Sleep mode
An optional per-patient mode that asks one question each morning — what time did you wake up — and shows you the regularity, never a score. No bedtime, no duration, no target.
Sleep is worth knowing about. Around 57% of people with eating disorders have sleep problems, insomnia is among the most commonly reported symptoms in both anorexia and bulimia, the relationship runs both ways, and patients with both tend to have more severe symptoms and a worse response to treatment.
Pippa is still a nutrition app, so this is a mode rather than a feature. Off by default, and off means there is no sleep surface anywhere in the patient's app.
What it asks
One question, once a day, in the morning:
morning 💛 when did you wake up? no wrong answer, and nothing is counted.
A time, or "not today". Skipping costs nothing and is never followed up.
This is the design, not a gap. Bedtime plus wake time is a duration, a duration invites a target, and a target invites a score — and a sleep score is a new number for a perfectionistic patient to fail at. Wake-time regularity is also the lever the intervention literature actually supports, so collecting less here is both the safer and the better-evidenced choice.
Why there is no score
The pattern has a name. Orthosomnia was described from a case series where chasing an ideal sleep score became the primary driver of the patient's distress: compulsive checking and sleep-related perfectionism, driven by the number a tracker produced. It correlates with perfectionism, obsessive-compulsive traits and health anxiety.
That is the same trait cluster that maintains an eating disorder. Handing this population a nightly grade is handing them a second thing to be failing at, and they are already fluent in that.
So the patient sees no score, no duration, no quality rating, no streak, no target and no comparison to yesterday. They log the time and Pippa says thank you.
Why there is no target bedtime
Adolescents phase-delay at puberty by up to two hours. Melatonin release shifts so the signal to sleep arrives around 11:30pm rather than 9pm, and sleep pressure builds more slowly through the evening. A target bedtime for a fifteen-year-old is a compliance metric they can fail for biological reasons, with no effort involved on their part.
The paediatric response to this was to move school start times, not to tell teenagers to try harder. It would be strange for an app to take the opposite view.
The coin
Logging pays the ordinary quiet participation coin, and it pays for logging, never for the time. Waking at 06:00 and waking at 11:00 pay exactly the same. Paying more for an earlier wake would turn this into the compliance scoring the mode exists to avoid.
What you see
A Sleep card on the patient page with the last fortnight of wake times, and the spread described in words:
- "Waking at a consistent time."
- "Wake time varies by about 45 minutes across most days."
- "Wake time varies by around 2.5 hours across most days."
The spread uses an interquartile range rather than earliest-to-latest, so a single early start for an appointment does not make a steady fortnight read as chaotic.
Regularity is the clinically meaningful signal here, and you can read a scatter without being handed a grade for it. Nothing on the card is styled as risk: an irregular week is information, not an alert.
One rule worth knowing
Nothing in this mode ever discourages eating late, and it never connects sleep to eating in either direction. No "eating late affects your sleep", ever.
Late eating is not a sleep-hygiene problem to solve here. Telling a patient in eating-disorder treatment not to eat before bed is prescribing restriction, and restriction is what drives the cycle the rest of the app works to interrupt.
Interactions
Sleep mode is orthogonal. Unlike structured, intuitive, ARFID and treatment mode — which are mutually exclusive and clear each other — sleep mode sits alongside any of them.
Android currently syncs the mode flag; the morning card is on iOS first.