Think of this scenario. A person dies alone in their apartment, and no one notices for days or weeks. Now multiply that by 76,941. This is how many people the Japanese police handled in 2025 who had died alone at home, and 58,919 of them were 65 or older.
But there’s something happening, not one you may expect. Loneliness in Japan, measured across four decades, has been rising, yet not among the elderly. The highest reported levels of loneliness are among people in their 30s, and on the only measure that cleanly compares with other rich countries, Japan is the least lonely of them all.
But why does it still find so many of its people dead and undiscovered? The answer is the difference between a feeling and a circumstance, and it is where this story starts.
Quick answer
Loneliness in Japan rose between 1983 and 2023, according to a 2026 meta-analysis of 183 datasets. But the increase was concentrated among adolescents and women; adults and older people showed no significant change. On the government’s own annual survey, loneliness has been flat since 2021 at around 4.5%.
40 years of data, 183 studies, one clear trend
How do you measure something as private as loneliness across four decades? You basically need a question that has been asked in the same way long enough for the answers to be compared. In Japan, that comes from the UCLA Loneliness Scale. It has a series of prompts formulated in the 1970s and applied from then on.
In April 2026, two researchers from Chuo University in Tokyo compiled all the Japanese studies they could find that used that scale. They ended up with 81 studies and 183 individual data sets ranging from 1983 to 2023. On plotting these results, they noted the scores rose over the four decades, highlighting Japan’s loneliness crisis.
The trend survives the obvious objections. For instance, after the researchers excluded unusual studies that could have affected the results and considered only studies that had employed the same version of the questionnaire, the increase could still be seen. The same upward trend was also there prior to 2019. So, the climb had already been underway for decades even before the pandemic.
Table 1 — Where the 40-year rise actually happened
| Group | Correlation with year | p | Datasets | Verdict |
| Overall | r = 0.428 | < 0.001 | 183 | Significant rise |
| Adolescents | r = 0.239 | 0.014 | 104 | Significant rise |
| Adults | r = 0.001 | 0.998 | 22 | No change at all |
| Older people | r = 0.224 | 0.331 | 21 | Not significant |
| Women | r = 0.552 | 0.012 | 20 | Significant rise |
| Men | r = 0.312 | 0.181 | 20 | Not significant |
Source: Homma & Takase, Frontiers in Psychology, 2026. Note the small dataset counts for adults, older people and the sex-split analyses — absence of a significant trend there may reflect low statistical power, not a genuinely flat line.
However, the rise in loneliness didn’t affect everyone. Teenagers and women got lonelier, but adults, as a group, did not change at all over 40 years. Older people showed no clear change either.
That should not be read as proof that older Japanese people are fine. The researchers had far more data on teenagers than on any other group: 104 sets of data, against just 22 for adults and 21 for older people. With less data, it is much more difficult to see if there is a trend. The honest answer to those rows is ‘no clear evidence of a rise,’ but not ‘evidence of no rise.’
They also examined the factors that rose with loneliness during the same period (such as a decrease in family size, marriage rates, and online time). They stress that these are patterns moving in parallel over 40 years, not causes of it.
The rise started long before the pandemic
It would be convenient if COVID explained all of this, but it doesn’t.
As noted earlier, the Japanese loneliness data showed an upward trend long before the pandemic hit. This removes the most obvious explanation for a recent rise. Whatever has been driving loneliness in Japan was already doing so before anyone had heard of the coronavirus.
The pandemic did leave a mark, though. Average loneliness scores climbed between 2017–19 and 2020–22, a real and measurable step up. But a step is not the same as a staircase. If you focus only on the COVID years, you see a jump. Looking back over the 40 years that precede it, you can see a steady upward trend, which the pandemic accelerated temporarily.
This matters for how the problem is understood and discussed. If loneliness is viewed as a side effect of a pandemic, then a pandemic-size response is called for: wait for normal life to return, and the problem will go away. That’s not what the Japanese data show. The rise predates the virus, and the underlying forces were in motion long before 2020 and have not reversed since.
The rise came from teenagers — not from the elderly
The numbers do not fit the story Japan tells about itself. In the national imagination, loneliness is something that happens at the end of life. The figures say it occurs much earlier, and to people with decades still ahead of them.
The 40-year trend tells you who. The rise is mostly seen among teenagers; adults and the elderly account for none of it. On the study’s most consistent measure, there is no clear evidence that Japanese elderly people became lonelier across four decades. The group most associated with the problem is the one the trend line leaves flat.
Japan’s own national figures tell a similar story in sharper relief. The Cabinet Office runs a statutory survey of 20,000 people aged 16 and over every year (since 2021), drawn by random sample. In the 2025 wave, 6.1% of people in their 30s said they felt lonely often/always. Among people in their 70s, the figure was 2.5%, which is the exact inverse of the popular image.
Table 2 — Loneliness by age, and the inversion at the end of life
| Age | Often/always lonely (2025) |
| 16–19 | 3.3% |
| 20s | 5.5% |
| 30s | 6.1% |
| 40s | 5.7% |
| 50s | 5.7% |
| 60s | 4.0% |
| 70s | 2.5% |
| 80+ | 3.0% |
Source: Cabinet Office of Japan, 2025 wave, n = 11,873. People aged 65+ report the least loneliness of any adult group — and account for 76.6% of the country’s solitary deaths.
Loneliness goes up during the 20s and 30s, then remains stable in the 40s and 50s, only to diminish in the 60s and 70s.
Infographic 1: Japan’s loneliest people are in their 30s, not their 70s
Notably, the flat performance of older people is based on a much smaller number of studies than the rise among the young, and so should be interpreted as not implying that they are doing well.
That reframes the problem by shifting focus to Japanese youth loneliness. Which leaves the question: If the young are the loneliest group, why do the country’s lonely deaths fall so heavily among the old? We will answer that later.
Japan’s own survey says loneliness hasn’t moved since 2021
There is a second set of numbers on Japanese loneliness, and it does not agree with the first.
The 2025 wave of the statutory survey run by the Cabinet Office was fielded in December and published on 14 April 2026. This is the same period the Chuo University meta-analysis appeared.
Its central question is blunt. It asks people how often they feel lonely, and offers five answers: often, sometimes, occasionally, hardly ever, and never. In 2025, 4.5% said they felt lonely often or always. A further 13.7% said sometimes. The largest single group, 41.6%, said hardly ever.
The striking thing is not the level, but the flatness.
Table 3 — Japan’s own national survey shows no recent trend
| Year fielded | 2021 | 2022 | 2023 | 2024 | 2025 |
| “Often or always” lonely | 4.5% | 4.9% | 4.8% | 4.3% | 4.5% |
| UCLA-3, highest band (10–12) | 6.3% | 7.1% | 6.9% | 6.5% | 6.5% |
| Valid responses | 11,867 | 11,218 | 11,141 | 10,871 | 11,873 |
Source: Cabinet Office of Japan, National Survey on Loneliness and Isolation, waves 1–5. Random sample of 20,000 people aged 16+ each year.
The percentage of people reporting loneliness on a regular basis has fluctuated from 4.3% to 4.9%, which is within the survey’s margin of error, over five waves in five years. The second row tells the same story, but using a different instrument (a short three-item version of the UCLA scale). It has increased from 6.3% to 6.5% over 5 years. In both measures, there is no trend that indicates a worsening situation.
This is the awkward fact at the centre of the loneliness debate. The five years shown in the Japan loneliness survey results are precisely the period everyone assumes were the worst: the pandemic, the disruption that followed, and the period when loneliness became a named government priority. Yet on Japan’s own national instrument, nothing moved.
It is worth being clear about some things. It does not mean the 2021 figure of 4.5% is low in any absolute sense. One in twenty-two Japanese adults reporting constant loneliness is a substantial number of people, and the survey also finds that 37.7% report some degree of loneliness when milder answers are included. It also does not indicate that loneliness is not a problem. The government’s own solitary-death figures, which we will come to, make that impossible to argue.
What it basically means is narrower and more useful: the headline measure of loneliness in Japan has been stable across the most recent five-year window, even as the long-run academic trend points upward. Both of those things are true. The question is how they fit together, and the answer says as much about how loneliness is measured as about how lonely Japan actually is.
Why two credible sources can disagree this sharply
The datasets are sound since they come from competent institutions. So how can one show a 40-year rise and the other show a flat line?
First, they measure different things over different windows. The meta-analysis pools study averages across four decades and is heavily weighted towards adolescents: 104 of its 183 datasets. The Cabinet Office survey follows a random national sample across five years. A long, slow rise concentrated in teenagers is entirely compatible with no movement in the national average over a recent short window.
The second explanation relates to statistical power. Where the meta-analysis finds no change (among adults and the elderly), it is working from far fewer studies. It is also worth remembering that the flat results are the least well-supported findings in the study, not the most.
Thirdly, the loneliness statistics Japan produces may be partly shaped by how the question is asked. The meta-analysis’s own authors flag this directly. Two factors significantly affected the scores they analyzed: the number of items in the scale used and the choice of Japanese translation. For adolescents, using the full 20-item version had a larger statistical effect than the year of the study itself; the questionnaire mattered more than the decade.
The authors raise a further possibility, and they deserve credit for it. Japanese people may simply have become more willing to admit loneliness over the past 40 years. As the subject became a public issue, with a government minister appointed to tackle it and an Act of Parliament passed, the stigma around saying “I am lonely” weakened. Rising reported loneliness is not necessarily the same thing as rising loneliness.
On the one clean comparison, Japan is the least lonely country
Is Japan the loneliest country? You have probably gotten the idea that it is. That claim is repeated constantly, but also, on the best available evidence, wrong.
The problem with comparing loneliness between countries is that most of them ask the question differently. In October 2025, the OECD published the first clean cross-country figures. It identified four countries that ask the question the same way: ‘how lonely do you feel in general,’ which makes their answers directly comparable. The results are not what the stereotype predicts.
Table 4 — On the only clean international comparison, Japan is lowest
| Country | Feel lonely often or always, 2023–24 |
| Japan | 5% |
| England | 7% |
| United States | 12% |
| Canada | 14% |
Source: OECD, Social Connections and Loneliness in OECD Countries, 2025. These four countries ask the question with the same wording; most OECD members use a different reference period and are not comparable.
Japan is the lowest of the four. Nearly three times fewer Japanese people report frequent loneliness than Canadians. The country that supposedly invented the lonely death reports less loneliness than England, the United States and Canada on the one measure that allows a fair comparison.
Infographic 2: The two countries with a loneliness minister report the lowest rates
This finding needs to be handled carefully, because it is easy to misuse. It does not mean Japan has no problem. Rather, it suggests the country’s issue is not the same shape as many assume. Loneliness, as the Japanese population describes it in surveys, is comparatively low. Measured isolation (living alone, dying alone, and withdrawing from society entirely) is comparatively high. Those are two different things, and Japan scores badly on one and well on the other.
That is the paradox. A country can be less lonely than its peers and still have a solitary-death crisis, because loneliness is what people feel, and isolation is what happens to them. Japan’s numbers suggest a society where many people are not especially lonely in the emotional sense, but where a significant minority are profoundly alone in the physical one. And the two groups overlap less than anyone would like.
211 people a day are found dead alone at home
The number is the one that made this story travel. In 2025, Japanese police handled the bodies of 76,941 people who had died alone at home. That is 211 people a day, and 37.6% of all bodies the police dealt with that year.
For most of Japan’s modern history, this kind of death had no official name and count. It was a private tragedy. In 2024, the Loneliness and Isolation Act took effect, and the government defined it for the first time. In May 2025, it published its first official estimate: 21,856 deaths in 2024 discovered eight or more days after death, 80% of them men, and 70% of them aged 65 or over.
The 2025 figures reveal more. Approximately 22,222 of the total bodies found had remained undiscovered for at least 8 days. About 58,919 were aged 65+, up 875 from 2024. The rest were younger.
This phenomenon is called “kodokushi.” The word carries a particular weight in Japan. It describes dying alone and going undiscovered for some time, the interval between death and discovery being the thing that measures how far a person had fallen out of the world.
Those intervals are worth sitting with. Over 22,000 people were dead for over a week before anyone found them. That does not reflect how many Japanese people felt lonely, but how many didn’t have anybody who noticed.
This is where the loneliness statistics and the kodokushi, lonely deaths in Japan, part company, and it is the crux of the entire subject. The surveys tell us Japanese people report comparatively little loneliness. The police records tell us thousands die in conditions of total isolation. Both are true, and the gap between them is where the real story of Japanese loneliness in 2026 lives.
The people who die alone are not the people who report loneliness
Here is the contradiction in its simplest form. People aged 65 and over report the lowest loneliness of any adult group in Japan. They also account for 76.6% of the 76,941 people found dead alone at home in 2025.
The two facts sit side by side in government data, and neither is disputed. The proportion of people feeling lonely often or always was lowest in the 70s (2.5%) and highest in the 30s (6.1%) in the Cabinet Office’s 2025 survey. In the police figures, 76.6% of those who died alone at home were aged 65+.
This means that loneliness and isolation are two distinct issues. Loneliness is a feeling. It’s the gap between the relationships you have and the ones you want. Isolation is a fact about your circumstances: how many people you see, how often, and whether anyone would notice if you died.
Japan reports comparatively low levels of loneliness and comparatively high isolation. That combination is what makes social isolation in Japan so dangerous. An older person may genuinely not feel lonely. They may have made peace with a quiet life, or never expected much company in the first place, while living with almost no one to check on them. The feeling is manageable, but the circumstances are not.
This is why the country’s loneliness statistics, however carefully gathered, cannot measure the thing that kills people. A survey can ask how you feel. It cannot tell you whether anyone would find you.
Poor health raises loneliness 41-fold. Living alone raises it threefold.
Why are Japanese people lonely? The most useful data that answers this question is not about age or sex, but circumstances.
The Cabinet Office survey asks people what their lives are like, then cross-references their answers with how lonely they say they are. The results are not subtle. They show that loneliness in Japan tracks health and hardship far more closely than it tracks whether someone lives alone.
Table 5 — Who is actually lonely: the gradients are enormous
| Circumstance | “Often or always” lonely | Comparison group | Ratio |
| Health “not good” | 32.9% | “good” 0.8% | 41× |
| No one to rely on | 23.4% | has someone 2.5% | 9.4× |
| No one to chat casually with | 25.0% | has someone 2.5% | 10× |
| Rarely eats with anyone | 17.3% | eats with others daily | 6.4× |
| Finances “very hard” | 14.7% | “very comfortable” 2.9% | 5.1× |
| Unemployed | 11.0% | employed (regular) 4.5% | 2.4× |
| Divorced / unmarried | 9.1% / 9.0% | married 2.4% | 3.8× |
| Lives alone | 9.8% | lives with others 3.4% | 2.9× |
Source: Cabinet Office of Japan, National Survey on Loneliness and Isolation, 2025 wave. Ratios are our own arithmetic on published percentages.
Check the first row again. Adults whose health is “not good” are 41 times more likely to report constant loneliness than those who describe it as “good.” This difference is much bigger than the increase in loneliness experienced by those living alone, where it is less than 3 times.
The pattern repeats down the table. Having no one to rely on and chat with roughly multiplies loneliness by 9.4 and 10, respectively. Rarely eating with another person multiplies it by 6.4, while financial hardship increases it fivefold. Being unemployed roughly doubles it.
This reframes why Japanese people are lonely in the first place. The popular explanation is demographic, comprising an ageing society, smaller households, and more people living by themselves. But living alone turns out to be one of the weaker predictors in the data. What matters far more is whether someone is healthy, has money, and if they have even one person to eat a meal with.
That has an uncomfortable implication for policy. A government that treats loneliness as a housing problem, or as an inevitable consequence of ageing, will miss most of the people suffering from it. The Japanese data suggest loneliness is better understood as a health and inequality problem; one that arrives alongside illness, poverty and the loss of daily human contact, and often goes unnoticed until it is too late.
The next question is what is coming. On the government’s own projections, the number of Japanese people living alone is set to rise sharply.
44% of Japanese households will be one person by 2050
The loneliness figures describe the present, while household projections describe what is coming.
Japan’s National Institute of Population and Social Security Research publishes regular projections of how the country will live. In this case, Japan single-person households are expected to increase sharply. From 38% in 2020 to 44.3% by 2050, reaching 23.3 million households out of 52.6 million. Within that total, the share headed by someone aged 65 or above will rise from 34.9% to 46.5%.
The elderly living alone are the fastest-growing segment. In 2020, 7.37 million people aged 65 or over lived by themselves, 13.2% of all households. By 2050, that is projected to reach 10.84 million; 20.6% of households, a 47% increase.
Simply put, in a little over 20 years, 1 in 5 Japanese households will be a single person aged 65 or over.
None of these happens in a vacuum. It occurs in a country that is steadily reorganizing itself into smaller and smaller units, with more people living alone and fewer people to notice them.
The projections also complicate the policy response. A government can fund helplines and community programmes, and Japan does. But the trend line is about household structure (later marriage, lower birth rates, longer lives, and adult children living further from their parents). Those are not issues a loneliness minister can solve, because they are not primarily problems to do with loneliness. They are the shape of a society that has been changing for decades.
The coming shift from widowed to never-married
For most of modern Japanese history, an elderly person living alone was almost always someone who had been married and outlived a spouse. That is not the future the projections describe.
The proportion of the Japanese aged 65 and older who live alone and never married will increase significantly. For men, it’s projected to climb from 33.7% in 2020 to 59.7% by 2050. The figure will increase from 11.9% to 30.2% for women over the same period.
We can’t ignore the shift. A widowed individual living alone typically still has some children, siblings, or in-laws who monitor their situation and note when something is amiss. Those ties may be uncommon in the case of a never-married person. The high elderly single-person households of 2050 will increasingly be those with no close relatives at all.
The reason is demographic. Japan’s marriage rate has been falling for decades, and the cohorts now approaching old age contain far more people who never married than any previous generation.
Today, the widowed elderly are the last generation for whom old age in Japan generally meant family. The never-married elderly of 2050 will be the first for whom it generally means no one.
The second country on earth to appoint a Minister for Loneliness
Japan did not stumble into this problem unnoticed. It was one of the first governments in the world to treat loneliness as a policy issue, not a private sorrow.
In February 2021, the Cabinet Office established an Office for Measures to Combat Loneliness and Isolation and appointed a minister to lead it. Only the United Kingdom had done anything similar. The first Japan minister of loneliness met his UK counterpart online that June during Loneliness Awareness Week, a small piece of diplomatic theatre that signalled how seriously Tokyo was taking the subject.
The machinery followed. The Act on the Advancement of Measures to Address Loneliness and Isolation was enacted in 2023 and took effect from April 2024. It placed responsibilities on both the national and local governments. The councils also had to coordinate the provision of support, and a headquarters was established in the Cabinet Office headed by the Prime Minister and comprising all cabinet ministers.
The results have been mixed. The Cabinet Office’s own survey has shown no worsening of loneliness since 2021, and no improvement either. Meanwhile, solitary deaths remain above 200 a day. A minister and an Act can put loneliness on the map. They cannot single-handedly go back in time and reunite people.
Questions people ask about loneliness in Japan
Is Japan the loneliest country in the world?
No. Only 5% of the Japanese answered that they often feel lonely. This is the lowest of the 4 countries surveyed. In comparison, 14% of Canadians stated the same. The problem with Japan is not that its people feel lonely; it is that many people are isolated.
How many people die alone in Japan each year?
About 76,941 individuals. This is based on statistics presented in 2025 by the police. Of these, 22,222 were found eight or more days after dying, and 76.6% were 65 or older.
What is kodokushi?
It means 孤独死, “lonely death”, where a person dies alone at home and goes undiscovered for some time. There was no official definition until the Loneliness and Isolation Act took effect in April 2024; the first official count followed in May 2025.
Are young or old Japanese people lonelier?
Young people are. Those in their 30s report constant loneliness at 6.1%, compared to 2.5% for those in their 70s. The rise over the past 40 years was mainly among adolescents.
What is hikikomori?
This is severe social withdrawal in which the affected individual remains at home for 6 months or more without work or school. It’s estimated that 1.46 million people aged 15-64 fall into this category.
Has loneliness in Japan actually got worse?
To some extent. Over the past 40 years, academic studies show it has worsened. But over the last five years, the government’s own annual survey shows no change. Both can be true because they measure different things over different periods.
Editorial note: If you feel lonely, there is support. The Japanese government’s loneliness and isolation support site can provide you with information on services and advice, while TELL Lifeline provides free, confidential English language support where needed.