Continuing an Effective but Metabolically Risky Antipsychotic vs. Switching
Olanzapine works better than anything either of these two patients has tried — and is quietly reshaping both of their metabolic health while it does. The same drug, the same tension, two very different answers.
C.W., a 33-year-old woman, describes the nine years before olanzapine as a slow loss of nearly everything she'd built as a young adult — a college program she never finished, an apartment she couldn't keep, friendships that faded as her untreated symptoms made ordinary contact harder to sustain. She was switched to olanzapine fourteen months ago after two prior antipsychotics failed to control her auditory hallucinations and disorganized thinking; olanzapine produced the first sustained remission of her illness, and she has since returned to part-time retail work and describes her quality of life as better than at any point since her diagnosis.
Over the same fourteen months, her weight has increased by eighteen pounds, her fasting glucose has risen from normal to prediabetic range, and her LDL has climbed into a range her primary care physician flagged as newly concerning at her last physical. Her presentation is the classic version of this dilemma: real, measurable psychiatric benefit set against real, measurable and still-progressing metabolic harm, in a patient young enough that the cardiometabolic trajectory she's on, left unaddressed, has decades to compound. She has told her care team plainly that she is frightened by both directions this could go — frightened of what changing anything might do to her remission, and frightened of what leaving it unaddressed might do to her health over the coming decades.
I'd hold on switching her. Two prior failures before olanzapine worked isn't a small detail — it's real evidence that another agent might not give her back what she has now, and she's told us directly how much this remission means to her.
I hear that this remission is fragile and hard-won, and I'm not proposing we risk it. But you're treating her labs like a static risk factor being monitored — they're actively trending worse over fourteen months with no sign of plateauing, and at her age that trajectory compounds for decades. This doesn't have to mean stopping olanzapine today. It means treating the metabolic problem as urgently as the psychiatric one, not as a secondary concern to revisit later once things get worse.
There's a real middle path, and it isn't just a delay tactic. De Silva and colleagues' 2016 meta-analysis of twelve placebo-controlled trials found metformin produced a real, significant benefit for antipsychotic-associated weight and lipid changes. I want to be honest about its limits, though: that same meta-analysis did not find a statistically significant benefit specifically for fasting glucose — and her glucose, not her weight, is the value trending most concerning. If her labs stabilize or improve with active management, that changes the calculus; if her glucose specifically keeps worsening despite it, that's a different, harder conversation about switching.
Agreed: continue olanzapine, add metformin, and refer to a dietitian, with metabolic labs rechecked in three months rather than at the next routine annual interval.
Explicitly left open: if her labs continue worsening despite active management at the three-month recheck, the team agreed a switch conversation returns to the table — not indefinitely deferred, just not decided today.
R.F., a 33-year-old man — matched to C.W.'s age deliberately for this comparison — works as a delivery driver, a job his wife says has become genuinely harder for him over the past year, though she attributes some of that to fatigue rather than his psychiatric symptoms alone. He was also switched to olanzapine after two prior antipsychotic failures, roughly the same window ago as C.W., and has had a comparably strong psychiatric response. Unlike C.W., R.F. came into this treatment already carrying a body mass index in the obese range, a pre-existing type 2 diabetes diagnosis managed with metformin, and an LDL already above goal on baseline statin therapy before olanzapine was ever started.
Since starting olanzapine, his weight has increased further, his A1c has risen despite an unchanged metformin dose, and his cardiologist, managing his pre-existing coronary risk profile, has raised direct concern about his now-compounding cardiometabolic burden. His wife, present at this visit, says she has watched his energy decline noticeably over the past several months and worries the fatigue itself may now be affecting his ability to keep up with his driving route. The pivot is not that his psychiatric benefit differs from C.W.'s — it doesn't — it's that he was never metabolically low-risk to begin with, and olanzapine is now accelerating an already-active disease process rather than initiating a new one in a previously healthy metabolic baseline.
This isn't the same decision as Case A wearing a different name. He came in with active diabetes and coronary risk already established — olanzapine isn't creating a new problem here, it's pouring fuel on one that was already burning, and his A1c rising despite unchanged metformin says the current approach isn't holding the line. Whatever helped stabilize her doesn't have the same room to work on a system that was already failing before olanzapine started.
I hear that, and I still don't think the psychiatric stakes are smaller for him than they were for C.W. — two prior failures, now a real remission. The CAMP trial — Stroup and colleagues' randomized study on switching for metabolic reasons, published in the American Journal of Psychiatry — found real improvement in weight and cholesterol, but it also found a meaningfully higher discontinuation rate in the group that switched compared to the group that stayed on their original drug. You're citing his labs failing on the current approach — I'd cite that switching itself isn't the risk-free upgrade it can sound like. It's trading one active risk for a different one, not removing risk from the equation.
Both of those risks are real, but I don't think they're the same shape. The discontinuation risk from switching happens once, at a moment we'd be watching closely and could catch early. Staying on a drug actively worsening his glycemic control on top of already-established coronary disease is an open-ended cost that compounds every month and doesn't announce itself the way a bad reaction to a new drug would. His risk isn't a future hypothetical the way C.W.'s still is — he already has established coronary disease, and I'd want a real switch trial seriously considered here, not held in reserve the way it was for her, precisely because his clock is running on a different, less forgiving track.
Agreed: begin a cautious cross-taper from olanzapine to aripiprazole, with metabolic labs and psychiatric symptom checks scheduled every four weeks through the transition rather than the standard interval.
The switch is reconsidered and olanzapine's psychiatric necessity is weighed again directly against his cardiometabolic trajectory, not simply reversed by default.
Aripiprazole becomes his maintenance regimen, with his cardiologist and endocrinologist both continuing active management of his pre-existing risk factors.