Social Connection Deficits Predict Mortality and Physical Disease
Social isolation and loneliness are associated with higher all-cause mortality across many meta-analyses (Nakou 2025; Naito 2023; Rico-Uribe 2018; Harris 2023) and with coronary heart disease and stroke (Valtorta 2016); lower social-contact frequency was associated with higher mortality, with moderate effect sizes and null subgroups (Shor 2015).
None, informational
- Persistent loneliness or isolation accompanied by low mood, hopelessness, loss of interest, or withdrawal lasting 2+ weeks: encourage contact with a primary care provider or mental health professional
- Thoughts of self-harm or that life is not worth living: direct to crisis line / emergency services immediately
What the evidence says
The sources are meta-analyses of social isolation, loneliness and mortality or disease. Nakou 2025 (meta-analysis) reports loneliness (HR 1.14), social isolation (HR 1.35) and living alone (HR 1.21) associated with higher all-cause mortality, with substantial heterogeneity; Harris 2023 reports social isolation associated with 32% higher and loneliness with 14% higher all-cause mortality; Naito 2023 reports a pooled HR of 1.33 for social isolation; Rico-Uribe 2018 reports loneliness a risk factor for all-cause mortality (HR 1.22), slightly stronger in men; Schutter 2022 reports loneliness associated with mortality in older adults with small effect sizes; Yu 2023 reports social disconnection linked to a substantial increase in all-cause mortality; Leigh-Hunt 2017 (systematic review of reviews) reports significant associations of isolation and loneliness with mortality and of isolation with cardiovascular disease. Disease-specific: Valtorta 2016 reports poor social relationships associated with 29% higher coronary heart disease and 32% higher stroke risk; Meng 2024 reports poor social relationships associated with stroke and post-stroke mortality; Barth 2010 reports low functional social support relevant to coronary heart disease prognosis; Levis 2025 reports loneliness and low workplace social support associated with higher type 2 diabetes risk, with social isolation and low structural or functional support non-significant; Abreu 2024 reports worse social environment associated with cardiometabolic disease. On contact volume: Shor 2015 reports lower contact frequency associated with higher mortality (HR 1.13) with no significant relationship for males or for contact with family members, and Pinquart 2000 reports quality of social contacts more strongly associated with well-being than quantity. These are associations from observational studies pooled in meta-analyses.
- Myth: this is settled cause-and-effect. Reality: the cited meta-analyses report associations from observational studies.
- Myth: only the number of friends matters. Reality: perceived quality and feeling connected matter as much or more.
- Do not tell any individual that connecting will guarantee a longer life
- Do not present observational associations as established cause and effect
- Agreement
- Direction recorded for 0 of 14 sources; not yet recorded for the rest.
- The claim's outcome
- Outcome not yet recorded
- Certainty
- No source states a certainty rating.
What each source measured and concluded (14)
- What it measured
- All-cause and cause-specific mortality associated with loneliness, social isolation, and living alone
- How it was read
- screened only, not extracted
- What the review says it includes
- Systematic review, meta-analysis, and meta-regression
- Studies and participants, as stated
- 86 studies (from 11,964 identified)
- Heterogeneity, as stated
- Loneliness (HR 1.14, 95% CI 1.10-1.18), social isolation (HR 1.35, 95% CI 1.27-1.43), and living alone (HR 1.21, 95% CI 1.13-1.30) increased all-cause mortality; substantial heterogeneity
- What it measured
- All-cause mortality associated with social isolation and loneliness
- How it was read
- screened only, not extracted
- What it measured
- Mortality risk associated with loneliness and social network size
- How it was read
- screened only, not extracted
- What the review says it includes
- Systematic review and meta-analysis
- Heterogeneity, as stated
- Loneliness associated with mortality (HR 1.10, 95% CI 1.06-1.14) and larger social network size protective (HR 0.96, 95% CI 0.93-0.99); large heterogeneity
- What it measured
- Risk of stroke and post-stroke mortality
- How it was read
- screened only, not extracted
- What the review says it includes
- Systematic review and meta-analysis
- Studies and participants, as stated
- 19 studies; 1,675,707 participants
- What it measured
- Association of loneliness with all-cause mortality
- How it was read
- screened only, not extracted
- What the review says it includes
- Systematic review and meta-analysis (random effects)
- Studies and participants, as stated
- 35 articles; 77,220 participants
- What it measured
- Public health consequences of social isolation and loneliness
- How it was read
- screened only, not extracted
- What the review says it includes
- Systematic overview (systematic review of systematic reviews)
- Studies and participants, as stated
- 40 systematic reviews
- What it measured
- Association of social isolation and loneliness with mortality
- How it was read
- screened only, not extracted
- What the review says it includes
- Meta-analysis (news/summary report of 90 prospective cohort studies)
- Studies and participants, as stated
- 90 prospective cohort studies; more than 2 million adults
- What it measured
- All-cause mortality
- How it was read
- screened only, not extracted
- What the review says it includes
- Systematic review and meta-analysis of cohort studies (random-effects)
- Studies and participants, as stated
- 1.30 million individuals
- Heterogeneity, as stated
- I²=76%)
- What it measured
- Cardiometabolic disease (type 2 diabetes, CVD) incidence and prevalence
- How it was read
- screened only, not extracted
- What the review says it includes
- Systematic review and meta-analysis
- Studies and participants, as stated
- 281 studies; 1108 associations (384 in vote counting, 180 in 14 meta-analyses)
- What it measured
- Incident coronary heart disease (CHD) and stroke
- How it was read
- screened only, not extracted
- What the review says it includes
- Systematic review and meta-analysis of longitudinal observational studies
- Studies and participants, as stated
- 23 papers (16 datasets); 4628 CHD and 3002 stroke events
- What it measured
- Subjective well-being (SWB)
- How it was read
- screened only, not extracted
- What the review says it includes
- Meta-analysis
- Studies and participants, as stated
- 286 empirical studies
- What it measured
- Myocardial infarction (etiologic); cardiovascular and all-cause mortality (prognostic)
- How it was read
- screened only, not extracted
- What the review says it includes
- Systematic review and meta-analysis of prospective studies
- Studies and participants, as stated
- >1700 papers screened (etiologic and prognostic studies)
- What it measured
- All-cause mortality
- How it was read
- screened only, not extracted
- What the review says it includes
- Meta-analysis and meta-regression
- Studies and participants, as stated
- 187 risk estimates from 91 publications; ~400,000 persons
- What it measured
- Incident type 2 diabetes mellitus (T2DM) risk
- How it was read
- screened only, not extracted
- What the review says it includes
- Systematic review and meta-analysis (random-effects)
- Studies and participants, as stated
- 18 studies
Every source below is shown with its study design and links to the original study. This is the audit trail.