Study setting, study design and study period
Wolaita Sodo town is the administrative capital of Wolaita Zone administration in Southern Ethiopia located at 380 km South from Addis Ababa. The town has 3 sub-cities and 11 lower administrative units. The total population of the town is estimated to be 182,607; from which 49% are females . According to the information obtained from Wolaita Zone Agriculture and education departments, the common staple foods in the area are cereals, roots, tubers, and vegetables and there are two private and five public secondary schools in the town respectively. A facility-based cross-sectional study was conducted among adolescent students from April to June, 2019.
Population and sampling
The source population for this study was all adolescents of secondary schools of Wolaita Sodo town and the study population are adolescent students of the selected schools. Pregnant adolescent girls, adolescents who were ill at the time of the study and with physical or visual disability were excluded from the study. A single population proportion formula was used to calculate the sample size with the following assumptions; 95% confidence level, 5% margin of error, an estimated magnitude of students’ academic performance of 72.8% taken from a similar study in Ethiopia , design effect of 2 and 10% non-response rate and the final sample size calculated is 670. There are seven schools in Wolaita Sodo town, two private and five public, and were stratified into public and private by assuming socio-economic differences among the families of the students and differences in teaching and learning resources between public and private schools. Among the seven secondary schools, one private and three public were randomly selected. The total sample size was allocated to the schools proportional to the number of students in each selected school. The study participants were selected by systematic sampling technique using the list of students enrolled in each school as a sampling frame. The sampling interval was determined by dividing the total number of students in the respective school grade level by the allocated sample size and was found to be five. The first participant was selected randomly by the lottery method, and then every fifth adolescent student was included in the study.
Data were collected using a structured interviewer-administered questionnaire. The questionnaire was developed and adopted from the Ethiopian Demographic and Health Survey (EDHS) validated tool and other related literature reviews [22,23,24] (Additional file 1). The questionnaire was pre-tested on 5% of the sample size on adolescent students from schools which were not selected for the actual data collection but no modification has been made. The data were collected by four data collectors and two supervisors after training was given for 2 days on the objective of the study, data collection procedures, anthropometric measurements, the confidentiality of the information and participant rights.
To ensure the reliability of anthropometric measurements, standardization test was done on five participants prior to actual data collection. First, the expert has taken the measurements and then the data collectors repeated the measurements on the same participants with some time intervals. The collected data were entered into ENA SMART software to check relative Technical Error of Measurements (TEM) and was found to be in the acceptable range, < 2.0%. Weight was measured using a portable digital flat Seca scale (Scale electronic scale, 770 Hamburg). Height was measured by Seca body meter (Seca 274 body meter). All measurements were taken three times, and the average was recorded as the final measurement. Academic performance and absenteeism data were taken from respective schools’ records.
The academic performance of the students which was calculated using two consecutive semesters mean mark scores out of 100.
Socio-economic and socio-demographic variables
Age of the adolescents, sex of adolescents, marital status of parents, educational and occupational status of parents and wealth status of the adolescents households. Wealth status was generated by using principal component analysis (PCA) and based on the results household wealth index/status was converted into quartiles and categorized as First, Second, Middle, Fourth, and Highest .
Nutritional status measurements and indices
Underweight– is BMI for age z-score (BAZ) of < − 2 standard deviation (SD) on the WHO growth reference cut-off point .
Overweight– was computed with BMI for age z-score (BAZ) of > + 1 SD on the WHO growth reference cut-off point .
Obesity– was computed with BMI for age z-score (BAZ) of > + 2 SD z-score based on the WHO reference cut-off point .
Stunting– is the height for age z-score (HAZ) of <− 2 SD on the WHO growth reference cut-off point .
Dietary diversity score
Dietary diversity was determined by using the Dietary Diversity Score (DDS). Three non-consecutive days 24-h recall of adolescents’ consumption of commonly consumed foods in the area was used to collect information on the DDS . Foods were categorized into 10 groups based on FAO recommendations ; starch stable food , vegetables, 3) fruits , meat , egg , fish and other sea foods , legumes, nuts and seeds , milk and milk products , oil and fats , sweets, spices, condiments and beverage . The response categories were “Yes” if at least one food item in the group was consumed and “No” when a food item in the group was not consumed. The results were summed and classified into < 4 food items and > 4 food items .
Alcohol consumption, the purpose of spending much time on the internet and being absent for 10% or more of school days for any reason in a calendar year.
Data management and analysis
Data were entered into Epi-Data version 3.1 and analyzed using Stata version 15 statistical software. Anthropometric data were analyzed using the WHO Anthro-plus software version 1.0.4 and nutritional status of the adolescents was determined using WHO reference 2007 cut-off point . Normality assumption was assessed for the dependent variable and the data were normally distributed (p-value is 0.77). Descriptive statistics such as frequencies, percentages, mean and standard deviation of the mean were done. Binary and multiple linear regression analysis were conducted to check the association between the dependent and independent variables. Variables with a p-value of less than 0.25 in the binary linear regression analysis were candidate variables for multiple linear regression analysis. Variables with the p-value < 0.05 in the multiple linear regression analysis were considered as statistically significantly associated with the dependent variable and parameter estimate (ß) with its 95% CI was reported.