1. Estimating risk of loneliness in adulthood using survey-based prediction models: A cohort study.
- Author
-
Elovainio M, Airaksinen J, Nyberg ST, Pentti J, Pulkki-Råback L, Alonso LC, Suvisaari J, Jääskeläinen T, Koskinen S, Kivimäki M, Hakulinen C, and Komulainen K
- Subjects
- Humans, Female, Male, Middle Aged, Adult, Aged, Finland epidemiology, Cohort Studies, Models, Statistical, Prospective Studies, Surveys and Questionnaires, Loneliness psychology
- Abstract
It is widely accepted that loneliness is associated with health problems, but less is known about the predictors of loneliness. In this study, we constructed a model to predict individual risk of loneliness during adulthood. Data were from the prospective population-based FinHealth cohort study with 3444 participants (mean age 55.5 years, 53.4% women) who responded to a 81-item self-administered questionnaire and reported not to be lonely at baseline in 2017. The outcome was self-reported loneliness at follow-up in 2020. Predictive models were constructed using bootstrap enhanced LASSO regression (bolasso). The C-index from the final model including 11 predictors from the best bolasso -models varied between 0.65 (95% CI 0.61 to 0.70) and 0.71 (95% CI 0.67 to 0.75) the pooled C -index being 0.68 (95% CI 0.61 to 0.75). Although survey-based individualised prediction models for loneliness achieved a reasonable C-index, their predictive value was limited. High detection rates were associated with high false positive rates, while lower false positive rates were associated with low detection rates. These findings suggest that incident loneliness during adulthood. may be difficult to predict with standard survey data., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2024 The Authors. Published by Elsevier Ltd.. All rights reserved.)
- Published
- 2024
- Full Text
- View/download PDF