A MEDITERRANEAN diet, with or without calorie restriction, improved dietary adherence and modestly reduced weight in older African American adults but did not produce long-term cognitive improvements, according to findings from a randomized controlled trial.
The Building Research in Diet and Cognition (BRIDGE) Trial evaluated the effects of an eight-month Mediterranean diet intervention with (MedWL) or without (MedA) a weight loss component in comparison to a control group. The study aimed to assess long-term changes in executive function, memory, attention, learning, and processing speed 6 months after the intervention concluded. Participants included 185 adults (91% African American) with a mean age of 66 years and average body mass index of 37.1 kg/m².
While no significant differences in cognitive outcomes were observed between groups from baseline to 14 months or from the end of intervention to follow-up, notable improvements in dietary adherence and weight were seen. Over the 14-month period, adherence to the Mediterranean diet increased significantly more in both intervention groups, by 3.2 points in MedWL and 3.4 in MedA, compared to just 0.2 points in controls. Additionally, the MedWL group achieved a greater mean weight loss of 3.8 kg, compared to 1.7 kg in the MedA group and 0.5 kg in controls (p 0<0.05). These results indicate that while cognitive benefits did not emerge, sustained lifestyle changes are possible. The trial demonstrates that older African American adults can successfully maintain dietary improvements and weight loss beyond the structured intervention period. The study adds important evidence to the limited body of research on long-term lifestyle interventions in diverse populations, highlighting that dietary adherence and weight outcomes may persist even when cognitive benefits do not. Reference McLeod A et al. Long-term outcomes of a Mediterranean diet with or without calorie restriction on cognition and cardiometabolic health: The building research in diet and cognition trial. Prev Med Rep. 2025;55:103099.
