Determinants of early marriage among married women in Injibara town, north West Ethiopia: community-based cross-sectional study

The study was conducted in Injibara town, Amhara region, Ethiopia from September to December 2018. Injibara is the administrative center of the Awi Zone in the Amhara Region, which is located 447 kms from the capital city of Addis Ababa.

The legal age of marriage is 18 years and above in Ethiopia. However, several girls go into marriage at younger ages in Injibara town. In the study area, marriages are usually organized by the families and heads, and heads from the husband’s family will communicate the spouse’s family and bid moneys as bride value. In addition, marriage also practiced through abducting girls and then sending community leaders and elders to the female’s family for negotiation and to give bride price. During this time, the female’s family will usually decide to accept the marriage and the provided bride price. However, harmful traditional practices such as rapping and abducting women are prohibited in the country.

Community-based cross-sectional study design was used.

All married women in Injibara Town were the source population. This study included all married women aged 15–49 years and who lived at selected kebeles in Injibara town. Those married women aged 15–49 years and who did not stay more than 6 months in the town was excluded.

The sample size was determined using a single population proportion formula by assuming a 95%CI, 4% marginal of error, and the prevalence of early marriage 83% [8]. After 10% of the non-response rate was added; the final sample size for this study was 373.

A multistage sampling procedure was adopted from previously published researches [8, 12] and employed to select study participants. Injibara town has five kebeles (the lowest administrative unit in the government administrative structure) and three kebeles out of five were selected by a simple random sampling technique with the lottery method. The selected kebeles were then subdivided into village/Got. Lists of the households in each village were obtained from the kebele administrative offices. First households were allocated proportionally in each selected village and then households were selected by simple random sampling method from each village by using the number of the household as a sampling frame. The first households were selected from the town using the town’s house number registration by lottery method. In cases of selected households with more than one eligible study subject, only a single respondent was chosen by the lottery method. In cases where no candidate respondent was found in the selected household, the data collectors have gone to the next household until they found an eligible study subject.

Data was collected using a structured interviewer administered questionnaires. The questionnaires were adapted and modified from previous similar literature after considering the local situations [9, 11, 13, 14]. Before pretest, the questionnaires were sent to two experts for evaluation and then some modification was made on the tool after receiving their suggestion. The questionnaires were designed first in English then translated to the local language Amharic by a language expert for data collection and transcribed back to English to check for consistency. Before the actual data collection period; the questionnaires were pretested on 5% of married women in Bahir dar town, thereby adjustment was made on the tool. Data was collected by four nurses and four health extension workers. Two- day training was given on the objective of the study, clarity of the tool and technique and time of interview for data collectors and supervisor prior to data collection. Respondents were asked their age retrospectively at which their marriage agreement was made between them and their first spouse.

Data were checked for completeness and inconsistencies. Then the data were coded and entered into EPI data version 3.1 then exported to SPSS version 23 for analysis. Descriptive and analytic statistics were computed.

Variables with p-values ≤0.2 in bivariate analysis remained in the model as potential confounders for the next level analysis. In multivariable logistic regression; statistical significance was considered at P < 0.05. Adjusted odd ratio (AOR) with 95% confidence interval (CI) was used to measure the strength of association between early marriage and predictor variables. The backward stepwise logistic regression method was used in multiple logistic regressions.

In this study, a total of 373 married women were interviewed. The overall response rate was 100%. Among the total respondents, 167 (44.8%) of respondents had first marriage below 18 years. The minimum and the maximum age at first marriage were 9 and 23 years with the mean age of 17 years with standard deviation of ±3.2 years (Table 1).

All of the respondents (100%) received a bridge price for their marriage and around 15.2% of the respondents left from school for marriage. The majority, 295(79.1%) of respondents didn’t gave their consent at their first marriage (Table 2).

Increase bonding between two families (74%), material benefits (20.4%), unable to cover to educate all children (3.8%) and ensuring virginity up to marriage (1.8%) were reasons for early girl marriage in Injibara town.

In the bivariable analysis, family size, ethnicity, the education status of the father, educational status of the husband, educational status of respondent and family income became significant at the level of 0.2.

However, non-formal educational level of the father [Adjusted Odd Ratio (AOR) =2.32; 95%CI = 1.33–4.05], family’s average monthly income <1000 Ethiopian birr [AOR = 2.32; 95%CI = 1.27–4.24], family size ≥7 [AOR = 3.59; 95%CI = 1.94–6.63] and non-formal education level of the respondents [AOR = 5.16; 95%CI = 2.87–9.28] were found to be associated with early marriage in multivariable logistic regression (Table 3).