In Post 3, I made the case that GLP-1 therapy is best understood as chronic disease management rather than a finite course of treatment — and that stopping means the weight returns. That framing raises a set of practical questions that I want to address directly here: What happens if the drug doesn't work well for you? Does your sex predict your outcome? And once you've reached your goal, how much medication do you actually need to stay there?
These are the questions that shape long-term treatment planning, and they deserve more nuance than most coverage of these medications provides.
Who Responds Best — and Why Some Don't
GLP-1 medications do not work equally well for everyone. The trials report mean weight loss, but the distribution around that mean is wide. In the STEP 1 trial, for example, while the average weight loss was 14.9%, individual responses ranged from minimal to over 25%.
We don't yet have good predictive biomarkers to tell you in advance where you'll fall on that distribution. What the data does suggest is that early response predicts later response: patients who lose at least 5% of body weight in the first 12–16 weeks of treatment tend to be responders; those who don't may be better served by dose adjustment, switching to tirzepatide if they started on semaglutide alone, or a formal reassessment of what's driving their weight.
I want to be honest about something I see in my practice: roughly 5–10% of patients are genuine non-responders in terms of weight loss, even on adequate doses of the more potent agents. This is real, and it deserves acknowledgment rather than dismissal. Interestingly, many of these patients do show meaningful improvements in metabolic markers — blood sugar, blood pressure, lipids — even without substantial weight loss. The drug is doing something biologically even when the scale doesn't move.
My working hypothesis for why some patients don't lose weight is that the pathways most responsible for their weight are distinct from the ones these medications primarily modify. Obesity is not a single disease — it has many biological subtypes, and GLP-1 and GIP receptors are not the only relevant levers. For patients who don't respond adequately after a thorough trial, I explore other medication options, and surgical evaluation is entirely appropriate for those who want it. A non-response to GLP-1 therapy is not a reason to abandon treatment of obesity — it's a reason to find the right treatment.
Early non-response is not a reason to push immediately to the maximum dose and wait indefinitely. If a patient has been on an adequate dose for 16 weeks without meaningful weight loss, a structured reassessment is warranted — exploring adherence, dietary patterns, concurrent medications that promote weight gain, underlying thyroid or hormonal contributors, and whether a different agent or surgical evaluation is more appropriate. GLP-1 therapy is not a passive prescription. It requires active follow-up.
Do Women Respond Differently Than Men?
Yes — and this is one of the most striking and underappreciated findings in the recent literature. Women consistently lose a greater percentage of body weight on GLP-1 medications than men, and this appears to be a genuine biological difference rather than a statistical artifact.
A 2026 systematic review and meta-analysis published in JAMA Internal Medicine analyzed sex-stratified outcomes across 64 randomized controlled trials involving 19,906 patients. Women lost a mean of 10.9% of body weight (95% CI 7.0–14.8%) compared to 6.8% in men (95% CI 4.6–9.0%), with 4 of 6 individual trials reaching statistical significance for this difference. Critically, sex was the only patient characteristic that demonstrated significant heterogeneity of treatment effect — efficacy did not differ meaningfully by age, race, ethnicity, baseline BMI, or baseline HbA1c. In the head-to-head SURMOUNT-5 trial comparing tirzepatide to semaglutide, weight reduction was approximately 6 percentage points greater among women than men in both treatment arms. A large Italian real-world cohort of 7,847 patients with type 2 diabetes followed for a median of 4 years confirmed that 66.5% of women versus 58.0% of men achieved at least 5% weight loss, and 40.0% versus 30.7% achieved at least 10% — independent of weight-adjusted drug doses. A Canadian real-world cohort found that female sex was independently associated with a hyper-response of greater than 15% total body weight loss, with an adjusted odds ratio of 1.92.
Two important nuances deserve mention. First, in the two trials that reported absolute kilogram losses rather than percentage losses, the sex difference was not statistically significant — suggesting that some of the percentage-based difference reflects women's lower average baseline body weight rather than a purely pharmacological advantage. Second, the proposed biological mechanisms remain under investigation but likely involve interactions between GLP-1 receptors and estrogen signaling, as well as pharmacokinetic differences in drug distribution and clearance between sexes.
The practical implication of this finding cuts in two directions. A woman who doesn't reach the average weight loss seen in mixed-sex trials hasn't necessarily failed treatment — she may be comparing herself to a benchmark skewed upward by female responders. And a man who loses less than a female partner on the same drug at the same dose isn't doing anything wrong — the biology appears to genuinely differ. I share these numbers with patients because understanding your own likely response range, not just the population average, is part of making a genuinely informed decision about treatment expectations and long-term planning.
Maintenance Dosing: Do You Need the Maximum Dose Forever?
The default clinical model for GLP-1 therapy is to escalate the dose gradually over several months until you reach the maximum approved dose, then continue indefinitely. That model made sense when the primary goal was understanding efficacy in clinical trials. It's less obviously correct as a long-term real-world management strategy, particularly when cost, tolerability, and patient preference all argue for using the minimum dose that achieves the goal.
The pharmacokinetics support a more individualized approach. Semaglutide has a half-life of approximately one week, which means it accumulates gradually to a steady state. Reducing the dose — or extending the interval between doses — doesn't eliminate drug exposure; it shifts the steady-state level. The question is how much shift is tolerable before weight begins to return.
A pharmacokinetic-pharmacodynamic modeling study found that switching from weekly to every-other-week dosing maintains approximately 75% of weight loss efficacy. Importantly, if the dose size is increased when frequency is reduced — injecting a larger amount less often rather than the same amount less often — approximately 100% of weight loss can be maintained. A retrospective case series of 30 patients who transitioned from weekly to every-other-week dosing after reaching a weight plateau found that weight actually continued to decrease slightly (74.1 kg to 72.4 kg over 36 weeks, p<0.01), with preserved body composition and metabolic improvements. The modeling authors noted that offering every-other-week dosing to twice as many patients at the same total cost would reduce national obesity burden more effectively than standard weekly dosing to fewer patients — a population health argument worth taking seriously.
The critical distinction, supported by the pharmacokinetics, is between extending the dosing interval and stopping entirely. Semaglutide's one-week half-life means that a full month without treatment results in near-complete drug washout — essentially a cold stop, with all the weight regain that entails. Extending the interval maintains some steady-state exposure. They are not the same thing.
There are currently no published studies evaluating a month-on/month-off or other cyclical stopping strategy for GLP-1 medications. The ADA 2026 Standards of Care mention "intermittent therapy" as a conceptual option but explicitly note it has not been validated in randomized trials. The ACC 2025 Expert Consensus Statement notes that missing three or more consecutive weekly doses may warrant a dose reduction upon restarting — but does not endorse planned intermittent cycling. If you want to explore dose de-escalation, do it with your physician and with active monitoring of weight and metabolic markers. Unplanned stop-start cycling is not the same as a structured maintenance protocol.
The 2026 ADA Standards of Care recommend that the treatment dose "may be less than the maximum approved dose" and endorse shared decision-making to find the lowest effective maintenance dose. I think this is the right framework. The goal of maintenance is not to be on the maximum dose — it's to be at your target weight with the fewest side effects and the most sustainable regimen. For some patients that's the full dose weekly. For others it may be a lower dose, or extended interval dosing. The evidence base for these alternatives is early, but the direction is clear, and it's where I expect the field to move as more real-world data accumulates.
The Embla Model: Lower Doses With More Support
A compelling real-world dataset from Denmark addresses a related but distinct question: what if you never escalated to the maximum dose in the first place, and instead paired a lower dose with intensive behavioral support?
The Embla program is a Danish digital weight management service that combined a "treat-to-target" semaglutide dosing approach — titrating each patient to the lowest dose that achieved continued weight loss, not automatically escalating to the maximum — with intensive behavioral therapy covering nutrition, exercise, and habit formation. In a retrospective cohort of 2,694 participants followed to week 64, mean weight loss was 16.7%, comparable to the STEP 1 trial result of 14.9% with standard maximum dosing — but achieved at a mean semaglutide dose of only 1.08 mg per week, substantially below the standard 2.4 mg maintenance dose. 98% of participants achieved at least 5% weight loss and 82% achieved at least 10%, consistent across all BMI classes.
The Embla data comes with significant limitations. It was retrospective and uncontrolled, with substantial attrition — only 17% of the original 2,694 participants remained at week 64, introducing real survivorship bias. There was no control group, and the relative contributions of lower medication dose versus the intensive behavioral support cannot be separated. Prospective randomized trials are needed to validate this approach. I find the hypothesis compelling and consistent with my clinical experience, but I hold it loosely until better evidence arrives.
The Embla model points toward what I think is the future of this field: individualized, treat-to-target dosing combined with robust behavioral and psychological support — not the current dominant model of escalating everyone to the maximum approved dose regardless of their individual response. For patients concerned about cost, side effects, or long-term commitment, this is meaningfully different from "you'll need the maximum dose forever." The right dose is the one that keeps you at your target weight with the fewest side effects and the most sustainable lifestyle architecture around it. The medication opens the door — what you build while walking through it determines how long it stays open.
Sources
- Wong M, et al. Reduced-Frequency GLP-1 Therapy Maintains Weight, Body Composition, and Metabolic Syndrome Improvements. Obesity. 2026.
- Wu CC, et al. Less Frequent Dosing of GLP-1 Receptor Agonists as a Viable Weight Maintenance Strategy. Obesity. 2025.
- Cengiz A, et al. Alternative Dosing Regimens of GLP-1 Receptor Agonists May Reduce Costs and Maintain Weight Loss Efficacy. Diabetes Obes Metab. 2025.
- American Diabetes Association. Standards of Care in Diabetes — 2026. Diabetes Care. 2026.
- Gilbert O, et al. 2025 ACC Expert Consensus Statement on Medical Weight Management for Optimization of Cardiovascular Health. J Am Coll Cardiol. 2025.
- Seier S, et al. Treat to Target in Weight Management With Semaglutide: Real-World Evidence From an eHealth Clinic (Embla). Diabetes Obes Metab. 2025.
- Wilding JPH, et al. Once-Weekly Semaglutide in Adults with Overweight or Obesity (STEP 1). N Engl J Med. 2021.
- Alexander GC, et al. Heterogeneity of Treatment Effects of GLP-1 Receptor Agonists for Weight Loss in Adults. JAMA Intern Med. 2026.
- Garvey WT, et al. Tirzepatide vs. Semaglutide for Obesity (SURMOUNT-5). N Engl J Med. 2025.