Wednesday, August 19, 2015

Coffee Intake, Recurrence, and Mortality in Stage III Colon Cancer: Results From CALGB 89803 (Alliance)

Here is a link to an article about coffee and colon cancer in the Journal of Clinical Oncology.  The article's conclusion reads:

Higher coffee intake may be associated with significantly reduced cancer recurrence and death in patients with stage III colon cancer.

The abstract suggests that the "may be" may be "may not be". :-)

Here is the abstract.

Observational studies have demonstrated increased colon cancer recurrence in states of relative hyperinsulinemia, including sedentary lifestyle, obesity, and increased dietary glycemic load. Greater coffee consumption has been associated with decreased risk of type 2 diabetes and increased insulin sensitivity. The effect of coffee on colon cancer recurrence and survival is unknown.

Patients and Methods
During and 6 months after adjuvant chemotherapy, 953 patients with stage III colon cancer prospectively reported dietary intake of caffeinated coffee, decaffeinated coffee, and nonherbal tea, as well as 128 other items. We examined the influence of coffee, nonherbal tea, and caffeine on cancer recurrence and mortality using Cox proportional hazards regression.

Patients consuming 4 cups/d or more of total coffee experienced an adjusted hazard ratio (HR) for colon cancer recurrence or mortality of 0.58 (95% CI, 0.34 to 0.99), compared with never drinkers (Ptrend = .002). Patients consuming 4 cups/d or more of caffeinated coffee experienced significantly reduced cancer recurrence or mortality risk compared with abstainers (HR, 0.48; 95% CI, 0.25 to 0.91; Ptrend = .002), and increasing caffeine intake also conferred a significant reduction in cancer recurrence or mortality (HR, 0.66 across extreme quintiles; 95% CI, 0.47 to 0.93; Ptrend = .006). Nonherbal tea and decaffeinated coffee were not associated with patient outcome. The association of total coffee intake with improved outcomes seemed consistent across other predictors of cancer recurrence and mortality.

Higher coffee intake may be associated with significantly reduced cancer recurrence and death in patients with stage III colon cancer.

Here are some reasons that strike me as justifying skepticism.

The study is a prospective study with no controls.  It is possible that patients who are doing the best and feeling the best drink more coffee.  In that case the causality runs from health to coffee, not from coffee to health.

131 items were examined for a possible relationship.  In a 131 variable regression, the probability that at least one variable will have a statistically significant estimated relationship with the dependent variable is high, even if the there is really no relationship between any of the 131 independent variables and the dependent variable.  The study quotes measures of statistical significance that appear to be unadjusted for this.  If so, they are wildly overstated.

No comments: