Exploring time trends in cancer incidence

Area of Science:

• Epidemiology
• Oncology
• Biostatistics

Background:

• Cancer incidence trends are crucial for public health planning and resource allocation.
• Understanding temporal changes in cancer rates by age and sex can reveal etiological factors and the impact of interventions.

Purpose of the Study:

• To analyze time-trends in the incidence of lung, stomach, intestinal, prostate, and breast cancer among White individuals in the United States from 1973 to 1987.
• To model cancer incidence using a log-linear function of diagnosis year to summarize trends and allow for extrapolation.
• To examine how these time trends vary across different age groups and sexes.

Main Methods:

• Log-linear regression modeling of cancer incidence against diagnosis year for specific age and sex groups.
• Analysis of cancer diagnosis data for White individuals in the United States between 1973 and 1987.
• Comparison of model results with nonparametric estimates to assess model performance.

Main Results:

• Stomach cancer incidence decreased by approximately 2% annually.
• Lung cancer incidence showed a decreasing trend in younger individuals (<40) and an increasing trend in older individuals (up to 8% in women aged 80).
• Prostate cancer incidence increase slowed with age, from ~6% in men <40 to ~2% in men >80. Breast cancer incidence increased 4-6% annually between 1980-1987 after a 1974 surge.

Conclusions:

• Cancer incidence time-trends exhibit complex age- and sex-specific patterns.
• The log-linear model effectively summarizes trends for most cancers studied, except breast cancer.
• Observed trends highlight the dynamic nature of cancer epidemiology and the need for continued monitoring.