Accepted answer
They are not the same mechanism, which is why they have different time courses. Nausea is largely a central emetic phenomenon that adapts. Constipation is largely a peripheral motility plus reduced-input phenomenon, and while the motility component does attenuate somewhat, the reduced-input component gets worse over time rather than better, because it tracks your food intake and your food intake is the point of the drug.
The four contributors, and what each does over time
| Contributor | Mechanism | Trajectory over months |
|---|
| Delayed gastric emptying and slowed small-bowel transit | Vagally mediated, direct receptor effects on gut smooth muscle and enteric neurons | Partially attenuates over weeks to months |
| Reduced total food volume | Less residue enters the colon; stool volume falls; distension-triggered propulsive reflexes are triggered less often | Worsens or persists, because intake stays low by design |
| Reduced fluid intake | Less eating means less drinking, and 20-30% of daily water intake normally comes from food | Persists unless actively corrected |
| Reduced dietary fibre in absolute terms | Someone eating 1,300 kcal eats less of everything, including fibre, even at unchanged fibre density | Worsens as intake falls |
That second row is the crux, and it is worth stating plainly: a substantial part of constipation on this class is not a drug effect at all. It is an eating-less effect. Which is a consequence of the drug, but a mechanistically different one, and it does not adapt because nothing is adapting to it. The colon requires a certain volume of residue to generate the distension that triggers mass movements. Cut intake by 40% and you cut that volume, and no amount of receptor-level tachyphylaxis will replace it.
Why the difference in adaptation
The emetic response is a defensive reflex with a threshold, mediated through the area postrema, and defensive reflexes habituate: that is what they do, and it is adaptive not to keep vomiting in response to a persistent stimulus. Colonic transit is not a reflex with a threshold. It is a continuous mechanical process whose rate depends on smooth-muscle activity, enteric neural coordination, luminal content and water. There is nothing there to habituate. The receptor-level component may desensitise; the mechanical consequence of a smaller input does not.
This also explains a confusing observation: people who eat more have less constipation, and the improvement is immediate rather than gradual. That is the input term changing, not the drug wearing off.
The dose question
Constipation is dose related in the trials, but more weakly than nausea and, interestingly, non-monotonically in the same way. SURMOUNT-1 reported constipation in 16.8%, 17.1% and 11.7% of the 5, 10 and 15 mg arms against 5.8% on placebo [1]. The 15 mg arm reported less constipation than the 5 mg arm. STEP 1 reported 23.4% on semaglutide 2.4 mg against 9.5% on placebo [2].
The non-monotonicity is probably explained by diarrhoea, which is also common on this class and rises with dose more consistently than constipation does. At the top of the range a larger fraction of people sit on the loose end of the spectrum, so the constipation column falls.
What that means for you: a lower dose might help and is a legitimate conversation, but expect it to be partial, because the intake term will not change much while appetite suppression is preserved. The largest lever on constipation specifically is usually not the dose. It is the volume and composition of what you eat.
Why your fibre attempt made things worse
Predictable, and it deserves its own answer, so briefly: adding insoluble bulking fibre to a slow-transit colon with inadequate fluid gives you a larger, slower, harder mass and more gas. Bulking requires adequate water and adequate transit to help, and on this class you frequently have neither. That is not a sign you need more fibre. It is a sign bulking is the wrong lever.
What actually has support here
In rough order of expected yield, on the specific problem of slow transit with reduced intake:
- Fluid, measured rather than intended. Not because water is a laxative, which it is not in a hydrated person, but because every other intervention on this list fails without it. Your fibre attempt almost certainly failed partly for this reason.
- An osmotic agent, most commonly a macrogol or polyethylene glycol preparation. This is the best-evidenced pharmacological approach for chronic constipation generally, and it works by mechanism rather than by stimulation: it holds water in the lumen, softening content and increasing volume without depending on residue. That makes it a good mechanistic fit for a low-intake state. Osmotic laxatives have the strongest evidence base in chronic constipation of any class.
- Magnesium salts, osmotic by the same principle. Cheap, widely used, and not benign in impaired kidney function.
- Soluble, low-fermenting fibre rather than insoluble bulk, if fibre at all. Discussed in the fibre answer below.
- Movement, modest evidence, good for you anyway, will not solve this alone.
- Protein intake, indirectly. High-protein low-volume eating is characteristically constipating, and this population is correctly told to prioritise protein. The two goals conflict, genuinely rather than through poor adherence.
And the thing worth checking first: a change from daily to every three or four days over eleven months is worth mentioning to a clinician once, both to exclude the other causes of altered bowel habit and because prescription options for refractory chronic constipation are more effective than anything on a shelf.
edited 12 Sept 2024 by Dr_Priya_Raghunathan — clarified the distinction between purity and content
7Framing part of it as an eating-less effect rather than a drug effect is the reframe that made this tractable for me. – h_villanueva 7 months ago 6The 15 mg arm reporting less constipation than the 5 mg arm is a nice example of a safety table containing something counterintuitive. – mz_4113 5 months ago The protein-versus-transit conflict is real and nobody warns you about it. – dead_volume 4 months ago add a comment