(030) The impact of Body Mass Index on colorectal cancer prognosis. A population-based cohort study in Denmark.

Background: Colorectal cancer (CRC) prognosis may be adversely affected by obesity, but evidence is limited. Obesity is related to diabetes, socioeconomic position (SEP), and lifestyle factors, which may also influence CRC prognosis.

Objectives: To evaluate the association of Body Mass Index (BMI) on CRC recurrence and mortality accounting for diabetes, cancer site and stage, SEP, and lifestyle factors.

Methods: We conducted a Danish population-based cohort study of stage I-III CRC patients who underwent curative intent surgery during 2002-2020. We ascertained data on BMI, tumor characteristics, surgery, comorbidities, SEP, lifestyle factors, and deaths from the Danish Colorectal Cancer Groups' clinical database and population-based registries. We used a validated algorithm to identify CRC recurrence. Stratifying by sex, we calculated 5-year cumulative incidences functions (CIF), age and calendar-year standardized incidence/mortality rates (SIR/SMR) per 100 person-years and used Cox regression to compute hazard ratios (HR) with 95% confidence intervals (95%CI) of CRC recurrence and mortality, using standard BMI categories and modelling BMI via spline functions. Models were adjusted for age, calendar-year, education, cohabitation, smoking and alcohol. Via stratification and interaction analyses, we examined the synergistic effects of cancer site, stage, diabetes, smoking, alcohol, comorbidities, education, and cohabitation status.

Results: Among 33,926 CRC patients (median [IQR] age 70 years [62-77], 66.4% colon cancer, 45.5% women), 5,908 developed CRC recurrence (CIF=19.8%), 6,662 died (CIF=40.1%). Recurrence rates were highest among people with underweight (BMI <18.5, SIR=5.99) and lowest for overweight (BMI 25-29.9, SIR=5.02), with little difference across strata. Underweight was associated with increased mortality in men (SMR 14.61, HR=1.84, 95%CI=1.46-2.34) and women (SMR 10.25, HR=1.71, 95%CI=1.49-1.98) compared with normal weight (BMI 18-24.9, SMRs 6.64 and 5.33, respectively). Mortality was lower in men with overweight (SMR 5.78, HR=0.88, 95%CI=0.82-0.95) and women (SMR 4.68, HR=0.89, 95%CI=0.81-0.98), whereas obesity class ll-lll (BMI >35) was associated with increased mortality in men (SMR 7.16, HR=1.27, 95%CI=1.08-1.51), but not women (SMR 4.69, HR=1.08, 95%CI=0.88-1.31). This was most evident in men with rectal cancer (SMR 9.22, HR=1.70, 95%CI 1.28-2.25). Mortality did not differ between normal weight and obesity class ll-lll in men with short education, mainly due to higher mortality in men with normal weight than in those with longer education. Spline regression models provided similar results.

Conclusions: We observed no clinically meaningful association between BMI and CRC recurrence. Yet, we observed a reverse J-shaped association between BMI and mortality. Regardless of subgroup, having underweight at CRC surgery was associated with increased mortality. For men, obesity class ll-lll was associated with excess mortality, especially for rectal cancer.