Project Summary/Abstract
The negative consequences of pediatric bipolar disorder (BP) are well documented, but the course, outcome,
and prognostic factors of BP as youth transit into adulthood remain unknown. To answer these questions, the
Predicting Adult Outcomes in BP Youth (PROBY), a multisite study (University of Pittsburgh, PI: B.Birmaher;
Brown University, PIs: M.Keller/S.Yen; UCLA, PI: M.Strober), proposes: 1) to extend the prospective follow-up
of the Course and Outcome of Bipolar Youth (COBY) cohort into adulthood, and 2) to use COBY data to
build risk calculators to predict short-(2 years), intermediate-(5 years), and long-term (10 years) personalized
mood course/outcome that capture transitions from adolescence through adulthood. We will study five key
domains relevant for BP: 1) course of mood and non-mood disorders; 2) psychosocial functioning; 3)
suicidality; 4) SUD; and 5) cognitive functioning. Risk calculation is an analytic approach that instead of
group prediction, yields individualized estimates of risk for a specified event in a given patient, based on the
presence of multiple risk factors. The risk calculators proposed in PROBY will have a true population impact by
predicting BP outcomes, similar to those used in cardiovascular disease and cancer, but rarely applied in
psychiatric disorders. The risk calculators will include well-established variables, in order to allow replication
and clinical utility. Also, these variables are easily ascertained in clinical settings and are part of standard
practice and thus, will be widely disseminable, and hold promise for utility in daily clinical practice. For
example, results may be used to: counsel patients/families about long-term prognosis; individualize treatment
by identifying patients likely to require prolonged treatments and those for whom medications may be reduced
with minimal risk; and inform early intervention/prevention and clinical/biological research. To pursue the
proposed study, 343 COBY subjects (BP-I=72.3%, BP-II=14.6%, BP-NOS=13.1%, mean age 26.3 (18.7–33.7),
47% female, 82% Caucasian, and on average, middle class), will have 2 additional clinical assessments over
the five years of study. These subjects represent 77% of the original sample (n=446) enrolled between 2000
and 2006 at the three study sites. PROBY will continue to prospectively collect data on categorical and
dimensional psychopathology, protective factors, psychosocial functioning (e.g., relationships, work, academics),
cognitive functioning, inflammatory markers, cardiometabolic conditions, exposure to stressful/traumatic life
events, and treatment (e.g., medication dosages, barriers). With 5 additional years of prospective observation,
the mean age of the sample will be 31.3 years old (23.7–38.7). Thus, PROBY will be the first study with
adequate sample size to answer questions about what happens to BP youth when they grow up. Furthermore,
PROBY will be the first study to build risk calculators to predict outcome in key domains relevant for BP and
validate these results using an independent adult BP sample (n=427) from the Collaborative Depression Study
(CDS; R01 MH025416). These risk calculators have the potential to widely transform clinical care of BP.