Objective To investigate the long-term association between four dietary quality indices and the risk of gastrointestinal (GI) cancer. Methods Baseline details of the dietary intake of participants, assessed by a single food frequency questionnaire from the prospective Lifelines population-based cohort were translated to diet quality scores using several dietary and dietary-lifestyle indices. Incident cases of GI cancer were then assessed by linkage to the Dutch nationwide histo-cytopathology registry. The association between GI cancer risk and diet quality (defined as higher quintiles on dietary indices compared to the first quintile) was assessed by multivariable Cox proportional hazard models. Results We included 72,695 participants aged 51.20 ± 8.71 years with a median follow-up to cancer diagnosis of 8 years (interquartile range 2 years). During follow-up, 434 colorectal cancers and 139 upper GI cancers were diagnosed. There was a significant reduction in colorectal cancer risk for high categories in the American Cancer Society (ACS) Index (hazard ratio 0.62; 95% CI 0.46–0.84). However, high dietary index scores were not associated with strong beneficial effects on upper GI cancer risk. Conclusion High quintiles on the ACS Index were associated with a significantly reduced risk of colorectal cancer. This index may be of use in a colorectal cancer prevention program.

Purpose To compare the differences in lifestyle behaviours between cancer survivors (CSs) and cancer-free participants in a large and representative population-based cohort. Methods We included 115,257 adults from the Lifelines cohort. Cancer status was self-reported, and health behaviours were measured (e.g. body mass index [BMI]) or assessed by questionnaire (e.g. physical activity, smoking, alcohol consumption, sedentary behaviour and diet). The data were then categorised for logistic regression analysis, stratified and adjusted by sex and age (< 55 vs ≥ 55 years). Results CSs (5473; 4.7%) were diagnosed 9 ± 8.5 years before data collection, were older (mean age 55.4 vs 44.4 years) and more often female (66.6% vs 33.4%) than the cancer-free participants. They were also more likely to be physically active and to have a better diet, and also less likely to be alcohol drinkers; but, were more likely to have a higher BMI, be former smokers and to be sedentary. After adjustment for sex and age, however, BMI was more likely to be normal, physical activity was more likely to be higher and smoking to be prevalent in CSs. Current smoking was also significantly higher among females and those aged < 55 years who were CSs than for those with no history of cancer. Conclusions In this population-based cohort, CSs have health behaviour comparable to those without a cancer diagnosis. Implications for cancer survivors Smoking cessation strategies should target all CSs, but efforts could yield greatest benefit if they target females and those younger than 55 years.

Diet quality among short- and long-term gastrointestinal (GI) cancer survivors with different tumor sites was investigated compared to a reference population cohort. Diet quality of GI cancer survivors (n = 307) was compared to an age- and sex-matched reference population with no history of cancer (n = 3070). All were selected from Lifelines, a population-based cohort. GI cancers were defined as having a history of cancer of the bowel, esophagus, or stomach. Diet quality was assessed by a self-administrated food frequency questionnaire in terms of: (i) Lifelines Diet (LLD) scores, where higher scores indicate higher diet quality; (ii) the adherence to dietary guidelines, quantified by the percentage of meeting dietary recommendations, as given by Dutch dietary guidelines; and (iii) the mean daily intake of food components. All analyses were adjusted for lifestyle factors. Diet scores in GI cancer survivors were not different from the reference population (OR = 0.97, 95% CI: 0.73–1.23). Stratification for time since diagnosis and tumor site gave similar results. The intake of vegetables, unsweetened dairies, and nuts and legumes was almost 50% lower than the recommended amount, and the mean intake of unhealthy food components was at least one serving/day among GI cancer survivors, as well as in the reference population. In the long run, GI cancer survivors do not differ from the reference population in their diet quality. In conclusion, both groups can improve their diet quality.