Logo for University of Iowa Health Care This logo represents the University of Iowa Health Care

Largest Longitudinal Brain Imaging Study to Date Carries Critical Implications for Treatment of Schizophrenia

Nancy Andreasen

by Jude Gustafson, Department of Psychiatry, UI Health Care

Internationally recognized UI Health Care schizophrenia expert, Nancy Andreasen, MD, PhD, recently published important new findings in the American Journal of Psychiatry in an article titled Relapse duration, treatment intensity, and brain tissue loss in schizophrenia: a prospective longitudinal MRI study. This study confirmed serious implications for antipsychotic dosing in the treatment of schizophrenia, and also warns of potential repercussions of antipsychotic use in treatment of other psychiatric disorders.

Statistics on prevalence of schizophrenia reflect a tremendous burden of disease. Schizophrenia is common, affecting 1.1% of the population of the United States, which in the current estimated total U.S. population of 315 million, represents 3,465,000 people. The World Health Organization estimates 24 million people suffer with schizophrenia worldwide .

Andreasen says, “Schizophrenia is one of the most severe mental illnesses psychiatrists treat, and it has a huge impact on society. Although some people with schizophrenia are highly successful and function well in society, outcomes are widely varied. It is associated with a fairly high suicide rate, a shortened life expectancy, and it affects people’s ability to interact within the community, to hold a job, and to have relationships. Years ago, most people with schizophrenia were hospitalized much of their lives. Now, thanks to better treatments, people are living in the community, but many are severely handicapped.”

Her new publication contains surprising results drawn from data derived from the largest, longest, and most fine-grained longitudinal study of schizophrenia that has ever been done. The study documented the evolution of structural brain changes seen in MRI scans following first-episode patients starting at the time of illness onset, and subsequently following up with sequential scans at 2, 5, 9, 12, and up to 15 years post onset. Clinical data on symptom severity and treatment were also collected at six-month intervals. The resulting data represent a unique trail of scientific evidence that allowed Andreasen’s research team to ask questions that previously could not have been answered, including queries about total brain tissue volumes; discrete measures of grey matter versus white matter; and whether age, duration of relapse, and antipsychotic dose had an effect on tissue volumes over time.

One important finding was that, on average, first-episode patients, at the time of onset, already had less brain tissue than age-matched healthy volunteers, which suggested something had happened prior to the onset of overt symptoms. With a subject population averaging 22 years of age at onset, Andreasen postulates that processes affecting brain development may go awry in people who develop schizophrenia.

She explains, “When does schizophrenia really begin? The answer is complicated, but with brain imaging technologies, we have learned a lot in recent years about human brain development throughout the lifespan. There are several studies, mine included, that show people with schizophrenia have smaller-than-average cranial size. Since cranial development is completed within the first few years of life, there may be some aspect of earliest development—perhaps things such as pregnancy complications or exposure to viruses—that on average, affected people with schizophrenia.”

“The normal human brain doesn’t complete maturation until about age 25. The brain is like a garden where, just in case a drought or something else could negatively affect growth, you plant more seeds than you need to harvest. Then, after the seeds sprout, you thin them out some so that the remaining plants can thrive. Similarly, grey matter actually over-grows, and a process called pruning sets in at about age 15 characterized by a gradual decrease in grey matter until about age 25. In parallel with that, the tracts of connections between brain cells known as white matter actually increase between age 15 and 25. So, another possible cause of schizophrenia could be things that happen while the brain is being sculpted into maturity, such as excessive pruning or failure of white matter myelination. Those processes are largely regulated by genetic code, but they’re also affected by environmental influences, and can be especially impaired by substance abuse.”

Another significant, yet puzzling, observation is that although people with schizophrenia, on average, tend to lose more tissue over time than healthy controls, the finding is not universal, with some participants’ data reflecting measures typical of the normal brain. Emphasizing this fact, Andreasen says, “I make this point for a reason. There’s a phenomenon in science called the ecological fallacy, which is the tendency to overgeneralize group findings to individuals, assuming every finding has to pertain to each individual in the group. It’s scientifically incorrect to overgeneralize, but it’s also psychologically wrong to overgeneralize, because we don’t want people to consider this a hopeless illness.”

Andreasen’s study was also in search of age-related brain volume changes and their potential relationship to schizophrenia. She initially hypothesized that tissue atrophy would increase over time and be greatest later in the illness, but data revealed exactly the opposite. In fact, tissue loss was greatest during the first two years after onset, then plateaued. This unexpected finding thus identified a critical period when appropriate, effective treatment might have the ability to attenuate tissue loss or other negative effects of the illness.

So, how do you treat an illness like schizophrenia? For about 40 years, medications have been used to block cell receptors for a chemical naturally produced in the brain called dopamine. These drugs help control what are known as positive symptoms of psychosis related to schizophrenia, which include delusions and hallucinations. Unfortunately, however, blocking dopamine has very little effect on negative symptoms of schizophrenia, such as flat emotions and diminished pleasure in everyday life. Andreasen’s own broader insights to the illness early in her career eventually revolutionized perspectives in psychiatry leading to changes in diagnostic criteria, and the constructs of negative symptoms are now widely recognized as a main component of diagnosis of schizophrenia.

Reflecting on past problems in treatment, she says, “Early on, when dopamine-blocking treatments initially became available, people believed higher doses would result in better long-term outcomes. But over time, serious side effects, especially movement disorders, resulted from high dosing, so they were generally reduced to try to avoid these problems.”

With this historical perspective in mind, Andreasen neither expected to find a major atrophy effect, nor an improvement in brain tissue volume related to treatment intensity, but data analysis refuted her hypothesis time and again. She laments, “We were very disappointed to find that a higher medication dose correlated with greater brain tissue loss. Again, the findings were not universal for each person or each type of treatment, but generally speaking, the greater the medication dose, the greater the amount of brain tissue loss.”

This observation could be reliably made because the research team created a novel metric construct called “dose years” to evaluate the data. In the same way studies on smoking use “pack years” to account for differences in the number of cigarettes smoked by individuals, the same basic principal can be used to account for dosage differences among patients. Dose years calculated with relation to MRI data showed an unmistakable, significant adverse effect of higher dose on brain tissue volumes.

Andreasen’s disappointment is evident as she says, “This was a very upsetting finding. We spent a couple of years analyzing the data more or less hoping we had made a mistake. But in the end, it was a solid finding that wasn’t going to go away, so we decided to go ahead and publish it. The impact is painful because psychiatrists, patients, and family members don’t know how to interpret this finding. Should we stop using antipsychotic medication? Should we be using less?”

Another question the group was trying to answer had never before been covered in research—the question of whether duration of relapse, as opposed to only number of relapses, was correlated with tissue volume changes in the brain. Again, using very detailed longitudinal records, the data were able to shed some light on this question, as well.

Coming from a background in the humanities, Andreasen understands the importance of defining concepts used in her research. She says, “I like to be precise about how I use language because solid definitions are critical to accurate data interpretation and results. The old fashioned way of gauging relapse in schizophrenia is a reduction in symptoms of 20 percent. This actually doesn’t make very much sense because if one person has very severe symptoms and they improve 20 percent, the person is still very sick. Whereas, if someone else has much less severe symptoms, it can be difficult to measure a 20 percent improvement because their symptoms just aren’t that bad to begin with.”

To correct this potential confound, her team created another model that was much more clinically useful. “We used a novel rating scale for measures of severity, and included only data for symptoms lasting at least two weeks. As for the duration of the relapse, we found that the number of relapses didn’t predict anything significant in terms of brain tissue volume change, which is rather [sardonic] because so much of the previous research tracked the number of relapses, but never paid any attention to how long they lasted. Our strategy enabled us to clearly see that duration of relapse correlates positively with brain tissue loss—so in other words, the longer the relapse, the greater the amount of brain tissue loss. Again, this is on average and doesn’t pertain to everybody.”

What have doctors traditionally done when a patient relapses? They increase the medication dose. Confident in her new findings, Andreasen recommends new approaches to treatment.

As if talking about an old, familiar friend, Andreasen says, “Thank heavens for statistics. They don’t solve every problem, but they can be tools to help us find answers and gain new knowledge if we are creative, and at the same time, careful to let the data reveal their own truths and show us when we’re wrong in our assumptions. With the help of a really good biostatistician, we were able to develop a model where we could control for treatment intensity while taking duration into account. In the end, we found that both of these things were important.”

“I feel strongly that we need to reassess the pervasiveness with which these medications are being used by doctors at present. We do need the treatments. We used to have hundreds of thousands of people chronically hospitalized. Now, most are living in the community, and this is thanks to the medications we have. People are able to work, they have better social relationships, and they have fewer symptoms. But antipsychotic treatment has a negative impact on the brain, so we must use common sense and prescribe these medications in the lowest possible dose to control symptoms. We must also think about how these medications are being used for things that have nothing to do with schizophrenia, such as behavioral problems in children or the elderly. We must get the word out that they should be used with great care, because even though they have fewer side effects than some of the other medications we use, they are certainly not trouble free and can have lifelong consequences for the health and happiness of the people and families we serve.”

The Study Research Team: 
Nancy Andreasen, MD, PhD, Andrew H. Woods Chair of Psychiatry, University of Iowa Health Care;
Beng Choon Ho, MD, Associate Professor of Psychiatry, Department of Psychiatry, University of Iowa Health Care;
Dawei Liu, PhD, Visiting Faculty, Department of Psychiatry, University of Iowa Health Care;
Anvi Vora, MD, Resident, Department of Psychiatry, University of Iowa Health Care;
Steven Ziebell, BA, Senior Research Assistant, Department of Psychiatry, University of Iowa Health Care