Background: This study aimed to investigate the demographical and clinical factors and their predictive powers before and at the end of the 6th and 12th month of lithium prophylaxis.
Methods: Subjects meeting the following criteria were included in the investigation: (1) bipolar patients (DSM-IV); (2) having at least a 3-year lithium prophylaxis; (3) being in either the 'definite poor' or 'definite good' response groups. Both groups were compared regarding sociodemographic and clinical variables.
Results: At the pretreatment point of the prophylaxis, four variables that could predict poor response with 74.1% power were severe episodes, higher ratio of mania/depression, psychotic index episode and being unmarried. At the end of the 6th month, the five variables having 84.89% predictive power for poor response were again the previous three variables and additionally bipolar I diagnosis and poor response to the first 6 months of lithium. At the end of the 12th month, the three variables for poor response had 91.37% predictive power and these were again the previous first two variables and a poor response to the first 12 months of lithium.
Limitations: This was a retrospective study; psychosocial stress was not evaluated by standardized criteria; and the predictive value of personality disorders could not be tested thoroughly.
Conclusions: This study suggests that it is possible to predict, rather reliably, the response to prophylactic lithium regarding clinical variables.