After gastric bypass, the low-prealbumin flag missed about three in four, or more, of the patients losing an outsized share of lean mass. In Lausanne, 178 patients' blood tests and body scans told different stories

People who have a gastric bypass lose fat, but they lose some lean tissue too, and clinics want early warning if that loss is too large. Prealbumin, a protein made in the liver, is still measured during follow-up, and a low result can be taken as a sign of poor protein nutrition. A team at Lausanne University Hospital asked whether it actually tracks lean mass, using dual-energy X-ray scans (DXA) from a long-running Swiss surgery cohort.

They followed 178 adults, 124 of them women, from before surgery through to six months, one year and three years. Prealbumin fell from an estimated 0.260 g/L to 0.203 g/L at six months, then partly recovered to 0.226 g/L at three years. The share with a result under the usual 0.20 g/L flag went from 3% to 44%, then settled at 22%.

The lean mass line looked quite different. It fell by an estimated 5.5 kg at six months, 5.9 kg at a year and 6.6 kg at three years, still sliding when prealbumin had turned back up. And the test did poorly at finding the people of most concern, defined as those whose lean mass made up more than 25% of the weight they had lost. A perfect test would score 1 on the standard measure of discrimination and a coin toss 0.50. Prealbumin scored 0.36, 0.28 and 0.48 at the three visits.

At the 0.20 g/L cut-off, the test picked up 27%, 12% and 25% of those patients, so between three quarters and nine tenths went unflagged. Scores below 0.50 might look as if low prealbumin pointed the wrong way, but the authors stress how few patients met the definition: 22, 8 and 12 at the three visits. The confidence intervals are wide, and the safe reading is 'not useful here' rather than 'reversed'.

A second group gave the same message from further out. In 310 people scanned a median of 8.8 years after surgery, 26% had low prealbumin, yet prealbumin wasn't meaningfully linked to any lean-mass measure once sex, age and time since surgery were taken into account. Inflammation, a usual suspect for low prealbumin, doesn't obviously explain the early dip either: median CRP was lower after surgery than before it, even at six months.

This doesn't show the test is worthless for every purpose, or that lean-mass loss doesn't matter. The authors' claim is narrower: if the reason for ordering prealbumin is to judge lean mass after bypass, these data don't support it, and a normal result shouldn't reassure anyone either. It comes from one centre, with no dietary, protein or activity data, and only 77 of the 178 had a three-year assessment, many because they hadn't reached that point yet. DXA also measures lean soft tissue rather than muscle alone. Scans, strength and a proper look at what someone eats are harder than a blood draw, but they ask the question more directly.