May 8, 2019   5 mins

It’s pretty much an uncontested truism that we are living through an “epidemic of loneliness”. Google searches for that exact term yield 100,000 results; you get another 100,000 for “loneliness epidemic”. It is a topic of regular concern both in Britain and the US.

The problem isn’t just that loneliness makes people sad; it’s also killing people, according to a 2017 report by the Jo Cox Commission. It claimed that loneliness (or “weak social connection”) is “as harmful to health as smoking 15 cigarettes a day”. It also suggests that it costs UK employers £2.5 billion a year.

But an “epidemic” is a disease that has spread suddenly – which by analogy implies a problem which has rapidly and recently increased in its magnitude. And I’m not at all convinced that is the case, or that the doomy claims about its health effects are well-supported.

One of my reasons for thinking this is a new study, available in preprint ahead of publication in the journal Psychology and Aging. It asks: has loneliness among older people got worse in recent years?

It found that “newer” old people (baby boomers born 1948-1965) are no more likely to think of themselves as lonely than “older” old people (born 1920-1947), and that older people have not become more likely to think of themselves as lonely in the last decade (2005 – 2016). There may be more lonely older people now, it says, but that’s because there are more older people in general. The average older person appears to be no more likely to be lonely than they were a decade ago.

I’m always wary of trusting any one study, though. All the more so when that study confirms my own beliefs. Also, it’s about elderly Americans, so its findings may not apply to, for example, younger Britons. But I think it is a good study: it has a large sample size and, most importantly, it has preregistered hypotheses, so that we know it probably can’t have p-hacked its way to exciting but false findings.

And my other reason for tentatively trusting it is that, everywhere else, the fact that loneliness is increasing just seems to be taken for granted. The BBC article I linked to earlier points to ONS data that finds that more older people are living alone, but it doesn’t say whether that’s just because there are more older people – and, of course, “living alone” is not the same as “lonely”.

The Jo Cox report says that nine million British adults are lonely, but doesn’t say whether that’s more than there used to be, and in fact says that “over the years, studies on loneliness have reached different conclusions” about whether it’s getting worse. Indeed, the report states that they have “found relatively consistent levels of chronic loneliness among older people – with between five and 15 per cent reporting that they are often or always lonely”. Which sounds to me like it’s not increasing.

Even scientific papers seem to have this problem. A paper in the Journal of Family & Consumer Sciences which calls loneliness “a private problem, a public issue” does the same thing, saying that loneliness is common, but making no effort to actually show that it has got more common. A letter in the Lancet does something similar, and quotes a magazine article in the Harvard Business Review as its evidence that it’s a problem.

I don’t claim that I’ve done a full systematic literature review, but my quick searches for “increasing loneliness” and “loneliness epidemic” on Google Scholar find lots of articles about how loneliness changes over a lifetime or how common it is now or how it correlates with various health outcomes, but I can’t see any that actually say that loneliness is more common now than it was 10 or 20 or 40 years ago.

You see surveys that find that young people are the loneliest age group, but I don’t know why that would be a surprise – surely most of us remember being lonely as teenagers, even those who seemed socially active from the outside. It doesn’t mean that people are getting lonelier, and it doesn’t seem to be a new thing.

The first study I found with a Scholar search of “loneliness and age” was from 1988 and found that high school students were the most likely to be lonely back then, too, long before Instagram or whatever we’re blaming it on these days.

And there’s also a lot of misinformation about it. The Jo Cox Report’s claim that loneliness is as bad as 15 cigarettes a day is, to put it baldly, false, even though it gets trotted out in reputable newspapers. The report references “Holt-Lunstad 2017”, which turns out to be a meta-analysis by Julianne Holt-Lunstad, which as far as I can tell makes no direct comparison to smoking. It does find that loneliness is associated with a 26% higher risk of death per year, but that’s about half as much as the increased risk of smoking between one and four cigarettes a day.

Furthermore, we can be extremely confident that smoking is the actual cause of many of those deaths. That’s not true of loneliness. It may be that, say, unhealthy people are less likely to get out and about, and so are more likely to be lonely. The Holt-Lunstad study tried to account for that sort of thing, but it cannot show causality, and there are loads of complex ways that loneliness and poor health could interact (poverty or substance abuse could lead to both loneliness and illness, for instance) that are not a simple story of “loneliness leads to death”.

As a nation, we are taking this “epidemic of loneliness” extremely seriously; last year, the Government actually appointed a “minister for loneliness” (I imagine she spends a lot of time in No 10, making sure Theresa’s all right). And we’re not the only ones. There was a plan to set up a fund, and for the ONS to develop a method of measuring loneliness. The US Surgeon General has declared an “epidemic of loneliness”. We’ve also got a Local Government Association guide to “combating loneliness” and a national “strategy for tackling loneliness” (which, again, seems to take it for granted that the problem has got worse, without actually making any effort to show as much).

And it’s part of a general feeling that society is more atomised – that we all spend our time scrolling through Instagram or playing video games, rather than sitting around a fire in the bosom of our family, sharing our jar of mead and wiping our elderly parents’ bottoms, as we did in the Before-Times. Things were always better in the past. But there’s not much evidence that technology is making us sadder, or lonelier.

That is not to say that being lonely is not terrible. And if there are simple, well-evidenced, cost-effective ways of reducing it, then great. But there are a lot of shaky – and in some cases plain false – claims going round about loneliness, and it’s worth addressing them.

First, I’m unconvinced (although I am open to being persuaded) that there is an “epidemic”, in the sense of a problem that has grown and spread. Maybe loneliness is an endemic disease in our society, but there doesn’t appear to be good evidence that it is worse than it was before. It could be that we’re just more willing to talk about it and acknowledge it as a problem – which may be a positive thing in its own right, of course; but it is not the same thing.

Second, the actual health impacts of loneliness are being greatly overstated. It’s not as bad as smoking by a long way – and, in fact, given the nature of the problem, it’s actually incredibly hard to tell whether “loneliness” is a cause of ill health, or a symptom, or a byproduct of the real cause, or what.

I’m not saying that we shouldn’t do anything about it. It could be that, even if it’s not getting worse, the problem is bad enough to be worth dedicating significant resources to (although, as always, every pound you spend reducing loneliness is a pound you can’t spend on neonatal intensive care units or diabetes drugs, so bear that in mind). But let’s have the conversation honestly, and stop saying “epidemic” and “as bad as smoking”.


Tom Chivers is a science writer. His second book, How to Read Numbers, is out now.

TomChivers