Description
- Abstract:
- Background: Loneliness is on the rise in the U.S., leading to increased risk of death equivalent to smoking ~15 cigarettes a day, decreased engagement in society and institutional trust, and increased risk of developing dementia. However, little is known about its practical impact on healthcare utilization. Objective(s): This study examines the association between loneliness and physician use with a nationally-representative sample of U.S. non-institutionalized older adults and their household partners. Methods: This analytical sample included 4,086 individuals from the Health and Retirement Study (HRS.) The dataset was weighted to reflect a nationally representative sample and account for non-response and complex survey design. Key variables include a binary categorical exposure (loneliness measured as 2+ on the UCLA loneliness scale), a binary outcome (having visited a doctor in the last two years), and covariates of interest based on previous literature. Data analysis includes unweighted frequencies, weighted percentages, unadjusted and adjusted odds of healthcare utilization in a logistic regression analysis. Results: There was no discernable difference in physician use between those who reported being lonely (91.35%) and those who did not report being lonely (92.25%.) Lonely respondents had 0.89 unadjusted odds (95% CI: .57 - 1.38) and .97 adjusted odds (95% CI: .61 - 1.53) of physician use compared to those who were not lonely, all else equal. Little reported high school education had an unadjusted .12 lower odds of seeking healthcare than those with a college degree or higher (95% CI: .08 - .19), all else equal. Black/African American respondents had an adjusted .49 lower odds of seeking healthcare than those who were White (95% CI: .32 - .74), all else equal. Financial strain “yes, but not upsetting” and “yes, somewhat upsetting” had .61 (95% CI: .43 - .88) and .59 (95% CI: .40 - .86) unadjusted lower odds of healthcare utilization than those with no strain, respectively. Limitations: A binary exposure and outcome might not be as sensitive to an association instead of using an ordinal outcome. Additionally, HRS psycho-social questions are distributed as a separate leave-behind questionnaire. This opt-in data collection process does not guarantee that our response sample is representative of the study population, or is generalizable to the U.S. older adult population. Discussion: These findings are not what I expected for a condition associated with poorer health outcomes; with loneliness decreasing health quality, I would expect an increase in healthcare utilization. One 2015 study in lonely older adult populations found a significant association between continuous loneliness and healthcare utilization, but not an association for a single time point measure of loneliness. Studying other self-reported measures of health could find stronger determinants of poor health outcomes in the context of loneliness.
Access Conditions
- Use and Reproduction
- All rights reserved
- Rights
- In Copyright
- Restrictions on Use
- All Rights Reserved
Citation
Gulley, Blythe,
"Evaluating Associations Between Loneliness and Healthcare Utilization"
(2024).
Biostatistics and Applied Data Analysis.
Brown Digital Repository. Brown University Library.
https://doi.org/10.26300/hzv8-8w19
Relations
Collection:
-
Biostatistics and Applied Data Analysis
This collection highlights student research from Brown's School of Public Health courses in Biostatistics and Applied Data Analysis. Databases used include:...- Behavioral Risk Factor Surveillance System
- National Survey of Children's Health
- National Health and Nutrition Examination Survey
- National Health and …