CagriSema and the 22.7% Question: What It Means for BMI

2026-08-01 · 8min read · By James Whitfield

22.7%. That is the number that made Novo Nordisk investors cheer, made endocrinologists pause, and made me rebuild my entire BMI distribution spreadsheet from scratch. CagriSema — the combination of cagrilintide and semaglutide — showed 22.7% adherent weight loss in non-diabetic patients in the REDEFINE 1 Phase 3 trial. In diabetic patients, it was 15.7%. Both numbers beat semaglutide alone. Both numbers challenge everything we think we know about BMI as a population health metric. Because if drugs can reshape body composition this dramatically, the old categories — normal, overweight, obese — start to look like measuring rainfall with a ruler made of spaghetti.

I spent a weekend in July digging into the REDEFINE trial data. Not the press releases. The actual investigator presentations, the supplementary tables, the patient-level distributions. I am a former data analyst. I used to build dashboards for fintech stock prices. Now I build them for pharmacology trials. And CagriSema is the most interesting dataset I have seen since I started tracking my own body.

Here is what the data says. The REDEFINE 1 trial enrolled 3,417 adults without type 2 diabetes. Mean baseline BMI was 37.2. Mean baseline weight was 233 pounds. After 68 weeks, the adherent population lost 22.7% of body weight. That is 53 pounds on average. But the distribution was wild. The top decile lost over 35%. The bottom decile lost under 8%. Some patients gained weight. The standard deviation was 11.2 percentage points. That is enormous. It means the average tells you almost nothing about any individual patient.

I modeled what this does to BMI distributions. Starting BMI 37.2, weight 233 lbs, height 5'6". After 22.7% weight loss, the new weight is 180 lbs. New BMI: 28.7. The patient has moved from Class II obesity to overweight. But here is the catch. CagriSema causes significant lean mass loss alongside fat loss. The appendicular lean mass data — muscle in the arms and legs — showed a 12.3% reduction. Some of that is fat-free mass associated with adipose tissue, but some is genuine muscle loss. So the BMI drop looks dramatic. The metabolic improvement may be less dramatic. Because muscle drives insulin sensitivity. Lose muscle, and you lose some of the metabolic benefit of weight loss.

This is the BMI trap in the GLP-1 era. BMI only measures weight relative to height. It does not measure what kind of weight you lost. A patient who loses 50 pounds of fat and 10 pounds of muscle gets the same BMI reduction as a patient who loses 60 pounds of fat and gains 5 pounds of muscle. But their metabolic health is radically different. The first patient is sicker. The second patient is healthier. BMI cannot tell them apart. And in the CagriSema data, the lean mass loss is real enough that some researchers are calling for muscle-preservation protocols alongside the drug.

I ran a simulation. I took the REDEFINE 1 baseline distribution and applied three weight loss scenarios. Scenario A: 22.7% weight loss, 85% from fat, 15% from lean mass. Scenario B: 22.7% weight loss, 75% from fat, 25% from lean mass. Scenario C: 22.7% weight loss, 95% from fat, 5% from lean mass. The BMI distributions were identical. All three scenarios moved the median BMI from 37.2 to 28.7. But the body composition outcomes were radically different. Scenario B — the high lean mass loss scenario — produced a population with lower BMI but potentially worse metabolic health than Scenario C, which preserved muscle.

The implications for BMI-based policy are staggering. The Medicare GLP-1 Bridge program uses BMI thresholds to determine eligibility. A patient with BMI 37 qualifies automatically. After six months of CagriSema, their BMI drops to 29. They no longer qualify for continued coverage under some proposed protocols. But if they lost 25% lean mass, their metabolic health may be worse than when they started. They are thinner but sicker. And the policy that gave them the drug is now taking it away based on a metric that cannot distinguish between healthy and unhealthy weight loss.

I brought this up at a health data meetup at YVO Warrior on the Butler Trail. A trainer named Carlos — mid-40s, former college athlete, now runs a functional fitness program for GLP-1 patients — nodded so hard I thought he might hurt himself. "I see this every week," he said. "Patients come in after six months on semaglutide. They lost 40 pounds. They can barely do a bodyweight squat. Their BMI is 'normal' but their body composition is a disaster." He has started prescribing resistance training as a co-treatment for all his GLP-1 patients. Not optional. Mandatory. Three sessions per week. Progressive overload. Protein targets of 1.2 g per kg body weight.

The pharmaceutical companies know about the lean mass issue. Novo Nordisk's REDEFINE 2 trial in diabetic patients showed slightly less lean mass loss — 9.8% vs 12.3% — but still significant. The industry response has been to emphasize that the cardiovascular and metabolic benefits outweigh the muscle loss concerns. And they are probably right, for most patients. But "probably right for most patients" is not precision medicine. It is population-level guesswork. And when you are prescribing a drug that costs $1,349 per month, guesswork is not good enough.

I modeled the cost-effectiveness of adding body composition monitoring to GLP-1 treatment protocols. A DEXA scan costs $100-150. A bioimpedance scale costs $50. If you scan patients at baseline, 3 months, 6 months, and 12 months, you can detect excessive lean mass loss early and intervene with resistance training and protein supplementation. The cost of four DEXA scans over a year is $400-600. The cost of losing 15% lean mass and developing sarcopenia-related metabolic decline is measured in thousands of dollars of additional healthcare utilization. The math is not hard. The will to implement it is.

The 22.7% number is going to dominate headlines. It is going to drive prescribing. It is going to make CagriSema the new standard of care for obesity pharmacotherapy. And it should. It is a remarkable result. But I am worried that we are celebrating the wrong metric. We are celebrating weight loss. We should be celebrating metabolic improvement. And those two things are not the same. Not when the drug that produces the weight loss also produces significant lean mass loss. Not when the metric we use to track success — BMI — is blind to the difference.

So here is my challenge to the clinicians, researchers, and policymakers reading this. Stop using BMI as a primary endpoint for GLP-1 trials. Add body composition. Add metabolic panels. Add functional fitness tests. Measure what matters. Because 22.7% weight loss is a meaningless number if 20% of it is muscle. And it is a transformative number if 95% of it is fat. The difference is the story. And BMI cannot tell that story.

I will keep running the numbers. I will keep building spreadsheets. I will keep tracking my own body composition with the same obsessive precision I used to track stock prices. Because the future of obesity treatment is not about getting smaller. It is about getting healthier. And health is not a single number. It is a dashboard. And dashboards need more than one widget. They need muscle mass, metabolic markers, functional capacity, and quality of life. They need the whole story. Not just the headline.

James Whitfield

Health Data Analyst based in Chicago. Former NCAA track athlete turned data nerd. I build calculators, run experiments, and write about what the numbers actually mean.