Common mental disorders among seasonal migrant farmworkers in Northwest Ethiopia

A community-based cross-sectional study was conducted from February to March 2019 in the western Gondar region, Ethiopia. The West Gondar zone (including Metema and West Armachiho districts) is one of the largest extensive agricultural development corridors of Ethiopia which attract large populations of seasonal migrant farmworkers. These are fertile agricultural areas with large-scale cash crops such as sesame, maize, cotton, and sorghum. As a result, a hundred thousand migrants and seasonal farmworkers travel to these areas for work during weeding (July to August) and harvesting seasons (September to November). The area is endemic to malaria and visceral leishmaniasis. The overall temperature ranges from 22 °C to 43 °C. Except for some mountainous regions, almost all of the land in these districts is lowland.

The study population was seasonal migrant farmworkers who migrated during agricultural seasons to these districts of the West Gondar zone during the data collection period. Seasonal migrant farmworkers are mainly engaged in cash crops such as sesame, cotton, but there is also substantial animal production. A seasonal migrant farmworker is an agricultural worker who is hired by a farm owned by someone else, usually hired on a part-time basis or a contract basis. Seasonal (non-migrant) farmworkers, on the other hand, earn a living in agriculture but remain in one location throughout the year. Migrants generally stay in the highlands of Ethiopia during the winter months and migrate northwest during planting or harvesting. The population of migrant farmworkers is ethnically diverse, with the majority stemming from the center of the Amhara region. Their ethnic composition therefore differs from that of this region of the country.

The sample size was calculated using a population proportion formula. Since no similar study was conducted previously in Ethiopia, 50% was used to estimate the proportion of the population of farmworkers who experience a CMDs. The size of the sample was decided based on the following assumptions: we fixed the significance level at 95%, the error margin at 5%, and the power (1-β) at 80%. With a design effect of two, the total sample size needed was 885, assuming a 15% non-response rate. First, two kebeles were purposefully selected based on the size (Delelo in the Metema district and Abderafi in the western Armachiho district). In the first stage of sampling, primary sampling units (PSUs)(Camps) were formed by segmenting the farms so that each PSU would have approximately 460 migrant workers. Camps within study clusters have been identified using line listings from ongoing farms. There are 174 farms in Delelo and 560 in Abderafi. The typical number of seasonal migrant workers in each farm is estimated to be 460, making the total number of migrant workers 80,040 in 174 farms. Lastly, the population of each camp was listed as living and hired by farm owners. Subsequently, systematic sampling was used to determine the required sample from each study site.

All data were collected through a questionnaire administered by the interviewer in Amharic. Interviewers participated in a 6-h training session and each completed practice interview before being approved to conduct study interviews. After informed consent was obtained from each participant, the interviewers worked in teams conducting one-on-one interviews with all camp residents who were interested in participating in the study. Interviews usually took place late in the afternoon or on the weekends when the workers were off duty at the safer place near the camps. No incentive was given to the participants. Regular supporting supervision was provided by the main investigators. Every day, the data were cross-checked for completeness and missing values by data collectors and supervisors.

A pretested and structured questionnaire was used to collect data. The tool was prepared first in English then translated into the local language (Amharic). Then it was re-translated to English by language experts to keep its consistency (Supplementary file 1). The questionnaire included demographic characteristics and other background information of the participants including sex, age, educational status, occupation before departure, family size, the residence of origin, marital status, average daily income, history of visits to farm areas, length of stay in months at farms, access to health information, sources of health information, working hours per day, working hours per week, heat stress and heat-related illness. The Self-Reported Questionnaire (SRQ-20) was used to evaluate the presence and prevalence of CMDs. The SRQ was originally designed by the WHO as a self-assessment scale but was also found to be suitable for interviewer-administration in settings of low literacy developing countries [31, 32]. The SRQ was cross culturally valid in Ethiopia among Amharic-speaking general populations; people in non-psychiatric treatment; and people in psychiatric treatment [33]. The optimal cut-off scores identified in previous studies were 4/5 and 6/7 for men and women, respectively. Each of the 20 items on the questionnaire receives a score of 0 or 1. A score of 1 indicates that the symptom was present during the last month before data collection; a score of 0 indicates that the symptom was absent. A cut-off point of 7/8 (7’yes’ a noncase, 8’yes’ a case) was used, which is the most commonly used cut-off point in developing countries [34]. It was used in previous community-based studies in Ethiopia [35]. The Cohen Perceived Stress Scale (PSS) was also used to assess perceived stress. The PSS, which is the most widely used psychological tool for measuring stress perception. The stress level is defined as a score of each item on a 5-point scale ranging from never (0) to almost always (4) indicating above 13 is considered to be psychological stress. The psychometric properties of the PSS has been tested in Ethiopian university students and found to be valid [36]. The PSS is a measure of the degree to which situations in one’s life are assessed as stressful. Items were designed to assess tap how unpredictable, uncontrollable, and overloaded respondents could find their lives. The scale also includes some direct queries about current levels of experienced stress ranging from 0 to 40 [37].

The data collected was encoded and entered into a computer using Epi-data version 3.1 programs and exported to STATA version 14. Descriptive statistics (frequency and percentages) were calculated and presented in the tables.

The protocol for the study was approved by the Ethical Review Committee of the College of Medicine and Health Sciences, University of Gondar. Moreover, we have received ethical approval from the Amhara Public Health Institute (APHI). Verbal informed consent was obtained from all respondents who were enrolled in the study. All participants have given written, informed consent to participate.

A total of 950 adults participated in the study. The majority of respondents were male (943/950; 99.26%). The median (± IQR) age of participants was 25 years (IQR:20–29) with an age range of 13 to 67 years. About 98% (932/950) of respondents were Orthodox Christian followers. The majority of respondents (832; 87.58%) were rural residents. More than 25.58% (243) were married. About 42.74% (406) had primary education and 10.21% (97) had secondary and higher education. Approximately half of the respondents (56.63%) earned a daily income of ETB 120 (5 USD). Most (772; 81.26%) had migrated more than twice to the study region (Table 1).

Nearly 60% of respondents (566) had access to health information from different sources. Mass media (328; 34.53%) and health workers (509; 53.64%) were common sources of health information. Approximately 72.74% (691) of respondents worked more than 8 h a day in the last month. Around 79.68% (747) of the respondents suffered from heat stress. Overall, 60.21% (572)) were affected by heat-related illness (Table 1).

The total prevalence of CMD and perceived stress was found to be 23.05% (219/950) [95% CI: 20.47–25.84] and 74.53% (731/950) [95% CI: 71.65–77.20], respectively. About 25.47% (242/950) of respondents reported mild stress, 58.95% (560/950) reported moderate stress and 15.58% (148/950) had extreme stress (Table 1).

Multivariate logistic regression analyses revealed that perceived stress levels, average daily income, receipt of health information, heat-related illness, and longer farm stays were significantly associated with CMDs.

The likelihood of having CMD were 1.53 times (AOR = 1.53, 95% CI: 1.10–2.15) higher for those with daily income below USD 5.00 than for those with daily income above USD 5.00. Respondents with moderate perceived stress (AOR = 3.18, 95% CI:1.18, 5.36) and severe stress (AOR = 16.15, 95% CI:8.96, 29.11) were more likely to experience CMDs than those with seasonal farmworkers suffering from heat-related illnesses were 1.6 times more likely to have CMDs than those who did not have a heat-related illness (AOR = 1.60, 95% CI: 1.11, 2.30).

Those seasonal migrant farmworkers who came to the farm for the first time were 62% less likely to experience mental health problems than those who came to the farm more than twice (AOR = 0.38, 95% CI:0.23–0.65). Seasonal workers with health-related information were 40% less likely to have CMDs than those without information (AOR = 0.60, 95% CI: 0.42, 0.85) (Table 2).