The obesity-drug story is usually told as a two-horse race between Novo Nordisk ($NOVO) and Eli Lilly ($LLY). That framing is increasingly incomplete. A cluster of Chinese developers now has late-stage obesity data that, taken at face value, matches the Western leaders on weight loss and beats them on tolerability. The interesting question for an investor is not whether that data exists - it does - but how much of it to believe, and what the West's response tells you.
The numbers, on paper #
Six incretin drugs from Chinese developers now have mature obesity readouts. Innovent's mazdutide (a GLP-1/glucagon dual agonist that Lilly originated and out-licensed to Innovent for China - so the "Chinese champion" is itself a Western molecule) is already approved in China and delivered roughly 19% weight loss at 60 weeks in the GLORY-2 trial, published in a top Western medical journal. Jiangsu Hengrui's HRS9531 and BrightGene's BGM0504 - both tirzepatide-class GLP-1/GIP duals - each reported about 19% in Chinese Phase 3 trials (GEMINI-1 and NCT06704581). Sciwind's ecnoglutide (13% at 40 weeks) and Gan & Lee's bofanglutide (about 17% at 30 weeks) round out the field - joined by United Laboratories' UBT251, a GLP-1/GIP/glucagon triple agonist (-19.7% at 24 weeks) that Novo Nordisk has already licensed (more below).
Click a column header to sort. A drug counts as Chinese if a China-based company developed it - so mazdutide (a Lilly molecule Innovent advanced) and UBT251 (now Novo-licensed) both belong, the same convention the trade press uses. The "Evidence" column is the tell: only two cleared an independent Western journal (JAMA, Lancet); below them sit a China-based peer-reviewed journal, a congress abstract, and - for the deepest headline numbers - unaudited press releases.
Placed next to the Western benchmarks - tirzepatide at roughly 22.5% and semaglutide at 15% - the Chinese drugs look like credible, cheaper equivalents. And on the side-effect axis they look better than credible: their trials reported adverse-event discontinuation rates of 0.7% to 2.9%, versus 6% to 7% for tirzepatide and semaglutide. On the chart below, that puts the entire Chinese cluster in the enviable top-left: high efficacy, almost no dropouts.
The tolerability number that looks too good #
A GLP-1 with 19% weight loss and a 0.7% dropout rate would be a genuine breakthrough - tolerability, not peak efficacy, is what limits real-world persistence. So the low numbers deserve scrutiny rather than applause. There are four reasons a Chinese trial can show a much lower discontinuation rate, and they are not equally reassuring.
Two are benign and mechanical. The Chinese trials ran 30 to 60 weeks against the West's 68 to 80 - fewer weeks means fewer cumulative adverse events and fewer chances to quit. Slower dose titration also improves tolerability honestly. A third is population: Chinese-only cohorts with lower baseline weight and different reporting norms for mild gastrointestinal side effects. The fourth is the one to worry about - softer adverse-event ascertainment, less independent trial monitoring, and, in at least one case, a number that has never been audited. BGM0504's 0.7% comes only from a company press release, not a peer-reviewed paper.
You cannot separate these causes from the outside. The honest position is that the Chinese tolerability edge is directional, not bankable. It is real that these drugs discontinue fewer patients; it is not established how much of that survives a 72-week, Western-monitored, apples-to-apples trial. Treat the AE-discontinuation gap as a hypothesis, not a moat - and note that ecnoglutide and bofanglutide did not disclose an AE-discontinuation figure at all.
One more read-the-fine-print item: the headline weight-loss figures use the flattering "efficacy" estimand (patients who stayed on drug), not the stricter treatment-policy estimand that includes dropouts. On the conservative basis, mazdutide is closer to 17% than 19%. Same direction of adjustment applies across the group.
The transparency gap shows up cleanly when you plot weight loss over time. The peer-reviewed drugs (mazdutide, ecnoglutide) come with a full weekly curve you can inspect; the press-release-only drugs collapse to a single dot - you cannot even see the shape of their descent. And notice how short the Chinese trials run (30-60 weeks) against tirzepatide's 72.
The real signal: the West is already buying #
Here is the part the skeptics miss. You do not have to fully trust the Chinese trial data to take these assets seriously, because the large-cap incumbents have already voted with their balance sheets. Even the two leaders have reached into China: $NOVO in-licensed UBT251, a GLP-1/GIP/glucagon triple agonist, from United Laboratories, and $LLY partnered with Laekna on LAE102, a muscle-preserving obesity antibody. Kailera Therapeutics licensed HRS9531 outside China from Hengrui and is running its own global Phase 3. AstraZeneca ($AZN) paid up to $18.5 billion to partner CSPC's obesity portfolio. Merck ($MRK) licensed an oral GLP-1 from Hansoh. Gan & Lee out-licensed bofanglutide to Lupin for global markets.
That is the tell. When Western pharma runs its own diligence and still writes nine- and ten-figure cheques for Chinese incretins - some mature, some preclinical - it is pricing the molecules as real, and re-running the pivotal trials under its own monitoring. The competitive threat to $NOVO and $LLY is therefore less about US market share and more about two things: pricing pressure in the emerging markets these drugs will reach first and cheapest, and the fact that the next wave of Western obesity pipelines increasingly originates in Chinese labs.
What would settle it #
The ambiguity resolves the moment a Chinese molecule is tested Western-style: longer, independently monitored, head-to-head against tirzepatide. Two such trials are already coming - Gan & Lee's US comparison against tirzepatide, and Kailera's global Phase 3 of HRS9531. Those readouts, not the domestic press releases, are the ones to underwrite. Until then, the right stance is engaged skepticism: the Chinese challengers are real competition and real M&A supply, but their cleanest numbers are the ones you should trust least.
We track every one of these assets, with the peer-reviewed and press-release data flagged separately, on the obesity drug comparison.