Matt McQueen

The field was left "at a crossroads, wondering in which direction to go next," said Matt McQueen, postdoctoral fellow in the Department of Epidemiology and lead author of a paper that offers the most convincing evidence yet linking specific chromosomal regions to the psychiatric illness. The paper is available now in the advance electronic version of the American Journal of Human Genetics (http://www.journals.uchicago.edu/AJHG/journal/contents/v77n4.html).

Bipolar disorder, a chronic and often debilitating condition, affects more than two million Americans, according to the National Institute of Mental Health (NIMH). Patients undergo dramatic mood swings, cycling from periods of euphoria and great vigor, to periods of utter exhaustion and intense sadness. Approximately 10 to 15 percent of people diagnosed with the disorder end up committing suicide. Psychiatrists have identified two primary forms: bipolar I disorder, which is the classic form of recurring mania and depression, and bipolar II disorder, which has less severe episodes of mania. Treatment usually includes medication.

No one knows what causes the illness, but there are clues. Bipolar disorder tends to run in families, suggesting a genetic basis. Yet, studies of identical twins have shown that one twin's having the disorder does not guarantee its presence in the other sibling, suggesting environmental factors. At this point, psychiatrists suspect a combination of genes and environment to be at play.

When he was a doctoral student, McQueen joined a team of researchers from the Study of Genetic Determinants of Bipolar Disorder, an NIMH-funded project based at Massachusetts General Hospital (MGH), the Broad Institute, and the University of Pittsburgh. The study included a meta-analysis subgroup headed by HSPH Professor of Biostatistics Nan Laird, one of McQueen's thesis advisors. The group made a very unusual decision. Instead of simply relying on summary information from published studies-as is usually the case-the team members went right to the source. They contacted investigators involved in 11 independent studies in several countries and secured raw data, then combined the information to build a pristine genome model to scan for candidate chromosomes.

"The use of original data made a significant difference in our ability to control for variation in several factors among the different data sets and to make the overall analysis much more consistent and powerful," said Laird.

For example, studies may include various diagnoses, such as schizoaffective disorder or even depression, that don't fall neatly into the categories of bipolar disorder I or II. "Different researchers use different terms, and if you are forced to rely on what people publish, you might be combining one study with bipolar I and another study with a related disorder," explained Laird. "Once we had the original data, we could clean that up."

Having the raw data was key to another major feature of the American Journal of Human Genetics paper-the creation of a consistent genomic map. Since the publication of the human genome in 2001, scientists have had a pretty good idea of where various gene markers fall, but, with the influx of new data, the locations of those markers can shift. As a result, each individual linkage study creates its own specific map of genetic markers, which likely does not correspond exactly to the maps of other linkage studies. McQueen and his team were able to build a common map for their paper based on raw data, clearing any inconsistencies that could throw up stumbling blocks.

The resulting analysis involved 1,067 families and 5,179 individuals from North America, Italy, Germany, Portugal, the UK, Ireland, and Israel, who had provided blood samples and family medical histories. The research team combined the data into a single genome scan and found strong genetic signals on chromosomes 6 and 8.

The team now plans additional genotyping in a new population to find specific genes underlying the mental illness.

"At this point, I think the study may help the field prioritize where to look for genes associated with bipolar disorder," said McQueen. "It also has demonstrated the usefulness of combining raw data from numerous studies to provide evidence that may have otherwise gone undetected."