What Your Trend Line Means

Cognitive performance is inherently variable. Sleep, stress, caffeine, illness, time of day, and dozens of other factors all shift your scores from session to session. A single low score is almost always noise — it tells you nothing reliable about your underlying cognitive health.

The trend line smooths this noise. By fitting a statistical curve through your data over weeks and months, Keel reveals the signal beneath the day-to-day variation: whether your average performance is stable, improving, or declining. This is the information that matters.

This is not unique to Keel — it is how longitudinal cognitive research works. Studies tracking Alzheimer's progression don't look at single-session performance. They look at the rate of change over time, measured across many assessments. Keel gives you that same longitudinal view of your own data.

A stable trend (flat or very gradually sloping) with normal variability is the expected pattern for a healthy adult. Some gradual decline in processing speed and reaction time is normal after age 60 — the question is the rate. A slope that is consistent with known age-related norms across the population is not a red flag.

An upward trend over 4-8 weeks often reflects genuine cognitive improvement from a change in your life: starting an exercise program, resolving an underlying health issue, recovering from illness, or improving sleep quality. It can also reflect practice effects if you are early in your Keel history — Keel distinguishes these by the phase of your baseline development.

A downward trend that persists over 4-6 weeks — not explained by a clear acute cause — is the pattern that warrants attention. Not panic, but attention: logging what changed in your life around the time the decline began, considering obvious reversible causes (new medication, sleep disruption, significant stress), and, if the trend continues, discussing it with a healthcare provider.

Keel's trend visualization uses deviation from your personal baseline as the primary signal. Days within your normal range are unmarked. Days significantly below your baseline trigger a low-signal indicator — not an alarm, but a data point. Sustained sequences of below-baseline days are flagged more prominently because sustained deviation is more meaningful than isolated dips.

The trend confidence band (the shaded area around your trend line) widens when you have less data (early in your Keel history or after a long break) and narrows as your baseline becomes more established. A wide confidence band means interpret cautiously; a narrow band means the signal is more reliable.

A downward trend in Keel does not tell you why your cognition has changed. It cannot distinguish between a reversible cause (poor sleep, stress, medication side effect, illness) and a progressive cause (early neurodegeneration). This distinction requires clinical evaluation — a complete history, examination, and likely neuropsychological testing and imaging.

Keel's role is to surface the trend so you have data when you go to that conversation. Rather than telling a doctor 'I feel like I've been forgetting more lately' — a highly subjective and unreliable report — you can show them a chart of your cognitive performance over six months. That is genuinely useful clinical information.