In many countries, a diagnosis of schizophrenia is close to a guarantee of ending up not in employment, education or training (NEET). In Denmark, that is not the case.
That is what figures from more than 65,000 people in Denmark living with schizophrenia show. We tracked their medical records for up to ten years after they received their diagnosis. Overall, the people in the study were in work, education or a vocational programme for 48.2 per cent of the weeks in which we observed them.
In other countries, typically 10 to 20 per cent of people with schizophrenia are in work, in education or on a placement.
Countries cannot be compared directly, however, because the studies use different diagnostic criteria, follow-up periods and definitions of work and education. So, we are not crowning Denmark 'world champion' at getting people with schizophrenia into jobs. But the figures show the country is doing something right.
Why are the Danish figures so high?
Part of the answer lies in the Danish welfare system. Universal healthcare, the Danish Fleksjobordningen (flexi-job scheme with subsidised posts with adapted hours and duties) and wage subsidies for employers create a safety net around the patient that is far less fragmented than in other countries.
We know from other research that people with schizophrenia experience more discrimination than people with other mental health conditions.
That makes the Danish figures even more remarkable, because they suggest that a strong support system can counteract some of the effects of stigma. This could be because a society which keeps people in the workplace also keeps seeing them as workers. That strengthens their identity and shrinks the social distance that otherwise feeds stigma.
Yet the gap between Denmark and other countries cannot be reduced to a single explanation. Nor do we claim that the Danish system is better in every respect.
Let us dig a little deeper into our findings and look at labour market attachment in periods with and without medication.
The first two years after diagnosis
Our analysis included 50,440 people who had both weeks with and weeks without recorded antipsychotic treatment. In the first two years after diagnosis, attachment to work or education was, in relative terms, around nine per cent lower in weeks with antipsychotic medication than in the same individuals' weeks without it.
In other words: a hypothetical patient, let us call her 'Ida’, has a better chance of being in work or education in the weeks when she is not taking medication during the first two years after her diagnosis.
That does not necessarily mean the medication makes Ida less able to hold down a job.
Our study is observational, not a randomised controlled trial.
Even though each person is compared with themselves, medication may be prescribed precisely in the periods when the illness is more active, and other circumstances that change over time can also influence the result.
In the hardest periods of illness, being in work or education is already more difficult. The figures therefore show an association, not proof that the medication is the cause.
We have tried to take this into account in our analyses, but in this type of study it can never be ruled out entirely.
A vulnerable window
The lower chance of being in work or education shrinks in years two to five, from around nine to around five per cent. After five years, the association reverses completely:
From that point, weeks with medication are associated with a roughly two per cent higher chance of work or education for Ida and the others in our study.
The results for years two to five surprised us.
One possible clinical interpretation is that the acute symptoms have often eased by this stage, so one might expect that stability to be converted into a job or a place in education.
The medical records did not measure symptoms directly, however, and weeks with medication remain associated with a lower rate of work or education.
That is what we interpret as the vulnerable window: the period in which clinical stabilisation may have begun, but working life has not yet caught up.
One possible explanation, which our study cannot test, is that side effects such as fatigue and reduced concentration can make the road back to work or education harder.
This is where services can do better.
Long spells out of work can take hold
Our study is not an argument for abandoning antipsychotic medication.
After five years, weeks with medication are associated with a slightly higher rate of work or education, and the treatment also protects against relapse.
But our results suggest that medication and vocational and social support should go hand in hand from the start if more people are to end up in education or work.
Vocational and social support should therefore be planned in alongside antipsychotic treatment.
Perhaps that is precisely what Denmark, with its flexi-jobs, education support and close coupling of treatment and employment services, already does better than most. But there is still room for improvement.
And it matters that as many people with schizophrenia as possible are in work, education or on placements, above all because schizophrenia typically first appears in early adulthood.
Other research (in Danish) shows that long-term unemployment in the early adult years can harden into permanent unemployment. That applies with particular force in psychosis.
A Finnish study that followed people with schizophrenia into midlife found that most never gained a lasting foothold in the labour market, and that the seeds were often sown in the early years of the illness.