Loneliness → All-cause mortality

Mechanism

Loneliness is the feeling that contact falls short of what a person wants. Isolation is a count of contacts. Living alone is a household. The three travel together. Wang, Gao, Han and colleagues (2023) pooled 90 prospective cohorts, 2,205,199 adults aged 18 or older (PROSPERO CRD42022299959). In the general population, loneliness tracked higher all-cause death (pooled effect size 1.14, 1.08 to 1.20) and higher cancer death (1.09, 1.01 to 1.17). Isolation was larger on all-cause death (1.32, 1.26 to 1.39) and also on CVD death (1.34). the record here is The 2.2 million is the whole review. Nakou, Dragioti, Bastas and colleagues (2025) stayed with older adults. Loneliness and all-cause death: 52 studies, 308,948 people, HR 1.14 (1.10 to 1.18). Isolation 1.35. Living alone 1.21. Heterogeneity was high (I-squared 84 percent). A prediction interval ran from 0.89 to 1.46. The Egger test flagged publication bias. Longer follow-up gave a smaller loneliness estimate. Yu, Steptoe, and Chen (2022) followed 35,254 people in the Chinese Longitudinal Healthy Longevity Survey (mean age 86.6) for a median 4.8 years. Isolation tracked death (HR 1.22, 1.18 to 1.25). Loneliness did not after health and low psychological well-being (HR 1.01, 0.98 to 1.04). Under age 80 it did (HR 1.15, 1.05 to 1.26). the record here The paper names N=35,254. Kraav and colleagues (2021) followed 2,588 Finnish men aged 42 to 61 for 23.2 years. Loneliness predicted all-cause death after the full covariate set. Cancer and CVD death moved with loneliness until lifestyle or HPL depression scores went in. Isolation predicted all-cause and injury death. That isolation path ran through loneliness and depression scores. the record here The paper names N=2,588. Authors name stress physiology, health behavior, and mood. The files do not isolate one path.

Caveats

The stamp is smoke because nobody randomized loneliness. These are cohorts and metas of cohorts. Observation alone cannot close the backdoor, so the stamp stays smoke. More papers of the same design do not settle the claim. A larger meta is still observation. Wang's 2.2 million mixes isolation with loneliness. This page is the loneliness arrow. Isolation is a different construct. So is living alone. Reverse causation sits on the backdoor: illness can cut contact and raise both the feeling and later death. Yu's very-old null after health and well-being is that problem in one sample. Nakou's prediction interval crosses 1. Self-report scales differ. Residual confounding from income, education, smoking, and drinking remains. Kraav's cause-specific associations moved when depression or lifestyle went in. This is a journal entry, not medical advice.

Effect

qualitative. Wang 2023: loneliness all-cause pooled ES 1.14 (1.08-1.20), 90 cohorts. Nakou 2025 older adults HR 1.14 (1.10-1.18), n=308,948. Sign is more loneliness, more death. Design does not identify.

Nodes

Named confounders

• Social capital
• Depression
• Disposable income
• Educational attainment
• Alcohol intake

Papers

• A systematic review and meta-analysis of 90 cohort studies of social isolation, loneliness and mortality.

  F Wang, Y Gao, Z Han, Y Yu, Z Long, X Jiang, Y Wu, B Pei · 2023 · Nature human behaviour

• Loneliness, social isolation, and living alone: a comprehensive systematic review, meta-analysis, and meta-regression of mortality risks in older adults.

  A Nakou, E Dragioti, NS Bastas, N Zagorianakou, V Kakaidi, D Tsartsalis, S Mantzoukas, F Tatsis · 2025 · Aging clinical and experimental research

• Social isolation, loneliness, and all-cause mortality: A cohort study of 35,254 Chinese older adults.

  B Yu, A Steptoe, Y Chen · 2022 · Journal of the American Geriatrics Society

• The effects of loneliness and social isolation on all-cause, injury, cancer, and CVD mortality in a cohort of middle-aged Finnish men. A prospective study.

  SL Kraav, O Awoyemi, N Junttila, R Vornanen, J Kauhanen, T Toikko, SM Lehto, S Hantunen · 2021 · Aging & mental health