Patterns of social determinants of health and their associations with health indicators: a latent class analysis

OBJECTIVES: Social determinants of health (SDOH) are unevenly distributed and frequently co-occur, contributing to persistent health inequities. Analyses that examine SDOH individually may overlook important heterogeneity in how demographic and social factors cluster within populations. This study applied latent class analysis (LCA) to identify patterns of intersecting SDOH and examine their associations with health indicators. STUDY DESIGN AND SETTING: Secondary analysis was conducted using 2023-2024 survey data from a Canadian sample (N = 9743). SDOH were assessed using the Screening for Poverty and Related Social Determinants to Improve Knowledge of and Links to Resources tool. LCA was used to identify distinct patterns of SDOH. Differences across latent classes in health care utilization, medication use, self-reported physical and mental health, and multimorbidity (presence of multiple chronic conditions) were examined using chi-square tests and Cramér's V (V) effect size statistics. RESULTS: Five-class solution (entropy = 0.89) best fits the data. The classes were characterized as (1) socially stable middle-aged Canadian-born adults (21%), (2) highly educated and economically stable long-term immigrants (16%), (3) Canadian-born adults with disability and financial challenges (13%), (4) younger, racially diverse newcomers with multiple social and economic challenges (9%), and (5) older Canadian-born adults with financial and social stability (41%). The strongest differences across the latent classes were observed for multimorbidity (V = 0.41), self-reported mental and physical health (V = 0.34 and 0.30, respectively), medication use (V = 0.27-0.34), and to a lesser extent with healthcare utilization (V = 0.10-0.31). Classes characterized by intersecting disability, financial challenges, and limited social support reported poorer self-rated health and higher multimorbidity, while socioeconomic stable classes reported more favorable health indicators. CONCLUSION: SDOH cluster in distinct patterns that are differentially associated with health indicators. Analytic approaches, such as LCA, that account for how social conditions co-occur and interact provide a more accurate representation of these complex realities and can inform efforts to address health inequities. PLAIN LANGUAGE SUMMARY: Social determinants of health such as income, housing, education, disability, and social support can affect people's health. These factors often occur together, but most research examines them one at a time. This can make it difficult to understand how combinations of social challenges influence signs of health. In this study, we analyzed Canadian's responses to a survey to identify different patterns of social determinants of health. We identified five distinct patterns with different combinations of demographic and social characteristics. Some patterns reflected multiple overlapping challenges related to disability, financial challenges, and unmet social needs. Individuals within these patterns were more likely to report poorer physical and mental health, multimorbidity (having multiple chronic conditions), and greater medication use. Our findings show that social determinants of health do not occur in isolation. Instead, individuals experience different combinations of social circumstances that are related to different signs of health. Understanding these patterns may help researchers, health care providers, and policymakers develop targeted approaches to improving health equity.