Low tierImaging StudyCitation verified
Carbohydrate Restriction-Induced Elevations in LDL-Cholesterol and Atherosclerosis: The KETO Trial
Matthew Budoff, Venkat S Manubolu, April Kinninger, Nicholas G Norwitz, David Feldman, Khurram Nasir - JACC: Advances, 2024
The cross-sectional baseline arm of the Keto-CTA study: in lean, metabolically healthy people whose LDL rose very high on a ketogenic diet, coronary plaque burden was no greater than in matched lower-LDL controls, and LDL-C did not correlate with plaque within either group.
Key findings
- No significant difference in plaque burden: CAC median 0 (IQR 0-56) vs 1 (IQR 0-49), P=0.520; CCTA total plaque score P=0.357.
- No correlation between LDL-C level and CCTA coronary plaque in either cohort.
- KETO group mean LDL-C 272 mg/dL (max 591) over a mean 4.7 years, vs 123 mg/dL in controls.
Why this evidence tier (Low)
- Risk of bias:
- Small, self-selected cohort matched to an external control group; cross-sectional baseline only.
- Precision:
- n=80; wide IQRs; low event/plaque counts limit precision.
- Directness:
- Measures plaque (a surrogate for events), not cardiovascular events; baseline (not longitudinal) here.
- Consistency:
- Single cohort; the longitudinal follow-up complicates the picture (see limitations).
- Funding / COI:
- Associated with citizen-science/low-carb advocacy (Feldman); registered trial NCT05733255.
Low certainty: a small cross-sectional imaging study with no outcome data; suggestive for the LMHR phenotype but far from definitive.
- Population:
- 80 lean, metabolically healthy adults (mean age 55.5) with carbohydrate-restriction-induced LDL-C >=190 mg/dL (mean 272, max 591), mean 4.7 years ketogenic; matched 1:1 to Miami Heart cohort controls (mean LDL-C 123 mg/dL).
- Conflicts of interest:
- Co-author Dave Feldman founded the Citizen Science Foundation and advocates around the LMHR phenotype; senior author Budoff is an established CV imaging researcher. See source for disclosures.
- Funding:
- Not reported on the fetched abstract; registered as NCT05733255.
Limitations
- Cross-sectional baseline; no cardiovascular event data.
- Small (n=80), self-selected, predominantly healthy volunteers; matched to an external control cohort.
- The one-year longitudinal follow-up (Soto-Mota et al., JACC: Advances 2025, DOI 10.1016/j.jacadv.2025.101686) was retracted by the authors and editors over methodology concerns, so reliable longitudinal plaque-progression data are not currently available; this cross-sectional baseline must not be read as evidence of long-term safety.