Depression and anxiety are the most common psychiatric disorders in the world. Around 300 million people suffer from depression, whereas 301 million have anxiety disorder. That’s nearly 8 percent of the global population.

Unfortunately, many of these people are prescribed drugs that have no effect, when they first visit their doctor. Nearly half of all patients experience no effect of the drugs first given to them, making recovery a lengthy affair, lasting weeks or sometimes months.

But maybe there’s a possible solution to this problem.

Researchers from Germany, Sweden and Denmark have developed a way to predict which drugs will have an effect on anxiety or depression by making a simple genetic test.

By using so called polygenic risk scores (PRS) the researchers are now able to predict which patients will benefit from which drugs. All it takes is a genetic test.

That said, the technology has only been demonstrated using genetic databases for research purposes. Not on real patients, as yet.

But lead author of the study, professor Fredrik Åhs from the Department of Psychology and Social Work, Mid Sweden University in Sweden, hopes to do clinical tests in a near future.

- We believe this technology could be used to develop more targeted tests. The long term goal is a test that doctors can use to choose the right medicine, and looking at our genes is one way of doing it. 

- We’re interested in looking into biomarkers as well. Hopefully, in the future, we’ll have a cheap and effective test that enables us to alleviate people’s suffering much faster.

Two years ago professor Doug Speed from Center for Quantitative Genetics and Genomics at Aarhus University was contacted by Fredrik Åhs. 

The Swedish professor wanted to use the polygenic risk scores that Doug Speed has developed. The professor from Aarhus is constantly trying to develop better ways to crunch human genetic data in order to use it in a variety of ways. He’s especially interested in what genetically affects our mental health.

- The last 10 years, we’ve been working towards using polygenic risk scores to predict disease. It’s very challenging because many diseases are caused by thousands of variations across the genome, he says and continues:

- It turns out that these polygenic risk scores can predict our response to drugs, which is a bit surprising, but a significant step forward.

Doug Speed has developed polygenic risk scores for a number of different diseases such as schizophrenia, anxiety, bipolar disorder and depression. All of these were used in the study.

Polygenic risk scores aren’t a diagnostic tool. They can’t tell if you have anxiety, depression or schizophrenia. Instead, they indicate if you are at risk of developing these conditions.

They are a useful tool, though. To see if our genetics play a role in our response to drugs against anxiety or depression, Fredrik Åhs took these polygenic risk scores and used them on the Swedish Twin Registry.

The registry is the largest of its kind in the world. With it, researchers are able to calculate exactly how much our genes or our lifestyle influence diseases, achievements, jobs or economy.

Because the people in the registry are twins, they are genetically very similar. This means that if all twins carrying a specific variation develop a specific disease, it strongly suggests a genetic cause.
 
In the registry, Fredrik Åhs located 2515 individuals who had been prescribed drugs for anxiety and depression. By looking at which drugs they were prescribed and if they changed into other drugs during the treatment, he was able to infer if the treatment had worked.

- We then looked at the polygenic risk scores of these individuals, and it became clear that if you had a higher risk score for depression or anxiety, drugs like benzodiazepine and histamines had a smaller effect, he says and adds:

- More research is needed, but hopefully, we’ll be able to develop accurate tests in the future that can predict which kind of drugs will most likely have an effect on you.

As it is with most science the results come with limitations. This type of study is only as good as the data, and in this case the data was very good, but not perfect, Fredrik Åhs explains.

- The data on the patient’s response and nonresponse to different drugs was based on which drugs were prescribed to them, not clinical notes. We can infer a lot from the prescription-data, but we can’t be sure if there was a slight bias, he says and continues:

- In other words, we don’t know exactly why they changed drugs. Was it because of side effects, lack of remission, or something else? We did compare our results with other studies that used clinical assessment, and they were consistent with ours.

The data on the use of drugs was challenging. Fredrik Åhs and his colleagues needed to limit the study data to a specific period of time. This might have biased the results a bit as well because some of the patients might have used other drugs before the start of the time period.

- This might have influenced the number of individuals who received only one drug in our study. Some of them could have received other drugs before that weren’t registered in our data. That’s one of the reasons why we want to do a clinical follow-up study, he says.