The open-label trials evaluated desipramine, nortriptyline, and bupropion sustained release.[Level of evidence: II] The studies included patients experiencing depressive symptoms after the deaths of their loved ones. The depressive symptoms were evaluated using the Hamilton Depression Rating Scale (HDRS). All studies evaluated intensity of grief using select grief assessment questionnaires.
Data from these studies suggest that antidepressants are well tolerated and improve symptoms of depression. Data also suggest that the intensity of grief improved but that the improvement was consistently less in comparison with the symptoms of depression. Limitations of these studies include open-label treatment and small sample sizes.
The only randomized controlled study conducted to date [Level of evidence: I] compared nortriptyline with placebo for the treatment of bereavement-related major depressive episodes. Nortriptyline was also compared with two other treatments, one combining nortriptyline with IPT and the other combining placebo with IPT. Eighty subjects, aged 50 years or older, were randomly assigned to one of the four treatment groups: nortriptyline (n = 25), placebo (n = 22), nortriptyline plus IPT (n = 16), and placebo plus IPT (n = 17).
The 17-item HDRS was used to assess depressive symptoms. Remission was defined as a score of 7 or lower for 3 consecutive weeks. The remission rates for the four groups were as follows: nortriptyline alone, 56%; placebo alone, 45%; nortriptyline plus IPT, 69%; placebo plus IPT, 29%. Nortriptyline was superior to placebo in achieving remission (P < .03).
The combination of nortriptyline with IPT was associated with the highest remission rate and highest rate of treatment completion. The study did not show a difference between IPT and placebo, possibly owing to specific aspects of the study design, including short duration of IPT (mean no. of days, 49.5) and small sample size. The high remission rate with placebo was another important limitation of the study. Consistent with previous open-label studies and for all four groups, improvement in grief intensity was less than improvement in depressive symptoms.
In summary, all of the antidepressant studies conducted to date suggest that the magnitude of reduction and rate of improvement in grief symptoms are slower than the decrease in magnitude and rate of improvement in depressive symptoms. One group of researchers  provides possible explanations for this phenomenon, arguing that depressive symptoms may be more responsive to pharmacological intervention because they are directly related to biological dysregulation and neurochemical changes. The other possibility is that the persistence of grief without depressive symptoms is not pathological-it might be a normal and necessary consequence of the bereavement process.