A patient with major depression is started on sertraline, fifty milligrams daily. Let me walk through what happens, week by week, at the molecular and circuit level.
Day 1. The medication blocks the serotonin reuptake transporter (SERT) at presynaptic terminals throughout the brain. Synaptic serotonin rises within hours. But the patient does not feel better. In fact, they may feel worse — anxious, nauseated, with disturbed sleep, possibly with some increase in suicidal ideation in adolescents and young adults. The reason: acute serotonin elevation is dysregulating before it is regulating.
Week 2. The elevated synaptic serotonin has activated presynaptic 5-HT1A autoreceptors on the raphe nuclei. Autoreceptors are negative-feedback receptors — they sense the level of their own neurotransmitter and adjust release accordingly. With autoreceptors strongly activated, raphe firing actually decreases, partially canceling out the SSRI's effect. The patient is in a transitional state. Some symptoms may be improving; others are not. The clinician is reassuring them to stick with it.
Week 4 to 6. The autoreceptors have downregulated — there are simply fewer of them, because they have been chronically activated and the cell has internalized them. This is the same phenomenon that produces opioid tolerance, but here it is therapeutic. With the autoreceptor brake released, serotonin release rises further. Postsynaptic 5-HT receptors also adapt. Hippocampal neurogenesis begins to increase. Synaptic plasticity in mood-regulating circuits — including connections between prefrontal cortex and amygdala — begins to shift. The patient begins to feel mood lift, though not all at once.
Week 8. The patient reports clear improvement. Mood is steadier. Sleep is better. Concentration has returned. The default mode network, which was stuck in ruminative self-referential loops, has begun to relax its hold. Amygdala reactivity to negative stimuli has decreased on functional imaging. Working memory and cognitive control have returned. The patient describes feeling more like themselves.
This time course is not a delay before the drug starts working. It is the drug working from day one, but at the level of slow plasticity rather than fast signaling. The patient feels better when the circuits have remodeled, not when the synapses first respond. The metabotropic nature of most serotonin receptors and the slow plasticity changes downstream of receptor activation are what produce this characteristic six-to-eight-week timeline.
This is why telling patients "give it six weeks" is mechanistically true, not just clinically pragmatic. The synapses respond on day one. The mood lifts when the circuits reorganize. Patients who stop the medication after two or three weeks because it isn't working are stopping before the actual therapeutic mechanism has had time to occur.
For partial responders, augmentation strategies that target different mechanisms can help. Bupropion (adds dopamine and norepinephrine reuptake blockade). Aripiprazole at low dose (D2 partial agonism). Lithium (multiple mechanisms including BDNF). Thyroid augmentation. Each adds a different molecular intervention that may complete the response the SSRI alone could not.
For non-responders, ketamine and esketamine offer a different mechanism entirely — rapid NMDA-mediated plasticity that works within hours and converges on the same network changes via different cellular routes. ECT remains highly effective for treatment-resistant depression, working through plasticity at large scale.
Hold the timeline. The SSRI starts working on day one. The patient feels it after weeks of slow circuit remodeling. The reassurance you offer at the two-week visit — this is normal, the changes that matter take time — is biologically accurate.