Social media, depression and suicide

This article was co-authored by Sinead Rhodes and guest writer Tracy Stewart

A link between use of social media services, such as Facebook and Instagram, and mental health is frequently discussed in the media. Invariably, an association between the two is interpreted as time spent online causing mental health problems. The current discussion centers on a study which examined depression and social media engagement. So how did the media handle reporting of this study?

Headlines varied from those that stayed close to what the study methodology could infer: ‘study links excessive digital screen time among teens to depression, suicide’, to the more inaccurate ‘screen time might boost depression, suicide behaviours in teens’ to the completely sensationalist headline ‘social media is making teen girls suicidal’. The latter conclusion cannot be inferred from the design of the study reported in this paper.

What did the study do?

The study involved a survey carried out by researchers at San Diego and Florida State University. The researchers analysed the results of two nationally representative surveys for U.S. teenagers aged 13-18 years conducted with over 500,000 teenagers. Different children were tested at each age point rather than the study involving a follow-up of the same children.

The surveys examined information relating to depression symptoms, and suicide and suicide related outcomes; in addition the researchers looked at National Statistics for suicide. The surveys also asked about time spent using electronic games, time spent on social media, and time spent on other non-screen activities such as exercise.

The researchers conducted what is known as a correlational analysis which looks for an association between two factors, between depression and suicide, and screen and non-screen activities. The researchers reported that “Adolescents who spent more time on screen activities were significantly more likely to have high depressive symptoms or have at least one suicide ­related outcome, and those who spent more time  on non-screen activities were less likely”. The authors also reported that depression levels and suicide increased between 2010 and 2015 alongside an increase in using electronic games and social media. The authors interpret the findings as suggesting that high levels of screen time is a risk factor for depression and suicide.

The difficulty with this interpretation lies in the fact that it is underpinned by a correlational and not a causal analysis. As we mentioned in our How To Research the Headlines series, correlation is not causation.

If participants were placed at random into two groups whereby group A spent a significant number of hours on screen time and group B did not, instead spending this time doing a non-screen activity, and group A had higher levels of depression and suicide when tested afterwards (where there were no other pre-existing differences between the groups), then we could infer that screen time was a risk factor for depression.

However, with the design the current study used, it is not possible to determine what the direction of effect is. It is equally likely that having depression, a core feature of which is withdrawing from social situations, leads teenagers to spend more time online possibly looking for support for their psychological functioning.

The researchers note in the discussion section of their paper that “It is possible…..that mental health issues increased for some other unknown reason and depressed  teens  were more likely to spend time on screens”. The problem is that the design of their study does not enable a directional, cause and effect, interpretation of the link between the two factors. Depressed girls in the study may have deliberately spent more time online at a time when they feel the need to physically withdraw from their peers, and they may of course have been looking online for sources of support rather than seeing information that was making them more depressed. So the depression may pre-date the screen use…….

Depression figures may have risen not because more parents were buying their teens iPhones, but because there is more awareness amongst teens of depression symptoms and less stigma about having these symptoms (but note we have a long way to go here yet). Regardless, it is imperative that the media are careful when reporting on depression and suicide especially in vulnerable populations such as teenagers. It is also essential that researchers engage fully with the media about their methodologies so that alternative explanations of the findings are considered.

Media reporting on this study also raises further issues. One concerns the incorrect reporting of mental health prevalence data in the media. One of the articles that covered the study also suggested it followed on from another they reported on in September. Linking back to that article it was suggested that “one in four British schoolgirls suffer from depression at the age of 14”. This high prevalence claim was first reported in the Guardian who asserted that “government funded research had found one in four girls to be clinically depressed by the age of 14 and that this has sparked new fears of an epidemic of poor mental health in young people in Britain”. The Guardian article claimed the research found 24% of 14-year-old girls and 9% of boys the same age have depression. In fact the research, which formed part of the Millennium Cohort Study, reported that 24% of girls and 9% of 14 year of boys had self-reported ‘symptoms’ of depression.

While symptoms of depression can, in some cases, lead to a diagnosis of clinical depression, prevalence rates of clinically diagnosed depression in adolescence is around 4-5%, not the 24% reported here. This example of course highlights a further issue – the cyclical nature of inaccurate, sensationalised headlines!