Anne Julia Maheux

Social media use has been linked to poorer sleep outcomes in adolescents, yet the mechanisms underlying this association remain unclear. Appearance‐related social media consciousness (ASMC)—a preoccupation with one's appearance online—could relate to offline behaviors, including sleep. This study examined the association between ASMC and various sleep behaviors using a parallel mediation structural equation model. Participants were 1158 U.S. 8th‐grade students ( M _age = 13.14, 52% girls, 36% Hispanic/Latine, 37% qualified for free or reduced‐price lunch) surveyed at two points across an academic year. Higher ASMC was indirectly associated with poorer sleep quality, increased sleep disturbances, and longer sleep onset latency through greater compulsive nighttime social media use. Additionally, higher ASMC was indirectly associated with earlier sleep onset times through longer morning routine durations. Findings suggest that appearance‐related experiences on social media have implications for adolescent sleep health and provide preliminary support for some pathways through which its effects may occur.

Generative artificial intelligence (GenAI) is rapidly transforming the context in which adolescents develop, presenting both opportunities and risks. Adolescence is a sensitive life period during which the accomplishment of key psychosocial developmental tasks—including social belonging, emotion regulation, autonomy, and identity formation—leads to lifelong consequences. Here, we synthesize interdisciplinary research on how adolescents’ use of GenAI chatbots such as ChatGPT may have unique impacts during this life stage. Although GenAI may provide low-stakes opportunities to simulate social interactions, scaffold emotional regulation, and support identity exploration in some contexts, it also threatens to displace or thwart essential developmental experiences. Risks include increasing social isolation, undermining distress tolerance, distorting expectations for human relationships, compromising identity and purpose development, and exposing youth to biased, inaccurate, or sexually inappropriate content. We call for a developmental science framework to guide future research, technology design, and policy. Longitudinal studies are needed—especially those that examine specific use cases, address all domains of development holistically, account for varying levels of vulnerability across stages and populations of adolescence, leverage interdisciplinary and academic-industry partnerships, and incorporate the perspectives of youth. As adolescents increasingly integrate GenAI into their daily lives, we must urgently address how these tools are shaping their transition to adulthood.

Importance: Artificial intelligence (AI) chatbot use is rapidly increasing, yet prospective evidence on its mental health implications during early adolescence remains absent.Objective: To examine longitudinal associations between young adolescents' AI use, AI use frequency, and AI use evaluations and internalizing symptoms.Design: Two-wave longitudinal cohort study; data collected December 2025 (Wave 1) and April–May 2026 (Wave 2; approximately 5-month follow-up).Setting: Twenty-two middle schools across 8 school districts in a large southeastern U.S. state.Participants: Parents of all enrolled students with valid email addresses were invited; 2,570 provided consent; 2,342 students assented and initiated the survey.Main Outcomes and Measures: School loneliness (PISA Sense of Belonging at School Index), social anxiety (Social Anxiety Scale for Adolescents), and depressive symptoms (Short Mood and Feelings Questionnaire). AI use and evaluations were measured using items developed by the research team.Results: Participants were 2,342 students in grades 6–8 (M age = 12.8, 47% boys, 49% other, 70% White, 20% Black, 12% Hispanic/Latino, 5% Asian). Prevalence of AI use in past month increased from 62% to 75% across waves. Any AI use was associated with higher social anxiety (concurrent b = 0.27, P < .001; prospective b = 0.22, P < .001) and depressive symptoms (concurrent b = 0.11, P < .001; prospective b = 0.12, P < .001). Frequency of AI use was associated with loneliness (concurrent b = 0.03, P = .007), social anxiety (concurrent b = 0.08, P < .001; prospective b = 0.07, P < .001) and depressive symptoms (concurrent b = 0.06, P < .001; prospective b = 0.05, P < .001). Factor analyses supported three evaluative dimensions: Emotional Intimacy and Companionship, Nonjudgmental Validation, and Overreliance. Emotional Intimacy and Companionship prospectively predicted greater school loneliness (z = 2.98; P = .003). No internalizing symptom variables prospectively predicted AI evaluations.Conclusions and Relevance: AI use and evaluations are related to negative mental health outcomes over time for young adolescents. Rapid policy responses and continued longitudinal research are needed to address the public health implications of AI use and companionship during early adolescence.

Generative AI is the most rapidly adopted technology in human history, with adolescents among its earliest and most active users. This systematic review synthesized 82 peer-reviewed empirical studies on adolescent (defined as mean age between 10–20) AI use published between 2020 and 2025. The literature is characterized by a focus on education, attitudes towards AI adoption, and frequency of AI use, with little attention to diverse use cases of AI (e.g., as a companion, for health information or advice) or diverse socioemotional outcomes (e.g., mental health, peer relationships). Methodologically, the field is characterized by mostly cross-sectional self-report designs (61% of published studies). Only two studies used longitudinal designs, 5 used experimental designs, and 8 focused on scale development and psychometrics. Participant recruitment has been geographically concentrated in Asia (66% of published studies), yet relatively diverse in terms of age within adolescence (29% of studies include early adolescents ages ~14 and younger, 54% include middle adolescents ages ~14–18; 49% include older adolescents ages ~18 and older). Findings were heterogeneous across topics, limiting cumulative conclusions. We synthesize the themes and methods across studies and outline a research agenda to advance a more theoretically grounded, methodologically rigorous, and developmentally informed science of adolescent AI use.

Artificial intelligence (AI) chatbots are increasingly embedded in adolescents’ daily lives, yet developmental science research struggles to keep pace with adoption. This mixed-methods study gathered developmental expert perspectives via a survey (n = 141 researchers and adjacent professionals working on issues related to youth and technology) and semi-structured interviews with a purposive subsample (n = 15) to inform a coordinated research and policy agenda on adolescent AI use. The study aims were to (1) identify priority AI use cases for future study, (2) identify priority research questions and methods, as well as barriers to future research, and (3) identify societal guardrails that developmental professionals recommend enacting now, while research is underway. Experts cautioned against vague "screen time" measures and agreed that nearly all use cases merit study but prioritized those with severe or irreversible immediate-term harms, AI literacy, AI as a social/relational actor, AI-generated (mis)information, and overreliance. No single methodological consensus emerged; participants advocated broadly for methodological pluralism and triangulation. Key barriers included the rapid pace of change in AI technology and use relative to academic timelines, limited industry transparency, funding gaps, and misaligned academic incentives. Proposed solutions included shared infrastructure, interdisciplinary consortia, and community partnerships. Experts underscored that responsibility for safe AI should not rest primarily on adolescents and families and identified three immediate guardrails: enforceable youth-centered regulation, developmentally informed product design, and multidimensional AI literacy. Findings clarify a preliminary direction for future research and advocacy and emphasize that coordination among researchers and between sectors is needed.

To systematically review and meta-analyze correlates of condom use among Black adolescents and assess the moderating roles of gender and age. A systematic search was conducted of studies published between January 2000 and February 2024 using Medline, CINAHL, PsycINFO, and Communication Source databases. Studies were included if they sampled adolescents from the United States, included an adolescent report of condom use behavior and one of 36 correlates of interest, included a bivariate statistical association between a correlate and condom use, were in English, and at least 90% of participants self-identified as Black. Data extracted included sample size, participant demographics, research design, and condom use measurement characteristics. Correlation coefficients and 95% CI were computed, and random effects meta-analyses were used. Forty-two studies with 47 independent samples were included (16,740 adolescents; weighted mean age = 16.16). Nine correlates were significantly associated with condom use. Condom communication with a partner had the largest mean effect size (r = .42, 95% CI; .13-.65; p = .006), followed by condom use intentions (r = .32, 95% CI; .21-.43; p ≤ .001). Although 61% of effects were heterogeneous, we were only able to examine moderation by age for two correlates (i.e., gender, substance use/abuse) due to a lack of studies. Moderation analyses were not significant. This study identified the strongest correlates of condom use among Black adolescents and highlighted notable gaps in previous research. Results can be used to guide future work with this population, including the development of sexual health interventions tailored for Black adolescents. (PsycInfo Database Record (c) 2026 APA, all rights reserved).

Research on appearance‐related avoidance behaviors and their links to anxiety in early adolescence is limited. This study used three‐wave data across one academic year to examine how appearance‐related avoidance behaviors (assessed at Wave 1) predict anxiety symptoms (measured at Waves 1–3) using zero‐inflated negative binomial (ZINB) mixed‐effects growth models. Participants included 1426 adolescents ( M _age = 12.2; 45.9% girls, 44.4% boys, 9.6% another gender; racially/ethnically diverse). Appearance‐related avoidance predicted both higher baseline anxiety and changes in anxiety over time, indicating a weaker increase in anxiety across waves among adolescents reporting higher avoidance. No gender or race/ethnicity differences were found in these associations. These findings highlight the relevance of appearance‐related avoidance in adolescent anxiety.

Generative artificial intelligence (genAI) is rapidly transforming the context in which adolescents develop, presenting both opportunities and risks. Adolescence is a sensitive life period, when the accomplishment of key psychosocial developmental tasks—including social belonging, emotion regulation, autonomy, and identity formation—has lifelong consequences. This perspective synthesizes interdisciplinary research on how adolescents’ use of genAI chatbots like ChatGPT may have unique impacts during this life stage. Although genAI may provide low-stakes opportunities to simulate social interactions, scaffold emotional regulation, and support identity exploration in some contexts, it also threatens to displace or thwart essential developmental experiences. Risks include increasing social isolation, undermining distress tolerance, distorting expectations for human relationships, compromising identity and purpose development, and exposing youth to biased, inaccurate, or sexually inappropriate content. We call for a developmental science framework to guide future research, technology design, and policy. Longitudinal studies are needed—especially those that examine specific use cases, address all domains of development holistically, account for varying levels of vulnerability across stages and populations of adolescence, leverage interdisciplinary and academic–industry partnerships, and incorporate the perspectives of youth. As adolescents increasingly integrate genAI into their daily lives, we must urgently address how these tools are shaping their transition to adulthood.