EVIDENCE MATRIX / FOUR FILES
Compare the Question, Not Just the Number
Mechanism, population, endpoint, and evidence maturity determine whether results can be compared—and what conclusions survive the comparison.
Start with the study question
These four compounds are often grouped under “metabolic peptides,” but they do not answer one shared clinical question. Semaglutide and tirzepatide are incretin-based approved medicines with large diabetes and weight programs [5][6]. Tesamorelin stimulates the growth hormone axis and was tested for excess abdominal fat in adults with HIV-associated lipodystrophy [11][15]. Retatrutide is a triple-receptor investigational agent with phase 2 results [17][20].
A fair comparison starts by asking what the investigators measured. Whole-body weight change is not the same as visceral-fat area, liver-fat fraction, a cardiovascular event, or a kidney composite. A placebo-controlled trial estimates an effect against no active treatment; a head-to-head trial directly compares options within one protocol; a safety review integrates a broader record; a community report describes experience without proving cause.
The table below is therefore an evidence map, not a winner board. It shows where direct comparison is strong, where only indirect context is available, and where later trials still have to do the work.
Evidence matrix
| Compound | Main receptor system | Best-established studied frame | Evidence maturity | What the current record can establish |
|---|---|---|---|---|
| Semaglutide | GLP-1 receptor | Weight management plus cardiovascular and kidney outcomes in defined high-risk groups | Approved; large pivotal and outcome trials; post-approval safety record | Causal efficacy for studied endpoints, including weight change, cardiovascular events, and kidney outcomes [2][3][4] |
| Tirzepatide | GIP + GLP-1 receptors | Type 2 diabetes and chronic weight management | Approved; phase 3 programs; direct active comparison; growing surveillance | Strong weight and glucose efficacy, including greater mean weight loss than semaglutide in SURMOUNT-5 [1][8][9] |
| Tesamorelin | GHRH receptor and downstream GH/IGF-1 axis | Excess abdominal fat in adults with HIV-associated lipodystrophy | Approved for a focused indication; pivotal and follow-up trials; pooled analysis | Reduction in visceral adipose tissue in the studied HIV population, with evidence on hepatic fat and reaccumulation after withdrawal [10][12][14][15] |
| Retatrutide | GIP + GLP-1 + glucagon receptors | Investigational obesity, type 2 diabetes, and metabolic liver-disease studies | Phase 2 published; phase 3 confirmation pending | Large preliminary signals in weight, glucose, and liver fat, without an approved or post-approval record [19][20][21] |
Where direct comparisons hold
The strongest direct comparison on this desk is SURMOUNT-5 [1]. It placed tirzepatide and semaglutide inside the same 72-week protocol in 751 adults with obesity but without type 2 diabetes [1]. Mean weight change was -20.2% with tirzepatide and -13.7% with semaglutide [1]. Because the population, calendar, and assessment framework were shared, the inference is cleaner than comparing separate placebo-controlled programs.
That result is still endpoint-bound. It establishes greater mean weight reduction under the trial conditions. It does not show that tirzepatide has superior kidney or cardiovascular outcomes; semaglutide's FLOW and SELECT trials directly tested those different questions [2][3]. SURPASS-2 also directly compared tirzepatide with semaglutide in adults with type 2 diabetes, finding larger glucose and weight reductions for tirzepatide at the studied regimens over 40 weeks [9].
Tesamorelin and retatrutide do not belong in those direct rankings. Tesamorelin trials measured visceral-fat change in HIV lipodystrophy [12][15]. Retatrutide's phase 2 program has no completed head-to-head trial against the approved drugs in this reference set [17][20].
Safety evidence changes with maturity
Common adverse effects can appear in pivotal trials; uncommon events and patterns may require far more exposure. Semaglutide has the broadest mature safety literature here, with gastrointestinal intolerance and biliary disease established while some cancer signals remain unresolved [5]. Tirzepatide has pooled randomized evidence showing an increased composite gallbladder or biliary risk but no statistically significant pancreatitis increase in that analysis [7].
Tesamorelin's key cautions follow its GH/IGF-1 mechanism, focused indication, and limited long-term generalizability. Its liver-safety monograph found it unlikely to cause clinically apparent liver injury, while pathway-specific concerns remain relevant [11]. Retatrutide's phase 2 trials identify gastrointestinal events and a dose-dependent heart-rate increase, but cannot yet supply rare-event or long-duration certainty [20][21].
Real-world sources serve different functions after that. Spontaneous reports can generate safety signals but often lack a reliable denominator. Routine-care cohorts can test effectiveness in less selected populations but remain vulnerable to confounding. Community reports are anecdotal, not clinical evidence; they are useful for hearing what participants notice, not for calculating incidence or making causal claims.
The responsible takeaway
Evidence maturity is not a synonym for effect size. Semaglutide's chief strength on this desk is breadth: multiple large outcome programs and a mature safety discussion [2][3][5]. Tirzepatide's distinctive strength is direct comparative efficacy within a robust phase 3 program [1][8][9]. Tesamorelin's strength is specificity: trials repeatedly address visceral fat in a defined HIV-lipodystrophy population [10][14][15]. Retatrutide's strength is the size and consistency of its early metabolic signals, paired with a large remaining confirmation burden [17][19][20][21].
The most defensible comparison preserves those differences. It avoids importing a trial result into an unstudied population, avoids treating a surrogate as a clinical outcome, and avoids turning a surveillance signal into a proven causal effect. That discipline is the difference between a list of impressive numbers and a usable literature review.