Loneliness in mid-life and older adults from ethnic minority communities in England and Wales: measure validation and prevalence estimates

We investigated the prevalence of loneliness among 1206 adults aged 40 + from six minority communities in England and Wales: Black Caribbean, Black African, Indian, Pakistani, Bangladeshi and Chinese. Replicating the approach from the previous studies, we demonstrate robust acceptability, reliability and validity for both the six-item De Jong Gierveld (DJG) and single-item loneliness scales in our six ethnic groups. The prevalence of loneliness using a single-item question (loneliness reported as often/always) ranges from 5% (Indian) to 14% (Chinese) compared with approximately 5% for the general population aged 40 + in Britain. Levels of loneliness are very much higher using the DJG scale. Using a loneliness threshold score of 5 +, the percentage ranged from 13% (Indian) to 36% (Chinese). We explored the importance of six established loneliness vulnerability factors for our sample using regression modelling. Three factors were not associated with loneliness-number of children, gender and health rating, and three factors were protective: younger age, being married and low financial strain. The addition of ethnicity did not change these relationships or enhance statistical power of our models. Being a member of the African Caribbean group was protective against loneliness but not for the other groups included in our study. We suggest that exposure to loneliness vulnerability factors rather than ethnicity per se or measurement artefact underpins differences in loneliness across ethnic groups.

Automated summary

The prevalence of loneliness varies among different ethnic groups, with the Chinese group having the highest levels. Factors such as marriage, younger age, and low financial strain are protective against loneliness, while number of children, gender, and health rating are not associated with loneliness. Ethnicity did not significantly impact loneliness vulnerability factors in this study.