Activation Levels, Cardiovascular Risk, and Functional Impairment in Remitted Bipolar Patients: Clinical Relevance of a Dimensional Approach

approaches, which, as a result, often lack sensitivity to detect subtle behavioral changes in BD patients, particularly during the remission phase [2]. Assessing BD patients using a dimensional approach (MAThyS), which is able to uncouple the assessment of activation levels from mood, we have demonstrated 2 subforms of BD depression, i.e., one characterized by behavioral inhibition/ emotional hyporeactivity and another defined by behavioral hyperactivation/emotional hyperreactivity [3], and that emotional reactivity is a relevant dimension for identification of BD patients at an increased risk for cardiometabolic dysfunction and suicide [4]. While some studies have identified changes in energy/motor activity in BD patients using single symptoms [5], the use of activation levels to distinguish clinical phenotypes among remitted BD patients has not yet been explored. Levels of activation are mediated by a set of interacting brain structures that process and respond to emotional/perceptual stimuli associated with motivation/reward. We aimed to characterize remitted BD patients based on levels of activation through 5 dimensions of behavior, i.e., emotional reactivity, sensory perception, psychomotor activity, motivation, and cognition, and to examine the heterogeneity of clinical phenotypes in terms of cardiovascular risk and psychosocial functioning. For this purpose, 979 remitted BD (according the DSM-IV criteria) were assessed via self-reported questionnaires and a semistructured personal interview conducted by trained psychiatrists in the French Network of Bipolar Expert Centers. Using a cluster analysis approach, patients were grouped according to their activation levels, measured with the MAThyS – a 20-item self-rated scale that assess levels of activation uncoupled from mood during the preceding week. It quantitatively evaluates 5 dimensions, i.e., emotional reactivity, sensory perception, psychomotor activity, motivation, and cognition, each of which can vary from hypoactivation to hyperactivation. The MAThyS score ranges from 0 to 200, and higher scores indicate increased levels of activation. Mood symptoms were evaluated using MADRS and YMRS, and psychosocial functioning was determined using the Functioning Assessment Short Test (FAST). Blood pressure (BP) was assessed and a fasting morning blood sample was taken from all of the patients to measure high-sensitivity C-reactive protein (hsCRP, a marker of lowgrade inflammation) and glucose levels. For detailed information about methods and results, see the online supplementary material. Of the 979 BD patients, 582 (59.4%) were female and 488 (50%) had BD-I, and the mean age was 41.2 years (±12.4) (online suppl. Table S1; see www.karger.com/doi/10.1159/000493690 for all online suppl. material). The use of a machine-learning algorithm enabled us to identify the following 4 clusters (C) of individuals with different clinical phenotypes: hypoactivation (C1), normal activation (C2), mixed activation (C3), and hyperactivation (C4) (online suppl. Fig. S1). Although all of the patients were in remission, as defined by the categorical approaches, 476 (48.6%) presented ab