The scale of the gender gap in adolescent mental health has grown alarming in recent years. Across wealthy countries, girls consistently report higher rates of depression, anxiety and self-harm than boys, and surveys show that internalizing problems—conditions in which distress is turned inward rather than expressed through outward behavior—have risen steeply among girls since the early 2000s. A recent correspondence in Nature Human Behaviour by Zui Narita and colleagues described the situation as a silent crisis, highlighting the worsening trajectory of girls’ mental health worldwide. Parallel work by Margreet de Looze and colleagues, published in the Journal of Child Psychology and Psychiatry, tracked psychological symptoms across 43 countries between 2002 and 2022 and found that the gender gap has been widening rather than closing, even as national-level gender equality has improved in many of those nations.

What makes the new study distinctive is its methodological rigor in moving beyond correlation. Rather than simply showing that girls who feel lonely or unhappy with their bodies also report more depression, the researchers employed a mediational g-formula, a sophisticated statistical framework developed by Shintaro Yamamuro and colleagues for estimating causal pathways when treatments and mediators vary over time and are measured repeatedly. The technique, originally applied to estimate how statin drugs affect cardiovascular disease through cholesterol lowering and anti-inflammatory action, allows researchers to model how an exposure—in this case, sex—acts on an outcome both directly and indirectly through intermediate variables. Mediation analysis of this kind is notoriously easy to get wrong; Narita and colleagues recently outlined the key pitfalls for psychiatric research in JAMA Psychiatry, and the new study appears designed with those cautions in mind.

The two mediators that emerged from the analysis tell a coherent story. Loneliness, the first pathway, has become one of the defining features of contemporary adolescence, and the evidence suggests it operates differently for girls than for boys. Adolescent girls who reported feeling isolated, disconnected from peers or lacking close confidants were more likely to develop depressive symptoms and suicidal thoughts, and loneliness was more strongly linked to girls’ internalizing problems than to boys’. This aligns with a broader literature suggesting that girls’ social worlds, while often more intimate and relationally oriented, may also expose them to greater relational stress, social comparison and exclusion, all of which feed vulnerability when support networks fail.

The second pathway, the desire for thinness, points to the persistent power of body-image ideals in shaping girls’ mental health. Despite decades of public discussion about body positivity and the documented harms of unrealistic beauty standards, the aspiration to be thin remains a potent driver of psychological distress in adolescent girls. The study’s findings indicate that this desire functions not merely as a symptom of poor mental health but as a causal route through which broader cultural and social pressures are converted into depression and suicidal ideation. Boys, by contrast, are far less exposed to thinness-oriented appearance pressure in this form, which helps explain why the internalizing gap between the sexes is so pronounced during adolescence.

The study’s design strengthened its conclusions in several important ways. By incorporating repeated measures across adolescence, the mediational g-formula could account for time-varying confounding—the reality that loneliness, body dissatisfaction and depression each influence one another over time in feedback loops that simpler statistical models mishandle. By grounding the choice of mediators in adolescents’ own accounts, the researchers avoided the common trap of testing pathways that are statistically convenient but psychologically irrelevant to young people. And by focusing on two of the most robust outcomes in psychiatric epidemiology—depressive symptoms and suicidal ideation—the study tied its explanatory pathways to outcomes of clear clinical and public health significance rather than to diffuse measures of general distress.

The implications of the findings extend well beyond academic debates about mediation. If loneliness and the desire for thinness genuinely explain a meaningful share of the sex difference in adolescent internalizing problems, then interventions targeting those two mechanisms could narrow the gap. School-based programs that build peer connection and belonging, for example, might be particularly protective for girls at risk of isolation. Efforts to counter appearance-based social comparison—including scrutiny of social media environments where filtered images and diet culture circulate—could similarly reduce one of the identified pathways. Clinicians and educators may also need to screen for loneliness and body-image concerns specifically in girls, rather than waiting for depressive symptoms to become severe before intervening.

At the same time, the researchers and commentators caution that mediation is not destiny. Explaining a disparity through two pathways does not mean those pathways are the only ones that matter, nor that modifying them will eliminate the gap. Structural factors—economic insecurity, exposure to violence, academic pressure, the digitization of social life—likely interact with loneliness and appearance concerns in ways that the current analysis cannot fully disentangle. The 43-country evidence from de Looze and colleagues suggests that even macro-level conditions such as gender equality do not straightforwardly reduce the gap, hinting that the mechanisms at work are subtle and culturally embedded. The mediational g-formula itself, while powerful, rests on assumptions about causality that can never be verified with certainty in observational data.

What the study does provide is a testable, youth-informed agenda for action. The silent crisis described by Narita and colleagues a year ago now has identifiable entry points: the social disconnection that isolates girls and the appearance ideals that devalue them. Because both pathways are potentially modifiable—through school climate, clinical screening, family communication and policy on the media environments adolescents inhabit—the findings convert a seemingly intractable sex difference into a set of concrete prevention targets. For the more than 3,000 adolescents whose data illuminated these mechanisms, and for the millions of young people they represent, that translation from description to explanation to intervention is precisely what the field has needed.

The work is also a reminder of the value of listening to adolescents themselves. The pathways investigated here were not imposed from the top down by adult assumptions about what troubles young people; they emerged from young people’s own accounts of their distress. As researchers continue to track the widening gender gap in adolescent mental health, studies that combine participatory design with rigorous causal methods of this kind offer the most promising route to interventions that actually reach the young people who need them most, before crisis becomes the first point of contact with mental health care.

Subject of Research: Mechanisms explaining sex differences in adolescent mental health, focusing on loneliness and desire for thinness as pathways to girls' higher rates of depressive symptoms and suicidal ideation.

Article Title: Why girls have worse mental health than boys

Article References: Why girls have worse mental health than boys. (2026). Nature Human Behaviour. https://doi.org/10.1038/s41562-026-02577-3

Image Credits: AI Generated

DOI: 10.1038/s41562-026-02577-3

Keywords: adolescent mental health, gender differences, depression, suicidal ideation, loneliness, desire for thinness, body image, mediation analysis, internalizing problems, Nature Human Behaviour, girls, worse