One in five Americans was lonely 'a lot of the day yesterday.' That number belongs on every health dashboard.

Gallup's latest well-being report, published today, is headlined by depression — 19.1% of US adults say they currently have or are being treated for it, roughly 51 million people, nearly double the 2015 figure — but the line we watch is the one underneath. The share of adults who felt lonely "a lot of the day yesterday" ran about 23% at the height of the pandemic, fell to 17–18% in 2022 through 2024, and has now leveled off at around 20%, where it stood in February. The two measures travel together: a third of people who were lonely yesterday report depression, against 13% of those who were not. The survey is a probability panel of 5,017 adults with a margin of error of about 1.4 points; it is self-report, one question, one day. It is also the most regular measurement of the fundamental with the strongest mortality evidence, and almost nobody tracks it.

Who is lonely. The report's depression breakdowns are the clearest signal: rates have risen most among adults under 45 — 18-to-29-year-olds went from 13% to 28% since 2017 — and among the lowest-income households, from 22% to 37%, while adults over 65 are the only group that improved. Gallup also cites its international comparison: young American men are two-thirds more likely than their peers in other rich countries to report significant loneliness on a given day. The pattern matters for this site because it is the inverse of the longevity customer. The people the industry sells to are older and affluent. The people whose connection is collapsing are young and poor, and no subscription is aimed at them.

Loree’s take

If I could put one number next to blood pressure on every patient's chart it would be this one, asked the way Gallup asks it: were you lonely a lot of the day yesterday? It predicts what I will see in the next five years better than most of the panel. The 20% is a population figure; the version that matters is the patient in front of me, and the question takes six seconds.

Ben’s take

There is a dashboard for everything in this industry — sleep stages, glucose, VO2 max, biological age — and not one for the measure with a 148-study meta-analysis behind it. That is not an oversight. A loneliness number cannot be improved by a purchase, so no product has a reason to show it to you. Gallup measures it every month. Treat it as the benchmark the rings left out.

Why it matters to a recovering optimizer. Twenty percent is the public number. The private one is yours, and it does not need a survey. If the honest answer to Gallup's question is yes more days than not, that is the finding to act on this month, ahead of anything a device has flagged. The intervention is a person.