Selecting an antidepressant

[FuturePsychiatrist827]

Hello, I am a soon to be Psychiatry PGY-1 resident. I am obviously still learning the fundamentals of psychiatry and want to pose a question to this forum. How do you go about selecting a particular antidepressant for a patient who is presenting with anxiety or depression?

Some "textbook" cases seem obvious to me, like in an elderly person who is losing weight and has depressed mood, mirtazepine would be a good choice. Or a person with major depression who smokes, burpropion would be a good choice.

In particular, I'm on a CA psychiatry rotation and there is a 17 year old female who has been diagnosed with GAD with dysthymia. Her anxiety is so bad that she will be dropping out of high school to finish online, and she is skeptical of anti-depressants due to fears of weight gain, but recognizes that her symptoms are impacting her ability to function. Her BMI is on the upper end of normal limits. How would I go about selecting the best antidepressant for her?

---

Members don't see this ad.

 

[mk04447]

Try bupropion.

 

[tr]

I usually base this decision on whether the patient is more lethargic/deactivated or anxious/hyperactivated.

I have a rough hierarchy of activation, which works something like this (bearing in mind that everyone is an individual and reactions vary):

Tricyclics<Paxil<Zoloft<Celexa<Lexapro<Prozac<Cymbalta<Effexor<Wellbutrin.

For people who are anxious and can't sleep I usually aim towards Zoloft. For people who are hypersomnic and can't get out of bed in the morning I aim towards Wellbutrin. If they've tried an antidepressant in the past sometimes they can say how it affected them and whether they feel that something more activating or more soothing than whatever they tried would be helpful.

Other than this there are specific indicators as you mentioned: elderly or underweight--> Remeron; multiple med interactions --> Lexapro; smoker or overweight --> Wellbutrin; chronic pain --> Cymbalta.

Tricyclics obviously promote weight gain, and I've seen a few people get weight gain on Zoloft, Paxil, and *maybe* Lexapro. Pretty much never on Prozac or SNRIs, and Wellbutrin promotes weight loss. For anxiety I usually aim towards the soothing end of the scale, but that's also the weight-gain end of the scale, so for someone with anxiety who is worried about weight gain I might aim for Prozac. It can be a little activating and does impair sleep but less so than SNRI/Wellbutrin, and while it can feel a little 'caffeinated' it does have demonstrated efficacy for anxiety, vs SNRIs/Wellbutrin which can often worsen anxiety. Also she's young, and Prozac has a good track record for children and adolescents. At 17 this person is very much in the age window for acute SSRI-induced SI so you definitely want to warn her about that before starting anything. Has she already maximized behavioral/psychotherapeutic approaches?

 

[DisorderedDoc417]

Comparative efficacy and acceptability of 21 antidepressant drugs for the acute treatment of adults with major depressive disorder: a systematic review and network meta-analysis
https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(17)32802-7/fulltext


There are guidelines to such things, I would recommend continuing to compare advantages (and side-effects) and trying to make decisions on a case-by-case basis. No one-size-fits all for depression as you know.

 

[Drrrrrr. Celty]

I usually base this decision on whether the patient is more lethargic/deactivated or anxious/hyperactivated.

I have a rough hierarchy of activation, which works something like this (bearing in mind that everyone is an individual and reactions vary):

Tricyclics<Paxil<Zoloft<Celexa<Lexapro<Prozac<Cymbalta<Effexor<Wellbutrin.

For people who are anxious and can't sleep I usually aim towards Zoloft. For people who are hypersomnic and can't get out of bed in the morning I aim towards Wellbutrin. If they've tried an antidepressant in the past sometimes they can say how it affected them and whether they feel that something more activating or more soothing than whatever they tried would be helpful.

Other than this there are specific indicators as you mentioned: elderly or underweight--> Remeron; multiple med interactions --> Lexapro; smoker or overweight --> Wellbutrin; chronic pain --> Cymbalta.

Tricyclics obviously promote weight gain, and I've seen a few people get weight gain on Zoloft, Paxil, and *maybe* Lexapro. Pretty much never on Prozac or SNRIs, and Wellbutrin promotes weight loss. For anxiety I usually aim towards the soothing end of the scale, but that's also the weight-gain end of the scale, so for someone with anxiety who is worried about weight gain I might aim for Prozac. It can be a little activating and does impair sleep but less so than SNRI/Wellbutrin, and while it can feel a little 'caffeinated' it does have demonstrated efficacy for anxiety, vs SNRIs/Wellbutrin which can often worsen anxiety. Also she's young, and Prozac has a good track record for children and adolescents. At 17 this person is very much in the age window for acute SSRI-induced SI so you definitely want to warn her about that before starting anything. Has she already maximized behavioral/psychotherapeutic approaches?

This is a pretty great explanation. I've only mildly had people explain the differences between anti-depressants or anti-psychotics with mostly just people saying that Prozac is more activating.

 

[FuturePsychiatrist827]

I usually base this decision on whether the patient is more lethargic/deactivated or anxious/hyperactivated.

I have a rough hierarchy of activation, which works something like this (bearing in mind that everyone is an individual and reactions vary):

Tricyclics<Paxil<Zoloft<Celexa<Lexapro<Prozac<Cymbalta<Effexor<Wellbutrin.

For people who are anxious and can't sleep I usually aim towards Zoloft. For people who are hypersomnic and can't get out of bed in the morning I aim towards Wellbutrin. If they've tried an antidepressant in the past sometimes they can say how it affected them and whether they feel that something more activating or more soothing than whatever they tried would be helpful.

Other than this there are specific indicators as you mentioned: elderly or underweight--> Remeron; multiple med interactions --> Lexapro; smoker or overweight --> Wellbutrin; chronic pain --> Cymbalta.

Tricyclics obviously promote weight gain, and I've seen a few people get weight gain on Zoloft, Paxil, and *maybe* Lexapro. Pretty much never on Prozac or SNRIs, and Wellbutrin promotes weight loss. For anxiety I usually aim towards the soothing end of the scale, but that's also the weight-gain end of the scale, so for someone with anxiety who is worried about weight gain I might aim for Prozac. It can be a little activating and does impair sleep but less so than SNRI/Wellbutrin, and while it can feel a little 'caffeinated' it does have demonstrated efficacy for anxiety, vs SNRIs/Wellbutrin which can often worsen anxiety. Also she's young, and Prozac has a good track record for children and adolescents. At 17 this person is very much in the age window for acute SSRI-induced SI so you definitely want to warn her about that before starting anything. Has she already maximized behavioral/psychotherapeutic approaches?

Wow! Thanks for the detailed response! This is a great learning tool for me. The patient's mother is very closed off to the possibility of medications helping, saying she is worried about her "committing suicide" while on an antidepressant, and I think the patient has adopted this mindset due to her mother's attitudes. The mother has tried all kinds of naturopathic therapy, and even tried some Eastern medicine (Reiki, trapped emotion release therapy (?)) I don't think the patient is open to psychotherapy helping right now, she SAYS she has tried it in the past and but it "didn't work". She is unsure of which therapy modality she utilized at that time. The patient is definitely more on the anxious/insomnia spectrum than the depressed/lethargic spectrum. She has a vague history of suicidal ideation ("I'm worried I will hurt myself, but I don't know how.") but no attempts or any sort of plan.

I like the idea of Prozac being more weight neutral for her, but would it be too activating for somebody with anxiety-induced insomnia? Maybe Zoloft would be the better choice, but I think she wouldn't like the weight gain side effects if there were any.

 

[MacDonaldTriad]

Try bupropion.

Please don't recommend bupropion for an anxiety disorder. It is the only antidepressant that has been proven to be ineffective in this indication.

 

[mk04447]

Please don't recommend bupropion for an anxiety disorder. It is the only antidepressant that has been proven to be ineffective in this indication.

No kidding.

 

[tr]

Wow! Thanks for the detailed response! This is a great learning tool for me. The patient's mother is very closed off to the possibility of medications helping, saying she is worried about her "committing suicide" while on an antidepressant, and I think the patient has adopted this mindset due to her mother's attitudes. The mother has tried all kinds of naturopathic therapy, and even tried some Eastern medicine (Reiki, trapped emotion release therapy (?)) I don't think the patient is open to psychotherapy helping right now, she SAYS she has tried it in the past and but it "didn't work". She is unsure of which therapy modality she utilized at that time. The patient is definitely more on the anxious/insomnia spectrum than the depressed/lethargic spectrum. She has a vague history of suicidal ideation ("I'm worried I will hurt myself, but I don't know how.") but no attempts or any sort of plan.

I like the idea of Prozac being more weight neutral for her, but would it be too activating for somebody with anxiety-induced insomnia? Maybe Zoloft would be the better choice, but I think she wouldn't like the weight gain side effects if there were any.

So if you think Prozac would be too activating you can move one down on the activation scale and try Lexapro. Weight gain is possible but unlikely, and it seems to hover just about neutral on the activation scale (a few people find it activating, others find it sedating, most find it neutral).

Even if she is not open to traditional psychotherapy, there are still lots of behavioral things you can do on your own. You can make sure her sleep hygiene is good (teenagers often have terrible sleep hygiene), encourage some basic behavioral activation (exercise/social engagement/outdoor time), and suggest psychotherapeutic alternatives like phone apps (Headspace, Calm, etc.)

 

[smalltownpsych]

Roughly about 99.9% of the anxious or depressed adolescents I have met with benefitted from psychotherapy whereas only about 25% seem to benefit from medications (might be exaggerating a little ). Resistance to that is interesting. It could be just that the other psychotherapist didn't know how to work with teens or it could relate directly to whatever the problem is. I would feel the need to delve further into it to figure out what is really going on.

 

[DisorderedDoc417]

So if you think Prozac would be too activating you can move one down on the activation scale and try Lexapro. Weight gain is possible but unlikely, and it seems to hover just about neutral on the activation scale (a few people find it activating, others find it sedating, most find it neutral).

Even if she is not open to traditional psychotherapy, there are still lots of behavioral things you can do on your own. You can make sure her sleep hygiene is good (teenagers often have terrible sleep hygiene), encourage some basic behavioral activation (exercise/social engagement/outdoor time), and suggest psychotherapeutic alternatives like phone apps (Headspace, Calm, etc.)

I was shocked to see how expensive lexapro still is even though it has been generic.....

 

[NickNaylor]

Unless I have a reason to select another agent, I will typically go with escitalopram. There is some evidence (a meta analysis) that it may be more effective than most other antidepressants, though I’m skeptical that the difference is clinically significant. It is also generally better tolerated in studies with respect to having lower rates of discontinuation.

But there is no definitive treatment algorithm I’ve come across that is truly evidence-based. All the SSRIs and SNRIs are more or less equivalent, and we have no real way to predict whether an individual will respond to or be unable to tolerate one particular agent. So it becomes a game of trial and error to figure out what works and what they’re able to tolerate.

Of course, there may be factors in each individual case that may sway me to choose one antidepressant over another. But those are nuances which you will learn about over the course of your training and are difficult to summarize, though the discussion about “activation” above is one example of this kind of nuance.

 

[J ROD]

Please don't recommend bupropion for an anxiety disorder. It is the only antidepressant that has been proven to be ineffective in this indication.

This. Do NOT use that. I would try and use Lexapro. Fairly weight neutral. And will work for anxiety. Prozac may be too activating and make her anxiety worse. Depends on person. Zoloft has more weight effects on average.

 

[FuturePsychiatrist827]

Unless I have a reason to select another agent, I will typically go with escitalopram. There is some evidence (a meta analysis) that it may be more effective than most other antidepressants, though I’m skeptical that the difference is clinically significant. It is also generally better tolerated in studies with respect to having lower rates of discontinuation.

But there is no definitive treatment algorithm I’ve come across that is truly evidence-based. All the SSRIs and SNRIs are more or less equivalent, and we have no real way to predict whether an individual will respond to or be unable to tolerate one particular agent. So it becomes a game of trial and error to figure out what works and what they’re able to tolerate.

Of course, there may be factors in each individual case that may sway me to choose one antidepressant over another. But those are nuances which you will learn about over the course of your training and are difficult to summarize, though the discussion about “activation” above is one example of this kind of nuance.

I'll have to some more learning about escitalopram. I've seen that it has less side effects than other anti-depressants from studying for boards and shelf exams. I have heard it is expensive though, even if you choose escitalopram over Lexapro?

 

[clausewitz2]

I'll have to some more learning about escitalopram. I've seen that it has less side effects than other anti-depressants from studying for boards and shelf exams. I have heard it is expensive though, even if you choose escitalopram over Lexapro?

In my area goodrx suggests it is available for about 10 dollars a month without insurance, so...I reckon it's not that expensive, although not quite the same as the 4 dollar list.

 

[NickNaylor]

I'll have to some more learning about escitalopram. I've seen that it has less side effects than other anti-depressants from studying for boards and shelf exams. I have heard it is expensive though, even if you choose escitalopram over Lexapro?

It is more expensive than the $4 list medications, but my understanding is that it isn’t terribly expensive. If escitalopram isn’t available, my second back-up is typically fluoxetine. Not ideal because of the activation, but I’m just more conservative and slow with the dose titration to try and minimize side effects. I believe citalopram is the only other SSRI on most $4 lists, and I generally try to avoid it due to the black box warning.

 

[Liquid8]

Comparative efficacy and acceptability of 21 antidepressant drugs for the acute treatment of adults with major depressive disorder: a systematic review and network meta-analysis
https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(17)32802-7/fulltext

There are guidelines to such things, I would recommend continuing to compare advantages (and side-effects) and trying to make decisions on a case-by-case basis. No one-size-fits all for depression as you know.

Agree that the decision generally comes down to the specifics of each case. Anyone can prescribe like an automaton based on guidelines, but the art of psychiatry is being able to elicit the specifics and knowing when (or when not) to apply your knowledge.

I remember the 2009 Cipriani study being particularly ground breaking and influential, as it was one of the early network meta-analyses in psychiatry which as you probably know allowed for head to head comparisons of different antidepressants. The most effective treatments from that paper were Escitalopram, Venlafaxine, Mirtazapine and Sertraline, and the most “acceptable” from a side effect perspective were Escitalopram and Sertraline. Reboxetine was down the bottom of the list, and that also coincided with released details about unsuccessful trials against placebo that had been buried by the manufacturer.

This more recent paper has Amitriptyline at the top in terms of response, with Mirtazapine, Duloxetine, Venlafaxine and Paroxetine rounding out the top 5. Of the SSRIs, after Paroxetine we have Fluvoxamine (7), Escitalopram (8), Sertraline (10), Fluoxetine (16) and Citalopram (17).

However, I’d still use Escitalopram as a first line SSRI for adults over paroxetine and fluvoxamine, primarily based on the relative lack of CYP interactions. Escitalopram, Citalopram and Sertraline are SSRIs with few, whereas Paroxetine, Fluoexetine and Fluvoxamine have many.

 

[thoffen]

I think the placebo benefit of having a rationale for choosing a particular antidepressant treatment and communicating this confidently in alliance with a patient's priority of symptoms trumps any actual clinical difference between medication options in the majority of cases.

 

[DisorderedDoc417]

Agree that the decision generally comes down to the specifics of each case. Anyone can prescribe like an automaton based on guidelines, but the art of psychiatry is being able to elicit the specifics and knowing when (or when not) to apply your knowledge.

I remember the 2009 Cipriani study being particularly ground breaking and influential, as it was one of the early network meta-analyses in psychiatry which as you probably know allowed for head to head comparisons of different antidepressants. The most effective treatments from that paper were Escitalopram, Venlafaxine, Mirtazapine and Sertraline, and the most “acceptable” from a side effect perspective were Escitalopram and Sertraline. Reboxetine was down the bottom of the list, and that also coincided with released details about unsuccessful trials against placebo that had been buried by the manufacturer.

This more recent paper has Amitriptyline at the top in terms of response, with Mirtazapine, Duloxetine, Venlafaxine and Paroxetine rounding out the top 5. Of the SSRIs, after Paroxetine we have Fluvoxamine (7), Escitalopram (8), Sertraline (10), Fluoxetine (16) and Citalopram (17).

However, I’d still use Escitalopram as a first line SSRI for adults over paroxetine and fluvoxamine, primarily based on the relative lack of CYP interactions. Escitalopram, Citalopram and Sertraline are SSRIs with few, whereas Paroxetine, Fluoexetine and Fluvoxamine have many.

Of course. Lexapro tops the list. Great option with little interaction. Just a sucky withdraw. Surprisingly mirtazapine scored highly as well.

 

[tr]

My clinical impression is that there is a great deal of variability in response to different antidepressants. Patient A may find Lexapro ineffective and Zoloft highly effective, while patient B finds Lexapro effective and Zoloft not. If you make up two study groups in which half of each group is an A-type and half of each group is a B-type, and give one group Lexapro and the other group Zoloft, you are likely to find very little difference in overall efficacy between them.

That being the case, I don't find most published comparisons very useful because they mostly don't sort patients out by baseline characteristics. Randomly assigning groups of participants to receive one drug or the other completely washes out the individual variability that might have predictive value for drug choice. As a result, mostly what is found is that the overall effect sizes are not too different between drugs.

I'd like to see someone sort patients out by biotype, as for example in
Resting-state connectivity biomarkers define neurophysiological subtypes of depression
and take a look at drug efficacy on that basis.

 

[DisorderedDoc417]

My clinical impression is that there is a great deal of variability in response to different antidepressants. Patient A may find Lexapro ineffective and Zoloft highly effective, while patient B finds Lexapro effective and Zoloft not. If you make up two study groups in which half of each group is an A-type and half of each group is a B-type, and give one group Lexapro and the other group Zoloft, you are likely to find very little difference in overall efficacy between them.

That being the case, I don't find most published comparisons very useful because they mostly don't sort patients out by baseline characteristics. Randomly assigning groups of participants to receive one drug or the other completely washes out the individual variability that might have predictive value for drug choice. As a result, mostly what is found is that the overall effect sizes are not too different between drugs.

I'd like to see someone sort patients out by biotype, as for example in
Resting-state connectivity biomarkers define neurophysiological subtypes of depression
and take a look at drug efficacy on that basis.

I think we can become as nuanced as we like, collecting data to fit one model vs another, but as long as we describe depression as a collection of symptoms without a known/correlated physiological state (identifiable with a ct scanner or lab) which we can measure independently of subjective reporting of symptoms, we will continue to run into this situation. As I’ve noted in other posts, if we were treating fever and found that in some people amoxicillin worked and for others doxycycline was perfectly effective, but then also found case after case where each was essentially worthless, and concluded they were more or less the same efficacy without any understanding of the mechanisms of microbes or antibiotic resistance, or even the underlying cause of systemic infection, we would be led to make silly, erroneous conclusions. Some day...

 

[thoffen]

I think we can become as nuanced as we like, collecting data to fit one model vs another, but as long as we describe depression as a collection of symptoms without a known/correlated physiological state (identifiable with a ct scanner or lab) which we can measure independently of subjective reporting of symptoms, we will continue to run into this situation. As I’ve noted in other posts, if we were treating fever and found that in some people amoxicillin worked and for others doxycycline was perfectly effective, but then also found case after case where each was essentially worthless, and concluded they were more or less the same efficacy without any understanding of the mechanisms of microbes or antibiotic resistance, or even the underlying cause of systemic infection, we would be led to make silly, erroneous conclusions. Some day...

If and when we get there I will be ecstatic. However, even if we learn what the appropriate targets are neurophysiologically, we are probably never going to do very well at delivering targeted interventions, especially chemical ones. Evolution wasn't so kind as to make all depression pathways and only those utilize a particular serotonin receptor subtype, for instance. Yet for reasons largely unknown medications which have the exact same mechanism of action have different efficacy on an individual basis. So maybe we'll find out more leading to better individualized treatment plans, but I am personally skeptical that it will really make much difference at all.

 

[clausewitz2]

If and when we get there I will be ecstatic. However, even if we learn what the appropriate targets are neurophysiologically, we are probably never going to do very well at delivering targeted interventions, especially chemical ones. Evolution wasn't so kind as to make all depression pathways and only those utilize a particular serotonin receptor subtype, for instance. Yet for reasons largely unknown medications which have the exact same mechanism of action have different efficacy on an individual basis. So maybe we'll find out more leading to better individualized treatment plans, but I am personally skeptical that it will really make much difference at all.

Not to mention the probably critical role of neuropeptides and other large molecules that are very difficult to get from someone's mouth/bloodstream into the CNS...

 

[whopper]

When you look at the vast amounts of data regarding antidepressants, there is no clear data showing one antidepressant as vastly superior to another in terms of efficacy. There is some data (that's corporate sponsored) that Vortioxetine (Trintellix) has better efficacy but it's not by much, plus the price of the medication is very expensive. For these reasons I tend not to give out this med first-line.

It basically comes down then, IMHO, trying the cheapest meds first with the least amount of side effects. This usually pushes me to give out Escitalopram or Citalopram first.

I discourage students and residents in picking a med as their favorite. I don't give a crap about how a resident likes a med. What matters is if the med had a positive effect on the patient and if it's tolerable and affordable. If some idiot doesn't like that cause Escitalopram's their favorite med, then they're egocentric and haven't been able to figure out that studies only show what a medication does to a group of people not the individual. I don't like Escitalopram anymore or less than any other medication other than that it's cheap and has lesser side effects for the majority. If a patient has a bad reaction to it then I tell the patient to get off of it. If it turned out that ProLexaphuck was the cheapest/safest med that gave the patient tremendous improvement I'd give the patient that one.

The notion of developing a personal feeling for a med as, "my favorite," is absurd.

There is data showing that some patients don't respond well to SSRIs (with the exception of Fluoxetine). So if a patient doesn't do well on Escitalopram I tend to avoid any SSRI except maybe Fluoxetine and try an SNRI next.

Although the TCAs are cheap (some aren't...check out the price of Clomipramine) I tend not to prescribe them due to side effects and because they're anticholinergic (raising risk of dementia)

I also avoid Paroxetine in general. Highest risk of side effects among the SSRIs (except perhaps Luvox), it's an anticholinergic, and high risk for discontinuation syndrome.

I tend to avoid the more expensive and newer antidepressants because of the price and because this creates an X-factor that's beyond my and the patient's control. I've had several experiences where an insurance company will abruptly discontinue coverage of a medication and it drives the patient and I nuts while we try to get them to cover it. I don't need that crap especially when this type of thing can take over an hour of work and come out of nowhere.

As for the other antidepressants, Mirtazapine-I tend to avoid unless the pt needs sleep or weight gain, Wellbutrin-Open to this med first line especially if there's smoking, ADHD, or need for energy.

I'll resort to the more expensive meds if the cheaper ones failed, at least 2-3 of them.

 

[SourceOfDenial]

I'm reading this case and not necessarily seeing a reason why an antidepressant is absolutely necessary, especially since she already says she's leery of taking them. Putting her on an ssri isn't a long term plan to treating her anxiety.

For mirtazapine there is a study that compared it to other antidepressants in I think an older cancer patient population(but I may be misremembering). Has a higher dropout rate than other meds due to somnolence, dry mouth, GI issues.

It seems to cause weight gain in the individuals that are already obese or don't want weight gain. Not sure of that mechanism, just clinical observation.

 

[WingedOx]

If and when we get there I will be ecstatic.

We won't.

And in the rare rare chance we do, it's hard to see the cost benefits being worth it.

 

[whopper]

Since it's been brought up above.

Mirtazapine-there's strong evidence it causes weight gain and somnolence in many. It also slightly drops neutrophils. Any med that ends in "pine" does including Clozapine, Olanzapine, etc, and while Clozapine is the major one the other ones do too though usually on a subclinical level. They usually are only of concern when the neutrophil count on the patient is already low or the patient is on several meds at once that can also slightly lower it-death by a thousand cuts though it's usually not death but several subclinical effects adding up to a clinically noticeable effect.

One should not be very hesitant to start antidepressants so long as one is confident that the patient is suffering a disorder where it's indicated for use. Unless it's a TCA or MAO-I, in general, the risk of problems is low (minus Fluvoxamine, Mirtazapine, Paroxetine and Venlafaxine), and depression/anxiety does have physiological implications that are not healthy mentally and physically. Depression and anxiety that is mild-it's indicated though at that level psychotherapy can also be used INSTEAD of the antidepressant, but at moderate to worse depression/anxiety medications are strongly recommended with or without psychotherapy.

Mirtazapine's benefits are that some need to gain the weight and/or sleep. It has no sexual side effects, and because it's not an SSRI it's an option as a second medication add-on to an SSRI if the first med is not working well enough.

 

[Merovinge]

I usually base this decision on whether the patient is more lethargic/deactivated or anxious/hyperactivated.

I have a rough hierarchy of activation, which works something like this (bearing in mind that everyone is an individual and reactions vary):

Tricyclics<Paxil<Zoloft<Celexa<Lexapro<Prozac<Cymbalta<Effexor<Wellbutrin.

For people who are anxious and can't sleep I usually aim towards Zoloft. For people who are hypersomnic and can't get out of bed in the morning I aim towards Wellbutrin. If they've tried an antidepressant in the past sometimes they can say how it affected them and whether they feel that something more activating or more soothing than whatever they tried would be helpful.

Other than this there are specific indicators as you mentioned: elderly or underweight--> Remeron; multiple med interactions --> Lexapro; smoker or overweight --> Wellbutrin; chronic pain --> Cymbalta.

Tricyclics obviously promote weight gain, and I've seen a few people get weight gain on Zoloft, Paxil, and *maybe* Lexapro. Pretty much never on Prozac or SNRIs, and Wellbutrin promotes weight loss. For anxiety I usually aim towards the soothing end of the scale, but that's also the weight-gain end of the scale, so for someone with anxiety who is worried about weight gain I might aim for Prozac. It can be a little activating and does impair sleep but less so than SNRI/Wellbutrin, and while it can feel a little 'caffeinated' it does have demonstrated efficacy for anxiety, vs SNRIs/Wellbutrin which can often worsen anxiety. Also she's young, and Prozac has a good track record for children and adolescents. At 17 this person is very much in the age window for acute SSRI-induced SI so you definitely want to warn her about that before starting anything. Has she already maximized behavioral/psychotherapeutic approaches?

I have the except same conceptualization of these medications and every once and awhile I wonder if it is just taught this way to make us feel we know our medications better, or if we really do know them better than uptodate. If another attending from a different area and a different generation has the same thoughts, I feel a lot better. So beyond educating others, thanks for the .