Two drugs that both blunt appetite, both raise resting pulse, and both get taken every day by the same people is a combination worth having evidence about. There is almost none. What exists instead is a pair of labels that never mention each other and a physiology that moves in two directions at the same time, which is a harder thing to summarize than either a green light or a warning.
The overlap is larger than the literature suggests
Attention-deficit/hyperactivity disorder is substantially over-represented among people seeking treatment for obesity. A systematic review and meta-analysis of candidates for metabolic and bariatric surgery pooled 14 studies covering 24,455 adults and found a prevalence of 8.94% and 9.90% depending on the case definition used, alongside 28.73% across three studies of 299 adolescents. The authors put that at roughly three times the general-population rate in adults and six times in adolescents, while rating the quality of the underlying studies moderate to poor.[1] Whatever the precise figure, the population buying a GLP-1 online and the population holding a stimulant prescription overlap heavily.
The co-prescribing literature, counted
Searching PubMed for tirzepatide together with methylphenidate returns nothing. Searching for the GLP-1 receptor agonist class against attention-deficit/hyperactivity disorder returns nothing. Searching for semaglutide with methylphenidate returns a single record, and it is a review of Alzheimer’s disease drug trials in China. No randomized trial, no pharmacokinetic study, and no cohort has examined this pair.
The labels are equally silent. The words amphetamine, methylphenidate and stimulant appear zero times across the semaglutide and tirzepatide weight-management labels. The lisdexamfetamine label’s table of clinically important interactions names monoamine oxidase inhibitors, serotonergic drugs, CYP2D6 inhibitors and agents that alter urinary pH.[2] No incretin appears on it, and no incretin would be expected to: amphetamine is not a CYP substrate of consequence, and lisdexamfetamine is a prodrug converted after absorption by red blood cells rather than in the gut, so a delayed stomach changes when it arrives and not how much of it does.
Both suppress appetite, on different clocks
The stimulant effect on weight is real, dose-dependent and documented in the label rather than inferred. In a four-week controlled trial in children aged 6 to 12, mean weight change from baseline was −0.9, −1.9 and −2.5 pounds at 30, 50 and 70 mg against a one-pound gain on placebo; in adolescents the same four weeks produced −2.7, −4.3 and −4.8 pounds against a two-pound gain.[2] Decreased appetite and decreased weight are listed among the most common adverse reactions at a rate at least twice placebo.
The same label states plainly that the drug “is not indicated or recommended for weight loss,” that “use of other sympathomimetic drugs for weight loss has been associated with serious cardiovascular adverse events,” and that its safety and effectiveness for treating obesity “have not been established.”[2] That sentence is the closest thing in either label to guidance about the stack, and it is aimed at the stimulant being used for the wrong purpose rather than at the pair.
The two effects also run on different clocks. A stimulant suppresses appetite across its dosing window and releases it when the dose wears off, which is why the evening is when stimulant-treated people eat. A GLP-1 suppresses intake continuously at a level that does not track the clock. Stacked, the result is not double suppression so much as a narrowed window in which eating happens at all — which matters mainly for the protein and fluid targets described in the protein and diet article.
Heart rate points the same way for both
Every stimulant label in the class carries the same sentence: CNS stimulants “cause an increase in blood pressure (mean increase about 2 to 4 mm Hg) and heart rate (mean increase about 3 to 6 bpm). Some patients may have larger increases.”[2] In the five-week fixed-dose adult trial in the methylphenidate label, dose-dependent mean increases of 3.9 to 9.8 bpm in standing pulse were recorded against 2.7 bpm on placebo.[3] A 2025 Cochrane review pooled 47 randomized trials covering 10,075 participants and put the heart rate increase at 3.71 bpm (95% CI 3.27 to 4.14), high-certainty evidence, with the effect sustained in trials running eight weeks or longer.[4] A separate meta-analysis in people with ADHD reached the same conclusion across 32 heart rate studies and 29 blood pressure studies.[5]
The GLP-1 labels report the same direction at a similar magnitude. Semaglutide produced mean increases in resting heart rate of 1 to 4 bpm against placebo in the weight-reduction trials, and in a pediatric trial in patients with normal baseline heart rate, maximum increases of 20 bpm or more occurred in 54% of treated patients against 39% on placebo.[6] Tirzepatide produced a mean increase of 1 to 3 bpm.[7] A network meta-analysis of eight tirzepatide trials found the increase dose-dependent: mean difference 1.82 bpm (95% CI 0.75 to 2.89) against the combined control, and 2.96 bpm (1.36 to 4.57) for the 15 mg dose against placebo specifically, with 15 mg exceeding 10 mg by 1.44 bpm and 5 mg by 2.53 bpm.[8]
Blood pressure points the other way, and that is the complication
If both drugs simply pushed the cardiovascular system in one direction the arithmetic would be easy. They do not. The Cochrane review found amphetamines raised systolic pressure by 1.93 mmHg (1.54 to 2.31) and diastolic by 1.84 mmHg (1.51 to 2.16) across 56 trials and 10,583 participants.[4] The semaglutide label’s own trial table, in the same row block that records the pulse increase, shows systolic pressure falling 6.2 mmHg on drug against 1.1 mmHg on placebo — a difference of −5.1 mmHg — alongside a heart rate change of +3.5 bpm against −0.7, a difference of +4.3 bpm.[6]
So one instrument adds and the other subtracts. A blood pressure cuff reading normal in somebody taking both drugs is not evidence that nothing is happening; it is the expected result of two opposing effects, and it says nothing whatever about the pulse underneath it. The label that cares about this asks for the right measurement: semaglutide’s instruction is to “monitor heart rate at regular intervals consistent with usual clinical practice,” to instruct patients to report palpitations, and, if a sustained increase in resting heart rate occurs, to discontinue.[6] The stimulant label asks prescribers to “monitor all VYVANSE-treated patients for potential tachycardia and hypertension.”[2] Two labels, two monitoring instructions, one pulse, and no mechanism by which either prescriber learns that the other drug exists.
What the withdrawal rates say that the means do not
Mean changes of a few beats and a few millimeters are easy to dismiss. The tolerability data are harder to. In the Cochrane analysis, participants taking amphetamines were more than twice as likely to leave the study because of adverse effects — risk ratio 2.69 (95% CI 2.13 to 3.40), an absolute increase of 4.3% over an average of one month, across 42 trials and 8,952 participants.[4] A GLP-1 dose escalation carries its own discontinuation curve. Nobody has measured what happens to either when they run at once, and the overlapping complaints — a racing pulse, poor sleep, dry mouth, reduced appetite, jitteriness — belong equally to both, which makes attribution by symptom unreliable in the one direction a reader most needs it. The same attribution problem applies to the sleep and anxiety effects covered in the mental health article.
The one dataset where both were given
The nearest published approach to this combination is not in ADHD at all. A retrospective open-label cohort in binge eating disorder compared patients prescribed semaglutide, patients prescribed lisdexamfetamine or topiramate, and patients prescribed semaglutide combined with one of those. The semaglutide-only group showed the greatest reduction in Binge Eating Scale scores, and adding the second drug did not improve on semaglutide alone — a result that held in the moderate-to-severe subgroup as well as the full sample.[9]
That is a chart review in a different diagnosis with no control group and no cardiovascular endpoint, and it cannot carry weight about whether the pair is safe. It is worth stating because it is the entire published experience of the combination, and because its one finding runs against the intuition that stacking two appetite-suppressing mechanisms produces more of the effect.
What this leaves for a cash intake
Nothing above argues for stopping a stimulant, and nothing establishes a hazard in the combination. What it establishes is that the one measurable thing the two drugs do in the same direction is raise resting pulse, that both labels ask for that measurement, and that the other vital sign moves in a way that will mask rather than reveal it. A reader taking both has an instrument available on a wrist or a phone and a baseline worth recording before the first injection.
The structural problem is that the instruction belongs to two prescribers, and a purchase made through a checkout creates neither of them. A form that asks whether a buyer takes any medication will get a correct answer about a stimulant only if the buyer thinks of it as relevant, and nothing about the GLP-1 marketing suggests it is — the pattern of intake gaps set out in the telehealth article. The stimulant label’s own instruction to avoid the drug in patients with known structural cardiac abnormalities or serious cardiac disease[2] depends on someone having asked, which is the same dependency described in the contraindications article.
Finally, the product distinction. Compounded semaglutide and tirzepatide are not FDA-approved and are not reviewed by the FDA for safety, efficacy or quality before they are dispensed, and what arrives in those vials is the subject of a separate page. The heart rate monitoring instruction quoted above is carried by a branded label; a compounded vial does not come with one, so a person stacking it on a stimulant has neither of the two sentences that would have told them what to watch. How claims here are established is set out in the methodology.