Share
6 min read
researchEvidence-based

Why cognitive variability predicts Alzheimer's decline better than test scores

This changes how we should think about monitoring our brain health.

T
· Reviewed by Dr. Kevin Tran, PharmD

Key takeaways · TL;DR

Day-to-day cognitive variability (your 'good days and bad days') is a stronger early predictor of Alzheimer's decline than average test scores in APOE4 carriers. Research from AAIC 2025 shows APOE4 carriers have more cognitive variability than non-carriers even when clinically healthy, and high-variability days correlate with more adverse driving events the same day. Smartphone apps make this trackable for the cost of an amyloid PET divided by thousands.

Definition

Day-to-day or within-day fluctuation in mental performance, distinct from average test scores and potentially a better early warning signal.

Standard neuropsychological testing captures a single snapshot of performance, but cognitive function naturally fluctuates across days based on sleep, stress, glucose, and other factors. Measuring this fluctuation (rather than averaging it away) can reveal reduced neural reserve earlier than single-point tests. For APOE4 carriers, variability is particularly informative because reduced metabolic reserve amplifies the impact of suboptimal days, making the signal detectable before traditional testing would flag abnormalities.

This changes how we should think about monitoring our brain health.

First, monitoring your brain health is critical to track the effectiveness of your interventions. What is tracked can be improved. Without tracking you are basically spinning your wheel.
I have written a few articles about tracking on our blog https://apoe4.co
If You’re Not Tracking, You’re Guessing


Ok- back to the video:

I just finished analyzing another six presentations from AAIC 2025, and the findings are both sobering and empowering.

The core insight: Your "good days and bad days" (or in other words the variability in your cognitive performance) might be a more powerful early warning signal than your average test scores.

And here's the part that hit home for me: we as APOE4 carriers show MORE cognitive variability than non-carriers even when we're clinically healthy. Even when traditional testing shows we're "fine."

What the research revealed:

📊 Dr. Katie Bangen (UC San Diego) tracked 818 people for 3 years. For APOE4 carriers , high variability at baseline predicted faster decline in real-world functioning (e.g. managing money, taking meds, handling complex tasks) before cognitive tests became abnormal.

📱 Dr. Andy Aschenbrenner (Washington University) used a smartphone app to track people 4x/day for a week. APOE4 carriers had more ups and downs across the week. And here's what's wild: on days when the app showed worse cognition, people had MORE adverse driving events THAT SAME DAY. More hard braking, more speeding, more sudden acceleration. (They analyzed 20,000+ car trips to prove this.)

But there's a silver lining: people seemed to know when they were having off days. They avoided risky nighttime driving on low-cognition days without even realizing why.

🌍 Dr. Laiss Bertola validated this in 9,000+ Brazilians over 8 years. Higher variability at baseline = higher odds of impairment eight years later.

⚠️ Dr. Andrew Kiselica revealed the nuance: variability scores are unreliable in asymptomatic people (mostly just noise), but become highly informative once symptoms appear. This means daily smartphone monitoring might be better for us in the asymptomatic stage.

Why this matters for us:

Unlike expensive biomarkers ($5K for amyloid PET, $1.5K for CSF testing), variability can be tracked through:

✓ Standard neuropsych tests

✓ Smartphone apps (minutes per day)

✓ Remote monitoring

✓ No invasive procedures

My questions for the community:

1. Have any of you noticed patterns in your "good days" vs. "bad days"?

2. Would you use a smartphone app to track daily cognitive variability if it were available?

3. For those who've had neuropsych testing, did your doctor ever mention your score variability, or just your averages?

Most Newsletters? One-way street.
How boring…
This is the Phoenix Community after all—so let's make it a two-way street.
Got a question? Feedback? Just want to say hi?
Hit reply.
I read every single one.

Filed under
researchtrackingcognition

Discussion

Join the conversation

Your email will never be published. Be respectful and constructive.

FAQ

Frequently asked questions.

What is cognitive variability and why does it matter for APOE4 carriers?
Cognitive variability measures the fluctuation in your mental performance across days or within a day, rather than just your average score on a single test. Research presented at AAIC 2025 by multiple teams shows APOE4 carriers have significantly more cognitive variability than non-carriers even when traditional testing labels them 'fine.' Dr. Katie Bangen (UC San Diego) tracked 818 people for 3 years and found that for APOE4 carriers specifically, high baseline variability predicted faster decline in real-world functioning (managing money, taking medications, handling complex tasks) before cognitive tests became abnormal. For carriers, variability may be the earliest actionable warning signal available.
Can smartphone apps detect early cognitive decline in APOE4 carriers?
Yes, and potentially better than single-visit neuropsych testing. Dr. Andy Aschenbrenner at Washington University used a smartphone app to test people 4 times per day for one week. APOE4 carriers showed more ups and downs across the week, and on days when the app showed worse cognition, people had MORE adverse driving events that same day (more hard braking, speeding, and sudden acceleration) across 20,000+ analyzed car trips. Interestingly, people seemed to implicitly recognize bad days and avoided risky nighttime driving without realizing why. Dr. Andrew Kiselica found variability scores are mostly noise in asymptomatic people but become highly informative once symptoms appear, suggesting smartphone monitoring may be best for the asymptomatic stage when frequent measurement averages out noise.
How does cognitive variability compare to expensive biomarkers?
Variability monitoring costs pennies per day on a smartphone versus roughly $5,000 for an amyloid PET scan or $1,500 for cerebrospinal fluid testing. It requires no invasive procedures, can be done at home, and generates many more data points per year than any clinical biomarker. Dr. Laiss Bertola validated the approach in 9,000+ Brazilians tracked over 8 years, finding that higher cognitive variability at baseline predicted higher odds of impairment 8 years later. For APOE4 carriers who want long-term tracking without repeatedly paying for expensive imaging, variability monitoring via standard neuropsych tests, smartphone apps, and remote cognitive batteries is a high-leverage option.
Why do APOE4 carriers have more cognitive variability than non-carriers?
The mechanism likely reflects the APOE4 brain's reduced metabolic reserve. APOE4 carriers have impaired glucose metabolism in memory-critical regions (20-25 percent lower in the hippocampus), reduced mitochondrial function, and heightened inflammatory responses. On days when sleep, stress, or glucose control are suboptimal, a non-carrier brain has spare capacity and performs normally. An APOE4 carrier brain has less spare capacity, so the same suboptimal day produces a measurable cognitive dip. This also means the variability signal gives carriers something actionable: by tracking which lifestyle factors correlate with their bad days, they can intervene on the specific drivers of their dips rather than chasing generic advice.
Keep reading

Related protocols.