Depression Treatment

Depression doesn’t feel like sadness. Sadness has a shape and a target and it moves through you. Depression sits down on top of everything and shrinks the world to the size of your bed. The texts you should answer, the meal you should eat, the shower you keep postponing… all of it pulls equal weight, which is none.

The biology is real. Serotonin, dopamine, norepinephrine, and the broader reward circuitry that runs on them go sluggish in ways you can measure on a PET scan. That’s why you can know you should get up and still not be able to make your body do it. The wiring isn’t delivering current at the right voltage. Calling that laziness is like calling a flat tire bad driving.

The cruelest part of the illness is that it lies to you about itself. It tells you nothing will work, you’re too far gone, treatment is for people whose problems are smaller than yours. Those thoughts feel like facts when you’re inside them. They’re symptoms. You can act before your brain gives you permission to.

The medication ladder, in the order I actually use it

Most people who land in my office with depression have already tried one antidepressant and decided antidepressants don’t work for them. Usually what happened is they got Lexapro 10mg, felt nauseous and weirdly tired for two weeks, didn’t feel obviously different by week four, and stopped picking up the refill. The STAR*D trial, still the biggest real-world depression study we have, showed that only about a third of people respond meaningfully to the first med they try. Most need two or three trials before something clicks. Knowing that going in changes the experience of the first one not working.

First line is almost always an SSRI. Zoloft, Lexapro, Prozac, Celexa. Cheap, studied to death, tolerable for most people. Starting doses are low (sertraline 25 to 50mg, escitalopram 5 to 10mg) and we titrate over a few weeks. The thing nobody tells patients clearly enough is that side effects peak in weeks one and two, and antidepressant effect usually doesn’t show up until weeks four to six. The window where it feels like a bad trade is real. That window is where most people quit.

Second line for me is usually an SNRI. Effexor (venlafaxine) at 75 to 225mg and Cymbalta (duloxetine) at 30 to 60mg add a norepinephrine kick to the serotonin work, which helps if there’s heavy fatigue or co-occurring chronic pain. Venlafaxine has a brutal discontinuation syndrome if you skip doses, which is worth knowing before you start it rather than after.

Then come the atypicals. Wellbutrin (bupropion) is its own category… acts on dopamine and norepinephrine, doesn’t tank libido, doesn’t cause weight gain, often hands people some energy back. Trintellix (vortioxetine) is newer, gentler on sexual side effects, sometimes helps cognitive fog more than the older drugs. Remeron (mirtazapine) is sedating and increases appetite, which sounds like a problem until you meet the patient who hasn’t slept or eaten in three weeks.

TCAs like nortriptyline and amitriptyline are older. They work, sometimes better than the newer drugs, but the side effect profile is heavier and the overdose risk is higher. We still reach for them when newer agents haven’t worked. MAOIs (Nardil, Parnate) are end of the line. Effective. Require a strict tyramine-restricted diet to avoid hypertensive crisis, so the patient has to be organized enough to follow it for years.

If you’ve tried one antidepressant and it didn’t work, you’ve sampled one rung of a ladder with eight rungs on it.

Behavioral activation, sleep, and the boring stuff

CBT has the most evidence for depression, and the piece inside CBT that does the heavy lifting is behavioral activation. The principle is counterintuitive and people hate it: action comes before motivation, not after. You don’t wait until you feel like going for a walk. You schedule the walk for 9 AM Tuesday, you do it badly, you do it again Wednesday, and your mood follows the behavior a few weeks down the line. Patients hate this because it asks them to do the thing while still feeling like garbage. The data on behavioral activation has been holding up in trials since the 90s, sometimes matching the full CBT package on its own.

Exercise. I know how that lands when you can’t get out of bed. Telling a depressed person to go for a run feels insulting and I get it. The data is annoying though. Aerobic exercise, thirty minutes, three to five times a week, shows up in meta-analyses with effect sizes comparable to a mild antidepressant for mild-to-moderate depression. Doesn’t replace medication for severe cases. Works as an add-on for almost everybody. Mechanism is probably some combination of BDNF, inflammation reduction, and behavioral activation hiding in a tracksuit.

Sleep is the one I push hardest. Depression breaks sleep and broken sleep deepens depression, and you’re stuck in the loop. Fixing sleep doesn’t fix depression by itself, but you can’t fix depression on top of unfixed sleep. Same bedtime every night, screens out of the bedroom, no alcohol in the four hours before bed, no scrolling at 2 AM. Boring. Effective.

When the first drug only half-works: augmentation

This is where a lot of community psychiatry stops and shouldn’t. If you’ve been on a reasonable dose of an SSRI for six to eight weeks and you’re partially better but not really better, the next move usually isn’t switching to a different SSRI. It’s augmentation. Add a second agent that hits a different system instead of swapping the first one out.

The cleanest augmentation move is adding bupropion 150 to 300mg to an SSRI. Different neurochemical system, often reverses the sexual side effects of the SSRI, frequently moves the needle for people stuck at partial response. I do this constantly.

Lithium augmentation is old, underused, and quietly impressive. We’re talking 300 to 600mg added to an antidepressant, not bipolar doses. The data goes back to the 80s. It’s cheap. It requires blood monitoring (lithium level, thyroid panel, kidney function every six months or so), which is why a lot of prescribers won’t touch it, but for the right patient it pulls people across the finish line.

Atypical antipsychotic augmentation is the newer answer for treatment-resistant cases. Abilify (aripiprazole) 2 to 5mg, Rexulti (brexpiprazole) 1 to 2mg, and Seroquel XR 150 to 300mg are FDA-approved as add-ons to antidepressants. They work. They also carry metabolic risk, weight gain, and movement-side-effect risk at higher doses, so we use the lowest effective dose and we check labs.

Treatment-resistant depression and the heavy artillery

The technical definition: failure to respond to two adequate trials of antidepressants from different classes. About a third of people with depression land here at some point. That’s not rare. That’s a third of the patients on my schedule any given week.

Ketamine changed the shape of this field. IV ketamine and the FDA-approved nasal spray (Spravato) can produce antidepressant effects within hours to days rather than weeks. It hits NMDA receptors, triggers glutamate release, drives rapid synaptic plasticity. Effects fade without repeat dosing, so it’s a series of infusions and then a maintenance protocol. Expensive, requires monitoring, and for the right patient it’s been the difference between functioning and not.

TMS (transcranial magnetic stimulation) is six weeks of daily outpatient sessions, around thirty minutes each, where a magnetic coil pulses over the left prefrontal cortex. No anesthesia, no cognitive side effects, no time off work. Response rates around 50 to 60% in treatment-resistant patients. Covered by most insurance now, which was not the case five years ago.

ECT (electroconvulsive therapy) has the worst PR of any treatment in psychiatry and is the most effective one we have. 70 to 80% remission rates for severe, treatment-resistant, or psychotic depression. The version we run now is nothing like the movies. Brief general anesthesia, muscle relaxant so the body doesn’t convulse, a very short controlled seizure under EEG monitoring. Main side effect is short-term memory loss around the treatment window, which usually resolves over a few weeks. I had a patient a while back, software engineer in his forties, two years of depression, four med trials and a full CBT course had done nothing. He’d been on short-term disability for nine months and was about to lose his job. Course of nine ECT sessions, twice a week, and by the end he was answering work emails on his phone in the recovery room. He went back to work the month after he finished and stayed well on a maintenance SSRI plus monthly check-ins.

If you’ve been depressed long enough that you’ve stopped believing treatment can help, that’s the depression talking, and it’s the most common reason patients stay sick for years. The ladder underneath you is real, and most of the rungs are ones nobody ever told you exist.