Head-to-head trials of low-carbohydrate and low-fat eating patterns have been run repeatedly, and the results converge on a narrow difference. The explanation is mostly about what participants actually do.
How the trials are built
Volunteers are randomized to one pattern or the other, given counseling and materials, and followed for months or a year with periodic measurements of weight and metabolic markers.
Randomization is what makes the comparison meaningful, since it distributes unmeasured differences between people evenly across the two groups rather than letting them cluster.
What randomization cannot do is make people eat what they were assigned. Participants remain free-living, buying and preparing their own food between visits.
Where the assigned diets converge
Reported intakes at the start of a trial usually differ sharply between groups. By the end they have moved substantially toward each other and toward habitual eating.
Both groups typically reduce calories early and drift back. The result is two arms that were designed to be different and finish looking more alike than intended.
A trial comparing two diets that participants only partly followed cannot detect a difference between those diets. It detects the difference between two attempts at them.
What both arms share
Almost every structured eating pattern removes a category of food, which removes a set of habitual eating occasions along with it. Attention to what is eaten rises in both groups.
Both arms also displace ultra-processed foods, because prepared products tend to be high in both refined carbohydrate and fat. Cutting either macronutrient cuts many of the same items.
The shared consequence — fewer eating occasions and more deliberate food preparation — plausibly accounts for a large part of what either arm achieves.
Why individual variation muddies the average
A near-tie in the mean is consistent with substantial variation underneath. Some people lose considerably more on one pattern, others on the other, and averages hide that.
Attempts to predict who responds to which pattern using insulin measures or genetic markers have not produced reliable classifiers, though the work continues.
Meanwhile, the variation within either group is consistently larger than the difference between groups, which is a strong hint about where the action lies.
What a tie actually tells a reader
The finding is not that macronutrients are irrelevant to physiology. Carbohydrate and fat have different effects on hormones, satiety and metabolic markers over the short term.
It is that at the scale of a year of real eating, sustainability dominates composition. The pattern a person keeps doing outperforms the one that tests better in a metabolic ward.
Anyone with diabetes, kidney disease or another condition affected by diet composition needs individualized guidance from a physician or dietitian rather than trial averages.