Most people, if you asked them to guess which generation reported the highest levels of loneliness in a large-scale national survey, would guess the oldest. The intuition is not unreasonable. The oldest generation has, on average, lost the most people. They are more likely to live alone. They are more likely to have retired from the working life that structured their weekdays for decades. They are more likely to have grown children who live in a different city, or country, or life stage. The popular image of the lonely person, in most cultures, is somebody in their seventies or eighties sitting quietly in a house that used to be full and is no longer.

What the largest published survey of loneliness in the United States found is the opposite. The loneliest generation, on the measurement instruments in wide clinical use, is the youngest one. And the gap is not small.

What the study actually measured

According to the 2018 Cigna U.S. Loneliness Index, an online survey of 20,096 American adults conducted by Ipsos on behalf of the health insurer Cigna between 21 February and 6 March 2018, using the twenty-item UCLA Loneliness Scale developed at the University of California, Los Angeles as the measurement instrument, the American population as a whole scored much higher on loneliness than any of the researchers had expected. Nearly half of the sample reported sometimes or always feeling alone. Nearly half reported sometimes or always feeling left out. A quarter reported rarely or never feeling as though anyone around them truly understood them.

The UCLA scale produces a numerical score from 20 to 80. The higher the number, the lonelier the person. Cigna’s researchers set a threshold of 43 as the point at which a respondent was classified as lonely. In 2018, 46 per cent of the survey population scored above it — not yet a majority, but close, and this was two years before the pandemic. (Separately, 54 per cent said they sometimes or always feel that no one really knows them well.) Within five years, the lonely share would cross into the outright majority.

The breakdown by generation was what surprised the researchers most. The Greatest Generation, aged 72 and older, scored a group average of 38.6. Below the loneliness threshold. Baby Boomers, aged 52 to 71, scored 42.4. Just below. Generation X, aged 38 to 51, came in at 45.1. Above the threshold. Millennials, aged 23 to 37, at 45.3. And Generation Z, then aged 18 to 22, scored an average of 48.3. Well above the threshold, and roughly ten points higher than the Greatest Generation their grandparents were part of.

Before continuing, we came across a video that you’ll find interesting if you relate to this article – regardless of age.

The pattern held after controlling for gender and race. It held across income levels. It held for people who described themselves as socially active as much as for people who did not. More than half of the Gen Z respondents identified with ten out of eleven of the specific feelings the UCLA scale associates with loneliness. Sixty-nine per cent said they felt that people were around them but not really with them. Sixty-nine per cent said they felt shy. Similar proportions reported feeling that no one really knew them.

The study did not, on its own methodology, establish why. Correlational surveys of this kind can measure a phenomenon but cannot cleanly identify what is producing it. What the researchers did rule out, using the same data set, was one of the most widely assumed candidate explanations. Social media use, which had been treated in popular discussion as the obvious cause of youth loneliness, was not correlated with loneliness scores in the survey. Very heavy users of social media scored 43.5 on average. People who reported never using social media scored 41.7. The difference between them was smaller than a rounding error on the overall population score.

What has happened since

The 2018 Cigna study, when its findings were announced, produced a small wave of media coverage and then, on the pattern of most public-health surveys, largely faded from the popular conversation. What has happened in the years since has quietly confirmed almost every part of the original result and, in some directions, made it worse.

According to the 2023 Vitality in America update published by the Cigna Group, drawing on continued survey data through the intervening five years, the overall American loneliness rate rose from 46 per cent in 2018 to 58 per cent by 2023, an increase of twelve percentage points over the five-year period covering the pandemic. The generational pattern held. Gen Z, by 2023 aged 18 to 26, remained the loneliest generation at 71 per cent. Millennials at 65 per cent. Gen X at 59 per cent. Baby Boomers, the group most affected by pandemic isolation, showed the sharpest year-over-year rise but still sat at 44 per cent, well below any of the younger cohorts. The Greatest Generation, on the follow-up data, remained the least lonely group in the country.

The pattern is not, on the continued survey record, an artefact of one particular moment in American cultural history. It has held across the strongest cultural disruption in living memory. Which means the finding it points at is real, well-measured, and probably not going away soon.

What the pattern does not do, on the primary-source record, is explain itself. The Cigna surveys measure. They do not explain. The candidate explanations that have been offered by researchers in the intervening years cover a wide range of proposed mechanisms. The specific developmental period Gen Z occupied during the pandemic, meaning the years of school and early independent living, may have been uniquely disrupted. The economic conditions this cohort faces, including housing costs, delayed household formation, and precarious employment, may be reducing the density of ordinary social contact this age group would historically have had. The specific way social interaction now happens, through digital rather than in-person channels, may not be substituting for the specific kinds of contact loneliness scales are measuring, even though the same surveys have failed to find a direct correlation between social media use and loneliness itself.

The honest answer, on the current state of the evidence, is that we do not fully know what is producing the pattern. What we do know, on the measurement, is that the popular image of the lonely person as an elderly widow in a quiet house is not matching what the surveys are consistently finding. The lonely person in the United States, on the accumulated survey record of the past six years, is more likely to be twenty-two, sharing a flat with two roommates, employed in some form, and reporting feelings of isolation the previous generation their age did not report at the same intensity.

Which leaves the whole picture with a quietly uncomfortable observation. Whatever combination of forces has produced the current generational pattern of loneliness in the United States has arrived, on the measurement, at exactly the group most modern culture has assumed would be least affected. The young adults who inherited the most connected communication infrastructure in history, who have more ways to reach other people than any previous cohort has ever had access to, who are the constant subject of concerns about being over-social rather than under-social, are also the loneliest group in the country. The two facts are not, on any straightforward reading of the data, contradictory. What they are, together, is a signal that something in the specific way this generation now lives is producing an outcome nobody was quite prepared for.

Kiran Athar is not a psychologist or a social scientist. She writes about mental health, culture, and the ordinary corners of life where the two intersect, drawing on peer-reviewed research and primary-source scholarship. If loneliness is affecting you or someone you care about, help is available in most countries through national mental health services.