Editor’s note: We talk a lot in this newsletter about which drug, which dose, which trial. We talk far less about the unglamorous work that actually determines whether a patient gets the benefit at all: staying on the medication. Most GLP‑1 discontinuations are not failures of the molecule—they are failures of side‑effect management or affordability. So I asked a registered dietitian who lives in this work every day to write down exactly how she keeps her patients on treatment. This is her playbook. — Michael Albert, MD

When a patient takes a GLP‑1 medication, what we are really after is long‑term health improvement. As responsible clinicians, we should not be satisfied by quick weight loss in the first few months. We should want to see a patient stay on the medication for the long run and achieve clinically significant weight loss, improved quality of life, and reduced risk of obesity‑related comorbidities. That means prescribing—and managing—with an eye on drug persistence rather than short‑term, observable change.

As a dietitian, my role has evolved to help patients achieve exactly that, and I can teach you what I have learned. Much of what I do is built on the pharmacokinetics and pharmacodynamics of GLP‑1s combined with the fundamentals of treating GI symptoms. Implementing these strategies, I have found that my own patients rarely discontinue a GLP‑1 for intolerance. Here is the playbook.

Set accurate expectations on weight loss. The first thing patients want to hear about is how much weight they will lose. Educate them on overall expected loss for their drug—roughly 15% for semaglutide and 20% for tirzepatide—and use those numbers to share a predicted end weight if they turn out to be an average responder. This sets realistic expectations for everyone, and especially for patients whose BMI will remain in the obesity category even after average success on a GLP‑1.

Set accurate expectations on side effects. This is the conversation I most want patients to hear. Patients need to walk into their first dose expecting nausea and constipation, and possibly diarrhea and acid reflux. Do not underplay these—if anything, err toward overplaying them. But pair that honesty with strong reassurance that you will do your absolute best to help them manage and treat whatever comes up.

Explain the usual patterns. Most patients find the first week on any dose brings the most side effects, with each subsequent week somewhat better. Many will have nausea for a few weeks on a given dose before it slowly subsides. And remind them that each dose increase tends to bring a renewed wave of symptoms—though the opposite happens too, where initiation is worse than any later increase.

After laying out the expected side effects, give a brief preview of the tools in your toolbox—the ones detailed below.

At the initiation visit, you may also choose to start prophylaxis. Some clinicians prescribe ondansetron whenever they initiate a GLP‑1; if you do, keep its constipating effect and QT considerations in mind and reserve standing use for patients who clearly need it. For patients prone to constipation, consider starting prophylactic PEG 3350, or at minimum give them a very low threshold to start it—say, the first missed daily bowel movement or the first day stools turn harder than usual.

Finally, schedule follow‑up about one week out so you can move quickly on whatever surfaces. At the very least, be reachable for a quick call or a portal message. If your practice runs that way, I encourage you to be the one to reach out if you have not heard from the patient within the week.

The goal of this phase is to help patients tolerate the medication well enough to stay on it for the long run. Given that, the main objective of your visits is simple: manage side effects.

Every other week is often sufficient until side effects are well controlled. That cadence may be necessary for a few months, after which you can space visits out. If your patient is working with a dietitian experienced in GLP‑1 care, much of this follow‑up can be handled there—but the prescriber should see the patient, or at least check in, before each dose renewal or escalation.

You must query about side effects. “You doing OK?” is inadequate. Even when a patient answers affirmatively, ask specific questions to surface symptoms they may be embarrassed to volunteer. We will happily talk about poop all day—but for many patients, that feels awkward. Here is how I ask:

Once you understand how they feel, help them spot a pattern. Many report worse nausea and constipation in the first day or two after injection, then improvement. Others feel equally rough all week.

That variation tracks the pharmacokinetics—an average ~7‑day half‑life for semaglutide¹ and ~5‑day half‑life for tirzepatide,² with some individuals at the faster end. When symptoms do follow a pattern, that's your cue for when to time anti‑nausea or constipation interventions.

The half‑life also explains why many patients feel somewhat worse over the first four weeks of each dose—there's still drug from the prior dose on board if their clearance is average or slower. Patients who feel better with each step may simply clear faster, or build tolerance more quickly than most.

GLP‑1 nausea comes from two places: receptors in the brain's vomiting center, and slowed GI motility. Treat it—both so patients can function and, crucially, so they don't discontinue. Treating vomiting matters even more. Persistent vomiting can cause electrolyte disturbances in the short run and esophageal damage in the long run, and it can drive central sensitization in which even minimal triggers provoke it. That's likely why some patients dry‑heave on an empty stomach, or gag while brushing their teeth.

Some patients like the idea of losing weight faster because they're vomiting, and resist ondansetron—or even ginger and peppermint. In reality there is almost certainly no long‑run difference in weight loss, but plenty of long‑run risk from frequent vomiting. Explain this when you firmly recommend treating it.

I tell patients we don't need nausea from dehydration stacked on top of nausea from the drug. The drive to drink falls along with the drive to eat on a GLP‑1—something also seen in the literature³—so fluid becomes something patients have to be deliberate about. For most, ~2 liters is enough; the real target is urine that stays light in color.

Temperature and carbonation are individual: cold soothes some and triggers others; fizzy helps some and nauseates others. Acceptable options include water, warm or chilled tea, seltzer, gelatin, broth, and diet‑sweetened drinks. Discourage regular soda (including ginger ale), juice, sweet tea, and sports drinks. If an electrolyte drink is warranted, choose one with ≤6 g sugar per reconstituted cup—about the concentration of oral rehydration solution—reserved for patients who are genuinely dehydrated.

Both calm nausea—ginger via prokinetic activity,⁴ peppermint by relaxing GI smooth muscle.⁵ For many patients these are the difference between unmanageable all‑day nausea and a subtle, manageable version. One caution: peppermint can worsen reflux by relaxing the lower esophageal sphincter.

A thumbnail‑sized nibble of crystallized ginger can soothe nausea on the spot and be repeated through the day; ginger tea covers both fluid and ginger needs. For peppermint, tea targets the stomach while enteric‑coated capsules reach nausea originating lower in the intestine. Peppermint candies help some patients. I'd avoid peppermint gum—chewing swallows air into the stomach.

Constipation itself causes nausea,⁶ so don't ignore it. It gets its own section below—just know the two are linked.

High‑fat foods slow GI transit and worsen nausea—a mechanism well described for dietary fat and gastric handling⁷—so they're a key dietary target. Explain it plainly: fatty food makes the gut move slower, so food sits and nausea follows. Common triggers are red meat, fried food, ice cream, and cheesy meals. Have patients avoid them, eat very small amounts, or eat them slowly to learn their tolerance.

Nighttime eating is the other big trigger—late dinners followed by morning vomiting. Normal circadian rhythm slows GI motility at night,⁸ so avoiding food in the evening (for many, after 5–6 pm) heads off late‑night nausea. A late, high‑fat meal is a “double whammy,” almost guaranteed to make a nausea‑prone patient sick.

Ondansetron relieves both nausea and vomiting.⁹ It comes as standard tablets, orally disintegrating tablets (better when swallowing itself provokes nausea), and an oral solution—useful when creativity is called for.

Teach patients to dose it ahead of a predictable pattern—not when they're already about to vomit. If mornings are the problem, take it on waking. Warn them it can cause constipation, especially with daily use, and consider starting a constipation intervention alongside it if you anticipate regular need. And remind them you don't expect them to need it forever—it's there to get them through escalation.

GLP‑1 constipation is multifactorial. Slowed motility leaves stool in the colon longer; the longer it sits, the more water is drawn out and the drier it gets. Add reduced overall intake and lower fiber intake, and constipation follows. This tends to be the long‑run problem—the one that outlasts the others.

Treatment starts with fluid, which is commonly a struggle in this group. Inadequate hydration drives constipation, so adequate intake—at least ~2 liters—is a mainstay.¹⁰

If a patient is already constipated, PEG 3350 for a few days will empty the colon. Once they're having regular, normal‑consistency movements, many can drop to every other or every third day, and some transition to a fiber supplement alone for maintenance.

Psyllium husk keeps constipation at bay once the colon has been cleared with PEG 3350—or you can start psyllium with the GLP‑1 to prevent the near‑inevitable constipation. Pair fiber with adequate water, or it can paradoxically make things worse.

My preferred routine: start psyllium once daily in the morning, then after a few days add a midday or early‑evening dose, up to three times daily. Some patients don't tolerate fiber at night—likely the combination of circadian and drug‑related slowing—so the third dose should land in the early evening, not late.

For bloating from psyllium, methylcellulose is my next choice. For patients put off by the thicker texture of either, wheat dextrin is worth a try.

Set expectations about texture. Psyllium and methylcellulose are powders mixed in water that thicken if they sit—so drink them immediately (unlike PEG 3350, which clears with a few minutes of mixing). Some patients mix psyllium with cranberry juice as a “jelly” over yogurt, or blend it into a smoothie—always with or immediately followed by a full cup of water. Once someone has experienced real constipation, they tend to make peace with the consistency.

Fiber capsules are impractical—roughly 15 psyllium capsules equal one powder dose, which gets expensive fast.

As a dietitian, I'd be remiss not to mention fiber—but it's genuinely hard to hit adequate fiber from food during escalation, when nausea and low appetite get in the way. Once water and protein are covered, patients can start prioritizing legumes, fruits, vegetables, and whole grains. Kiwifruit¹¹ and prunes¹² have specific evidence for constipation—great places to start, alongside other interventions or as prevention.

Some patients get slight help from docusate and sennosides—usually those who won't take PEG 3350 (taste) or fiber (texture). But patients relying on docusate and sennosides alone consistently report ongoing constipation, because neither adequately treats it.

There's evidence for physical activity—especially walking—in constipation.^13,14 A short 10‑minute walk after a meal (eating itself stimulates motility) can prompt the urge to go. I describe it to patients as a therapeutic walk, not exercise: a slow stroll is enough to get things moving.

Constipation is more common than diarrhea on GLP‑1s, but diarrhea is hardly rare. In my practice it usually shows up at initiation and with dose escalation, whereas constipation tends to persist. Some patients have diarrhea without constipation; others alternate.

The diarrhea likely relates, at least in part, to bile‑acid malabsorption: when bile reaches the colon instead of being reabsorbed in the small intestine, it pulls in water and electrolytes and triggers cramping. (I'd hold this as the leading working mechanism rather than settled fact.)

Cutting high‑fat foods—meat, fried food, cheesy meals—can blunt bile release and the resulting diarrhea. Smaller portions and smaller meals overall help too.

Sugar alcohols are a common and under‑recognized trigger—and patients often choose the very products that contain them (diet drinks, protein bars, “sugar‑free” items) for the calorie savings. Other patient‑reported triggers include dairy, juice, soda, and cruciferous vegetables. I wouldn't have patients avoid these by default—better to identify true triggers through observation.

Fluid won't stop the diarrhea, but it's essential for preventing and treating the dehydration it causes—these patients may need more than 2 liters. Water alone suffices for most. Many feel better symptomatically with electrolyte drinks, and those with more severe or persistent losses benefit from solutions resembling oral rehydration solution. A simple homemade ORS is 2 Tbsp sugar, ½ tsp table salt, 4 cups water; it's most palatable chilled and, in mild‑to‑moderate dehydration, can rehydrate as effectively as IV fluids if the patient keeps it down. Where available, a commercial ORS removes the risk of mixing the ratios wrong.

My patients often reach for OTC bismuth subsalicylate or loperamide. Bismuth helps attenuate diarrhea for most. Loperamide should be used for only a day or so—it slows motility and can flip the patient into severe constipation.

For persistent diarrhea, or patients who alternate, a prescription bile‑acid sequestrant can help and can be used proactively in those who tend toward diarrhea.

Fiber works like a sponge: in constipation it holds water in the colon; in diarrhea it soaks up the excess. Because it cuts both ways, it's a useful adjunct whether a patient runs toward one symptom or alternates between both.

Reflux is common on GLP‑1s. Delayed gastric emptying leaves food in the stomach longer, and the resulting pressure pushes contents past the lower esophageal sphincter. In patients who vomit frequently, the esophagus can stay irritated, producing baseline pain or burning.

Smaller meals are key to avoiding the gastric fullness that drives reflux. Limiting irritants like spicy food and coffee helps somewhat.

Ginger's prokinetic effect aids gastric emptying, which helps reflux. Peppermint is double‑edged: it relaxes stomach muscle (reducing upward pressure on the sphincter) but also relaxes the sphincter itself (making reflux easier). My patients tend to find ginger reliably helpful for reflux, while peppermint helps some and worsens it in others.

Calcium carbonate is the go‑to PRN for many patients and settles reflux quickly. For those with constant reflux, famotidine can lower the baseline, with calcium carbonate layered on as needed.

The point of GLP‑1 therapy is to treat the chronic disease of obesity. Because the disease is chronic, we aim to keep patients on treatment for the long run—keep that in mind when you prescribe, and say it out loud before patients start and at every dose conversation.

Patients experience escalation in a few common ways. Some feel renewed side effects at every step—occasionally as bad as week one of dose one, often not quite that bad. Others sail through with no renewed symptoms. All of these are normal and expected.

For those with renewed symptoms, many feel OK by week four of a given dose and are ready to move up. If not, it is entirely reasonable—probably wise—to repeat a dose level a second or even third round before increasing. The goal is persistence, not the fastest route to the top dose or to weight loss. Slightly quicker weight loss is not worth trading away tolerability. I want patients feeling at least “pretty OK”—manageable nausea, managed constipation, minimal vomiting—before I escalate. And as you go up, keep setting accurate expectations, paired with the reassurance that you can help them through it.

As months two, three, and beyond arrive, patients may fixate on the scale and worry that they have not lost enough. They'll tell me, worried, that they lost nothing on the first dose, or only 3 pounds in two months. Reassure them: many people lose no weight on the first two doses of semaglutide, and only minimal weight on the first dose of tirzepatide. Little or no loss over the first two months is normal.

As time passes, calculate their percent loss and compare it to expectations—about 15% on semaglutide and 20% on tirzepatide, at top dose, after more than a year. Always interpret progress in the context of current dose and total time on drug.

If weight is dropping too low, stop escalating; if they're at the top dose, step it down. “Too low” isn't defined by BMI alone—a careful weight history may reveal a stable pre‑gain baseline. Either way, use clinical judgment, the patient's functional status, and how they actually feel to guide the conversation about a healthy weight for them.

By one to two years, most patients are near their maximum loss. This is not the time to stop. Obesity is a chronic disease and is meant to be treated as one. The goal is persistence—fewer comorbidities and lower risk of new ones—not just a number on the scale.

The field has seen an influx of poor-to-mediocre care over the past few years. You have an opportunity to set yourself apart.

Our patients—so often blamed and shamed for their weight—deserve care that is both high-quality and compassionate. That means more than writing a prescription. It means thoughtful, attentive management of side effects and long-term support that allows these medications to do what they are meant to do: improve health and help patients thrive for years to come.

REFERENCES