Prevalence and predictors of malnutrition during adolescent pregnancy in southern Ethiopia: a community-based study

A community-based cross-sectional study was conducted among pregnant adolescents in Kore district, southern Ethiopia on March 2018. The district is divided into 21 rural and 4 urban kebeles (smallest administrative unit) and its estimated population is 156,035, of which, more than 95% live in the rural kebeles. The food production system in the area is characterized by mixed-crops and livestock production. Barley is among the major food crops produced in the area. Most of the communities in Ethiopia including the one in the study area, are characterized by patriarchal norms and high gender disparity. The study was concerned to the issues faced by pregnant adolescents aged 15–19 years living in a rural area in relation to malnutrition.

The sample size was estimated by assuming a 50% malnutrition, a 95% confidence interval, a 5% margin of error, and a 10% non-response rate and this yielded a sample size of 422. The selection of the study subjects was occurred on two stages -cluster and individual. To select the kebeles (cluster level), the list of all rural kebeles in the district was considered as a sampling frame and ten kebeles were randomly selected. The individual level was involved selecting all pregnant adolescents within the selected kebeles. Following the selection of clusters, pregnant adolescents were identified by self-report, from the records of health extension workers (HEWs) providing a community based maternal and child services in each kebele and urine testing when pregnancy was doubtful. Pregnant women were identified from the report of health extension workers in the kebeles Total of 436 pregnant adolescents were identified from the selected kebeles and all of them were recruited recognizing that little difference from the sample size estimate.

Data were collected using interviewer-administered questionnaire and mid upper arm circumference (MUAC) measurement. Initially questionnaire was written in English and translated to the local languages then back translated to English. The questionnaire include information on socio-demographic and-economic characteristics, dietary and lifestyle/work related practices, and obstetric and healthcare history.

Improved eating habits during pregnancy was assessed using the items adapted from the national nutrition survey conducted by the Ethiopian Health and Nutrition Institute in 2009. Pregnant adolescents who responded affirmatively to at least either of the two items asking whether they had (1) increased their meal amount and (2) increased their meal frequency were classified as having improved dietary practices during pregnancy. Pregnant adolescents were asked about received support during pregnancy: (1) material aid (2) assistance with tasks (3) advice or information and (4) listening while one expresses beliefs or feelings [15]. For example, to assess task support, pregnant adolescents were asked, “In this pregnancy, did you get help from anyone with things you had to do such as errands, household tasks, or childcare?” Responses were recorded as yes or no. Then, a value of 1 was assigned if the subjects answered yes to any of the support items and 0 otherwise. The values were summed to produce a perceived support scores. The scores were dichotomized as received support for pregnant adolescents having a value of 1 and above and did not receive for those who had a value of 0.

We use the definition of malnutrition during adolescent pregnancy in this study to refer to MUAC of pregnant adolescents below 22 cm [16]. A MUAC of each adolescent mother’s left arm at the mid-point between the tips of the shoulder (olecranon process) and the elbow (acromion process) was measured using inelastic insertion type tape. Measurement of MUAC was made while the woman stood-up relaxed with her hand hanging down. After checking that the tape was applied with correct tension (neither too loose nor too tight), the MUAC of the woman was read and recorded to the nearest 0.1 cm. MUAC measurement was taken with a pair of data collectors following the recommended measurement technique [17].

The data were collected by trained female data collectors. A 2 day intensive training prior to the pre-test and an additional day of training was given with the final version of the questionnaire before the actual interviews. Trained supervisors and principal investigators conducted field supervision. All completed questionnaires were collected by respective supervisors and checked overnight and doubtful interviews and MUAC measurement were repeated.

The basic characteristics of the study participants are given in Table 1. A total of 425 (97%) pregnant adolescents were included in the analysis. Their mean age was 17.1 (SD ±1.2) years. Two-thirds (66.1%) of the study participants were in 15–17 years age group and 68.2% of them were illiterate. Nearly half (48.8%) of the study participants were in the third trimester of their pregnancy and more than half (51.1%) of them were primigravidae (Table 1).

The mean MUAC (± SD) of the respondents was 22.3 cm (± 2.8). Of 425 pregnant adolescents studied, 112 had MUAC < 22 cm. This gave a prevalence of 26.4% malnutrition (95% CI: 23.0–29.7%). The detail of malnutrition distribution with the examined characteristics of study participants is given in Table 1.

As Table 1 shows, the higher proportion of malnourished pregnant adolescents were living in the households that did not own any livestock compared to those who owned any livestock (30.0% vs 23.8%; P = 0.01). The proportion of malnutrition was 29.3 and 25.7% respectively in pregnant adolescents in the third and second trimesters as compared to 17.7% among those in first trimester (P = 0.01). Furthermore, support in pregnancy (P = 0.015), dietary practice in pregnancy (P = 0.004), workload in pregnancy (P = 0.01) and pregnancy intention (P = 0.02) were statistically associated with malnutrition. However, women’s age (P = 0.10) and their educational status (P = 0.21) were not statistically associated with increased risk for malnutrition.

In the bivariate analysis, unprotected drinking water source (COR = 1.54, 95% CI = 1.23–1.85), not owning any livestock (COR = 1.39, 95% CI = 1.08–1.76), unintended pregnancy (COR = 1.34, 95% CI = 1.10–1.57), being in the second (COR = 1.56, 95% CI = 1.02–2.40) or the third (COR = 1.86, 95% CI = 1.22–2.78) pregnancy trimester and excess physical work in pregnancy (COR = 1.54, 95% CI = 1.17–1.76) were positively associated with malnutrition risk. Improved dietary practice in pregnancy (COR = 0.68, 95% CI = 0.55–0.89) and received support in pregnancy (COR = 0.70, 95% CI = 0.53–0.94) were negatively associated with malnutrition risk. However, no differences were observed in malnutrition risk between age groups, educational status, antenatal visits and previous pregnancies (Table 2).

After adjusting for the potential confounding effects of all other variables considered in the multivariable logistic regression model, livestock ownership (AOR = 1.67, 95% CI = 1.26–2.19), pregnancy intention (AOR = 1.36, 95% CI = 1.08–1.65), excess physical work in pregnancy (AOR = 1.29, 95% CI = 1.02–1.62) and pregnancy trimesters were positively associated with malnutrition risk. Improved dietary intake in pregnancy (AOR = 0.46, 95% CI = 0.33–0.63) and had support in pregnancy (AOR = 0.59, 95% CI = 0.43–0.82) were negatively associated with malnutrition risk (Table 2).

However, age (AOR = 0.81, 95% CI = 0.51–1.28), antenatal visits (AOR = 1.85, 95% CI = 0.92–3.70), previous pregnancies and educational status were not showed significant associations with malnutrition risk. In addition, the associations between farm land ownership (AOR = 0.85, 95% CI = 0.58–1.23) and access to improved drinking water source (AOR = 1.20, 95% CI = 0.94–1.53) and malnutrition risk observed in the bivariate analysis were not maintained in the multivariable analysis (Table 2).