Abstract Body: Introduction: Mild cognitive impairment (MCI) is a major public health challenge that affects up to 18% of individuals aged 60 or older. Prior studies have shown that soy and milk protein supplementation reduces blood pressure compared to carbohydrate. Given that hypertension is a major risk factor for dementia, it is important to investigate whether these benefits from protein supplementation translate to cognitive improvements. Additionally, evidence suggests that replacing ultra-processed foods with whole foods, particularly those containing soy protein, may enhance cognitive outcomes. We aimed to evaluate the impact of carbohydrate, soy protein, and milk protein supplementation on MCI risk, as indicated by a validated metabolomic risk score (MRS), in the Protein and Blood Pressure (ProBP) trial.
Hypothesis: We hypothesize that the three dietary supplementations, particularly soy protein, would attenuate the MRS.
Methods: The ProBP study was a double-blind, crossover randomized controlled trial involving 80 participants. Each participant received 40 grams daily of carbohydrate, soy protein, or milk protein supplementation for eight weeks in random order, with three-week washout periods between interventions. Untargeted metabolomic profiling was conducted at baseline and post-intervention to calculate the MRS. We used paired t-tests to assess net changes in the risk score after each intervention, and a mixed-effects model to compare the impacts of the three dietary interventions.
Results: All three interventions significantly reduced the MRS for MCI (Table). Soy protein demonstrated the largest effect (net change = -0.207; P = 0.006), followed by milk protein (net change = -0.182; P = 0.008) and carbohydrate (net change = -0.137; P = 0.038). However, the effects of the protein supplementation were not significantly different from carbohydrate.
Conclusion: Soy protein, milk protein, and carbohydrate supplementation all significantly reduced the MRS for MCI, with soy protein showing the greatest effect.