In May, US Surgeon General Vivek Murthy issued a health advisory warning that excessive social media use could lead to harmful effects on children’s mental health. Assistant Professor of Psychiatry and Human Behavior Jacqueline Nesi, PhD, agrees, though she says the full picture is more complicated.

Although social media can certainly play a role in mental health, it’s important to remember that mental illness is almost never the result of a single factor. At this point, the research suggests that the role of social media in youth mental health depends on how teens are actually using it. If it is being used for social connection—keeping up with friends, messaging, even meeting new friends—that can be protective for mental health. When it’s passive scrolling, posting and seeking validation in the form of likes or views, or viewing harmful content, that can be more complicated.

Another consideration is the individual differences between teens themselves. Those who are having trouble socially, taking risks, or getting into trouble offline are more vulnerable to negative impacts online. Teens who are more resilient are more likely to take advantage of social media’s benefits.

That said, we do have some evidence that social media can play a negative role in mental health for some teens. One category of risk is time spent. For some teens, excessive social media use can get in the way of activities that are important for mental health, like sleep, in-person social interaction, or physical activity. The other risk is exposure to problematic content—anything from cyberbullying to hate speech to content promoting dangerous behaviors like self-injury or disordered eating.

It’s important to recognize there can be benefits of social media as well. It can promote social connection, particularly for youth who are marginalized in their offline lives. For example, for LGBTQ+ adolescents, social media can provide a place for identity affirmation and connection.

The majority of studies on social media and youth mental health are correlational; they can’t necessarily indicate causation. But just because the research isn’t settled doesn’t mean we shouldn’t take action. I would say—and the surgeon general’s advisory makes this point—that if we want to take a safety-first approach, we can still be thinking about how to protect youth from the known risks of social media.