Ultra-processed food consumption is associated with the acceleration of biological aging in the Moli-sani Study.

Background: Healthy diets have been inversely associated with biological aging. However, the nutritional content is only one aspect of the overall food health potential, and more recently, increasing attention has been paid to nonnutrient food characteristics, such as food processing.

Objectives: To examine the association of ultra-processed food (UPF) consumption with biological aging measured by circulating blood biomarkers.

Methods: Cross-sectional analyses were conducted on 22,495 participants enrolled in the Moli-sani Study (2005-2010, Italy). Food intake was assessed by a 188-item food frequency questionnaire. UPF was defined according to the Nova classification and calculated as the ratio (weight ratio; %) between UPF (g/d) and total food eaten (g/d). Diet quality was assessed by the Mediterranean Diet Score (MDS; ranging 0-9). A deep neural network approach based on 36 circulating biomarkers was used to compute biological age (BA), and the resulting difference (Δage = BA - chronological age)-an index of biological aging-was tested as dependent variable in multivariable linear regression analyses including known risk factors.

Results: The mean Δage in this population was -0.70 (standard deviation ±7.70) years. In multivariable-adjusted analyses, higher intake of UPF was associated with accelerated Δage [β = 0.34 years; 95% confidence interval (CI): 0.08, 0.61 for the last fifth compared with the first). This association was not linear (P value for overall association <0.001; P value for nonlinearity = 0.049). Inclusion of the MDS into the model slightly attenuated this association by 9.1% (β = 0.31 years; 95% CI: 0.04, 0.59).

Conclusions: A diet rich in UPF was associated with an acceleration of biological aging in a large sample of Italian adults. The poor nutritional composition of highly processed foods weakly accounted for this association, suggesting that biological aging could be adversely influenced by nonnutrient characteristics of these foods.