Low-carb traded lower triglycerides for higher LDL

Carrying excess weight can impair cardiometabolic health, and cutting carbohydrates is one of the most popular routes people take to address it. The usual argument is that low-carbohydrate eating does more than shed weight — that it improves the underlying markers of cardiovascular risk. That claim is testable, but it needs trials that hold calories equal so the comparison is about macronutrients rather than about who ate less. Flow-mediated dilation, a measure of how well arteries widen in response to blood flow, is a particularly interesting outcome because it reflects vessel function rather than a blood lipid number.

The authors searched PubMed, Scopus and Web of Science through May 2023 for randomised clinical trials comparing hypocaloric low-carbohydrate diets with hypocaloric low-fat diets in adults with overweight or obesity, with flow-mediated dilation as the primary outcome. Ten studies reporting flow-mediated dilation, totalling 475 participants, were meta-analysed using weighted mean differences. Secondary outcomes covered systolic and diastolic blood pressure, C-reactive protein, HDL, LDL, triglycerides and total cholesterol, plus anthropometric and body composition measures. A separate analysis isolated very-low-carbohydrate diets from the broader low-carbohydrate group.

Across all low-carbohydrate trials, flow-mediated dilation fell by 1.04% relative to low-fat diets, but the confidence interval crossed zero and the result was not significant (p = 0.10). When very-low-carbohydrate diets were analysed on their own, the reduction reached 2.12% and was significant (p = 0.001). Low-carbohydrate diets produced no different effect from low-fat diets on body composition or anthropometry, C-reactive protein, blood pressure, HDL or total cholesterol. Two lipids did diverge: triglycerides fell by 19.94 mg/dL more on low-carbohydrate diets, while LDL rose by 20.00 mg/dL more.

The findings complicate the idea that low-carbohydrate eating is inherently better for cardiovascular risk under matched calorie restriction, since most markers simply did not separate. They fit the familiar pattern of triglycerides improving and LDL worsening on carbohydrate restriction, but add a vascular-function signal that has had far less attention. The very-low-carbohydrate result rests on a subset of an already modest evidence base, so it needs confirmation in trials designed around vessel function. It also leaves open whether the LDL rise and the flow-mediated dilation change translate into differences in actual cardiovascular events.

If the goal is weight and body composition under a calorie deficit, this pooled evidence gives little reason to expect low-carbohydrate to beat low-fat. It is a reason to think the lipid trade-off is real: better triglycerides, worse LDL, on average. The vascular finding is worth knowing about mainly for the most extreme versions of carbohydrate restriction, and even there it comes from a small pool of trials. None of this speaks to how any individual responds, and blood markers are not the same thing as clinical outcomes.