Evaluating Rural Ethiopian Youths’ Willingness and Competency to Promote Literacy Regarding G × E Influences on Podoconiosis Open Access

Advances in genomic technologies and environmental exposure measurement (e.g., lifestyle, toxins, social determinants) have advanced our understanding that virtually all diseases are jointly influenced by interactions of genes with environment (G × E) [1‒3]. Indeed, there are numerous ways that genetic predisposition interacts with environmental exposures to influence health outcomes (e.g., additive, multiplicative interactions) [4, 5]. For the purposes of this report, we consider G × E interactions broadly to occur when the expression of a trait is influenced by the interplay of genes and the environment [6]. For example, environmental exposures can “turn” genes on and off in ways that increase or decrease health risks [5]. Moreover, genetic influences may only manifest in the presence of an environmental exposure. Thus, reducing or eliminating exposures is particularly beneficial for those who have inherited susceptibilities to a specific health condition.

Lay understanding of G × E interactions is low generally and even more so in communities with limited literacy [7]. Our prior work and that of others show that lay audiences commonly misperceive genetic influences to operate deterministically such that inherited susceptibility inevitably results in the health condition and cannot be prevented [7‒11]. Promoting accurate understanding of G × E influences is important because beliefs about the causes of health conditions are associated with what, if any, precautionary actions are taken to reduce disease risk [12]. For example, beliefs that a health condition cannot be prevented are generally negatively associated with individuals taking any precautionary actions [13, 14].

Podoconiosis is a neglected tropical disease with confirmed genetic and environmental influences on its pathogenesis [15]. The disease is caused by barefoot exposure to the red clay soils of volcanic areas [15‒17]. Recent nationwide mapping in Ethiopia suggests the overall disease prevalence is 4% [18]; 43% of the population is at risk [18]. Our research was conducted in the Wolaita Zone of Southern Ethiopia where 1 in 20 people are estimated to be affected by the condition [16, 19]. Podoconiosis poses substantial socio-economic burden, particularly to subsistence farmers in Ethiopia who routinely work barefoot. Individuals who develop the condition experience progressive bilateral swelling of the lower legs, oozing wounds with an offensive odor [16, 20], and physical disfigurement. Affected individuals often experience social and self-stigma that excludes them from social activities and marital relationships with members of unaffected families [21‒24].

The condition clusters among families (63% heritability) [25, 26]. Genome-wide association testing has implicated variants at the HLA class II locus on chromosome 6 associated with immune response to increase risk of developing podoconiosis [15, 26]. However, the condition is entirely preventable if genetically high-risk individuals consistently protect their feet by wearing shoes and practicing foot hygiene [27, 28].

Our research team has been working for over a decade with communities in the Wolaita district [15]. In a series of qualitative studies, we have found that both affected and unaffected individuals acknowledge that podoconiosis runs in families [25, 29]. However, this understanding is counterbalanced by beliefs that the condition is also contagious, which they reason to explain the clustering of cases [30, 31]. Community members believe the condition can be passed on by sharing shoes, bath water, and other activities in close proximity with someone who is affected. We found this belief also to be associated with strong external and internal stigma and less inclination to endorse shoe wearing and foot hygiene as effective for prevention [32].

Based on these findings, we conducted a randomized community-based intervention trial described in detail elsewhere [33]. We tested a three-arm design with all arms receiving general activities to raise community awareness (e.g., posters at the local markets), and in two arms, affected and neighboring households received a home-based session provided by adult lay health workers to promote shoe wearing and foot hygiene. In addition, in the third arm, households received education on the process by which G × E influenced podoconiosis and its amenability to prevention. We found that unaffected neighbors who received the G × E education were significantly more likely to self-report decreased levels of stigma against their affected neighbors [31, 33]. Based on these findings, we began to test sustainable dissemination avenues for this education as described below.

Engaging youth has been consistently shown to be effective for disseminating health messages on sensitive issues and expanding reach of health-related services [34‒38]. Numerous studies in low- and middle-income countries (LMICs) show that youth-led education interventions can reduce youth’s risky sexual behaviors, violence, substance use [39‒41], and reproductive health knowledge [42].

To be effective at building genomic literacy, youth would need to be open to the notion that genetic influences are not deterministic of health outcomes. To this end, we surveyed youth in the Wolaita district to assess their understanding of G × E contributors to podoconiosis [30]. Half of the youth surveyed accurately endorsed the joint contributions of both gene and environment influences on podoconiosis and that the condition could be prevented. However, many youth also held misconceptions about contagion that could reduce their competence as lay educators.

Surprisingly, the majority of studies of lay peer educators have not described or assessed the requisite qualities that influence youth competency to serve in this role [43‒45]. One systematic review of school-based peer education interventions suggested that the most effective peer educators have leadership skills [46], a minimal level of related confidence, some technical competency, and communication skills.

In the context of health, peer educators who demonstrated an interest in the health condition [47], had relevant extracurricular experience [48, 49], particularly involving leadership [47, 49], and exhibited willingness to dedicate time to service were most successful [47, 49]. Most recently, notions of practical knowledge and practical wisdom have been suggested to be key for youth leaders to be effective [50]. Peer self-assessments of competency have also been shown to be important [51]. Indeed, these findings are supported by evidence that self-reported competency including achievement motivation, self-efficacy, and collaborative ability are not only important but may also be attributes associated with willingness [52].

We present data from a feasibility study of implementing a peer educator approach to increase G × E literacy among youth. We pose the following research questions: (1) What attributes are associated with youths’ competency to disseminate G × E explanations of podoconiosis? (2) What attributes are associated with youths’ willingness to disseminate G × E explanations of podoconiosis? and (3) Do these attributes moderate the association between competency and willingness?

Data come from a cross-sectional survey conducted of 377 youth living in communities in the Wolaita zone, where podoconiosis is endemic, of Southern Ethiopia. To represent the 10% of families affected by the condition, 41 were from families affected by podoconiosis. Data collection was conducted between August and September 2017. A detailed description of the survey and sampling methods is provided elsewhere [30].

Youth were eligible if they were aged 15–24 years and resided in one of the two study communities (Tome Gerera and Sura Koyo). We conducted a census to identify households with eligible youth to create a sampling frame and draw a random sample. Power analyses indicated 347 youth to be an optimal sample size to detect differences of plus or minus five units in survey responses. Of the total 3,542 youth enumerated, 5% could not be reached; 377 youth were approached to complete the survey that was administered by trained interviewers.

We used a systematic qualitative process to inform survey development. This process is described in detail elsewhere [8]. Briefly, we engaged a team of scientist experts and a group of lay individuals (youth living in areas of endemic podoconiosis) and asked each to create a schematic mental model of the factors they believed caused podoconiosis. The expert’s mental model was used as the gold standard [8]. We identified misconceptions and gaps in knowledge based on differences in the youth and expert models. These “gaps” guided our survey items intended to quantify the prevalence of these differences in the target communities. Examples included extent of endorsement of contagion, supernatural influences, inheritance, shoes as adequate prevention, and avoidance of personal contact. We also assessed whether the youth knew anyone personally affected by podoconiosis (yes/no), had knowledge of how gene and environment factors influenced podoconiosis development, and self-rated confidence to explain G × E contributors to podoconiosis. Lastly, the survey assessed factors described in the literature to influence competence. This included youth’s age, gender, formal education (yes, no), reported engagement in civic activities (e.g., participation in youth clubs), ever having participated in health-related civic activities (e.g., interactions with health extension work) (yes/no), or having any civic leadership role (yes/no).

Written informed consent was obtained from all participants. Thumbprints were taken from those who were not able to sign the information sheet and consent forms. For youth under the age of 18 years, consent was obtained from their parents or guardians.

Willingness was measured using four indicators with three scale response categories (1 = not willing, 2 = undecided, 3 = willing, e.g., How willing would you be to volunteer to educate the community about the causes of podoconiosis? range [4–12]). The Cronbach alpha for the willingness scale was 0.95. Responses to this item were highly skewed; thus, we coded willingness as a dichotomous variable where 0 represented not willing and 1 represented more willing using the median (11) as a cut point.

Objective Competency was a composite score drawn from the sum of three separate scales: accuracy of G × E knowledge, number of misconceptions endorsed about causes of podoconiosis, and confidence to explain these influences to peers. The G × E knowledge variable was measured by responses to four true/false questions (scored from 0 to 4 correct, e.g., “A person who has relatives with podoconiosis will certainly get the disease”). Misconception endorsement was assessed using 11 true/false responses (range 0–11 correct, e.g., Podoconiosis can be transmitted when people make physical contact with the blood or body fluids of a patient”). Confidence to explain was measured with 9 items assessed on a 3-point scale (agree – 1, undecided – 2, and disagree – 3, e.g., “I am confident that I could explain to other people why some individuals develop podoconiosis and others do not”). A point was given for each agree response and 0 for disagree and undecided responses, with a range from 0 to 9. We derived a competency composite score by summing up the three scores (range 0–24). We tested the competency score as a continuous variable and dichotomized variable (0 for below the median of 13 to indicate “lower” competency and 1 for above the median to indicate “higher” competency).

Bivariate associations between predictor variables and dependent variables (willingness and competency) were assessed using the Statistical Software package (SAS) version 20. Binary logistic regression models were fit to examine associations with the main and interaction effects of covariates with the binary outcome of willingness (not willing and willing). Bivariate analysis was conducted to identify significant differences in competency by each attribute using an independent t test for binary independent variables and analysis of variance for three-level categorical age.

Of the 377 youth who participated (100% response rate), the mean age was 17.7 (SD = 2.3, range 15–24); 52% were female and 95% reported having some formal education. Over half of the respondents (58.6%) reported that they had participated in extracurricular/civic activities and 12% reported having played a leadership role in civic activities in their community. Forty-four participants (11.7%) reported that they had friends who were affected by podoconiosis. The majority (79.8%) said they had been visited by health extension workers (HEWs) in the prior 12-month period (Table 1).

Competency level was significantly associated with most of the characteristics measured (Table 2). Males had significantly higher competency levels than females (mean difference: 1.89; 95% CI, 1.10, 2.68). Youth with no formal education had lower competency than those who had attended some formal schooling (mean difference: −3.54; 95% CI, −5.35, −1.72). Youth from affected families had lower competency than youth from unaffected families (mean difference: −1.68, 95% CI: −2.99, −0.37). Youth who had no friendship attachment with affected individuals had lower competency than those who reported having friends affected by podoconiosis (−1.63, 95% CI: −2.84, −0.38). Age, however, was not associated with competency level (F = 1.18, p = 0.28).

Youth who had not participated in civic engagement had significantly lower competency scores than those with civic engagement experience (mean difference: −1.85, 95% CI: −2.65, −1.05). Youth who reported no community leadership roles had lower competency than those who had experience in leadership (−1.83; 95% CI: −2.11, 0.99). Youth who had not been visited by the HEWs over the prior 12-month period had lower competency than those who had contact with HEWs (−2.49, 95% CI: −3.47, −1.42).

Due to the skewness of the willingness score, we used a nonparametric Wilcoxon rank-sum test (Mann-Whitney U) to examine associations between youth characteristics and willingness. Median willingness to participate in G × E literacy building was significantly associated (Table 3) with age – older youth were more willing than younger youth, gender – male youth were more willing than females, education – youth with some formal education were more willing than those with no formal education, civic engagement – youth with civic connections were more willing than those without civic engagement, and leadership experience – youth with leadership experience were more willing than those with no such experience (p = 0.01, p = 0.01, p = 0.005, p = 0.04, and p = 0.02, respectively). We also found significant associations of median willingness with objective competency – youth who were more competent reported greater willingness to serve as lay G × E educators than those who scored as less competent (p < 0.001).

Variables showing statistically significant associations with competency and willingness were further analyzed using a binary logistic regression model (high vs. low willingness) to determine whether any of the youth characteristics moderated the association of competency with high willingness. First, individual covariates (i.e., competency, gender, education, leadership role, civic engagement, and age category) were included in the model to see whether the individual main effect of each predictor was significantly associated with willingness. The main effects of all variables were significantly associated with willingness (p < 0.05, Table 4). Competency remained significantly associated with willingness after controlling for the other covariates.

We added an interaction term to the model to test whether the association of competency with willingness varied by level of the other covariates. Our findings showed no interaction effect between competency and other variables in predicting high willingness (p > 0.05). However, examining the interaction in a 2 × 2 table, one can see that the association of competency and willingness did vary by gender and reported civic engagement. Among those with low competency, the odds of being willing were 1.87 times higher for males compared to females (OR 1.87; 95% CI, 1.02, 3.40). Among those with low competency, the odds of being willing were 2.3 times higher for those who had experience in civic engagement as compared to those who did not have experience in civic engagement (OR 2.30; 95% CI, 1.27–4.18).

Findings of this study suggest various attributes to consider in identifying potential youth innovators/early adopters who would be willing to serve as peer G × E educators. Based on bivariate analyses, demographic (e.g., gender, education) and civic experiences (leadership experiences, interactions with HEWs) were positively associated with youths’ objective competency to serve in this role. These findings are consistent with youth peer educators using single dimensional competency measures in other health contexts [46]. However, our findings that affected youth had generally lower competency are not consistent with research showing that those living with HIV/AIDS have had high competency in other similar settings [53]. For instance, the role of people living with HIV in providing peer support has been widely accepted and encouraged in Thailand for many years [54]. Measurement differences make it difficult to compare findings. Prior studies have used a variety of indicators of competency (e.g., self-evaluated, single item). Our measure was multi-dimensional with strong internal consistency.

Characteristics associated with willingness were similar to those for competency. Indeed, competency and willingness were significantly associated. A few characteristics specific to podoconiosis with respect to willingness are worth noting. As with competency, we found that affected youth were also less willing to serve as peer educators than unaffected youth. Our prior qualitative research has consistently shown that social stigma surrounding podoconiosis can make these roles aversive for affected individuals [22, 55]. However, in the context of HIV, a highly stigmatized condition, this barrier has been overcome and peer volunteers with HIV have been shown to be effective educators [56]. It is possible that greater acknowledgment of the preventability of podoconiosis facilitated by improved G × E literacy could lead affected youth to be willing to service as lay health educators.

Youth with some formal education were more willing to participate in literacy-building activities. This finding was consistent with past research about qualities of peer educators for HIV prevention in South Africa [57]. It also indicates the possible benefits of integrating peer leader training in school setting as alternative pathway to disseminate GxE information.

Consistent with other research [58], youth’s civic and leadership experience was associated with greater willingness to serve as peer G × E educators. Participation in civic engagement may give youth the opportunity to acquire confidence and skills that enhance their willingness to help disseminate G × E messages.

On average, males had higher competency and willingness to participate than females. Similar results have been found for gender differences associated with youths’ willingness to participate in an agriculture education program in Nigeria [59]. Such gender differences could be reflections of the extant patriarchal and male-privileging cultural norms found in the study communities and many LMICs. Limited exposure of women and girls to civic roles compared to their male counterparts may diminish girls’ views that peer education is an appropriate role for them. Moreover, boys tended to be willing even when their competency was assessed to be low. Taken together, proactive and gender-tailored recruitment approaches will be needed to encourage girls to be peer educators and build their confidence that they can be successful in this role, whereas willing boys may benefit from training to ensure their competency.

The strong positive association of competency and willingness is encouraging in that those most likely to express willingness are also likely to be reasonably competent. This is consistent with assertions that competency is antecedent to voluntary participation [60]. For instance, a recent study [61] on participation of medical students in COVID-19 activities in Saudi Arabia found positive moderate correlation between willingness to work as part of the healthcare workforce during the pandemic and self-perceived competence score. This finding also suggests that youth who are willing may also be competent. However, as noted above, this self-selected group will likely systematically exclude girls and those with fewer civic experiences. Thus, relying on self-reported willingness alone is not optimal; use of some objective competency measures specific to the context may be advisable.

A notable strength of the study was that 100% of youth in the assembled sampling frame agreed to and completed the survey. Survey items related to competency were developed based on a systematic mental model exercise. This is also the first study to address these questions by designing a competency composite measure to evaluate youth to serve as G × E peer educators. Currently, there is scant research on participation of the youth as lay disseminators of G × E information. The consistency of our findings with studies of youth peer educators in other contexts suggests that this as yet untapped workforce could be proactively engaged and trained to promote understanding of the preventability of this highly stigmatized condition with the potential to redress stigmatizing beliefs.

However, these results should be considered against a few limitations. The sample was drawn from only two districts of Southern Ethiopia and results may not be generalizable to other districts or countries. We measured competency as only one time point and thus cannot assess the reliability of the competency score. Youth were not provided with a description of what would be expected of them in the role of youth educator. Thus, their responses are based on their own conception of what the role would entail. Despite the comprehensiveness of our survey, we did not assess youths’ perceptions of barriers to being effective in disseminating G × E information on podoconiosis. Lastly, the data were collected in 2017, and misconceptions and understanding of G × E effects on podoconiosis may have shifted.

This report describes an attempt to conceptualize a composite set of skills youth need to be competent genomic literacy disseminators. These results should be replicated. Moreover, a next step is to test associations of predisposing characteristics with the composite assessment in different contexts where G × E influences are pronounced. Additionally, repeated assessment of the composite to assess its reliability will be key.

Youth peer educator programs could have an important role in disseminating improved G × E literacy among LMICs. Our study suggests that many youth are both willing and competent to be peer educators. However, proactive recruitment and training strategies will be needed to ensure that the broadest group of youth has the opportunity to serve, particularly girls and those without formal education.

We would like to thank the youth who participated in the study. Authors also thank officials in the Wolayta Zone and Woreda administrations for the support rendered to the study team during data collection and dissemination of the study. We would also like to thank all individuals who took part in facilitating the study and providing insightful comments to enrich the study report. We are also grateful to Dr. Deborah McFarland for her role as a member of the first author’s thesis process.

Ethical approval was obtained from the Institutional Review Board of the College of Health Sciences, Addis Ababa University, approval number AAU-MF-03-008. All participants read the information sheet and a consent form, were given the opportunity to ask questions, and provided voluntary and an informed written consent to participate in this research. For youth under the age of 18 years, consent was obtained from their parents or guardians. Thumbprints were taken from those who were not able to sign the information sheet and consent forms. The Review Board approved the consent procedure for taking thumbprints.

Authors have declared that no competing interests exist.

C.M. conceived and designed the study. D.A. wrote the first paper draft with input from C.M. and circulated to other co-authors K.E., G.T., G.D., and R.M. for review. K.E. and D.A. developed draft interview instruments and participated in data collection. K.E. contributed in coding the data and R.M. helped D.A. in data analysis. All authors read and approved the final version of the paper.

All the relevant data are within the paper. Supplementary data will be shared on request to the corresponding author. Further inquiries can be directed to the corresponding author.