Study record
Hierarchical random forest model, inflammation and oxidative stress as predictors of the atherogenic index of plasma and diabetes progression
Jelinek HF, Muteir I, Al-Aubaidy H · Sci Rep · 2025
What this study reported
Applied random forest and hierarchical random forest models to predict the atherogenic index of plasma in adults stratified by HbA1c. Reported that mitochondrial redox markers including humanin ranked among the top predictors in the normoglycaemic group, that oxidative stress markers were more informative in prediabetes, and that MOTS-c and IL-1β were key features in the group with diabetes, with waist-to-height ratio the main contributing variable overall.
Limitations
Cross-sectional design with no reported sample size or external validation cohort, so stage-specific feature rankings may not be stable; MOTS-c appears only as one feature in a machine-learning model rather than as a prespecified outcome.
Study record
| Study model | Human observational |
|---|---|
| Substance studied | Endogenous MOTS-c measuredNaturally occurring MOTS-c was measured; nothing was given. |
| Species | human |
| System studied | Adults stratified by HbA1c into normoglycaemic, prediabetic and diabetic groups with clinical, oxidative, inflammatory and mitochondrial biomarkers |
| Population | adults grouped as normoglycaemic, prediabetic or with type 2 diabetes by HbA1c |
| Design | Cross-sectional cohort with machine-learning modelling |
| Sample size as reported | not reported |
| Control group | normoglycaemic group as reference stratum |
| Route (data field only) | Not reported |
| Endpoints measured | atherogenic index of plasma; MOTS-c; humanin; p66Shc; GSH and 8-OHdG; IL-1β and model R2/MSE |
| Pathways measured | Oxidative stress · Inflammation · Lipid metabolism |
| Research areas | Metabolic · Cardiovascular · Methods |
| Evidence tier | B — Other human study |
| Publication types | Journal Article |
Appraisal
Each dimension is recorded separately and never combined into a single score. A study can be well controlled and still tell you little about people.
- Relevance to MOTS-c
- Peripheral
- Model applicability
- Human
- Sample size
- Not reported
- Controls
- Limited
- Outcome validity
- Exploratory
- Support for its own conclusion
- Weak
Provenance
- Bibliographic fields retrieved from PubMed E-utilities efetch (title, journal, date, authors, DOI) on 2026-11-11.
- Annotated from the published abstract.
- Editorial review: approved (2026-11-11).
- Classification follows the site methodology; the record describes the study, not what MOTS-c does in people.