Cinnamon in diabetes studies: why species and product change the interpretation

The results suggest that biological activity and composition do not always coincide. All samples showed inhibition of the studied enzymes, with varying performance between aqueous extracts, decoctions, and ethanolic extracts. Ethanol made certain components and enzymatic activity more evident, but this reflects the extraction method and not the effect of a beverage or capsule taken by a person.

Trans-cinnamaldehyde, the characteristic aromatic substance in cinnamon, was detected at levels between 7.73 and 333.69 mg/g. Coumarin, a natural compound that can pose safety concerns at high exposure levels, reached up to 43.73 mg/g. The highest levels were observed in some sticks of C. cassia, C. burmannii, and C. loureiroi. An herbal preparation inhibited the enzymes but did not show the main markers being sought: a signal compatible with an as-yet-unidentified composition, not proof of efficacy.

The primary relevance concerns the quality of the product. Before discussing potential metabolic effects, it is necessary to know which species was used, which part of the plant the preparation contains, and which substances are actually present. This information is essential because clinical studies use powders, extracts, and capsules that are not always equivalent.

Research also combines two aspects that are often kept separate: in vitro activity and safety. A sample may interfere with an enzyme while simultaneously containing amounts of coumarin that require caution. Therefore, it is not correct to use laboratory data as proof that a commercial product is suitable for managing type 2 diabetes.

This work does not measure glycated hemoglobin, glycemia, complications, or adverse events in people. It is not a randomized clinical trial and does not allow for the establishment of effective doses, duration of use, or an individual benefit-risk ratio. Enzyme inhibition is an initial mechanistic clue, not a clinical outcome.

The seven samples represent only a small portion of the available products. Furthermore, the composition can change between batches, suppliers, harvests, and storage conditions. The coumarin values describe the analyzed samples and not all cinnamon on the market. The 2024 study by Mancak and colleagues does not even demonstrate that a product lacking the measured markers is adulterated: it signals a composition that requires further verification.

Evidence in humans primarily confirms the need to distinguish between formulations and contexts, rather than the efficacy of the individual samples analyzed by Mancak. A meta-analysis of 24 randomized trials found average reductions in fasting glycemia, insulin resistance, and glycated hemoglobin compared to controls [1]. This supports the hypothesis that certain preparations may modify metabolic markers, but the heterogeneity between products and studies limits any conclusion applicable to all cinnamon.

In a trial involving people with prediabetes, 54 participants were assigned to either cinnamon or a placebo for 12 weeks. The difference in fasting glycemia between groups was 5 mg/dL, and no serious adverse events emerged [2]. The result supports the plausibility of an effect in a specific formulation, but it does not verify the species, purity, or safety of the commercial products studied by Mancak.

Another randomized study included 109 adults with type 2 diabetes; 89 completed 90 days. In the cinnamon group, glycated hemoglobin decreased by 0.83 percentage points, compared to 0.37 in the control group, with a fairly evident difference between groups [3]. This is a limited clinical confirmation, because the model does not make all available formulations equivalent and does not resolve the standardization problem highlighted in the original work.

Results are not uniform. In a controlled study of 25 postmenopausal women with type 2 diabetes, 1.5 g/day of C. cassia for six weeks did not improve insulin sensitivity, glucose tolerance, or lipid profile compared to placebo [4]. The small sample size and short duration limit generalizability, but they suggest a possible dependence on population, dose, and product.

A meta-analysis of five randomized trials, totaling 282 participants, also found no significant changes in glycated hemoglobin, fasting glycemia, or lipids [5]. This synthesis reinforces the need for caution: effects observed in test tubes or in certain trials are not automatically reproducible. No directly conflicting evidence sufficiently relevant to the chemical characterization of the seven samples was identified; this does not prove the absence of contrary results.