Sir: Deciding whether to prescribe antidepressants for the depressed patients with bipolar spectrum disorders remains a challenge for clinicians. Two most-feared potential consequences of using antidepressants in patients with bipolar spectrum disorders are triggering a manic switch and/or generating a rapid cycling course of the disease.1 Experts have long tried to recognize the depressed patients who are at a greater risk for switching to mania; thus, several symptoms have been suggested as “soft signs” of bipolarity, including hypersomnia, psychomotor retardation, young age at onset of depression, family history of bipolar disorder, drug-induced hypomania, seasonality of the episodes, and diurnal variation of the symptoms.2 However, the list is not yet complete, and many cases still cannot be correctly predicted. As a result, clinical findings that could be used as new predictors of bipolar spectrum disorders could be helpful for future interventions. Case report. Ms. A was a 27-year-old married woman who experienced an aggravation of depressive and anxiety symptoms for 6 months beginning in August 2005. Symptoms included depressed mood, loss of interest, repeated crying spells, loss of concentration, insomnia, decreased appetite, irritability, worry, and preoccupation with death. Past history revealed no significant finding other than mild depressive and anxiety symptoms in the previous 4 years and no episodes of mania or hypomania. She had no suggestive symptoms for a bipolar spectrum disorder except for very brief periods (about 30 seconds) of euphoria and inflated self-esteem, which occurred just before falling asleep, in a semiconscious state between wakefulness and sleep. During these periods, she experienced an exuberant sense of power, and she believed that she was able to “move mountains.” Family history for bipolar disorder was negative. The patient met the criteria for major depressive disorder based on DSM-IV-TR and received 5 mg/day of citalopram (February 2006), and after 5 days, the dose was increased to 10 mg/day, in addition to 1.25 mg/day of alprazolam, in 3 divided doses. One week later, the patient showed significant improvement. In addition, alprazolam had reduced the sleep latency, and the patient did not experience or remember any sleep-related hypomanic symptoms. Instead, she reported having distinct periods of euphoria and talkativeness 15 to 20 minutes after taking each dose of alprazolam. These periods would happen after all doses of alprazolam and would completely resolve after about 10 minutes, and the patient would return to her depressed baseline. The patient did not experience any other side effects other than a mild sedation during the day. The alprazolam dose was reduced to 0.75 mg/day to decrease sedation and then tapered to complete cessation in the next 3 weeks, and citalopram was continued. Sleep-related hypomanic symptoms did not recur after reducing or discontinuing alprazolam. The patient was followed for 6 more months and remained euthymic (August 2006). Recent studies suggest that the lifetime prevalence of bipolar spectrum disorders in the general population is more than 6%,3 that is, much higher than what was previously speculated.2 It is also reported that about 12.5% of the patients initially diagnosed as having depression will be reclassified as having a bipolar spectrum disorder in their course of illness.4 This patient reported experiencing distinct brief periods of euphoria and inflated self-esteem before sleep onset, which could potentially be related to bipolar spectrum disorders. Even though we did not have enough evidence to accept it as a new soft sign of bipolar spectrum disorders before starting the medications, transient hypomanic symptoms induced by taking al-prazolam suggested that “sleep-related periods of euphoria” could be assumed as one of the predictors of bipolar spectrum disorders. We are aware that had the patient been followed without decreasing the dose of alprazolam, she might not have developed a bipolar disorder. Obviously, this could not be performed because of ethical considerations. Future studies could tell more about the diagnostic value and effect of each of the soft signs on the long-term course of the disease.