Key moments, in order
Depression is probably the biggest umbrella term in psychiatry. Depression is a symptom, not one diagnosis: perhaps the biggest mistake early clinicians make is hearing depressive symptoms and concluding major depression.
Depression of personality should not be overlooked. It looks like, and even feels to the patient like, major depression, but the life course and onset show it is not the same thing.
Depression of the personality looks and even feels like major depression to the patient, but the life course and onset tell them apart. Serotonergic drugs give it little benefit, and not the kind you see in a classic first episode, so reaching for the depression algorithm misses the diagnosis.
The classic case: a 40-year-old with her first episode, low mood all day every day, low motivation, early awakening and leaden paralysis. Put her on a little sertraline and she will start responding right away.
The primary mimic is persistent or major depression. Several core criteria resemble a depressive episode, and not everyone knows to disclose, or is comfortable disclosing, their traumas.
For unipolar depression, SSRIs come first: low risk, higher efficacy, and a side effect profile that is not too bad.
Somebody with a bipolar disorder can become agitated or worse on a serotonin reuptake inhibitor, so be careful.
We generally prefer sertraline. There is more room for titration (four steps up to 100 and four up to 200, against two up to 10 and two up to 20), and it tends to have fewer interactions.
Our four first-line antidepressants are escitalopram, sertraline, mirtazapine and bupropion, chosen by patient profile, and we very rarely deviate from them.
It is one of our early antidepressants, but we push sertraline or escitalopram first and think of bupropion more as an augmenting agent. It is fine to start with for a patient who does not want the SSRI side effects of a little weight gain or sexual problems.
A great choice when sleep is a big issue and the patient is underweight or not concerned about weight, and almost everyone with a true major depressive disorder has a sleep problem. The weight gain risk is why it is used rarely.
It is a great augmenting choice for low energy and for anorgasmia from SSRIs, and it seems to help anorgasmia even without a mood disorder.
We do not use it first line because enough people get jitteriness, insomnia or an appetite drop they do not want. For patients who are not too anxious, sleep fine and eat too much, it becomes a very nice choice.
As a rule of thumb, an adequate trial is at least three quarters of the FDA maximum dose. Side effects are the reason to switch before you get there.
Many patients are scared to push the dose. Find out why and talk about it: there can be good reasons, and a little education may make them more comfortable.
If an adequately dosed medication is not working, you either switch or augment, and the research does not show one approach is better. We prefer to augment, because one medication usually does not treat all the ill effects of depression right away.
50 mg is the minimum effective dose for the average patient.
One instinct is to switch first to keep the medication count down, and augment when there is a partial effect but side effects make you not want to push the dose higher.
Much of standard practice comes from a landmark study that tested switching or augmenting after non-response, with a psychotherapy arm as well.
How fast you go depends on severity and how soon you can see the patient. For moderate depression with a four-week follow-up, aim to have them at 50 mg by the next visit.
Find out what they have taken before: that tells you roughly what dose they respond to. If they have never had one, start low, go slow, and follow up soon.
A common start is 25 mg for one week, then 50 mg if tolerated, staying there for a month before judging the effect. Prescribe 50 mg tablets and have the patient break the first four in half, and email the instructions after the visit so they do not have to remember everything.
If there is no rush, start 2.5 mg for one week, then 5 mg. Many patients have a harder time going from 5 straight to 10.
Another approach gets patients to 10 mg fairly fast, partly because 5 mg tablets can be hard to get from local pharmacies.
In unipolar depression the usual side effects are nausea, diarrhea and some stomach cramping for three to five days, coming back briefly with each dose increase. For someone very sensitive to nausea, you could consider ondansetron early on, but most people are fine.
Patients with a little bipolarity, even if they will never have real bipolar disorder, or with very high anxiety can react to fast titration.
Use the PHQ-9 to track symptoms rather than as a screener. It covers a lot of the symptoms and is shorter and easier than the HAM-D or BDI, so you can do it in a visit.
Scale numbers give a baseline and can show patients they have improved when their subjective sense lags.
Scales also support honest conversations about stalled treatment, for example telling a patient a year in with no change that it may be time to try something new.
We do not find SNRIs more effective than SSRIs (at most a very marginal improvement) and they have slightly more side effects, even if just duloxetine's slight blood pressure rise. Many patients have a really tough time getting off them, and venlafaxine's withdrawal is strong and uncomfortable enough that we avoid it entirely.
Use an SNRI first line only when a comorbidity gives it a second job: a diagnosed migraine disorder or chronic pain with a good medical explanation that is not adequately treated, or fibromyalgia. Then you might as well use a dual-use medication. In some cases true ADHD counts, but not apparent ADHD. They may also be a little better for hot flashes.
Stopping it abruptly can produce a withdrawal so bad it looks like a medical illness to the patient and caretakers, even with a little delirium.
You do not get the norepinephrine effect until a relatively high dose, which undercuts the point of choosing it.
Formative early experiences change each clinician's approach, for good and bad. A few early patients who could not tolerate escitalopram, for example, can make you go slower.
Do not be afraid to switch to another SSRI. The mechanisms are not totally the same, and someone can fail one and respond to another.
It lasts so long that it is great for patients who forget or skip doses: give a higher dose, and the long half-life still gives an effective serum level when they skip a few doses a week. Start-and-stop patients can get used to it and take it a few times a week (not as a PRN, just whenever they remember).
We tend not to use it. It is anticholinergic and makes people sleepy, it causes weight gain, and it is probably the worst SSRI for pregnancy. Even where it helps, as in anxiety, these hold it back.
Its reputation for anxiety probably comes from sedation, not from being better than the others. Anxiety and rumination tend to get worse at night, so anything that makes people sleepy helps, and that does not have to be Paxil: a little gabapentin can do it.
We reserve it pretty much for OCD, and we do not think it is that special for OCD, even though some people do.
There is little point to it: escitalopram is the active enantiomer. The one thing citalopram does better is prolong QTc. On the FDA data, escitalopram up to 20 mg did not significantly raise QTc.
Whether it is activating or sedating varies from patient to patient. Take it at night if it makes you sleepy, in the daytime if it activates you.
Some patients tolerate citalopram better than escitalopram, against what we are taught. One patient felt very uncomfortable on escitalopram with bad sexual side effects, then did great on citalopram.
When the same drug works differently at different times, that is usually because a lot more is going on in the person's life than one medication.
Do not let expectations of how a drug should behave override what the patient reports. We do not understand as much as we think.
We are willing to give a sleeper on the first visit because the sleepers we start are never the habit-forming ones: trazodone, gabapentin or hydroxyzine, which support deep, healthy sleep.
For someone on an adequately dosed SSRI or SNRI, add-on options include bupropion, mirtazapine and buspirone, and then the second-generation antipsychotics, which are very popular in the US but high risk.
As antidepressant augmenters they work, and not just because people were misdiagnosed with bipolar disorder: they have many unique receptor effects an SSRI lacks. They are popular in the US but high risk.
Many antipsychotics are depressogenic and are not great long term, with bad metabolic effects and risks of EPS and tardive dyskinesia.
They are called antipsychotics because of a historical or FDA-approved use in psychotic disorders, but their receptor effects are so wide that the name is almost a misnomer. They are extremely effective, and a strong medication is strong in its side effects too, so use them with caution when lower-risk options exist.
A perfectly fine medication, but expensive and not particularly more effective than others, the traditional SSRIs included (though we have not used it a lot). For cost reasons we would not go to it first, but someone who has a hard time tolerating medications may tolerate it better than the others, so it is worth a try.
Do not underdose it. It is often written off because it was never adequately dosed. Titrate fairly aggressively and do not be afraid to take it to the max, but do not expect the moon and the stars: it gives a modest benefit. We have seen it work better at the higher dose range.
Give it at least twice a day. The half-life is short, and you need an adequate serum level.
A typical climb: 15 and 15, then 15 and 25, then 25 and 25, which puts you about 10 mg under the usual FDA daily maximum.
Behavioral activation, exercise, light exposure and psychotherapy are the most important 'medications' for depression.
When therapy is not possible, at least have them work through a workbook. Feeling Good or Feeling Great is a very nice one for the classic melancholic patient, and The Happiness Trap is the other we always recommend.
Providers often try medication after medication when people need to be doing something: a purpose in life, and small goals they achieve.
The medication is not the treatment; it is part of it. It is not a happy pill. What it probably does is make it easier for you to change.
Prescribe from a gatekeeper stance (I give you the meds, the meds treat your thing) and you will get much worse outcomes.
They have many more side effects: dangerous in high doses, QTc prolongation, anticholinergic effects like dry mouth and urinary retention, and a lot of sedation.
The problem with MAOIs is the risk of a hypertensive crisis with foods containing tyramine, and many of the best foods (the ones at a wine party) have some.
We reserve them for patients who have failed even the SSRI plus second-generation antipsychotic combination.
One view we have discussed holds that MAOIs are, point blank, the most effective antidepressants, and that instead of switching medications again and again, these patients deserve an MAOI trial earlier in the algorithm.
They have no independent efficacy in depression, even though they are given a lot, including early in SSRI starts.
For depression it is symptomatic treatment without much evidence of long-term benefit. Anyone feels better on a stimulant, and takes it away and they feel more depressed.
The main exception in depression is geriatric depression with apathy, avolition and low energy that probably reflects something neurocognitive, such as early dementia. A little stimulant is okay there.
Prevention is the best strategy: avoid weight-gain-inducing medications.
For SSRI sexual side effects, buspirone and bupropion are the most helpful add-ons. Otherwise your choices are pretty much lowering the dose or switching the medication.
For bruxism, the safest step is probably a mouth guard. A sleeper is something you could consider.
Balance giving information against the nocebo effect: say the nausea is going to be awful, and the nausea will be awful.
Some side effects, like nausea, GI upset and feeling a little uncomfortable, are as bad as they will be at the first dose. They level out in about three days, and people usually feel a little normal by about a week.
On average, patients may gain about five pounds over long-term treatment, and even patients working hard at it find those five pounds really tough to lose.
Treat to remission, then wait six months at minimum and preferably a year before making changes. That is prudence, not data: even after remission, an early stop or decrease can bring the depression back. Let the patient guide it, and ask what they have changed about themselves and their life since starting.
Probably half of depressions are single episodes, and half of people with depression will have a recurrence if they come off medication.
When people get worse on antidepressants, it is mostly, in our view, because they have a more bipolar-type condition and develop antidepressant-related agitation and sleep problems. That is probably why we see more suicidal ideation in people under 25 on antidepressants.
A good rapport and a patient who trusts you is probably infinitely more important than the nitty-gritty of which medication you pick.
Many patients are trained to expect you to throw a medication at them. Unless you take active steps to show that treatment is more than the medication, that is what they will assume.
Each moment is a point made in the episode, linked to the second it was said. The topic links open everything the podcast has said on that topic.