Drawing on data from the Population Assessment of Tobacco and Health Study, a large nationally representative survey of tobacco use in the United States, the researchers tracked adolescents ages 12 to 17 from 2014 through 2024. That wider lens captured the entire arc of the youth vaping era, from the initial rise of e-cigarettes through the explosive surge of 2018 and 2019 to the shifting patterns seen today. Importantly, the researchers did not count casual experimentation.

The study focused on what is called current established use, meaning participants had to report using the product regularly and within the past 30 days. That makes the figures a conservative estimate of the problem. From 2014 through 2018, vaping followed the historical pattern, with higher rates among boys and among older teens ages 15 to 17 compared with younger adolescents ages 12 to 14.

Then came the dramatic surge of 2018 and 2019, when rates climbed sharply among both sexes and the longstanding gender gap largely disappeared. Rates fell by 2021, but what happened next depended on who was looking at the numbers. Among younger adolescents ages 12 to 14, vaping rates stayed relatively low and fluctuated from year to year, with girls generally showing higher rates than boys after 2021, although vaping among girls declined in the most recent data.

Brownstein cautioned against reading that dip as evidence that the problem has faded. Even though the rate dropped from 2023 to 2024, she noted, it remains high, and there are still many middle school girls vaping. Among older teens ages 15 to 17, the shift was far more pronounced: regular vaping among girls climbed steadily from 4.9 percent in 2021 to 7.8 percent in 2024, approaching the peak seen during the 2018-2019 surge, while 4.6 percent of boys the same age were regular users in 2024.

Against those vaping trends, the decade of data also reveals a genuine public health triumph: traditional cigarette smoking among adolescents has collapsed. By the final wave of data, fewer than half of 1 percent of both boys and girls were established smokers. Toll described the decline as a massive public health win, and Sanford suggested that the changing patterns indicate vaping may be displacing at least some cigarette smoking among young people.

E-cigarettes do expose users to fewer known toxicants and carcinogens than combustible cigarettes, which is part of why they may help adults who smoke transition away from cigarettes. That does not make them harmless, particularly for adolescents who otherwise might never use nicotine. Because e-cigarettes are relatively new, their long-term health effects remain unknown, and for teenagers, regular vaping can establish nicotine addiction while the brain is still developing.

The devices also make patterns of use harder to measure. Unlike cigarettes, which have a natural stopping point, a single cartridge can deliver hundreds or thousands of puffs, allowing users to vape repeatedly throughout the day, and the devices are easy to conceal. Sanford described seeing that challenge firsthand with his patients: when asked how much they vape, many struggle to answer clearly because they cannot track individual sessions or how long a device lasts.

What the study cannot yet say is why girls are now vaping more. Because the analysis is observational, Brownstein emphasized that the researchers can only point to associations, not conclude that any single factor caused the increased use among girls. Still, she called the patterns fascinating and said there is far more work to delve into.

Sanford agreed that understanding what drives use is one of the most important next questions, asking whether there have been shifts in motivation, marketing, social influences or perceptions. Answering those questions could ultimately help researchers develop more targeted prevention messages and cessation programs tailored to the factors influencing use among girls and among young people more generally. The findings carry implications well beyond the research community.

Toll warned that some complacency has set in since the height of the JUUL crisis, and urged pediatricians, parents, coaches and others who regularly interact with teenagers to stay alert to youth vaping, especially among female patients. Sanford said the changing trends point to where attention and resources should go, in terms of prevention efforts, treatment development and public health policy. The message from the research team is clear: the youth vaping problem has not gone away, it has simply changed shape, and the alarm bells now ring loudest for girls without falling silent for boys.

Subject of Research: Gender trends in adolescent vaping and cigarette smoking in the United States

Article Title: Girls vaping at higher rates than boys, national study finds

Article References: Girls vaping at higher rates than boys, national study finds. Original publication (Bibliographic details unavailable.)

Image Credits: AI Generated

DOI: Not provided

Keywords: youth vaping, adolescents, nicotine, e-cigarettes, JUUL, gender differences, JAMA Pediatrics, public health, MUSC Hollings Cancer Center, PATH Study, Girls, vaping

The research team, led by Evan Krueger, an assistant professor in the Tulane University School of Social Work, drew on a nationally representative survey of adolescents between the ages of 12 and 17. Rather than focusing solely on identity markers or demographic categories, the investigators examined a behavioral variable that is often overlooked in disparities research: whether a teenager has anyone they would confide in when confronted with a serious problem. This shift in focus, from who a teen is to who a teen can reach, allowed the researchers to quantify the relationship between perceived social support and mental health risk with unusual precision.

The numbers are striking. Among gay and lesbian teens who reported having no trusted person to turn to, an estimated 61 percent had experienced depression in the past year. Among those with at least one confidant, that figure dropped to 35 percent. The disparity widened further when the researchers looked at suicidal ideation: 66 percent of gay and lesbian teens without a trusted person reported having seriously considered suicide in the past year, compared with 27 percent of those who had someone to talk to. In epidemiological terms, the presence of a single supportive relationship was associated with a reduction in reported suicidal thoughts of nearly forty percentage points, a magnitude of effect that few individual-level interventions in adolescent mental health can approach.

Across the full sample of sexual minority youth, the pattern held consistently. Having someone to talk to was linked to lower rates of depression, lower rates of suicidal thoughts and lower rates of having made a suicide plan in the past year. Krueger noted that while this protective association appeared across the board, for many LGBQ+ youth it made an even bigger difference. That observation matters because sexual minority adolescents already carry a disproportionate burden of mental health risk, driven in large part by stressors that are social rather than biological in origin, including stigma, discrimination and bullying layered on top of the ordinary challenges of adolescence.

The study also illuminated a second, equally important finding about where teenagers seek help. When asked whom they would approach with a serious problem, roughly eight in ten heterosexual teens said they would turn to a parent. Among gay, lesbian and bisexual teens, that figure fell to about six in ten. On average, LGBQ+ youth reported feeling less able to turn to their mothers or fathers, a gap that likely reflects a combination of fear of rejection, uncertainty about how their families will react, and in some cases experiences of outright non-affirmation at home. The result suggests that for a substantial share of sexual minority adolescents, the most conventional source of support in a crisis is not reliably available.

What the researchers found instead was a picture of resourcefulness. LGBQ+ teens turned more often to friends, romantic partners and other adults in their lives, and they were more likely than their heterosexual peers to rely on several different people rather than a single confidant. Krueger emphasized that this does not mean these young people do not need their parents; rather, it shows that when support is not found at home, many LGBQ+ youth actively seek out connection and support elsewhere. From a developmental science perspective, this diversification of support networks can be adaptive, but it also raises questions about consistency and depth, since friends may not fully replace a stable, supportive adult in a young person’s life.

The findings dovetail with a growing body of evidence on the role of schools in adolescent wellbeing. Previous research has found that the presence of supportive school personnel, including teachers, counselors and other staff, is associated with lower suicide risk among LGBQ+ youth. Krueger and his colleagues suggest that these adults may play pivotal roles in filling the support gap, fostering psychological safety and a sense of belonging for young people who lack affirmation at home. In practical terms, a counselor who keeps an open door or a teacher who signals acceptance may function as the trusted confidant whose presence the new study links to sharply reduced risk.

The research may also help explain why broader population-level indicators of teen mental health have recently begun to improve. New data from the Centers for Disease Control and Prevention have documented gains in adolescent mental health, and experts have credited part of that improvement to school-based efforts that connect students with support. If the mechanism behind those efforts is, at least in part, the creation of trusted relationships, then the Tulane study provides a plausible causal scaffold for why such programs work: they increase the number of young people who have someone to turn to, and that single change is associated with large reductions in depression and suicidality.

The study’s co-author, Isaac Wright, a recent graduate of the School of Social Work, joined Krueger in pointing toward the implications of the work. The researchers are careful to note that friends, however vital, may not fully substitute for a consistent, supportive adult, and that schools, families and communities all share responsibility for ensuring that no teenager faces a serious problem alone. The message for intervention design is correspondingly broad: mentorship programs, inclusive school climates, family acceptance initiatives and community organizations all contribute to the same underlying variable that the study identifies as protective.

Ultimately, the research reframes a longstanding disparity in a way that invites action rather than resignation. The entrenched gaps in depression and suicide risk between sexual minority and heterosexual adolescents are well documented, but this study suggests that strengthening supportive relationships is one concrete way to begin narrowing them. As Krueger put it, if researchers and practitioners can find ways to facilitate meaningful, supportive relationships, that is one way to start reducing some of the longstanding, entrenched disparities in mental health outcomes. Supportive relationships benefit all children, but for LGBQ+ youth who are less likely to find support at home, cultivating them may be the difference between struggling in isolation and having a lifeline when it matters most.

Subject of Research: Social support and mental health outcomes among LGBQ+ adolescents in the United States

Article Title: Having someone to talk to linked to better mental health among LGBQ+ teens

Article References: Having someone to talk to linked to better mental health among LGBQ+ teens. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: LGBQ+ youth, adolescent mental health, depression, suicidal ideation, social support, JAMA Pediatrics, Tulane University, help-seeking behavior, school support, sexual minority health, peer relationships, public health

The study, led by corresponding author Tracy Vaillancourt of the Faculty of Education at the University of Ottawa, examined whether what the researchers describe as affective generative artificial intelligence use, meaning the use of AI tools for emotional rather than practical purposes, was associated with poorer mental health among youth. The central question was whether this behavior merely overlaps with known correlates of distress, such as loneliness or a diminished sense of mattering to others, or whether it represents something separate that clinicians and educators should pay attention to in its own right. The findings point to the latter: emotional reliance on AI emerged as a unique marker of psychological distress, independent of both mattering and loneliness.

This distinction matters because loneliness and a lack of mattering have long been recognized as powerful predictors of poor mental health outcomes in young people. Loneliness captures the painful gap between the social connection a person wants and the connection they actually experience, while mattering reflects the sense of being significant to others, of being noticed and valued by the people around them. Both constructs are closely tied to depression, anxiety, and suicidal ideation in adolescents. If turning to AI for emotional support were simply a symptom of loneliness or of feeling invisible, it would arguably add little new information. The fact that the association with distress persists even after accounting for these factors suggests that affective AI use is tapping into something additional about a young person’s inner life.

The researchers frame their conclusions around a practical and increasingly urgent distinction: the difference between functional AI assistance and the use of algorithmic interfaces as a digital refuge for emotional needs. Functional assistance covers the many legitimate and often beneficial ways young people interact with generative AI, such as getting help with homework, brainstorming ideas, answering factual questions, or learning new skills. Affective use, by contrast, involves bringing emotional burdens to a machine, asking a chatbot for comfort, confiding in it about painful feelings, or treating it as a substitute source of understanding and support. The study suggests that clinical frameworks and digital literacy programs should draw a clear line between these two patterns of behavior, because they may carry very different implications for a child’s wellbeing.

The timing of this research is significant. Generative AI tools have moved from research laboratories into everyday life at remarkable speed, and children and adolescents are among their most enthusiastic adopters. Conversational agents are now available around the clock, respond instantly, never judge, and can simulate empathy with convincing fluency. For a young person who feels isolated, misunderstood, or burdened, the appeal of such an interlocutor is easy to understand. Yet the same qualities that make these systems attractive also raise questions about what it means when a developing mind begins to route its emotional needs through an algorithm rather than through family, friends, teachers, or mental health professionals.

From a clinical standpoint, the study’s findings carry implications for how practitioners assess and respond to psychological distress in youth. If asking an AI for emotional support is a unique marker of distress, then simply asking a young patient whether they use AI tools, and how they use them, could provide clinicians with diagnostically meaningful information that might otherwise go uncollected. A teenager who reports using chatbots primarily for schoolwork presents a very different picture from one who reports turning to them when feeling sad, anxious, or alone. The authors argue that clinical frameworks should be updated to make this distinction explicit, so that affective AI use is recognized and explored rather than overlooked or lumped together with benign technology use.

The implications extend beyond the clinic into schools and public health more broadly. Digital literacy programs have traditionally focused on topics such as online safety, privacy, misinformation, and screen time. The present findings suggest that these programs should also help young people understand the difference between using AI as a tool and using it as an emotional refuge. Teaching children and adolescents to recognize when they are seeking comfort from a machine, and to reflect on what that impulse might be telling them about their own emotional state, could turn a passive risk factor into an opportunity for self-awareness. Equally, such programs could help adults, including parents and teachers, understand that a child’s growing attachment to conversational AI may be less about fascination with technology and more about unmet emotional needs.

The study is published alongside an editorial and an editor’s note in the JAMA Network, reflecting the significance that the journal’s editors attach to the topic. Cross-sectional research of this kind captures a snapshot in time, which means it can identify associations but cannot establish whether emotional distress drives young people toward AI, whether heavy affective AI use contributes to distress, or whether both patterns are shaped by other underlying factors. The authors are careful to describe the relationship as a marker rather than a cause, a distinction that matters for interpretation. What the design does establish is that the association is real and independent: even when loneliness and mattering are held constant, young people who seek emotional support from generative AI show higher levels of psychological distress.

That independence is the study’s most striking contribution. It suggests that affective AI use is not reducible to the social deficits that mental health professionals already screen for. A young person might report feeling connected to others and valued by them, yet still turn to an algorithm for emotional support, and that behavior alone would flag elevated distress. This makes the behavior potentially valuable as an early signal, something visible on the surface of everyday technology use that hints at struggles happening beneath it. In an era when parents and educators often struggle to detect internalizing problems in children, precisely because conditions such as depression and anxiety tend to be hidden rather than disruptive, a marker embedded in routine digital behavior could prove genuinely useful.

The broader conversation about AI and youth mental health is only beginning, and this study adds an important empirical data point to a debate that has so far been dominated by speculation. As generative AI becomes more emotionally sophisticated, more personalized, and more deeply woven into the daily routines of young people, the questions raised by this research will only grow in urgency. The authors’ central message is measured rather than alarmist: the goal is not to condemn AI use among youth, much of which is functional and harmless, but to recognize that when children and adolescents begin treating algorithmic interfaces as a refuge for their emotional lives, adults should pay attention. Distinguishing between a young person using AI to finish an assignment and one using it to fill an emotional void may become one of the more consequential skills for the clinicians, educators, and families raising children in the age of artificial intelligence.

Subject of Research: The association between affective use of generative artificial intelligence and psychological distress in children and adolescents

Article Title: Affective generative artificial intelligence use and youth mental health

Article References: Affective generative artificial intelligence use and youth mental health. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: generative AI, youth mental health, psychological distress, loneliness, mattering, JAMA Pediatrics, adolescents, digital literacy, chatbots, emotional support, cross-sectional study, clinical frameworks