Headlines in many countries now speak of a loneliness epidemic. The phrase carries weight: it suggests something widespread, rising and dangerous. Some of the evidence behind it is strong, some is shakier, and learning to tell the difference is a useful skill for reading any alarming trend.
The alarm
In May 2023 the United States Surgeon General, Vivek Murthy, issued an advisory titled Our Epidemic of Loneliness and Isolation. It described loneliness and isolation as a public health epidemic and cited surveys suggesting that about half of American adults reported measurable loneliness even before the pandemic. Later that year the World Health Organization launched a Commission on Social Connection. Its 2025 report estimated that about one in six people worldwide is affected by loneliness and linked it to roughly 871,000 deaths a year, though estimates of this kind rest on many assumptions.
Signs of decline
Some of the strongest evidence concerns time rather than feelings. The American Time Use Survey asks thousands of people each year to record how they spent the previous day. According to figures cited in the advisory, time spent with friends in person fell from about an hour a day in 2003 to about 20 minutes in 2020. The 2020 figure was depressed by the pandemic, but the decline was steep well before it.
Friendship surveys point the same way. In a 2021 survey by the Survey Center on American Life, the share of Americans reporting no close friends had risen from 3 percent in 1990 to 12 percent, while the share with ten or more close friends fell by more than half. Households have changed as well. About 29 percent of American households now hold one person, up from 13 percent in 1960, and similar shifts have occurred across Europe and East Asia.
Not only the old
Loneliness is often pictured as a problem of old age. In 2018 the BBC ran an online Loneliness Experiment that drew about 55,000 responses. The group reporting loneliness most often was people aged 16 to 24: about 40 percent said they felt lonely often or very often, compared with about 27 percent of those over 75. The sample was self-selected, so the exact figures should be treated with care, but other surveys in several countries have also found high loneliness among young adults.
Reasons for doubt
Skeptics raise several points. In 2006 a well-known study using the General Social Survey reported that the share of Americans with no one to discuss important matters had jumped from about 10 percent in 1985 to about 25 percent in 2004. It was widely quoted. The sociologist Claude Fischer and others argued that the jump was probably a quirk of the survey, caused partly by changes in how questions were asked and recorded, and later analyses largely supported that doubt.
Long-run data on loneliness itself are also patchier than data on behavior. Several studies of older adults in Britain and elsewhere found loneliness levels fairly stable across decades. Critics add that the word "epidemic" implies a sudden outbreak, while loneliness is as old as humanity, and that treating an ordinary feeling as a disease may push people toward clinics when what they need is company.
Reading the claims well
One way through the debate is to separate measures. Trends in isolation, such as time spent alone and the number of people living alone, are clearly rising in many countries. Trends in felt loneliness are less certain and depend heavily on survey wording.
It also helps to remember that living alone is not the same as being lonely. Older people in southern and eastern Europe, where living with family is more common, often report more loneliness than older people in northern Europe, where living alone is widespread. Expectations, health, money and the strength of community life all shape whether solitude feels like freedom or abandonment. A rise in one-person households may reflect wealth and independence as much as distress.
None of this makes the problem imaginary. A steady fall in shared time and close friendships deserves attention whether or not it fits the word "epidemic." The point is to weigh each claim on its own evidence.
Worked example: a doubling headline
Suppose a news site announces: "Loneliness has doubled in a generation." A careful reader asks four questions.
First, what was measured: a feeling, such as loneliness, or a behavior, such as time with friends? The headline says loneliness, but the article turns out to cite the growth in one-person households, which is a different thing.
Second, was the question asked the same way both times? If the earlier survey asked about "feeling isolated" and the later one about "feeling lonely," the comparison may say more about wording than about people.
Third, who answered? An online poll that invites lonely people to take part will find plenty of them.
Fourth, what do other measures show? If time-use data, friendship surveys and health studies all point the same way, the claim gains strength. Here, time-use data do show less time with friends, so there is a real trend, but "doubled" overstates what is known about the feeling itself.
