I have some version of the same conversation every week now. A patient is starting semaglutide or tirzepatide, and partway through the visit they ask the question I suspect most of us are hearing constantly: What happens if I ever stop?
For a while I gave the honest-but-incomplete answer, that we’d cross that bridge when we got there. I don’t anymore, because I think the real answer has less to do with pharmacology than with fiber, and I think most of us haven’t built our practices around that fact yet.
Here’s what happens when patients stop, and it’s blunter than most of us say out loud in the exam room. In the STEP 1 trial extension, participants who discontinued semaglutide regained roughly two-thirds of their lost weight within a year, along with a reversal of the metabolic gains that came with it. These drugs work only for as long as they’re taken. We are heading toward a moment, and industry analysts project U.S. spending on GLP-1 receptor agonists could eventually top $100 billion a year, where almost nothing in how we counsel patients, or how payers structure coverage, is built to change that arithmetic.
What I don’t think gets enough airtime, even among colleagues, is a simple biological fact: GLP-1 is a hormone our own gut already makes, released in response to what we eat. One of its strongest natural triggers is dietary fiber. When gut bacteria ferment fiber from legumes, vegetables, whole grains and fruit, they produce short-chain fatty acids that stimulate the body’s own GLP-1 secretion through the same receptor pathway the injectable drugs engage pharmacologically. In a randomized crossover study my colleagues and I ran, a single plant-based meal produced a GLP-1 response comparable to a DPP-4 inhibitor, measured against a conventional calorie-matched meat sandwich.
Real physician voices, twice a week
Free, and one click to unsubscribe.
I find that useful to sit with, because it reframes the drug and the diet as the same lever, pulled two different ways, rather than competing options. Sleep and activity modulate the same axis. The drug can open a window of appetite and metabolic regulation. Diet is plausibly what keeps that window open once the prescription ends.
We haven’t built our practice, or our payment models, around that. Any of us can prescribe the drug in under a minute. Getting a patient in front of a dietitian for structured, ongoing nutrition counseling, the piece of this the evidence says matters most for what happens after, is a different story. Coverage for medical nutrition therapy is inconsistent across Medicare, Medicaid and commercial plans, and what exists is usually capped at a handful of visits, if it’s covered at all. Sleep intervention gets even less attention than nutrition, clinically and financially, despite plausibly working through the same pathway. Most of our patients leave the office on a GLP-1 and go home to a food environment and a follow-up structure that was never built to sustain what the drug started.
I’m not raising this to relitigate whether these drugs are worth prescribing. The problem is that we’re funding the half of this equation with the least durable effect on its own, and treating the other half as optional.
A few things I’ve changed in my own practice: I now bring up fiber and food pattern at the same visit where I discuss starting a GLP-1. I document the discontinuation-regain data explicitly when I counsel patients on the long game, not just the drug’s mechanism. And when a patient tells me a dietitian referral wasn’t covered, I’ve started pushing back on that denial the way I would for a diagnostic test I considered medically necessary, because I think it is.
None of that fixes the coverage-design problem on its own. Three changes would, and I’d like to see more of us pushing for them:
- Conditioning GLP-1 coverage on concurrent, reimbursed nutrition and behavioral support, so the intervention that sustains the drug’s benefit is funded alongside it.
- Extending coverage frameworks explicitly into the multi-year maintenance phase, rather than concentrating benefit design on the acute weight-loss phase as most plans do now.
- Building quality measures around durable, multi-year metabolic improvement rather than prescription initiation, so health systems have an actual incentive to build the support infrastructure this requires.
We are about to spend more on a single drug class than most of us have prescribed in an entire career. Whether that spending buys durable health, or an indefinitely renewing prescription for tens of millions of patients, depends on choices about coverage and counseling being made right now. I don’t think the next real advance in obesity treatment is a new molecule. I think it’s deciding to fund what makes the molecules we already have keep working after the prescription ends, and that starts with something as unglamorous as fiber.
This essay is cited in the KevinMD record on GLP-1 drugs.
Hana Kahleova is an endocrinologist.
