Prospective study of dietary changes in cancer survivors for five years including pre- and post- diagnosis compared with those in cancer-free participants.

The number of long-term survivors after a cancer diagnosis is increasing. Few investigations have compared survivors' diets to their original pre-diagnosis dietary pattern or with the patterns of cancer-free controls. We examined the dietary changes in survivors for five years (i.e. before to after diagnosis) in cancer survivors, comparing them with cancer-free controls in a prospective cohort study in Japan. Using 1995-1998 for the baseline and 2000-2003 for the follow-up survey, a validated food frequency questionnaire was administered to 33,643 men and 39,549 women aged 45-74 years. During the follow-up period, 886 men and 646 women had developed cancer. Participants that had not been diagnosed with cancer served as controls. There was a greater decrease in the calorie intake (median change: - 168 kcal/d [Interquartile range: - 640, 278]) in male cancer survivors compared to controls (- 33 kcal/d [- 453, 380], P < .001). On comparison with cancer-free controls, multiple linear regression analysis revealed a significantly larger reduction in energy-adjusted ethanol intake for male cancer survivors (β =  - 0.36). There was no difference in changes in fruit and vegetable or red meat intake and no other significant differences in dietary changes between survivors and controls for either gender. This suggests that most dietary changes in survivors after cancer diagnosis are not systematically different from those that occur in people without a cancer diagnosis.