What the research found

Researchers conducted a six-month randomized controlled trial comparing three diets in people with obesity: ketogenic (4% carbohydrates), Mediterranean (50% carbohydrates), and low-fat plant-based (70% carbohydrates). All participants received prepared meals calibrated to produce equivalent weight loss (~10%), and the study included advanced metabolic testing — MRI, DEXA scans, and specialised challenge tests — beyond typical clinical assessments.

Despite similar weight loss across groups, the ketogenic diet produced larger improvements in several cardiometabolic markers: liver fat accumulation, fasting blood glucose, triglycerides, inflammatory markers, and insulin resistance scores. A particularly notable finding was the 21-point improvement in lipoprotein insulin resistance (LP-IR) on keto, compared with 7 points on Mediterranean and 3 points on low-fat diets. Since LP-IR tracks cardiovascular risk more strongly than LDL cholesterol in observational studies, this suggests metabolic benefits beyond weight loss alone.

The ketogenic group also showed sharper reductions in liver-derived glucose and fat synthesis — two processes that drive high blood sugar and triglycerides in metabolic dysfunction. Interestingly, neither LDL nor ApoB changed significantly differently between keto and low-fat diets, despite the keto group consuming roughly 50 times more dietary cholesterol. The authors attribute this to baseline obesity; people with higher BMI tend not to experience LDL rises when adopting low-carbohydrate eating.

Why it matters for you

If you're tracking triglycerides, fasting glucose, or insulin resistance markers, this trial suggests carbohydrate restriction may shift these metrics more efficiently than Mediterranean or plant-forward approaches — at least in the context of intentional weight loss. The emphasis on liver fat and endogenous glucose production is particularly relevant: excess hepatic fat correlates with poor metabolic health markers you'd monitor (HOMA-IR, fasting insulin), and the keto group saw steeper declines here.

The LP-IR finding deserves attention if you're optimising beyond conventional lipid panels. Many MyKine users measure LDL and ApoB, which remained stable on keto despite higher saturated fat intake. But LP-IR — which reflects how insulin resistance affects lipoprotein particle metabolism — improved substantially more on the ketogenic approach. This hints that total cardiovascular risk may improve even if LDL doesn't decline, provided insulin sensitivity markers move favourably.

One practical caveat: this trial was conducted in people with BMI ~40 and significant metabolic dysfunction. Your own lipid response to keto depends partly on your starting body composition and insulin sensitivity. Someone already lean or metabolically healthy may see different LDL or ApoB responses than the obese cohort studied here.

Caveats

• Short duration: Six months is moderate for a nutrition trial; longer-term sustainability and effects unknown
• Controlled feeding: Meals were provided, eliminating real-world adherence challenges and food quality variability
• Specific population: Results apply to people with obesity and metabolic dysfunction; generalisation to leaner or healthier individuals is uncertain
• No active control: No comparison to standard medical nutrition therapy or other interventions
• Intermediate markers: Improvements in biomarkers don't directly prove reduced cardiovascular events; observational data on LP-IR and risk are separate from this trial's outcomes