The patient has taken the same antidepressant for two years. The depression responded. Follow-up was appropriate. At the most recent review, physician and patient agree to attempt a gradual reduction. The plan is familiar: reduce the dose and evaluate.
Ten days later, the patient reports anxiety, insomnia, and irritability. The physician recognizes the picture. It is read as a sign that the underlying depression had not fully resolved. The full dose is restarted. Symptoms settle within days.
What happened has another possible reading.
The symptoms the patient reported at ten days are consistent with the nervous system’s response to antidepressant withdrawal: autonomic activation, insomnia, rebound anxiety. They are not necessarily the return of depression. They may be the echo of the medication leaving.
The difference between the two readings is not academic. One leads to reinstating treatment and a conversation about chronicity. The other leads to reviewing the speed and shape of tapering.
Guy and colleagues document that patients who experience antidepressant withdrawal symptoms are often dismissed, misdiagnosed as relapse, or attributed a new medical condition. The pattern has a clinically understandable logic: withdrawal symptoms and depressive relapse can overlap in the patient’s subjective report. Both can produce anxiety, insomnia, and dysphoria. Distinction requires a criterion the standard follow-up model does not always call forward: speed of onset and somatic texture.
Withdrawal symptoms often appear quickly, within days, and may include marked somatic components: dizziness, electric sensations, sweating, gastrointestinal symptoms. Depressive relapse usually has another temporality: weeks, with the cognitive and affective profile the physician recognizes from prior episodes. That difference in texture and speed is the available clinical criterion before the picture is assigned to one cause or the other.
The mechanism explaining why a standard reduction can produce this picture lies in the relation between dose and receptor occupancy. Horowitz and Taylor, in the Maudsley Deprescribing Guidelines, describe that relation as hyperbolic rather than linear. Early milligram reductions may produce modest occupancy changes, while the final milligrams can produce disproportionately large changes. Reducing from 20 mg to 10 mg is not equivalent to reducing from 10 mg to zero. The nervous system does not experience it as half the change.
That asymmetry underlies hyperbolic tapering: progressively smaller reductions in the final stages, at speeds determined by the patient’s response rather than by a fixed calendar. A linear taper was not designed around that asymmetry.
None of this implies that antidepressants are the problem, that the patient can or should discontinue them, or that depression does not require pharmacological treatment when indicated. It means that discontinuation requires the same clinical precision as initiation: reading of the patient profile, pace adjusted to response, and a criterion for distinguishing what the nervous system is expressing from what depression would express if it returned.
Constitutional reading is pertinent here. Two patients tapering the same medication at the same speed may have different trajectories. A Vata profile may show high autonomic reactivity: intense withdrawal symptoms after modest changes, rapid onset, irregular recovery between steps. A Pitta profile may hold stability at first and then decompensate when reduction enters the steeper final part of the pharmacological curve. A Kapha profile may tolerate reductions that would destabilize another patient, but may need more time between steps for the system to assimilate change. This variability is not unknowable. It can be read before the symptom returns to the consultation.
Transcendental Meditation is discussed here only as modulation of activation. Walton and colleagues associate regular practice with reductions in stress-related neuroendocrine activation, including cortisol. During antidepressant tapering, such modulation does not replace the tapering protocol and does not treat depression. It may reduce the amplitude of autonomic activation on which withdrawal acts, making the reduction window more stable for a patient already in that process.
The physician who can distinguish the echo of the medication from the signal of depression does not make fewer decisions. They make decisions from a more precise reading. The speed of taper, the size of each reduction, the patient profile that determines how much space the nervous system needs between steps: that is clinical work, not an automatic protocol.
The clinician who recognizes one of their own patients in this article already has the right question.
It is not whether the patient can stop the antidepressant. It is at what speed, with which criterion of distinction, and with what reading of the profile that determines how this specific nervous system responds to change.