Adult loneliness in 2026 — every statistic you need, sourced

If you've heard "loneliness is a public health emergency" and wondered whether the data actually backs the claim, this is the citation page. Every number here is from a primary source, with the link.

The case is more solid than I expected when I started looking. The convergence across independent datasets is unusual for a public-health story.

The Surgeon General's 2023 advisory

The single most authoritative U.S. document on this is the May 2023 Surgeon General Advisory titled Our Epidemic of Loneliness and Isolation. It's an 82-page report from then-Surgeon General Vivek Murthy. The headline claim:

> "Lacking social connection is as dangerous as smoking up to 15 cigarettes a day, and even more dangerous than the well-known risks associated with obesity and physical inactivity."

That mortality-equivalent comes from a 2010 meta-analysis by Holt-Lunstad et al., which pooled 148 studies covering 308,849 individuals. The effect size of social isolation on all-cause mortality was a hazard ratio of approximately 1.50 — translating to roughly the cigarette-equivalent the Surgeon General cited.

The 2023 advisory framed loneliness as a structural public-health problem, not a personal failing — comparable to how tobacco was reframed in the 1960s. It called for a "national strategy to advance social connection," explicitly naming urban design, workplace policy, education, and technology platforms as part of the solution if they're built to encourage off-platform interaction (and part of the problem if they aren't).

Pew Research: 1 in 5 Americans report no close friends

A 2024 Pew Research Center survey found that roughly 1 in 5 American adults reports having no close friends. The share has approximately doubled since the early 1990s. Among men under 30, the figure is closer to 1 in 4.

What's specifically growing: the share of people who say they have zero close friends. The "many close friends" category has shrunk too, but more slowly. The center of the distribution is sliding toward isolation.

The Pew data also breaks down by living situation. The starkest gap: adults living alone (a category that has grown enormously over the past 30 years) report close-friend numbers about half those of adults living with a partner or family. Living alone isn't the cause, but it's a strong correlate — and the share of U.S. adults living alone has gone from about 13% in 1960 to about 28% in 2025.

Cigna's Loneliness Index

Cigna has been running a U.S. Loneliness Index since 2018, using the UCLA Loneliness Scale (a validated 20-item psychometric instrument). The 2023 results:

• 58% of U.S. adults are considered lonely (UCLA score ≥ 43).
• 18–24 year olds report the highest rates — 79%. This contradicts the common framing of loneliness as primarily an elderly issue.
• Adults working remotely full-time report higher loneliness than hybrid or fully on-site workers, even after controlling for age and household composition.

The Cigna data is the strongest evidence for the "young adults are lonelier than the elderly" framing — a finding that surprises a lot of people but is well-replicated across multiple instruments now.

Holt-Lunstad et al.: the mortality data

The 2010 meta-analysis from Holt-Lunstad, Smith, and Layton is the foundational mortality paper. Pooled across 148 studies and 308,849 individuals, the headline finding:

> "Individuals with adequate social relationships have a 50% greater likelihood of survival compared to those with poor or insufficient social relationships."

The 2015 follow-up extended the analysis to 70 prospective studies and found the same effect size — social isolation increases mortality risk by 26%, loneliness by 26%, and living alone by 32%, independently of other risk factors.

These numbers are the basis for the Surgeon General's "15 cigarettes a day" claim. The math is roughly: smoking 15 cigs/day produces a hazard ratio in the 1.4–1.6 range; chronic isolation produces a hazard ratio of ~1.5. Same order of magnitude, same kind of population-level effect.

The "third place" decline

Sociologist Ray Oldenburg coined the term third place in 1989 — somewhere that isn't home (first place) or work (second place) but where casual social life happens. Bars, churches, civic clubs, barbershops, libraries, public parks.

Subsequent work (notably Robert Putnam's Bowling Alone, 2000, with periodic updates) has documented a steep decline in third-place participation across nearly every category in the U.S. since the 1970s. The most-cited stats:

• Membership in fraternal organizations (Elks, Rotary, etc.): down ~80% since 1970.
• Religious-service attendance: down ~30% since 1970.
• Bowling-league membership (the metaphor in the book title): down ~80% since 1980 even as bowling itself has remained popular.
• Average number of "non-relative people you discuss important matters with": down from 2.94 (1985) to 2.08 (2004) to lower numbers in newer GSS waves.

The drivers are complicated — TV, suburbanization, two-income households, longer commutes, screen time. But the macro-level effect is that the structural infrastructure of friendship has thinned out significantly over a generation.

The COVID-19 effect

Pre-pandemic baselines from 2018 already showed elevated loneliness. The 2020–2022 lockdowns made it worse. But the surprising finding is that post-pandemic loneliness has not fully reverted even though lockdowns ended.

Cigna's index in 2018 was 54%. In 2020 (the pandemic year) it spiked to 61%. By 2023, with lockdowns long over, it had only come back down to 58% — significantly above the pre-pandemic baseline.

The most-cited explanation: the pandemic didn't just cause temporary isolation, it changed durable behaviors (remote work, meal delivery instead of restaurants, streaming instead of cinema) in ways that reduced the ambient amount of low-stakes social contact.

What to do with these numbers

Two things.

First: if you're feeling lonelier than you used to, the data says it's not just you. Roughly half the adults around you feel similarly, and the structural drivers are real. That's not an excuse not to do anything about it — but it should kill the shame, which is itself one of the bigger barriers to action.

Second: the fixes are mostly small and recurring, not big and dramatic. Show up to the same yoga class for three months. Join the recurring book club, not the one-off speaker event. Take the same walk at the same time every week and nod at the same neighbors. Friendships and weak-tie networks rebuild from the bottom up, not the top down.

If you want a structured tool that helps with the rebuilding, Are We Friends? is the friendship app I built around exactly this research — personality-matched introductions, AI-planned first hangs, end-to-end encrypted DMs. The full product story is at /loneliness and /how-it-works.

But you don't need an app for the basics. Show up. Show up again. Make the second ask. Repeat.