Chronic Loneliness Raises Dementia Risk By 38 Percent
A Universite de Bordeaux analysis of the Paquid cohort, published November 17 in Aging and Mental Health, found that older adults who reported loneliness at three or four visits over eight years developed dementia at a 38 percent higher rate, while occasional loneliness showed no association.
Edited and fact-checked by Stephen Sprague
The Bordeaux team's next question is whether breaking a chronic loneliness pattern changes the risk, and this cohort cannot answer it because it only observed the pattern.
Researchers at the Universite de Bordeaux followed older adults for three decades and separated loneliness into three patterns before they measured dementia. The study, published on November 17, 2026, in the peer reviewed journal Aging and Mental Health, uses the French Paquid cohort, which followed 3,777 older adults from 1988 to 2018, and loneliness was assessed repeatedly over eight years with a single question before participants were grouped as never lonely, occasionally lonely or chronically lonely.
Never lonely participants reported no loneliness at any visit and made up 48.3 percent of the cohort, occasional loneliness meant a report at one or two visits and covered 34.4 percent, and chronic loneliness meant a report at three or four visits and covered 17.3 percent. By 1995 and 1996, at the start of the dementia follow up phase, 1,540 participants remained eligible for the analysis.
Of those 1,540 eligible participants, 560, or 36.4 percent, developed dementia, and the incidence ran from 33.1 percent in the never lonely group to 49.1 percent in the chronic loneliness group, 131 of its 267 participants. After adjustment, chronic loneliness carried a hazard ratio of 1.38, a 38 percent higher risk, with a 95 percent confidence interval of 1.03 to 1.86, and occasional loneliness showed no significant association with dementia risk.
Co-author Karine Peres, a senior researcher at the Bordeaux Population Health Centre, said the findings challenge the assumption that any experience of loneliness poses equal danger to cognitive health. A brief lonely stretch after a move or a bereavement did not track with dementia in this cohort, and the persistent pattern did, which is the distinction the authors draw for prevention. The study is observational, so it shows an association across twenty years of follow up, and it cannot prove that loneliness itself caused the extra cases.
A single question cannot capture the quality of a person's relationships or whether they had meaningful activities in their lives, and the authors name that measurement as the study's main limit. The cohort is French and older, which bounds how far the result travels, and the Paquid data are available on reasonable request to the study coordinators at the Bordeaux Population Health Research Center. The work was supported by the Caisse Nationale de Solidarite pour l'Autonomie through a national research program on autonomy.
The conversation.
Loading comments…
A little more perspective.
Reading is free. Verified members can comment and reply for $2 a month.