15 Biomarkers That May Predict SuperAger Brains

A new study used machine learning and blood data to identify biomarkers linked to extraordinary cognitive health. The ability to think, learn, and remember clearly as you age, supported by brain structure, function, and lifestyle factors like sleep, diet, and exercise after 65.

What if a simple blood test could flag whether your brain is aging better than average? That’s the premise of a small but compelling new study published in Nature Scientific Reports. Researchers set out to decode what makes “SuperAgers,” people over 65 with memory and mental function rivaling folks in their 40s, so resilient. And they found clues not in the brain, but in the blood.

Using advanced machine learning techniques on just 81 individuals, the team identified 15 blood biomarkers that could predict exceptional cognitive performance with over 75% accuracy. These markers weren’t exotic. They were largely metabolic, inflammatory, or related to liver, immune, or vascular health, all things your standard blood panel might already be checking. What matters is the pattern.

While more research is needed before doctors can act on these findings, this study offers a fresh way to think about aging well: cognitive resilience The ability to recover quickly from stress or setbacks.

Paper:

https://www.nature.com/articles/s41598-025-01477-2

1. Glucose
Bidirectional effects were observed. Levels below 92 mg/dL supported attention, visual memory, and visuospatial skills. Interestingly, levels above 100 mg/dL were associated with better verbal memory and frontal executive function.

2. HDL Cholesterol
Higher HDL (>54 mg/dL, particularly >60 for women) consistently contributed to SuperAger classification and supported performance across language, memory, and frontal domains. This threshold exceeds standard cardiovascular targets.

3. ALT (Alanine Aminotransferase)
Levels above 17 U/L were positively associated with SuperAger status, especially in frontal and visuospatial domains, suggesting liver health may quietly support brain resilience

4. MCHC (Mean Corpuscular Hemoglobin Concentration)
Lower values (<34 g/dL) were associated with SuperAger status, though the mechanism remains unexplored in the study.

5. AGE (Advanced Glycation End-products)
Higher levels (>15 ng/mL) surprisingly aligned with SuperAger classification, challenging conventional thinking and indicating the need for more research.

6. oxLDL (Oxidized LDL)
Lower levels (<51 ng/mL) were linked to SuperAger status and better performance in visual and visuospatial domains.

7. CD36
A receptor involved in lipid metabolism and immune signaling, CD36 levels above 40 ng/mL were associated with better visual memory and visuospatial performance.

8. Insulin
Insulin sensitivity: How effectively your body uses insulin, which regulates blood sugar levels Learn More, not just level, may be key. Insulin showed up as an important signal, but more insulin in your blood wasn’t better; it was actually tied to weaker performance in certain thinking skills. What really matters is how sensitive your body is to insulin, or how well your cells respond, because that’s what keeps blood sugar stable and the brain fueled.

9. Phosphorus
Associated with visuospatial function, phosphorus likely plays a role in neuronal signaling and plasticity.

10. LDL Cholesterol
Its impact varied by cognitive domain. Extremely low levels (<70 mg/dL) were not universally beneficial, suggesting the need for balance.

11. RDW(CV) (Red Cell Distribution Width)
Lower variation (<12.5%) in red blood cell size was associated with better visuospatial performance.

12. Chloride (Cl)
Levels above 107 mmol/L exceeded typical clinical ranges and contributed to SuperAger prediction, possibly via roles in neurotransmission.

13. Leptin
Included as a key predictive biomarker. While domain-specific effects were not detailed, its role in metabolic regulation suggests relevance.

14. Vimentin
A cytoskeletal protein involved in inflammation. Your body’s response to an illness, injury or something that doesn’t belong in your body (like germs or toxic chemicals). Learn More, Vimentin was part of the final predictive model and linked to visuospatial and visual domains.

15. RAGE (Receptor for Advanced Glycation End-products)
Also included in the final set. RAGE–AGE interactions are known to trigger neuroinflammatory cascades, suggesting relevance to cognitive aging.

1 Like

Blood Tests for a Sharp Old Brain? A Korean Study Says Maybe, But the Signal Is Thin

Researchers at Ewha Womans University in Seoul drew blood from 81 healthy people aged 65 and over, 39 of whom performed on cognitive tests like adults in their 40s (“SuperAgers”) and 42 of whom performed as expected for their age. They measured 55 blood markers, from routine cholesterol and glucose panels to specialist assays for inflammation and lipid transport proteins. Conventional statistics found only four markers that differed between the groups, and none of those would survive correction for multiple testing. The team then applied machine learning with feature selection, synthetic data generation using a large language model, and SHAP interpretability, arriving at a 15-marker panel that classified SuperAger status with 76 percent accuracy and an AUC of 0.739 on a held-out set of 33 people. Notably, Alzheimer’s proteins and classic inflammatory cytokines showed no group difference at all. The metabolic markers did most of the work.

The interesting question in cognitive aging is not why some people decline. It is why a small minority do not. SuperAgers, people past 65 whose memory and executive function match adults thirty years younger, have been studied mostly through expensive brain scans that show thicker cortex and preserved connectivity. What nobody had done was ask a simpler question: does a SuperAger’s blood look different from everyone else’s?

This Korean team asked it. They recruited 81 healthy elderly volunteers from a Seoul hospital, sorted them by performance on a standard neuropsychological battery, and ran 55 blood measurements on each. The design was deliberately practical. If you could spot cognitive resilience from a routine health-checkup panel, you would have a cheap screening tool that works at population scale, unlike MRI.

The headline result is a negative one, and it is the most informative part of the paper. Tau protein, amyloid precursor protein, IL-6 and TNF-alpha, the biomarkers that dominate dementia research, showed no difference between the two groups. Neither did LDL, triglycerides, CRP, kidney function, or most of the panel. Whatever distinguishes a SuperAger is not visible through the lens the field has been using.

What did emerge, faintly, was metabolic. SuperAgers had lower fasting glucose (98 versus 106 mg/dL), lower HbA1c (5.75 versus 6.00 percent), higher HDL cholesterol (59 versus 56 mg/dL), and marginally lower oxidised LDL. Each of these differences is small, and each carries a p-value that the authors obtained using one-tailed tests without correcting for 55 comparisons. Taken individually, none is convincing.

The authors’ argument is that the individual markers are the wrong unit of analysis. Cognitive resilience, they suggest, is a pattern across many markers rather than any single one. To test that, they trained a gradient boosting classifier and found the combination did carry real information: roughly 76 percent classification accuracy where guessing the larger group would give 52 percent.

The catch is that they got there partly by inflating a 49-person training set with 70 synthetic patients generated by a language model. Synthetic data can stabilise a model. It cannot create information that was not in the original 49 people. The paper is a well-motivated proof of concept, not evidence that a blood panel predicts cognitive reserve. [Confidence: High]

Insights

The differences that showed up are real in direction but small in magnitude. Statisticians measure this with Cohen’s d, which asks how far apart two groups are relative to how spread out people are within each group. A d of 0.4 means the two bell curves overlap by about 84 percent. Practically, if you picked one SuperAger and one typical ager at random, the SuperAger would have the better fasting glucose only about 61 percent of the time, versus 50 percent by coin flip.

Every blood difference in this study sits between d = 0.25 and d = 0.41. That is the magnitude of a modest population trend, not a personal predictor.

What the data weakly supports, and what is already supported by much stronger evidence elsewhere: keeping fasting glucose in the high 80s to low 90s rather than the low 100s, keeping HbA1c below roughly 5.8, and keeping HDL above 54 mg/dL rather than merely above the standard clinical cutoff of 40 to 50. The recurring theme is that clinically normal is not the same as optimal, and the SuperAgers clustered at the favorable end of normal rather than outside it.

What this study does not support: any claim that AGE, RAGE, CD36, vimentin or leptin levels tell you anything about your brain. Those markers made the panel on algorithmic ranking alone and showed no group difference.

Context and Source

  • Open Access Paper: Machine learning based prediction of cognitive metrics using major biomarkers in SuperAgers
  • Authors: Hyo-Bin Lee, So-Yeon Kwon, Ji-Hae Park, Bori Kim, Geon-Ha Kim, Jang-Hwan Choi, Young Mi Park
  • Institution: Ewha Womans University (Departments of Computational Medicine, Molecular Medicine, and Artificial Intelligence) and Ewha Womans University Mokdong Hospital, Seoul, South Korea
  • Journal: Scientific Reports, volume 15, article 18735
  • Impact evaluation: The impact score of this journal is 3.9 (2024 Journal Citation Reports two-year JIF; aggregator sources report a 2025 figure near 4.9, which I have not been able to confirm against the primary JCR record), evaluated against a typical high-end range of 0 to 60+ for top general science, therefore this is a Low-to-Medium impact journal. Scientific Reports is a high-volume, soundness-only megajournal: peer review assesses technical validity rather than novelty or importance, so publication here carries little signal about a finding’s significance.

Biomarker Data: Effect Size Extraction

A note on reading these numbers. Cohen’s d expresses a group difference in units of the natural spread within groups. By convention, 0.2 is small, 0.5 is medium, 0.8 is large. A more intuitive companion measure is the common language effect size, which answers: if you draw one person from each group at random, how often does the SuperAger come out ahead? At d = 0.4 the answer is 61 percent. At d = 1.2 it is 80 percent. Chance alone gives 50 percent.

I calculated all values below from the reported means, standard deviations and t-statistics in Tables 1 and 2. The authors reported none of them.

Cognitive domain differences (the phenotype itself)

Domain Typical (n=42) SuperAger (n=39) Cohen’s d Common language
Visual Memory (RCFT delayed recall z) -0.12 +0.88 1.21 80%
Verbal Memory (SVLT delayed recall z) +0.32 +1.06 0.84 72%
Visuospatial (Rey CFT copy z) -0.26 +0.28 0.73 70%
Language (K-BNT naming z) -0.06 +0.34 0.48 63%
Attention (digit span forward z) +0.08 +0.39 0.31 58%
Frontal (Stroop color reading z) +0.58 +0.79 0.28 58%

The phenotype is driven overwhelmingly by memory and visuospatial performance. Attention and frontal executive function barely separate the groups, which is unusual for a SuperAger cohort and suggests the classification is capturing a memory-specific rather than a global cognitive advantage.

Blood biomarker differences

Marker Typical SuperAger Absolute delta Relative Cohen’s d Reported p (one-tailed)
HbA1c (%) 6.00 5.75 -0.25 -4.2% 0.37 to 0.39 0.044
Glucose (mg/dL) 106.10 98.38 -7.72 -7.3% 0.41 (see note) 0.035
Chloride (mmol/L) 104.90 105.85 +0.95 +0.9% 0.41 0.036
Free T4 (ng/dL) 1.30 1.17 -0.13 -10.0% 0.15 to 0.41 (see note) 0.034
HDL cholesterol (mg/dL) 55.55 59.13 +3.58 +6.4% 0.30 0.089
EPO (mIU/mL) 12.08 10.56 -1.52 -12.6% 0.35 0.060
oxLDL (ng/mL) 52.37 51.00 -1.37 -2.6% 0.31 0.085
Insulin (uIU/mL) 9.06 7.71 -1.35 -14.9% 0.23 to 0.25 0.148
Adiponectin (ug/mL) 35.85 44.24 +8.39 +23.4% 0.25 0.136
RDW (%) 12.60 12.75 +0.15 +1.2% 0.27 0.115
AGE (ng/mL) 13.56 12.23 -1.33 -9.8% 0.20 0.183
RAGE (pg/mL) 85.54 80.95 -4.59 -5.4% 0.17 0.224

Every blood effect size falls in the small range, 0.15 to 0.41. The largest metabolic effect (glucose, chloride) reaches only d = 0.41, which corresponds to 84 percent distributional overlap between SuperAgers and typical agers.

Exceptionally underwhelming. If you have to LLM-create synthetic subjects to enrich your sample, and the effect sizes are barely perceptible, it really doesn’t inspire confidence, especially if comparing to real life Super Agers you are only “capturing” a slice of the cognitive advantage. I mean, if you torture the data viciously enough, it’ll confess to anything you want.

I think we need to look elsewhere, this study is not it when it comes to identifying what distinguishes Super Agers from normal agers. Because the phenotype is distinctive enough - the cognitive advantage sticks out as very significantly off several standard deviations, meanwhile this data implies the markers they found are barely different. Something doesn’t add up. Sure, there’s likely a group effect, but even so, this is too weak a signal if it’s even real.

I’m not getting excited here. We need something much better.

1 Like