Anxiety and depression, suicidal thoughts, cyberbullying, poor self-image, alcohol, cannabis and other psychoactive substance use… social networks exploit the vulnerabilities of young people and thus contribute to amplifying certain disorders of which they are victims.
This is the conclusion of a major report by the French Agency for Food, Environmental and Occupational Health & Safety (ANSES), which analyzes the mechanisms of action of these digital tools configured to target, for commercial purposes, the specific characteristics but also the vulnerabilities linked to adolescence.
Olivia Roth-Delgado and Thomas Bayeux are part of the ANSES team that coordinated this work. They present the main findings of this report.
The Conversation: The ANSES report on the “effects of digital social media use on the health of adolescents” is unprecedented. Why?
Olivia Roth-Delgado: This expertise represents five years of work and the analysis of over 1,000 articles. It is unprecedented in its originality and scope, which, to our knowledge, are unparalleled for public bodies like ANSES.
For the first time, certain social media design mechanisms have been linked to effects on adolescent health. These mechanisms are known as dark patterns – which we have translated as “deceptive interfaces”.
Adolescence is a vulnerable period because the brain is still maturing. Teenagers experience changes in how they feel and manage emotions, and in the brain’s reward system. They are also more sensitive to social situations, which can lead to risky behaviors when they are around their peers. It is also a time of heightened vulnerability to mental health problems .
Thomas Bayeux: During adolescence, a culture develops that encourages interaction with others, an appetite for communication, a construction of self that leads to testing norms, etc. All these arguments led us to focus on the 11-17 age group, during which all these changes take place.
The role of ANSES, as a public health agency, is to assess health risks. However, in the chapters of the report dedicated to the use and maintenance of intergenerational relationships, the expert opinion discusses the potential beneficial effects of social networks, through the motivations that lead to their use during adolescence.
The report highlights particularly worrying effects linked to social media use during adolescence, such as anxiety and depression, suicidal thoughts, and self-harm. What are the underlying mechanisms?
O. R.-D.: Among the mechanisms we have highlighted and studied are these deceptive (or manipulative) interfaces and content personalization algorithms. All of these correspond to attention-grabbing mechanisms that keep users on the social network by presenting them with increasingly targeted, even extreme, content.
If a teenage girl has, for example, searched for content on self-harm for the first time, this type of content will be presented to her repeatedly and can trap her in a spiral of difficulties.
T. B.: Capturing attention serves the economic model on which these digital platforms are based. It allows them to have a large amount of data and to monetize it, and also contributes to the sale of advertising space.
Digital platforms therefore have an interest in keeping people on the social network through the two strategies we have mentioned: on the one hand, the personalization of content through increasingly efficient algorithms that lock those who use them into an information loop and, on the other hand, through the highlighting of the most impactful content.
Deceptive interfaces ( dark patterns ) employ techniques that we all know: likes , notifications, infinite scrolling , videos that play automatically, etc.
The adolescent period resonates strongly with these strategies implemented by social networks. At ANSES, we see major public health challenges here, because supply and demand are, in a way, meeting. And the combination can be explosive!
Regarding these mental health disorders, but also harassment, alcohol, tobacco, cannabis use and other risky behaviors against which you warn, does this mean that social networks amplify pre-existing phenomena?
O. R.-D.: Absolutely. Social networks are a social space. They act as a sounding board for problems present in society, such as gender stereotypes, incitement to drug use, etc.
TB: Social networks contribute to socialization and adolescent development; they represent a continuation of the offline world, with its positive aspects and drawbacks. There is no clear barrier between what happens offline and on social networks.
Should the rules that exist to protect minors in society also apply to social media?
O. R.-D.: This is the founding principle of the Digital Services Act , the European regulatory framework that aims to regulate the online content of very large platforms, according to the formula: “What is illegal offline is illegal online.”
T. B.: This concern motivates one of the key recommendations of the ANSES report, namely that those under 18 should only have access to social networks designed and configured to protect minors. Our aim is not to ban them. But technical solutions must be implemented to ensure that social networks are safe spaces for teenagers, and ANSES insists on holding platforms accountable in this regard.
Furthermore, for a teenager, discussing their digital practices with others—peers, parents, teachers, educators—can be beneficial. This does not absolve public authorities and digital platforms of the need to adopt collective strategies to ensure that social networks are safe spaces for teenagers.
The report mentions links between social media use and certain disorders, but it does not establish a direct causal relationship. Why?
O. R.-D.: The question of causality remains complex. We rely on extensive and well-documented expertise. Our methodology is sound, but it is not based on a weight-of-evidence approach. Nevertheless, we demonstrate robust associations between social media use and the disorders we have discussed, and we uncover the underlying mechanisms in a highly relevant manner.
In the case of sleep, for example, several mechanisms are at play. When teenagers check social media in the evening before bed, the blue light from screens delays sleep onset because it stimulates what is called cognitive arousal, thus shortening their sleep duration. The long-term effects of chronic sleep deprivation on physical and mental health are well-documented. Furthermore, when using social media, emotional stimuli can also disrupt sleep. As we can see, there is a body of evidence. But the specific effects of social media on a teenager’s sleep also depend on how they use it.
Similarly, the nature of the content presented plays a major role in the development of anxiety and depression or suicidal thoughts. A two-way effect must also be considered. Let me explain: a teenager who is already psychologically vulnerable will tend to spend more time on social media. The content presentation algorithm will detect this emotional fragility and present them with emotionally charged content. And that’s where they will find themselves trapped in a harmful spiral. Establishing causality is more difficult in the presence of feedback loops and two-way effects.
Finally, regarding the alteration of self-image, we also have a compelling body of evidence based on the same type of mechanisms, using repeated exposure to content that glorifies thin women and muscular men.
Girls appear to be more sensitive than boys to the negative effects of social media. How can this be explained?
T. B.: That’s one of the major findings of this study. Girls clearly represent a high-risk group on social media regarding all health effects, not just the impact on self-image. They are more present on social media than boys, are more often harassed, victims of gender stereotypes, and subjected to social pressures… Girls also pay more attention to what happens on social media, the comments posted there, etc.
LGBTQIA+ communities also represent a group at risk on social media. They are particularly more likely to be victims of cyberbullying, which carries risks, especially in terms of mental health.
The ANSES report points out that time spent on social media is not the only criterion to be taken into account.
T. B.: Usage time is a useful metric, but not enough to fully address the issue. Time allows us to study certain health issues, such as sedentary lifestyles, even though mobile digital devices are proliferating for accessing social networks. Quantifying usage duration also proves valuable when discussing late-night exposure that can impact sleep, for example.
However, we also know that understanding how people use social media is essential for studying certain health effects. It’s important to know what people do on the social network: post, like , read comments, edit photos, for example, and what emotional engagement is associated with these practices. These approaches shouldn’t be seen as opposing forces, but rather as complementary.
Your report relies on research that has not, or only minimally, studied the impact of the latest digital tools, such as TikTok or “AI companions.” Can we assume that these new technologies increase the risks to the mental health of adolescent girls and boys?
O. R.-D.: ANSES’s expertise was based on more than a thousand articles published primarily between 2011 and 2021. Due to the combined timeframe of research and expert review, we were led to study technologies that have indeed evolved. Nevertheless, we relied on a common set of mechanisms, such as deceptive interfaces ( dark patterns ) and content personalization algorithms, to which we linked health effects.
Therefore, our conclusions and recommendations can be extended to more recent social networks. Regarding artificial intelligence and “AI companions,” ANSES recommends that future expert assessments examine the issue.
In your recommendations, you propose to involve adolescent girls and boys in programs developed to prevent risks.
O. R.-D.: ANSES proposes involving young people themselves in the work, because they are the ones who know their motivations for using social media, and they are the ones who develop and disseminate new practices. It therefore seems important to us to include them in the discussions and the development of guidelines, along with educators and parents. They will be more inclined to respect rules that they have helped to formulate. Among its recommendations, ANSES also mentions promoting spaces for dialogue among young people to share their experiences of what happens to them online.
T. B.: And once again, we remind you that Anses does not recommend banning social networks, but rather thoroughly reviewing their design so that they do not harm the health of adolescent boys and girls.
Author Bios: Olivia Roth-Delgado is Head of scientific projects and Thomas Bayeux is Socio-economic project manager both at French Agency for Food, Environmental and Occupational Health & Safety (ANSES)
