Characteristics of the study cohort of 15
058 men (51%) and 14
412 women (49%) are presented in . The mean age at cohort entry, that is, time of first recorded prescription for low-dose aspirin, was 70 years, with only 6% of the cohort being younger than 50 years at entry. The cohort members accrued a total of 121
562 person-years, with an average length of follow-up since the first recorded prescription for low-dose aspirin of 4.1 years (range, 0–9 years); 80% of subjects received two or more prescriptions for low-dose aspirin, and 32% received 10 or more prescriptions.
Characteristics of users of low-dose aspirin recorded in the Prescription Database of North Jutland County, Denmark, between 1989 and 1995
Overall, we observed 2381 cancer cases compared with 2187 expected, yielding a SIR of 1.09 (95% CI, 1.05–1.13) (). No apparent risk reductions were observed for cancers of the gastrointestinal tract, including cancers of the colon (SIR, 0.9; 95% CI, 0.7–1.1), rectum (SIR, 1.0; 95% CI, 0.8–1.2), and stomach (SIR, 1.1; 95% CI, 0.8–1.3). Increased SIRs were found for cancers of the kidney (SIR, 1.4; 95% CI, 1.1–1.7) and brain (SIR, 1.7; 95% CI, 1.3–2.2). Slight risk elevations were seen for several smoking-related cancers, including cancers of the buccal cavity and pharynx (SIR, 1.3; 95% CI, 0.9–1.7), urinary bladder (SIR, 1.2; 95% CI, 1.0–1.4), oesophagus (SIR, 1.3; 95% CI, 0.9–1.9), larynx (SIR, 1.5; 95% CI, 0.9–2.4), and leukaemia (SIR, 1.3; 95% CI, 1.0–1.6), whereas the SIR for lung cancer was close to expectation (SIR, 1.1; 95% CI, 0.9–1.2). The SIR estimates for site-specific cancers were similar in men and women, except for cancer of the kidney (SIR, 1.6 (95% CI, 1.1–2.1) in men vs 1.1 (95% CI, 0.7–1.6) in women) and non-Hodgkin's lymphoma (SIR, 1.4 (95% CI, 1.0–1.9) in men vs 0.9 (95% CI, 0.5–1.4) in women).
Standardised incidence ratios (SIR) and 95% confidence intervals (CI) for cancers at selected sites, stratified by sex,a among users of low-dose aspirin in North Jutland county, Denmark, 1989–1997
The SIRs did not vary markedly when stratified by age. The SIR for total cancer was 1.14 (95% CI, 1.07–1.22) among persons below 70 years at cohort entry and 1.06 (95% CI, 1.01–1.11) among older persons (data not shown). Similarly, for specific cancer sites, including colorectal cancer and kidney cancer, the SIR estimates did not differ substantially between these two age groups.
presents SIR estimates, stratified by number of prescriptions for low-dose aspirin and years of follow-up, for total cancer, selected gastrointestinal cancers, and kidney cancer. Overall, stratification by duration of follow-up yielded SIRs for total cancer of 1.2 (95% CI, 1.1–1.3) for less than 1 year of follow-up, 1.1 (95% CI, 1.0–1.1) for 1–4 years of follow-up, and 1.1 (95% CI, 1.0–1.2) for 5–9 years of follow-up. Similar SIRs were observed for total cancer in the four categories of prescription frequency. The SIRs for colorectal and stomach cancer decreased slightly with both increasing number of prescriptions and increasing years of follow-up, but none of the trends was statistically significant. No consistent trends in risk estimates were observed for cancers of the oesophagus or kidney. Among persons with 10 or more prescriptions, the SIR for kidney cancer was of the same magnitude in the 1- to 4-year (SIR, 1.6; 95% CI, 0.7–3.1) and 5- to 9-year (SIR, 1.7; 95% CI, 0.9–2.9) follow-up periods.
Standardised incidence ratios (SIR) and 95% confidence intervals (CI) for selected cancer sites among users of low-dose aspirin, stratified by number of prescriptions and years of follow-up, North Jutland county, Denmark, 1989–1997
The overall increased SIR for brain cancer was entirely due to an excess in the first year of follow-up (SIR, 4.7; 95% CI, 3.4–6.3) among persons receiving one prescription (SIR, 6.2; 95% CI, 4.1–9.1) or 2–4 prescriptions (SIR, 3.5; 95% CI, 2.0–5.7) (data not shown). For the remaining sites of interest, including cancers of the lung, breast, prostate, bladder, and ovary, stratification by number of prescriptions and years of follow-up revealed no consistent trends (data not shown).