You might not put too much thought into what words you include in your tweets, but now scientists are saying that an algorithm might be able to predict a potential mental illness based on what you say when using the social media site. It shouldn’t come as a surprise that certain things can be said for someone when they use social media, as you tend to talk about yourself quite a bit, but there hadn’t been a scientific link before.

Certain things that you say in your tweets could be a signal that you might suffer from a mental illness such as post-traumatic stress disorder or depression. Over 200 people participated in the study that had about half of the users with a diagnosed mental illness, while the other half was determined to be mentally healthy. The algorithm was then set in place that tracked particular words that both groups were using.

As it turned out, people who suffered from mental illnesses were using words such as “death” and “never” more frequently than the healthy group that generally used more positive words such as “happy.” The results of the tweets were studied in the 200 days leading up to their diagnosis, and another 200 days that followed. There was a strong link between depression and PTSD judging by the words that were used, and the algorithm was able to predict which people were diagnosed with a mental illness with strong accuracy.

The study was conducted by Andrew Reece and his team of researchers who said that “Our findings strongly support the claim that computational methods can effectively screen Twitter data for indicators of depression and PTSD. Our method identified these mental health conditions earlier and more accurately than the performance of trained health professionals, and was more precise than previous computational approaches.”

One of the researchers, Katharina Lix, said that “We hope that our research will eventually help improve mental health care, for example in preventive screening. We could imagine clinicians using this technology as a supporting tool during a patient’s initial assessment, provided that the patient has agreed to have their social media data used in this way. However, before we get to that point, the technology needs to be validated using a larger sample of people that’s representative of the general population. We want to emphasize that any real-world application of this technology must carefully take into account ethical and privacy concerns.”

The research team is hoping that these types of algorithms can be used to help people get treatment sooner rather than later. This is especially true with younger people such as teenagers that might be more vulnerable to mental illness. It’s estimated that more than 12 percent of teenagers in the United States suffer from depression, while adults came in at under seven percent.

Researchers concluded their study by saying that “This report provides an outline for an accessible, accurate, and inexpensive means of improving depression and PTSD screening, especially in contexts where in-person assessments are difficult or costly. In concert with robust data privacy and ethical analytics practices, future models based on our work may serve to augment traditional mental health care procedures. More generally, our results support the idea that computational analysis of social media can be used to identify major changes in individual psychology.”