Early research on the pharmacotherapy of pathological grief responses centered on TCAs , since this class of medication was first-line for depressive symptoms until the advent of SSRIs and other newer-generation pharmacologic agents. Jacobs et al. [8] conducted the first TCA study on bereavement-related depression, in an open-label trial of ten widows and widowers (age range 26–65 years, 80% women). When participants were rated on a global symptom improvement scale after 4 weeks of treatment with desipramine (dose range 75–150 mg/day), four participants were rated as “very much improved” and three as “much improved.” One participant dropped out due to side effects. All of the improved participants also showed significantly reduced depressive symptoms, but only three of the seven reported a significant reduction in grief intensity.

The effect of nortriptyline on bereavement-related depressive symptoms, sleep quality, and grief intensity was examined by Pasternak et al. [9] in another open-label trial of 13 bereaved spouses (mean age 71.1 years, 61.5% women). In their analysis of study completers, it was found that depressive symptoms were significantly reduced after a median treatment period of 6.4 weeks at a nortriptyline mean dose of 49.2 mg/day. Grief intensity also improved, but only marginally so (9.3% decrease in grief intensity).

Reynolds et al. [11] performed the only randomized controlled trial on a tricyclic antidepressant for bereavement-related depression. Eighty older adults (mean age 66.1 years, 72.5% women) were randomized into one of four arms: nortriptyline alone (n = 25), placebo alone (n = 22), nortriptyline plus interpersonal therapy (IPT) (n = 16), or placebo plus interpersonal therapy (n = 17). Participants were required to meet criteria for MDD and to have high grief intensity on a grief symptom scale following the death of a spouse. Following 18 weeks of treatment at a mean nortriptyline dose of 66 mg/day, 69% of participants in the nortriptyline plus IPT group achieved remission of depression. In comparison, 56% achieved remission with nortriptyline alone, 45% with placebo alone, and 29% with placebo plus IPT. Controlling for age, nortriptyline was found to have a significant effect over placebo for bereavement-related depression, but no effect was found for IPT or for nortriptyline plus IPT. Additionally, no differential effect was found for any treatment condition on grief intensity.

In sum, trials assessing the efficacy of TCAs for pathological grief responses are few. In the three that have been conducted, TCAs were uniformly found to be effective in the treatment of bereavement-related depressive symptoms, as expected based on prior literature and clinical results. However, all three trials also found that TCA pharmacotherapy was either marginally efficacious or not efficacious in reducing pathological grief symptoms. Of interest, this evidence supports the hypothesis that pathological grief reactions are a distinct clinical entity with a different treatment response profile than depression. Thus, as suggested by these trials, it may be difficult to evaluate the efficacy of pharmacological agents for CG in the context of comorbid depression.

The advent of SSRIs coincided with the operationalization of CG. Zygmont et al. [12], in an open-label trial, examined the use of paroxetine in the treatment of 15 individuals with CG after the loss of a loved one (mean age 57 years, 73.3% women). These participants were concurrently enrolled in a study for psychotherapy treatment development. After 16 weeks of treatment with paroxetine at a median dose of 30 mg/day, grief symptoms decreased by 53%, while depression symptoms decreased by 54%. When compared to a separate, ongoing trial of nortriptyline for bereavement-related depression , it was found that results were similar: both agents yielded reductions in depressive and grief symptoms; and in both cases, there was a greater improvement in depressive symptoms than in grief symptoms. Moreover, since the paroxetine group was more heterogeneous than the nortriptyline group, with great chronicity and comorbidity of illness, the authors suggested favoring paroxetine in clinical practice.

Shear et al. [13] completed another open-label trial of escitalopram on 17 participants diagnosed with CG. The study duration was 16 weeks, with an escitalopram starting dose of 10 mg/day and an option to increase the dose to 20 mg/day in the fourth week. At the end of 16 weeks, grief symptoms had improved by 35% in the study completers and only 24% in the intention-to-treat sample (defined as having attended at least one session).

Simon et al. [14] prospectively examined four patients with CG in a case series on the use of open-label escitalopram on the treatment of complicated grief (mean age 41.8 years, 100% women). The study was conducted for 10 weeks at a flexible dosing range of escitalopram from 10 to 20 mg/day. At study completion, 100% of participants responded with a rating of “very much improved” on a global symptom improvement scale. Both grief scores and depression scores were also significantly improved.

Hensley et al. [15] conducted another open-label trial of escitalopram on 30 individuals with bereavement-related depression. Of these participants, 14 met diagnostic criteria for CG in addition to MDD, while 15 only met criteria for MDD. After a 12-week trial of flexibly dosed escitalopram (dose range 10–20 mg/day), 66% of subjects experienced a 50% or greater reduction in depressive symptoms. When analyzed by CG diagnosis, grief scores were reduced by 21% in the CG group and by 39% in the uncomplicated grief group; this difference was not statistically significant. As measured by a treatment response of “very much improved” or “much improved” on a global symptom improvement scale, 83% of the whole sample experienced improved depressive symptoms, while 45% experienced improved grief symptoms.

The first randomized controlled trial of an SSRI for CG was recently conducted [4]. In this study, the effect of citalopram on CG symptoms was assessed with and without concurrent psychotherapy (CGT). The total sample consisted of 395 adults across four sites in the USA (mean age 53, 78% women). Subjects were divided into four groups: citalopram alone (n = 101), placebo alone (n = 99), citalopram with CGT (n = 99), and placebo with CGT (n = 96). Over 20 weeks of treatment, citalopram was flexibly dosed to its maximum allowable daily dose, which decreased during the course of the study from 60 to 40 mg/day due to an alteration of US Food and Drug Administration guidelines for this medication. The resultant median dose of citalopram was 40 mg/day. Global symptom improvement assessments anchored in CG were administered every 4 weeks, with treatment response defined as a rating of “much improved” or “very much improved.” At the end of the treatment period, CGT was found to lead to a significant improvement in grief symptoms, but the addition of citalopram did not affect CGT outcome. Depression symptoms were noted to decrease significantly when citalopram was added to CGT, and conversely adding CGT to citalopram also significantly improved citalopram outcome. However, there was no significant difference between citalopram and placebo at either week 12 or week 20 on grief symptom severity.