In searchh of better clinical predictors of functioning in bipolar disorder: a follow-up study

STREJILEVICH S.A., MARTINO D.J., TEITELBAUM J., FASSI G., MARENGO E., IGOA A., AIS.

Conferencia; Ninth International Conference on Bipolar Disorder; 2011

Introduction: Although more than half of bipolar disorder burdenis due to occupational and social impairment, most research hasfocused on outcome measures which have not shown strongassociation with their level of functional impairment. In this studywe tested several ‘‘traditional’’ clinical variables (i.e. number ofepisodes, time spent with symptoms, time spent asymptomatic,history of switch, exposure to di!erent psychopharmacologicaltreatments, etc) and a novel combination of clinical variables(Mood Factors) regarding their correlation with overallfunctioning in a sample of outpatients in naturalistic conditionsof treatment along (on average) 3 years of follow-up.Methods: Clinical data from 56 bipolar I and II patients wasretrospectively analyzed. Inclusion criteria were: a)>12 months oftreatment in our program, b) no interruptions of treatment duringsuch period, c) euthymia > 8 weeks at the time of the assessment(< 8 points in HAM-D and < 6 points in YMRS), d) no otherAxis I comorbidity apart from GAD. Course of illness was weeklyassessed by means of the previously reported life chart techniqueand overall functioning by the Functioning Assessment Short Test(FAST). Age at illness onset, length of illness, previous manic/hypomanic and depressive episodes were obtained from clinicalcharts and patient interview. Spearman bivariate correlations werecomputed to assess the relationship between di!erent measures offollow-up and FAST scores (confidence interval 95%). Linearregression models were used to determine the best clinical predictorof functioning.Results: Age at onset, subsyndromatic manic and depressivesymptoms, current YMRS score, percentage of weeks withsymptoms and the novel measure Mood Instability Factor(number of all mood swings over the time followed) weresignificantly correlated with FAST score. After multivariateregression, only MIF emerged as an independent predictor offunctioning (Adjusted R2= 0,288 ± 5, 7).Discussion: Clinical measures that assess mood instability werebetter predictors of functional recovery than those which focusedon either intensity or time spent with symptoms.