We found that brains shrink in winter
It may be hard to believe a bipolar patient had impulse-control issues because of a medication and not because of the bipolar disorder itself. But with millions of data points pointing to a problem, it was hard to ignore.
At the Olin Neuropsychiatry Research Center at Hartford Hospital, we recently found something unexpected with our data. In a set of MRIs from 3,279 healthy adults over 15 years, our study found that certain parts of the brain appear to get larger in summer and smaller in winter. Other parts of the brain appear to do the opposite.
This study seemed to show that something seasonal was happening. For people suffering from seasonal affective disorder, a study indicating that the brain itself is changing is, well, mindblowing. This study was only possible because of the availability of a dataset large enough to observe the effect. Such datasets often surprise scientists. We need to do more with them.
It turns out that many diseases have a relationship to season: Migraines, multiple sclerosis, Alzheimer’s and schizophrenia are just a few. One theory is the connection between viral infection and autoimmune diseases. Viral infection is a known trigger of autoimmune diseases, such as type-1 diabetes, and bacterial infection is known to cause a range of anxiety symptoms in children.
There are simply more infections in winter than in summer, so you are more likely to develop an autoimmune disease in the winter. It’s a weird concept that mental health disorders could be triggered by an infection, but it’s the kind of theory that needs more research through datasets.
Many mental health disorders are diagnosed through subjective checklists, and psychiatric medications are often prescribed to treat the disorders. Many medications are effective, but some gain foothold as treatments with little scientific evidence — few datasets — for their efficacy, causing unforeseen problems.
An example is the antipsychotic Abilify. It was approved to treat schizophrenia in 2002, and depression in 2007, in adults. But then the drug maker then received a multi-million-dollar fine for, among other things, illegally promoting Abilify for unapproved uses with children and with geriatric patients suffering from dementia-related psychosis.
Despite the fine, the marketing was working. By 2017, the drug was the eighth bestselling in 25 years, with well over $50 billion in sales.
The full picture of the drug, however, was not so rosy. With millions of prescriptions, there were reports of people developing pathological gambling habits while taking the drug. Some patients allege they lost hundreds of thousands of dollars.
It may be hard to believe a bipolar patient had impulse-control issues because of a medication and not because of the bipolar disorder itself. But with millions of data points pointing to a problem, it was hard to ignore. Analysis of the Food and Drug Administration’s adverse-events database showed a clear link between Abilify and impulse-control problems, and the drug received a black box warning. Big data was saving people from what might have otherwise been a hard-to-identify drug side-effect.
If a drug can cause impulse-control issues in adults, what would that look like in a child? Children taking Abilify experienced significant weight gain and metabolic issues at greater rates than the adults for which the drug was approved.
Developing a vaccine is easy, but fixing mental health is hard. A common tagline at the end of scientific papers on mental health is the phrase “but more research is needed.” This phrase is there because more research is needed.
Datasets exist that are large enough to test many theories about the causes of mental illness and to test the efficacy of treatments. The data is available now, and the answers to our questions are in the data. We must invest more resources to explore the data. Datasets can surprise us and even save us.