Background and Purpose: Patients with Bipolar I Disorder recurrence experiences mood

variation between manic and depression during the time. Hence, that is need to the longitudinal

study on Bipolar Disorder patients. This study aims to evaluate the effective factors on Bipolar I

Disorder frequent recurrence in 5 years longitudinal study using generalized estimation

equations (GEE) method.

Materials and Methods: Data were collected with repeated measurements on 255 Bipolar I

Disorder patients in mazandaran, Iran, in a longitudinal study between 2007 and 2011. The

outcome variable is Bipolar I Disorder recurrence, and the predictor variables are as follows:

sex, age of onset, family history (Grade 1), economic status and education level. In this paper,

SAS PROC GENMOD was used to apply GEE regression to the assessment of parameters

corresponding to the factors causing recurrence.

Results: The age was among 13-55 years and the average of age of onset was 24.1 years. Almost of

patients were male and had economic status with (upper/middle) deciles and also had a diploma and

under diploma education level. The results of GEE method showed that the covariate of family

history (Grade 1) increased the odds of recurrence (odds ratio [OR] >1; P < 0.0500); and age of onset

decreased the odds of recurrence in patients with Bipolar I Disorder (OR <1; P < 0.0500).

Conclusion: Predictor variables in recurrence Bipolar I Disorder include first-degree relatives’

psychiatric family history and age of onset. Understanding this factors, and educate patients, and

their families are valuable for the prevention and planning the treatment.