Consumption of Ultra-Processed Food Is Independently Associated with Premature Mortality in Cancer Survivors: A Prospective Analysis from the Moli-sani Study in Italy ^†

Background and objectives: There is poor knowledge on the role of diet in relation to mortality among cancer survivors. Available studies have mainly focused on diet quality, regardless of food processing, a well-described risk factor in numerous cohorts worldwide. We therefore examined the association of food processing with all-cause mortality in a sample of cancer survivors from the Moli-sani Study (2005–2010, Italy) and analysed biological pathways potentially underpinning these associations. Methods: Longitudinal analysis was performed on 799 men and women (mean age 63 ± 12 y) with a history of cancer at enrolment, followed for 11.8 y (median). Food intake was assessed using a 188-item FFQ. The Nova classification was used to categorize foods according to increasing levels of processing: (1) minimally processed food (e.g., fruits, meat); (2) culinary ingredients (e.g., butter, sugar); (3) processed food (e.g., canned fish, bread); (4) ultra-processed food (UPF; e.g., carbonated drinks, processed meat). We then calculated the proportion (%) of each Nova group on the total weight of food (g/d) by creating a weight ratio. The modified Food Standards Agency Nutrient Profiling System (FSAm-NPS) dietary index was used to assess overall diet quality. Shared biological risk factors for chronic diseases were analysed as potential mediators through change-in-estimate method. Results: In multivariable-adjusted COX analysis controlled for known risk factors and diet quality, a 5% increment of UPF intake in the diet was associated with 14% increased risk of premature mortality (HR = 1.14; 95%CI 1.01–1.29; p = 0.03), independent of diet quality; HRs associated with higher intakes of either unprocessed/minimally processed food, or culinary ingredients or processed food were, respectively, 0.94 (0.88–1.01; p = 0.10), 0.90 (0.75–1.06; p = 0.21 for 1% increment), and 1.02 (0.95–1.09; p = 0.56). Serum C-reactive protein levels and resting heart rate accounted together for 58% (p value < 0.01) of the association of UPF with mortality. Discussion: In a general adult population, increasing the dietary share of UPF was associated with higher risk of premature death among cancer survivors, independent of diet quality. This association was largely explained by altered levels of inflammation and resting heart rate. Further large cohorts are warranted to possibly confirm these findings and extend knowledge on the biological mechanisms underpinning these associations.