Background

Approximately one-third of the African population is between the ages of ten and twenty-four, the highest proportion of youth anywhere in the world [1], [2]. This means that one-third of the population in this region is in the process of transitioning to adulthood, a transition that includes, among other monumental changes, the beginning of a person’s sexual and reproductive life [3]. It is during this transitional period that many young women in sub-Saharan Africa (SSA) first enter into motherhood. SSA has the highest level of adolescent fertility in the world, which significantly contributes to the region’s lifetime average of 5.1 births per woman [4], [5].

Though childbearing is a natural part of the transition to adulthood, evidence shows that a significant portion of adolescent fertility is unintended–either unwanted or mistimed–across countries in SSA [6]. Unintended adolescent pregnancy poses a clear public health challenge, as both unintended pregnancy and pregnancy at a young age are associated with adverse health outcomes for mother and newborn. It is estimated that one-third of all unintended pregnancies in Africa end in induced abortion [7]. This is disconcerting given that the majority of abortions in the region are performed under unsafe conditions due to restrictive abortion laws, lack of knowledge, and inaccessibility of services [6]–[8]. Young women 15–24 account for over half of all unsafe abortions as well as over half of all maternal deaths from unsafe abortions in the region [9], [10]. About a quarter of the growing number of deaths among 10–24 years old women in SSA is due to maternal mortality [11]. Among women who carry their unintended pregnancies to term, adolescents face an increased risk for obstetric complications such as obstructed labor, anemia, and preeclampsia, which can result in long-term morbidities and even death [12]–[14]. Evidence also suggests that women with unintended pregnancies are less likely to seek appropriate prenatal care, placing them at higher risk for complications before and after delivery [15]–[17]. This lack of appropriate prenatal care has implications for the children born to adolescent parents, as they tend to have higher incidences of low birth weight, prematurity, stillbirth, and neonatal mortality [18]–[21]. In addition to acute health risks, adolescent mothers may also face social consequences of unintended pregnancy, including poor educational and employment prospects, marital instability, single parenthood, social ostracism, and an increased risk of getting into or remaining in poverty [4], [22].

The population of sub-Saharan Africa is growing faster than anywhere else in the world, and the strains of population growth are increasingly felt in the region’s urban areas. As sub-Saharan Africa undergoes rapid urbanization, the population of adolescents transitioning to adulthood in informal settlements–or “slums”–is quickly rising [23]–[27]. Slum communities are characterized by a lack of basic infrastructure, poor schooling facilities, high risk of sexual and gender-based violence, high levels of substance abuse and poor livelihood opportunities, all of which negatively impact the health and wellbeing of residents [28]–[30]. Even with overall reductions in the proportion of the SSA population living in poverty, the slums in SSA continue to grow at an alarming rate. Between 1990 and 2010, the proportion of urban residents in SSA living in slum settlements declined from 70 to 62%, yet the population of slum dwellers doubled from 103 to 200 million [27].

Despite the significant proportion of adolescents residing in slum communities, little attention has been paid to the sexual and reproductive health (SRH) challenges they face during their transition to adulthood within this harsh environment. While it is well established that there are social determinants of SRH–which may include education, poverty, social norms, and health systems–little is known about the extent to which living in extreme environments, like slums, impacts SRH, especially during this key developmental period [31]. Studies of adolescents living in informal settlements in Bangladesh, Brazil and Kenya provide some of the few insights into the reproductive health challenges facing this population and suggest that adolescent girls in these communities are especially at risk for unplanned pregnancies given poor education, insecurity within the community, and low-levels of autonomy stemming from cultural norms and the desperation of poverty [32]–[35]. Evidence from slum communities in Nairobi indicate that adolescents living in the slums engage in riskier sexual behaviors than their peers in non-slum parts of cities, including early sexual debut, transactional sex, and multiple sexual partnerships [28], [35], [36]. Additionally, adolescents’ knowledge about contraception is inadequate, and access to contraceptive methods is limited in slum communities, impinging an adolescent girl’s ability to control her reproductive life [37]. This is especially troubling, since adolescents are already at a disadvantage when it comes to accessing contraceptive services in SSA, independent of poverty. In Kenya, for example, 45% of all sexually active unmarried adolescents have an unmet need for contraception [38]. Because the urban poorest population tends to have the greatest unmet need, these estimates are likely even greater among adolescents in the slums [37].

There are adolescents, however, that transition to adulthood in the slums without experiencing a major adverse reproductive health event. Studies from Brazil and Kenya suggest that there are protective factors, such as having a father present in the home, that are associated with delayed sexual debut and fewer incidences of unintended pregnancy among adolescents in slum communities [34], [39]. Further research may be able to elucidate additional factors that help adolescents in these communities lead healthy and fulfilling sexual and reproductive lives.

This paper seeks to expand on this limited body of research by examining the factors associated with unintended pregnancies among young women aged 15–22 living in two informal settlements in Nairobi city, Korogocho and Viwandani. The use of prospective data allows us to investigate the association between pregnancy experience and status between two waves of data collection and various relevant factors captured during the first wave. Unlike other similar studies, we therefore control for the temporal ordering of the explanatory factors and the outcome variables. The findings may help in understanding the circumstances and factors surrounding unintended pregnancies among young people in order to inform policy development, design and implement appropriate interventions, and identify areas for further research. Understanding and identifying ways to address the poor reproductive health outcomes among adolescents and young adults is central to the attainment of the Millennium Development Goals (MDGs) and other development indicators in Kenya and other sub-Saharan African countries. As urbanization continues unabated in the region, the wellbeing of the urban poor will increasingly drive national development indicators. Preventing unintended pregnancies among young women would greatly reduce maternal and infant mortality and disability adjusted life years (DALYs) among this vulnerable age group, increase educational and livelihood opportunities for girls, and ease the strain of population growth on communities and nations. Beyond reducing their adverse health consequences, averting unintended pregnancies among adolescents and young female adults living in urban slums could also contribute to slowing rates of urban growth which is substantially brought about by natural increase in SSA [40]. Ezeh et al. [37] show that in Kenya and other countries in SSA, mistimed and unwanted fertility significantly contribute to the high fertility among the urban poor.

Data and Methods

Results

Descriptive findings

Sample description.

Table 1 presents some key socio-demographic characteristics of the study participants at the time of the Wave 1 survey. In total, 849 girls aged 15–22 observed at Waves 1 and 2 are considered for analysis in this paper. About 78% are aged 15–19, a bit more than half (53%) are from Viwandani, 72% are Catholic or Other Christian; and 78% have never been married. In addition, 49% were still in school at the time of Wave 1 survey; 47% have reached at a secondary level of education.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Percentage of female adolescents (15–22) who reported having initiated sex, by socio-demographic characteristics (as of Wave 1).

https://doi.org/10.1371/journal.pone.0101034.t001

Sexual initiation.

Findings indicate that about 37% of young women aged 15–22 have already initiated sex in their lives, with the proportion being significantly higher among those aged 20–22 (Table 1). Young women living in Viwandani are marginally less likely to report having initiated sex; 36% compared to 39% of their Korogocho counterparts reported sexual debut. Young women who are still in school are less likely to have initiated sex; the proportion of those who have already experienced sex is lower among those with at least a secondary level of education (30% vs. 44%). When considering religion, Muslims exhibit the lowest proportion of those who have initiated sex. With respect to ethnic group, the highest percentage of sexually experienced is observed among young women from Luo and Luyha ethnic groups. As expected, the proportion of those who reported sexual debut is much higher among currently married young women (98% vs. 25%). Young women who live with neither parent are much more likely to have initiated sex. As indicated in Table 1, similar trends of bivariate associations are observed among unmarried female youth. Results on the sexual experience by parental presence among the married women (not shown) indicated that the majority of them do not live with their parents (85%), and that 99% of those young women have initiated sex.

Pregnancy experience and pregnancy intention.

Table 2 shows data on young women’s pregnancy experience and their pregnancy intention at Wave 2 of young women who have already initiated sex by Wave 1 or between Wave 1 and 2, by socio-demographics. About two-thirds of sexually experienced female youth (69%) have ever been pregnant by Wave 2, with the percentage being higher among those aged 20–22. The proportion of those who have already been pregnant is higher among those with primary education. Ever-married young women are significantly more likely to report having experienced a pregnancy (93%) compared with those who have never been married (55%). Those who have had their first sexual experience with their husband/partner, or never used contraception at first sex are proportionately more likely to have ever been pregnant. The percentage of ever pregnant young women is higher among those who live with neither parent, and those who have initiated sex before age 15.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Pregnancy experience and pregnancy intention for young women aged 15–22 who have ever been pregnant.

https://doi.org/10.1371/journal.pone.0101034.t002

In addition, findings show that overall, 41% of pregnancies were unintended, with 26% being mistimed and 15% unwanted. The percentage of young women with unintended pregnancy is estimated at 51% and 31% for 15–19 and 20–22 years old, respectively. Adolescents who have never been married exhibit higher levels of unintended pregnancy (61% vs. 20% for those who are currently or formerly married). The proportion of unintended pregnancy is higher among those who had their first sexual encounter with their boyfriend as compared with those who first experienced sex with their husband/partner (50% vs. 19%). In addition, the percentage of those who have had an unintended pregnancy is higher among those who initiated sex before age 18 and those who live with at least one parent.

Multivariate analysis

Results from multivariate analysis for timing of pregnancy experience (Cox model) and pregnancy intention (logistic regression) are presented in Table 3. Estimations indicate that being in school is negatively associated with the timing of pregnancy experience as young women who are still in school are significantly less likely to have had a pregnancy. Young women from Luhya ethnic group get pregnant earlier as they are as twice as likely to become pregnant compared with their Kikuyu counterparts. Early initiation of sex increases the chances of pregnancy as those who initiated sex before age 18 have had a pregnancy episode earlier than those who experienced first sex from age 18. Young women who live with their mother only or both parents experience a pregnancy later than those who live with neither parent. In addition, marital status, schooling status, age at first sex and living arrangements are the only factors that are significantly associated with unintended pregnancy among these young women. In fact, young women who have initiated sex before age 18 are more likely to experience an unintended pregnancy but the coefficient is statistically significant only for those whose sexual debut occurred between ages 15 and 17. Also, being in school reduces the chances of experiencing an unintended pregnancy (at 10% level). Young women who were married are much less likely to have had an unintended pregnancy. Interestingly, adolescents living with their father only or both parents are much more likely to have had an unintended pregnancy than their counterparts who live with neither parent.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Factors associated with timing of pregnancy experience (Cox regression) and unintended pregnancy among young women aged 15–22 (Logistic regression).

https://doi.org/10.1371/journal.pone.0101034.t003

Discussion

In this paper, we use data from baseline and second round survey of a three-year prospective study, to examine the pregnancy experience and intention of young women aged 15–22 in two informal settlements in Nairobi. About two-thirds of sexually experienced adolescents have ever been pregnant by Wave 2. For 41% of adolescents, the pregnancies were unintended, with 26% being mistimed and 15% unwanted. Findings indicate a significant association between a set of factors including age at first sex, schooling status, living arrangements and timing of pregnancy experience. In addition, marital status, schooling status, age at first sex and living arrangements are the only factors that are significantly associated with unintended pregnancy among the young women. These findings contribute to filling the gap in understanding and meeting the sexual and reproductive challenges and needs of young women living in resource-poor urban settings such as informal settlements in the country.

Overall, our findings show that early sexual initiation was prevalent as about 37% of female youth aged 15–22 had already initiated sexual intercourse, though the proportion of sexually experienced adolescents vary significantly with age, marital status and schooling status. Young women were less likely to initiate sex if they were currently in school, with at least a secondary education and never married. Most notably, the odds for young women currently in school reporting a pregnancy experience are substantially lower than for those who are out-of-school, suggesting that pregnancy may lead to school drop-out among young women. Earlier studies in Kenya have reported that 13,000 girls drop out of school annually due to early childbearing [50]. Although school environment seems to be protective against childbearing, it may also be true that both drop-out and pregnancy are influenced by a same set of factors. This findings call for more school-based studies to help disentangle the association. This result mirrors studies done in other African context were initiation of first sex (before age 18) was less likely during adolescence if the girl was educated or exposed to mass media [35], [51]. Similarly, other studies indicated the role of education in reducing fertility in Kenya and elsewhere in sub-Saharan Africa [4],[52]. This indicates that keeping girls in school and formal education in general might be an effective strategy to enable young girls delay sexual debut. Increased education might therefore act as a protective mechanism that not only prevents early sexual initiation, but more education improves the understanding by young women of the need to protect themselves from unintended pregnancy.

Moreover, about two-thirds of girls with previous sexual experience had experienced a pregnancy; 41% of them reported having had an unintended pregnancy. In addition, our results indicated that age at first sexual intercourse was associated with timing of pregnancy experience on one hand and pregnancy intention on the other. Young women whose sexual debut occurred before age 18 got their pregnancy earlier than those who initiated sex from age 18. In addition, young women who had initiated sexual intercourse between the ages of 15–17 had significantly higher odds of experiencing an unintended pregnancy, indicating that early sexual initiation was associated with adverse SRH outcomes. Early sexual debut is much likely to automatically increase young women’s exposure to the risk of pregnancy, especially in the slum settlements where use of contraception is limited. More than often, young people’s first sexual experience is unplanned, a pattern that predisposes them to irregular and ineffective contraceptive use. Extant literature indicates that premarital sexual activity among adolescents is increasing but less than half are using any modern contraception to avoid pregnancy [53], [54]. It is reported that barriers to accessing health care such as lack of decision-making power, access to and control over resources, and socio-cultural norms regarding adolescent sexual behavior and childbearing contribute to low use of contraception [55]. Moreover, the nature of services offered, who and the attitude of who is offering the services reportedly influences young people’s attitude toward contraceptive use, as well as their ability to express their needs and consequent health seeking behavior [56]. Provider bias and attitude toward adolescent therefore discourages young people from accessing essential contraceptive information and services. This finding highlights the fact that targeted programs are needed to reach adolescents and young adults with a range of SRH information and services at different stages - before they initiate sex or as sexually active unmarried adolescents. Moreover, reproductive health interventions including uptake of family planning which often do not reach adolescents, should be scaled up to include adolescent friendly services, with targeted responses required. Given that schooling was shown to be protective, bringing comprehensive sexuality education into schools and moving beyond the fear that sex education for young people encourages risky behaviors is important. In fact, previous studies show that providing sex education to young people in schools has the potential to be protective, especially in delaying early sexual initiation [57].

Findings also indicate that ethnicity had statistically significant effect on timing to pregnancy between Wave 1 and 2; young women from Luhya ethnic group are more likely to become pregnant than those from Kikuyu tribe. This finding is consistent with higher levels of fertility observed in Western Kenya where Luhyas come from. This may partly due to Kikuyu women’s higher access to education, higher involvement in economic activities that empower them to have a better control over their reproductive lives. Also, Kikuyus are known to be less pronatalist than their counterparts from Luo and Luhya ethnic groups [29], [58], [59].

Living with a mother only or both parents reduced the risks of becoming pregnant. Evidence from other SSA settings indicates that parental monitoring and control reduces the likelihood of involvement in problem behavior and is likely to lead to delayed onset of sexual experience [60], [61]. In particular, living with both parents is associated with lower risk of sexual experience, and therefore lower chances of getting pregnant. It also appears that having the mother in the household is important in preventing pregnancy experience, suggesting that the father’s absence does not deter the mother’s ability to control her daughter’s activities. However, on the other hand, living with a father only or both parents was significantly associated with higher chances of experiencing an unintended pregnancy. In a study on transition to first sex by adolescents in Kenya, Kabiru et al. [35] found that high level of parental monitoring was associated with greater odds of transitioning to first sex for females in the slums, which is positively associated with higher levels of unintended pregnancy. It is noted that the social context of slum environment especially cramped living conditions and shared living space, might expose children to early sexual activity [62]. Furthermore, perception by adolescents of parents being overly controlling has been shown to instead increase risky sexual behavior [63]. Our findings therefore highlight the fact that ways to adequately involve parents in sexual and reproductive health interventions targeted towards adolescents must be explored and tested, especially in resource-poor settings such as slums.

Current marital status is significantly associated with unintended pregnancy, with currently married adolescents being much less likely to have had an unintended pregnancy. However, adolescent childbearing is associated with increased risk of adverse health outcomes, regardless of whether a pregnancy is intended or not. Adolescent fertility still has significant implications not only for the health of the adolescent girl but also for larger demographic processes, especially high fertility rates in sub-Saharan Africa. It is reported that adolescent girls are more likely to die during pregnancy and delivery, because of their physical immaturity [64]. Moreover, younger mothers are also more likely to have babies with significant mortality and morbidity risk. For example, a study of 76 countries showed children born to mothers between the ages of 12–20 had a greater risk of dying before age five, being stunted, underweight or suffering from anemia [64]. Overall, this study underscores the importance of looking at reproductive outcomes of early sexual initiation, the serious health risks early fertility entail, especially among out-of school girls, and sexual activity in general. This provides greater impetus for addressing adolescent and young females’ reproductive behaviors, especially in resource-poor settings such as slums. In particular, it is paramount to identify and dismantle barriers that prevent adolescents and young women living in slum areas, from accessing existing family planning services.

While the prospective nature of the data used helps partially overcome the causality issue that is often encountered when using cross-sectional data, findings in this paper should be interpreted in light of some limitations. For instance, given the relatively short duration between the two waves of data collection, not many episodes of pregnancy were observed among adolescents. As a result, the analysis of pregnancy experience and status is based on a relatively small number of cases. In addition, our analysis could be affected by unobserved potential confounding variables that may influence both the observed explanatory variables and pregnancy experience/status. Finally, retrospective self-reporting of the status of pregnancies might be biased as mother’s feelings about pregnancy might change over time.

Despite these inherent limitations, the findings of this study have far-reaching implications for reproductive health of poor adolescents and young women in Kenya. Two clear policy and programmatic implications arise from our findings. First, it is important to design appropriate and realistic programs that are focused on reducing the number of unintended pregnancies; thereby helping reduce the ensuing induced abortions among adolescents and young women living in poor urban areas. These programs should be cognizant of the fact that these individual level behavior factors are exacerbated by an environment affected by extreme poverty. Poverty has been considered a root cause of poor SRH outcomes especially in slum environments. Targeted multipronged interventions which address the economic, education and gender consideration of young people will therefore have the most direct influence on their behavior.

Although the Kenyan government has made significant strides in addressing the sexual and reproductive health (SRH) needs of young people, including passing significant policies such as the National School Health Policy, the implementation of this policy and its reach to young people is still low. The accelerated implementation of this and other youth related policies to include a comprehensive school health program to address youth SRH and broader health needs are of essence. Finally, finding innovative ways to demystify comprehensive sexuality education and looking for innovative delivery channels by which comprehensive sexuality education can be brought to schools is a necessity. Reaching young women in their early adolescence, while they are still in school, will have a potential of not only reducing risky sexual behaviors, but reduce unintended adolescent fertility, which will have significant impact on their wellbeing.