Omnivorous and plant-based dietary patterns: a comparative analysis using data-driven and index-based approaches

Background and objective Plant-based diet (PBD) followers are growing worldwide. The aim was to explore the dietary profile of PBDs, taking the omnivorous (OMN) diet as a reference. Methods A cross-sectional study was carried out. A total of 760 participants were included, of whom 161 self-identified as vegan, ovo-lacto-vegetarian, or pesco-vegetarian. Intakes of 175 foods, assessed through a food frequency questionnaire (FFQ), were adjusted for energy intake. The intake of 32 food groups was compared across diet types using Kruskal–Wallis and post-hoc tests. Adherence to established a priori dietary patterns (Pro-vegetarian, Mediterranean, and EAT-Lancet) diet was assessed. A posteriori dietary patterns were characterized using principal component analysis (PCA), k-means and hierarchical clustering, amongst others. Results Per 1000 kcal, daily mean intakes of legumes (> 48.2 g), vegetables/mushrooms (> 209.9 g), nuts (> 10.2 g), and whole grains (> 24.7 g) were higher in the PBD groups than OMN ( p < 0.001). The three a priori dietary indices were inter-correlated (rho = 0.4–0.6) and OMN scored lower than the PBD groups ( p < 0.001). PCA revealed a PBD-like pattern (variance: 15.0%), an unhealthy (7.2%), and a mixed animal-PBD pattern (5.5%). The clustering methods identified similar groupings of the PBD participants, and a group of PBD-like OMN (N = 61). The latter, compared with the other OMN patterns, showed significantly lower intakes of milk (48.0 vs. > 95.2 ml), red meat (13.9 vs. > 17.8 g), processed meat (9.0 vs. > 14.0 g), and precooked foods (9.0 vs. > 12.2 g), alongside higher intakes of whole grains (25.5 vs. < 20.8 g), nuts (13.8 vs. < 8.3 g), and vegetables (207.7 vs. < 161.3 g). Conclusions Thus, food choices between OMN, vegetarians and vegans differ, beyond the exclusion of animal foods. PBDs present common characteristics despite their diversity, and some OMN share these dietary features.