study). From the cohort, 454,691 men aged over 40 years were included in final analysis. Subjects were classified as two group by their occupation (blue collar : 354,651, white collar : 100,852), which were guided by Korean National Standard Classification of Occupation. Cancer death and incident cases were identified by the link with the database from Korean National Statistical Office (KNSO) and the Korea Central Cancer Registry (KCCR). Mortality and incidence of cancer and relative risks were estimated with Standard Poisson Regression Model in each job strata in different smoking status after adjustment for all related confounders such as age at enrollment, residence, smoking status, amount of alcohol consumption, body mass index, fasting glucose level, leisure time physical activities, dietary preference and family histories of cancer.
 >During 6 years follow up (from January 1, 1996 to December 31, 2002), 6,831 and 1,632 cancer death and 11,388 and 2,721 cancer incidence in blue and white collar, respectively, were identified. The frequency distribution of each cancer incidence and death was not different between blue and white collar. There were more current smoker in blue collar and more former smoker in white collar. In multivariate model adjusting demographic and life style factors, mortality of all cancer (Relative Risk (RR) :1.07, 95% Confidence Interval (CI) : 1.02-1.13), incidence of total cancer (RR : 1.08, CI : 1.04-1.13), smoking related cancer (RR : 1.07, CI : 1.02-1.12), upper aerodigestive cancer (RR : 1.20, CI :1.01 - 1.44), and stomach cancer (RR : 1.10, CI : 1.01-1.19) were significantly higher in blue collar. Stratification for smoking status highly related with cancer risk, the increased risks for incidence of all cancer and smoking related cancer in blue collar still remain in current smoker and never smoker but not in former smoker. However, the significantly higher risk in blue collar for incidence of upper aerodigestive tract cancer was founded in former smoker only.
 >This study confirmed the job based socioeconomic class difference in cancer mortality and incidence with the proper adjustment for potential confounding factors including smoking. It is suggested that reducing socioeconomic inequality through improving lifetime circumstances might help reduce the cancer rates and would also contribute to reducing inequality in health, even if intervention program for life style change such as smoking cessation, healthy diet, and prevention infection produce greatest benefit.

Sixth AACR International Conference on Frontiers in Cancer Prevention Research-- Dec 5-8, 2007; Philadelphia, PA