From a public health perspective, determinants of diets are crucial to identify, but they remain unclear in Switzerland. Hence, we sought to define current dietary patterns and their sociodemographic and lifestyle determinants using the national nutrition survey menuCH (2014⁻2015, n = 2057). First, we applied multiple factorial analysis and hierarchical clustering on the energy-standardised daily consumption of 17 food categories. Four dietary patterns were identified ("Swiss traditional": high intakes of dairy products and chocolate, n = 744; "Western 1": soft drinks and meat, n = 383; "Western 2": alcohol, meat and starchy, n = 444; and "Prudent": n = 486). Second, we used multinomial logistic regression to examine the determinants of the four dietary patterns: ten sociodemographic or lifestyle factors (sex, age, body mass index, language region, nationality, marital status, income, physical activity, smoking status, and being on a weight-loss diet) were significantly associated with the dietary patterns. Notably, belonging to the French- and Italian-speaking regions of Switzerland increased the odds of following a "Prudent" diet (Odds ratio [95% confidence interval]: 1.92 [1.45⁻2.53] and 1.68 [0.98⁻2.90], respectively) compared to the German-speaking regions. Our findings highlight the influence of sociodemographic and lifestyle parameters on diet and the particularities of the language regions of Switzerland. These results provide the basis for public health interventions targeted for population subgroups.
Keywords: 24-h recall; clustering; dietary survey; language region; multinomial logistic regression.