Risky Sexual Behavior Among Orphan and Non-orphan Adolescents in Nyanza Province, Western Kenya

Abstract

A household survey of male and female adolescents was conducted to establish whether orphanhood or other factors contribute to risky sexual behavior. Results show that orphanhood was not associated with risky sexual behavior. Sleeping in a different house from the household head and attending social activities at night were positively associated with sexual activity and transactional sex among boys and girls. Older adolescents were more likely to be sexually active while urban residents, and those who perceived their caregivers as able to provide for their basic needs, were less likely to have ever engaged in sex. Condom use at last sex was associated with older age and having talked with parents/caregivers about sexual risks. Interventions should address these predictors, promote risk reduction among all adolescents irrespective of orphan status, and strengthen parents’/guardians’ capacity to discuss sexuality with adolescents and to provide for their basic needs.

Resumen

Una investigación entre adolescentes masculinos y femeninos fue hecha para determinar si ser niño huérfano es un factor fuerte en el comportamiento sexual de riesgo o son otros factores. Los resultados demuestran que ser niño huérfano no es asociado directamente con el comportamiento sexual de riesgo. Dormir en otra casa que la casa de la cabeza de familia y participar en actividades sociales nocturnos (por ejemplo en discotecas) han sido asociados positivamente con las relaciones sexuales entre muchachos y muchachas por dinero u otras cosas. Los adolescentes mayores eran más sexualmente activos. Los adolescentes urbanos y los adolescentes que encuentran sus padres o sus acogedores en condiciones de poder satisfacer sus necesidades básicas demostraban menos comportamiento sexual de riesgo. El uso de condones la última vez que los adolescentes tuvieron contacto sexual había sido asociado con una edad más alta del adolescente y con conversaciones de riesgos sexuales con los padres o los acogedores. Es recomendable desarrollar intervenciones dirigidas a estos indicadores de comportamiento de riesgo que promuevan la reducción de este comportamiento independiente de ser huérfano o no, que fortalezcan la capacidad de los padres o los acogedores para hablar de temas sexuales con los adolescentes y que los ayuden a satisfacer sus necesidades básicas.

Appendix

Operational Definitions of Key Variables

Sexual Behaviors

Three sexual behaviors were measured: ever had sex; if yes, condom use at last sex; and ever engaged in sex in exchange for money, gift, or favor (transactional sex).

Adolescent Socio-demographic Characteristics

These included orphan status, age, gender, engaging in income-generating activity and residence (urban vs. rural). An adolescent was categorized as a non-orphan if both parents were alive and an orphan if one or both biological parents were dead. Engaging in income-generating activity was defined as having worked for pay or engaged in business/trade the previous month.

Adolescent Socio-cultural Behaviors

Three socio-cultural behaviors were measured: attending funeral discos/other night social activities in the past 12 months; normal location for sleeping, either in the same house as the household head or in a different house; and discussions between parent/caregiver and the respondent about sexual and reproductive health.

Perceived Parent/Caregiver Capacity to Provide Adolescent Basic Needs

Eleven items were identified through prior qualitative research, including school fee/levies, school uniform, soap, body oil, shoes and others). Each item was measured on a three-point scale: most of the time [2], sometimes [1], or never (0). Items were interrelated and subjected to factor analysis which revealed one dimension (Cronbach’s alpha = 0.88). Items were added which resulted in a scale range from 0 (no capacity) to 22 (capacity to provide all). To provide descriptive statistics and establish preliminary associations between variables, the scale was re-coded into three categories of low (0–8), moderate [9–15], and high [16–22] based on the frequency distribution. Original scale values were used in multivariate logistic regression analyses.

Psychosocial Support at Home

Six items (e.g. having someone at home to talk to about their feelings, experience of scolding/ridicule, beating, discrimination) assessed the level of psychosocial support at home on a one-dimension scale (Cronbach’s alpha = 0.83) ranging from 0 (no support) to 6 (high support). This scale was adapted from a social support scale developed for a study among secondary school students on HIV and AIDS life skills in Kwazulu Natal, South Africa [40].

Discussions between parent/caregiver and respondent about SRH

Three items were measured on a one-dimension scale (Cronbach’s alpha = 0.98) ranging from 0 to 3, identified through prior qualitative research. Adolescents were asked if their parents/guardians had ever discussed preventing pregnancy, ways to prevent sexually transmitted diseases, or HIV/AIDS infection and prevention.

HIV/AIDS knowledge

Eight items were measured using a scale (Cronbach’s alpha = 0.74) adapted from a 12-item HIV/AIDS knowledge scale which had been validated in a study among school-going adolescents in Yemen [41] and which was itself adapted from the Multiple Indicator Cluster Survey Questionnaire [42].