Correlates of Health Literacy among Farmers in Northern Thailand

Low health literacy is a barrier to public health efforts worldwide. Agricultural workers have an elevated risk for lower health literacy, with important health implications because of their potential exposure to harmful chemicals. The Asian Health Literacy Survey (AHLS) has been developed and translated for use in several different Asian countries and is standardized for easy comparisons across regions. However, it has not been translated for use in Thailand. The purpose of this study was to (1) to determine the health literacy of rural Thai farmers in Northern Thailand, and (2) identify correlates of health literacy within this group. Internal consistency of the Thai AHLS translation was “excellent” (alpha = 0.92). Descriptive results showed that health literacy was relatively high (M = 34.98/50, SD = 6.87). Education, income, working as a village health volunteer, age, length of time farming, no chemical use in farming, health, and pesticide screening were statistically significant correlates of health literacy (R2 = 0.19). Thai farmers had higher health literacy than reported for several other Asian countries. Results may be used to inform the design of future health promotion programs.


Introduction
Having sufficient health literacy is important when individuals manage health problems, as well as in everyday life [1]. For optimal health, people "need to be able to find, understand, and use health information and services" [2]. Individuals with low health literacy may struggle navigating health care systems, especially when unfamiliar with medical terms, how bodies function, how to interpret numbers or risk to make healthcare decisions, and when they have little formal education or complex conditions that require extensive care. Ultimately, poor health literacy can have negative effects on health, resulting in poor outcomes such as disease progression, poorer overall health status, and higher rates of hospitalization and emergency services use [3][4][5]. Identifying correlates of low health literacy can help practitioners design more effective health education programs.
Many individuals have limited ability to understand and apply health information. Title V of the United States 2010 Patient Protection and Affordable Care Act (p. 518) defines health literacy as "the degree to which an individual has the capacity to obtain, communicate, process, and understand basic health information and services to make appropriate health decisions." Health literacy involves the importance of efforts to increase health literacy among such workers [21,22]. Thus, farmers in Thailand, who have been identified as having incomes and education below the national average and who face potential chemical exposure, represent a particularly important potential risk group [25].
Prior research has assessed health literacy in a variety of ways. One major European study used the Test of Functional Health Literacy in Adults (TOFHLA) to measure overall health literacy among patients in outpatient settings in two public teaching hospitals [26]. Results indicated that only half the participants answered more than 80% of the questions correctly, 15% of patients could not understand basic prescription information printed on bottles, and 37% could not understand instructions about taking medication on an empty stomach [26]. Research examining health literacy among a nationally representative sample of US adults 16 years of age and older who spoke English or Spanish, including those in prison, found that 14% had "below basic", 22% had "basic", 53% had "intermediate", and 12% had "proficient" health literacy [27]. More recent pilot research conducted in the United States measured health literacy among a nationally representative internet panel of adults using the Health Literacy Skills Instrument, with a different set of outcome categories [28]. In all, 38% of participants had "below basic", 22% had "basic", and only 40% had "proficient" health literacy.
Developing standardized health literacy scales and translating them into multiple languages can help advance health literacy research. Unfortunately, relatively fewer scales have been designed for use in Asia compared to other regions. An exception is the Health Literacy Survey (HLS-EU-Q47), which was developed for use in Europe and then translated as part of a population-based study surveying residents of six Asian countries in seven languages: Indonesian, Kazakh, Russian, Malay, Myanmar/Burmese, Mandarin, and Vietnamese [12]. Translated questionnaires were pilot-tested and evaluated by experts from each country, with standardized scoring to enable comparisons. Thus far, the AHLS has not been translated and tested for use in Thailand, however. The present study extends this work to Thailand by translating the AHLS and administering it to rural Thai residents.
The purpose of this study was to: (1) to determine the health literacy of rural Thai farmers in Northern Thailand, and (2) identify correlates of health literacy within this group.

Study Design
The study used a descriptive-analytic cross-sectional survey design. Multivariate analysis was used to control for confounding.

Setting
This project examined farmers in the On Tai subdistrict, Sankampaeng district, Chiang Mai province in northern Thailand. The area is typical of rural farming communities of northern Thailand. Data were collected via face-to-face interviews.

Participants
A volunteer sample was recruited. To be eligible, individuals must have been: (1) at least 18 years of age, (2) residents of the On Tai subdistrict (Tambon), (3) able to speak Thai, and (4) currently employed as farmers. Recruitment was conducted through the local community health center registry list of farmers from all 11 villages in Tambon On Tai. There were 405 registered farmers, all of whom were invited to participate. Of those invited, 344 volunteered (85%), and all completed the questionnaire.

Measures
The study collected health literacy information via a Thai translation of the Asian Health Literacy Survey (AHLS) and several demographic items.

AHLS
The AHLS (HLS-EU-Q47) was originally created to measure health literacy in European countries and was later translated and validated for Asian populations [12,29]. In this study, AHLS items were translated to Thai by members of the CMU Faculty of Medicine who speak and write in Thai and English, using the translation-back translation method [30] (See Table A1).
The instrument consisted of 47 items and measured four competencies (access/obtain, understand, appraise/judge/evaluate, and apply/use health information) that fell within three domains (health care, disease prevention, and health promotion). Each item was rated on an ordinal scale ranging from 1 to 4 (1 = very difficult, 2 = difficult, 3 = easy, 4 = very easy). An overall health literacy score was calculated using the formula provided by Duong and colleagues [12]: Index = (mean − 1) × (50/3). Standardized scores ranged from 0 to 50, with 0 representing the lowest level of health literacy and 50 representing the highest. To reduce bias, only data for individuals who completed at least 42 of 47 items (~90%) were included in the analysis.

Demographics
Participant demographics were collected, including sex, age, religion, marital status, education level, income, household size, farming practices, and overall health status [31].

Data Collection Procedures
Data were collected by Chiang Mai University (CMU) medical students and Research Institute for Health Science research staff, under the supervision of study investigators. Training consisted of study protocol review, item-by-item questionnaire review, and supervised roleplay. After questionnaire translation and interviewer training were conducted, the questionnaire was pre-tested among a small group of farmers from outside the study area.
Farmers were invited to participate by word of mouth and instructed to gather in a central location within the village at a designated time. Interviewers met with participants privately, described the study, obtained verbal informed consent, and then conducted the interview, which took approximately 30 min. Data collection was anonymous. A small gift valued under $5 was offered to participants.

Data Management and Analysis
Data were analyzed using SPSS version 26.0. Standard multiple linear regression was used to identify correlates of health literacy. For categorical variables included in the regression analysis, the reference groups were those who: had lower education, had lower income, were not village health volunteers, were female, used chemicals in farming, rated their health as other than very good (i.e., very bad, bad, moderate, or good), had not been screened for pesticides, and had less farming experience (i.e., 20 years or fewer). Information from a small number of participants who elected to skip some items was excluded from the analysis.

Ethical Considerations
Procedures were implemented in accordance with the Helsinki Declaration of 1975. The study was a collaboration between investigators from Chiang Mai University (CMU) in Thailand (#COM-2560-047911-4791, 7/26/17) and California State University, Fullerton (CSUF) in the United States (#HSR-17-0193, 6/21/17), and was approved by Institutional Review Boards at both institutions. Verbal informed consent was collected and documented by interviewers due to the non-sensitive nature of the information collected.

Participant Characteristics
Participant characteristics are reported in Table 1. Age ranged from 35 to 84 years (M = 60.59, SD = 7.07). The median number of years having worked as a farmer was 30, ranging from 1 to 70 years. In general, participants had fairly low levels of education and income, and about half worked in more than one job. All but 2% attended at least some school (6/343). Half reported having another form of employment in addition to farming (175/336, 52%). Of these, nearly half (163/336, 48%) had some general type of employment, including working as a daily laborer, in construction, and with handicraft. About a third reported they owed money and did not have enough to live their lives (124/343, 36%); about a quarter (74/343, 22%) reported having more than enough to live their lives with the ability to save money. In the prior 12 months, about two-thirds of participants reported using herbicides (229/341, 67%) and pesticides (224/341, 66%), about a quarter used fungicides (89/341, 26%), while others reported using no chemicals in their farming (67/341, 20%). Most used fertilizer (333/340, 98%); about half reported using both chemical and organic fertilizers (171/333, 51%). More than half indicated they had been screened for pesticides in the past (223/341, 65%), most of whom had been screened within the previous 5 years (189/209, 90%). Of those tested, about half reported blood screening levels that were "normal" (115/223, 52%). Reasons given for not being screened included not knowing the test was available, not having time, not using pesticides, lacking interest, not going to the doctor, being afraid of needles, and not believing they were at risk.

Asian Health Literacy Survey
The internal consistency reliability for the Thai AHLS was excellent (Cronbach's alpha = 0.92). The overall mean AHLS score for the sample was 3.10 (SD = 0.41), and the standardized AHLS score [12] was 34.98 (SD = 6.87). Data were fairly complete. Most items (42/47) were skipped by 3 or fewer respondents (<1%). One exception was the question, "On a scale from very easy to very difficult, how easy would you say it is to judge how information from your doctor applies to you?", which was skipped by 9 of 344 total respondents (<3%). The next most frequently skipped item was, " . . . how easy would you say it is to understand the leaflets that come with your medicine?", which was skipped by 5 respondents (<2%). Although even the lowest response rates are still fairly high, they may reflect items that were more difficult for respondents to understand and answer. Table 2 presents responses for each AHLS competency subarea. For assessing and obtaining health information, finding information about reducing noise and pollution (69%) was the competency most frequently rated as "fairly easy" or "very easy." For understanding health information, the competencies of understanding doctor and pharmacist instructions on how to take prescribed medicine (96%) and understanding what doctors say (96%) were most frequently rated as fairly/very easy. For appraising, judging, evaluating health information, the competency most frequently rated as fairly/very easy was judging how housing conditions help one stay healthy (93%). For the ability to apply and use health information, following medication instructions (97%) was most frequently rated as fairly/very easy (95%).

Correlates of Health Literacy
A standard multiple linear regression was conducted to determine if education, income, having worked as a village health volunteer, sex, age, length of time working as a farmer, use of chemicals in farming, health rating, and pesticide screening were significant correlates of overall health literacy. Results indicated a statistically significant proportion of variation in health literacy (19%) was predicted by the variables in the model, F(9, 322) = 8.23, p < 0.001, suggesting a medium to large effect. Specifically, those with higher education, higher income, experience working as a health volunteer, younger age, more experience farming, no use of chemicals in farming, better self-rated health, and prior screening for pesticide exposure had higher general health literacy, controlling for the other variables in the model. Sex was not statistically significantly correlated with health literacy (See Table 3).

Discussion
Cronbach's alpha (0.92) indicated the translated AHLS had excellent internal consistency. The overall standardized AHLS mean score of Thai farmers in this study was 34.98. Compared to prior research, study participants had a higher standardized health literacy score than those in Indonesia (31.4), Kazakhstan (31.6), Malaysia (32.9), Myanmar (31.3), Taiwan (34.4), and Vietnam (29.6) [12]. Significant correlates of higher health literacy included more education, higher income, experience working as a village health volunteer, younger age, more experience farming, no chemical use in farming, better perceived health, and prior pesticide screening. These findings are consistent with prior research identifying lower education and older age as important risk factors contributing to lower health literacy, and the use of chemicals in farming and compromised health status as outcomes [15][16][17][18][19][20][21][22]. The fact that lower health literacy was associated with not having been screened for pesticide exposure, that 20% of respondents reported blood test screening results considered "at risk" or "unsafe/dangerous", and that another 18% were unsure of their results reinforces the need for developing and implementing interventions designed to increase health literacy among farmers [23,24]. Results contradict prior research that found lower age associated with lower health literacy [22].
The majority of health literacy research has been conducted with English-speaking populations. Relatively fewer studies have focused solely on measuring general health literacy, and not all instruments are appropriate for all settings. For example, the TOFHLA focuses on reading comprehension for healthcare and medical terms, but it is somewhat more complex and structured as a self-administered test, making it more challenging to use with rural populations [26]. The AHLS is relatively easy to use and has been translated into several Asian languages, but previously had not been translated to Thai. The present study helps address this gap.
There are several implications of this research. The results can be used by local community leaders and health care professionals to tailor health literacy education efforts, which can lead to improved health promotion programs and overall quality of care. Specifically, health literacy efforts should target those who: have lower education, have lower income, are older, use chemicals in their farming, have not been tested for pesticide exposure, have less experience farming, rate their health as poorer, and have not worked as a village health volunteer. Although participants had an overall higher standardized health literacy score compared to scores reported for other Asian countries, their health literacy can continue to improve. This is especially true for those identified as potential targets of health literacy promotion efforts, having characteristics correlated with lower health literacy. Improved health literacy can help people overcome barriers they face when navigating the healthcare system, resulting in more effective health promotion programs and better community health [12].
The current study adds knowledge to a growing research area in Thailand. Health care professionals may use results to improve current programs and overall farmer health. The study had a high response rate; 344 of 405 registered farmers participated (85%); thus, the sample can be viewed as an approximate representation of the farmer population in this subdistrict, with potential self-selection bias. Consistent with prior research in six other Asian countries, the Thai translation of the AHLS showed high internal consistency [12]. It should be noted that study participants represent farmers who were recruited from a specific region in Northern Thailand, and results cannot be generalized to the entire country. Although the sample reported lower levels of formal education, clinic staff reported that this group of farmers had participated in prior health-related research studies, and their health literacy may have benefitted from these experiences.
Future health literacy research should include other regions and other populations in Thailand to provide a more complete picture nationwide. The AHLS should be translated to other languages, and additional research should be conducted to better understand health literacy differences across Asia, as well as globally, so these differences may be addressed more effectively.

Conclusions
This study advances prior health literacy research by translating the AHLS to Thai and testing it among residents of rural Thailand. The translation demonstrated excellent internal consistency and should be tested in other Thai communities. Future health literacy efforts in this community should target those who have lower education, have lower income, are older, use chemicals in their farming, have not been tested for pesticide exposure, have less experience farming, rate their health as poorer, and have not worked as a village health volunteer. Acknowledgments: The authors thank Chiang Mai University, medical students and staff, and the Research Institute for Health Sciences for assisting with data collection; Nisit Wattanatchariya and Ratana Sapbamrer for their guidance and support; and Victoria van Twist for data assistance.

Conflicts of Interest:
The authors declare no conflict of interest.   Note: The scale was published by Duong and colleagues, Appendix A [12].