Sexual behaviors, such as transactional sex, early sex debut, intergenerational sex, inconsistent condom use, and sexual violence, that are associated with poorer health outcomes are often conceptualized as sexual risk behaviors (SRBs) (1). SRBs are reported as some of the major public health problems experienced by adolescents (10–19 years old) globally (2). In sub-Saharan Africa (SSA), SRB partly explains the unacceptably high incidence of HIV infection among adolescents and young women (25% in 2017) (3), and highest burden (19%) of teenage pregnancy compared to any other region (4). For example, a study conducted among school-going South African adolescents found that the odds for ever being pregnant were higher for girls who had two or more sexual partners (5). A cohort study among young women (15–24 years old) in Zimbabwe found that intergenerational relationships were significantly associated with concurrent relationships and an increase in HIV incidence. Moreover, the association between age disparity and HIV incidence did not change over time (6). Most countries with the lowest median age of sexual debut are in SSA (7). In Kenya, the median age for first sexual activity is 17 years, and a significant proportion of adolescents (26% of boys and 19% of girls) have initiated sex by the time they are 15 years of age (7).

Neurobiological developmental processes of the adolescent brain and various socio-ecological factors can contribute to risk-taking during adolescence. There is evidence that a heightened propensity for risk-taking during adolescence partly results from the curvilinear development of motivational behavior arising from the temporal gap between arousal of the socio-emotional system and the full maturation of the cognitive control system (8, 9). The early development of the socio-emotional system coincides with the onset of puberty (between 9 and 14 years); a stage during which impulsive sensation-seeking increases. It is explained that synaptic remodeling during puberty results in diminished inhibitory control, thus the heightened reward salience. Synaptic remodeling peaks by the age of 15 years, after which it begins to decline (9, 10). During late adolescence (i.e., 16 years and above), impulsive sensation-seeking begins to decline toward adulthood (i.e., until mid-20s), owing to the gradual development of the cognitive control system, which improves the self-regulation capacity of an adolescent (10). From a socio-ecological perspective, the engagement of adolescents in SRBs in SSA has been attributed to various factors that include individual level factors, such as male gender, level of religiosity (11–15), and substance use (16–19); interpersonal level factors such as peer influence (14, 20), household poverty, level of maternal education (11–15), and parenting behavior (17, 21–23); community level factors such as neighborhood factors, living arrangement and parenting behavior (17, 21–23), social cultural norms (24), and level of social support (17); and broader societal or public policy factors, for example, the level of accessibility to adolescent-friendly services coupled with the nature of the policy environment (25, 26) have been linked to the level of engagement of adolescents in SRBs in SSA.

Psychological factors, namely, self-esteem and hopefulness, have been associated with health risk behavior (27). It is important to understand how these individual psychological factors relate to the different SRBs, so as to design interventions that are most effective in protecting and promoting the sexual health and well-being of adolescents (28). Both hope and self-esteem are also concepts that carve out from the frameworks of reasoned action (29) and planned behavior (30) as well as positive psychology (31, 32). These approaches seem to agree that individuals tend to behave in a certain manner by considering one's attitude, values, beliefs, and perceived consequences of the behavior. An existing body of evidence suggests that self-esteem is a basic feature of mental health and has also been described as a buffer against negative influences, which actively promotes healthy functioning (33). A study found that stability of self-esteem was associated with self-regulatory styles, self-concept clarity, and goal-related affect (34). Although high self-esteem is generally believed to protect against unhealthy behavior (35), it is also reported that some people with high self-esteem engage in risk behavior, possibly because they employ a variety of self-serving cognitive strategies, such as over-estimating the prevalence of that behavior and by avoiding to think about the consequences, which in turn enables them to cope with the inconsistency between their risky behavior and their positive self-perception (36). Indeed, for SRB in adolescence, there have been some studies indicating protective effects of self-esteem (37–39), self-esteem as a risk for engagement in SRB (40), and the lack of significant link between self-esteem and SRB (41, 42). Indeed, the nature of the relationship between SRB and self-esteem remains unclear and warrants further investigation (43), especially in the SSA context where this research is still limited. Hope is conventionally understood as the perceived capability of a person to initiate pathways toward successful goal pursuit and to motivate themselves to use these pathways (44). The evidence indicates that adolescents with higher levels of hope experience a lower risk for internalizing behavior problems (45). It is posited that when operating from a hopeless perspective, adolescents are more likely to replace long-term goals, for example, maintaining a healthy lifestyle and good academic performance, with immediate benefits of engaging in risky behaviors (46). Unlike other forms of health risk behavior, such as substance use and injury-related behavior (47–51), the role played by hope in the sexual behavior of adolescents has been scantily researched (46, 52) and even much more scarcely researched within the SSA context. Furthermore, depression during adolescence is a common global health concern (53, 54). An extensive body of literature also shows that depression is associated with SRB during adolescence, whereby adolescents who report multiple SRB face an elevated risk of depression (55–57), which further underscores the need to investigate these factors simultaneously in studies on adolescent health.

Adolescents, especially in marginalized settings such as poor rural households and urban-poor informal settlements of Kenya, experience a myriad of psychosocial risk factors, which may impact immensely on their sexual reproductive health and rights. So far, there have been no reported studies within rural settings and urban informal settlements of Kenya that examine the important role of psychological factors of self-esteem and hopefulness on the occurrence of SRB during adolescence. Accordingly, the purpose of this study was to investigate patterns of SRB and whether the occurrence of SRB is associated with self-esteem and hopefulness among adolescents from rural and urban-poor settings in Kenya.

Of the 314 adolescents who participated in the study, we successfully appended data on SRB and other study relevant variables of 296 adolescents. We did not come across any emancipated minors during the recruitment. Of the 296 participants, 55% were from urban informal settlements, and the overall mean age was 14 years (SD = 1.6). We did not find statistical significant differences in age between the urban and rural adolescents. The proportion of females was significantly greater (58%) in the urban sample, whereas the rural sample had a significantly higher proportion of males (57%) [P < 0.01]. All the adolescents attended school. There was a greater proportion of participants in upper primary and secondary education level among the rural adolescent (91%) compared with those from the urban setting (61%) [P < 0.001]. Overall, 44 (15%) of the participants reported engagement in some form of sexual behavior, and most of them (61%) were males. Most (43%) of those who engaged in some form of sexual behavior were secondary school students, whereas 30 and 27% were in upper and lower primary education levels, respectively. Of the 44 participants who engaged in some form of sexual behavior, 11% did not experience any of the five forms of SRB, 57% reported engagement in only a form of SRB, equal proportion (14% for each) engaged in two and three forms of SRB, and 4% engaged in four forms of SRB. In total, the adolescents who did not experience any of the five forms of SRB together with those who were abstinent (i.e., never experienced any SRB) comprised 87% of the overall sample in this study.

In this study, we found that most of the adolescents had not experienced any of the five forms of SRB. Given their young age, it is not surprising that they were not sexually active. Of note, a small proportion (about 13%) who experienced SRB is unacceptably high among this young subpopulation. Moreover, 1/3 of those adolescents experienced multiple or co-occurring forms of SRB, which may suggest the presence of multi-faceted sources of risk and vulnerability to sub-optimal adolescent sexual and reproductive health within this group. Noteworthy is that the co-occurrence of health risk behaviors (including multiple forms of SRB) has been previously reported among adolescents in rural Kenya (75). Another study conducted among adolescents from informal settlements of Nairobi reported the co-occurrence of early sex debut and involvement with multiple sexual partners among 14% of participants (76). This finding highlights the need for well-targeted sexual health interventions to concurrently address multiple sources of risk and specific needs of the adolescents within distinct vulnerability profiles.

The findings from this study indicate that risky sexual experiences among the adolescents in this study setting are characterized by a high occurrence of sexual violence, coupled with intergenerational and transactional sex. We found that a large proportion (32%, i.e., 14 adolescents) of the total number of sexually active adolescents had engaged in intergenerational sex and about 43% (i.e., 6 adolescents) of those who engaged in intergenerational sex also experienced transactional sex. Moreover, 1/4 (i.e., 11 adolescents) of the sexually active adolescents indicated that they did not use condoms during their most recent sex. The findings on non-condom use practices, exploitive practices such as transactional sex and intergenerational sex, coupled with experiences of adolescents of coerced sex are corroborated by other studies on Kenyan adolescents in both rural (14, 77) and urban-poor settings (78–81). For instance, a study on 1,365 orphaned adolescents in rural western Kenya found that overall, 9% had experienced sexual coercion and an equal proportion had ever engaged in transactional sex (77). Another study involving 1,754 adolescents from two informal settlements in Nairobi reported that 10% of the sexually active females had engaged in intergenerational sex (an additional 10% had older partners but were not sure of the age difference), and 18% of sexually active females were not at all willing to have sex during their first debut. In the same study, close to 62% (both males and females) of those who had ever had sexual intercourse indicated that they did not use any form of contraception or used a traditional method during their first sexual intercourse (78). As the sample is relatively small, we anticipate some variations in the proportions of adolescents reporting SRB if a much larger sample is utilized. Nonetheless, the findings importantly highlight that among a considerably large portion of young adolescents in this study setting, sexual intercourse often involves little control over safe sex practices, for example, condom use, but also high vulnerability to sexual exploitation and violence (82). These findings stress the need to strengthen the implementation of and the update of existing laws and policies addressing sexual and gender-based violence in addition to other forms of injustices experienced by adolescents within the Kenyan setting (83). Researchers have called for the need to design interventions with cognizance of the structural and social drivers, which underlie transactional and intergenerational sex, as well as the recognition that sometimes girls in transactional relationships may not consider themselves as being exploited (84).

Another key finding was that self-esteem and hopefulness were not statistically significant predictors of SRB once adjusted for depression, education level, gender, and feeling unsafe. Nonetheless, based on the results from the bi-variate analysis, we found a significantly higher score of hopefulness among those who had not experienced SRB, suggesting some potential protective effects of hopefulness against SRB. The result showing that adolescents who engage in SRB report greater hopelessness than those who do not is corroborated by findings from studies conducted among young women from rural South Africa (52) and adolescents from low-income urban communities (46, 85). While this effect was not statistically significant according to the P-value in the multi-variate analysis model, its confidence interval in the model is relatively narrow (0.94–1.01), which may suggest that the study is possibly underpowered to detect an independent direct effect of hopefulness. Similar to the findings on self-esteem, studies on adolescents from Uganda (41) and the United States (42) did not find a significant association between self-esteem and sexual behavior of adolescents. It is plausible that most of the adolescents from both of the rural and urban-poor study settings experience shared psychosocial risks, such as household poverty and suboptimal living conditions, which potentially impact their self-esteem and hopefulness in comparable magnitude. This is also supported by the absence of differences in household socioeconomic status between the adolescents from the rural and urban-poor settings in the analysis. Indeed, poverty and income disparities have been shown to impact self-esteem and hopefulness among adolescents (52, 86, 87). However, in contrast to the findings, studies conducted on adolescents in Nigeria (37), USA (28, 38), and the Netherlands (39) showed that higher self-esteem was associated with lower engagement in SRB. There are also other studies that have found higher self-esteem among adolescents with greater engagement in SRB (40). It is also plausible that hopefulness is a mediating or intermediate factor that could potentially explain mechanisms through which other factors lead to SRB. Such investigation was beyond the scope of this study; however, from exploratory analysis, we observed that in adjusted multivariate analyses, the variable “hopefulness” had remained statistically significantly associated with SRB until the introduction of the variable “feeling unsafe in the neighborhood” into the multivariate model. As there was not a confounding effect or interaction, it is likely that the effects of feeling unsafe in the neighborhood mask those of hopefulness in explaining SRB of adolescents. This assumption is partly supported by the findings from an adolescent cohort study in the United Kingdom, which suggest that perceived neighborhood safety and emotional problems of adolescents influence each other partly because of experiences of risky behavior (88). The findings of this study indicate that the nature of the relationship between SRB and the psychological aspects of self-esteem and hopefulness during adolescence remains unclear and therefore requires more rigorous research. We recommend the use of larger sample sizes in future studies. Higher-powered studies could potentially identify the effect of hopefulness at a statistically significant level. However, it is highly likely that the effect size of hopefulness would still be smaller compared with other predictors such as depression and fearfulness.

The finding of this study indicates that lower education attainment among adolescents who experienced SRB compared with those who had not experienced SRB is corroborated by findings from other adolescent health studies within the Kenyan context (89, 90). It is also reported that while most young adolescents (i.e., those who are 0–14 years old) have not initiated sexual intercourse, some are already exploring intimate relationships without adequate information to make informed decisions about their sexual and reproductive health (91). Higher educational attainment may be one way through which adolescents learn more about their developing bodies and obtain more exposure to sexual health information, which could possibly lessen the vulnerability to multiple SRBs. Therefore, keeping adolescents in school is beneficial. Moreover, early adolescence presents an opportune window of time to provide comprehensive sexuality education, which improves informed decisions about their sexual health (91).

Another key finding was that both depression and feeling unsafe (fearfulness) were associated with the occurrence of SRB. First, the findings on an elevated level of depression and/or feeling unsafe in the neighborhood among adolescents who are vulnerable to health risk behavior (including SRB) are corroborated by similar findings within rural (66, 72) and urban-poor (92) settings of Kenya. Overall, an extensive body of literature indicates that adolescents who experience risky sexual behavior (such as violent and exploitive experiences) experience elevated levels of mental illness (56, 93, 94), as well as negative emotional outcomes such as low self-esteem and diminished positive outlook. Mental illness is also shown to negatively impact on the level of self-esteem and positive outlook of adolescents (95, 96). When left unaddressed, depression and internalizing problems can escalate into life-threatening public health consequences such as self-harm, suicide, serious psychotic conditions, and other mental health-related mortalities (97–99). The findings stress the need for integration of mental health screening, treatment, and prevention services into routine adolescent health promotion programs within this study setting. Addressing the mental health of adolescents in these settings requires more holistic approaches that take into proper consideration various social determinants of health and sources of daily stressors, for example, household poverty, insecurity, crime, and marginalization (100).

This study crucially builds onto a currently under-explored topic on the role of the psychological factors of hope and self-esteem in adolescent sexual and reproductive health, especially in the sub-Saharan context where adolescent health outcomes remain suboptimal amidst scanty adolescent specific data (101). The focus on adolescents from rural and urban poor settings of Kenya is also timely, as adolescents from such settings tend to be ignored and yet are likely to face disproportionately greater vulnerability to suboptimal sexual health outcomes as compared with their peers from other settings. The findings should, however, be interpreted with caution, since a relatively small sample size of 296 was studied. Although comparable with the sample sizes in a number of studies on SRB of adolescents (28, 37, 102), the sample may present some limitations in generalization and detection of some statistical differences. All measures of SRB, self-esteem, and hopefulness were self-reported, and there is a possibility that some level of social desirability bias and recall bias may have been introduced. However, self-reported assessment of health behavior and psychological well-being of adolescents is conventionally utilized and considered reliable (103). Besides, we used measures that have been previously psychometrically tested and pre-piloted within the study setting (58, 70). The study being cross-sectional by design meant that it was not possible to ascertain the direction of the associations explored in this one.

About one-tenth of the adolescents from the rural and urban-poor study context have experienced SRB, among whom 36% report multiple forms of SRB. Interventions addressing SRB of adolescents are likely to be most effective when they are age-appropriate and address multiple drivers of SRB, to concurrently tackle multiple forms of SRB. Noteworthy is that sexual experiences of adolescents were largely marked by sex coercion, age, and economic asymmetry, suggesting the urgent need to design interventions with cognizance of the structural and social drivers of transactional and intergenerational sex, and to prioritize more stringent implementation of laws and policies addressing sexual and gender-based violence plus other forms of injustices faced by adolescents. A need to incorporate mental healthcare and prevention components in adolescent sexual and reproductive health programs within this study setting is crucial. Future studies, preferably with larger sample sizes and longitudinal in design, are needed to further explore the patterns and trajectories of SRB in addition to the role played by psychological factors of self-esteem and hopefulness during adolescence, especially among poor marginalized settings.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by Research Ethics Committee of the Aga Khan University, Nairobi, Kenya. Written informed consent to participate in this study was provided by the participants' legal guardian/next of kin.

CNe, AA, and GS contributed to conception and design of the study. CNa and VA organized the database. DS performed the statistical analysis. DS and AM wrote the first draft of the manuscript. MK, VA, JD, CNa, and VK contributed important details in the methodology sections of the manuscript. All authors contributed to manuscript revision, read, and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.