Latent Profiles of Health Literacy and Their Association with Sociodemographic Characteristics Among Karen Refugees in Buffalo, NY

Background and Purpose: Health literacy (HL), defined as the ability to obtain, process, and understand basic health information, is a major determinant of an individual’s overall health and healthcare utilization (Institute of Medicine, 2004). The level of HL among refugees, in particular, may be influenced by where they are in their awareness/knowledge of the overall acculturative process with respect to the healthcare system. Low HL is not only associated with a major reason for medication non-compliance but also greater utilization of emergency services, frequent hospitalization, and a higher risk of death (Berkman et al., 2011). In order to explore how the HL scores can assist in identifying subgroups, whose sociodemographic characteristics can be used to develop tailored HL educational content delivery for maximum efficiency and effectiveness, this study aims to (1) identify different profiles based on three subscales (functional, communicative, and critical) of a health literacy measure, and (2) examine the sociodemographic characteristics associated with the emerged profiles.

Methods: We analyzed survey data of 201 Karen refugees living in Buffalo. Health literacy was assessed using a modified version of the All Aspects of Health Literacy Scale (AAHLS; Chinn & McCarthy, 2013). We employed latent profile analysis (LPA) to identify profiles using three subscales from the AAHLS. Then, with the emerged profiles from LPA as an outcome variable, we conducted multinomial logistic regression to examine the association with sociodemographic characteristics. Sociodemographic variables included in the multinomial logistic regression were gender (0=male), age (in years), English language proficiency (1=poor to 4=excellent), self-rated health (1=excellent to 5=poor), marital status (0=single), religious participation (0=never to 6=Daily), and education (0=No college degree). All analyses were conducted using Mplus version 8.6 and SPSS version 28, respectively.

Results: Three profiles of Karen refugees emerged from the LPA including “resettling” (those who scored low in functional HL, high in communicative HL, and low in critical HL, 64.6 %), “acculturating” (those who scored average in functional HL, low in communicative HL, and average in critical HL, 19.0 %), and “integrated” (those who scored high in functional HL, average in communicative HL, and high in critical HL, 16.4 %). The results of multinomial logistic regression showed that, compared to “resettling” refugees, “integrated” refugees were more likely to be female (OR=4.131, CI=1.034-16.498), have higher English language proficiency (OR=6.963, CI=2.832-17.117), and poorer physical health (OR=1.997, CI=1.037-3.845). There were no statistically significant factors differentiating Karen refugees between resettling and acculturating refugee profiles.

Conclusions: Three emerged profiles showed unique patterns across functional, communicative, and critical subscales of HL that broadly resembled the characteristics of the acculturative stages, which prompted our naming convention for each profile (Resettling, acculturating, and integrated). The regression that followed further resulted in showing the consistent and expected association between profiles and selected sociodemographic variables. The findings from this study illustrate the need to pay attention to the individual’s acculturative level and use the information to figure out which HL-domain healthcare providers should selectively target for tailored educational efficiencies.