Background: Colorectal cancer is the third most commonly diagnosed cancer in the United States, and the number of survivors is growing. Current lifestyle recommendations for survivors are largely based on studies of cancer incidence, so research into modifiable factors possibly impacting survival is a public health imperative. Therefore, we examined relationships of recommended lifestyle behaviors to five-year total and cancer-specific mortality in a cohort of colorectal cancer cases.

Methods: Prediagnostic lifestyle factors were assessed at baseline in 1995-1996 in the NIH-AARP Diet and Health study. Cancer incidence and first course of therapy were determined by linkage to state cancer registries. A total of 6,318 incident colorectal cancer cases were identified between 1996 and 2006 (mean age=69; 68% female). Date and cause of death were determined by linkage to the National Death Index through 2008. Lifestyle factors examined included healthy diet (upper two quintiles of Healthy Eating Index 2005; HEI 2005), normal body mass index (BMI; 18.5-25 kg/m2), physical activity (≥20 minutes moderate to vigorous activity ≥ 3 times/week), moderate alcohol intake (≥2 drinks/day for men, ≥1 drink/day for women), and nonsmoking. Factors were examined individually and as a combined lifestyle score in which participants were awarded 1 point for each recommendation met (range 0-5). We estimated relative risks (RRs) and 95% confidence intervals (CIs) of overall and colorectal cancer death using Cox proportional hazards models with age as the underlying time metric. We analyzed colon and rectal cancer separately, and a multivariable model was adjusted for time from baseline to colorectal cancer diagnosis, sex, education, stage at diagnosis, surgery, radiation, chemotherapy, and family history of colon cancer.

Results: Among 4,368 colon cancer cases, 1,322 died within five years of diagnosis, including 862 colorectal cancer deaths, 121 cardiovascular disease (CVD) deaths, 138 deaths from cancers other than colorectal cancer, and 201 deaths from other causes. Compared to smokers, nonsmokers at baseline had about 40% lower risk of total death (0.59; 0.51-0.69) and colorectal cancer death (RR 0.62; 95% CI 0.52-0.74). Compared to normal weight individuals, obese (BMI≥30) individuals at baseline had significantly increased risk of total mortality (1.22; 1.05-1.42), but not colorectal cancer mortality (1.14; 0.95-1.37). Compared to the individuals with the lowest combined score (0-1), those with the highest score (5) had lower risk of total (0.63, 0.48-0.82) and colorectal cancer mortality (0.79, 0.59-1.07).

Among 1,680 rectal cancer survivors, 468 deaths occurred within five years of diagnosis, including 303 colorectal cancer deaths, 45 CVD deaths, 52 deaths from cancers other than colorectal cancer, and 68 deaths from other causes. Compared to the lowest, the highest quintile of HEI 2005 scores was significantly associated the reduced total (0.59; 0.41-0.84) and colorectal cancer mortality (0.65; 0.42-1.00). Higher combined lifestyle scores were also associated with significantly lower risk of total (0.53. 0.32-0.88) and colorectal cancer mortality (0.40, 0.21-0.76).

Conclusion: Several modifiable pre-diagnostic lifestyle factors, alone and in combination, were related to survival among colon and rectal cancer cases. Further studies are warranted to examine the effect of changes in lifestyle factors after diagnosis on colorectal cancer prognosis and survival.

Citation Format: Colleen Pelser, Hannah Arem, Pfeiffer Ruth, Joanne Elena, Catherine Alfano, Albert Hollenbeck, Yikyung Park. Prediagnostic lifestyle factors and five-year survival after colon and rectal cancer diagnosis in the NIH-AARP Diet and Health Study. [abstract]. In: Proceedings of the Eleventh Annual AACR International Conference on Frontiers in Cancer Prevention Research; 2012 Oct 16-19; Anaheim, CA. Philadelphia (PA): AACR; Cancer Prev Res 2012;5(11 Suppl):Abstract nr A89.