Patients with colorectal cancer who had a history of cannabis use disorder faced a significantly higher risk of mortality within 5 years compared with those who didn't have the disorder, according to new research.

Investigators examined the health records of 1,088 patients across six University of California medical centers between 2012 and 2024. They found that 34 patients had a documented diagnosis of cannabis use disorder (CUD) prior to their cancer diagnosis.

“Patients with a history of CUD (n = 34) had a substantially higher 5-year mortality rate (55.88%) compared to patients without CUD (5.05%),” said study author Raphael E. Cuomo, of the University of California, San Diego School of Medicine.

Statistical models showed that the risk of mortality within 5 years was higher among the patients with CUD. In unadjusted analysis, the odds ratio (OR) was 24.40 (95% confidence interval [CI] = 11.39–52.34, P < .001). After adjusting for age, gender, and disease severity using carcinoembryonic antigen (CEA) levels, the adjusted OR remained elevated at 10.52 (95% CI = 5.76–19.22, P < .001).

Cox proportional hazards modeling yielded similar results. The unadjusted hazard ratio (HR) was 10.52, and after adjustment for CEA, the HR was 6.39 (95% CI = 2.85–14.31, P < .001). When cancer stage was used instead of the CEA as a measure of disease severity, the association was weaker and not statistically significant.

The data were sourced from the University of California Health Data Warehouse, which stores standardized electronic health records. The investigators verified that CUD diagnoses occurred prior to cancer diagnosis to reduce the chance of reverse causality. Patients had a mean age of 59 years, and about 45% were female. Among those with CUD, the median time between diagnosis of CUD and cancer diagnosis was 135 days.

The investigators noted several limitations, including the potential for the misclassification of CUD based on electronic records and the inability to track cannabis use after cancer diagnosis. They also acknowledged possible residual confounding from unmeasured factors such as socioeconomic status, lifestyle, and co-occurring substance use.

“These results suggest that preexisting CUD may have a negative prognostic impact on cancer outcomes, warranting further investigation into potential biological, behavioral, and health care–related mechanisms,” Dr. Cuomo concluded.

The study didn't receive external funding and reported no conflicts of interest.

Source: Annals of Epidemiology