Why you wake up tired: what a sleep test does and does not measure

You go to bed on time, you put in your hours, and still the morning is a blow. Usually that is not about the number of hours, but about what happens within those hours and about how little of it you take in yourself. How long you take to fall asleep, how often you wake up, how rested you get up and what the night costs you during the day: that is what decides whether you recover, and that is exactly what you remember badly. A sleep test measures how much these complaints trouble you, and gives you an indication of where you stand today in this area, the first of the 11 health areas the full scan assesses. This is not a diagnosis. It is insight, to see where you stand today. And to help you make better choices in your food and your lifestyle. What this scan does not do: measure a night, point to a cause or make a diagnosis. Only a doctor can do that last one.

Enough hours, and still tired: what may be going on

In the Netherlands, sleeping badly is the rule rather than the exception. In a nationally representative sample of 2,089 Dutch people between 18 and 70, more than three in ten reported a general sleep disturbance and more than four in ten reported too little sleep.⁠² Most of those people are not lying awake all night. They sleep, just not well enough to get up rested.

The hard part is that you do not really feel a shortfall like that. In a laboratory study, 48 healthy adults slept four, six or eight hours for fourteen nights in a row. At four and six hours the losses in performance piled up further every day, while after the first few days the participants hardly rated themselves as sleepier.⁠³ You get used to the feeling, while the shortfall stays.

More hours is not automatically the answer here. In a summary of sixteen population studies, both short and long sleep were linked to higher mortality: short sleep with 12 percent more, long sleep with 30 percent more.⁠⁴ That is a link, not a cause, but it does say that “longer” is a poor measure. Recovery is the measure.

Why how your night feels is a poor yardstick

With sleep, you are not there yourself. You know exactly how you feel during the day; what happened at night is something you have to reconstruct afterwards. Research into the perception of sleep shows that people with sleep complaints systematically overestimate how long they lie awake and underestimate how much they really sleep, compared with objective measurements.⁠¹

That is why no serious sleep instrument asks about one night. The Pittsburgh Sleep Quality Index, built in 1989 for research and the clinic, looks at a month and at seven components, from how long you take to fall asleep to how you function during the day.⁠⁵ A pattern over weeks, split into the components that matter: that is what you need in order to know whether things are getting better or worse.

What a sleep test measures

A sleep test such as the BioProfile sleep scan asks five questions about the past two weeks: how long you take to fall asleep, how often you wake up at night, how often you get up rested, how satisfied you are with your sleep, and how much poor sleep hinders you during the day. Each question is given points, the points are added up and converted into a sleep score from 0 to 100, with the reading band it falls into. Higher is better.

The questions are modelled on the Insomnia Severity Index, a scale of seven questions about the same two weeks. That scale was tested in 959 people from the general population and 183 people in treatment, with an internal consistency of 0.90 and 0.91,⁠⁶ and it moves in step with sleep diaries and with sleep studies.⁠⁷ Our five questions follow the construct and the look-back period of that scale, in our own words. They are not a copy of it, so the score is not a validated clinical score. What you get is a structured picture of your sleep burden, and a starting value you can measure again two weeks later.

What a sleep test does not measure

A questionnaire reads the sleep burden you describe. The cause of it stays out of view. It cannot tell you whether something is physically disturbing your sleep, such as your breathing, pain or hormonal changes, and it is not a sleep study.

A sleep test does not give a diagnosis either. The European guideline for insomnia puts questionnaires and sleep diaries at the front of the assessment by a doctor, alongside the conversation itself; a sleep study in the laboratory is there to detect other sleep problems if there is reason for it, such as pauses in your breathing during sleep.⁠⁸ So a questionnaire is a good first sieve, and a first sieve is something other than a result. A low score means that looking further makes sense, and nothing more. Loud snoring with pauses in breathing, or falling asleep unintentionally during the day, are reasons to go to a doctor and not reasons to work on a score.

What you do with your score

Most people start with the solution: to bed earlier, the phone away after ten, a new mattress. All sensible, and three weeks later you still do not know whether it did anything, because you never recorded where you began. A score taken today is that starting value. Repeat the test two weeks later, on the same questions, and you no longer have a feeling about your nights but two numbers.

If you want to change something in those two weeks, choose something whose effect has been measured. Coffee is one: 400 milligrams of caffeine, taken six hours before bedtime, still disturbed sleep measurably compared with placebo.⁠⁹ Screen light is a second: people who read on a light-emitting screen in the hours before bed instead of from a paper book took longer to fall asleep, produced less melatonin and were less alert the next morning.⁠¹⁰