A new brain imaging study of 231 older adults is raising questions about whether standard vitamin B12 blood tests and deficiency cutoffs are missing early damage in the aging brain. Led by researchers at UC San Francisco, the work found that people in their seventies with “normal” total B12 levels but lower active B12 showed slower thinking speed and more white-matter injury, suggesting current guidelines may not fully protect older brains. The latest publicly available guideline benchmarks for vitamin B12 were set years ago, so these findings arrive in a period when formal thresholds have not recently been updated.
Why scientists say vitamin B12 guidance matters now
The tension begins with who is being studied. The UCSF team examined 231 participants with a median age of 71 in the BrANCH cohort, focusing on how different forms of vitamin B12 relate to central nervous system health according to the primary UCSF study. Instead of looking only at total serum B12, the researchers measured holotranscobalamin, or holo-TC, which represents biologically active B12 circulating in the blood.
Lower holo-TC was linked with delayed visual evoked potentials, which are electrical signals the brain produces in response to visual stimuli, according to the same Annals of Neurology report. The study also found that lower holo-TC was associated with slower processing speed in an age-dependent way and with greater white matter hyperintensities, which are lesions seen on MRI that are often interpreted as signs of small vessel or tissue injury in the brain. These associations appeared even though the participants’ average total B12 level was reported as well above the typical U.S. deficiency cutoff, according to a related Institutional summary.
That gap between “normal” total B12 and low active B12 matters because current clinical practice often defines deficiency using total serum B12 alone. The Authoritative vitamin B12 fact sheet from the NIH notes that clinical subnormal serum B12 cutoffs often vary around 200 to 250 pg/mL, and describes 150 to 399 pg/mL as a borderline range where additional testing is recommended according to the NIH fact sheet. In this range, methylmalonic acid, or MMA, is used as a sensitive functional marker to confirm deficiency, but holo-TC is not yet a routine part of most lab panels.
The hypothesis that raising holo-TC above a specific threshold such as 150 pmol/L could speed up choice-reaction time by 15 to 20 percent within a year in older adults with total B12 between 250 and 400 pg/mL is not directly tested in the available sources. The BrANCH data show that lower holo-TC is associated with slower processing speed and delayed visual signals, and the NIH materials describe MMA as a functional marker in the 150 to 399 pg/mL range, but no randomized trial has yet confirmed that targeted holo-TC elevation produces a specific percentage improvement in reaction time. The idea of drawing such a trial from Nationally representative sampling frames such as NHANES remains a proposal rather than a documented study.
The evidence behind concerns about current vitamin B12 thresholds
The strongest evidence challenging current practice comes from the detailed biomarker work in the BrANCH cohort. In that study of 231 older adults with a median age of 71, lower holotranscobalamin was statistically associated with delayed visual evoked potentials, slower processing speed in an age-dependent pattern, and greater white matter hyperintensities according to the Annals of Neurology report on vitamin B12 fractions and CNS injury. The associations were observed even when total B12 levels were not in the deficient range by standard cutoffs, which suggests that tissue-level B12 availability can fall short before serum totals flag a problem.
Guideline documents show how those serum totals became the reference point. The Canonical Dietary Reference Intakes chapter on vitamin B12 sets the Estimated Average Requirement and Recommended Dietary Allowance and states that adults 51 years and older should meet much of their vitamin B12 requirement from fortified foods or supplements containing crystalline B12, because absorption from natural food sources can decline with age, according to the Dietary Reference Intakes. That chapter provides the framework for current intake advice but does not incorporate holo-TC thresholds or MRI findings such as white matter hyperintensities into its definitions of adequacy.
Diagnostic reviews add another layer. A peer-reviewed overview of biomarkers for vitamin B12 deficiency explains that total serum B12 can be an imperfect proxy for tissue status and that MMA is often treated as a more sensitive functional marker, particularly when serum B12 is in the borderline range, according to a review of biomarkers and MMA. That review supports the NIH fact sheet’s emphasis on MMA for 150 to 399 pg/mL serum B12, yet holo-TC is still not widely used in primary care despite its direct measurement of active B12.
Population-level data from the National Health and Nutrition Examination Survey, or NHANES, show that vitamin status and cognition are already being linked in national samples. An analysis of the NHANES 2011 to 2014 dataset examined cognitive test performance in older adults in relation to combined folate exposure, including unmetabolized folic acid, and B12 status according to a Nationally representative study of folate and B12 interactions. Another peer-reviewed analysis of the same 2011 to 2014 NHANES cycles looked at supplement intake, folate and B12 biomarkers, and cognitive impairment among older adults and was described as useful for understanding how supplement patterns relate to biomarker distributions and cognitive outcomes in that age group, according to a Peer NHANES analysis.
These NHANES studies do not measure holo-TC or white matter lesions directly, but they show that when B12 status is low in the presence of higher folate exposure, cognitive performance on standard tests tends to be worse. That pattern aligns with the BrANCH finding that lower active B12 is linked to slower processing speed and delayed brain responses, even when total B12 appears adequate. Together, the cohort and national survey data suggest that relying solely on total serum B12 and broad intake recommendations may leave a subset of older adults with subtle but functionally important brain changes that are not caught early.
What remains unresolved about protecting older brains
Despite these warning signs, several key questions remain open. The latest guideline-setting documents for vitamin B12, including the Dietary Reference Intakes chapter and the Authoritative NIH fact sheet, do not incorporate holo-TC thresholds, MRI white matter findings, or visual evoked potential delays into their criteria according to the Dietary Reference Intakes and the NIH fact sheet. There is no evidence in the available sources that committees have formally revisited the 2.4 microgram-per-day RDA for older adults in light of the BrANCH data or that they have debated adding holo-TC or MMA-based thresholds to intake targets.
The BrANCH cohort is cross-sectional in nature, so it documents associations between lower holo-TC, slower processing, delayed visual signals, and greater white matter hyperintensities in 231 people with a median age of 71, but it does not show whether raising holo-TC reverses or slows those brain changes according to the Annals of Neurology UCSF study. There are no longitudinal intervention trials in the provided record that track white-matter lesion burden or visual evoked potentials after targeted holo-TC elevation in this age group.
Similarly, while NHANES 2011 to 2014 data have been analyzed for links between B12 status, folate exposure, and cognitive test performance, there is no trial in these sources that randomizes older adults with total B12 between 250 and 400 pg/mL to different B12 strategies and then measures a 15 to 20 percent change in choice-reaction time, as imagined in the hypothesis. The idea that such an effect size could be detected using an NHANES-style sampling frame remains speculative, as the Nationally representative analyses to date have been observational according to the folate and B12 NHANES dataset study.
For readers, the unresolved science translates into a practical gap. Current advice from the Dietary Reference Intakes and NIH materials tells adults 51 and older to rely on fortified foods or crystalline supplements for much of their B12 intake and to treat serum B12 below about 200 to 250 pg/mL, or a borderline range of 150 to 399 pg/mL with elevated MMA, as potential deficiency, according to the Canonical DRI chapter and the Authoritative NIH fact sheet. Yet the UCSF findings show that even when total B12 is comfortably above that deficiency line, lower active B12 can be linked to slower mental processing and more white-matter damage in the brain.
The next developments to watch will be whether guideline bodies such as those behind the Dietary Reference Intakes or federal nutrition policy sites such as dietaryguidelines.gov revisit vitamin B12 language for older adults, and whether researchers launch randomized trials that test whether raising holo-TC or correcting MMA-defined deficiency improves reaction time, visual evoked potentials, or white-matter lesions. Until those data arrive, the evidence base shows a clear signal that current vitamin B12 thresholds may not fully guard against subtle neurocognitive decline in aging, even if they prevent the most obvious signs of deficiency.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
