Research
Everything the Center has published, most recent first. 257 items.
20 April 2026
A close look at how the American Psychological Association's 2026 review of AI companionship hedged its conclusions, and whether the hedge holds up against a longitudinal study published two months later.
19 April 2026
Social prescribing has become the default policy response to loneliness across several health systems, but the randomised evidence behind it is thin, and where it exists it does not favour the model most often funded.
16 April 2026
National estimates of loneliness among older adults rely on surveys that require a landline, an internet connection, or an address on file. That sampling frame tends to exclude the very people it is trying to count.
13 April 2026
A review of effect sizes, not p-values, across the loneliness intervention literature: what social prescribing, befriending, and structured behavioural programmes actually deliver, and how those numbers compare to the mortality risk they are meant to offset.
12 April 2026
AARP's 2018 and 2025 surveys of adults 45 and older used the same instrument, which makes them one of the few genuine before-and-after comparisons in the loneliness literature. Most other series cannot make that claim.
5 April 2026
A new longitudinal study finds chatbot use predicts later loneliness, and loneliness predicts later chatbot use. The finding is more precise than most coverage of AI companionship allows, and policy has not yet absorbed it.
31 March 2026
A study in Science this year ties remote work to rising isolation and mental distress. The proposed mechanism runs from time alone to loneliness to worse outcomes, and the evidence for each link is uneven.
24 March 2026
The Surgeon General's 2023 comparison of social disconnection to heavy smoking traces back to a single 2010 meta-analysis. What that analysis measured, and did not measure, matters more than the headline.
20 March 2026
A 2025 reanalysis of federal survey data split loneliness into severity tiers rather than a single yes/no measure, and the choice of where to draw the line changes the headline prevalence figure by a factor of two or three.
17 March 2026
Self-report loneliness surveys disagree by tens of percentage points depending on instrument and year. Time-use diaries and health-system administrative data tell a steadier, less noisy story about actual isolation.
16 March 2026
Young adults report the highest loneliness rates of any age group across nearly every major survey. The cross-sectional design behind almost all of that evidence cannot say whether this is a cohort effect, an age effect, or a measurement artifact -- and policy needs to know which.
6 March 2026
For six years running, the major reviews of social isolation and loneliness in older adults have flagged the same missing evidence: controlled trials of interventions. The gap has barely narrowed.
5 March 2026
The observational link between social disconnection and mortality is one of the most replicated findings in behavioral epidemiology. The intervention literature that should confirm it is thin enough that a null result would barely register — and that asymmetry is the story.
25 February 2026
The Cigna Group found 57% of Americans lonely; a HINTS-6 analysis found 37% moderately to severely lonely. Both are recent, both are national. The gap comes from what each instrument counts as loneliness, not from a change in the underlying reality.
24 February 2026
National strategies addressing youth loneliness lean heavily on the social-media explanation, but the survey evidence points to at least three other contributing factors that current policy does little to address.
21 February 2026
Most evidence on loneliness in older age comes from single-point-in-time surveys. A review of what longitudinal designs add -- and why the difference matters for policy aimed at this group.
13 February 2026
The 2022 HINTS-6 data, reanalyzed and published in 2025, split loneliness into four severity states rather than a binary yes/no. The change makes prevalence numbers harder to compare, not easier.
10 February 2026
A new study mines 2024 Household Pulse Survey data on loneliness and isolation by demographic group. Whether it can be compared to other prevalence figures, or even to itself over time, is a separate question.
7 February 2026
Survey after survey finds young adults the loneliest age group, yet the same data show them among the least structurally isolated. The gap is the finding, not a contradiction to explain away.
6 February 2026
Governments are writing technology-specific loneliness policy on the back of evidence that mostly cannot distinguish correlation from cause, or isolation from loneliness. That gap has consequences for what gets funded.
31 January 2026
A new cross-national study puts global loneliness patterns in one frame, but the headline average obscures how unevenly loneliness falls on young adults and low-income countries.
23 January 2026
A study published January 15 asks which socially isolated older adults go on to feel lonely and which do not, sharpening a distinction the field has argued for since 2010.
22 January 2026
National surveys consistently find older adults report less loneliness than the young, while clinical measures find them more structurally isolated. The instruments are measuring different things, and conflating them misdirects policy.
15 January 2026
Clinical recommendations increasingly call for routine loneliness screening among young adults, but the randomized evidence behind that recommendation was almost entirely generated in older-adult populations.