How is loneliness measured as a public-health number?

What is loneliness, and how is it different from social isolation?

The World Health Organization defines loneliness as "the painful feeling that arises from a gap between desired and actual social connections". It defines social isolation as an objective lack of enough connections.

The European Commission's Joint Research Centre draws the same line. Isolation is an objective state, such as the number of contacts a person has. Loneliness is a subjective experience.

Which questions and scales measure loneliness?

The UK Office for National Statistics recommends four questions; where there is room for only one, a direct question: "How often do you feel lonely?" The answers are: often or always, some of the time, occasionally, hardly ever, never. This is the wording of the Community Life Survey. It is a recommendation for surveys, not a diagnostic tool.

The Office for National Statistics also recommends three items based on the UCLA Loneliness Scale (Hughes and colleagues, 2004). They ask how often a person lacks companionship, feels left out, and feels isolated from others. Each answer scores 1 to 3, so the total runs from 3 to 9. There is no cutoff for "lonely".

The three-item scale was tested in the US Health and Retirement Study in 2002, with 2,182 respondents. Its internal consistency (alpha) was .72. In a second study it correlated .82 with the full revised UCLA scale, which means the three questions ranked people much as the longer scale did. Both tests were on US adults aged 50 and over.

A 2022 review of loneliness scales by Maes and colleagues notes that the original UCLA scale has 20 items. Short forms exist with 8 items (two different sets), 4 items and 3 items. The review warns: "It is not always clear which items were used in a particular study."

How do countries and global bodies count lonely people?

The WHO Commission on Social Connection reported in 2025 that 1 in 6 people worldwide experience loneliness — a modelled estimate for 2014–2023, people aged 13 and over. Among people aged 13 to 29 the share is 17–21%. It is about 24% in low-income countries and about 11% in high-income countries. These figures are modelled from 23 data sets that asked in different ways, with uncertainty intervals; for many countries there is only one source.

Meta and Gallup asked people in 142 countries between June 2022 and February 2023. About 1,000 people aged 15 and over answered in each country (500 to 3,000 in some). 24% said they felt very or fairly lonely. People aged 19 to 29 were lonelier than people aged 65 and over. This rests on one question, “In general, how lonely do you feel?”

The OECD listed 49 national surveys in OECD countries in 2024; loneliness was included in 39% of them. This was an incomplete list of surveys, not a check of their quality.

Why do the numbers differ between surveys?

The OECD found little agreement on indicators. Surveys vary in phrasing, answer options and the period people are asked to recall.

About the three-item scale, the UK Office for National Statistics states: "There is no standard accepted score for which a person would definitely be considered lonely." It also says some groups, such as men, may under-report loneliness on the direct question. It still calls that question the best measure available. The size of the under-reporting is not given.

The Joint Research Centre notes that loneliness can be measured directly, by asking about feelings over a period, or indirectly. Because of stigma, people may under-report when asked directly. The size of this effect is not given.

The review by Maes and colleagues (2022), a content analysis of the eight most-used loneliness scales, found that all of them include items that do not measure loneliness. In the UCLA scale there are two such items. That item analysis looked at wording; the authors summarised the scales' measurement properties separately, based on published data.

This is why the two world figures above should not be set side by side: the WHO's 1 in 6 is a model of 23 data sets for 2014–2023 from age 13, partly built on Meta-Gallup data; Meta-Gallup's 24% is one survey in 2022–2023 from age 15.

The link between isolation and early death is covered on our separate page, “Weak social ties — few contacts or loneliness — are linked to earlier death”.