# The Old Cheap Drug Keeps Winning

By Nora Bissett, Pricing Editor (https://glpwatchdog.org/authors/nora-bissett) | Published: 2026-08-20 | Updated: 2026-11-01

Canonical page: https://glpwatchdog.org/research/the-old-cheap-drug-keeps-winning

> A network meta-analysis of 41 pediatric trials found semaglutide producing the largest BMI reduction, and concluded an older, cheaper combination may be the better option.

## Key points

- Semaglutide −20.40% mean reduction; phentermine–topiramate −18.35%.
- But responders at ≥10%: 734 per 1,000 for phentermine–topiramate against 399 for semaglutide.
- A larger mean with fewer responders describes a different distribution, and nobody explains it.
- Second independent analysis to rank phentermine–topiramate above semaglutide once cost or tolerability enters.
- This is not a recommendation. Phentermine–topiramate is a stimulant combination with its own contraindications, including teratogenicity.
- Semaglutide, liraglutide and phentermine–topiramate are each FDA-labeled for obesity from age 12; exenatide is not.

A network meta-analysis of 41 trials in children and adolescents found semaglutide producing the largest reduction in BMI. Its authors nonetheless conclude that **phentermine–topiramate**, an older and cheaper combination, may be the optimal addition to lifestyle treatment.[1] The numbers underneath explain why, and they contain a discordance nobody accounts for.

## The two ways of measuring disagree

The review pooled 41 randomized trials covering 3,923 children and adolescents, comparing drug treatments added to lifestyle modification against lifestyle modification alone. It reported both the average reduction and how many additional people crossed meaningful thresholds.

*Against lifestyle modification alone. Reduction is in the 95th BMI percentile; responders are additional patients per 1,000 person-years.[1]*

| Drug | Mean reduction | Extra reaching ≥5% | Extra reaching ≥10% |
| --- | --- | --- | --- |
| Semaglutide | **−20.40%** (95% CI −24.22 to −16.58) | 500 | **399** |
| Phentermine–topiramate | −18.35% (−22.26 to −14.45) | **554** | **734** |
| Liraglutide | Significant, smaller | — | — |
| Exenatide | Significant, smaller | — | — |

> The larger average effect produced fewer people crossing the higher threshold.

At the 10% threshold, phentermine–topiramate produced nearly **twice** as many additional responders as semaglutide (734 against 399), even though its mean reduction was two percentage points smaller.

## What that discordance implies, and what nobody explains

> **A larger mean with fewer responders is a statement about the shape of the distribution, not the size of the effect.** It is what you get when one drug produces very large reductions in a subset while another produces moderate reductions in more people. The averages then favor the first while the threshold counts favor the second, and both are correct.

That is the most plausible reading, but it is not established. The review does not explain the pattern, and network meta-analyses can also generate discordant continuous and binary estimates for technical reasons: different studies contributing to different outcomes, different follow-up lengths, different definitions of response. We are reporting the discordance rather than resolving it.

It matters practically because the two measures answer different questions. “How much will the average child’s BMI fall” and “what are the chances this child gets a meaningful result” are not the same, and a family is usually asking the second.

## The same drug came out ahead in adults, for a different reason

This is the second independent analysis to place phentermine–topiramate above semaglutide on a measure that is not raw efficacy. The American College of Physicians’ economic review found **semaglutide to be low value compared with phentermine–topiramate** in adults; that review is covered in [we know what they do, not what they’re worth](https://glpwatchdog.org/research/we-know-what-they-do-not-what-theyre-worth).

The two analyses differ in population, method and endpoint, yet they converge on the same underrated option. Neither analysis says the older drug produces more weight loss. Both say that once tolerability or cost enters the comparison, the ranking changes.

The pediatric reviewers make the tolerability point explicitly, citing it as the reason for their conclusion. Gastrointestinal side effects and discontinuation are the recurring cost of the GLP-1 class across the research on this site. In adolescents, where treatment may run for years and adherence is harder, that cost weighs more.

## Why this is not a recommendation

> **Phentermine–topiramate is a stimulant combination with substantial contraindications of its own**, including teratogenicity, which carries particular weight in adolescents who may become pregnant. It is not a gentler alternative; it is a different drug with a different risk profile (see how every [FDA-approved appetite suppressant](https://glpwatchdog.org/research/fda-approved-appetite-suppressants) compares on its label), and choosing between them is a clinical decision about a specific young person. Taking phentermine alongside a GLP-1 is a separate question again, covered in [phentermine and a GLP-1 together](https://glpwatchdog.org/research/phentermine-and-a-glp1-together); phentermine on its own against a GLP-1 is in [phentermine vs a GLP-1](https://glpwatchdog.org/research/phentermine-vs-a-glp1).

- **The outcome is BMI, not health.** These are surrogate measures in a group whose outcomes lie decades away (see [what changed in the teenagers’ bloodwork](https://glpwatchdog.org/research/what-changed-in-the-teenagers-bloodwork)).
- **Long-term safety in growing bodies is unstudied** for every drug here.
- **Not every drug here is FDA-approved for adolescents.** Semaglutide, liraglutide and phentermine–topiramate are each labeled for obesity from age 12;[2][3][4] exenatide is not labeled for weight management.[5]
- **Trials in this population are small.** With 3,923 participants across 41 trials, the average is fewer than a hundred per trial.

For what these drugs do to weight in adolescents specifically, see [GLP-1 drugs in teenagers](https://glpwatchdog.org/research/glp1-drugs-in-teenagers); for the diabetes question in the same age group, [tirzepatide in ten- to seventeen-year-olds](https://glpwatchdog.org/research/tirzepatide-in-ten-to-seventeen-year-olds).

## Frequently asked questions

### Which weight drug works best in children and adolescents?

Semaglutide produced the largest average reduction in the 95th BMI percentile, at 20.40% against lifestyle modification alone. But phentermine-topiramate produced more additional responders at both the 5% and 10% thresholds, and the reviewers conclude it may be the optimal adjunct because of tolerability.

### How can a drug have a larger average effect but fewer responders?

It describes a different distribution: very large reductions in a subset against moderate reductions in more people. That is the most plausible reading, but the review does not explain the pattern, so it should be held loosely.

### Should a child be on phentermine-topiramate instead?

That is a clinical decision and this is not a recommendation. It is a stimulant combination with its own contraindications including teratogenicity. Like semaglutide and liraglutide, it is FDA-approved for adolescents 12 and older with obesity.

### Is this the first analysis to favor the older drug?

No. An independent economic review for the American College of Physicians found semaglutide low value compared with phentermine-topiramate in adults. The population and method differed; the underrated option was the same.

### How strong is the evidence?

The review rests on forty-one trials covering 3,923 children, fewer than a hundred per trial on average. They measured BMI rather than health outcomes, and long-term safety in growing bodies is unstudied for every drug involved.

## References

1. Luo L, Huang T, Yan Z, et al. Pharmacotherapy for Children and Adolescents With Overweight or Obesity: A Systematic Review and Network Meta-Analysis of Randomized Controlled Trials. Diabetes, Obesity and Metabolism. 2026. PMID: 42310900. https://pubmed.ncbi.nlm.nih.gov/42310900/
2. VIVUS LLC QSYMIA (phentermine and topiramate extended-release) capsules, CIV — prescribing information (section 1, Indications and Usage). DailyMed, U.S. National Library of Medicine. 2026. https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=40dd5602-53da-45ac-bb4b-15789aba40f9
3. Novo Nordisk SAXENDA (liraglutide) injection — prescribing information (section 1, Indications and Usage). DailyMed, U.S. National Library of Medicine. 2026. https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=3946d389-0926-4f77-a708-0acb8153b143
4. Novo Nordisk WEGOVY (semaglutide) injection and tablets — prescribing information (section 1, Indications and Usage). DailyMed, U.S. National Library of Medicine. 2026. https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=ee06186f-2aa3-4990-a760-757579d8f77b
5. AstraZeneca Pharmaceuticals LP BYETTA (exenatide) injection — prescribing information (section 1, Indications and Usage). DailyMed, U.S. National Library of Medicine. 2025. https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=53d03c03-ebf7-418d-88a8-533eabd2ee4f

---

Cite as: GLP Watchdog, Nora Bissett, "The Old Cheap Drug Keeps Winning", published 2026-08-20, updated 2026-11-01. https://glpwatchdog.org/research/the-old-cheap-drug-keeps-winning

Canonical HTML page: https://glpwatchdog.org/research/the-old-cheap-drug-keeps-winning

Medical disclaimer: Nothing here is medical advice. We report on this subject; we do not treat anyone. Whether to begin a medication, supplement, or treatment, whether to adjust it, and whether to stop are questions for a clinician licensed to answer them, who knows your history.
