Skip to navigation Skip to main content Skip to footer

Approved research

Social isolation and health - what are the mechanisms?

Principal Investigator: Professor Marko Elovainio
Approved Research ID: 14801
Approval date: December 1st 2015

Lay summary

According to the loneliness model of Cacioppo, lonely individuals would (1) engage in poorer health behaviours than others, (2) show altered cardiovascular activation, (3) show chronically elevated levels of hypothalamic pituitary adrenocortical activation, and thus lower glucose tolerance, diminished cortisol regulation, poor sleep and decreased cognitive performance, and (4) report more frequent mental health problems. All these four factors are risk factors for increased cardiovascular diseases, diabetes and mortality. This research projects aims to investigate the association between social isolation and health and test whether pathways including health behaviours, physiological factors, mental health and socioeconomic position would explain the associations Study aims to provide knowledge how social isolation affects health, which will help to create better prevention programs to heal individuals who are socially isolated. We will define ?social isolation? based on the following UK Biobank questions: 709 Number in household 6141 How are people in household related to participant 1031 Frequency of friend/family visits 6160 Leisure/social activities 2110 Able to confide An individual will be defined as ?socially isolated? if he/she lives alone, doesn?t participate in any social activities and can rarely confide to other people. The primary analyses will be performed on all-cause mortality and incidence of major diseases (e.g., coronary heart disease, stroke, cancer and diabetes). In secondary analyses, we will examine whether loneliness is linked with health outcomes (cardiovascular heart disease, diabetes and mortality) through pathways suggested by the loneliness model, including behavioural risks (drinking, smoking, physical activity and diet), physiological risks (blood pressure, obesity, handgrip cholesterol, sleep), mental health (depression and anxiety) and socioeconomic position (education, employment status, income). Tested using hierarchical regression models, structural equation modeling and bootstrap-models. Full cohort will be used.