AbstractBackgroundUltra‐processed foods (UPF) are ready‐to‐eat/heat foods that derive from a series of industrial processes. These products constitute a large and increasing amount of the food supply in high and middle‐income countries. Accumulating evidence point to an association of UPF intake with increased mortality and adverse health outcomes, including cardiovascular and metabolic diseases. However, whether UPF consumption is associated with cognitive function is not clear. We investigated the associations between UPF intake and cognitive decline, in a cohort of older adults at high‐risk for dementia due to type‐2 diabetes (T2D).MethodThe study sample included initially non‐demented T2D older adults (≥65y), from the Israel Diabetes and Cognitive Decline (IDCD) study, who had complete information on nutrition at baseline and at least three cognitive assessments (mean follow‐up 5.3±1.5y). Nutritional intake was evaluated by a validated Food‐Frequency Questionnaire and foods were categorized as UPF based on NOVA classification. Percent of calories from UPF were calculated from total caloric consumption, first as overall UPF intake and then for specific food groups. Global cognitive functioning and four cognitive domains (episodic memory, working memory/attention, semantic categorization, and executive functions) were assessed. Mixed effect models were used to examine the link between UPF intake (top vs. bottom quartiles) and change in cognitive function while adjusting for age, sex, education, smoking, physical activity, BMI, HbA1C, duration of diabetes, total cholesterol and blood pressure.ResultOf the total sample (N=578; mean age at baseline 71.3±4.5y, 61% men), 144 (top quartile) consumed >31% kcal from UPF out of total kcal. Total UPF consumption was not related to cognitive function. However, consumption of ultra‐processed fat (e.g. mayonnaise) was associated with faster decline in global cognition, executive function and memory (ß=‐0.028±0.010; p=0.009, ß=‐0.031±0.014; p=0.022 and ß=‐0.035±0.017; p=0.046, respectively). Additionally, increased intake of ultra‐processed meat (e.g. sausages) was associated with a faster decline in executive functions (ß=‐0.029±0.013; p=0.031). UPF derived from dairy products and carbohydrates were not associated with cognitive change.ConclusionThis pioneering study suggests a possible association of ultra‐processed meat and fat intake with accelerated cognitive decline in older individuals with T2D, independently of sociodemographic, lifestyle, cardiovascular and T2D‐related factors.