Oral semaglutide carbon estimates split wide apart at the lowest dose
Pharmaceutical climate accounting rarely reaches the level of a single product, yet oral semaglutide now has two separate carbon estimates that do not fully agree. A comparative analysis in BMJ Open examined Rybelsus using the medicine carbon footprint, or MCF, framework and manufacturer-reported life-cycle assessment, or LCA, data. The exercise matters because GLP-1 receptor agonists are being used more widely for type 2 diabetes and obesity, and the sector’s greenhouse gas burden is increasingly part of clinical and policy discussions.
The comparison used cradle-to-gate boundaries, meaning the analysis focused on emissions up to the point the product leaves manufacturing rather than downstream use or disposal. Within that scope, the two approaches produced different values across the 3 mg, 7 mg, and 14 mg tablets. The gap was largest at the lowest dose and narrowed as tablet strength increased.
The lowest-strength tablet carried the biggest modeling gap
Annualised MCF estimates were 8.9 kg CO₂e per year for the 3 mg tablet, 20.0 kg CO₂e per year for the 7 mg tablet, and 39.3 kg CO₂e per year for the 14 mg tablet. The corresponding European LCA values were 17.0, 25.2, and 39.9 kg CO₂e per year. The mean symmetric percentage difference was -29%, indicating that the MCF framework generally returned lower estimates than the manufacturer-reported LCA across the three strengths.
The most pronounced separation appeared at 3 mg, where the symmetric percentage difference reached -62.6%. By contrast, the 14 mg estimate nearly converged, with a difference of -1.5%. In practical terms, the modeling approaches looked far less alike when emissions were divided across a low-dose tablet than when they were spread across a higher-dose product with a larger active ingredient contribution.
Lower-dose products can expose fixed emissions more starkly. Packaging, excipients, and device-related components do not shrink in lockstep with the amount of active drug, so their relative weight rises when the medicine content is smaller. In a formulation such as oral semaglutide, that asymmetry can make the carbon profile look very different depending on the accounting method and the assumptions built into each model.
Why fixed manufacturing emissions matter in peptide medicines
Oral semaglutide is a peptide-based medicine, and peptide production carries a manufacturing footprint shaped not only by synthesis of the active ingredient but also by excipients, packaging, and the broader industrial process. When product-specific data are incomplete or proprietary, modeling frameworks become essential. The MCF method offers a faster route to an estimate, but it is only as strong as the assumptions used to allocate emissions across ingredients, packaging elements, and production steps.
The manufacturer LCA likely had access to product-specific production data that are not publicly available in full detail. That can make the estimate more complete, but it can also make independent replication difficult. The methodological gap is therefore not simply a numerical disagreement; it reflects two different ways of translating industrial inputs into a climate figure. For clinicians and health systems trying to compare therapies, those differences can affect how a product is positioned in sustainability assessments.
The contrast was especially visible where the dose strength was lowest. At 3 mg, the MCF estimate was roughly half the manufacturer-reported value, a pattern consistent with fixed emissions dominating the calculation when less active ingredient is present. At 14 mg, the estimates nearly aligned, suggesting that dose-related allocation effects may diminish as the active drug becomes a larger component of the formulation profile.
Carbon-footprint transparency remains uneven across medicines
This analysis was built around a simple but under-addressed problem: product-level carbon footprints are rarely disclosed in a form that allows direct comparison. Pharmaceutical emissions contribute materially to healthcare’s total greenhouse gas burden, yet the field still lacks consistent disclosure standards. Without harmonised boundaries, two legitimate methods can produce figures that are not directly interchangeable.
The investigators compared the annualised estimates descriptively and calculated absolute differences and symmetric percentage differences across the three dose strengths. That approach kept the focus on method-to-method divergence rather than trying to force a false sense of precision. For a medicine like oral semaglutide, where dose strengths vary and formulation components may carry large fixed burdens, the choice of accounting method clearly shapes the result.
One sentence captures the central issue: the same tablet can look materially more or less carbon intensive depending on how emissions are assigned.
What the findings can and cannot support in practice
The authors conclude that MCF provides a rapid, accessible, and indicative estimate when proprietary data are unavailable. That is a useful message for early screening, internal benchmarking, or preliminary sustainability review. It is not a substitute for a full product-specific LCA when detailed manufacturing data exist and decision-makers need a more exact estimate.
The report also calls for improved transparency and harmonised boundaries to strengthen comparability across pharmaceutical carbon footprint methods. Without shared conventions, low-dose products may be disproportionately sensitive to accounting choices, and comparisons across manufacturers or across GLP-1 receptor agonists may become misleading. In a field increasingly asked to document environmental performance, method alignment is as important as the final number.
The findings should not be generalized beyond this formulation. The comparison applies specifically to the Rybelsus version of oral semaglutide and may not extend to other semaglutide formulations or to other GLP-1 receptor agonists. Even within the same therapeutic class, differences in delivery system, packaging, manufacturing pathway, and dose structure can shift the carbon profile enough to change the estimate materially.
The narrowest takeaway is methodological, not promotional
The most useful reading of this paper is not that one method is universally right and the other wrong. The useful point is that carbon accounting in pharmaceuticals is still vulnerable to boundary choices, allocation rules, and data access. In oral semaglutide, those choices mattered most at the lower dose and mattered least at the highest dose studied.
What remains unresolved is how quickly the field can move toward standardized, transparent product-level disclosure that allows independent comparison without forcing every institution to rebuild the same estimate from scratch. Until that happens, carbon footprint data for medicines will remain partly method-dependent, and the clinical sustainability conversation will continue to rely on estimates that are informative but not fully interchangeable.