thinking through the mechanism here because the SURMOUNT trials reported constipation at something like 17% but I don’t have a clean head-to-head with sema at equivalent weight-loss doses in front of me. GLP-1 slowing gastric emptying is well characterized. What’s less clear to me is whether the GIP agonism in tirzepatide adds anything on top of that, or whether it’s neutral for colonic motility. GIP receptors are expressed in enteric tissue but most of the motility literature is GLP-1 focused, and I haven’t seen colonic transit time data that isolates the GIP contribution specifically. if the constipation burden is proportionally worse on tirz than sema at comparable weight loss, that would suggest GIP is doing something downstream. if it’s about the same, then it’s probably just the gastric emptying effect and you’d expect interventions to generalize. the metformin switch also complicates things anecdotally, because metformin’s diarrhea mechanism (bile acid reabsorption, microbiome shifts, maybe AMPK in gut tissue) is different enough that stopping it doesn’t just restore baseline motility, you’re overlaying a new mechanism on a gut that was operating abnormally for years. anyone know of published colonic transit data comparing the two? that seems like the actual question.
edit: realized I said that wrong
the inference framework here is reasonable but “comparable weight loss” is where it breaks down. tirz achieves greater weight loss than sema at doses with different GLP-1 receptor occupancy profiles, so any head-to-head at matched weight loss is actually comparing mismatched doses, which confounds the GIP attribution logic pretty badly. you can’t cleanly isolate the GIP contribution when the GLP-1 component isn’t held constant. separately, there’s some preclinical signal that GIP agonism may actually oppose GLP-1 on certain gut motility endpoints, not compound it, which would cut against the “GIP adds downstream constipation” hypothesis rather than support it. i went looking for colonic transit time data isolating the GIP arm specifically and came up mostly empty, same as you, which is the actual gap. the SURMOUNT vs STEP comparison is also messy because trial populations aren’t matched.
The dose-matching critique is fair, and “mismatched receptor occupancy at equivalent weight loss” is a genuinely tricky confound, I’m not trying to wave that away. But the preclinical signal you’re citing on GIP opposing GLP-1 at gut motility endpoints, I’d be cautious about leaning on that too hard, bc GIP receptor expression in human enteric tissue doesn’t map cleanly onto the rodent models that most of that work comes from, and “opposing on certain endpoints” in vitro or in animal gut preps doesn’t necessarily tell us what the net colonic transit effect looks like in people on chronic dosing. It’s possible GIP is attenuating the GLP-1 motility effect, it’s also possible those preclinical endpoint differences wash out or reverse in human tissue, and we don’t really have the mechanistic clarity to use that as a counter-argument yet. The gap in human colonic transit data is the same problem from a different angle: we’re both reasoning from incomplete evidence, and I’d rather hold the GIP-opposes hypothesis as “plausible but unconfirmed” than use it to explain away the SURMOUNT constipation numbers, bc that feels like picking the mechanistic story that fits rather than the one we’ve actually demonstrated.
one angle that hasn’t come up: colonic transit time is only part of the picture. the motility literature tends to focus on gastric emptying and upper GI, but stool water content and colonic fermentation patterns matter too, and those are more likely to be downstream of the microbiome and bile acid shifts priyahalliday mentioned re: metformin. so even if we eventually get human colonic transit data that isolates GIP contribution, it still won’t fully explain the constipation phenotype, bc constipation isn’t just slow transit. the other thing i’d add: my n=1 is that the constipation burden on tirz was worse at lower doses for me than at higher doses, which is the opposite of what you’d expect if gastric emptying slowdown is the primary driver (since that effect tends to attenuate as GI adaptation occurs). that’s not evidence of anything, but it made me less confident that “it’s just gastric emptying” is the complete story. afaik there’s no published colonic transit data comparing the two at matched endpoints. that absence is itself informative, bc it means the mechanistic argument is running ahead of the measurement infrastructure.