The cause of the excess mortality in people who have been diagnosed with schizophrenia has been much debated. Is it due essentially to the disease itself, or other factors, including the drugs used in its treatment?

Now, long-term data from Swedish population registries shows that the higher mortality is not due to antipsychotics. In fact, the highest overall mortality – compared with controls from the general population matched for age and gender – was seen in patients with no exposure to antipsychotics. This suggests that mortality risk in schizophrenia is due to factors other than medication, Dr Mitterndorfer-Rutz and colleagues conclude.

The fact that everyone in Sweden has an identification number means it is possible to link an individual’s history of diagnosis, receipt of prescribed drugs and social insurance status with time and cause of death – and to control for many possibly confounding socio-demographic variables. These registries have been mined for unique insights into schizophrenia-associated excess mortality.

The five year death rate among first episode psychosis patients was 7.5%

Data were gathered prospectively over the period 2006-2013 and included approximately 29,000 people with a diagnosis of first-episode psychosis. Over five years of follow-up, 7.5% of the cohort died. The overall mortality rate was 4.8 times that of age and sex-matched members of the general population. A third of the deaths were due to cardiovascular disease (CVD), 17% to cancers, and 11% to respiratory disease.

U-shaped curve relates exposure to mortality

Crucially for the debate on causation, those people with first-episode psychosis who received no antipsychotic medication had a mortality risk ten times that of the general population. Investigators divided those who had received antipsychotics into three groups, depending on their cumulative exposure to drug treatment. The reduction in excess mortality for people with low and moderate exposure to antipsychotics was similar. The hazard ratio for people with high cumulative antipsychotic exposure and approached that of people with no exposure at all.

People with first episode psychosis who had no antipsychotics had a mortality rate 10 times that of the general population. Lack of antipsychotics may be the most important treatment-related factor that we can address.

Ellenor Mittendorfer-Rutz and colleagues were also interested in the relationship of risk to the class of antipsychotic medication that people had been taking. Excess mortality was lowest in those with schizophrenia – relative to the general population – among patients who had been prescribed second-generation long-acting injectable agents.

A puzzling paradox

Robin Murray (Kings College, London, UK) accepted the importance of the Swedish findings but pointed to an intriguing paradox. On the one hand, use of antipsychotics at adequate doses is associated with reduced mortality excess. On the other hand, we have convincing data that they contribute to obesity, metabolic syndrome and diabetes – all of which are major cardiovascular risk factors; and CVD is the greatest contributor to shortened life expectancy among people with schizophrenia.

Life expectancy with schizophrenia is 15-20 years less than in the non-affected population, according to data from Nordic countries.

Data from a UK study show that at the time of diagnosis, 17% of people with first-episode psychosis were obese (i.e., they had a BMI of 30 or more). This was the same proportion as in the general population, suggesting that people with first-episode psychosis have no baseline predisposition to overweight. But, after one year of treatment, 29% of the first-episode psychosis patients were obese. And in a group of patients who had had a diagnosis of schizophrenia for 17 years, and who had been treated for much of that period, the obesity rate was 50%.

IMPaCT had little impact

In an attempt to address unhealthy eating and other adverse lifestyle factors, a large UK intervention study (IMPaCT) was undertaken. Despite 2.3m euros and eighteen months of effort, there was no measurable effect. This suggests to Professor Murray that prevention is better than attempts to reverse weight gain, and that weight-sparing antipsychotics may be preferable. He also advocates for the use of the lowest dose for the shortest time compatible to achieving good mental health.

In the search for better physical health, no-one must be left behind

Jeanette Westman, also from the Karolinska Institute in Stockholm, Sweden, described the reduction in life expectancy among psychiatric patients as catastrophic. Swedish registry data show that people with schizophrenia:

• who die from CVD are on average ten years younger than people in the wider population dying from the same cause
• who have CVD are less likely than members of the general population to survive after hospital admission
• are less likely than members of the general population to receive acute hospital care
• have poorer five-year cancer survival rates than the wider population.

Dr Westman and colleagues have also tried using the IMPaCT program. As in the UK experience, it did not result in significant health gains.

We are still searching for pragmatic and effective interventions to address the poor physical health of those with mental health problems. But the aim is clear – in the search for better health, no-one must be left behind, she urged.