The Tau Movement: The More Active the Body, the 'Quieter' the Alzheimer's Markers—and the Better the Memory

Alzheimer's disease (AD) is the leading cause of dementia in the elderly; there are still no effective drugs that radically change the course of the disease. Physical activity is one of the few modifiable factors that is consistently associated with better brain aging and a lower risk of cognitive decline.

In a large Korean study of 25 memory centers (n=1,144, average age 71 years), people with higher physical activity had lower levels of plasma markers of neurodegeneration and Alzheimer's disease — pTau-217 and NfL — and better cognitive tests. The effect was especially noticeable in participants aged 65+ and in those with pre-existing cognitive impairment. The study is published in JAMA Network Open.

What is already known

• Blood biomarkers have become a reliable “window” into the pathology of bronchial asthma:
  • pTau-217 reflects tau pathology;
  • NfL (neurofilament light chain) - degree of neuronal damage/neurodegeneration;
  • GFAP - astrocytic activation/neuroinflammation;
  • Aβ42/40 ratio - amyloid cascade.
• Observational studies and small interventions have shown that active people fail tests less often and later, and that vascular function, sleep, and neuroplasticity may improve.
• However, the relationship “movement ↔ molecular markers of AD” has been studied fragmentarily: cognitive tests, PET/CSF, small samples were most often looked at; plasma pTau-217 and NfL were rarely assessed, and correction for the real amyloid load was even less common.

Where is the gap?

• It is unclear to what extent actual weekly activity (not just program participation) is associated with blood pTau-217/NfL/GFAP/Aβ42/40 levels in the general clinical population – in healthy individuals, with MCI, and with dementia.
• It is unclear whether this association is independent of PET amyloid (centiloid), age, education, and vascular factors.
• It is not clear who benefits more from this potential treatment: the “healthy” elderly or those with MCI/dementia.
• There is little evidence as to whether the effects of activity on cognition are mediated in part through reductions in tau pathology/neurodegeneration (mediator pathways).

What did they do?

• Who: 1144 people with different cognitive status (normal, MCI, Alzheimer's dementia), South Korea.
• How activity was assessed: International questionnaire IPAQ → total MET-min/week; divided into quartile groups from Q1 (minimum) to Q4 (maximum).
• What was measured in the blood:
  • pTau-217 is the “signature” of tau pathology in Alzheimer’s,
  • NfL - neurofilament light chain, a marker of neuronal damage,
  • GFAP - astrocyte response (neuroinflammation),
  • Aβ42/40 - amyloid ratio.
• Cognition: MMSE and CDR-SB.
• Analytics: Multivariate models adjusted for age, sex, PET amyloid formation and load (centiloid), and vascular factors.

Main results

• Plasma markers. Compared with the least active (Q1), the most active (Q4) had lower pTau-217 (estimate -0.14; p = 0.01) and lower NfL (-0.12; p = 0.01). Q3 was also significant for NfL (-0.10; p ≈ 0.047).
• Amyloid and GFAP. No associations were found with Aβ42/40; for GFAP the trend was weakened after adjustments (borderline significance).
• Cognition: All more active groups had higher MMSE (~+0.8–0.94 points) and lower CDR-SB (better everyday functioning).
• Who it helps the most: in people aged 65+ and those with cognitive impairment, activity was more strongly associated with both "chemistry" (lower than pTau-217, NfL, GFAP) and tests. In the cognitively intact group, the connection between activity and pTau-217 was most clearly visible.
• How it might work: Mediation analysis showed that part of the effect of activity on cognition is mediated through pTau-217 (~18–20% of the indirect effect) and NfL (~16% for MMSE). That is, physical activity may influence tau pathology and neurodegeneration, and the rest is a direct contribution through vascular, neuroplastic, and other mechanisms.

Why is this interesting?

• Not just prevention, but also “biology.” It’s not about “whoever is more active has a better test,” but about the connection with molecular markers of Alzheimer’s disease in the blood. It’s especially important that the associations persisted after taking into account the amyloid load on PET, but Aβ42/40 was not associated with activity — a hint that movement may have a stronger effect on tau/neurodegeneration than on amyloid.
• Window of opportunity. More pronounced connections in 65+ and in those with existing disorders indicate: it is not too late to start, even when problems are noticeable.

What this does not prove

• The design is cross-sectional: we see associations, not causal evidence. Reverse causality is possible (poorer cognition → less movement).
• Activity - self-report (part - from the words of caregivers), errors are possible.
• One country, one health care system - let's generalize with caution.

What to do today

• Move regularly. WHO guidelines: 150–300 minutes of moderate or 75–150 minutes of vigorous aerobic activity per week + 2 days of strength training. Walking at a “conversational speed”, Nordic walking, exercise bike, swimming are good starters; add balance exercises.
• Routine is important. Divide it into 5-6 short sessions a week; even 10-15 minutes makes sense if it’s systematic.
• For MCI or dementia: choose simple, safe exercises, involve your family/physical therapy instructor; monitor your blood pressure, pulse, and hydration.

Conclusion

Physical activity in the elderly is associated not only with better tests, but also with “silent” markers in the blood — lower pTau-217 and NfL, especially in those aged 65+ and with cognitive impairment. This is not yet proof of causality, but the signal is powerful: movement is one of the most realistic ways to slow the path to cognitive decline, acting both “through the blood” and directly through the vessels, plasticity and endurance of the brain. Now we need longitudinal and interventional studies to translate associations into proven recommendations for “how much, how and to whom”.