Yes. If we restrict the question to unipolar major depression in older adults, roughly ≥60–65, and especially ask what has geriatric-specific evidence rather than simply extrapolating from younger adults, the hierarchy is clearer than one might expect.
Antidepressants
For ordinary late-life major depression, SSRIs and SNRIs remain effective, though the average drug–placebo separation is modest. Canadian geriatric guidelines specifically recommend sertraline or duloxetine as first-line choices, with escitalopram and citalopram as alternatives. They also regard venlafaxine, bupropion, mirtazapine, and vortioxetine as reasonable options.
My practical ranking of the evidence would be approximately:
|
Drug |
Geriatric evidence |
Clinical comment |
|
Sertraline |
Good |
Probably the safest default SSRI; relatively few CYP/DDI problems |
|
Duloxetine |
Good |
Particularly attractive with pain; some geriatric trials/meta-analyses show a robust efficacy signal |
|
Escitalopram |
Good |
Effective/well tolerated; QT issue becomes relevant at higher doses/older age |
|
Citalopram |
Good |
Effective, but QT ceiling makes it less convenient |
|
Venlafaxine XR |
Good |
Clearly effective; useful in more severe depression, but BP, withdrawal, SIADH issues |
|
Vortioxetine |
Quite good |
One particularly nice 65 RCT; possible cognitive benefit |
|
Mirtazapine |
Reasonable |
Particularly useful with insomnia, anorexia/weight loss |
|
Bupropion |
Reasonable |
Useful with fatigue/apathy, hypersomnia, sexual dysfunction; relatively low hyponatremia burden |
|
Nortriptyline |
Effective, older evidence |
It works, but cardiac/orthostatic/anticholinergic burden generally pushes it to later lines |
There is a particularly interesting geriatric RCT of vortioxetine in patients ≥65 with recurrent MDD: vortioxetine was superior to placebo and was generally well tolerated.
One caution regarding the often-made statement that “SNRIs are better in the elderly”: I would not say that. They work, but comparative safety is somewhat worse. An AHRQ review found more adverse events with duloxetine/venlafaxine than placebo, and duloxetine was associated with increased falls in one longer trial. SSRIs, vortioxetine, and bupropion generally looked somewhat cleaner on tolerability, although the evidence base is thin.
I would generally avoid paroxetine in geriatric depression because of anticholinergic burden, and usually avoid choosing fluoxetine de novo because of its very long half-life and interaction problems. TCAs work, but they are seldom first choices.
And the SGAs? Here the answer is much simpler.
Aripiprazole is in a class of its own for geriatric augmentation.
It is the SGA for which we have convincing, direct randomized evidence in late-life treatment-resistant depression.
The important Lenze et al. trial took patients ≥60 who had failed to remit with venlafaxine and randomized them to aripiprazole augmentation versus placebo. Remission was approximately:
44% aripiprazole vs 29% placebo — NNT ≈ 6.
The mean final aripiprazole dose was only about 7 mg/day.
The price was mainly akathisia and parkinsonism rather than major metabolic toxicity; akathisia occurred in roughly a quarter of patients.
Then came the very useful OPTIMUM trial in 2023. In adults ≥60 with treatment-resistant depression, augmenting the existing antidepressant with aripiprazole produced greater improvement in psychological well-being than switching to bupropion, with a numerically higher remission rate. It also had an important practical advantage over bupropion augmentation: fewer falls.
So if you asked me:
“I have an otherwise medically reasonable 72-year-old with a partial antidepressant response. Which SGA has the best evidence for augmentation?”
I would answer low-dose aripiprazole, quite decisively.
What about the other SGAs?
Quetiapine: there is geriatric evidence that quetiapine XR monotherapy can improve late-life depression, but that is not the same thing as demonstrating effective augmentation. The Canadian guideline specifically distinguishes this: quetiapine has geriatric monotherapy evidence, whereas aripiprazole has geriatric augmentation evidence. Sedation, orthostasis, metabolic effects, and falls make me substantially less enthusiastic about quetiapine in an 80-year-old unless insomnia/agitation gives it a special rationale.
Brexpiprazole: good adult MDD augmentation evidence, but I am not aware of comparably persuasive geriatric-specific RCT evidence. Thus I would call it plausibly effective by extrapolation, but not evidence-equivalent to aripiprazole in geriatric depression.
Cariprazine: same problem, only more so—adult adjunctive MDD evidence, little reason presently to prefer it specifically in late-life depression.
Risperidone: some small augmentation studies and older literature exist, but the geriatric evidence is far thinner than for aripiprazole, with greater concern about EPS/prolactin.
Olanzapine: I would reserve it for special circumstances. It has evidence in psychotic depression and olanzapine/fluoxetine in TRD generally, but it is not my choice for routine nonpsychotic geriatric augmentation because of metabolic, anticholinergic, and sedation burdens.
So I would summarize the SGA evidence rather brutally as:
Aripiprazole >>> quetiapine / brexpiprazole / risperidone / olanzapine / cariprazine
—not necessarily because the others don’t work, but because aripiprazole is the one for which we actually have unusually good late-life depression data.
One further point I think is clinically important: in geriatric TRD, the question is no longer necessarily “switch or augment?” The OPTIMUM study makes a fairly strong case that when there has been a meaningful partial response, augmentation—particularly with aripiprazole—is often preferable to throwing away the antidepressant response and switching.
And this strikes me as one of those rare areas in geriatric psychopharmacology where the evidence actually gives us a fairly actionable answer rather than merely saying, “start low, go slow.”
