A neutral primary endpoint is not the same as a reassuring result on every measure. This appraisal examines a six-month L-carnitine trial in metabolic syndrome, separating its main plaque-volume finding from exploratory concerns about arterial narrowing.
What the study reported
The double-blind trial randomized 177 participants with metabolic syndrome and carotid plaque. Complete scans were available for 157: 76 of 89 assigned L-carnitine and 81 of 88 assigned placebo. There was no significant between-group benefit in total plaque-volume change. Stenosis increased more with treatment, but the authors identify that outcome as not prespecified. Nominally unfavorable cholesterol findings did not survive the reported multiplicity adjustment. Primary report, Methods and Results.
Why the participant counts matter
Randomized and analyzed are different denominators. Reporting only the smaller number hides the fact that some allocated participants did not contribute complete imaging data. Our interpretation is that this limits how confidently the complete-scan comparison can be treated as representative of the entire randomized group.
Randomization protects the initial allocation; it does not automatically remove bias introduced by missing observations. The direction and size of any resulting bias cannot be inferred from unequal missingness alone. This limitation neither proves the stenosis result false nor establishes that it would persist in every analysis.
Plaque volume and stenosis answer different questions
Plaque volume measures the amount of plaque, whereas stenosis describes narrowing of the vessel lumen. A summary should not collapse them into a single claim that the artery either improved or worsened in every respect.
A non-prespecified finding can raise a legitimate safety question, but it is not confirmation of the trial’s primary hypothesis. Likewise, a nonsignificant primary result does not prove equivalence or the complete absence of an effect. These distinctions matter when favorable and unfavorable results coexist.
Subgroups, cholesterol and clinical events
The paper reports no significant treatment-by-subgroup interactions. It also reports no important between-arm differences in collected clinical adverse events. Primary report, subgroup and safety sections.
A result reaching significance in one subgroup but not another does not, by itself, demonstrate different treatment effects between them. Testing that difference directly is a separate analysis. The cholesterol finding similarly needs its multiple-testing qualification rather than presentation as a confirmed isolated result.
The imaging concern should not be translated into a claim of proven increased stroke or mortality risk. Conversely, the absence of a clear event difference in this experiment is not a guarantee of long-term safety.
What this record contributes
The balanced conclusion is no demonstrated plaque-volume benefit, accompanied by an exploratory stenosis concern. Proposed metabolite explanations remain hypotheses here, not established mediators of the imaging result. The selected population, formulation and follow-up also prevent this trial from settling every question about other carnitine exposures.
Source and review scope
Johri AM and colleagues. Progression of atherosclerosis with carnitine supplementation: a randomized controlled trial in the metabolic syndrome. Nutrition & Metabolism, 2022. PubMed record; publisher record.
Prepared with AI assistance. Main-report methods, results and discussion checked October 1, 2026. This page does not reproduce supplementary-table estimates or claim an independent audit of registry history. No independent medical expert review is claimed. Educational research appraisal only, not treatment advice or a product-quality endorsement.
Explore the study library and the guide to subgroup analyses and interaction tests.