Understanding Social Isolation in the Ageing Population
of the UK

Social Isolation has been a renewed topic of discussion over the past few with the pandemic setting a toll on the mental health of people. Unlike loneliness, which is often used to describe the emotional or mental state of an individual, social isolation is used to describe the physical isolation of an individual, that is it defines how much an individual mingles or maintains relations with the people around him. Social isolation has been linked to several adverse health effects, both physical and mental, and many of these effects happen independently of loneliness. For instance, this has been linked with increased risk of cardiovascular diseases and mortality.
My research aimed to make use of the English Longitudinal Study of Ageing (ELSA) dataset to understand the association social isolation had with demographic factors like age, sex and levels of education and behavioural factors like alcohol consumption and internet use.

Evaluating Social Isolation

The ELSA dataset did not have a specific variable that quantifies social isolation. The index I used here is based on the works of Steptoe et al., where they created a score which ranges from 0 to 5, where a score of 5 indicated extreme isolation and 0 represented a non-isolated individual. The score was constructed by assigning a point if the participant was not married or living with anyone, one each if they maintained less than monthly contact with friends, family, and children (This included emails/texts, telephonic and direct communications) and a final one if they were not part of any society or clubs. This index was then further recoded into two categories for our analysis, where a score of zero was kept as ‘Not Isolated’ and a score of 1 was considered as ‘socially Isolated’.

Chi-square tests were used to understand if social isolation is experienced equally in groups sorted based on various demographic and socio-economic factors discussed above. A binary logistic regression was then done to see if these factors could be used to classify people into isolated and non-isolated groups.
The study revealed significant levels of social isolation among aging individuals in the UK, with some experiencing extreme isolation.

The study confirms that social isolation tends to vary with age, with the highest share of isolated individuals belonging to the age group 60-75. However, this is partly due to the higher number of participants in this age group. The proportion of socially isolated individuals increases across age groups, with those over 90 having the highest proportion. The regression analysis showed that individuals over 90 are 3.6 times more likely to be isolated than those aged 60-75.
Regarding gender, the chi-squared test aligned with previous studies, showing that different genders experience social isolation differently, with females having the highest proportion of isolated individuals. However, gender was not a significant factor in the regression analysis.
The results related to internet use were mixed, with the chi-squared test and regression analysis indicating that moderate internet users are more likely to be isolated compared to those who use it rarely or frequently.
The chi-squared test examining the association between social isolation and education levels revealed that a significant portion of isolated individuals completed their education at ages 15 and 16. However, this trend was also observed among non-isolated individuals. Regression results show that individuals who completed their education at age 19 or were still studying are less likely to be isolated than those who completed it at ages 18 and 17, though the difference was not pronounced.
The chi-squared test identifies associations between alcohol consumption and social isolation, contrary to expectations. Both isolated and non-isolated individuals tend to be moderate drinkers, and alcohol consumption levels were not significant for classifying isolation.

To conclude, the study found that social isolation among respondents in the ELSA study exhibited some common characteristics seen in global studies. Age played a significant role in social isolation, highlighting the need to address isolation issues among the aging population. Internet use showed mixed results, whereas education levels had associations with isolation, particularly among those who completed their education in their mid-teens. The study emphasized the importance of researching social isolation in older individuals, given its adverse health effects and underassessment in this demographic.

A full report of this project can be downloaded from here.