Abstract
Background and Objective:
This study assesses health literacy and other health care barriers among African immigrants in Southwest Ohio using survey questions and the short assessment of health literacy translated into four major languages spoken within the community.
Methods:
Engaging 59 Southwest Ohio participants, the Short Assessment of Health Literacy (SAHL) survey and open-ended questions regarding health care barriers were administered.
Results:
The research identified education, family dynamics, and employment status as significant predictors of health literacy levels, with nearly half of the participants exhibiting limited health literacy. Participants also preferred one-on-one health education sessions.
Conclusions:
The study highlights language barriers as a significant obstacle to health care system navigation, with a preference among participants for one-on-one learning sessions about health information. The findings suggest a critical need for culturally and linguistically tailored interventions to improve health literacy and access to preventive health care services. Future research with larger sample sizes over longer periods is essential to fully understand the health literacy needs of this complex demographic and to inform targeted health care provider interventions.
Background
Navigating the U.S. health care system is complex, particularly for immigrant populations who face additional barriers such as language, cultural differences, unfamiliarity with the system, and low socioeconomic status. These challenges are best understood through the lens of the social determinants of health, which include factors such as economic stability, education, health care access and quality, and social context, each playing a critical role in shaping health equity and outcomes.
Among these factors, health literacy, the ability to find, understand, and use health information, is a critical yet often overlooked barrier to care. Low health literacy is more common among socioeconomically disadvantaged groups and is linked to limited use of preventive services and ineffective management of chronic conditions.1–3
In recent years, the immigrant, refugee, and asylum seeker population in Southwest Ohio has grown significantly, 4 bringing new urgency to understanding and addressing the specific health care challenges faced by these communities. African immigrants, in particular, represent a uniquely vulnerable group, often confronting a constellation of barriers beyond language, such as fear of institutional systems, social isolation, limited transportation, and a lack of culturally responsive care.5–11 There are also cultural beliefs and stigma around disease, including bewitchment, which may disproportionately affect women and self-advocacy.12–14 Some research underscores the pivotal role of health literacy in mitigating these challenges and improving health outcomes.15–18 Yet, the health literacy needs of African immigrants in the United States, and especially in the Midwest, remain underexplored in research and policy.
The intersection of limited health literacy with cultural, structural, and U.S. health care racism creates a challenging landscape for African immigrants, leading to compounded health inequities. Addressing these issues requires targeted interventions that promote health literacy, empower individuals to navigate health care systems, and combat the migration trauma, plus cultural and structural barriers that contribute to their marginalization. Cultural racism, expressed through anti-immigrant sentiments and violence, can lead to social isolation and mistrust within communities. This environment may hinder African immigrants from seeking information about health services or engaging with health care providers, thus limiting their health literacy. When individuals are uncertain about their conditional citizenship rights or fearful of repercussions due to their immigration status, they may avoid seeking necessary medical care, exacerbating health disparities. A diminished sense of belonging, stemming from both cultural racism and limited health literacy, 19 can lead to mental health issues such as anxiety and depression. 20 This can diminish personal health advocacy, effective health care system navigation, and resource utilization to challenge discriminatory health care practices. 21
Recognizing the central role of health literacy in bridging these health care gaps, this study utilizes the short assessment of health literacy (SAHL) as a validated tool that offers a quick and reliable way to measure health literacy across diverse populations.22,23 By translating the SAHL into four major languages spoken by African immigrants in the region, this research provides a culturally and linguistically sensitive assessment of health literacy levels.
A health literacy survey alone may not fully capture the specific needs of this Midwest African immigrant population. Adding a qualitative survey with open responses can help better understand local immigrant concerns regarding health information and the U.S. health system.
The objective of this research is two-fold: (1) to evaluate health literacy among African immigrants, refugees, and asylum seekers in Southwest Ohio, and (2) to identify key barriers to health care access in this population through open-ended/check-box survey questions. By doing so, this study aims to inform targeted interventions and policy recommendations that can improve health care navigation, preventive service utilization, and, ultimately, health outcomes for African immigrant communities.
Methods
Participants
Adults aged 18 and older who self-identified as African immigrants living in Southwest Ohio were eligible to participate. As responses reflected the general health literacy experience of the community, individuals with formal health care training were excluded. Participants were recruited through community engagement facilitated by Ebenezer Healthcare Access and limited by participant availability and willingness. Community referrals included faith-based (e.g., churches) and health organizations in Dayton, Ohio.
Baseline demographic information was collected for each participant, including age, sex assigned at birth, length of time in the United States, country of origin, language(s) spoken, income range, employment status, marital status, number of children in the household, and highest education level attained. In addition, given the exploratory nature of this study, no predetermined sample size was set.
Data collection
The University of Dayton Institutional Review Board (IRB) approved the study, including protocols and consent materials translated into four languages. Surveys were administered by trained personnel affiliated with Ebenezer Health care Access, who had prior immigration experience and fluency in participants’ preferred languages. Administrators received specific instructions on using the SAHL and ethical research practices. Informed consent was obtained in person, by phone, or virtually (via Zoom/Webex), following IRB-approved procedures.
The SAHL tool, used with permission from the Agency for Healthcare Research and Quality (AHRQ), was translated by Vocalink Global into four spoken languages among the local African immigrant population: Kinyarwanda, Swahili, French, and English. SAHL assesses comprehension of 18 common medical terms using a word-association format (e.g., “kidney” with “urine,” “fever,” or “don’t know”), and assigns one point per correct response (see Table 1). AHRQ guidelines classify scores 0–14 as low health literacy and 15–18 as higher literacy. 22
SAHL Survey Question and Instructions: “Read the Top Word Out Loud. From the Two Words Underneath, Which of the Two Words is More Similar or “Like” the Top Word? If You Don’t Know or Are Not Sure, Please Say ‘I Don’t Know.’ Don’t Guess.”
SAHL, short assessment of health literacy.
The SAHL’s reliability and validity have been established in English and Spanish, and its effectiveness has been maintained across translations into other languages, such as Dutch, using Cronbach’s alpha and item response theory. 24 This ensures the tool’s applicability across different linguistic backgrounds and allows it to meet diverse community needs without compromising its assessment integrity.17,18,22–24
A few qualitative open-ended/check-box questions (see Supplementary Appendix) were asked after the SAHL survey to gather deeper insights into participants’ experiences with the health care system, which were not fully explored through standardized tools.
Measures and analysis
All quantitative analyses were conducted using the programming language R and its integrated development environment RStudio. 25 To examine associations between participant characteristics and health literacy (SAHL ≤14 vs. >14), we applied Pearson’s chi-squared test to each categorical variable, where cell counts were sparse, p-values were estimated using a Monte Carlo simulation with 2,000 replicates. 26 To further explore the influence and interactions among these variables, a Random Forest model was employed. Figure 1 illustrates the importance of demographic and social variables in predicting health literacy scores. This model handles complex, nonlinear relationships and interactions between predictors. The robust model accommodation for categorical and continuous variables is ideal for exploratory analyses involving a diverse sample.
Variable importance ranking from the Random Forest model predicting SAHL scores among African immigrants in southwestern Ohio (n = 59). Predictors are ordered by their relative contribution to model accuracy, with education level emerging as the most influential. These insights highlight key factors that may guide the development of targeted health literacy interventions. SAHL, short assessment of health literacy.
Responses to open-ended survey questions were analyzed using themes for qualitative data analysis. Two experienced researchers independently performed initial coding to identify fundamental elements within responses. Categorical themes were then developed by comparing responses across participants. Regular comparisons were held to reconcile interpretation differences and to enhance inter-rater reliability. The final codes and themes were grounded in participants’ original narratives to preserve contextual authenticity. 27
Results
Participant characteristics and patterns by health literacy
Table 2 summarizes the baseline characteristics of the 59 African immigrant adults who participated in the study, stratified by health literacy group (SAHL ≤14 vs. >14).
Sociodemographic Characteristics of Participants, Stratified by Health Literacy Status (SAHL ≤14 vs. >14)
Bold-italic value represent significant p-values.
Values are reported as counts and column percentages. Pearson’s chi-squared test was used to assess associations between categorical variables. Asterisks (*) indicate p-values obtained using a Monte Carlo simulation with 2,000 replicates, due to sparse cell counts. 26
“Other” languages include English, French, Fulani, German, Kinyabwisha, Luganda, and Swahili. “Other” countries include Belgium, Iran, Kenya, Mauritania, South Africa, Tanzania, and unlisted responses.
A committed life partner is defined as a monogamous, long-term relationship between two people who have not been legally married.
Within both groups, participants were fairly split by sex. Most participants in the higher literacy group were aged 35–44 (47.1%), while those in the lower literacy group were more commonly 24–34 (48.0%). Age was significantly associated with health literacy (p value = 0.0185), with higher scores observed in the 35–44 and older age groups.
Most participants across both groups were married or partnered, with no significant differences in marital status. Similarly, the length of time in the United States, the number of children, and employment status were not significantly associated with health literacy.
Kinyarwanda was the most frequently spoken first language in both groups. A greater proportion of participants with higher literacy reported speaking more than two languages (79%) compared with those with lower literacy (56%). Although not statistically significant, this difference in additional languages spoken approached marginal significance (p value = 0.1002), suggesting a possible relationship between multilingualism and health literacy.
Education was significantly associated with health literacy (p value = 0.0425). Most participants with higher literacy scores had completed high school or above, while lower literacy scores were more frequent among those with only elementary or middle school education in this study, supporting existing literature that links formal education with improved health literacy.5,15,28 No significant associations were observed for income or country of origin, though Rwanda was the most commonly reported origin in both groups.
Predictive modeling using Random Forest
Figure 1 displays the variable importance rankings from the Random Forest regression model, indicating the relative influence of each predictor in estimating SAHL scores. Education emerged as the most important variable, underscoring its foundational role in shaping health literacy. Employment status and the number of children in the household also contributed meaningfully to the model, suggesting that work context and family responsibilities may influence an individual’s ability to access and process health information. Other relevant variables, including age, time in the United States, and prior country of residence, reflect how life stage and cultural integration may shape literacy patterns. While income, first language, multilingualism, marital status, and sex had smaller effects, their inclusion further supports the multifactorial nature of health literacy in this population.
Figure 2 displays a scatterplot comparing each participant’s actual SAHL score with the score predicted by the Random Forest model. The diagonal line represents perfect agreement between the observed and predicted values. Most points cluster closely around the line, indicating reasonable predictive performance. This clustering also highlights model consistency, with multiple participants sharing identical actual and predicted scores. The model achieved a mean squared error of 1.91, translating to an average deviation of 1.38 points and only a 9.4% error relative to the mean SAHL score. The R-squared value of 0.7795 suggests that the model explains approximately 78% of the variance in SAHL scores, providing substantial explanatory power for health literacy outcomes in this context.
Scatterplot comparing predicted and observed SAHL scores from the Random Forest model. The 45-degree line represents perfect prediction. The clustering of points around the line indicates reasonable model performance, though scores at the higher end show slight underprediction. This plot visualizes the model’s accuracy in estimating individual health literacy. SAHL, short assessment of health literacy.
Figure 3 offers additional interpretability by showing the marginal effect of two top predictors, education level and employment status, on predicted SAHL scores. The red line represents the average predicted SAHL score across the sample. In the left panel, participants with at least some secondary education show higher predicted scores than those with only elementary schooling. The effect plateaus somewhat after a bachelor’s degree, suggesting diminishing returns from advanced degrees in this context. The right panel shows predicted literacy scores across employment types. Although the scores of self-employed and retired participants are marginally higher, overall employment status seems to have a rather small impact (standard deviation of 0.28). These plots provide insights into how specific demographic factors shape health literacy outcomes, offering potential leverage points for tailoring interventions.
Partial dependence plots showing the marginal effect of
Qualitative insights: Barriers and preferences in health care access
Table 3 summarizes 69 open-ended responses describing participants’ confusion with the U.S. health care system. Responses were grouped into four overarching areas: language and communication barriers, health care system navigation and cost, medical understanding and decision-making, and those reporting no confusion.
Themes Identified from Open-Ended Survey Responses on Confusion Navigating the U.S. Health Care System
Responses were brief and exploratory in nature. Themes are not mutually exclusive.
The most commonly cited issue was language and communication, with 27.1% referencing general language barriers, 11.9% mentioning difficulty reading information, and 5.1% noting they did not speak English. One female participant shared that shyness affected their speaking ability, especially when unsure about language use. Another respondent explained, “Most of the time I don’t understand due to language barrier,” highlighting how language can hinder comprehension during health care interactions. A few participants specifically pointed to gaps in interpreter services, such as “The issue of not having interpreters who explain well.”
Several participants also mentioned challenges with the health care system logistics, including understanding costs (13.6%), finding providers (6.8%), and scheduling appointments (6.8%). One participant reflected, “The most confusing thing about health care is choosing the right health care provider, especially when you come into a new country,” underscoring how unfamiliarity with the system may affect newcomers. A smaller number expressed confusion about diagnoses, medication, or general medical knowledge.
To examine variation by health literacy, we categorized responses by SAHL score (≤14 vs. >14), as shown in Table 4. Language-related concerns appeared at similar rates across both groups. However, those with higher scores more frequently reported confusion related to system navigation and cost, while those with lower scores more often reported no confusion. Among the latter group, most had at least two children, suggesting that caregiving experience or life context may influence perceived ease of navigation.
Distribution of Open-Ended Confusion Themes by Health Literacy Group (SAHL Score ≤14 vs. >14)
Note: Participants could mention more than one area of confusion. Percentages may not sum to 100%.
Responses were grouped into four key areas. Percentages reflect the proportion of participants within each group who reported at least one concern in that category. Participants could report multiple concerns.
SAHL, short assessment of health literacy.
About 45% (n = 27) of participants preferred one-on-one interactions when receiving health information. This preference may reflect the value of trust, individualized support, and language accommodation, particularly in contexts where literacy levels and communication needs vary. These insights suggest potential directions for future interventions, but should be interpreted with caution given the limited depth of responses.
Discussion
This study highlights the complex factors influencing health literacy among African immigrants in Southwest Ohio. Using a linguistically adapted administration of the SAHL survey in four languages, we explored how education, age, language skills, and family structure relate to literacy outcomes in this underrepresented population.
Age and education were significantly associated with SAHL scores. Participants aged 35–44 and those with postsecondary education were more likely to demonstrate higher health literacy. Multilingualism showed a marginal association, suggesting that speaking multiple languages may support health care navigation. Across all participants, the mean SAHL score was 14.6 out of 18, with 42.4% classified as having limited health literacy based on a cutoff of 14.
Random Forest modeling reinforced these findings and revealed additional predictors such as employment and the number of children. While sex and marital status were not statistically significant, they contributed modestly to a model that explained nearly 78% of the variance in SAHL scores. These findings suggest that health literacy is shaped by interlocking social, economic, and linguistic factors.
Education appeared to be the most influential predictor in the model, suggesting that programs should be tailored to different literacy levels. 23 Effective strategies may include audiovisual tools (e.g., medical brochures) in participants’ preferred languages, the use of multilingual community health workers for navigation assistance and home visits, and the establishment of multilingual hotlines for health-related inquiries. In-person education with trusted community leaders may assist health navigation, while recognizing that group education sessions may be more effective due to the same culture prevalent in the target population.19,20 Additionally, the prominence of employment and family structure as predictors suggests that flexible program delivery, such as weekend sessions, mobile health visits, or virtual engagement, may better support working adults and caregivers. These implications offer direction for designing culturally responsive, accessible health literacy interventions within African immigrant communities.
While the SAHL survey provided a standardized measure, some items may have posed cultural or linguistic challenges. Participants were instructed to respond with “Don’t know” if unsure, and survey administrators did not provide definitions, which aligned with SAHL guidelines. Reproductive and gastrointestinal terms such as hemorrhoids, syphilis, miscarriage, hormones, and constipation had the highest “Don’t know” response rates (see Supplementary Appendix), which may indicate unfamiliarity or discomfort, even though no visible distress was observed during administration.
Qualitative responses, while brief, offer additional context. About 37% cited language-related challenges, including difficulty reading health materials or communicating with providers. Navigation and cost-related confusion were also common, especially among participants with higher literacy scores. Notably, those with lower SAHL scores more often reported “no confusion,” which may reflect underreporting, social desirability bias, or reliance on family networks.19,20 Given the brevity of responses and small sample, these patterns should be interpreted with caution.
One female respondent cited “shyness” as a barrier, which is insufficient to draw broader cultural or gender-based conclusions. This highlights the need for richer data to explore how institutional mistrust, gender norms, or acculturation may influence care-seeking behavior. Future qualitative studies using interviews or focus groups are essential to deepen the understanding of how African immigrants experience health care navigation and interpret health information. 14
Limitations include a small sample size, geographic focus, and language translation constraints that may limit generalizability. Some medical terms lacked direct translations, requiring interpretive decisions. Involving local medical translators helped address dialectal differences and culturally specific beliefs. Selection bias may also be present, as participants were recruited through a trusted non-profit organization, potentially increasing openness compared with less-connected immigrants. Additionally, qualitative data were collected through brief survey responses rather than in-depth interviews, which may limit the richness of insight. Such formats may yield vague responses, introduce respondent bias, and risk survey fatigue.
Despite these limitations, this study offers insight into health literacy among African immigrants in the United States, with nearly half exhibiting limited literacy. This rate was lower than that of a recent study in Massachusetts, which utilized a different literacy tool and had most respondents from West Africa. 23 Most participants here were from East Africa, such as Rwanda and Kinyarwanda-speaking countries, with similar culture and traditions. 29 Participants with lower health literacy may reduce participation in preventive and general health care. While echoing prior research, this study reinforces the need for larger, multi-site investigations including more diverse African populations. Future studies should also consider longitudinal or mixed-methods approaches to explore how health literacy evolves over time and shapes health care behaviors. A deeper understanding of these dynamics can guide the development of culturally responsive, effective interventions within African immigrant communities.
Conclusion
This study provides preliminary insight into the health literacy challenges faced by African immigrants in Southwest Ohio. It highlights the potential role of education, language access, and personalized communication in improving health literacy among African immigrants. The findings suggest a need for culturally and linguistically tailored interventions that move beyond written materials to emphasize direct, one-on-one engagement. As the population of African immigrants continues to grow, so does the importance of addressing health literacy as part of broader efforts toward equitable health care access. Future research should build on these exploratory findings to inform more comprehensive, scalable strategies that promote preventive care and support immigrant communities in navigating the U.S. health care system.
Footnotes
Acknowledgments
The study authors would like to thank the Agency for Healthcare Research and Quality for allowing the use and translation of the Short Assessment of Health Literacy Survey. In addition, the authors would also like to thank Ebenezer Healthcare Access staff and directors for their assistance with reviewing translation materials and broadening the number of our participants.
Authors’ Contributions
A.C. contributed to the study conception, data collection, qualitative analysis, article preparation, article writing, and article revision. Y.J.C. contributed statistical analysis, qualitative analysis, article writing, and article revision. P.H. contributed to the statistical analysis and initial article writing. A.M. contributed to data collection, qualitative analysis, and initial article writing.
Data Availability
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
Ethics Approval
The University of Dayton Institutional Review Board granted study approval, including a participant consent protocol, on July 23, 2020.
Participant Consent
Participants were given an IRB-approved consent form in their preferred language prior to survey participation. For in-person surveys, a printed copy was provided for participants to read and sign. For virtual surveys, the consent form was either displayed via screen sharing and read aloud by the survey administrator, who then confirmed verbal agreement to participate. In a few phone-based cases, the full consent form was read to the participant, and verbal consent was obtained. These verbal procedures were explicitly approved by the IRB and designed to ensure informed, voluntary participation.
Permission to Reproduce Material from Other Sources
Permission was given by AHRQ to translate and use the SAHL Survey.
Consent to Publish
All authors consent to publish this article as written in the submission.
Author Disclosure Statement
Amy Christopher is a non-paid board member of Ebenezer Healthcare Access, involved in recruiting participants. Alexis Muhumure became a non-paid board member of Ebenezer Healthcare Access during the article writing process (after the data collection and analysis phase). All other authors have no relevant financial or nonfinancial interests to disclose.
Funding Information
Translation of the SAHL survey into various languages was supported by the University of Dayton Faculty Fund. Partial funding for open access provided by the University of Dayton and Ebenezer Healthcare Access Libraries Open Access Fund. The authors declare that no other funds, grants, or other support besides survey translation were received during the preparation of this article.
Abbreviations Used
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
