Chronic undernutrition (stunting) is detrimental to academic performance among primary schools of adolescent children: a randomized cross sectional survey in Southern Ethiopia

Objective Despite several decades of work in improving childhood nutrition in Ethiopia, stunting remains a major public health concern with lack of evidence on its effect on school performance. Therefore, this study aimed at determining prevalence of stunting and the impact on academic performance among schoolchildren attending primary schools in Meskan District of Southern Ethiopia. Results During October 2016, we interviewed primary school children and their parents, collected anthropometric measurements of children, and conducted school record reviews among 408 randomly selected children attending ten primary schools. Univariate and multivariable linear regression were performed to assess stunting associated with academic performance among schoolchildren. A total of 378 primary schoolchildren were involved in the study giving a response rate of 93%. The prevalence of stunting was 16.9%. Grade repetition, place of residence, class rank of students and absenteeism from the class was negatively associated with the academic performance. In general, the present study children had relatively high prevalence of chronic undernutrition. We found that school underperformance was significantly high in the study area. The study underlines the need for nutrition interventions targeting rural areas to improve children’s academic performance.


Introduction
Poor academic achievement has been a major area of concern for educators, parents, and school children for more than three decades. There are several reasons for children to underperform at school such as age at enrollment, absenteeism, concentration in the classroom, psychosocial factors, illness and nutritional problems. Nutritional problems are among major factors that affect learning and school performance [1][2][3]. Some local studies conducted in Ethiopia revealed the relationship between academic performance and stunting. A study from Southwest Ethiopia found a significant correlation between stunting and academic performance [4]. Another study from Northwest Ethiopia showed that the low level of educational performance was significantly higher among the stunted children than that of the normal children [5].
Over the past decade, Ethiopia has shown encouraging progress in reducing undernutrition. However, as local studies conducted in Ethiopia revealed the baseline levels of chronic undernutrition (stunting) still remain high. According to the Ethiopia National Nutrition Program (NNP) baseline survey conducted in 2010, 23% of adolescent girls were stunted [6]. Study conducted in West Ethiopia indicates that prevalence of stunting was 17% [7]. A cross sectional study in Eastern Ethiopia shows that the prevalence of stunting was 25.5% [8]. In Northeast Ethiopia, the overall prevalence of stunting among the adolescents (10-19 years) was 15.5% [9]. A school based cross-sectional study in Northwest Ethiopia had revealed the overall prevalence of stunting among adolescent girls was 33.1% [10]. Furthermore, the prevalence of stunting among adolescents was 28.5% [11].
Although chronic undernutrition is a public health problem in Ethiopia, there is little documented information on the nutritional status of schoolchildren and the impact on their academic performance in Southern Ethiopia. Recently, the Ethiopian Ministry of Education recognized the impact of poor health and nutrition on children's ability to learn, their school attendance, and concentration [12]. The national gross enrolment rate (GER) in primary school (grade 5-8) was 71.1% and 74.2% in Southern Ethiopia. The national repetition rate among primary schoolchildren was 6.7%, slightly higher in males (7.1%) than females (6.2%). The national dropout rate was 10.1% [13]. This study aimed to determine rates of chronic undernutrition (stunting) and correlation with academic performance among primary school children in Meskan District, Southern Ethiopia.

Study design, period and setting
A school-based randomized cross-sectional survey was employed from October to November, 2016 in primary schools of Meskan District in Southern Ethiopia, a food insecure district. Up to 150 undernutrition cases are reported monthly from all health facilities of the District [14]. According to the 2015 population projection of central statistics authorities (CSA), the total population of the Meskan District is 411,933, of these 204,037 are males (CSA, 2013) [15]. Farming is the main occupation. In the 2015-2016 academic terms, 6745 schoolchildren were attending grades 5 to 8.

Population and sampling
The source population included all primary school children in Meskan District, with children from randomly selected primary schools constituting the study population. Primary school children, who did not intend to transfer schools during the intervention period, were included. Exclusion criteria included; schoolchildren with obvious congenital or physical growth measurements. We excluded schoolchildren whose parents did not provide written consent or were unwilling to participate.
The sample size was calculated using StatCalc in Epi Info 7. The following parameters used to determine the sample size: 23% of adolescents were stunted [6], 95% confidence level, 4% margin of error and adding 5% for nonresponse, the final sample size was 408 schoolchildren.
Ten primary schools were selected from the forty primary schools using proportional to size (PS) allocation method. Number and list of students were obtained from school. From each selected schools study participants were selected by simple random sampling based on sampling frame existing in the schools (students' roster).

Measurements
A structured questionnaire was developed by reviewing relevant literature, such as the nutrition baseline survey report for the National Nutrition Program of Ethiopia by Ethiopian Public Health Institute (EPHI) [6]. The questionnaire was first prepared in English, then translated into the local language (Amharic), and back translated into English by language experts to check for consistency. The questionnaires were administered in Amharic. The questionnaire was tested on 20 schoolchildren (not included in the sample) selected from primary school in Butajira town, and refined on the basis of the feedback obtained from the pre-test.
Height was measured in an erect position using a calibrated wooden height-measuring board (Stadiometer, Shorr Productions, Olney, MD, USA) with a sliding head bar while children were barefoot. Their height was measured in duplicate and recorded to the nearest 0.1 cm or in triplicate whenever the deviation between the first two measurements was > 0.5 cm and the average of all measurements was taken. The age of children in completed years was obtained from student's records and confirmed from their parents. In this study, stunting was defined as height-for-age z-score (HAZ) of equal to or less than minus two standard deviation (− 2 SD) below the mean of a reference standard [16].
The overall subjects the students were given in the academic year 2015-2016 were considered to evaluate the academic performance of the students. Annual average grade score was evaluated by taking the result of two consecutive semesters of the year. To assess the relationship of chronic undernutrition (stunting) on academic performance, average marks of the overall subjects the students received were divided into two categories, poor scores and good scores, based on a cut-off mark of 50%. This cut off point was determined by considering the pass mark set by federal ministry of education, Ethiopian [13].

Study variables
The outcome variable in the study was academic performance, operationalized as students' average marks of the overall subjects. Stunting in children was considered as an independent variable in this study. Durable household assets, home animals, housing conditions and vehicles were considered in the construction of household relative wealth index (RWI) using a principal component analysis after checking all the assumptions. RWI divided and ranked into quintile (lowest, second, middle, fourth and highest).

Data management and statistical analysis
The data were entered in double, cleaned and checked for missing values and outliers, and analyzed using IBM SPSS statistics for windows, version 23.0. The Z score value for height-for-age was calculated using the WHO AnthroPlus software [17].
Descriptive analyses were performed using frequencies and percentages for categorical variables. Variables which were significant at p-value < 0.2 in the bivariate linear regression model analyses were candidate for entering into the multivariable linear regression model to identify the independent predictors for academic performance. A p < 0.05 was considered statistically significant. We present the results of the linear regression as parameter estimates (ß), p-values and 95% confidence intervals.

Results
We identified 408 primary school children in the study; however, 30 schoolchildren (17 boys and 13 girls) declined participation, leaving 378 (93%) schoolchildren paired with their parents from ten primary schools who were enrolled in our study. The study children (n = 378) had girl to boy ratio of 1.21:1. The mean age of the children was 12.8 years (SD ± 1.3) with a range of 10 to 15 years. Two hundred two (53.4%) of the schoolchildren were attending grade 6 and while a few students 30 (7.9%) were attending grade 8. Majority of the study participants 273 (72.2%) were from families who were currently lived in rural areas. Two hundred five (54.2%) of them had fathers who attended primary school. Two hundred twenty eight (60.3%) of enrolled students had fathers who were employed as farmers.
One hundred five (27.8%) respondents were categorized in the fourth wealth quintile (Table 1).
The majority of the schoolchildren 357 (94.4%) reported that they have good relationships with their peers, and 369 (97.6%) had good relationships with their families. Two hundred thirty nine (63.2%) schoolchildren reported being bullied by one of their families and 233 (61.6%) reported being punished at school. A majority of the parents 303 (80.2%) reported that they have to perform non-school work at their home ( Table 2). Three hundred twenty five (86%) of the parents reported that they regularly attend the school meeting (parent-teacher meeting). The majority (n = 345, 91.3%) of the parents said that they always discuss their children schooling with students while 34.9% of the parents support their children in doing their homework (Table 2).
In bivariate linear regression (Table 3), grade repetition and dietary diversity score of students were positively associated with average grade scores, while residence in  rural, class rank of students and absenteeism from the class were negatively associated with the average grade scores. After adjusting for height for age z-score (HAZ) and all other variables in the multivariable linear regression model (Table 3), grade repetition, residence in rural areas, student class rank and class absenteeism were negatively associated with the average grade scores.

Discussion
The overall prevalence of stunting in this study was 16.9%, similar to a study conducted in West Ethiopia (17%) [7] and Southwest Ethiopia (16%) [18]. But higher than Addis Ababa (7.2%) [19], East Ethiopia (8.9%) [20], Adama Ethiopia (12.6%) [21], Gonder (12.9%) [22] and Northeast Ethiopia (15.5%) [9]. The possible reason for this discrepancy might stem from differential dietary intake, socioeconomic and cultural differences rather than differences in their genetic potential to achieve maximum height. However, the prevalence of stunting among schoolchildren was lower than National study (23%) [6], Tigray Ethiopia (26.5%) [23], Northern Ethiopia 28.5% [11], and East west Ethiopia (24%) [24]. The current study revealed underperformance when compared with a study conducted in Uganda and India of the primary school children was achieved better academic performance [25,26]. This lower underperformance in the study schools may be attributed to the poor course delivery system and curriculum. The finding of current study also indicated poor academic performance than that study conducted in Southwest Ethiopia had scored better academic performance [4].
The finding of this study showed that grade repetition, class rank and absenteeism were associated with academic performance of students. Consistent with this study finding, a study done in Ethiopia, India and Malaysia, chronic undernutrition is associated with grade repetition [4,5,26,27]. The finding showed that place of residence was negatively associated with the academic performance, similar in other parts of Ethiopia, Sub-Saharan Africa and Western Africa [4,[28][29][30][31][32][33][34].

Conclusion
In general, the present study children had relatively high prevalence of chronic undernutrition. We found that school underperformance was significantly high in the study area. But, chronic undernutrition was not associated with poor academic performance of grade. The majority children were from subsistence farming households and consume foods grown in their local area. In study settings, the production of foods are mainly using for cash crops. However, there is ample evidence that children's academic performance outcome can be affected by other factors that were not captured in this study, including micronutrient deficiencies, lack of food availability in their households, medical problems, below average intelligence, specific learning disabilities, attention deficit hyperactivity disorder, emotional problems, a poor sociocultural home environment, psychiatric disorders, or environmental factors, and parasitic infections. The study underlines the need for nutrition interventions targeting rural areas to improve children's academic performance of grade.

Limitations
• This study used only anthropomorphic measurements and did not assess the micronutrient status of study participants. • There may have been differences in the evaluation system for schoolchildren academic performance among the study schools. • Furthermore, physical performance capacity and motor skills were not measured in this study. Authors' contributions TW and TB designed and supervised the study and ensured quality of the data and made a substantial contribution to the local implementation of the study, and TB assisted in the analysis and interpretation of the data. TW, the corresponding author did the analysis and wrote the manuscript and had the responsibility to submit the manuscript for publication. TW and TB highly participated in revising this paper. Both authors read and approved the final manuscript.

Ethics approval and consent to participate
To conduct this study, ethical approval was obtained from Institutional Review Board (IRB) of College of Health Sciences, Jimma University (ref: RpGe/4086/2016). Parents/guardians of the study participants were informed about the objective of the study, risks and benefit, privacy and confidentiality of information they provided. Parents/guardians of the schoolchildren were told that they have the right to refuse to participate at all or can interrupt the question at any time if they feel discomfort to respond for the question. Written and signed informed consent to participate in this study was obtained from the parents/guardians.