Thyroid Hormone Treatment Efficacy in Bipolar Disorder with Irregular and Sparse Cyclical Longitudinal Data

• Activity Number: 525
• Type: Contributed
• Date/Time: Wednesday, August 3, 2011 : 10:30 AM to 12:20 PM
• Sponsor: Section on Nonparametric Statistics
• Author(s): Brian Calimlim*+ and Catherine Ann Sugar and Patricia Walshaw
• Companies: University of California at Los Angeles and University of California at Los Angeles and University of California at Los Angeles
• Keywords: Bipolar Disorder ; Markov Chains ; Permutation Tests ; Cyclicity ; Longitudinal Data ; Depression/Mania

Abstract

Bipolar Disorder (BD) is a psychiatric illness in which a person experiences episodes of elevated mood and energy, known as mania, and episodes of lowered mood and interest, known as depression. Due to its complex nature, challenges arise in analyzing BD data. Irregular cyclical patterns, varying in amplitude and phase by subject, make it difficult to develop simple parametric models, while the dual nature of BD requires a multi-component outcome measure. Further complications exist in treatment trials as efficacy can be manifested in many ways. In a study of thyroid hormone treatment efficacy for refractory rapid cycling BD, 32 patients were randomized to receive triiodothyronine (T3), thyroxine (T4), or a placebo plus usual medication. Depressive and manic symptoms data were longitudinally gathered before and after treatment. We present an analysis of this study and strategies taken to handle data irregularity and sparsity. A Markov chain approach is used for modeling, while permutation tests are developed to determine statistical significance. Preliminary results suggest patients receiving T3 show reductions in mania while those on T4 improve with regard to depression.