The most useful sundowning tracker is a written log of what happened just before the episode — the Alzheimer's Association's own tip sheet tells caregivers to make notes about what happens before sundowning events and use them to identify triggers. Beyond that notebook, three tools do real work: a printable A-B-C chart to structure the notes, the NPI-Q questionnaire to turn weeks of observation into scores a clinician can read, and wrist-worn activity monitors as supporting evidence.
Sundowning is the late-afternoon and evening rise in confusion, agitation, restlessness or distress common in dementia — the Alzheimer Society of B.C. estimates about 66% of people living with dementia experience it. Tracking matters because the pattern, not the single bad evening, is what points to a fixable cause.
Medical information disclaimer: This article is for general educational purposes only and does not provide medical advice, diagnosis, or treatment. Always consult a physician or other qualified health professional about symptoms, medications, tests, or treatment decisions.
Table of Contents
- Start with a time-and-conditions log
- The A-B-C chart, and where to get a free one
- Turning your notes into something a clinician can use
- Do wearables and sensors help?
- What to rule out before blaming the hour
- Frequently Asked Questions
Start with a time-and-conditions log
The lowest-tech tool is the one both major organizations recommend first. The Alzheimer Society of B.C.'s sundowning handout directs caregivers to record the time of day the behaviour occurs and the conditions just before or during it, so patterns can be matched to causes you can actually change. In practice that means noting a handful of things each evening: the clock time the restlessness started, what room the person was in, the light level, whether anyone had just asked them to do something, whether a meal or nap had happened, and who else was present.
Two weeks of those notes usually reveals something a single day cannot — that episodes begin when the overhead lights go on, or always follow a 4 p.m. bath, or cluster on days with afternoon visitors. Keep it short enough that you will actually do it at the end of a hard day. A phone note or a kitchen calendar works as well as anything purpose-built.
The A-B-C chart, and where to get a free one
The standard structured format for behaviour logging is the A-B-C chart: Antecedent, Behavior, Consequence. The Penn Memory Center's caregiver course supplies a blank chart and defines antecedents as what happened immediately before the behaviour, explicitly including demands placed on the person. That last point is the one caregivers most often miss. A request to change clothes, take a pill, or come to the table is an antecedent, and recording it is how a "random" 5 p.m.
outburst turns out to be a predictable response to being hurried. A ready-made form saves you designing one. The Registered Nurses' Association of Ontario publishes a free printable A-B-C charting template with date/time, antecedent, behaviour and consequence columns in its long-term care toolkit. This is not only record-keeping. A study of assisted-living staff trained in the ABC approach, published in an NIH-hosted peer-reviewed report on the STAR program, significantly reduced residents' affective and behavioral distress — the act of structured observation changes how caregivers respond.
Turning your notes into something a clinician can use
A log of evenings is hard for a doctor to absorb in a 15-minute appointment. The validated tool for that handoff is the NPI-Q, a caregiver-completed questionnaire in which the informant rates each neuropsychiatric symptom's severity from 1 to 3 and their own distress from 0 to 5. It is designed for exactly this constraint.
The original validation study by Kaufer and colleagues in the Journal of Neuropsychiatry and Clinical Neurosciences reports that it finishes in five minutes or less, and that its severity total correlated 0.91 with the much longer clinician-administered Neuropsychiatric Inventory. A blank NPI-Q form is distributed free by the Alzheimer's Association. Filling one out before an appointment means arriving with numbers for both the symptoms and your own strain, rather than trying to summarize a month of evenings from memory.
Do wearables and sensors help?
They can, as supporting evidence. A 2024 feasibility study in Alzheimer's & Dementia, involving agitated late-stage dementia care-home residents, reported 96.6% acceptance of wrist-worn actigraphy and 88.0% wear compliance over four weeks — so the devices stay on, which had been the open question. Accuracy improves when devices are combined rather than used alone.
In the MODERATE study at Oregon Health & Science University, detection of nocturnal agitation reached an AUC of 0.74 with a wearable alone, 0.71 with passive infrared motion sensors alone, and 0.83 when both were combined. The limit is important: this research measures agitation in general, in small samples, and no validated consumer instrument or cleared device targets sundowning specifically. Treat wearable data as a timestamped supplement to your log, not a replacement for it.
What to rule out before blaming the hour
A pattern in your log is most valuable when a clinician uses it to check for a physical cause. Mayo Clinic's guidance on sundowning and the BrightFocus Foundation's tips for managing it point to a short list worth excluding first: Any of these can produce evening agitation that looks like sundowning and resolves once treated. Bring the log and an NPI-Q to the appointment and ask directly whether an infection or pain source has been ruled out before a behavioural medication is considered.
- Pain, including untreated arthritis or dental pain
- Infection, especially a urinary tract infection
- Hunger or thirst late in the day
- Fatigue from a long day or a disrupted nap schedule
- Poor or dim lighting as daylight fades
Frequently Asked Questions
How long should I track before looking for a pattern?
The 2024 actigraphy feasibility study ran four weeks, and that is a reasonable benchmark. Two weeks of written notes is usually enough to see whether episodes cluster by time, room or activity.
Is there an app made specifically for sundowning?
No validated consumer instrument or cleared device targets sundowning specifically. A general note-taking app holding A-B-C entries is as good as anything marketed for it.
Should I record the consequence as well as the trigger?
Yes. The A-B-C format includes it because what happened after — you left the room, offered food, turned on a lamp — tells you which responses actually shortened the episode.





