🟠 Moderate Evidence
Cognitive decline in older adults involves far more than chronological age: vascular health, systemic inflammation, metabolic stress, and dietary patterns all contribute to whether the brain remains resilient. A 2025 systematic review and dose-response meta-analysis published in a peer-reviewed journal examined whether soy consumption—particularly fermented varieties—might protect against neurocognitive disorders by acting on multiple pathways simultaneously.
Key takeaways
- Higher total soy intake was associated with an 8% lower odds of developing major neurocognitive disorders, based on data from 68,691 participants across 6 prospective Asian cohorts
- Natto (fermented soy) showed the strongest protective signal, with a 14% reduction in neurocognitive disorder odds for each additional gram of daily intake
- Protective effects were more pronounced in adults without prior stroke history, suggesting a vascular or microcirculatory mechanism
Study at a Glance
| Source | Peer-reviewed systematic review and meta-analysis (PMID: 40917087) |
| Study type | Systematic review and dose-response meta-analysis of prospective cohorts |
| Sample size | N = 68,691 participants |
| Population | Community-dwelling adults from Japan and Singapore; follow-up 5–20 years |
| Country | Japan and Singapore |
Soy intake and neurocognitive protection across fermented and non-fermented forms
Odds reduction for neurocognitive disorders by soy type, from 2025 meta-analysis (PMID: 40917087)
Source: Systematic review and meta-analysis, PMID 40917087 (2025) | Georgian Medical Journal News
Natto’s fermentation-derived compounds may explain stronger protection
The meta-analysis found that natto—a traditional Japanese fermented soy product—showed the strongest and most consistent association with reduced neurocognitive risk, with each additional gram of daily intake corresponding to a 14% reduction in odds. This effect size exceeds that of non-fermented soy forms, suggesting that fermentation itself may unlock or generate bioactive compounds not present in unfermented soy.
Natto contains several fermentation-derived components implicated in brain and vascular health. These include vitamin K2 (menaquinone-7 or MK-7), bioactive peptides generated during bacterial fermentation, and metabolites from probiotic organisms. Each of these agents acts on distinct but interconnected pathways: vitamin K2 supports endothelial function and vascular calcification prevention; peptides modulate neuroinflammation and blood-brain barrier integrity; and probiotic metabolites influence systemic and local inflammatory tone. Together, they address multiple mechanisms relevant to age-related cognitive decline—a discovery supported by research in vascular and neuroinflammatory biology.
Vascular health emerges as a likely mechanism
A particularly revealing finding was that protective effects were more pronounced in adults without prior stroke history. This pattern suggests that soy’s cognitive benefits may operate primarily through vascular and microcirculatory mechanisms rather than through direct neuroprotection alone. Individuals with established cerebrovascular disease may have crossed a threshold where dietary interventions offer less protection.
This observation aligns with a growing body of evidence linking vascular dysfunction—including endothelial impairment, arterial stiffness, and microinfarction—to cognitive decline in aging. Vascular contributions to cognitive impairment and dementia (VCID) have become a major focus of cognitive neuroscience, and dietary compounds that improve vascular health—such as those in fermented soy—may offer preventive value. The fact that soy’s effect was strongest in the stroke-free group suggests a preventive rather than therapeutic window.
Observational design and dietary self-report limit causal inference
The strengths of this analysis—large sample sizes (68,691 participants), long follow-up periods (5–20 years), and prospective design—are offset by important limitations. Dietary intake was measured via food-frequency questionnaires or dietary recalls, methods prone to recall bias and measurement error. The cohorts are also region-specific, reflecting habitual soy consumption patterns in Japan and Singapore (40–181 g/day), which far exceed typical Western intake. Generalizing these findings to Western populations consuming soy primarily as supplements or occasional foods may be premature without separate evidence.
Furthermore, observational cohorts cannot establish causality. Participants who consume more soy may differ in unmeasured ways—education, physical activity, overall diet quality, or healthcare-seeking behavior—that independently protect cognition. Residual confounding in dietary epidemiology remains a persistent challenge.
Higher soy intake, particularly in fermented form (natto), was consistently linked with lower long-term odds of major neurocognitive disorders in Asian populations, with natto showing a 14% reduction per additional gram of daily intake.
— Systematic review and dose-response meta-analysis (PMID: 40917087), 2025
What this means
Frequently asked questions
Does this mean I should eat more soy to prevent dementia?
The evidence suggests an association between higher soy intake and lower neurocognitive risk in Asian populations with habitual consumption of 40–181 grams per day. However, this is observational data and cannot prove causation. Soy should be considered one component of a dementia-prevention strategy that prioritizes vascular health, cognitive activity, sleep, and overall diet quality. If you enjoy soy foods, the data support including them; if you do not, there is no evidence that supplementation alone will prevent cognitive decline.
Why does natto show stronger effects than other soy foods?
Natto’s fermentation process generates vitamin K2 (MK-7), bioactive peptides, and probiotic-derived metabolites that are either absent or present in lower amounts in unfermented soy. Each of these compounds may act on pathways relevant to vascular and neuroinflammatory health. Fermentation also increases bioavailability of isoflavones. However, the meta-analysis cannot distinguish whether natto’s superiority reflects its unique chemistry, cultural practices around its consumption, or unmeasured confounding in the Asian cohorts studied.
Can these findings apply to Western populations?
Not directly, without additional evidence. The cohorts come from Japan and Singapore, where soy consumption is part of traditional diets consumed since childhood. Western soy intake is lower and often from processed sources (isolates, supplements) rather than whole foods. A separate body of research is needed to test whether isolated soy isoflavones or fermented soy products show similar cognitive benefits in Western populations. Until then, these data are most applicable to populations with habitual soy intake similar to that in the studied cohorts.
Future research should prioritize randomized controlled trials testing fermented soy extracts or natto supplementation in both Asian and Western populations, alongside mechanistic studies of how vitamin K2 and soy peptides influence vascular and neuroinflammatory biology relevant to aging. The convergence of vascular health, dietary polyphenols, and cognitive resilience is increasingly central to understanding how to extend healthy cognitive aging. This meta-analysis contributes a valuable regional perspective that complements Western dietary and neuroscience literature.
Source: Systematic review and dose-response meta-analysis (PMID: 40917087), 2025. Read more in New Studies and explore cognitive health topics in Explainers.
Was this article helpful?
Disclaimer. This article is health journalism intended for general information and education. It is not medical advice and is not a substitute for professional diagnosis or treatment. Always consult a qualified healthcare provider about your individual circumstances. Full disclaimer →
Related Coverage
Editorial standards. This article was produced under the GMJ News editorial process, with oversight by the GMJ Editorial Board. Our editorial process. Spotted an error? Contact the editorial team.
