About a decade ago NIH held a series of workshops to develop a Roadmap for Biomedical Research. We brought experts to Bethesda to suggest new directions for technology development, training, and a range of other areas. In one such meeting, on re-engineering clinical research, many of the nation’s leading clinical scientists debated how to improve our approach to research on diseases. Well into the two-day workshop, Dr. Eugene Braunwald stopped the discussion when he suggested he could solve the nation’s problems with clinical research in a single word. Dr. Braunwald, then in his mid-70’s, was one of America’s most respected cardiologists and thought leaders, so his pronouncement was met with awe and anticipation. After a carefully timed, dramatic pause, Dr. Braunwald said simply, “Sweden.”
I remembered this moment recently reading a new report on co-morbidity and mortality in persons with schizophrenia in Sweden.1 What Dr. Braunwald was thinking about was the Swedish universal health care system and population registries, making Sweden the ideal setting for epidemiology, the study of patterns and determinants of disease in populations. For a disorder like schizophrenia, virtually every affected person is identified and the predictive value (accuracy) of diagnosis is over 94 percent. Unlike in this country, there are few barriers to care based on insurance, access, or geography.
Which is why you should look at these new data on mortality. Crump and colleagues followed a Swedish national cohort of over 6 million adults between 2003 and 2009 to detect mortality and illness based on the results of every outpatient or inpatient visit nationwide. Among the 8,277 people with schizophrenia, men died 15 years earlier and women died 13 years earlier than the rest of the population. This early mortality was not due to suicide, but to cardiovascular disease, cancer, and pulmonary disease. The adjusted hazard ratio (increased risk) for mortality from ischemic heart disease in women with schizophrenia was a stunning 3.3—in other words, women with schizophrenia have over 3 times the risk of dying from heart disease compared with women in the general population—and for men with schizophrenia it was 2.2. Importantly, heart disease and cancer were not more common in people with schizophrenia, but mortality from these diseases was increased markedly.
These numbers might not be so surprising in the United States, where recent studies report early mortality in people with serious mental illness ranging from 8 years2 to 27 years.3 But in a country with a far more effective health care system, one might have expected much better health outcomes. In fact, in the Swedish study, people with schizophrenia were seen nearly twice as often for medical care as the general population. Yet even with these extra visits, heart disease and cancer went undetected: only 26.3 percent of people with schizophrenia who died of heart disease and 73.9 percent who died of cancer had been diagnosed previously. Another surprise—treatment with antipsychotic medications, which might have been considered a risk factor for cardiovascular disease, actually lowered the risk. The highest risk was among those not treated with antipsychotic medication.
What’s the lesson for the United States? The authors end this new report with this reflection: “Underdetection of important causes of mortality in schizophrenia patients in Sweden, despite universal health care, raises the question of whether it may be an even larger problem in countries without universal health care.” Indeed. While we are hopeful that the implementation of mental health parity and new, integrative care approaches such as medical homesExternal Link: Please review our disclaimer. will close the gap on early mortality and under-treated co-morbid conditions in the United States, the Swedish data suggest this may not be so easy.
Recent reports in the United States have documented the success of smoking cessation and weight loss programs tailored for people with serious mental illness.4,5 These are important signs of progress in an area that has been woefully neglected for too long. But the Swedish report suggests the problem, at least for heart disease and cancer, is more a lack of detection than a lack of treatment. Indeed, those with schizophrenia who had been diagnosed with ischemic heart disease had only a slightly higher mortality risk and those diagnosed with cancer had no higher mortality risk than people without schizophrenia who had been diagnosed with these diseases.
The lesson is that systemic changes in health care are necessary but may not be sufficient to reduce mortality from co-morbid diseases in people with serious mental illness. We will also need to build in better detection of heart disease, cancer, and pulmonary disease as well as better management of diabetes. To this end, recent initiatives by NIMH will support testing of both existing and newly developed innovative service interventions to reduce health risk factors and premature mortality in people with serious mental illness. We will also be issuing funding announcements to support research aimed at reducing delays in early detection and referral to services of individuals experiencing first episode psychosis. Our intention is to obtain actionable information that can be rapidly applied to change practice across the continuum from screening to detection to treatment.
These new numbers from Sweden should remind us that serious mental illness is a health disparity issue. One way to think about losing 13 – 15 years of life expectancy is to realize that people with serious mental illness have not benefitted fully from the gains in longevity over the past half century. We frequently say “no health without mental health” to stress the importance of treating mental illness as a pathway to better health outcomes in society. For those with schizophrenia, even in the most advanced health care system in the world, we are still facing early mortality from lack of diagnosis and treatment of medical illnesses.
1 Crump C et al. Comorbidities and mortality in persons with schizophrenia: a Swedish national cohort study. Am J Psychiatry. 2013 Mar 1;170(3):324-33. Doi: 10.1176/appi.ajp.2012.12050599.
2 Druss BG et al. Understanding excess mortality in persons with mental illness: 17-year follow up of a nationally representative US survey. Med Care. 2011 Jun;49(6):599-602. doi: 10.1097/MLR.0b013e31820bf86e.
3 Colton CW, Manderscheid RW. Congruencies in increased mortality rates, years of potential life lost, and causes of death among public mental health clients in eight states. Prev Chronic Dis. 2006 Apr;3(2):A42.
4 Compton MT, Daumit GL, Druss BG. Cigarette smoking and overweight/obesity among individuals with serious mental illnesses: a preventive perspective. Harv Rev Psychiatry. 2006 Jul-Aug;14(4):212-22.
5 Daumit GL et al. A behavioral weight-loss intervention in persons with serious mental illness. N Engl J Med. 2013 Mar 21. [Epub ahead of print]