Risky Sexual Behaviors in Emerging Adults: Longitudinal Changes and Within-Person Variations

Abstract

This study examined longitudinal changes in condom use inconsistency and sexual encounters involving alcohol use and the within-person associations of these two risky sexual behaviors with other personal and contextual factors. Data were drawn from a sample of college students who completed surveys on four occasions across 3 years and included 317 participants (48 % male; 32 % African American, 28 % Latino American, and 40 % European American) who reported recent penetrative sexual activities on at least one of the occasions. Multilevel models revealed that, although condom use inconsistency increased and then leveled off over time, sexual encounters involving alcohol use showed a linear increase. Moreover, at times when students held more negative attitudes toward condoms than usual, they used condoms less consistently than usual; at times when students felt more anxious about HIV/AIDS than usual, they had more sexual encounters involving alcohol use than usual; and at times when students were involved in a serious relationship, they used condoms less consistently and had fewer sexual encounters involving alcohol use than usual. Findings demonstrate the utility of a developmental perspective in understanding sexual behaviors, the importance of examining the unique correlates of different risky sexual behaviors, and the distinctiveness between within-person versus between-person associations.

Appendix

Equations for Multilevel Models of Risky Sexual Behaviors

Level 1 equation.

$$ \begin{aligned} {\text{Risky sexual behaviors}}_{\text{ij}} = \,& \beta_{{0{\text{j}}}} + \beta_{{ 1 {\text{j}}}} \times {\text{time}} + \beta_{{ 2 {\text{j}}}} \times {\text{time}}^{ 2} + \beta_{{ 3 {\text{j}}}} \times {\text{WP hedonistic outcome expectancies}} \\ & + \beta_{{ 4 {\text{j}}}} \times {\text{WP anxiety about HIV}}/{\text{AIDS}} + \beta_{{ 5 {\text{j}}}} \times {\text{WP romantic involvement}} + r_{\text{ij}} \\ \end{aligned} $$

Level 2 equations.

$$ \begin{aligned} \beta_{{0{\text{j}}}} = & \gamma_{00} + \gamma_{0 1} \times {\text{gender}} + \gamma_{0 2} \times {\text{African American}} + \gamma_{0 3} \times {\text{Latino American}} \\ & + \gamma_{0 4} \times {\text{Greek affliation}} + \gamma_{0 5} \times {\text{parental education}} + \gamma_{0 6} \times {\text{BP hedonistic outcome expectancies}} \\ & + \gamma_{0 7} \times {\text{BP anxiety about HIV}}/{\text{AIDS}} + \gamma_{0 8} \times {\text{BP romantic involvement}} + \gamma_{0 9} \times {\text{study completion}} + u_{\text{oj}} \\ \end{aligned} $$

$$ \begin{aligned} \beta_{{ 1 {\text{j}}}} =\, & \gamma_{ 10} \\ \beta_{{ 2 {\text{j}}}} =\, & \gamma_{ 20} \\ \beta_{{ 3 {\text{j}}}} =\, & \gamma_{ 30} \\ \beta_{{ 4 {\text{j}}}} =\, & \gamma_{ 40} \\ \beta_{{ 5 {\text{j}}}} =\, & \gamma_{ 50} \\ \end{aligned} $$

Note. WP = within-person; BP = between-person. Residual terms are absent for variables treated as fixed. Subscript i indicates occasions within individual j, and j indicates individuals.