Figure. Age-specific death rates of rectum cancer and colon cancer plotted against the period of birth as cohort-age contours. W Indicates white subjects; NW, nonwhite subjects. A, B, D, and E, Age-specific death rates of rectum and colon cancer plotted as cohort-age contours; the curve of each 10-year age group is labeled by the number representing its central year, for instance, 80 indicating the age group 75 to 84 years. C and F, Standardized cohort-mortality ratios of rectum and colon cancer plotted against the year of birth and stratified by W and NW subjects.
Lee BY, Sonnenberg A. Time Trends of Mortality From Colorectal Cancer in the United States: A Birth-Cohort Analysis. JAMA Intern Med. 2013;173(12):1148-1150. doi:10.1001/jamainternmed.2013.656
Author Affiliations: Division of Gastroenterology and Hepatology, Portland Veterans Affairs Medical Center (Dr Lee and Sonnenberg), and Oregon Health and Science University (Dr Lee and Sonnenberg), Portland.
The occurrence of colorectal cancer in the United States is decreasing.1 We sought to determine whether the decrease began prior to increased colorectal cancer screening.
Data on US colorectal cancer mortality from 1931 through 2007 were obtained from the Centers for Disease Control and Prevention. Age-specific death rates from rectum and colon cancer per 100 000 living population were calculated for consecutive 10-year periods and 10-year age groups. For instance, the total number of deaths from 1980 until 1989 among subjects aged 55 to 64 years was divided by the corresponding number of total US population of the same age group and living during the same time period when the deaths occurred. The age-specific death rates were plotted against the period of death as period-age contours and against the period of birth as cohort-age contours. An age-standardized cohort mortality ratio (SCMR) was used as a summary statistic of the overall mortality associated with each consecutive birth cohort.2 The SCMR was calculated similarly to the standardized mortality ratio (SMR) as the ratio of observed over expected number of deaths in each birth cohort. The individual SCMR value indicates to what extent the SCMR associated with an individual birth cohort lies above or falls below the overall average ratio of 100%.
The period-age contours of rectum and colon cancer show a concurrent rise and fall among old and young age groups, respectively (graphs not shown). In general, such divergent patterns among period-age contours of consecutive age groups are highly suggestive of an underlying birth-cohort phenomenon.3 To delineate such a cohort phenomenon, the Figure shows age-specific death rates of rectum cancer (upper panels) and colon cancer (lower panels) replotted against the period of birth as cohort-age contours. In both cancer types, the individual cohort-age contours align in patterns that resemble hyperbolas with an initial rise and subsequent decline associated with consecutive periods of birth. The oldest age groups participated in the initial rise and parts of the subsequent decline. The younger age groups contributed mostly to the recent decline. The location of the peak associated with the highest mortality from rectum or colon cancer appeared to occur before the turn of the century (1900s). Overall, the alignment of all the individual cohort-age contours to form a single overarching pattern suggests that risk of death from colon cancer was strongly associated with the period of birth.
The 2 panels on the right of the Figure graphically summarize the individual cohort-age contours from the 4 panels on the left as time trends of the corresponding age-standardized cohort mortality ratios. Every point of the SCMR curves represents an approximation of the average death rate among individuals belonging to different age groups but being born during the same time period. In appreciating the SCMR curves, one should focus more on their temporal behavior and peaks location than on the actual heights reached by the individual curves. As evidenced by the SCMR curves, mortality from rectum and colon cancer increased among consecutive generations born in the 19th century and then declined in all subsequent generations during the 20th century. The rise and fall were statistically significant in all 4 subgroups of white and nonwhite persons with rectum and colon cancer.
Current screening efforts against colorectal cancer were introduced in time such that subjects born around the mid 20th century are the first to benefit from such screening. Our birth-cohort analysis clearly demonstrates that mortality from colorectal cancer already started to decline among subjects born 50 or more years earlier. Cohort effects are generally caused by risk exposure during early childhood that influences subsequent disease behavior throughout life. The acquisition of Helicobacter pylori infection during early childhood with the ensuing risk for future development of peptic ulcer or gastric cancer represents a typical example for a cohort effect in digestive diseases.2 Infection with H pylori is also a known risk factor for the development colonic neoplasms.4- 6 The consistency of the positive associations between H pylori and colorectal neoplasms, as evidenced by multiple case-control studies, and the similarity in their birth-cohort patterns both suggest that exposure to H pylori may have influenced the long-term time trends of colorectal cancer in the United States.
The present epidemiologic analysis reveals that the time trends of mortality of colorectal cancer in the United States have been shaped by an underlying birth-cohort pattern. Americans born around the beginning of 20th century were the ones most likely to develop and die from colorectal cancer. The risk of colorectal cancer has continuously dropped in all subsequent generations. Although there is good evidence to show that various screening techniques protect against colorectal cancer, a detailed analysis of the long-term trends of colorectal cancer clearly demonstrates that the decline of colorectal cancer in the United States started long before any of the various screening techniques became widely used.
Correspondence: Dr Sonnenberg, Division of Gastroenterology, Portland VA Medical Center, P3-GI, 3710 SW US Veterans Hospital Rd, Portland, OR 97239 (firstname.lastname@example.org).
Published Online: April 29, 2013. doi:10.1001/jamainternmed.2013.656
Author Contributions:Study concept and design: Sonnenberg. Acquisition of data: Lee. Analysis and interpretation of data: Lee. Drafting of the manuscript: Lee. Critical revision of the manuscript for important intellectual content: Lee and Sonnenberg. Statistical analysis: Sonnenberg.
Conflict of Interest Disclosures: Dr Sonnenberg was supported by a grant from Takeda Pharmaceuticals. No funding was obtained for the present study.