Observational and genetic evidence disagree on the association between loneliness and risk of multiple diseases.
Yannis Yan LiangMingqing ZhouYu HeWeijie ZhangQiqi WuTong LuoJun ZhangFu-Jun JiaLu QiSizhi AiJi-Hui ZhangPublished in: Nature human behaviour (2024)
Loneliness-the subjective experience of social disconnection-is now widely regarded as a health risk factor. However, whether the associations between loneliness and multiple diseases are consistent with causal effects remains largely unexplored. Here we combined behavioural, genetic and hospitalization data from the UK Biobank to examine the associations of loneliness with a wide range of non-overlapping diseases. During a median 12.2-year follow-up, loneliness was associated with greater risks in 13 of 14 disease categories and 30 of 56 individual diseases considered. Of the 30 diseases significantly associated with loneliness, 26 had genetic data available for Mendelian randomization (MR) analyses. After Benjamini‒Hochberg correction and multiple sensitivity analyses within the MR framework, non-causal associations were identified between genetic liability to loneliness and 20 out of the 26 specific diseases, including cardiovascular diseases, type 2 diabetes mellitus, obesity, chronic liver diseases, chronic kidney disease, most neurological diseases and the other common diseases. Genetic liability to loneliness was only potentially causally associated with the remaining six diseases. Socioeconomic factors, health behaviours, baseline depressive symptoms and comorbidities largely explained the associations between loneliness and diseases. Overall, our study revealed a dissociation between observational and genetic evidence regarding the associations of loneliness with multiple diseases. These findings suggest that loneliness may serve as a potential surrogate marker rather than a causal risk factor for most diseases tested here.
Keyphrases
- social support
- depressive symptoms
- chronic kidney disease
- healthcare
- genome wide
- mental health
- type diabetes
- risk factors
- gene expression
- magnetic resonance
- machine learning
- dna methylation
- electronic health record
- risk assessment
- coronary artery disease
- artificial intelligence
- brain injury
- peritoneal dialysis
- glycemic control
- high fat diet induced