Establishing a Baseline

What a baseline is

A baseline is a record of how things are before you change anything: your usual bedtime, your usual walking, how you usually rate your energy. Without it, you have nothing fair to compare against. It is the third step after defining a clear question and selecting a meaningful outcome.

This is information only, and it applies to ordinary habits. It does not set targets or normal ranges for anyone.

Why one reading is not a baseline

Bodies and habits vary from day to day. In a study of daily resting heart rate, most people's readings remained relatively stable over the short term, but 20% had at least one week in which their resting heart rate fluctuated by 10 beats per minute or more. [1] The researchers also found a small seasonal trend. [1] So a "before" number from one morning can look very different from your true usual level.

A single measurement also may not capture daily life. A medical guide notes that to find a resting heart rate you must have been resting for at least 10 minutes. [2] If a measurement depends on conditions like this, record the conditions every time.

Night-to-night swings are normal for sleep measures, too. One review noted substantial night-to-night variability even in sleep apnea severity, which has raised concern about misdiagnosis from a single night. [3] That is a clinical example, not a consumer one, but the lesson carries over: a single night or day is a weak basis for a conclusion. See sleep trends vs individual nights and baseline variation and normal fluctuation.

Building a useful baseline

• Record before you change anything. Keep your habits as they usually are while you observe.
• Record enough days to see your own range. The aim is to learn how much your outcome normally bounces around. There is no single correct number of days; more days show more of your normal variation.
• Use the same method each time. Same rating scale, same time of day, same device worn the same way.
• Note unusual days (illness, travel, late nights) rather than deleting them. See recording context and confounding factors.
• Do not start the change during a bad patch. An unusually bad or good period tends to be followed by readings nearer your average anyway; see regression to the mean in personal data.

Example: a bedtime routine

Suppose you want to compare a calm wind-down with your usual evenings. Your baseline is your usual evenings, with a morning rating of how rested you feel written down each day. Only after seeing how that rating naturally moves would you introduce the calmer routine. The NHS, for example, lists going to bed only when sleepy, waking at the same time every day and relaxing before bed as ways to help insomnia. [4] That gives context for the routine, but whether it works for you is the very thing the comparison tries to show.

When a baseline is a sign to see someone

If your baseline itself worries you, for example you are consistently exhausted, short of breath on ordinary walks or have chest discomfort, do not run an experiment. Speak to a clinician; see talking with a health professional.