Cohort analysis of colorectal cancer mortality in the Republic of Serbia over the period 1971-1996

Colorectal cancer is one of the most frequent malignant neoplasms in both sexes within developed countries. In the Republic of Serbia(Serbia) colorectal cancer mortality in 1971 ranged 5 in females, and 4 in males; it became the second leading malignancy in 1982 in females (after breast cancer), and in 1992 in males (after lung cancer). The objective of this descriptive-epidemiological study was to examine colorectal cancer mortality in Serbian population, particularly the effect of cohort variations on death rates in defined age groups over the period 1971-1996. Mortality rates were calculated from unpublished national vital statistics data of the Institute of Statistics of the Republic of Serbia. To estimate the age effect on colorectal cancer mortality, specific death rates were computed for cohorts born between 1892-96 and 1972-76, and died at subsequent time periods. The mortality rates were adjusted by direct method, using the world standard population. Confidence intervals (CI) for death rates were assessed with 95% level of probability. In time trend analysis of mortality, Fisher's test was used as a significance test for linear regression coefficient. In the study period (1971-1996), a share of all digestive tumors in cancer mortality has decreased from 42.0% to 32.3%. However, the mortality risk of colorectal cancer and its share in cancer mortality have increased. For example, in men, the share of colorectal cancer in digestive cancer mortality increased from 20.7% (1971) to 32.8% (1996) and in overall cancer mortality from 7.5% to 10.5%. In women, the share of colorectal cancer in digestive cancer mortality increased from 23.0%(1971) to 35.6%(1996), and in overall cancer mortality from 8.5% to 11.6%. The average colorectal cancer age-adjusted death rates (1971-1996) were 11.2 per 100,000 men (95% CI: 10.1-12.3), and 8.3 per 100,000 women (95% CI: 7.7-8.9). The secular linear mortality trends showed significant increase both in males (y = 11.2 + 0.2x; ? = 0.000), and females (y = 8.3 + 0.1 ?; ? = 0.000). The highest rise in age-specific death rates, according to linear mortality trends, was observed in males over 65 years (7.8% annually), and females between 60 and 69 years (5.9% annually). In cohort analysis of age-specific rates in males, younger birth cohorts were compared with older ones. The increasing colorectal cancer mortality risk has been observed for ages over 40, with statistical significance in age groups over 45. In ages between 45 and 59, and over 60, the youngest birth cohorts were at 2 and 2.5-fold higher cancer mortality risk than birth cohorts of the oldest generations. For example, the age specific colorectal cancer death rates in a 70-74 year group were 2.5-fold higher in men born between 1922 and 1926 (139.3/100,000) than in cohorts born 25 years earlier (58.7/100,000). In cohort analysis of age-specific rates in females, changes in the age under 50 were not so expressive. In all age groups over 50, women of younger generations were at 2-fold higher cancer mortality risk than the oldest ones. The age specific colorectal cancer death rates in a 65-69 year group were doubled in women born between 1927 and 1931 (61.0/100 000), than in cohorts born 25 years earlier (30.5/100 000). According to the present mortality trends, the further increase in colorectal cancer death rates especially in the ages over 40, should be expected in future generations. Consistent increase in mortality risk in all younger birth cohorts of older ages, as well as in successive five-year age groups of the observed generations, could reflect the continuous increase in colorectal cancer incidence attributed to predominantly environmental exposures.