Discontination of Antidepressants in Remitted Depression

Antidepressants (ADs) are effective in the treatment of moderate-to-severe depression and anxiety disorders and are recommended by national and international clinical guidelines, although concerns about over-prescribing persist. In Italy, AD consumption increased by over 10% between 2014 and 2020. Selective serotonin reuptake inhibitors (SSRIs) account for approximately 70% of AD prescriptions and around half of total AD-related expenditure. Across most countries, AD prescribing has increased steadily, largely driven by use in general practice and by longer treatment duration.

For individuals experiencing a first depressive episode, evidence-based guidelines recommend continuing AD treatment for 6-9 months after clinical response. Meta-analytic evidence suggests a similar risk of relapse for maintenance periods ranging from six months to over one year. Treatment duration may vary according to illness severity and recurrence, with at least two years commonly recommended in cases of multiple episodes with functional impairment. In Italy, approximately one in three individuals continue AD treatment beyond 12 months, raising concerns about potentially inappropriate long-term or indefinite prescribing.

Prolonged AD use is associated with an increased risk of long-term adverse effects, including sexual dysfunction, weight gain, emotional blunting, and cardiovascular conditions. Discontinuation of ADs may also lead to withdrawal symptoms, which can be clinically relevant in a substantial proportion of patients. Several clinical factors appear to increase the risk of withdrawal, and it has been suggested that, for some individuals, disabling withdrawal symptoms rather than relapse may be the primary reason for unsuccessful discontinuation.

Randomized controlled trials indicate that discontinuing ADs is associated with a higher risk of relapse compared with continuation. However, discontinuation schedules used in many trials are rapid or abrupt, potentially increasing withdrawal symptoms that may be misclassified as relapse and thereby inflating relapse estimates. There is broad agreement that gradual tapering may reduce withdrawal symptoms, but most clinical guidelines provide only generic recommendations, often implying a linear dose reduction approach that remains standard in routine practice.

Alternative tapering strategies have been proposed, most notably hyperbolic tapering, which involves proportional dose reductions over longer periods. This approach is based on the nonlinear relationship between dose and receptor occupancy and may require tapering periods ranging from several months to over one year. Hyperbolic tapering often necessitates very low doses that are not readily achievable with standard marketed formulations, raising feasibility, regulatory, and policy challenges. To date, no randomized trials have directly compared different tapering strategies.

Against this background, there is a clear need for experimental evidence comparing hyperbolic and linear tapering approaches in routine clinical practice. The DISCARD study is a pragmatic, multicentre randomized trial designed to address this gap by comparing hyperbolic versus linear tapering in adults with currently remitted depressive disorders who have been taking ADs for at least six months and for whom discontinuation is clinically appropriate.

Participants will be recruited across four university psychiatric units in Italy, with methodological support provided by a fifth academic centre. General practitioners and other specialists will be involved in identifying potentially eligible individuals. After providing informed consent and completing baseline assessments, participants will be randomized to either a linear or a hyperbolic tapering strategy according to predefined schedules tailored to the starting dose. Follow-up assessments will be conducted regularly over a 36-week period using validated measures of depressive symptoms, anxiety symptoms, withdrawal symptoms, quality of life, and social functioning. Outcome assessors and the biostatistician will remain blinded to treatment allocation.

The primary outcome is the proportion of participants who fail to discontinue the AD within the predefined tapering schedule or who re-initiate AD treatment during the 16 weeks following discontinuation. Secondary outcomes include non-adherence to the tapering schedule, occurrence and severity of withdrawal symptoms, clinical relapse, early study discontinuation, and use of predefined rescue strategies. This study is designed to generate pragmatic evidence on the comparative effectiveness, feasibility, safety, and cost-effectiveness of hyperbolic versus linear tapering strategies in everyday clinical practice.

Inclusion criteria

• 18 years old or above;
• diagnosed with a depressive disorder, single episode (ICD-11, 6A70) or recurrent (ICD-11, 6A71);
• currently taking a selective serotonin reuptake inhibitor (SSRI), serotonin and norepinephrine reuptake inhibitor (SNRI), tricyclic antidepressant (TCA), or vortioxetine for the treatment of depression;
• the current AD has been taken for at least 6 months;
• the current AD has been on a stable dose over the last 2 months;
• a score ≤9 on the PHQ-9 and ≤5 on the GAD-7 at study enrolment;
• DSM-5-TR criteria for a depressive episode are not met at the time of recruitment;
• no clinical evidence of moderate-to-severe symptoms in the last 6 months, as assessed by the recruiting clinician;
• discontinuing the AD is clinically indicated by the recruiting clinician, and agreed to by the participant in a shared decision-making process;
• uncertainty about which discontinuation strategy would be best for the participant;
• the participant is willing to sign the informed consent to participate to the study.

Exclusion criteria

• comorbid schizophrenia-spectrum disorders, bipolar disorder, or dementia, as formally diagnosed by a psychiatrist, neurologist, geriatrician, or other specialists;
• current treatment with more than one AD at therapeutic doses;
• current treatment with ADs of other classes (e.g., mirtazapine, agomelatine, bupropion, for which the evidence on the risk of withdrawal is unclear), alone or in combination with other ADs;
• conditions or medications that contraindicate the use of any AD according to the Summary of Product Characteristics of included ADs (synthetically reported in Table 1) (e.g., current symptoms of mania);
• current treatment with benzodiazepines above the dose of 2.5 mg equivalents of lorazepam per day (corresponding to clonazepam 1.2 mg/day; alprazolam 1.2 mg/day; diazepam 19 mg/day);
• pregnancy or willingness to become pregnant.