Social isolation and loneliness each raised the probability of physical and mental illness among adults over 50, according to an 18-country study published in BMC Medicine on November 10.
Loneliness carried a far stronger association with mental disorders than isolation did — and the gap between the two numbers is where the clinical story lies.
The study pooled data from 94,122 participants aged over 50 across 18 countries. Previous research had linked disconnection from others to chronic illness, cognitive decline, and all-cause mortality, but had not investigated the effects of social isolation and loneliness on health in a multinational fashion.
What the researchers found splits into two distinct patterns. For physical disease, isolation and loneliness behaved almost identically. For mental health, they did not.
- Being cut off from other people — whether you feel lonely or not — was linked to a higher chance of developing physical illness in this study.
- Loneliness, the felt experience, showed a much stronger association with mental disorders than social isolation, the measured circumstance.
- The numbers describe probability: an association across large groups, not a prediction for any one person.
The study’s pooled average marginal effects, or AMEs, estimate how much more likely an outcome becomes for people exposed to a given factor. For social isolation, the probability of physical disease rose by 0.007 (95% CI, 0.003–0.011) and of mental disorders by 0.026 (95% CI, 0.014–0.037).
Loneliness tracked closely on the physical side: 0.011 (95% CI, 0.006–0.016). On the mental health side, it diverged sharply — a pooled AME of 0.243 (95% CI, 0.214–0.271), roughly nine times the isolation figure. In plain terms, people who reported feeling lonely had a substantially higher probability of a mental disorder diagnosis than people who were merely socially isolated.
It suggests the two conditions are distinct constructs: someone can be socially embedded and still feel alone, or isolated and content.
By drawing on 18 countries, the study’s design tests associations between social isolation and loneliness with physical and mental health in middle-aged and older adults across 18 different countries.
The study appears in BMC Medicine, published ahead of print on November 10. Yudong Miao, Zhanlei Shen, Dongfang Zhu and others conducted the analysis.
The study design cannot determine whether loneliness causes mental disorders, whether emerging mental illness makes people feel lonelier, or whether both flow from shared causes.
This is an observational study. The study’s mental-health association was the strongest signal in the data (pooled AME, 0.243).
- Miao Y, Shen Z, Zhu D, Zhang J, Wu J, Yang F, et al. BMC Med. 2026 Nov 10. doi: 10.1186/s12916-026-05369-8.