Targeting food addiction through an intensive energy-reduced Mediterranean diet lifestyle intervention: evidence from the 3-year PREDIMED-plus trial

Food addiction (FA) has been proposed as a behavioural phenotype linked to obesity, potentially influencing adherence and response to lifestyle interventions. Its role within long-term, multicomponent weight-loss programmes remains unclear. This study evaluated the impact of an intensive lifestyle intervention based on an energy-reduced Mediterranean diet (MedDiet) on FA symptomatology and examined whether baseline FA predicts adherence and weight-loss outcomes. A total of 448 older adults (mean age = 65.3 ± 4.6 years) with overweight/obesity and metabolic syndrome from the PREDIMED-Plus Cognition sub-study were randomised to an intensive intervention (IIG; energy-reduced MedDiet, physical activity, behavioural support) or a control group (CG; ad libitum MedDiet). FA (measured using the Yale Food Addiction Scale 2.0), anthropometric measures, MedDiet and physical activity adherence, and quality of life (QoL) were assessed at baseline, 1- and 3-years follow-up. Longitudinal changes were analysed using repeated-measures ANOVA. Logistic and Cox regression models evaluated predictors and time-to-achievement of clinically meaningful outcomes. Structural equation modelling (SEM) explored 3-year pathways linking anthropometric measures, MedDiet adherence, QoL, and intervention group according to FA status. FA severity decreased in both groups, but reductions were sustained only in the IIG, which showed lower FA scores than the CG at 3 years ( p = 0.011). The IIG increased the likelihood of achieving clinically meaningful reductions in weight and waist circumference, particularly among participants without FA at baseline (≥ 8% weight loss: OR = 2.87, 95%CI 1.99–4.14; ≥5% waist circumference reduction: OR = 3.01, 95%CI 2.20–4.11). Improvements in MedDiet adherence were markedly greater in the IIG, with effects strongest among individuals without FA (OR = 5.82, p < 0.001). SEM indicated that adherence mediated long-term improvements in weight, waist circumference, and QoL only in individuals without FA. An intensive energy-reduced MedDiet-based intervention improves FA symptoms and cardiometabolic outcomes. However, baseline FA appears to limit adherence and attenuates long-term benefits. FA may represent a clinically relevant phenotype for personalising obesity treatment and enhancing intervention effectiveness. International Standard Randomized Controlled Trial Number (ISRCTN): 89,898,870. Registration date: 24 July 2014.