Variation in the Relationship Between Past Cancer History and Cognitive Performance by Mediterranean Diet Adherence

Objectives: To determine whether the association between past cancer history and cognitive performance varies by Mediterranean Diet (Med Diet) adherence. Methods: We used National Health and Nutrition Examination Survey data from 1999-2002 and 2011-2014 and adjusted linear regression, with an interaction term (CancerHx*Med Diet) to estimate whether the association between past cancer history and cognitive performance varied by Med Diet adherence. Med Diet adherence was defined using the alternative Mediterranean Diet (aMed) score which gives 1 point for intake above the sample-specific median for food groups highly abundant in a Med Diet, such as fruits and vegetables, and 1 point for intake below the sample-specific median for food groups in low abundance in a Med Diet, such as red and processed meats (range 0-9). aMed score was split at the median into low and high adherence. Cognitive performance was measured as the number correct on the Digit Symbol Substitution Test (DSST), which assesses processing speed, attention, and working memory. Models were adjusted for demographics (e.g., age, race, gender), comorbidity burden, smoking status, total kilocalories, day of recall, time of year of recall, and testing environment. Results: Among the 4,869 participants who had dietary and cognitive data, 733 were cancer survivors. Participants mean age was 69.5 years (SE=0.2), and the most common cancer reported was prostate (24%). The average aMed score was 3.7 (SE=0.1) in survivors and 3.6 (SE=0.1) in participants with no cancer history. In stratified analyses, past cancer history was less associated with DSST score among those with low adherence to a Med Diet (ß= 0.8, 95%CI: -1.2, 2.9) compared to those with high adherence (ß= -1.4, 95%CI: -2.8, -0.005). Conclusions: Past cancer history was not associated with cognitive performance among those with low adherence to a Med Diet but among those with high adherence, past cancer history was negatively associated with cognition. Prospective studies should be conducted to confirm these results. Funding Sources: National Cancer Institute.