Researchers identify key proteins and pathways connecting social relationships to cardiovascular disease, stroke, and mortality, paving the way for targeted interventions.
Study: Plasma proteomic signatures of social isolation and loneliness associated with morbidity and mortality. Image Credit: David Michael Bellis/Shutterstock.com
In a recent study published in Nature Human Behaviour, a group of researchers investigated the proteomic signatures linking social relationships, particularly social isolation and loneliness, to health outcomes and mortality.
Background
Social relationships are essential for well-being and survival, yet social isolation and loneliness are linked to significant health risks, including morbidity and mortality, comparable to smoking and obesity.
These effects are mediated through mechanisms such as inflammation, antiviral responses, and dysregulation of the sympathetic nervous and hypothalamic-pituitary-adrenal systems.
Proteins, as functional components of biological processes and key drug targets, hold promise in explaining these links. Emerging evidence suggests proteomic changes may mediate the health effects of social relationships.
Further research is needed to identify specific proteins and pathways to improve disease prediction, prevention, and intervention strategies.
About the study
The United Kingdom (UK) Biobank cohort consists of over 500,000 participants aged 40-69 years, recruited from 22 centers in the UK between 2006 and 2010. Participants provided extensive data, including genome-wide genotyping, magnetic resonance imaging, electronic health records, and blood and urine biomarkers.
Ethical approval was obtained from the National Information Governance Board for Health and Social Care and the North West Multi-Centre Research Ethics Committee. Social isolation and loneliness were assessed using validated scales.
Social isolation was defined based on living arrangements, social contact frequency, and participation in activities, while loneliness was evaluated by feelings of loneliness and frequency of confiding in close others.
Proteomic profiling was performed on plasma samples from 53,026 participants using Olink Explore assays, yielding data on 2,920 proteins after stringent quality control. Genotyping was conducted on 487,409 participants using accurate methods to ensure high-quality data, retaining 9,910,057 Single-Nucleotide Polymorphisms (SNPs) and 337,138 participants post-quality control.
Structural Magnetic Resonance Imaging (MRI) data and blood biomarkers provided additional phenotypic endpoints.
Health outcomes were ascertained through linked health records, focusing on cardiovascular disease (CVD), dementia, type 2 diabetes(T2D), depression, and stroke. Covariates included demographic, behavioral, and physiological factors.
Statistical analyses incorporated proteomic and genomic data, identifying associations, co-expression networks, and potential causal relationships using Mendelian randomization and mediation models.
Study results
The primary study population comprised 42,062 participants from the UK Biobank, with a mean age of 56.4 years (±8.2), and 52.3% were female. These individuals had quality-controlled proteomic data and comprehensive behavioral data, including measures of social isolation, loneliness, and all relevant covariates.
Among them, 9.3% (3,905 participants) reported being socially isolated, while 6.4% (2,689 participants) reported feeling lonely.
Over a median follow-up of 13.7 years (±2.1), ending in November 2022, there were 2,695 cases of CVD, 892 cases of dementia, 1,703 cases of T2D, 1,521 cases of depression, 983 cases of stroke, and 4,255 deaths.
Proteome-wide association studies (PWASs) of 2,920 plasma proteins revealed 776 proteins significantly associated with social isolation and 519 proteins related to loneliness in models adjusted for age, sex, site, technical factors, and genetic principal components.
After additional adjustments for ethnicity, education level, household income, smoking, alcohol consumption, and body mass index, 175 proteins remained significantly associated with social isolation and 26 proteins with loneliness at the Bonferroni-corrected threshold.
Key findings included the strong association of growth differentiation factor 15 (GDF15) with social isolation and proprotein convertase subtilisin/kexin type 9 (PCSK9) with loneliness.
Notably, most identified proteins were positively associated with social isolation and loneliness, while a few, such as C-X-C motif chemokine ligand-14 (CXCL14), emerged as protective factors.
Shared proteomic patterns were observed, with 22 proteins common to both social isolation and loneliness. Protein-protein interaction networks revealed significant interactions among identified proteins, particularly in immune pathways and complement systems.
Further analyses highlighted the causal links between loneliness and five proteins, including adrenomedullin (ADM) and asialoglycoprotein receptor 1 (ASGR1), supported by Mendelian randomization and colocalization analyses.
These proteins were significantly associated with diverse health outcomes, including CVD, stroke, and mortality. Mediation analyses demonstrated that ADM notably mediated the relationship between loneliness and multiple diseases, including CVD, dementia, and mortality.
Conclusions
To summarize, using data from 2,920 plasma proteins in over 40,000 UK Biobank participants, this study identified proteins and networks associated with these social constructs.
Many proteins were linked to inflammation, antiviral responses, and complement systems, with more than half prospectively associated with CVD, T2D, stroke, and mortality over a 14-year follow-up.
Mendelian randomization analyses suggested loneliness causally influenced five proteins, with ADM and ASGR1 further validated by colocalization. These proteins also mediated relationships between loneliness and major health outcomes, underscoring their biological relevance.
New long COVID index highlights five symptom subtypes