Objective

To identify predictors of loneliness and social isolation experienced by people with dementia at baseline and over time.

Methods

Using data from the Improving the experience of Dementia and Enhancing Active Life (IDEAL) cohort study (2014-2018), we examined the prevalence and predictors of loneliness and social isolation in 1547 people with mild-to-moderate dementia over 24 months. Loneliness was measured using the six-item De Jong Gierveld Scale at baseline and 24 months and social isolation by the six-item Lubben Social Network Scale at baseline, 12 and 24 months. Generalised linear mixed effects models examined possible predictors of loneliness and social isolation including individual characteristics, depression, cognition, cultural participation, and neighbourhood characteristics.

Results

At baseline 35.4% of people with dementia were categorised as being lonely and 28.8% as socially isolated, increasing to 39.3% and 32.0% 2 years later. Over the 24-month follow-up none of these predictors were associated with changes in social isolation scores. Only perceived neighbourhood trust was associated with change in loneliness longitudinally. At baseline, depressive symptoms, living alone, smaller social networks and lower neighbourhood trust were associated with greater loneliness. Cross-sectionally, loneliness and lower cognitive ability were associated with greater social isolation, and greater cultural participation, more green and blue spaces nearby and higher neighbourhood trust were associated with lower social isolation scores.

Conclusions

The findings highlight the importance of the local environment and cultural participation for people with dementia. Enhancing interactions with the local neighbourhood through initiatives such as dementia friendly communities may help to reduce loneliness and social isolation.