Words may predict treatment success for depression sufferers

Scientists at the University of Portsmouth have published the first study to provide a promising predictor of whether the most commonly prescribed therapy – guided self-help – will help a person recover.

The results are likely to reduce drop-out rates of those referred to mental health units and speed up the decisions on which treatment to offer patients. They are also likely to help people diagnosed with depression or anxiety because, according to the researchers, dropping out of therapy or therapy not working may make the original problem worse.

Guided self-help requires people to reflect upon their feelings and to try to make sense of events and experiences that have lead to a mental health problem. It works for those with the least serious mental health problems because being asked to take a step back, reflect and write things down helps many people establish a link between cause and effect. It tends to be the first therapy prescribed by GPs for people suffering from depression or anxiety.

Dr. Joerg Zinken and colleagues in the Department of Psychology studied the language and phrasing used by people referred to mental health specialists by their GPs. Their study, published in Psychiatry Research, is the first to find a clear link between syntactic relationships in people’s writing – the details of grammar and sentence construction – and the likelihood of guided self-help working.

Guided self-help is prescribed for those with mild to moderate symptoms of depression or anxiety. It is cognitive behavioural therapy in which patients work through a workbook on their own and meet a mental health support worker once a week. It is much cheaper than one-to-one therapy with a qualified therapist.

Dr. Zinken said: “The most exciting thing about these results is we have found the way people express themselves can help predict how likely a person is to complete a program of guided self-help.

“Reducing the drop-out rates is important to improve the availability of mental health treatment as well as to avoid making people’s depression and anxiety worse through prescribing unsuitable therapy.”

Dr. Zinken explored two ways in which it might be possible to better predict the chances of guided self-help working for patients – the words patients use when describing themselves and their condition, and the syntax they use, or way they construct their sentences.

The researchers examined the statements given to mental health care workers by 100 patients about their problems they were facing. Of the 100, about a third either didn’t take up any therapy or were offered a different form of therapy; a third completed the guided self-help programme and about a third dropped out (not unusual for this form of treatment). Of those who completed the programme, 11 were still considered to be depressed or suffering from anxiety after treatment and 15 had shown signs of improvement – giving a 24 per cent chance of such therapy working based on this sample.

As an example of language use, a patient who makes a statement like this is likely to later gain no benefit from undergoing guided self-help: “Feeling tired. Worrying too much, no contact with grandchildren.”

Whereas an example of language used by a person who is likely to show improvement in their mental health after undergoing guided self-help is: “I constantly feel tired because I worry too much. I want to have contact with my grandchildren again.”

Dr. Zinken said: “Patients like those in the first example who write in a telegram style with very short isolated statements tend not to benefit from guided self-help, while those who try to relate past events to one another and to their own feelings about these events, as in the second example, are more likely to benefit.

“In terms of a therapy working, how people link their sentences together to show the relationships between the things they are talking about has never been studied before. Previous research has concentrated on trying to identify which variables (including the severity of the disorder and the motivation of the person to get better) affect the chance of any particular therapy working. But until now, scientists were in the dark as to how to identify which people were most likely to find guided self-help the most effective.

“There are a high number of drop-outs in guided self-help and anything that enables mental health professionals to establish in advance who is most likely to benefit from a particular form of therapy would save time, free up resources and enable people to be directed towards the best form of therapy for them.

“Software for assessing syntax in the way we did in the study does not currently exist, but is within the scope of what can be done; so if further research finds that these findings are reliable, it might be worth thinking about developing such a software.

“Those diagnosed with mild to moderate forms of depression or anxiety are usually referred to guided self-help quickly, but those who are prescribed more intensive treatment of talking with a therapist one-to-one can wait up to two years for treatment to start.”