Depressive symptoms were measured with the 10-item Center for Epidemiologic Studies Depression Scale, while anxiety symptoms were assessed with the 7-item Generalized Anxiety Disorder Scale. Focusing on participants living alone, the team used latent profile analysis to sort individuals into distinct symptom profiles, then mapped the relationships among individual symptoms within each profile using network analysis. The latent profile analysis identified two subgroups: a low-risk class and a high-risk class.

This means that among older adults living alone, depression and anxiety symptoms grouped into two recognizable patterns, one in which symptoms were minimal and another in which both depressive and anxious symptoms were substantially elevated. Within the high-risk group, the network analysis pointed to three symptoms with the highest expected influence, a measure of how strongly a symptom connects to the rest of the network. These were sadness, measured by the CESD-10, and two anxiety symptoms from the GAD-7: feeling nervous or anxious, and experiencing uncontrollable worry.

These symptoms appear to be the most tightly woven into the web of comorbid distress in this subgroup. The comparison between the two classes yielded a nuanced picture. Overall network structure and global strength—the total connectivity among symptoms—did not differ significantly between the low-risk and high-risk groups.

However, six individual edges, meaning specific symptom-to-symptom connections, did differ between the classes, as tested with a Network Comparison Test. This combination of findings suggests that while the general architecture of how depression and anxiety symptoms relate to one another may be similar across risk levels, the intensity of particular symptom connections shifts in the high-risk group. The study's methodological toolkit was extensive.

Model selection for the latent profiles relied on fit indices including the Akaike information criterion, the sample-size-adjusted Bayesian information criterion, the Bayesian information criterion, the bootstrap likelihood ratio test, and the Lo-Mendell-Rubin adjusted likelihood ratio test. Network estimation used the extended Bayesian information criterion graphical least absolute shrinkage and selection operator method, and the stability of centrality estimates was checked with correlation-stability coefficients. The scale of the dataset is one of the study's strengths.

With more than eleven thousand older adults drawn from a large longitudinal survey, the analysis captures a substantial population in which living alone is a recognized risk factor in aging, associated with adverse psychological outcomes including depression and anxiety. According to the authors, the findings offer a theoretical basis and potential targets for intervention in the comorbidity of depression and anxiety symptoms among older adults living alone. If sadness, nervousness, and uncontrollable worry are the most influential nodes in the high-risk network, they represent candidate focal points for future research on this population.

Subject of Research: Depression and anxiety symptom profiles and symptom networks among older adults living alone in China

Article Title: A latent profile analysis and network analysis of depression and anxiety symptoms among older adults living alone

Article References: Hui, Q., Li, Y., Zhu, R., Luo, B., Liang, Y., Che, X., & You, X. (2026). A latent profile analysis and network analysis of depression and anxiety symptoms among older adults living alone. BMC Psychology. https://doi.org/10.1186/s40359-026-05503-4

Image Credits: AI Generated

DOI: 10.1186/s40359-026-05503-4

Keywords: depression, anxiety, older adults, living alone, latent profile analysis, network analysis, CESD-10, GAD-7, comorbidity, Chinese Longitudinal Healthy Longevity Survey, aging, mental health