Socio-Ecological Associations of the Development of Sexual Behavior in Young Adolescent Girls in the Rural Southern Region of Malawi

Adolescent girls are more likely to experience early sexual debut than boys. However, the developmental context of their sexual behaviors is under-investigated. Using the socio-ecological model and Malawi Schooling and Adolescent Survey, we investigated factors crucial in determining the development of sexual behaviors of 416, 14-year-old girls in rural southern Malawi. We applied Bivariate Logistic Regression analysis to determine associations. Results showed that 353 (84.9%) experienced sexual intercourse, 60 (18.4%) had multiple sexual partners, and 32 (9.1%) used condoms or hormonal contraceptives. Participants’ educational background, desire for higher education, reproductive health knowledge, and being monitored by teachers in school were positively associated with healthy sexual behaviors. Having a boyfriend, lack of schooling support, and being invited to teachers’ homes were positively associated with risky sexual behaviors. Therefore, promoting adolescents’ formal education, sexual health literacy, and safety in schools should be essential components of research and biobehavioral interventions targeting young adolescents in Malawi.


Introduction
While sexual debut is part of normal human development (Greenberg et al., 2017), it can result in undesirable outcomes, mainly when sexual behavior involves having sexual intercourse at too early an age or without due attention to the risks involved (Maswikwa et al., 2015;Pringle et al., 2017).Although all adolescents may engage in early sexual activities, adolescent girls are particularly more likely to having their first sexual intercourse at an earlier age because they tend to enter puberty early compared with boys (Pringle et al., 2017).With early maturing, sexuality and sex-based gender roles begin to assume greater importance in how girls are viewed socially (Ellsberg, 2017), and are more likely to be exposed to harmful practices, such as early and forced marriage, which obligates them to participate in harmful and nonconsensual sexual behaviors (Ellsberg, 2017).Thus, for most girls (∼7%-55%) worldwide, their first sexual experience is forced or coerced (Centers for Disease Control and Prevention, 2021).
Globally, adolescent girls who start having sex at the age of 14 or younger are more likely to have multiple sexual partners, not use any protection, and have an increased risk of having an early pregnancy, HIV infection, and other sexually transmitted infections (STIs), including human papillomavirus, which causes cancer of the cervix later in their lives (Alawode et al., 2021;Epstein et al., 2014;McClinton et al., 2022).While adverse outcomes are observed globally, the consequences of early sexual debut are more in sub-Saharan Africa and other least-developed countries than in developed countries.For example, the most recent available data in 151 countries shows that childbearing among adolescent girls 10-14 years old is higher in 11 countries in sub-Saharan Africa and Bangladesh, estimated at more than six per 1000 girls ages 10-14, compared to regions such as Latin American and Caribbean countries that have adolescent birth rates between one and five Muheriwa Matemba et al.
births per 1000 girls aged 10-14 (United Nations, 2020).In the United States, the birth rate of girls aged 10-14 is estimated at .2 per 1000 girls (Michelle, 2022).Due to early pregnancy, maternal and infant mortality is high, many drop out of school, and suffer childbirth complications such as eclampsia, obstructed labor, cesarean sections, puerperal sepsis, and obstetric fistulas (Stampfel, 2014).In addition, infants born to these mothers are at higher risk of low birth weight, prematurity, and severe neonatal conditions (Ganchimeg et al., 2014).In addition, adolescent girls who experience their first sexual intercourse early have a higher risk of lower self-esteem and depression due to their tendency to internalize negative emotions (Park & Yang, 2017).Low self-esteem and depression influence other unhealthy behaviors, such as alcohol and drug use, that propagate risky sexual behaviors (Lara & Abdo, 2016).
Adolescent girls in rural areas in Malawi face more challenges than those in urban areas which increases their risk of premature sexual debut.For example, over-age school enrollment is high in rural than in urban areas due to the remoteness of schools and long distance as parents do not allow their children to walk a distance of over 4 km when they are not accompanied by older siblings (Ravishankar et al., 2016).Half of the enrollment in the first grade is over eight years old.In addition, most schools are inaccessible during rainy season, which leads to absenteeism and inadequate learning causing more children to repeat grades, making most of them to be enrolled in grades below their age.Consequently, adolescent girls in the rural areas delay in receiving Life Skills Education course, particularly comprehensive sexual education which is offered later in the educational system (Ravishankar et al., 2016).Malawi delivers comprehensive sexual education at the secondary school level, usually in the third year, the time the girls are 15 years or older.By that time, over half would have experienced a sexual debut and would be sexually active, and others would have dropped out of school or would still be in primary school due to repeating grades in (Ministry of Gender Children Disability and Social Welfare [MoGCDSW] et al., 2014;National Statistical Office and ICF., 2017).
Adolescent girls in Malawi also face a lot of barriers in accessing youthfriendly health services.These barriers include long distance to health facilities, long waiting times and inconvenient opening hours, poor attitudes of health workers towards youth, lack of privacy and confidentiality, and lack of participation in developmental programs (Feyisetan et al., 2014;Munthali et al., 2015).Many lack empowerment and decision-making power over their sexuality due to poor education, limited access to information and resources, culture, and low socioeconomic status (Feyisetan et al., 2014;Malawi Ministry of Health, 2015a).As a result, 14.8% of adolescent girls initiate sex before the age of 15 in the rural areas in Malawi compared to 10.1% of girls in the urban areas.Furthermore, in rural areas, 31% of adolescent girls ages 15-19 have begun childbearing in the rural area compared to 21% in the urban area (National Statistical Office and ICF., 2017).
Many studies associate the development of sexual behaviors in older adolescents mainly with factors included in the five domains of the socioecological model: intrapersonal, interpersonal, institutional, community, and public policy (Adamek et al., 2019;Harper et al., 2013;Muchimba, 2019).It is unknown which factors are most important for young adolescent girls in the rural parts of Malawi, where adolescent birth rates and new HIV infections and other STIs are high.Besides, most factors associated with the development of sexual behaviors in young adolescents are studied in isolation.Studying the complex interplay between intrapersonal, interpersonal, and societal factors in the development of sexual behaviors of young adolescent girls allows for a better understanding of the range of factors that make young adolescent girls vulnerable to or protected from risky sexual behaviors.
In addition, in Malawi, and in many other parts of the world, knowledge of the development of sexual behaviors of young adolescent girls (10-14 years old) depends on what is known about older adolescents and women 15 years and older.Young adolescents ages 10-14 differ in many ways from older adolescents at ages 15-19 including in biological and intellectual development, early sexual experiences and their consequences (United Nations, 2020).Sexually active young adolescent girls often face obstacles to accessing contraceptives and health services, thereby increasing the risk of unintended pregnancy and unsafe abortions, and the youngest mothers are the most likely to die due to complications of pregnancy and childbearing (Hervish & Clifton, 2012).However, little is known about factors most important in influencing their sexual behaviors and even less is known about their sexual and reproductive health.Consequently, research and programs targeting the sexual and reproductive health of young adolescents are sparse in comparison to those focused on older adolescents (United Nations, 2020).With the average age at the onset of puberty and sexual maturation, which continues to fall to as low as 10 years or even younger for girls in various settings (Finer & Philbin, 2014;Leone & Brown, 2020), it has become increasingly important to understand factors associated with the development of sexual behaviors in young adolescents.
Moreover, in fewer studies that include younger adolescent girls in Malawi (Dancy et al., 2006(Dancy et al., , 2014;;Nash et al., 2019), investigators usually combine older and young adolescent girls and often combine males and females, making it challenging to define the behaviors unique for young adolescent girls.Therefore, because past research in this study, does not only focus on young adolescent girls, the terms such as adolescents, young people, young women, or youth are used where necessary to reflect this combination.The term young adolescents refer to those aged 10-14 years; young people, young women and youth refer to the 10-24 group, the term young adults refers to those between the ages of 15-24.This study used the Malawi Schooling and Adolescent Study (MSAS) data to determine factors associated with the development of healthy and risky sexual behaviors among young adolescent girls in the rural southern region of Malawi.The study examined the intrapersonal, interpersonal, institutional (school), and community-level factors of the socio-ecological model.The aim was to identify the most critical factors in the development of sexual behaviors among young adolescent girls, the interplay between these factors, and to suggest potential prevention strategies to improve the sexual health of the girls.Healthy sexual behavior in this study refers to the reduced likelihood of having both sexual intercourse and multiple sexual partners by the age of 14, and the increased likelihood of using protection, including condoms and hormonal contraceptive pills and injectables to avoid early pregnancy and sexually transmitted infections (STIs) when young adolescent girls experienced their first sexual intercourse.Risky sexual behavior in this study is defined as the increased likelihood of having first sexual intercourse and multiple sexual partners by the age of 14, and the lesser likelihood of using proctection against early pregnancy, HIV infection and other STIs when they experienced their first sexual intercourse.

The Conceptual Framework of the Study
We used the socio-ecological model for health promotion (McLeroy et al., 1988) to organize factors associated with the development of sexual behaviors according to their level of influence on healthy and risky sexual behaviors.A strength of this model is that it views behavior as a function of the relationship between the individual and contextual factors occurring at multiple levels of influence within the social environment, and addresses the importance of directing interventions at changing intrapersonal, interpersonal, institutional, community and public policy factors which support unhealthy behaviors.(McLeroy et al., 1988).The socio-ecological model for health promotion was particularly appropriate for understanding the development of sexual behaviors for young adolescent girls in this study because evidence shows that sexual behavior is a product of multiple factors (Adamek et al., 2019;Kar et al., 2015).Figure 1 shows how the socio-ecological model for health promotion was applied in this study.We selected the factors in each level based on their associations with the young adolescent girls' sexual behaviors as documented in the literature and in the parent study.

Sexual Behavior Among Adolescent Girls
Recent studies in developing countries show that among 252 million female adolescents, an estimated 38 million (∼15%) 15-19-year-old adolescents are sexually active, (Darroch et al., 2016).In Africa, two-thirds of adolescents in this age group have experienced sexual intercourse and 43% are married (Darroch et al., 2016).A UNICEF study reports that 11% of adolescents under 15 years of age are sexually active, with some children as young as five years old being exposed to sexual activities directly or indirectly (UNICEF, 2016).Condom use among adolescents remains relatively low, despite a high number of multiple sexual partners, especially in countries most affected by the HIV epidemic, including Eswantin, Lesotho, and Botswana (Elflein, 2020).In Malawi, studies show that 34.4% of adolescent females 15-19 years old are sexually active, with 13 to nearly 18% having their first sexual experience before age 15 (National Statistical Office and ICF., 2017;Seff et al., 2021).According to the recent national demographic and health survey, the rate of sexual debut before 15 years is exceptionally high in the southern region of Malawi, currently estimated at 16.8% (compared to 10.8% in the central and 14.7% in the northern region), and 58% of those who are sexually active do not use condoms and other contraceptive methods (National Statistical Office and ICF., 2017).Forced sexual debut is also common in Malawi.Decker et al. (2018); Ministry of Gender Children Disability and Social Welfare [MoGCDSW] et al. (2014) found that 25% of sexually experienced adolescent girls ages 15-19 years characterize their sexual debut as forced, well above estimates pooled across all available data globally (15%) and within the African region (21%).

Factors Associated with Development of Sexual Behaviors in Young Adolescent
Girls.Factors associated with whether and when young adolescent girls start to engage in sexual activities have been described at multiple levels, from their personal characteristics to their interpersonal relationships and environment Figure 1.The social ecological model for health promotion as applied to the development of young adolescent girls' sexual behavior study (McLeroy et al., 1988).
Intrapersonal Level Factors.At the intrapersonal level, a number of factors including educational attainment and aspirations, marriage expectations and knowledge of sexual and reproductive health have been associated with the development of sexual behaviors among adolescent girls.The capability to abstain from risky behaviors has been reported to be associated with the adolescent girls' educational attainment and academic aspirations, intentions and socio-cultural beliefs related to sexual behavior and marriage.In a systematic literature review of evidence to determine the impact of education on sexual behavior of adolescents and adults ages 10 to under 60 years in sub-Saharan Africa, Zuilkowski and Jukes (2011) revealed that educational attainment or enrollment strongly suggest a protective relationship for young women.The study attributed the delay in experiencing first sexual intercourse to students enrolled in schools having a more structured schedule, supervision during the school day, and fewer free hours to spend time with a partner.Enrolling in school delayed the immediate sexual gratification in exchange for a long-term educational or professional goal.The results of this study showed that simply keeping adolescent girls in school extended their abstinence period (Zuilkowski & Jukes, 2011).
In addition, an 11-year longitudinal study showed bidirectional effects between academic goals and achievement with initiating sexual intercourse at a younger age (Schvaneveldt et al., 2001).The study found that adolescents with high educational goals and achievement delayed sexual debut because of the perceived risks including that pregnancy and STIs might jeopardize their plans for the future.Conversely, adolescents who engaged in sexual intercourse at young ages appeared to have reduced interest in academic achievement and goals.However, the specific educational variables most strongly related to adolescent sexual intercourse in this study differed substantially by race and gender (Schvaneveldt et al., 2001).With regard to the intentions to enter into marriage, Mbadu Muanda et al. (2018) in a study conducted in the Democratic Republic of Congo, found that adolescent girls' intention to marry and their preferred age of entering into marriage was influenced by the traditional custom of awarding parents when girls marry while a virgin.This practice motivated some girls to delay sexual practices.In Malawi, the association of the young adolescent girls' academic aspirations and the development of their sexual behavior has not been established.
Comprehensive knowledge of sexuality before adolescents become sexually active is also reported to be associated with the development of their sexual behaviors.Studies report that providing information to young adolescents helps them make informed decisions about the sexual behaviors they adopt, approach relationships with more self-confidence, increase condom use and prevent early pregnancies, HIV infection and other STIs (Kyilleh et al., 2018;Sokhanvar et al., 2014;UNAIDS, 2016a).For many years, the adolescents' knowledge related to sexual and reproductive health has focused on how to engage in safer sex and prevent early pregnancy, HIV infection and other STIs (Clifton et al., 2016;Costa et al., 2018;Greenberg et al., 2017;McManus & Dhar, 2008;Serovich et al., 2009).
Although this basic knowledge of sexual and reproductive health is relatively high in many countries, there are still considerable gaps in comprehensive basic knowledge among adolescents and often knowledge is not consistent with their sexual behaviors (UNAIDS, 2016b).For example, the level of knowledge of STIs, including HIV infection, depends on the type of STIs, with many young people knowing HIV infection more than other STIs.In Uganda, 95% of adolescents 10-14 years knew that HIV infection is an STI, and only 37% knew other STIs besides HIV infection (Kemigisha et al., 2018).Equally, a report of a population-based survey conducted across Africa between 2011 to 2016 showed that, only 36% of adolescent girls and young women in East and Southern Africa, and 24% in West and Central Africa had comprehensive and correct knowledge about HIV infection (UNAIDS, 2016b).Idele et al. (2014) also reported that only 26% of adolescent girls 10-19 years old in sub-Saharan Africa have comprehensive and correct knowledge on HIV infection, and this knowledge was consistent with an increase in early sexual activities and early childbearing.While a cross sectional study in eight sites in sub-Saharan Africa among adolescents ages 10-19 found that 37.3% of the 7116 adolescents included in the study lacked knowledge of menstruation and 55.9% lacked knowledge of other STIs except HIV (Finlay et al., 2020).
In a study about pregnancy prevention and contraceptive use knowledge among 13-19 year old adolescents in Ethiopia, Birhanu et al. (2018) found that only 13.5% mentioned contraceptives as part of sexual and reproductive health services.In Ghana Kyilleh et al. (2018) in a qualitative study among adolescents ages 10-19 to explore their knowledge about reproductive health and choices, results showed that 26.3% of the 80 adolescents (15 in school and 6 out of school) mentioned abstinence and using contraceptives as methods of preventing unintended pregnancies.Most adolescent girls in this study reported using local herbs and wearing beads around the waist during sexual intercourse as methods of preventing pregnancy.The lack of knowledge in this study made adolescent girls vulnerable to unsafe reproductive health practices and inappropriate choices.However, the specific knowledge levels and practices among young adolescent girls aged 14 and under was not reported.
In Malawi, the results of a study by Mwale and Muula (2018) conducted in the rural and urban areas of the Northern Malawi (Nkhatabay, Mzimba and Mzuzu city) among adolescent boys and girls ages 10-19 showed that a majority of respondents had adequate knowledge of HIV infection dynamics (mode of transmission, prevention, and care).Knowledge of biomedical prevention protocol such as abstinence which was estimated at 96.4% and voluntary counseling and testing was estimated at 93.3%.However, knowledge for most adolescents, especially in rural settings, was not consistent with their expected risk reduction outcomes.About 46% initiated sexual intercourse between the ages of 10-14, and 34.2% used a condom at sexual debut.However, the study did not report what exactly young adolescent girls below 15 years knew about HIV infection and how this awareness related to the onset of their sexual behaviors.
Similarly, Kaphagawani and Kalipeni (2017) in Zomba district among 13-19-year-old pregnant teenagers found that contraceptive awareness among participants was at 73.5%, but contraceptive use was at 17.5%.Fears of side effects, sterility, disapproval from sexual partners and condoms disappearing in a woman's body were some of the reasons given by some teenagers for noncontraceptive use.Similarly, Nash et al. (2019) in a study conducted in Mulanje district among adolescent girls 10-18 years and their caregivers found that, although participants were universally aware of the negative consequences of early child bearing and were against childbearing before age 18, all reported that many girls are having babies before age 18 in their village.Caregivers reported introducing sexual health conversations when they notice that the girl is sexually active and mostly negatively influenced adolescents' knowledge of contraceptive use.Adolescent girls reported that their parents strongly disapproved of contraceptive use, believing they are harmful to health and future childbearing in nulliparous girls (Nash et al., 2019).However, results in these studies did not show how much information young adolescent girls 14 years old and younger have related to pregnancy and use of contraceptives, and how that supported the development of their sexual behaviors.
Interpersonal Level Factors.At the interpersonal level, studies have shown that family support including their engagement in adolescent schooling and the relationships they have with their peers including the romantic partners significantly influence adolescent girls' lives and sexual decision-making.The Center for Disease Control and Prevention (2018) indicates that the family's support of schooling including parent engagement in school makes it more likely that children and adolescents will avoid unhealthy behaviors, such as sexual risk behaviors and the use of tobacco, alcohol, and other drugs, which increase the adolescents' risk in engaging risky sexual behaviors.However, there is a dearth of studies on research that examine the association of parental support of schooling and the development of sexual behaviors among young adolescents.A lot of studies have focused on the association of parental engagement in school and academic achievement (Goodall, 2018;Kim, 2019;Wilder, 2014).In Malawi, a qualitative study among parents and teachers revealed a lack of support and involvement in children's schooling (Foster & Child, 2020).However, the association of family support of adolescent girls' schooling and early sexual debut has not been established.
The influence of romantic partners has been reported to be powerful in determining the young adolescent girls' sexual behavior.Adamek et al. (2019) found partners' expectations and intentions to have sexual activities as a risk factor for early sexual debut, particularly for females.The study found female adolescents often reported the first sexual intercourse to be influenced by the need to meet their romantic partners' expectations.Nevertheless, the type and the dynamics of romantic relationship also play a key role in the development of sexual behaviors among adolescents.Kaufman-Parks et al. ( 2023) in a study among 7 th to 11 th graders in Ohio found that respondents reporting on romantic relationships of longer duration were less likely to report casual sex.Specifically, for each unit increase in relationship duration, respondents were 12.9% less likely to report having casual sex in the last two years.Conversely, each unit increase in relationship uncertainty with the primary relationship partner was associated with an 11.1% increased likelihood of the respondent engaging in casual sex.Nevertheless, and perhaps most intuitively, romantic relationships play a key role in the development of sexuality among adolescents.However, there is apparently absence of studies in Malawi that report association of romantic relationship and sexual debut in Malawi.
Institutional (School) Level Factors.At the school level, research shows perceived school safety to be significantly associated with the development of sexual behavior, with those with positive perceived school safety being reported to delay sexual debut (March & Serdar, 2010).In Malawi, Muheriwa Matemba et al. (2022) found a positive association between sexual initiation, engaging multiple sexual partners and not using a protection during the first sexual intercourse with experiencing school-based violence such as being teased, punched, or touched in private areas or hearing sexual comments from schoolmates or teachers.However, there is a dearth of studies that investigate the association between other school safety issues such as monitoring young adolescent girls in school or the practice of inviting adolescent girls in the teachers' homes with the development of sexual behavior among adolescent girls.Kadzamira and Moleni (2008) in a participatory action research in Machinga district in the southern region in Malawi, found that 32.5% of the girls in the study identified availability of a sufficient number of teachers to ensure supervision of pupils as a characteristic of a safe environment but the study did not explore if the girls were monitored in school by the teachers to make sure nothing wrong happens to them and whether monitoring of girls in school was associated with the development of their sexual behaviors.Also, 20% of the girls in the same study reported being assigned to do household chores at the teachers' houses as a punishment for reporting late to school or from break (Kadzamira & Moleni, 2008).Similarly, the proportion of girls whose first sexual experience was a result of being invited in the teachers' home and the association of this safety issue and the development of young adolescent girls' sexual behaviors has not been established in Malawi despite this behavior being the most common practice among teachers in the rural areas.
Community Level Factors.At the community level, studies have associated the development of sexual behaviors with societal norms, particularly in the rural areas.The way girls are socialized to gender norms and sexual roles has shown to influence them to develop healthy or risky sexual behaviors.An ethnographic study conducted in rural northern Tanzania by Wamoyi et al. (2015) showed that parents socialized their daughters in a way of seeing their role as that of serving men while boys were socialized to be independent, and assertive.This socialization made girls develop submissive sexual roles and disadvantaged girls in sexual negotiations, making it harder to resist unwanted sexual activities and insist on condom use (Wamoyi et al., 2015).In Malawi, such socialization is reported to take place during traditional initiation ceremonies.Traditional initiation ceremony is a cultural practice that marks the child's entry in adulthood (Munthali et al., 2018).It is a practice where boys and girls from 6-18 years old are taken to the bush or to a particular house in a community for a period of three to 4 weeks to be given advice on different matters (Munthali et al., 2018).While traditional initiation ceremonies aim at character building including educating young people good behaviors, studies have reported these ceremonies to encourage early sexual debut as girls are advised to have sexual intercourse to mark passage to adulthood (Munthali et al., 2018;Nash et al., 2019).However, the proportion of young adolescent girls having their first sexual intercourse as a result of this traditional initiation ceremony and the association of this cultural traditional practice and the development of sexual behaviors among young adolescent girls has not been established.
Long distance to school is another factor that has been reported in sub-Saharan Africa to expose young people to early sexual activities.The school the child attends is defined as "far" if a child has to travel more than 30 minutes to reach it, irrespective of mode of transport (Hall, 2023).Traveling long distance to and from school every day put girls at risk of having risky sexual relationships and also exposes them to rape and kidnapping.In South Africa, 26% of adolescent girls are reported to have experienced some form of sexual abuse including being touched inappropriately in private areas and rape (Hall, 2023).In Malawi, many children in the rural areas cover a distance of over 4-10 km (km) to get to school from their villages (Sangala, 2019).Covering this long distance to school may feel unsafe due to the increased threat of gender-based violence, especially those who must travel long distances (Bisika et al., 2009;Sangala, 2019).However, little is known if there is a significant association between distance to school and the development of sexual behaviors among young adolescent girls.

Methods
This exploratory study was a secondary data analysis of the Malawi Schooling and Adolescent Study, a six-round longitudinal study conducted between 2007-2013 by the Population Council of New York in collaboration with the University of Malawi (Hewett & Mensch, 2009).The study explored the mechanisms linking school quality and educational outcomes to sexual behavior and the acquisition of HIV and herpes simplex-2 (HSV-2).The parent study enrolled 2649 boys and girls ages 14-17, of which 1764 boys and girls were enrolled in primary school and 885 were not enrolled in school.The participants who were enrolled in school were randomly selected from the enrollment rosters in 59 selected primary schools in Balaka and Machinga districts, while those who were not enrolled in school were drawn from the communities surrounding the selected schools.
Data for the parent study were collected in structured face-to-face interviews conducted by the study enumerators in a private place in their household or somewhere in the community.Sensitive information including sexual behavior and violence were obtained through the audio-computer assisted self-interview (ACASI).With ACASI, software was designed so that the respondent heard both the question and the response categories through headphones, and answered each question by pressing a number on a keypad or computer keyboard.The advantage of ACASI over face-to-face interviews was that neither the investigator nor anyone else in the area where the interview was conducted heard the question or response, thus increasing privacy and reducing the potential for social desirability bias.All questionnaires were translated into Chichewa and Yao the local languages in Malawi, and back translated into English.All interviews were conducted in the local language either Chichewa or Yao (Hewett et al., 2006).
The accuracy of data collection and entry were determined by the trained local study coordinators.The local study coordinator checked all completed questionnaires for completeness and accuracy.Data entry was done locally at the project central research site.Data were entered using CSPro data entry software, controlling for valid responses, skip patterns, and internal consistency.A sub-sample of questionnaires were double entered for each data entry staff to evaluate the validity and reliability of the entered data.If the percentage of errors for a particular data entry staff person exceeded 5% all questionnaires were re-entered and the staff member retrained.Data were exported into Stata for further internal consistency checks and analysis and Muheriwa Matemba et al.
were stored on secure notebook computers with multiple layers of security, requiring login at the computer and application level.The data were backedup daily on external media, encrypted and stored in a secure environment.The ACASI data was merged with the data collected via face-to-face interviews using identifying ID codes for each respondent (Hewett et al., 2006).
The current study used a Correlational design and examined the baseline data of 416 adolescent girls who were 14 years old and in grades 4-8 when they were enrolled in the study.The study explored factors at the intrapersonal, interpersonal, institutional and community levels of the socioecological model that might have been associated with the development of sexual behaviors of 14-year-old girls that were enrolled in school during the first phase of the study.We limited the current study to 14-year-old girls because this was the youngest age group in the parent study and sexual development and behavior change rapidly between the ages of 14-17 and findings might vary across age groups.

Measures
Explanatory Variables.The explanatory variables were grouped by the level of the socio-ecological model.At the intrapersonal level, the variables assessed were educational background, age at first enrollment in school, schooling and marriage expectations, and reproductive health knowledge.At the interpersonal level, the variables examined were household support of schooling and boy/girl relationship.With regard to institutional level factors, the study examined two variables that were related to school safety: whether the teachers monitored the girls to make sure they are not getting into trouble and whether the girls were invited in the teachers' homes for a reason other than to perform chores.At the community level, the variables examined were distance to school which was determined by the time it took the girls to walk to their school and participation in the traditional initiation rite.Table 1 shows the variables at each level of the socio-ecological model, the specific items for each variable and response options.Some variables particularly those related to age, grade enrolled in school and those that assessed knowledge of reproductive health were transformed into dichotomous variable and were coded "1" and "2".All variables with "No" and "Yes" responses were coded "0" and "1".
Outcome Variable.Sexual behavior was the primary outcome and had three indicators: sexual debut, having multiple sexual partners, and the use of protection from pregnancy or STIs.Sexual debut in this study was defined as the first experience of full penetrative sexual intercourse, forced or not forced (Hewett et al., 2006).Two main questions ascertained sexual debut.The first question asked about the age at which the girls first experienced full penetrative sexual intercourse, and the second asked about the number of boyfriends, hit and runs, relatives or teacher who were not a boyfriend with whom the girl has ever had sex.Therefore, to determine sexual debut a variable was created by programming instruction indicating that if age of first sex was indicated 0, meaning that the girls had never had sex, and the total number of boyfriends, hit and runs, teacher, or anyone else is equal to zero that meant the girl has never had sex and was coded 0; and if the responses of these questions were not equal to zero then the girl had sexual intercourse and was coded 1. Engagement of multiple sexual partners was ascertained by asking the girls who reported having had sex, the total number of people they ever had sex with in their life time.This dichotomous variable was coded 1 for those who had one sexual partner and 2 for those who reported having had multiple sexual partners, ranging from 2 to 5. The use of protection was ascertained by asking the girls if they used any protection the first time they had sex.Table 2 below shows the outcome variable and the specific questions that were asked to determine the sexual behaviors of the girls as well as the response options.

Analysis
We conducted descriptive statistics and used frequencies and percentages to characterize the variables (Curtis et al., 2016).We then analyzed the missing values by running Little's test and determined that overall, 3.1% of the values were missing.However, given that <5% missingness is considered a lower threshold below which missing data handling techniques, such as multiple imputation, provide little benefit (Madley-Dowd et al., 2019), we applied listwise deletion, and the excluded cases ranged from 25-45.In addition, several variables related to sexual behavior had a substantial number of nonresponses due to skip patterns for non-applicable questions (e.g., only girls who reported ever having sex were asked about protection at first sex).We used Bivariate Logistic Regression to explore the association between intrapersonal, interpersonal, institutional and community factors and sexual behavior.We did not employ the adjustments to control for experiment-wise Type 1 error due to the exploratory nature of the analysis.We avoided using pvalue cut-offs and wording related to the concept of statistical significance following the recent guidelines advocated by the American Statistical Association (Wasserstein et al., 2019).Alternatively, the relationship between the predictor and outcome variables was estimated using Odds Ratios (OR) and interpretation of confidence intervals associated with an increase in exposure to sexual behavior (Szumilas, 2010), given a 5% probability of making a type 1 error.

Results
Table 1 presents the results of the descriptive analyses of the sociodemographic characteristics and the selected socio-ecological factors.Over 70% of the girls were in senior primary school, and the mean age of entering first grade one was 6.5 ± 1.1.About 35% of the adolescent girls participated in a traditional initiation ceremony, and their reproductive health knowledge was diverse.
Table 2 shows the sexual behaviors of young adolescent girls.Although 85% of the girls reported having experienced sexual intercourse with at least one person, fewer than one-third reported incidents around their first sexual experience.Nearly 6% of these girls reported having concurrent sexual partners.Of the 32 girls who reported using protection during their first sexual intercourse, 24 (75%) used a condom, 4 (12.5%)used hormonal injectables, 2 (6.3%) used pills, and the other 2 (6.3%) used other methods.

Associations of the Development of Sexual Young Adolescent Sexual Behaviors
Table 3 shows the associations between the selected socio-ecological factors and the sexual behaviors of the young adolescents.Associations at the Intrapersonal Level.When subscales were entered in the model as independent predictors, at the intrapersonal level, the results showed that, the girls who were in senior primary (grades 6-8) were 69% less likely to have multiple sexual partners (OR = .31,95% CI = .18,.55)and over six times more likely to use protection (OR = 6.44, 95% CI = 1.54, 27.00) as compared to the girls who were in junior primary (grades 4-5).The girls who expected to attain secondary school level of education were 81% less likely to initiate sexual activities (OR = .19,95%CI = .06,.58),67% less likely to have multiple sexual partners (OR = .33,95% CI = .14,.79),and nearly four times more likely to use protection to prevent early pregnancies and STIs including HIV infection (OR = 3.90, 95% CI = .15,.57)compared with the girls who never had the desire to attain secondary school.Also, the girls who knew abstinence as a method of preventing HIV infection were 61% less likely to initiate sex (OR = .39,95% CI = .16,.99)compared to the girls who did not recognize abstinence as a method of preventing HIV infection.Those with knowledge of the menstrual cycle were 58% less likely to engage multiple sexual partners (OR = .42,95% CI = .18,.99)compared to the girls who never had basic knowledge about the menstrual cycle.
On the other hand, girls who started grade one at an older age when they were 7-10 years old were nearly twice likely to have multiple sexual partners (OR = 1.94, 95% CI = 1.12, 3.35) compared to the girls who started school by the age of 6, the recommended age of starting school in Malawi.The girls who lacked knowledge of pregnancy prevention such as use of contraceptive pills (OR = 2.60, 95% CI = 1.42, 4.77), injectables (OR = 2.68, 95% CI = 1.53, 4.69) and condom use (OR = 1.84, 95% CI = 1.01, 3.37) were more likely to have multiple sexual partners compared to those who possessed such knowledge.In addition, the girls who lacked knowledge of condom use were also 78% less likely to use protection than those who knew about condom use as a method of contraception (OR = .22,95% CI = .07,.73).
Associations at the Interpersonal Level.At the interpersonal level, results in Table 3 show that, the girls who reported being discouraged from attending school were twice more likely to initiate sex (OR = 2.07, 95% CI = 1.01, 4.23) compared to the girls who were not discouraged from attending school.While the girls who were assisted with schoolwork at home were 42% less likely that to have multiple sexual partners (OR = .58,95% CI = .34,.99)and nearly three times more likely to use protection at first sex (OR = 2.96, 95% CI = 1.01, 8.78) compared to the girls who were never assisted with schoolwork at home.Additionally, the girls who reported ever having had a boyfriend, were more than 5 times likely to have initiated sex by the age of 14 (OR = 5.06, 95% CI = 2.81, 9.11), nearly three times more likely to have multiple sexual partners (OR = 2.98, 95% CI = 1.22,4.34) and 89% less likely to have used a protection the first time they had sex (OR = .11,95% CI = .05,.23)compared to the girls who reported never having had a boyfriend.The girls who first had a boyfriend at an older age at 13-14 years were six and half times more likely to have used a protection the first time they had sex (OR = 6.53, 95% CI = 1.24, 34.44) compared to their counterparts who had the first boyfriend when they were younger (10-12 years).
Associations at the Institutional level.The results showed that the girls whom the teachers monitored in school to make sure that nothing wrong happened to them were 63% less likely to have multiple sexual partners (OR = .37,95% CI = .20,.66)compared to the girls who reported the teachers not monitoring them in school.However, we found the results of safety experiences with using protection counter-intuitive.The odds of using protection among the girls who reported being monitored in school were 59% lower (OR = .41,95% CI = 1.00, 1.02), indicating a lesser likelihood of using protection than the girls who were not monitored in school.With being invited in the teachers' homes, we found that the girls who were invited in teachers' homes to do things other than chores were nearly three times more likely to have multiple sexual partners (OR = 2.85, 95% CI = 1.02, 8.14) than the girls who were never invited in the teachers' homes, and were 59% less likely to use protection the first time they had sexual intercourse (OR = .41,95% CI = 1.00, 102).In addition, though the results were not conclusive based on the wider confidence intervals, the results showed that the girls who were invited in the teachers' homes were nearly twice more likely to initiate sex (OR = 1.74, 95% CI = .59-5.16) and were 45% less likely to use protection the first time they had a sexual encounter (OR = .55,95% CI = .50,1.71).
Associations at the Community Level.The results in Table 3 show that based on the wide confidence intervals, the associations at the community level were inconclusive.However, the results showed that, the girls who reported participating in a traditional initiation ceremony at an older age, 11-14 years old, were 13% less likely to initiate sexual activities (OR = .87,95% CI = .40,1.86), 48% less likely to have multiple sexual partners (OR = .52,95% CI = .21,1.31) and were nearly four times more likely to use protection the first time they had sexual intercourse (OR = 3.90, 95% CI = .41,37.40); compared to those who participated in traditional initiation ceremony between the ages of 7-10 years old.

Discussion
This study analysed the Malawi Schooling and Adolescent Study to explore the socio-ecological factors most important in determining the development of sexual behaviors among 14-year-old adolescent girls in rural southern region of Malawi.Specifically, the study examined the sexual behaviors of young Muheriwa Matemba et al.
adolescent girls and factors at the intrapersonal, interpersonal, institutional and community levels of the socio-ecological model associated with the development of their sexual behaviors.The results showed that, nearly 85% of the girls reported experiencing sexual intercourse.Less than 20% had multiple sexual partners, and just below 10% used condoms or hormonal contraceptives to prevent early pregnancy and STIs during their first sexual experience.The study also revealed factors most critical in determining the development of sexual behaviors of young adolescent girls in the rural southern region of Malawi.These factors included educational background particularly being in senior grade, desire to attain higher education, preferred age of entering into marriage, knowledge of menstrual cycle, and awareness of methods for preventing pregnancy, HIV infection and other STIs at the intrapersonal level; household support of schooling and boy/girl relationship at the interpersonal level, and being monitored in school by the teachers and being invited in the teachers' homes at the institutional level.However, the results of the associations were inconclusive at the institutional level of the socio-ecological model.
The finding that 84.9% of the girls reported having sexual intercourse by the age of 14 is of great concern considering the high rates of early childbearing in Malawi currently estimated at 136 per 1000 births among adolescents girls ages 15-19 years (World Health Organization, 2018), and high numbers of new HIV infection cases among adolescents, currently estimated at 2500 infection per year (The World Bank, 2022), and with adolescent girls constituting one third of new HIV infections (Government of Malawi, 2018).This proportion of sexual activity among young adolescent girls is far above any other sexual activity rate ever reported in previous studies in Malawi among adolescent girls below the age of 15, which has been estimated to range from 13%-18%.While this finding requires further exploration, it can be explained by the increase in the number of girls in premarital union and partnerships most common in the rural areas of Malawi including in Balaka and Machinga districts estimated at over 50%-60% (Makwemba et al., 2019;Poulin, 2007).In this study the young adolescent girls started having boyfriends (love partners) from the age of 10.Also, among the those who reported the relationship they had with the person they first had sex with, 11.4% indicated spouse, 8.6% indicated fiancé or promise to marry, while 45.7% indicated boyfriend, suggesting that at the age of 14, the adolescent girls had already developed serious partnership.A previous study conducted in Balaka district, one of the study settings, revealed that when a girl accepts the love proposal or to be in union or partnership with a man she is regarded to have accepted sexual activities, and adolescents were reported to become sexually active around the age of 15 (Poulin, 2007).Hence, Malawi needs adequate support structures for young adolescent girls in the rural parts if the country is to set the stage for positive long-term sexual development.
Notable, however, was the lack of adequate response to age at first sexual intercourse.Only 19.5% of girls reported the age at which they first experienced full penetrative sex, but 84.9% of the girls reported having experienced full penetrative sex by the age of 14.This finding might relate to the way the questions were asked to reveal young adolescents' sexual behavior.The girls were asked once the age at which they first experienced full penetrative sex, but they were asked a series of questions to determine if they ever had sex and with how many people.The girls were asked separately if they ever had sex with their boyfriend, relative, teacher or as hit and run.These questions were programmed to show whether the girl had had sexual experience.This approach was appropriate because eventually, it showed the proportion of those who have had sex which revealed a significant increase.However, the participants were never asked about how old they were when they first had sex with each of the different categories of persons they reported.Instead, they were asked about the total number of distinct categories of people (boyfriend, teacher, relative or hit and run) they have had sex with, suggesting that more emphasis was put on exploring the engagement of multiple sexual partners.Having a clear understanding of the age the young adolescent girls first experience sexual intercourse was essential to determine the appropriate age to introduce interventions to reduce the early sexual debut among the young adolescent girls as previous research shows more adverse outcomes to adolescent girls who experience first sex at the younger age (Alawode et al., 2021;Epstein et al., 2014;McClinton et al., 2022).

Intrapersonal Level Determinants of Sexual Behavior Development
The findings revealed that, at the intrapersonal level, the young adolescents' educational background, schooling and marriage expectations, and their knowledge of the physiology of the reproductive system as well as pregnancy and HIV prevention were important determinants of the development of their sexual behaviors.The finding that being in senior primary school and having desire to attain secondary school education, get married at the age of 20 and over and after completing secondary school and some college were associated with the development of healthy sexual behaviors is consistent with what Zuilkowski and Jukes (2011) found that showed educational attainment to be a protective factor for young women to develop risky sexual behaviors.These results mean that when girls set goals to attain higher education, they focus on their education and have less time to engage in sexual activities.These findings reveal the importance of taking steps to motivate girls to pursue higher education and delay marriage.Encouraging girls to attain higher education is particularly important for rural populations because these girls lack well-educated role models, an important factor for keeping girls in school and delaying early sexual activities (Chikhungu et al., 2020).Only 3.7% of women complete secondary school, and only .9%progress beyond this secondary level of education in rural areas in Malawi (National Statistical Office and ICF., 2017), suggesting that more needs to be done to motivate young girls.
Additionally, at the intrapersonal level, the results revealed that, starting formal education at an older age was positively associated with having sexual intercourse and multiple sexual partners, and was negatively associated with the use protection during the first sexual intercourse.Late enrollment in this study resulted in some girls being enrolled in fourth and fifth grades at the age of 14, the age they were supposed to be in secondary school.This finding shows that these girls were developmentally at a stage of experimenting with sex, but without the information and skills they needed to navigate this terrain because sexual health education is almost non-existent in the lower grades in primary schools in Malawi (Ministry of Gender Children Disability and Social Welfare [MoGCDSW] et al., 2014).This might explain why basic knowledge of preventing early pregnancy and HIV infection was low among girls in this study.For instance, knowledge of pregnancy prevention ranged from 37%-55%, and only 8% of the girls knew that having one sexual partner was a method of preventing HIV infection and 51% of the girls knew that using a condom during sex was a method of preventing HIV infection.However, having basic knowledge of menstrual cycle was negatively associated with having multiple sexual partners, while the lack of knowledge of preventing pregnancies and HIV infection was positively associated with having multiple sexual partners.These findings therefore point to the need for a mandatory primary school entry age of 6 in Malawi and comprehensive sexual and reproductive health education from the early years among young adolescent girls to adulthood in the rural areas.Comprehensive knowledge about sexuality before adolescents become sexually active helps them to make informed decisions about the sexual behaviors they adopt, approach relationships with more self-confidence, increase condom use and prevent early pregnancies, HIV infection and other STIs (Kyilleh et al., 2018;UNAIDS, 2016a).
Of course, many factors can be attributable to the late enrollment and being enrolled in a grade lower than age in Malawi, which makes young adolescent girls lack access to comprehensive sexual health education at the ages they greatly need the information.Regrettably, most of these factors are beyond the control of the young adolescents, their parents, or the teachers.For example, a recent study by the Malawi Ministry of Gender Community Development and Social Welfare and UNICEF (2020) shows that school availability and accessibility remains a challenge.Currently, the majority of children in primary schools in the rural areas still walk a distance of over 4-10 km to get to school from their homes (Sangala, 2019).Although the relationship of distance to school and the development of sexual behavior was inconclusive in this study, it may not feel safe to travel such distances to school because of the increased threat of gender-based violence (Bisika et al., 2009;Sangala, 2019).Moreover, many parents may not be willing to allow their younger children to start school early until they are older and parents are comfortable that the child will be able to walk a longer distance to school considering the lack of transportation for school going children due to increased poverty in families (Nash et al., 2019) and poor educational systems.Malawi needs the government and the community to work together to ensure that Malawi's children have access to school, because doing so puts the most vulnerable girls in age-appropriate grades.

Interpersonal Level Determinants of Sexual Behavior Development
At the interpersonal level, the results showed that household support of schooling and having a boyfriend were important determinants of the development of sexual behaviors among the rural adolescent girls in Balaka and Machinga districts.The finding that some girls were discouraged from attending school in this study was consistent with what Foster and Child (2020) found in a qualitative study in Malawi that parents do not support their children's schooling.Discouraging the girls from attending school is particularly detrimental because girls who are discouraged from getting educated are likely to substitute as their life goal getting married and having children to gain social status, and consequently embark on early sexual activities (Perianes & Ndaferankhande, 2020).Moreover, in rural Malawi, when the girl is discouraged from attending school, parents might have already identified a potential husband, or want the girl to consider marriage to reduce family poverty (Chimombo et al., 2000).
Additionally, this study found a positive association between being discouraged from attending school and experiencing early sexual intercourse.In contrast, being helped with school lessons in the home was negatively associated with experiencing early sexual intercourse and having multiple sexual partners, and was positively associated with using protection during the first sexual intercourse.These findings are critical in a Malawian setting where parental involvement in children's schooling is reported to be scanty (Foster & Child, 2020), and few teachers give students home work in primary schools.But then, the Malawi Ministry of Gender Community Development and Social Welfare and UNICEF (2020) that found the pupils who were given assignments to complete at home at least once a week to be more likely to be motivated and participate actively in school, which can help them delay in engaging sexual activities.It is therefore essential that schools, community agencies, and organizations make a concerted effort to promote school and household support of children and young adolescents' schooling in meaningful ways (Center for Disease Control and Prevention, 2018).Likewise, parents and teachers must be initiative-taking in their engagement with schooling of adolescent girls.
Results at the interpersonal level also showed that having a boyfriend (romantic partner) was associated with an increased likelihood of an early sexual debut and having multiple sexual partners, and the lesser likelihood of not using protection against a pregnancy and STIs during the first sexual intercourse.These results are consistent with previous studies that showed an association between early sexual debut and having romantic relationships (Adamek et al., 2019;Kaufman-Parks et al., 2023). Vanoss Marín et al. (2000) also found young adolescent girls with a boyfriend (mean age 11.5) to be 30 times more likely to have sex compared to those who report not having a boyfriend.These findings suggest the need to empower girls with information on relationships, dating, gender norms, communication, and negotiation skills for them to be able to navigate relationships and prevent risky sexual behaviors.In this study, girls started having romantic relationships at the age of 10.Equipping young girls with knowledge and skills to recognize inappropriate behaviors in a relationship, information about the language they need to speak when the behavior is not right, and information about boundaries, consent, secrets and how to say no, can help them understand what is allowed and empower them to speak up and seek help.
However, simply empowering young girls while excluding boys and men may not really be effective.Hence, boys and men also need training and empowerment to protect and honor girls, and demonstrate genuine and sacrificial love if they are to enter into a relationship.Amoo et al. (2022) emphasize that to save a girl child, the society must train boys and men.Currently, in Malawi and many countries in sub-Saharan Africa, initiatives, campaigns, programs or interventions to reduce early sexual initiation and its consequences including early pregnancies, HIV infection and other STIs are concentrated on girls and women, despite boys and men being involved in sexual activities that lead to all these adverse consequences (Amoo et al., 2022).It is therefore imperative to train, educate and coach boys in sexual and reproductive health education from an early age to protect and care for girls and women starting with their sisters.Increasing the exposure of boys and men to sexual education and counselling can motivate them to be supporters of delayed sexual initiation and safe sex, as well as supporters of only wanted pregnancy, wanted fatherhood, harmonious living and girlfriend or partner empowerment, rather than sexual violence (Amoo et al., 2022).

Institutional Level Determinants of Sexual Behavior Development
At the institutional level, we found monitoring of students in school by the teachers to make sure that nothing was going wrong was associated with the lesser likelihood of having multiple sexual partners and using protection.
While being invited in the teachers' home was positively associated with having multiple sexual partners.These findings imply that teacher-student interaction and relationship is a strong determinant of the development of sexual behaviors in young adolescent girls in the rural Malawi.The finding that some girls reported being monitored in school by the teachers was encouraging as it ensured the girls of their safety in school which has been reported to be missing in Malawian schools (Kadzamira & Moleni, 2008).Monitoring students in schools should be promoted as school safety perceptions have been reported to have a protective effect against risky behaviors (Lessard et al., 2021).
The finding that some girls were invited in teachers' homes to do things other than household chores is consistent with what Kadzamira and Moleni (2008) found that girls were assigned chores at teachers' houses during class as a punishment for coming late from break or for misbehaving.Assigning young adolescent girls chores at the teachers' houses is an unacceptable practice and needs to be discouraged because it puts girls at risk of being coerced by teachers into engaging in risky sexual behaviors.It is therefore imperative that young girls be empowered to recognize healthy and unhealthy relationships with the teachers and be able to know when and where to report suspicious teachers' behaviors.These results also suggest the need to develop strategies that can help teachers as well to recognize their own limitations when interacting with young adolescent girls.The young girls can have the power to prevent sexual activity but many may not be able to overcome the power dynamics of their teachers.Therefore, Malawi as a nation must develop policies and guidelines that protect young girls.

Community Level Determinants of Sexual Behavior Development
Another notable finding of this study is that participation in the traditional initiation between the ages of 11-14 showed the lower likelihood of engaging in risky sexual behavior compared to participating in the traditional initiation ceremony at a younger age between 7-10 years old.Though inconclusive, this finding reveals a crucial point that requires further research.Studies in Malawi have predominantly associated participation in traditional initiation ceremonies with early sexual initiation (Munthali et al., 2018;Nash et al., 2019).This finding reveals valuable information that can be used to inform parents, initiation counselors and traditional leaders to discourage participation in initiation ceremonies among very young adolescents who may not be able to discern what is wrong and what is right.As Nash et al. (2019)

Limitations and Future Directions
The parent study was a longitudinal study that was conducted from 2007-2013, whereas this study was a cross-sectional analysis.Future research should examine the development of sexual behavior at different time points to clarify factors associated with the development of sexual behaviors in young adolescent girls.This study explored associations using bivariate logistic regression.Future studies should also compute multivariate logistic regression to determine the magnitude of the relationship between the development of sexual behaviors and the socio-ecological factors assessed in this study.Also, the study only included girls who were 14 years old and enrolled in school.Excluding girls who failed to attend school and those who were older may have missed the most sexually vulnerable girls.There is also need for future studies of adolescent boys to build a foundation of knowledge about their sexual health and development.This information may help inform how to intervene in the sexual and reproductive health of young adolescent girls and teach boys about the importance of delaying sexual activity.
Also, the study did not explore the factors at the policy level of the socioecological model that would be associated with the development of sexual behaviors of young adolescent girls.It is well known that in addition to individual, interpersonal, institutional, or community-related factors, the adolescents' sexual behaviors occur as a result of the underlying patterns of social systems including economic, legal, and political factors.If these structural elements remain constant, there is a limit to how much impact changes in knowledge, norms, intentions or skills at the individual, household, school, or community level can have.Future studies should explore structural systems including economic, legal, and political factors that can influence the development of sexual behaviors in young adolescent girls.
Another limitation pertains to the methods used to collect data for the original survey, which did not offer an opportunity to explore key areas of knowledge.For example, some girls reported that their first sexual experience was at the age of 6, but the information was not available about the circumstances of the sexual experiences at this younger age.Future research should use mixed methods to clarify many other factors associated with the development of sexual behavior in young adolescent girls.Not all components of a complex developmental process can be described by quantitative methods alone.

Implications
This study revealed a number of significant issues crucial to the healthcare workers' support of parents, communities, and young adolescent girls to develop healthy sexual behaviors and prevent them from experiences that put them at risk of developing risky sexual behaviors.These results can help the health care professionals involved in school health programs, in addition to conducting general health assessments and screening for diseases, to also screen for risky sexual behaviors and identify strategies to help young adolescent girls with sexual development.Health care professionals using existing structures in communities can develop programs that provide young adolescent girls with information and skills and empower them to face and overcome the challenges inherent in adolescence and developing relationships.The finding that delayed school enrollment in primary school was associated with having multiple sexual partners constitute an important target for improving the school infrastructure in Malawi.This knowledge can help the Ministry of Education and Finance and Economic Planning build adequate primary schools and provide school transport for children.Also, the results that being discouraged from attending school and being helped with lessons at home were important determinants of the development of sexual behavior has the potential to influence a change in the way parents and teachers engage in the teaching and learning of children and adolescents in schools.However, more research is needed to find strategies to empower teachers to engage young adolescents in their learning and to promote parent engagement in schools in rural Malawi where literacy rates are generally low.
This study also has a lot of implications for research.The inadequate reporting of the age of first sex and other circumstances of the young adolescents' first sexual experiences, suggests the need for rigorous studies to develop high quality and reliable measures for the sexual behaviors of those who are inclined to giving conflicting information about their sexual behaviors, and who may be reluctant to talk about their sexual experiences.The findings that a lot of adolescent girls experienced sexual intercourse before the age of 15 calls for extensive research to document the sexual behaviors of young adolescents.A proper documentation of sexual behaviors of adolescents below 15 years of age can inform policymakers to consider including significant sexual and reproductive health information in sexual health education materials for this age group which is currently scanty.Our results also suggest the need for Ethnographic and longitudinal studies that include girls from 6 years of age to establish socio-ecological factors associated with sexual debut in young adolescents.Studies are also needed that explore how to empower boys and men to protect young girls from developing risky sexual behaviors.
Results of this study show that knowledge of some components of the sexual and reproductive health was low and there is a significant association between the lack of reproductive health knowledge and the occurrence of risky sexual behavior.Together, these two findings suggest the need for the educational policy in Malawi that advocate for initiating comprehensive sexual education early in the primary school rather than waiting until the girls are in the third year of secondary school when the students are 15 years of age or older and already sexually active.With these findings, the Ministry of Education can also re-organize and strengthen the Life Skills Education course taught in primary schools and empower teachers with knowledge, skills, and positive attitudes to help young adolescent girls set educational and career goals, beginning at their entry into primary school.This improvement could motivate girls to focus on their education to achieve their career goals, which the study found was linked to the development of healthy sexual behavior.
This study also found that participation in traditional initiation ceremony at an older age was associated with lower odds of engaging in risky sexual behavior.Further research is therefore needed to explore age at entry in traditional initiation ceremony and impact on the development of sexual behaviors among adolescent girls.Research is also needed to review the messages that are shared with adolescents during traditional initiation ceremonies and develop strategies and guidelines to bridge the gap between traditional knowledge and conventional knowledge about growing up and sexuality.There is also a need to regulate traditional initiation ceremonies, so that they give age-appropriate information that helps young adolescent girls develop healthy sexual behaviors.

Conclusion
Our analysis of the Malawi Schooling and Adolescent study reveals socioecological factors associated with the development of sexual behaviors, particularly early sexual initiation, engagement of multiple sexual partners, and use of protection to prevent pregnancy and sexually transmitted diseases.The findings at the intrapersonal, interpersonal and institutional levels revealed areas that need further research and intervention to promote young adolescent girls' formal schooling, sexual and reproductive health literacy and school safety.The results also reveal the need to reorganize the implementation of comprehensive sexual education in Malawi and make sure that both boys and girls have access to age-appropriate comprehensive sexual education from a younger age and be empowered to avoid risky sexual behaviors.The non-conclusive results at the community level suggest the need for research to explore further the cultural traditions and practices and the physical and structural infrastructure and how they impact the sexual development of both young adolescent girls and boys.Thus, there is a need for more studies and better interventions that act in synchrony across intrapersonal, interpersonal, institutional, and community levels of the socioecological model if Malawi is to achieve universal access to sexual and reproductive health care services among young adolescents and prevent early sexual debut.

Table 1 .
Descriptive Statistics of the Variables of Interest and Actual Question Items (N = 416).

Table 2 .
Descriptive Statistics of Sexual Behaviors of the Young Adolescent Girls and Questions Used to Measure Sexual Behavior (N = 416).Script: "We will ask you questions about people with whom you have had sex.What we mean by sex is full penetration.We want to find out about everyone you have had sex with in your lifetime, even if it was a "hit and run,"and even if it involved someone like a teacher, employer, or stranger."also, when we ask about sex, we mean someone with whom you have had sex even if you did not want to.For instance, if you were forced by verbal or physical force to have sex."
noted, initiation counselors remain entrenched information sources in Malawi.Incorporating them in efforts to promote healthy sexual development among young adolescents remains essential.Efforts to provide them with training on accurate information on sexual and reproductive health could leverage an existing channel.