Dizziness gets one line on a side-effect list and covers at least four distinct events: too little fluid, too little glucose, too little blood pressure on standing, and too little food. Each has a different number behind it, a different time course, and a different response. Sorting them matters more than the total, because three of the four are fixable at home and the fourth is a prescription review nobody on a cash plan is offered — the kind of gap also visible in what an intake fails to ask.
What the trials recorded
In the pooled semaglutide 2.4 mg weight-reduction trials, dizziness was reported by 8% of 2,116 treated adults and 4% of 1,261 on placebo.[1] A separate table covering the 7.2 mg dose reports dizziness — including postural dizziness — at 1% on placebo, 5% at 2.4 mg and 6% at 7.2 mg.[1]
The tirzepatide program recorded 2% on placebo and 4%, 5% and 4% at 5, 10 and 15 mg.[2] That flatness is the first useful clue. In the same table and the same patients, diarrhea climbs from 19% to 23% across those doses; dizziness does not move. Whatever is producing it is not scaling with exposure the way the gastrointestinal effects do.
The label also reports a separate composite for low pressure itself — blood pressure decreased, hypotension and orthostatic hypotension pooled — at 0% on placebo and 1%, 1% and 2%across the three doses.[2] An order of magnitude below the dizziness figure, which is the second clue.
The blood-pressure effect is real, and larger than most people think
An individual-patient-data meta-analysis pooled three randomized trials of semaglutide 2.4 mg over 68 weeks, covering 3,136 participants — 2,109 treated and 1,027 on placebo. The difference in systolic change between arms was −4.95 mmHg (95% CI, −5.86 to −4.05).[3]
A much larger meta-regression across 184 trials of GLP-1 receptor agonists and SGLT-2 inhibitors in type 2 diabetes put the class effect lower, at a weighted mean difference of −2.856 mmHg systolic and −0.898 mmHg diastolic, both P < 0.001. Its more useful result is the mechanism: HbA1c reduction was not independently associated with either systolic or diastolic change, while weight reduction was, at β = 0.821 for systolic and 0.287 for diastolic.[4] The pressure follows the weight, not the glucose. The cardiovascular consequences of that are in the heart article.
Two things about that fall that reverse the obvious reading
The first is who it happens to. The expectation going in was that a pressure-lowering effect would be larger in people with higher baseline pressure. It was not. Against −4.95 mmHg overall, the reduction was −4.78 in participants with hypertension, −4.09 in those with baseline systolic above 140 mmHg, and −3.16 in apparent resistant hypertension — where the confidence interval, −8.69 to 2.37, crosses zero.[3] The fall is essentially flat across baseline pressure, and smallest in the group that needed it most.
The second is that the measured number understates the drug. Over the same trials, the antihypertensive treatment intensity score decreased in the semaglutide arm relative to placebo, by 0.51 (95% CI, 0.71 to 0.32 lower).[3] Participants were coming off blood-pressure medication as they went, and the observed 4.95 mmHg is what was left after that unwinding. The authors say as much, and attribute the pressure reduction substantially to weight loss.
Put those together and the blood-pressure literature stops being an explanation for dizziness and becomes something more uncomfortable. The trials produced their modest pressure numbers because a clinician was watching and reducing other prescriptions. A patient on two antihypertensives who buys a compounded vial online, loses fifteen percent of their body weight, and has nobody reviewing the rest of their medication list is not in the same situation as the trial participant whose intensity score fell by half a point.
Why the timing points at fluid instead
If a falling pressure were the dominant cause, dizziness should grow as weight is lost, because that is what the pressure follows. The reported timing runs the other way. A pharmacovigilance analysis of 28,953 neurological adverse-event reports across six agents found dizziness, presyncope and hypoglycemic unconsciousness among 19 identified signals, with a median time to onset of 32 days (interquartile range 7 to 122) and 45.28% occurring within 30 days of starting.[5]
Thirty days into a titration is when almost no weight has been lost and almost all of the gastrointestinal effect has arrived. Six randomized trials of tirzepatide in adults without diabetes give the size of that effect: vomiting at a relative risk of 5.94 (95% CI, 4.50 to 7.85), diarrhea at 2.92 and nausea at 3.11 against placebo.[6] Both labels carry a warning for acute kidney injury due to volume depletion and state that most reported cases followed gastrointestinal reactions causing dehydration; in the pooled tirzepatide trials, acute kidney injury was reported in 0.5% of treated patients against 0.2% on placebo.[2] The bowel half of that is covered in the diarrhea article, and what to do during an acute illness in the sick-day article.
Too little food is a mechanism in its own right
The fourth candidate needs no pathology at all. A randomized crossover trial gave 30 adults with obesity 12 weeks of semaglutide escalated to 1.0 mg and measured what they ate when offered free choice. Energy intake at lunch fell 1,255 kilojoules below placebo (P < .0001), and across every free-choice meal in the day the total fell 24% — a difference of 3,036 kilojoules — with less hunger, fewer cravings, and resting metabolic rate adjusted for lean mass unchanged.[7]
A quarter less food is also a quarter less of whatever fluid, sodium and carbohydrate arrived inside it, and a person who has skipped a meal without noticing is lightheaded for the most ordinary reason there is. That version of the symptom responds to eating something, which is both the cheapest test available and the reason a structured eating pattern is worth keeping — the subject of the diet article.
Hypoglycemia is a co-prescription problem, not a drug problem
In a trial of semaglutide in adults with type 2 diabetes and a body-mass index of 27 or above who were not on insulin, hypoglycemia below 54 mg/dL was reported by 6% against 2% on placebo, across 403 and 402 patients.[1] The tirzepatide equivalent was 4.2% against 1.3%.[2]
Without diabetes, the picture is both smaller and less well measured. The tirzepatide trial in adults without type 2 diabetes recorded plasma glucose below 54 mg/dL in 0.3% of treated patients and none on placebo, and states there was no systematic capturing of hypoglycemia.[2] The semaglutide label says the same of its weight-reduction trials, and reports from a cardiovascular outcomes trial of 8,803 treated and 8,801 placebo patients that serious hypoglycemia occurred in three semaglutide-treated patients against one on placebo.[1]
Two groups sit outside that reassurance. Both labels direct considering a dose reduction of a concomitant insulin or insulin secretagogue such as a sulfonylurea when starting, because the risk of hypoglycemia rises when they are combined.[1][2] And patients with a history of bariatric surgery, already a risk factor, had more serious hypoglycemia on semaglutide than placebo: 2.3% (2 of 87) against 0% (0 of 97).[1] Those specific interactions are set out in the insulin-user article.
Heart rate moves in the direction that complicates all of this
A drug producing symptomatic low pressure would be expected to leave the pulse compensating upward, and the pulse does rise — but it rises in everyone, not only in the dizzy. Semaglutide produced mean increases in resting heart rate of 1 to 4 beats per minute against placebo, with maximum increases from baseline of 10 to 19 bpm in 41% against 34% and of 20 bpm or more in 26% against 16%.[1] Tirzepatide produced a mean increase of 1 to 3 bpm against none on placebo.[2]
A chronotropic effect that common cannot be read as a barometer of anyone in particular, and the labels direct monitoring heart rate at intervals rather than treating it as a symptom. A resting pulse that has climbed along with lightheadedness is still worth a prescriber’s attention, particularly in the age group covered in the article on adults over 65, where a dizzy episode is a fall rather than an unpleasant minute.
What none of this decides
Dizziness on these drugs is roughly twice as common as on placebo and remains a minority experience. It does not scale with dose, the hypotension-specific terms run far below it, and its reported onset sits in the window when fluid losses are at their worst — which points at volume and intake rather than at hemodynamics for most people. Fluid, salt where a prescriber permits it, and standing up slowly address that version. The general tolerability context is in what the trials recorded.
The blood-pressure finding still matters, in a different place than expected: it identifies the person whose antihypertensives should be reviewed as the weight comes off, and that review is exactly what a prescription filled through a cash telehealth plan does not include. Every figure above comes from randomized trials of FDA-approved product, while most sellers on this site dispense compounded versions that are not FDA-approved and are not reviewed by the FDA for safety, efficacy or quality before they are dispensed.
Fainting, dizziness with chest pain or palpitations, confusion, sweating and shakiness together, or lightheadedness severe enough to prevent standing are not the ordinary version of this. Each of those belongs with a prescriber or an emergency service the same day rather than with more water.