Adherence to Mediterranean Diet and physical activity practice in general population: an intersectional analysis of inequalities by sex and economic status

Glòria Reig-Garcia¹Joan Martínez-Sancho²Alícia Baltasar-Bagué¹Glòria Mateu-Figueras¹Maria Buxo Pujolras²Ruth Martí-Lluch³Lluís Zacarías-Pons³Josep Puig⁴Rafel Ramos³Jose Manuel Fernández-Real^1,2Josep Garre-Olmo^1*

• ¹University of Girona, Girona, Spain
• ²Institut d'Investigacio Biomedica de Girona Dr Josep Trueta, Girona, Spain
• ³Institut d'Investigacio en Atencio Primaria de Salut Jordi Gol, Barcelona, Spain
• ⁴Hospital Clinic de Barcelona, Barcelona, Spain

Background: Inequalities in physical activity (PA) practice and adherence to Mediterranean diet (MD) between sex or socioeconomic groups have been described. We analyzed inequalities using an intersectional approach combining sex and economic difficulties (ED) in a population-based adult sample from Girona (Spain). Methods: Cross-sectional study including 1,425 inhabitants. ED were assessed with a self-reported question on perceived ability to make ends meet and dichotomized (with/without ED). Sex and ED were employed to form four intersectional positions (men/women×with/without ED). PA was assessed using the Minnesota-Regicor questionnaire, total weekly metabolic equivalents of task were calculated, and PA was dichotomized as low (first quartile) and not low. Adherence to MD was measured with the PREDIMED questionnaire and dichotomized as non-adherence/adherence. Sociodemographic and lifestyle characteristics were registered. We estimated crude and adjusted prevalence differences in low PA practice and non-adherence to MD corresponding to joint, referent, and excess intersectional inequalities using linear binomial regression models with identity link, which directly estimate prevalence differences in percentage points (pp). The discriminatory accuracy of the models was assessed using the area under the receiver operating characteristic curve (AUC-ROC). Results: The prevalence of low PA was higher among women and increased by the presence of ED (26.7% vs. 40.9%). Overall, 27.8% reported ED and the four intersectional positions ranged from 10.5% to 40.3%. The adjusted joint disparity in low PA between women with ED (doubly disadvantaged) and men without (doubly advantaged) was 20.8pp, of which 11.6pp represented excess intersectional inequality. For non-adherence to MD, only the referent disparity for sex was statistically significant (8.54pp). The AUC-ROC ranged from 0.62 for low PA to 0.67 for nonadherence to MD, with small increases when adding the intersectional groups. Conclusion: In this study, inequalities in adherence to MD were modest, mainly by sex and with low discriminatory accuracy, suggesting that population-wide interventions, complemented by attention to sex differences, may be more appropriate than targeted strategies. In contrast, low PA was markedly higher in women with ED than in men without. This highlights the importance of intersectional inequalities in PA and the need for universal policies that also prioritize socioeconomically disadvantaged women.