Two gym shakes, one winner by 0.9 kg of 'muscle'. The authors won't crown it, and the scan may have been counting water too

Walk down the sports nutrition aisle and you will find carbohydrates sold by size. Big molecules, the pitch goes, empty from the stomach faster and refill tired muscles sooner. There is some short-term lab work behind that idea. Whether it adds up to more muscle after weeks of training is a different question, and a trial published in Nutrients on 25 October 2026 tried to test something close to it.

A team at Beijing Sport University enrolled 36 untrained men aged 18 to 25, all of normal weight, and 33 finished. Everyone lifted three times a week for eight weeks under a coach, in sessions of about 90 minutes, and everyone took one 50 g sachet a day in water, straight after training or with breakfast on rest days. Two groups got a shake with 12.5 g of protein and about 32 g of carbohydrate. One recipe was built on maltodextrin, maltooligosaccharides and glucose (the 'low molecular weight' mix), the other added isomaltulose, cyclic dextrin, a modified starch and fructose (the 'high molecular weight' mix). The third group got what the paper mostly calls a placebo: a sachet that was mainly maltodextrin.

All three groups gained. On a whole-body DXA scan, 'muscle mass' rose by 0.8 kg with the control sachet, 1.5 kg with the low molecular weight shake and 2.4 kg with the high molecular weight one, and fat changed little. Tape measures, ultrasound and strength tests in the squat, deadlift and bench press also favoured both shakes over the control. Read quickly, that looks like a ranking, with the bigger carbohydrates on top at three times the control's gain.

The first gap has an ordinary explanation. The control sachet was carbohydrate and the other two contained protein, so this was never a test of a shake against nothing, as the authors themselves spell out. The men's recorded protein during the trial averaged 1.42 g per kg of body weight a day in the control group, against 1.60 and 1.69 in the shake groups. Those differences were not statistically significant, and the paper doesn't say plainly whether the sachets are counted, but they sit in an interesting place. A large 2018 meta-analysis found extra protein stopped adding fat-free mass at around 1.62 g per kg a day, so the 'adequate protein' described in this paper may not have been quite adequate in the control group. That same analysis, pooling 49 trials, put the average benefit of protein supplements at 0.30 kg of fat-free mass. The gaps here are two to five times that, in groups of 10 to 12.

The second gap is the one a supplement label would love, and it is weak. The high molecular weight shake beat its rival on total body mass and total muscle mass and on nothing else: not arm, chest, thigh or hip measurements, not muscle thickness on ultrasound, not strength, not hormones. The authors call the finding exploratory. They point out that DXA can be swayed by stored glycogen and the water that travels with it, which is the very thing a carbohydrate recipe might change. They also note that the two recipes differed in several ways at once, so molecular weight can't be singled out. Their conclusion is that nothing here justifies recommending one over the other.

Two sums in the tables are worth a raised eyebrow, and these are our own arithmetic, not points the authors make. In the low molecular weight group, the three regional gains (0.5, 0.6 and 0.7 kg) add up to 1.8 kg, more than the 1.5 kg reported for the whole body. And 89.3 g of protein divided by the 68.6 kg average body weight gives about 1.30 g per kg, not the 1.21 in the table, which would fit a man of about 74 kg. Neither sum is proof of anything serious, and there may be dull explanations. But when a claim rests on a difference of 0.9 kg, the bookkeeping matters. The trial was also supported by a sports nutrition research institute, and three authors work for a company called Beijing Competitor Sport Science, which the paper declares.

What should someone new to the gym take from this? The plainest result is the one nobody sells: eight weeks of coached lifting built measurable size and strength in every group, including the one drinking carbohydrate alone. Getting enough protein across the day looks sensible and is backed by far larger bodies of evidence than this trial, and 12.5 g is about what two large eggs provide. Paying extra for a particular carbohydrate is not something this study supports. It was small, ran for eight weeks, included only healthy young men and analysed only those who finished.