Labels are a necessary part of the diagnosis and treatment of mental ill-health – after all, it's hard to treat a problem when you can't name it. They are the central foundational pillar of international mental health practice: the current version of the Diagnostic and Statistical Manual of Mental Disorders DSM-5 lists over 300 of them.

But there's increasing evidence to suggest that labels don't always help. One of the first experiments to separate the effect of labels from the behaviours that led them to be applied took place ten years ago. Therapists were divided into three groups and then watched a video of a woman describing symptoms of panic disorder. One group had only basic background information; another group was given a description of her behaviour that could reasonably have led to borderline personality disorder being diagnosed; and members of a final group were incorrectly told that the woman had a formal BPD diagnosis.

Those who'd been given the label were more likely than members of the other two groups to evaluate the woman's problems and her prognosis negatively. In other words, it seems to be the label and not the symptoms that caused their pessimism.

It's not just therapists whose behaviour and judgements are affected by these labels, which may exacerbate symptoms - or even cause new ones - in those to whom they're applied. Many clusters of symptoms are normal responses to difficult experiences like bullying or family breakdown, but pathologising these responses through labels of mental ill-health can undermine resilience and erode self-esteem. A label may lead sufferers to believe that the cause of their symptoms is innate and fixed, leaving them lacking a sense of agency. They can become vigilant to further symptoms. Labels of mental illness are also linked to lower use of positive coping strategies and greater catastrophising.

Labels affect the way that those who hold them are treated by other people, too. While labels may be helpful for some, and get them the support they need, others experience more stigma. The shame of being labelled whether by themselves or other people can worsen symptoms.

Adolescence may bring particular vulnerability to the negative effects that can arise from labelling mental health symptoms. It's a time when a quest for identity is constant. We don't yet fully understand what happens when a mental ill-health label hooks itself into a burgeoning sense of identity that becomes increasingly fixed as teenagers move towards early adulthood.

Researchers in one recent study attempted to measure the impact of self-labels by tracking adolescents, all of whom had been assessed as having mental health problems, over two years. The study looked at whether adolescents agreed with the statement 'I have a mental illness'. Those who self-labelled as having a mental illness had lower self-esteem than those who didn't label themselves this way. Some adolescents later dropped the mental illness label - after which their self-esteem rose once more.

The philosopher Ian Hacking came up with a theory of so-called 'looping' effects. Through these, having a label makes people take on its features - even those that they may not previously have inhabited before the label was assigned. They also bring their own unique features to the label, expanding its definition and leaving a wider range of symptoms for those who are diagnosed in future to inhabit.

What do looping effects mean for adolescents and others who take on labels of mental ill-health? Scholars diverge on this point. On the one hand, the gradual bloating of the DSM has meant that the meaning of mental disorder 'has not so much looped as spread in an ever-expanding vortex'. On the other, researchers have suggested that adolescents and young adults have 'looped' the meaning of mental ill-health labels from psychiatric categories into cultural ones.

Where these analyses contradict each other is in their implications: the second analysis sees the transformation from psychiatry into culture as a positive step that removes the weight of diagnostic labels and makes them descriptors of normal human experience.

This more positive slant misses the numerous studies, some of which I've referred to above, showing the negative effects of labels and their potential to worsen symptoms in teenage populations as well as adult ones . It also misses some of the broader impacts - if a wider range of normal experience is being caught within diagnostic labels, it dilutes the support available to those who really need it. Mental health services are stretched. Some teenagers have severe mental health conditions. The tension between these two facts is potentially made worse by labelling those whose reactions are a rational response to life's normal challenges.

What does this mean for therapy? It means, for some, being more cautious about the application of labels to developing minds. It means being careful about using symptom checklists that may be interpreted by some teenagers as representing normal life and others as representing severe problems. It means keeping the possibility in mind that mental health awareness-raising efforts may, in some cases, backfire by causing teenagers to label themselves and to experience worsening symptoms.

It also means ensuring that normal parts of the human condition are not stamped with a label that may be difficult to remove. It's the laws of physics that prevent blackberry jam morphing into rhubarb, not the label stamped on the front of the jar. The same can't be said of adolescents when it comes to mental health and ill-health: teenagers are not jam.