Prozac (Fluoxetine)

Prozac came out in 1987. Almost forty years later it’s still on the short list I reach for, and not because of nostalgia. Fluoxetine has a profile that nothing else in the SSRI class quite matches, and most of what makes it useful comes down to one boring pharmacokinetic fact: it leaves the body slowly.

The drug itself has a half-life of four to six days. Its active metabolite, norfluoxetine, sticks around for nine to sixteen days. Compare that to sertraline at roughly a day, or paroxetine at less than that. What this means in practice is that Prozac essentially tapers itself. You can stop it cold on a Friday and the drug levels drift down over the next two or three weeks while the patient barely notices. With Paxil you’d be in withdrawal by Monday morning, sweating through a shirt and feeling like someone left a tuning fork in your skull.

That single property changes how I prescribe it. It’s the SSRI I pick when I’m reasonably sure the patient is going to be lousy about taking pills on schedule, when they travel a lot, when they’ve already had discontinuation hell from another SSRI, or when I genuinely don’t know if we’ll be sticking with this medication for the long haul.

Why the long half-life keeps mattering

Discontinuation syndrome from short-half-life SSRIs is real and frequently underestimated. Paxil is the worst offender. Effexor (technically an SNRI but same mechanism issue) is right there with it. Patients describe brain zaps, vertigo, GI distress, irritability that goes from zero to wanting-to-fight-a-stranger in about ten seconds. The drug isn’t dangerous in withdrawal the way benzos are, but it’s miserable, and it traps people in medications they’d rather be off.

Prozac essentially doesn’t do this. I’ve had patients miss a week of doses on vacation and feel completely fine. I’ve had patients stop it abruptly against my advice, and the worst they reported was feeling somewhat flatter for a few weeks. The mechanism is just the drug staying in the system long enough to taper itself.

There’s a clinical trick that comes out of this. When I need to get a patient off Paxil or Effexor and they keep failing the taper, I’ll sometimes cross them over to a single dose of fluoxetine, let the Prozac wash in while we drop the offending drug, and then taper the Prozac slowly over months. The long half-life smooths the transition. It’s not novel, plenty of psychiatrists do it, but it’s a useful maneuver when somebody’s been stuck on Paxil for years and can’t get off without coming apart.

The activating profile in week one

Prozac is more activating than most of the SSRIs. Sertraline can be activating too, but fluoxetine is the one most likely to make an already anxious patient feel worse before they feel better. Jittery. Wired. Sleep gets weird for the first ten days. Some people describe it as feeling like they had too much coffee for a week straight.

I had a woman come in last spring, mid-30s, mostly generalized anxiety with a depressive overlay. She’d been doing alright on no medication for years but a job change had tipped her over. I started her on 10mg of Prozac because she was the kind of anxious patient I’d usually start lower with, and she still called me on day four convinced something was very wrong. Heart pounding at night, intrusive thoughts, couldn’t sit still. None of this was the depression worsening. It was just the drug ramping up.

We held the dose, added a small amount of hydroxyzine for sleep, and by day twelve she’d evened out. By week six she was the most functional she’d been in two years. But if I hadn’t warned her in advance that week one might feel rough and that we’d ride it out, she would have quit by day five and added Prozac to her list of medications that “didn’t work.”

The drug working and the drug feeling like it’s working are two different things, separated by about four weeks.

For pure anxious patients with no depressive component, I tend to lean toward escitalopram or sertraline instead, just to avoid that activating bump. But Prozac for an anxious-depressed patient still earns its place, you just have to set expectations honestly. Patients tolerate a rough week far better when they’ve been told to expect a rough week.

One of the few SSRIs we actually have for kids

This is where Prozac really has no competition. It’s the only SSRI with FDA approval for both pediatric depression (age 8 and up) and pediatric OCD (age 7 and up). Lexapro picked up an adolescent depression indication later, but for the under-12 crowd, fluoxetine is essentially the only thing with the regulatory paperwork.

The data behind that approval is actually decent, which isn’t true of every pediatric psychiatric indication. The TADS trial in the early 2000s looked at fluoxetine versus CBT versus the combination versus placebo in adolescent depression, and the combination of medication plus therapy did best. Fluoxetine alone outperformed therapy alone in the acute phase, which surprised a lot of people at the time.

In practice I’m conservative with pediatric prescribing. I don’t put kids on SSRIs casually. But when a child genuinely needs a medication, fluoxetine is what I default to. The long half-life helps here too, because pediatric adherence is even more inconsistent than adult adherence, and a missed dose or two with Prozac doesn’t matter. With a short-half-life agent, the kid would be cycling through withdrawal symptoms every weekend.

The interaction landscape

Prozac and norfluoxetine are both strong inhibitors of CYP2D6 and moderate inhibitors of 2C19 and 3A4. This matters more than people realize.

The classic problem is co-prescribing with tamoxifen, which depends on CYP2D6 to become its active metabolite. Block 2D6 and you reduce the cancer drug’s efficacy. Oncologists generally know this. Some psychiatrists don’t. I always check the medication list and avoid Prozac in any breast cancer patient on tamoxifen.

The other one is opioids. Codeine and tramadol both require 2D6 for activation, so Prozac essentially neuters their analgesic effect. I had a guy last winter, late 40s, on 40mg of fluoxetine for two years, who came back from a kidney stone in the ER absolutely furious that the tramadol they’d sent him home with did nothing for the flank pain. He’d swallowed six pills over twelve hours and was still pacing the kitchen at 3 AM. Nobody in the ER had checked his med list against the metabolism pathway. That’s the pattern. The patient gets labeled drug-seeking or pain-tolerant when the actual problem is the SSRI is sitting on the enzyme that’s supposed to turn the prodrug into something useful.

It also raises levels of tricyclics, certain antipsychotics (Risperdal especially), and some beta blockers. If I’m adding Prozac to a patient already on a complicated regimen, I check every drug they’re taking against the 2D6 and 2C19 substrate lists. Not because anything catastrophic usually happens, but because mild toxicity is easy to miss and easy to prevent.

Why it’s still around

New psychiatric drugs come out every few years, and most of them are slightly tweaked versions of older ones with worse data and a bigger marketing budget. Prozac has been out long enough that the patents are decades expired, the generic costs about four dollars a month, and the long-term safety profile is as well-characterized as any psychiatric drug in existence.

It isn’t the right SSRI for every patient. Pure anxiety presentations often do better on escitalopram. Patients who need rapid response shouldn’t start with anything in this class. Anyone on tamoxifen or chronic opioids needs a different option. But for a fair slice of the depressed and OCD and anxious-depressed patients I see, Prozac at 20mg remains a reasonable first move that I rarely regret making.

The pills are green and white, the bottle costs less than a sandwich, and the drug has been quietly fixing people since the Reagan administration. There are worse things to start with.