Some doctors now start patients on lower antidepressant doses because research increasingly shows that standard starting doses cause intolerable side effects in a large proportion of patients, leading them to quit treatment before it has a chance to work. Initial SSRI therapy fails in roughly 50 percent of patients due to either lack of efficacy or side effects they cannot tolerate, according to a 2024 study published in Frontiers in Pharmacology. That failure rate has pushed clinicians — particularly those treating older adults and people with cognitive concerns — toward a more cautious approach: begin with a smaller dose, monitor the response, and increase gradually.
This shift is not simply about being conservative for its own sake. A growing body of evidence, including landmark research in The Lancet Psychiatry, demonstrates that SSRIs deliver their therapeutic benefits at the low-to-medium end of their licensed dose ranges, while higher doses mainly increase dropout rates from side effects. For someone caring for a parent with early-stage dementia who also needs treatment for depression, for instance, the difference between a starting dose of 25 milligrams of sertraline versus 50 milligrams can determine whether they stay on the medication long enough for it to help. This article examines the clinical reasoning behind lower starting doses, the role of genetic testing, what the research says about dose-response curves, how withdrawal concerns factor in, and what new guidelines in 2025 mean for patients and caregivers navigating antidepressant treatment.
Table of Contents
- Why Are Standard Antidepressant Starting Doses Failing So Many Patients?
- What Does the Dose-Response Research Actually Show?
- How Pharmacogenomic Testing Is Changing Prescribing Decisions
- How Withdrawal Concerns Are Reshaping How Doctors Begin Treatment
- Special Risks for Older Adults and People With Cognitive Decline
- New Guidelines and FDA Activity in 2025
- Where Personalized Antidepressant Dosing Is Headed
- Conclusion
- Frequently Asked Questions
Why Are Standard Antidepressant Starting Doses Failing So Many Patients?
The traditional approach to prescribing antidepressants assumed a relatively uniform starting point: pick the recommended dose from the label, hand the patient a prescription, and schedule a follow-up in four to six weeks. that model was built on clinical trial averages, not individual biology. The problem is that human bodies metabolize these drugs at wildly different rates. The CYP2D6 and CYP2C19 genes, which encode liver enzymes responsible for breaking down most antidepressants, vary dramatically across the population. Patients classified as poor metabolizers accumulate higher drug concentrations in their blood, meaning a standard dose can behave like an overdose in their system — producing nausea, dizziness, insomnia, and the kind of side effects that make people flush their pills down the toilet within the first week. Consider two patients prescribed the same 50-milligram dose of sertraline. One is a normal metabolizer and processes the drug at a predictable rate. The other is a poor metabolizer whose body clears the drug far more slowly, effectively experiencing a much higher exposure.
The second patient wakes up nauseated, cannot sleep, and calls the office asking to stop. This is not a rare scenario. The common side-effect rate with standard dosing runs between 30 and 50 percent, and for many patients, those first few weeks of misery are enough to end treatment permanently. Starting at a lower dose — say 25 milligrams — gives that second patient a chance to adjust before titrating upward. The challenge has been that most prescribers historically lacked the tools to identify which patients needed the lower starting point. As of 2017–2018 data, fewer than 1 percent of depressed patients received CYP2D6 or CYP2C19 pharmacogenomic testing through insurance. That number is growing, but slowly. In the meantime, the pragmatic response from many clinicians has been to treat every new patient as if they might be a poor metabolizer — start low and adjust based on what happens, rather than risk losing the patient to side effects in the first two weeks.
What Does the Dose-Response Research Actually Show?
One of the most influential pieces of evidence behind the lower-dose movement came from a Lancet Psychiatry analysis examining how antidepressant benefits scale with dose. The researchers found that SSRIs display dose-dependent benefits only up to the low-to-medium end of their licensed dose ranges. Beyond that threshold, increasing the dose does not meaningfully improve depression outcomes — it mainly increases the likelihood that patients will drop out of treatment because of side effects. In practical terms, this means a doctor who starts a patient at a high dose in hopes of a faster response is more likely to lose that patient from treatment entirely than to achieve a quicker recovery. This finding has particular relevance for older adults, including those with or at risk for dementia. Depression is both a symptom and a risk factor in cognitive decline, making effective treatment genuinely important.
But older adults are also more sensitive to medication side effects, more likely to be taking other drugs that interact with antidepressants, and more vulnerable to falls and confusion caused by adverse reactions. Clinical guidelines for this population now explicitly recommend starting at 50 percent of the standard adult dose and titrating gradually. For sertraline, that means beginning at 25 to 50 milligrams daily rather than the standard 50-milligram adult starting dose, according to recommendations from the Psychopharmacology Institute. However, there is an important caveat embedded in the geriatric prescribing literature that sometimes gets lost: “start low, go slow, but go.” The point is not to leave patients on a sub-therapeutic dose indefinitely. Starting low is a strategy for getting patients onto and tolerating their medication so that the dose can be raised to an effective level. A patient left on 25 milligrams of sertraline for six months without reassessment is not receiving cautious care — they are receiving inadequate care. The goal is always to reach the minimum effective dose, just without the collateral damage of starting too high.
How Pharmacogenomic Testing Is Changing Prescribing Decisions
Pharmacogenomic testing — sometimes called gene-guided prescribing — offers a more precise alternative to the guess-and-check method that has dominated psychiatry for decades. By analyzing a patient’s CYP2D6 and CYP2C19 gene variants, clinicians can classify them as poor, intermediate, normal, or ultrarapid metabolizers before writing a single prescription. A poor metabolizer might need half the standard dose or a different drug altogether, while an ultrarapid metabolizer might burn through a standard dose so quickly it never reaches therapeutic levels. The practical impact can be significant. Gene-guided dosing can help avoid the 30 to 50 percent side-effect rate associated with standard dosing, preventing weeks of nausea, dizziness, or insomnia that lead many patients to abandon treatment. For a 78-year-old with mild cognitive impairment who is already anxious about taking new medication, those first few weeks of side effects can be the difference between staying in treatment and refusing all future attempts.
A University Hospital Frankfurt observational study, launched in April 2025 and running through June 2026, is actively investigating how pharmacogenetic testing affects depression treatment outcomes, which should add to the evidence base. The limitation worth noting is access. Despite growing adoption, pharmacogenomic testing remains underutilized. Insurance coverage is inconsistent, costs can run several hundred dollars out of pocket, and many primary care physicians are not trained to interpret the results. If your loved one’s doctor has not mentioned pharmacogenomic testing, it is reasonable to ask about it — but do not assume its absence means negligent care. Many clinicians achieve similar outcomes through careful clinical observation and the start-low approach, which functions as a practical workaround when genetic data is unavailable.
How Withdrawal Concerns Are Reshaping How Doctors Begin Treatment
It may seem counterintuitive that worries about stopping a medication would influence how doctors start it, but this connection has become one of the strongest arguments for lower initial doses. Up to 50 percent of patients experience discontinuation symptoms when stopping antidepressants, according to the 2024 CANMAT guideline update. A 2024 systematic review estimated the absolute risk increase for withdrawal symptoms at approximately 15 percent, with a number needed to harm of 6 to 7 patients — meaning that for every 6 or 7 people who take antidepressants, one additional person will experience withdrawal symptoms beyond what a placebo group would report. Not all antidepressants carry equal withdrawal risk. Paroxetine, venlafaxine, desvenlafaxine, and duloxetine have the highest self-reported withdrawal rates, according to a June 2025 update from the Therapeutics Initiative at the University of British Columbia.
Starting these medications at lower doses creates a smaller physiological footprint, which makes eventual tapering safer and easier. Compare this to a patient started at the maximum recommended dose of venlafaxine: when the time comes to discontinue, the taper must cover far more ground, takes longer, and carries a higher risk of debilitating withdrawal symptoms including brain zaps, vertigo, and rebound anxiety. A 2025 Lancet Psychiatry meta-analysis of 76 randomized controlled trials involving more than 17,000 participants reinforced this logic from the other direction. The researchers found that slow tapering over more than four weeks, combined with psychological support, prevented relapse in 1 out of every 5 patients compared to abrupt discontinuation. The takeaway is that how you start antidepressant treatment shapes how you can end it. Physicians who are now thinking about the full arc of treatment — initiation through maintenance through discontinuation — naturally gravitate toward lower starting doses as part of a more sustainable prescribing strategy.
Special Risks for Older Adults and People With Cognitive Decline
The “start low, go slow” principle was originally developed with older adults in mind, and the reasons go beyond general caution. Aging changes how the body processes drugs in several ways: reduced liver and kidney function slows drug clearance, changes in body composition alter drug distribution, and the aging brain becomes more sensitive to psychoactive substances. For someone with dementia or mild cognitive impairment, these pharmacological realities are compounded by the difficulty of reporting side effects accurately. A person who cannot reliably describe how a medication makes them feel is at greater risk of suffering in silence through adverse effects that a younger, cognitively intact patient would immediately flag. There is also the interaction problem. Older adults with cognitive decline are frequently taking other medications — cholinesterase inhibitors, blood pressure drugs, blood thinners — that can interact with antidepressants in ways that amplify side effects or reduce effectiveness.
Starting an antidepressant at a lower dose provides a window to observe for interactions before the drug reaches its full concentration. A geriatric psychiatrist treating depression in a patient already on donepezil, for example, has to account for the fact that both medications affect serotonin pathways. Beginning with a reduced antidepressant dose is not timidity; it is recognition that the pharmacological environment in this patient’s body is already crowded. The warning here is against the opposite extreme. Some caregivers, having heard about medication risks in dementia, resist antidepressant treatment entirely. Untreated depression in someone with cognitive decline accelerates functional decline, worsens behavioral symptoms, and increases caregiver burden. The goal is not to avoid medication but to use it wisely, and lower starting doses make that possible for patients who might otherwise be denied treatment out of fear.
New Guidelines and FDA Activity in 2025
The year 2025 has brought meaningful updates to the prescribing landscape. The FDA-approved landscape now includes newer agents and updated label changes in psychiatry that reflect the broader trend toward personalized dosing. These label changes increasingly acknowledge that one-size-fits-all dosing recommendations fail a significant portion of patients and that starting-dose flexibility should be part of standard practice, not an off-label workaround.
A June 2025 Therapeutics Letter update specifically encourages re-evaluation of prescribing practices in primary care, particularly around starting doses and long-term use. For families navigating the intersection of depression and cognitive decline, this is relevant because primary care is often where antidepressant prescriptions originate — before a patient ever sees a psychiatrist or neurologist. If your loved one’s primary care doctor is still defaulting to the old approach of standard-dose-first, the new guidelines provide a basis for a conversation about alternatives. Bring the topic up directly: ask whether a lower starting dose might be appropriate and whether pharmacogenomic testing is available.
Where Personalized Antidepressant Dosing Is Headed
The trajectory of antidepressant prescribing is clearly moving toward greater individualization, and the tools to support that shift are maturing. Pharmacogenomic testing is becoming cheaper and more widely available. Research like the ongoing University Hospital Frankfurt study is building the evidence base for gene-guided dosing in real clinical settings rather than idealized trial conditions. And the deprescribing movement — which focuses on how to safely reduce or stop medications — is feeding back into initiation practices, creating a more coherent philosophy that considers the entire medication lifecycle.
For families dealing with dementia and depression simultaneously, this evolution is welcome but incomplete. The gap between what the research supports and what happens in a busy primary care office remains wide. The most practical step available right now is also the simplest: ask the prescribing doctor whether a lower starting dose is appropriate, whether pharmacogenomic testing might be useful, and what the plan is for monitoring response and adjusting the dose over time. The evidence strongly supports that starting low and going slow — while still going — produces better outcomes for more patients, especially those whose brains are already vulnerable.
Conclusion
The shift toward lower starting doses of antidepressants reflects a genuine change in how the medical community understands these medications. The old model of prescribing a standard dose and hoping for the best produced failure rates near 50 percent and left millions of patients either suffering through side effects or abandoning treatment entirely. Research now clearly shows that therapeutic benefits peak at the low-to-medium dose range, that genetic variation means there is no true standard patient, and that how treatment begins directly shapes how safely it can eventually end. For caregivers and families in the dementia space, this matters deeply.
Depression worsens cognitive decline, and effective treatment depends on the patient actually tolerating the medication long enough for it to work. Lower starting doses, pharmacogenomic testing where available, and careful upward titration give older adults and cognitively vulnerable patients the best chance of benefiting from antidepressants without being driven away by early side effects. Talk with your loved one’s physician about whether their current dosing approach reflects these newer findings — and remember that cautious initiation is not the same as inadequate treatment. The goal is still to reach an effective dose. The route there is just getting smarter.
Frequently Asked Questions
What does “start low, go slow” actually mean in practice?
It means beginning an antidepressant at roughly half the standard adult dose — for example, 25 milligrams of sertraline instead of 50 — and increasing the dose gradually over weeks based on the patient’s response and tolerance. Clinical guidelines emphasize adding “but go” to this phrase, meaning the doctor should still titrate upward to reach an effective dose rather than leaving the patient on a sub-therapeutic amount indefinitely.
Is pharmacogenomic testing covered by insurance?
Coverage varies widely. As of 2017–2018 data, fewer than 1 percent of depressed patients received CYP2D6 or CYP2C19 testing through insurance, though adoption has been increasing. Some insurers now cover it after a patient has failed one or two medications, while others still consider it experimental. Out-of-pocket costs can range from roughly $200 to $500 depending on the testing company.
Which antidepressants are hardest to stop taking?
Paroxetine, venlafaxine, desvenlafaxine, and duloxetine carry the highest self-reported withdrawal rates, according to a June 2025 update from the Therapeutics Initiative. Up to 50 percent of patients experience some form of discontinuation symptoms, and these particular medications tend to produce the most severe withdrawal effects.
Does a lower starting dose mean the medication will take longer to work?
It may add one to two weeks to the timeline for reaching a therapeutic dose, since the titration process involves gradual increases. However, starting at a standard dose and then having to stop due to side effects — or switch to a different medication entirely — costs far more time. The net effect of starting low is usually faster arrival at sustained, tolerable treatment.
Should someone with dementia take antidepressants at all?
Depression worsens cognitive decline and increases behavioral symptoms, so leaving it untreated carries real risks. Antidepressants can be appropriate for people with dementia, but the prescribing approach needs to account for increased drug sensitivity, potential interactions with other medications, and the patient’s reduced ability to report side effects. Lower starting doses and closer monitoring are especially important in this population.
