Higher Vitamin C Levels Are Tied to Larger Brain Volume and Stronger Memory Networks

Vitamin C is not typically the first nutrient researchers reach for when studying brain structure. Yet a large cohort study published in PLOS One involving 2,044 older Japanese adults found that higher plasma vitamin C levels were associated with greater gray matter volume, greater white matter volume, and measurably different connectivity in the brain's default mode network (DMN), a set of regions central to memory, future planning, and self-reference. The effect sizes are modest, but they held up after adjusting for age, sex, diabetes, smoking, education, and other lifestyle factors, which is exactly the kind of result that earns a second look.

Plasma vitamin C associates with gray and white matter brain volume

Participants in the Iki-Iki Health Promotion Project, a population-based study of community-dwelling residents in Hirosaki City, Japan, had their blood drawn after an overnight fast and their brains scanned on a single 3-Tesla MRI scanner. The mean age was 70.0 years and 61.1% of participants were women. Researchers measured plasma vitamin C using a commercial assay, then processed T1-weighted brain images to calculate the ratio of gray matter volume (GMV) and white matter volume (WMV) to total intracranial volume.

The result: higher plasma vitamin C was positively associated with both GMV/ICV ratio and WMV/ICV ratio, with p-values below 0.001 for both. To put those numbers in context, diabetes was associated with a standardized coefficient of -0.240 for gray matter, and smoking with -0.121 for white matter. Vitamin C's association, while smaller in magnitude, runs in the protective direction and is statistically comparable to other lifestyle factors researchers routinely track. As one review of ascorbic acid's role in the central nervous system notes, vitamin C is recognized as a potent antioxidant that scavenges reactive oxygen species in the brain, which operates at a high metabolic rate and is therefore especially vulnerable to oxidative damage.

Voxel-wise analysis pinpointed where the gray matter associations were strongest: the posterior cingulate cortex, middle cingulate cortex, medial prefrontal cortex, and inferior temporal gyri. Every one of those regions sits within or adjacent to the core DMN.

Default mode network connectivity shifts with vitamin C status

The study went a step further using Source-Based Morphometry, a multivariate technique that extracts spatially independent gray matter structural networks from MRI data, rather than examining single voxels in isolation. Independent component analysis produced 147 structural networks; neuroradiologists identified three as DMN-related (anterior DMN, posterior DMN-I, and posterior DMN-II), with 100% inter-rater agreement on classification.

Higher plasma vitamin C was positively associated with anterior DMN loading (standardized β = 0.097, p < 0.001) and positively associated with posterior DMN-I loading (β = 0.082, p < 0.001). It was negatively associated with posterior DMN-II loading (β = -0.111, p < 0.001). All three associations survived False Discovery Rate correction, with adjusted q-values below 0.001.

The opposing directions are not contradictory. The anterior DMN and posterior DMN-I components both declined with age in this sample, suggesting they represent normal structural integrity that erodes over time. The posterior DMN-II component increased with age, suggesting it reflects an aberrant, compensatory pattern of structural change. Vitamin C appears to support the healthy components and suppress the aberrant one, a pattern the authors interpret as helping preserve normal DMN architecture while resisting age-related deterioration.

Research on structural and functional brain network coupling underscores why this matters: individual differences in cognitive performance are significantly linked to the characteristics of both structural and functional brain networks. A nutrient that shifts those networks in a predictable direction is worth understanding precisely.

How 2,044 participants and a single 3T scanner kept the data clean

One reason to take these findings seriously is methodological consistency. All brain scans came from a single 3-Tesla scanner running a standardized protocol, which eliminates scanner variability, a common headache in multi-site neuroimaging. Participants with MRI distortions, missing T1 images, Alzheimer's disease diagnoses, or incomplete lifestyle data were excluded before analysis. Blood draws were taken after an overnight fast, limiting the noise introduced by recent dietary intake.

As large cohort studies are designed to do, the sample of 2,044 provided enough statistical power to detect associations of modest effect size while adjusting for a long list of covariates simultaneously. That list included age, sex, years of education, Mini-Mental State Examination (MMSE) score, medical history (diabetes, hypertension, hyperlipidemia), and lifestyle factors including smoking.

Key limitations before reading too much into this

The study is cross-sectional: everyone was measured at one point in time, so there is no way to determine whether higher vitamin C causes larger brain volume or whether people with healthier brains happen to have higher vitamin C, perhaps because they eat better overall. The authors acknowledge this directly.

Vitamin C was measured only once. Plasma vitamin C can fluctuate with recent diet, illness, and seasonal fruit availability. A single measurement may underestimate the true association, or it may introduce noise that dilutes it. Factors that complicate vitamin C research include age, disease status, and transport dynamics that make plasma levels an imperfect proxy for brain concentrations.

The cohort is also ethnically homogeneous and highly educated, both of which limit how far the findings generalize. Residual confounding from total dietary quality, BMI, and socioeconomic factors beyond education cannot be ruled out. Notably, plasma vitamin C did not directly correlate with MMSE scores (ρ = -0.011, p = 0.631), which means the study is not claiming that taking more vitamin C will improve your performance on a cognitive test tomorrow.

Longitudinal studies needed to establish whether vitamin C protects aging brains

What the research does establish, provisionally, is an association between a person's circulating vitamin C and the structural integrity of brain regions that matter for cognition in older adults. The anterior DMN and posterior DMN-I components that tracked with higher vitamin C were themselves positively correlated with MMSE scores, creating an indirect chain: vitamin C status relates to network structure, and network structure relates to cognitive performance.

Humans cannot synthesize vitamin C and must obtain it through diet, primarily fruits and vegetables. The brain concentrates it aggressively: cerebrospinal fluid vitamin C levels run more than twice as high as plasma levels, which signals that the brain treats it as a priority cargo. Whether boosting intake in adults with low plasma levels would actually shift brain structure over time is a question only a longitudinal intervention trial can answer. The authors call for exactly that: repeated measurements, longer follow-up, and cohorts that span a broader range of ages, ethnicities, and educational backgrounds.

For now, the evidence is emerging rather than established. But for researchers interested in modifiable nutritional factors that track with brain structure in aging populations, vitamin C has moved a step closer to the center of that conversation.

Evidence rating

EMERGING — 78/100

Caveats:

• Cross-sectional design — association only, causation cannot be established.
• Single plasma vitamin C measurement; repeat measures were not available.
• Cohort is exclusively older Japanese adults with high educational attainment, limiting generalizability.
• Residual confounding possible; total dietary intake, BMI, and socioeconomic factors were not fully controlled.
• Funding was not disclosed in the article.

Large cross-sectional cohort (n=2,044). Demonstrates associations between plasma vitamin C and brain structural measures but cannot establish causation. Findings are novel in this imaging context and have not yet been independently replicated; single-timepoint vitamin C measurement limits precision.

References

1. Haruka Nagaya, Keita Watanabe, Tomohiro Shintaku, et al. (2026). Plasma vitamin C levels are associated with brain structural networks on MRI: A large cohort study. PLOS One. https://doi.org/10.1371/journal.pone.0348504
2. Adriana Covarrubias-Pinto, Aníbal Acuña, Felipe Beltrán, Leandro Torres-Díaz, Maite Castro (2015). Old Things New View: Ascorbic Acid Protects the Brain in Neurodegenerative Disorders. International Journal of Molecular Sciences. https://doi.org/10.3390/ijms161226095
3. Johanna L. Popp, Jonas A. Thiele, Joshua Faskowitz, Caio Seguin, Olaf Sporns, Kirsten Hilger (2025). Structural-functional brain network coupling during cognitive demand reveals intelligence-relevant communication strategies. Communications Biology. https://doi.org/10.1038/s42003-025-08231-4
4. Research Design: Cohort Studies – PMC – NIH. NIH. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4104944/
5. Alexander Michels, Balz Frei (2013). Myths, Artifacts, and Fatal Flaws: Identifying Limitations and Opportunities in Vitamin C Research. Nutrients. https://doi.org/10.3390/nu5125161