The Moderating Role of Testosterone in a Values-Alignment Healthy Eating Intervention

• In Event: Poster Session 01
  In Poster Session: PS 01 Topic - Prevention and Interventions

Abstract

In adolescence, the driving factors of myopia (valuing immediate rewards more so than distal rewards) and need for autonomy often clash with tedious but important positive behaviors (e.g. eating healthily). Specifically, having authority figures demand they stop eating junk food (autonomy threatening) so that they can grow up healthy (distant/abstract reward) often falls on deaf ears. Previous work has established that providing a more immediate symbolic reward (fighting for social justice and displaying independence) through an expose on junk food companies leads middle schoolers to adopt healthier eating habits. Yet, what developmental mechanisms are underlying the effectiveness of this approach? In other words, why might a values-alignment approach be especially effective amongst adolescence, for whom traditional behavioral change messages fail more often than not?

We proposed that Testosterone was the driving factor in the expose’s effectiveness as it indicates a shifting value structure that closely aligns with the intervention message. Testosterone is a gonadal hormone that surges to higher levels during pubertal development and has been linked to increased reward sensitivity to cues related to status and respect (Eisenegger, Haushofer, & Fehr, 2011; Braams et al., 2015). Previous work indicated that these effects are more consistent in males more so that females, thus our current analyses focus on male participants. Thus, in a pre-registered moderation analysis we predicted that males who were higher in Testosterone (i.e. further along in pubertal development) would show greater treatment effects.

In order to test this prediction, we administered the expose to eighth grade students and collected cafeteria purchasing data from the year (total K = 58, N = 10,320). Across 2 days, participants gave 2 mL of saliva which were assayed for testosterone via LC-MS and averaged across days. The main outcome consisted of cafeteria purchases; for each day (180 total) students received a 1 for unhealthy purchases, 0 for no purchases, and -1 for healthy purchases (higher values mean unhealthy).

As predicted, we found a three-way interaction of condition x time (pre/post intervention) x testosterone was significant b condition x time x testosterone = -0.040, p =.007. We conducted a region of significance test to determine at what theoretical cutoff point of testosterone would there be no longer be a treatment effect (i.e. difference between control and expose). We found that at only above 0.36 SD above the sample mean of testosterone do we observe significant treatment effects but not at lower levels (fig. 1).

Additionally, we ran a moderated mediation model to explore if feelings of autonomy associated with healthy eating (i.e. “Eating healthy is a way to stand up to people who are trying to control us”) affected cafeteria purchases and how testosterone moderated this relationship. Results indicated a partial mediation model. The indirect effect of autonomy was moderated by testosterone levels such that only at high levels of testosterone (ROS = +0.25) was the conditional indirect effect of autonomy significant (fig. 2). Implications for intervention work are that values-alignment interventions may be especially affective with more pubertally mature adolescents.

Authors

• Fortunato Nicholas Medrano, University of Texas at Austin
  Presenting Author

• Christopher J. Bryan
  Non-Presenting Author

• Robert Alan Josephs, University of Texas at Austin
  Non-Presenting Author

• David Scott Yeager, University of Texas at Austin
  Non-Presenting Author