A five-item, two-week screen for subjective wellbeing. Free, fast and widely used, but easy to mistake for a diagnosis.
| Full name | World Health Organization Five Well-Being Index (WHO-5) |
| What it measures | Subjective wellbeing. Mood, energy and interest over the past two weeks |
| Items and time | Five items, about a minute |
| Response scale | Six points, from at no time (0) to all of the time (5) |
| Recall period | The last two weeks |
| Score range | 0 to 25 raw, reported as 0 to 100 |
| Developed by | WHO Regional Office for Europe, first published 1998 |
| Cost and licence | Free. No permission needed |
| Languages | More than 30 |
| Ages | Adults. Used with children from about nine years |
| Best used for | Frequent, repeated measurement. Screening for low wellbeing |
| Pair with | A broader measure of flourishing, and something that captures what people can do |
The WHO-5 is five statements about how you have been feeling over the past fortnight. It takes a minute, it is free, it exists in more than thirty languages, and it has been used in well over two hundred published studies. If you only know one wellbeing measure, this is probably the one.
And that is the problem with it, because it gets asked to do things it was never built for.
The instruction reads: please indicate for each of the five statements which is closest to how you have been feeling over the last two weeks.
Each is answered on a six-point scale: all of the time (5), most of the time (4), more than half of the time (3), less than half of the time (2), some of the time (1), at no time (0).
Notice that every statement is positive. There is no item about feeling low, tense or tired. The scale was designed that way on purpose, so that it measures the presence of wellbeing rather than the absence of symptoms, and so that people who are unwell in other ways can still answer it without it feeling like a depression questionnaire.
The WHO-5 came out of a WHO Regional Office for Europe project in primary care in the late 1990s, led by the Danish psychiatrist Per Bech. The starting point was a longer WHO wellbeing questionnaire, which was cut to ten items and then to five, keeping only the items that behaved as a single dimension. It was first published in 1998, and the Psychiatric Research Unit at the Mental Health Centre North Zealand, in Denmark, has looked after it since. The WHO now hosts the translations.
Add the five item scores. That gives a raw score from 0 to 25. Multiply by four and you have the percentage score from 0 to 100 that nearly all the literature reports, where 0 is the worst imaginable wellbeing and 100 the best. There is no reverse scoring and no weighting. If a person skips an item, the usual practice is to treat the questionnaire as incomplete rather than to prorate, because with only five items each one carries a lot of the score.
Three thresholds do most of the work. A percentage score of 50 or below (a raw score under 13) is the WHO’s own marker for poor wellbeing, and the point at which it recommends looking further for depression. Not diagnosing, looking further. A score of 28 or below has been used in the research literature as the level typical of people with major depression. And a change of ten points on the 0 to 100 scale is the usual threshold for a change that matters clinically. Three points is statistically detectable in a big enough sample and means very little to the person.
For context, general population means in European studies tend to sit well above the 50 line. The Danish general population mean is 70, and Topp and colleagues collected comparable figures from several other European countries in their review. When the scale is used as an outcome in a trial, Bech’s group argued that the sensible target is to bring people back to the general population mean, not just to show a difference from the control group. That is a useful habit for any wellbeing measurement: ask whether people are back in the normal range, not only whether the number went up.
In a group, I would always look at the share of people below 50 as well as the mean. A stable mean can hide a growing tail. In one workplace study of nearly 2,800 employees, about a third were below 50, which tells you more than the average would have.
Topp, Østergaard, Søndergaard and Bech reviewed 213 studies in 2015, and this is the paper to read if you want the evidence in one place. A few things stand out. Item response analyses (Rasch models, in younger and in older people) show that the five items form one dimension and that each adds something, which is what justifies adding them into a total. Internal consistency is commonly reported above 0.8.
As a screen for depression against a clinical interview, with the cut-off at 50, the pooled sensitivity across eighteen studies and 5,823 people was 0.86 and the specificity 0.81. It is responsive to change, which the review shows through six controlled trials across medicine where it was the outcome measure. And it has some predictive validity: cardiology outpatients scoring below 50 at baseline had higher mortality over the following six years than those at or above 50.
The field that has used it most is diabetes care, where low wellbeing turned out to predict poorer glycaemic control. It has also been used in adolescent psychiatry and in paediatric samples down to nine years, where it still screened reasonably (sensitivity around 0.75, specificity around 0.9 in one study of children and teenagers).
There is one WHO-5. There is no short form, because it is already the short form, and no long form in use. The WHO’s site holds the English original and translations in more than thirty languages, and the scale is free for anyone to use without asking. It has been used from around age nine upward, though the validation work is mostly in adults.
It is short enough to repeat every fortnight without anyone minding, which is rare. It is free and translated. It is unidimensional, so the total means something. It moves when things change. And because the items are all positive, it sits comfortably in settings where a symptom checklist would feel wrong, which is most workplaces and most products.
The WHO-5 measures how life has felt lately: mood, energy, rest, interest. It does not measure meaning, relationships, character, autonomy, whether a person is learning or growing, or how they judge their life as a whole. If you want flourishing in the fuller sense, this is one slice of it, and the slice most sensitive to the last two weeks.
Because so much of the literature uses it to find depression, people read a low score as a diagnosis, and it is not one. A score under 50 is a reason to ask a person how they are. In a workplace or a product, it should never become a label attached to an individual, and it should not be collected from people without their knowing what it is for.
Then there is the question that matters most when the people being measured are also using an AI system. The WHO-5 will faithfully register that people feel better. It will not register whether they feel better because something has got easier in a way that is good for them, or because effort and friction have been removed in a way that will cost them later. A system that takes work off someone’s plate will often lift their mood. The WHO-5 sees the lift and nothing underneath it.
A fortnightly self-report also cannot attribute a change to anything. Without a measure taken before the system arrived, or a group who are not using it, or repeated readings over time, a change in WHO-5 scores is a change, and that is all you can say.
In groups who are already doing well, scores bunch toward the top of the scale and there is little room to show improvement. If your baseline mean is in the high seventies, do not expect the WHO-5 to show you much good news.
I use it when I want a quick, repeatable reading of how a group is travelling and I intend to take that reading often. It is the right instrument for the short cycle, every two to four weeks, in a study that runs for months.
On its own it is a thermometer. It tells you the temperature and nothing about the cause. Pair it with something broader and slower for the long cycle (the Flourishing Measure is the obvious choice) and with at least one measure of what people can actually do, so that a rising WHO-5 is not mistaken for evidence that nothing is being lost.
The English original and the translations are hosted by the WHO. The English sheet is at cdn.who.int. The review is Topp, C. W., Østergaard, S. D., Søndergaard, S. and Bech, P. (2015), The WHO-5 Well-Being Index: a systematic review of the literature, Psychotherapy and Psychosomatics, 84, 167–176, doi.org/10.1159/000376585. It is open access.