Sleep disturbance and social reward processing as characteristics linking minority victimization and suicidal ideation in youth

Abstract

Adolescence is characterized by heightened sleep disturbances (e.g., poor sleep quality and irregular/insufficient sleep) and sensitivity to social feedback that may exacerbate suicidal ideation (SI). Victimization experiences (e.g., bullying, humiliation) can contribute to sleep disturbances and SI, particularly among minoritized youth (e.g., sexual/gender, racial/ethnic minorities). However, sensitivity to social reward, despite social challenges, may buffer against the effects of victimization on sleep and SI. In a diverse sample of youth at varying suicide risk, we examined sleep disturbance as a mediator of victimization and SI, and if neural response to social reward moderated the link between victimization and sleep disturbance. Ninety eight youth (14-22 years old; 50% sexual and/or racial/ethnic minority) with varying SI severity provided self-report data on past-six-month identity-related victimization, past-week sleep disturbance, and past-month SI. Seventy four youth completed an fMRI task involving receipt of social feedback. Region-of-interest analyses examined ventral striatum (VS) activity during positive feedback. Mediation and moderation effects were examined using linear regressions. Sleep disturbance mediated the association between identity-related victimization and SI: higher victimization was associated with worse sleep disturbance, predicting more severe SI. Moderation analyses revealed a positive association between victimization and sleep disturbance at lower but not higher levels of VS response to social reward. Sleep disturbance occurring in the context of social stress heightens vulnerability for SI, particularly among minoritized youth. Greater neural sensitivity to social reward buffers against the effects of victimization on sleep, with implications for mitigating SI. Findings suggest potential mechanisms and individual difference factors underlying minority health disparities.

Keywords: adolescence; health disparities; neuroimaging; sleep; social reward; social stress; suicide; victimization.

Figure 1

Description of fMRI task design. Photos with green-colored background indicate rewarding social feedback while those with white-colored background indicate ambiguous feedback. Task procedure: Before scanning, participants rated 40 photos (50% female) of other youth (age-matched to participants) based on how much they would like these peers (rated 1-“not at all” to 9-“very much”). Participants were told that these peers were other participants in the research study and were unaware that these photos were from a standardized database of facial stimuli. Participants had their photos taken, as part of a cover story, and were informed that they would also be rated by other unfamiliar “peers.” Each participant had a personalized stimulus set of 32 photos based on their likeability ratings: 16 photos with rewarding feedback (i.e., informed that they were rated positively by other peers) and 16 with ambiguous feedback (i.e., informed that peers had not rated them yet). To increase salience of social feedback, rewarding feedback comprised the eight highest-rated (mutual liking) and eight lowest-rated (received liking) photos (50% female) and ambiguous feedback comprised 16 moderately-rated photos (50% female) for each participant. Task design: A block design comprising four rewarding and four ambiguous feedback blocks was used to present stimuli. Each photo appeared three times over the eight blocks. Each photo provided either rewarding or ambiguous feedback but not both, and order of stimulus presentation was counterbalanced. Each block comprised 12 photos, lasting for 84-s. To reduce habituation and predictability of the block design, 2 of the 12 photos in each block were of the opposite stimulus type (e.g., 2 rewarding feedback stimuli in the ambiguous block). Each photo was presented for 3-s with a jittered inter-stimulus interval (1, 3, 5, or 7 s), and blocks were separated by an 8-s interval. Participants were instructed to press a button whenever they saw each photo as an attention check during the task. After scanning, participants were debriefed on task deception and informed that the facial stimuli were derived from a standardized database (Martinez and Benavente, 1998) and that their photos were not actually rated by peers.

Figure 2

(A) Describes the mediation of identity victimization and suicidal ideation by sleep disturbance. All regression coefficients presented are unstandardized. (B) The Johnson-Neyman regions of significance plot depicting the interaction between identity victimization and ventral striatum (VS) activation to social reward in predicting sleep disturbance. The line represents slope coefficients (between identity victimization and sleep disturbance) across values of the moderator (VS activation). (A) Mediation of Identity Victimization and Suicidal Ideation by Sleep Disturbance. **p < 0.01; CI = confidence interval. (B) Identity Victimization × VS Activation to Social Reward Predicting Sleep Disturbance. Darker shaded regions indicate region(s) of significance using the Johnson-Neyman technique, i.e., range of VS activation values (x-axis) when the slope between identity victimization and sleep disturbance (y-axis) is significantly different from zero (p < 0.05). n.s. = not significant (p > 0.05).