Loneliness as a risk factor for dementia

Abstract

AbstractBackgroundThe importance of social relationships in health has been recognized over the years. Growing evidence in the field suggests that the feeling of loneliness, characterised by dissatisfaction with personal relationships, correlates positively with the outcome of dementia (Holwerda et al., 2014; Rafnsson et al., 2020; Sundström et al., 2020; Sutin et al., 2020). Systematic review has found mixed results but follow up times have generally been short and sample sizes small (Balouch et al., 2019; Penninkilampi et al., 2018).MethodUsing data from the UK Biobank, we will examine the association between loneliness and dementia. Between 2006 and 2010, twenty‐two recruitment centres have collected baseline data from over 500,000 participants all across the UK, providing socioeconomic and ethnic heterogeneity (Sudlow et al., 2015). This diversity in participants allows the generalisability of association between baseline characteristics and health outcomes (Sudlow et al., 2015). At baseline, all participants completed a mental health questionnaire which included questions about loneliness and quality of relationships. Our study will thus comprise the data of thousands of participants with a follow‐up of over 14 years.We will use appropriate regression models, such as cox proportional hazard models, to study the impact of loneliness on the subsequent development of dementia, adjusting for age, sex, education, physical health and lifestyle factors, as well as social contact and quality of relationships.ResultData analysis is yet to be done. According to preliminary data analysis, we are expecting to find a positive correlation between loneliness and the onset of dementia, as illustrated by previous studies. We will conduct multiple cox proportional hazard models, adjusting for basic socio‐demographic, clinical and behavioural covariates. In addition, we will exclude all participants with a diagnosis of dementia within three years of baseline to eliminate reverse causation bias.ConclusionLoneliness as a risk factor for dementia.