Tiredness is one of the most common things people say about these drugs and one of the least visible things in the trials that tested them. That is not a contradiction to be resolved in favor of whichever side sounds more trustworthy. It is a statement about two measuring instruments that are blind in different places, and the honest version of this page is about which number came from which. The broader incidence picture in what the trials recorded is dominated by gastrointestinal terms for the same reason.
What the randomized record says
The largest synthesis of obesity pharmacotherapy to date pooled 262 randomized trials covering 99,791 participants across 19 drugs, with follow-up from 12 to 172 weeks, and it reported fatigue as one of its 24 outcomes.
Fatigue risk did increase, and the analysis names where. The largest signal belonged to naltrexone-bupropion, at a risk ratio of 8.9 — an absolute increase of 331 more cases per 1,000 people over a year. Orforglipron followed at 3.4, or 100 more per 1,000, and cagrilintide-semaglutide at 3.2, or 92 more per 1,000.[1]
Those three are a non-GLP-1 combination, an oral small-molecule agonist and a dual-molecule combination. Once-weekly injectable semaglutide and tirzepatide — the two products nearly every telehealth prescription is written for — are not among the agents the analysis names as raising fatigue risk. The signal sorts by agent rather than by drug class, which is the finding most consumer coverage flattens away.
The result in the same paper that complicates it
The same analysis looked at quality of life across 43 trials and 45,663 participants, and found that no drug improved it past the threshold anyone would notice: every mean difference came in under 5 points against a minimally important difference of 10.[1]
That sits awkwardly beside both stories. If these drugs were producing widespread exhaustion, quality-of-life scores would have dropped, and they did not. If losing a fifth of body weight were transforming how people feel day to day, those scores would have risen, and they did not do that either. The instrument recorded very little movement in either direction, which is itself worth knowing before reading a weight figure from the tirzepatide results as a promise about energy.
Why both instruments under-read this symptom
A trial captures adverse events by coding what participants volunteer. Gastrointestinal events get volunteered because they are acute, unusual and easy to name. Tiredness in a population that is dieting, exercising more and being weighed every few weeks is diffuse, expected, and easy not to mention — and no trial in this class made it a prespecified endpoint with an instrument designed to detect it.
Spontaneous reporting databases fail in the opposite way. They accept any report from anyone, which is why they surface effects trials miss, and they have no denominator at all — the semaglutide analysis that produced the clearest gastrointestinal signals worked from 5,442 case reports with no way to say how many people were taking the drug.[2] Such a database can rank how distinctive a symptom is. It cannot say how common one is, and a page quoting it as a rate is misreading its own source.
The calorie deficit, tested directly
The deficit is real and it is large. In a 20-week randomized trial, energy intake at a free-choice meal was 35% lower on semaglutide 2.4 mg than on placebo — 1,736 against 2,676 kilojoules, P < 0.0001 — alongside reduced hunger and increased fullness.[3]
Whether a sustained deficit makes people tired has been tested on its own. CALERIE 2 randomized 220 healthy adults to two years of 25% calorie restriction or to eating freely. The restricted group lost 7.6 kg against 0.4 kg, and their scores went the other way from the folk expectation: mood improved, tension fell, general health improved with an effect size of 0.75, and sleep duration improved at month 12. Greater percentage weight loss was associated with increased vigor.[4]
The trial’s own conclusion was that calorie restriction had some positive effects and no negative effects on health-related quality of life. Three caveats belong with it: the cohort was not obese, the deficit was deliberate and gradual rather than imposed by nausea, and nobody in it was titrating a drug. A deficit produced by appetite suppression during dose escalation is not the same exposure, and the shape of that phase is set out in the timing article.
Dehydration has never been measured here
No trial of these drugs has published what happens to fluid intake when meals shrink by a third, and a large share of daily water arrives with food. The claim is plausible and untested.
The closest indirect check runs through the kidney, because sustained volume depletion shows up as acute kidney injury. A cohort drawn from a global network matched 12,123 adults with obesity and without diabetes starting a GLP-1 receptor agonist against 12,123 who started no glucose-lowering drug, and found the treated group at lower risk of acute kidney injury over five years.[5] That result should be read with its own warning label: the same analysis reports an all-cause mortality hazard ratio of 0.23, a 77% reduction that no randomized trial has come close to, which is the classic fingerprint of healthy-user selection rather than of drug effect. What it can support is narrow — mass clinically significant dehydration is not showing up — and dehydration during a stretch of vomiting is a separate and real problem handled in the sick-day article.
Protein and lean mass are a better-founded candidate
The same network meta-analysis reports that tirzepatide reduced fat mass the most, by 25.7%, and also reduced lean mass the most, by 8.3%.[1] Losing skeletal muscle while eating a third less protein is a mechanism with an obvious relationship to how strong someone feels climbing stairs, and it is measurable in a way fatigue is not. The body-composition figures are in the muscle article and the intake question in the diet article.
What does not exist is a trial connecting the two. The 2026 international consensus on nutritional support during treatment produced 52 statements and states plainly that they were primarily derived from indirect evidence, with direct evidence urgently required.[6] Protein targets attached to a telehealth prescription are borrowed from bariatric practice, not measured in this population.
The hypoglycemia case is narrower than it sounds
Low blood sugar is the explanation most often offered for mid-afternoon exhaustion, and for most people buying these drugs it is the wrong one. A meta-analysis of 144 placebo-controlled trials found no increase in any hypoglycemia with a GLP-1 receptor agonist used alone (RR 1.77; 95% CI, 0.91 to 3.46), added to metformin (1.24; 0.80 to 1.91), or in triple therapy (1.13; 0.67 to 1.91). Severe hypoglycemia was rare and not increased in any comparison.[7]
The reversal is in the same table. Metformin monotherapy did raise the risk of any hypoglycemia (1.73; 1.02 to 2.94), and starting two agents simultaneously raised it further (3.56; 1.79 to 7.10).[7]The hazard tracks the rest of the regimen rather than the GLP-1, which is part of why the contraindication list in the eligibility article is a question about a full medication list.
Even the co-prescription case behaves unexpectedly. SCALE Insulin randomized 396 adults with overweight or obesity and insulin-treated type 2 diabetes to liraglutide 3.0 mg or placebo for 56 weeks. Weight change was −5.8% against −1.5% (treatment difference −4.3%; 95% CI, −5.5 to −3.2), and more hypoglycemic events occurred on placebo than on liraglutide despite a treat-to-target protocol, because insulin requirements fell.[8] The risk in that combination is manageable by adjusting the insulin, which is a prescriber’s job and not something a checkout page performs.
What would close the gap
A trial that asked about fatigue on a validated instrument, at fixed intervals, separately during escalation and at maintenance, with protein intake and body composition recorded alongside. None exists. Until one does, the accurate statement is that the randomized record shows fatigue rising with specific agents that are not the weekly injectables, that the reporting databases cannot produce a rate, and that the most plausible mechanisms are consequences of rapid weight loss rather than of receptor occupancy.
Two limits close this out. Every figure above was generated with branded product at labeled doses, while most sellers reviewed here dispense compounded versions, which are not FDA-approved and are not reviewed by the FDA for safety, efficacy or quality before dispensing. And fatigue that arrives with shortness of breath, a racing heart, fainting, or an inability to keep fluids down is not the ordinary version of this and warrants prompt medical contact rather than a protein shake.