The conversation about antidepressants and therapy almost never begins where it should. It begins, in my experience of the room, with the question of which one is the right one, as though backing one means doubting the other. Most accounts of treatment for depression keep that framing in place. What they leave out is that medication and therapy work through different mechanisms, on different timelines, and that the practical question is rarely which to pick. It is which combination, in what order, for whom.
What is missing from the usual conversation
The public debate about depression treatment tends to collapse into a competition between pills and talking. Whichever side a person is more familiar with shapes the question they bring to the GP or psychologist. Are antidepressants safe? Does therapy do real work, or is it just a place to vent? What I notice, again and again, is that these questions almost never sit at the right altitude. They are framed as if the answer were going to settle the matter, when the more useful question is one most people have not been given the language to ask.
The evidence base for both interventions is substantial, though the way the evidence has been communicated has not always been helpful. A 2018 network meta-analysis pooled 522 randomised controlled trials and found all 21 antidepressants studied to be more effective than placebo at the acute treatment stage, with effect sizes that varied by drug but were clinically meaningful across the class (Cipriani et al., 2018). The evidence for psychological therapies is similarly strong, and a more recent meta-analysis of treatment durability found that the gains from psychotherapy, and the combination of psychotherapy and medication, tend to hold up well over time, while medication alone is more vulnerable to relapse after stopping (Voderholzer et al., 2024). I sit with that comparison carefully. It does not mean medication is less useful. It means the two do different kinds of work and tend to leave different things behind.
Different mechanisms, different timelines
What is most useful to understand, if the conversation is going to land somewhere honest, is that medication and therapy operate on different parts of the system, on different schedules. Antidepressants alter the chemical signalling environment of the brain over weeks. The effect on mood and energy is usually gradual, often noticeable in some form by the fourth to sixth week, and the work of finding the right drug and dose can take longer than that. The mechanism is biochemical, and it acts on the state the depressive system has settled into.
Therapy works on something different. It addresses the patterns of thought, behaviour, and relationship that have either contributed to the depression or that the depression has reinforced. Sessions take time precisely because that work is slow. New behavioural patterns require repeated practice. New ways of relating to thoughts require many small applications across daily life. The gains tend to accumulate gradually, and they tend to stay when treatment ends, because the change has been built into the person's way of moving through the world rather than added to it from outside.
For someone whose depression is severe enough that engaging with therapy is not yet possible, who cannot get out of bed or take in what is being said, medication often makes the work of therapy more available by lifting the floor. For someone whose depression is closely tied to identifiable life patterns, therapy may do work that medication alone cannot. For many people, both are useful at once. The 2024 meta-analysis of enduring effects found that the combination of psychotherapy and antidepressants produced the most reliable benefit at long-term follow-up (Voderholzer et al., 2024).
What the current guidance says, and what it does not
The clearest framing of the decision sits in the current English-language guidance (National Institute for Health and Care Excellence, 2022a). For less severe depression, guided self-help and brief individual therapy sit at the top of the list, and medication is not the first step. For more severe depression, individual cognitive behavioural therapy is recommended, often in combination with medication. For chronic or recurrent depression, the combination is preferred, and longer-term medication is often discussed. The Australian and New Zealand college guidelines reach broadly similar conclusions, with somewhat more emphasis on the role of medication in moderate-to-severe presentations and on the importance of treating to remission rather than to partial response (Malhi et al., 2021).
What the guidance shares is that the decision is contextual. The same depression in two different lives can call for two different starting places. A person who has tried medication before and tolerated it well will weigh things differently than someone for whom the side-effect profile of an SSRI was difficult. A person with strong family history may make different decisions than someone for whom this is a first episode after a defined precipitant. What the guidance does not say, and what I think is worth saying out loud, is that the decision is not philosophical. It is clinical, and it belongs in a conversation with a GP or psychiatrist, alongside whatever psychological work may be running in parallel.
Stopping medication is its own conversation
Coming off antidepressants is the area that has shifted most in the recent guidance, and the area I find people are most often unprepared for. A systematic review found that for people in remission after maintenance antidepressant treatment, the relapse rate after discontinuation was substantial, with a meaningful fraction of people experiencing return of symptoms within several months (Kato et al., 2021). The current NICE guidance on safe prescribing now includes detailed recommendations on gradual tapering and on recognising withdrawal symptoms, particularly with shorter-half-life antidepressants where stopping abruptly produces a constellation of symptoms that can be mistaken for relapse (National Institute for Health and Care Excellence, 2022b).
What I have come to hold carefully, in conversations about this, is that the decision to stop medication is not a decision to make alone, in a low moment, or to defer on the basis of a single appointment. It belongs with the prescriber. For people who have done therapy alongside medication, there is often more flexibility, because the patterns the therapy worked on may have shifted enough that the medication is no longer doing the same load-bearing work. The timing, the rate of taper, and the support around the taper all matter, and the guidance is now clear that this is a process to manage with care, not a decision to make on a single visit.
Two framings worth putting down
Two framings I would set down, gently. The first is that medication is a sign of weakness, or of having failed to try hard enough. The chemical signalling changes documented in depression are real, and treating them is not an admission of anything. The second is that therapy is a luxury, or a soft alternative when the "real" treatment is medication. The evidence base for psychological therapies is among the strongest in the field, and the durability of those gains is one of the clearer findings in the modern literature.
What stays with me, after these conversations, is how often the right question is rarely which to pick. It is what is happening in your life, what your particular shape of depression is doing, and what combination of supports is likely to meet you where you are. That conversation takes time, and it tends not to fit into a fifteen-minute appointment. It can begin with a GP, or with a psychologist, or with both at once. What matters most is that the person sitting with the decision is not left to make it on their own.
For how we work with depression, what sessions cost, and how to begin, see Working with depression and mood in adults.
