Attitudes towards early marriage and related factors among adolescent girls in Ethiopia

Background Early marriage, although it remains a common practice in sub-Saharan Africa, has detrimental effects on women’s sexual and reproductive health, social networks, and educational attainment. We sought to investigate the association between attitudes towards early marriage and marital status among adolescent girls in Ethiopia. A cross-sectional study was conducted among 3,013 adolescent girls aged 13-17 years of age, living in four districts of the Oromia region of Ethiopia. Sociodemographic characteristics and were participants’ attitudes towards early marriage were measured using a pre-tested structured questionnaire. Abstract Background: Early marriage, although it remains a common practice in sub-Saharan Africa, educational attainment. We sought to investigate the association between attitudes towards early marriage and marital status among adolescent girls in Ethiopia. Methods: A cross-sectional study was conducted among 3,013 adolescent girls aged 13-17 years of age, living in four districts of the Oromia region of Ethiopia. Sociodemographic characteristics and were participants’ attitudes towards early marriage were measured using a pre-tested structured questionnaire. Results: The prevalence of early marriage among adolescent girls was 8.3% and approximately 87.0% of girls had an overall positive attitude towards early marriage. Compared to adolescent girls who were unmarried, those who were married or living with a partner were less likely to believe they should have a say whether they want to marry or not (OR=0.32; 95%CI: 0.18-0.57), believe they should get to decide when to marry (OR=0.52; 0.31-0.88), and believe they should have the final decision over the decision to marry (OR=0.48; 0.32-0.73). Conclusions: Interventions to prevent early marriage should strengthen women decision-making processes and promote women’s agency.

has detrimental effects on women's sexual and reproductive health, social networks, and educational attainment. We sought to investigate the association between attitudes towards early marriage and marital status among adolescent girls in Ethiopia.
Methods: A cross-sectional study was conducted among 3,013 adolescent girls aged [13][14][15][16][17] years of age, living in four districts of the Oromia region of Ethiopia. Sociodemographic characteristics and were participants' attitudes towards early marriage were measured using a pre-tested structured questionnaire.
Results: The prevalence of early marriage among adolescent girls was 8.3% and approximately 87.0% of girls had an overall positive attitude towards early marriage.
Compared to adolescent girls who were unmarried, those who were married or living with a partner were less likely to believe they should have a say whether they want to marry or not (OR=0.32; 95%CI: 0.18-0.57), believe they should get to decide when to marry (OR=0.52; 0.31-0.88), and believe they should have the final decision over the decision to marry (OR=0.48; 0.32-0.73).
Conclusions: Interventions to prevent early marriage should strengthen women decisionmaking processes and promote women's agency.
Keywords: early marriage, marriage, child marriage, attitudes Background Adolescence is a crucial period in development, during which increasing social awareness, autonomy, and experienced-based learning mark the physical transition into adulthood (1)(2)(3). However, many young girls in patriarchal societies are denied this period of adolescence (2). Instead, they take on roles as wives and mothers due to early marriage, defined as marriage that occurs before the age of 18 (1,2,4). Early marriage effects the health and well-being of adolescent girls including through increased risk of HIV and sexually-transmitted infections (1, 3), reproductive complications (5), social isolation and low educational attainment (6).
Recent literature suggests expectations, attitudes, legality and acceptability of early marriage practices vary widely according to perceived social norms and normative beliefs on marriage and gender roles, socioeconomic position, educational attainment, religiosity, and region (7)(8)(9). Employment opportunities for educated women and access to health services also impact prevalence of early marriage (1,7,8). High socioeconomic position and level of education are associated with decreased likelihood of early marriage (1,7).
Alternatively, women with lower socioeconomic status are more likely to endorse early marriage as a means to alleviate the economic burden of early rearing (2).
Few epidemiologic studies have focused on the attitudes of adolescent girls towards early marriage. There is a clear need for research that acknowledges the regional and districtlevel differences in early marriage across sub-Saharan Africa. In sub-Saharan Africa, the age of marriage is steadily rising; however, this region still has the highest rate of early marriage in the world (4). In Ethiopia, over 40% of 20 to 24-year-old women reported having been first married before the age of 18 (7). Among 15 to 49 year-old women, the median age of marriage is 16.5 years old in Ethiopia (1). Ethiopia is a country with over eighty ethnic groups, cultural norms, practices and perspectives are subject to wide variations. It is critical to examine the attitudes of adolescent girls, which inform the development of meaningful intervention strategies for decreasing the prevalence of early marriage. Our objective was to investigate attitudes towards early marriage and related factors among adolescent girls in Ethiopia. We utilized a baseline quantitative survey to examine girls' attitudes and behaviors in regard to marital timing, attitudes in marital decision-making, and perceived benefits of early marriage.

Study participants and setting
This quantitative study conducted from June 2016 to August 2016 among adolescent girls in Ethiopia. The study took place in four districts or 'woredas' (Boke, Chiro, Doba, and Mesela) of the West Hararghe Zone in the Oromia Region. A census of all households was conducted in each district to identify households with eligible adolescent girls between the ages of 13-17 and male and female references. Survey respondents were chosen using a simple random sampling technique. Participants consisted of three individuals per household: girls between 13 and 17 years of age and both a male and female reference from the same adolescent girls' household. If more than two influential adult family members were identified, then the two most influential family members were selected based on the findings of the prior qualitative assessment. If there was more than one eligible adolescent in the household, the oldest adolescent was selected.
Data were collected using a structured and pre-tested questionnaire. All study tools were developed in English and then translated into Affan Oromo, the local language. Open Data Kit, an electronic data collection program was used to record data. Data collectors and supervisors were recruited and trained for three weeks. Data collectors submitted completed data to supervisors on a daily basis. Validation was performed to assess competency levels of data collectors and supervisors in data collection by sampling predetermined responses. Supervisors reviewed data and submitted it to the server at the Addis Continental Institute of Public Health (ACIPH). The data management team at the ACIPH regularly reviewed data and provided feedback to the supervisors in the field.

IRB approval and informed consent
All study procedures were approved by the Addis Continental Institute of Public Health Ethical Review Board. All necessary permissions were obtained from the Oromia Regional Health Bureau and West Hararghe Zone. Informed consent was obtained from all participants. For participants below the age of 15 years old, additional parental/guardian 6 consent was sought. All interviews among participants took place in a private setting.

Attitudes towards marriage
Adolescent girls' attitudes toward marriage were measured on a four-point Likert scale as follows: "(1) agree a lot," "(2) agree a little," " (3) disagree a little," " (4) disagree a lot." Individual attitude towards marriage questions were collapsed to agree vs. disagree with the statement. Based on the cumulative score and question wording, respondents were further classified as having a positive or negative attitude towards early marriage (10,11). Answers corresponding to negative attitudes towards early marriage were coded as 0 and those corresponding to positive attitudes as 1. The total marriage attitude score ranged from 0-6. A median split was used to dichotomize those below the median (negative attitude) vs. those above the median (positive attitude). For the purposes of our analysis, we defined positive attitude as ≥ 1. The reference group is always the negative attitude towards early marriage (Supplementary Table 1).

Additional covariates
Sociodemographic characteristics were measured in relation to adolescent marital status using a household questionnaire. Sociodemographic characteristics were examined using pretested questionnaires. Adolescent girls were asked were asked their age (in years), religion (Muslim, Orthodox Christian, Others), occupation (Farmer, Household Work, Student, Others), source of income (no vs. yes), wealth index (Lowest, Second, Middle, Fourth, or Highest), household food insecurity (food secure, mildly food insecure, moderately food insecure, severely food insecure). Marital status was dichotomized as never married vs. married or living with a partner.

Analytic population
Analyses for this study were restricted to girls aged 13-17. A total of 3,238 adolescent girls were approached for an interview. 52 adolescent girls were excluded due to missing 7 information on consent to participate in the survey. A total of 3,186 girls were interviewed. For this analysis, 67 girls were excluded due to the marital status of 'divorced' and 20 were excluded due to the marital status of 'separated'. Adolescent girls were also excluded due to answers of 'don't know' or 'refuse' on marriage attitude questions as follows: 'Adolescent girls should marry after they turned 18' (N=30), 'Adolescent girls should have a say whether they want to marry or not' (N=26), 'Adolescents girls should get to decide when to marry' (N=22), 'Adolescent girls should decide who to marry' (N=21), 'Adolescent girls should have the final decision over the decision to marry' (N=26), and 'Adolescent girls should talk to their parents before making the decision to marry' (N=24). After exclusions, 3,013 participants were included in the analysis.

Statistical analysis
Sociodemographic characteristics was represented as counts and percentages (%). For continuous variables, P-value was calculated using an independent-sample t-tests (ANOVA), and for categorical values, P-value was calculated using Chi-square tests. The frequency of participants' negative or positive attitude towards marriage was also represented as counts and percentages for each attitude item. Multivariable logistic regressions were used to calculate odds ratios (ORs) and 95% confidence intervals (CIs) for associations between adolescent girls' attitudes towards marriage and their marital status. The following a priori confounders were included in the models: age (continuous), source of income (no/yes), household food security (food secure/mildly food insecure/moderately food insecure/severely food insecure). Statistical analyses were performed using IBM SPSS Statistics 25.0 (SPSS Inc, Chicago, Illinois).

Results
Sociodemographic characteristics of adolescent girls are shown in Table 1. The mean age 8 of girls was 14.84 (SD=1.36) ( Table 1). The prevalence of early marriage was 8.3% (N=249).
The majority of the adolescent girls were Muslim (87.8%), had attended school (82.3%), were currently students (52.0%), and had no source of income (89.7%), and were severely food insecure (64.8%). Compared with adolescent girls who had never married, those who were married or living with a partner were more likely to be older, never attended school, have an occupation of household work or housewife, have a source of income and were more likely to be from a food secure household (p<0.001; Table 1). Adolescent girls' attitudes towards early marital attitude questions are shown in Table 2.
The majority of adolescent girls agreed adolescent girls should have a say in whether they want to marry (84.5%), decisions on when to marry (85.8%), and decisions on who to marry (86.2%). Approximately 75.5% of adolescent girls agreed that they should have the final decision over the decision to marry, yet, only 64.1% of all participants believed that adolescent girls should talk to their parents before making the decision to marry. Overall, 87.0% of adolescent girls held a positive attitude towards early marriage. Table 3 shows the association between adolescent girls' attitudes towards early marriage and their marital status. After adjusting for sociodemographic characteristics, those who were married or living with a partner were more likely to disagree with the statement 'adolescent girls should wait to marry after they turn 18 years old' (aOR=1.39; 95%CI: 0.93-2.07). Additionally, married girls were less likely to disagree with the statement 'adolescent girls should have a say whether they want to marry or not' (aOR=0.32; 95%CI:0.18-0.57), and 'adolescent girls should get to decide when to marry' (aOR=0.52; 95%CI:0.31-0.88) compared to girls who were never married. Similarly, those who were married or living with a partner were more likely to disagree with the statement 'adolescent girls should decide who to marry', although it did not reach statistical significance (aOR=0.64; 95%CI:0.39-1.05). Adolescent girls who are married are significantly less likely to disagree with the statement 'adolescent girls should have the final decision over the decision to marry' compared to girls who were unmarried (aOR=0.48; 0.32-0.73). Compared to those who are unmarried, girls who are married or living with a partner are also less likely to disagree with the statement 'adolescent girls should talk to their parents before making the decision to marry' although this did not reach statistical significance (aOR=1.01; 95%CI: 0.74-1.38). Compared to those who were never married, adolescent girls who were married or living with a partner had a similar positive overall attitude towards marriage (aOR= 1.49, 95%CI: 1.03-2.15; Table 3).

Discussion
Overall, our findings from a total of 2,861 adolescent girls in Ethiopia demonstrate an association between attitudes towards early marriage and marital status of adolescent girls. The prevalence of early marriage was 8.3%, and 87% of adolescent girls held an overall positive attitude towards early marriage. Compared to adolescent girls who were unmarried, those who were married or living with a partner were less likely to believe they should have a say whether they want to marry or not (aOR=0.32; 95%CI: 0.18-0.57), believe they should get to decide when to marry (aOR=0.52; 0.31-0.88), and believe they should have the final decision over the decision to marry (aOR=0.48; 0.32-0.73).
Our findings on the impact of attitudes towards marriage and adolescent girls' marital status align with previous studies. In a population of ever-married women aged 20-29 years in Ghana (N=1,349), child marriage was associated with lower odds of believing that one's life is determined by their own actions (OR=0.42; 95% CI: 0.25-0.72 among women 20-24 years and OR=0.54; 95% CI: 0.39-0.75 among women 20-29 years) (12). Among adolescent pregnant women (N=78) in Bangalore, India, the most common reasons given for early marriage and early pregnancy was traditional practices (64%) and family pressure (57.7%) (13). Among women between the ages of 20-24 in Indonesia (N=6,578), higher income, education, and media exposure had protective effects, while rural residence was a risk factor for early marriage (14). Among 4445 parents or guardians in the Amhara Region of Ethiopia, community beliefs regarding the perceived social norms of early marriage, potential benefits of delayed marriage, and exposure to mass media depicting the legal minimum age of marriage were importance for supporting adolescent girls right and ability to delay marriage (7). Our study has important strengths including well-trained interviewers, random sampling process, and a large sample size. There are also a few limitations. First, the crosssectional design of this study has contributed to general knowledge regarding the prevalence of early marriage specific to the West Hararghe Zone. However, the crosssectional study design means that the temporal relationship between marital status and girls' marital attitudes cannot be fully determined. Second, social pressures may hinder adolescent girls' ability to openly discuss reproductive health and attitudes towards marriage. To mitigate this, data collectors interviewed each respondent separately to ensure confidentiality of survey responses. Third, all questionnaires were translated from 11 English to Oromifaa for data collection. There is a potential for discrepancies in word significance and meaning due to the translation of questionnaire items and participant responses into both Amharic and English. This effect was minimized by validity tests performed by data collectors. Questionnaires were also evaluated for appropriateness of questions, language flow and understandability of the questions by both the respondents and the data collectors.
Given the physical, mental, and social health implications of early marriage, we sought to assess adolescent girls' attitudes towards the practice. Our study can inform future community level interventions in the region. Early marriage threatens to disempower adolescent girls by hindering their access to reproductive control, decision-making power, secondary education, and career fulfillment. National campaigns have portrayed early marriage as a coercive and harmful practice that is detrimental to women's health. Yet, community-driven interventions may be more effective by utilizing the narratives of married adolescent girls to examine the personal implications of early marriage. Through the implementation of community-based intervention models that establish female empowerment and cultural norms, adolescent girls may be more likely to adopt roles as decision-makers in their household and thus escape the practice of early marriage.  Participants were informed about their rights to refuse participation and/or withdraw their consent at any time. All interviews took place in private settings.

Consent for publication
Not Applicable as this manuscript does not include details, images, or videos relating to individual participants.

Competing interests
The authors declare that they have no competing interests.