2,711 Autopsies Show Alzheimer's Brain Changes Are Rare Before 70

The evidence is an observational autopsy study in people, posted as a preprint on medRxiv. A preprint has not been peer reviewed, so details beyond it are unverified.

The researchers used the Biobank for Aging Studies of the University of Sao Paulo. Decedents were sampled through the city's centralized autopsy service, which performs about 15,000 autopsies a year and verifies about 16% of deaths in the city. The cohort covered 2004 through 2025. It included adults with complete amyloid, tau, dementia-rating and coexisting-pathology assessments. Mean age was 73.4 years, the range was 23 to 111, and 1,340 (49.4%) were female.

Amyloid and tau are two proteins that build up in Alzheimer's disease. The team staged each using standard tissue measures: Thal phase for amyloid and Braak stage for tau tangles. They defined the range a biomarker could detect as Thal phase 3 to 5 together with Braak stage IV to VI.

Of the 2,711 decedents, 531 (19.6%) met both thresholds. Another 270 (10.0%) met the amyloid threshold only, 207 (7.6%) the tau threshold only, and 1,703 (62.8%) neither. So the two markers disagreed in 17.6% of cases, according to the authors.

Age mattered a great deal. The combined state was found in 0% of those aged 23 to 50. It rose to 5.5% at ages 61 to 70, 14.5% at 71 to 80, 36.5% at 81 to 90, and 45.9% at 91 to 111.

Cognitive impairment, defined here as a Clinical Dementia Rating above 0, tracked tau stage. It was present in 88 of 541 people (16.3%) at Braak stage 0 and 155 of 179 (86.6%) at stage VI.

The second finding concerns other causes. The authors counted contributing coexisting brain disease, which amyloid and tau tests do not detect. Impairment was present in 225 of 1,467 people (15.3%) with neither type of pathology. It was present in 204 of 349 (58.5%) with coexisting disease alone, 216 of 339 (63.7%) with amyloidtau pathology alone, and 165 of 192 (85.9%) with both.

In a statistical model of dementia severity, coexisting disease had the largest unique contribution (partial R2 0.088). Braak stage accounted for the largest share of total explained variance. These are associations in a cross-sectional study. They do not show that either pathology caused the impairment.

For reading a test result, the authors conclude that these tissue-based frequencies can help interpret positive and negative biomarker results. In plain terms, a positive result is less common in younger people. Impairment in someone with a negative result, or a positive one, may still involve other brain disease.

This adds a different angle to our October 20 report on a mouse study of brain immune cells and sleep in Alzheimer's. That work was in animals. This one looks at human brain tissue and suggests Alzheimer's is often not the only factor.

Next to watch: peer review of this preprint, and whether other autopsy cohorts with different populations find the same age pattern and the same role for coexisting disease.

Faster Epigenetic Aging Linked to Higher Death Risk Across 381,000 Adults, Meta-Analysis Finds Sleep Shortened and Shifted Later as Electricity Reached Argentine Communities, Study Finds Exercise for Mild Cognitive Impairment: Network Meta-Analysis Ranking Could Not Be Confirmed FDA Clears AI Tool Estimating Alzheimer's Risk in Mild Cognitive Impairment Low Activity Plus Weak Grip Linked to Higher Cognitive Impairment Risk Review of 248 Studies Ranks Protein Plus Strength Training Top for Sarcopenia Amylin Drug Petrelintide Cut Weight Up to 9.2% in Diabetes Trial Hearing and Vision Loss Linked to Body-Mind-Memory Decline, Partly via Loneliness NAD+ and Calorie Restriction May Work Through Brain Cell Antennae in Mice Heart Amyloid Drug Tafamidis Linked to Healthier Blood Vessel Cells Eye Gene Therapy Linked to Vision Gains in Stargardt, Heads to Geographic Atrophy Trial Stair-Climb Test Gave Only Rough Estimate of Aerobic Fitness Body-Clock 'Circadian Depression' Subtype Found in a Quarter of Patients Epigenetic Age Acceleration Tracks Kidney Disease Progression in Type 1 Diabetes Most Epigenetic Aging Clocks Fail to Predict Dementia, Meta-Analysis Finds Review in Public Health Asks Whether Exercise Timing Affects Heart and Metabolic Markers Fasting Mimicking Diet Programme for Diabetes and Obesity Reaches UK Clinics AI Drug Firm Pathos Raises $237 Million, Shifts Toward Alzheimer's APA Survey: Most Americans Accept Chronic Stress, Nearly Half Withdraw Socially Tiny Peptide Alleviates Artery Calcification in Mice by Changing Cell Energy Use Hormone Therapy Linked to More Frequent, Longer Migraines After Menopause Meta-Analysis Examines Which Exercise, and How Much, Relates to Sleep in Older Adults 2,711 Autopsies Show Alzheimer's Brain Changes Are Rare Before 70 Does Pregnancy's Biological Aging Fully Reverse? A Hypothesis Awaits Testing Diet, Exercise and Empagliflozin Matched on Blood Sugar in Older Diabetics Review Finds Ultra-Processed Food Subtypes Differ Sharply in Diabetes Risk AI-Designed Peptide Targeting Zombie Cells Heads Toward First Human Testing Blocking Lactate Exporter MCT4 Improved Aged Egg Quality in Lab Study Aquatic and Dance Exercise Accompanied Lower Inflammation in Obese Older Women Aging Brain Cells Keep Their Identity, Challenging Epigenetic Drift Theory