Thirteen Years After a Sleeve, the Lipid Panel Holds — Except Where It Doesn't

A surgical team in Larissa, Greece, went back to people it had operated on between 2005 and 2009 and asked for a blood sample. Eighty-six adults agreed, and their lipid panels were compared with the ones drawn before their laparoscopic sleeve gastrectomy, a mean of 13.4 years earlier. Mean total body weight loss across the interval was 28.8% — the kind of durable result a sleeve is supposed to deliver.

Triglycerides fell from 130.7 to 91.2 mg/dL. HDL cholesterol rose from 45.2 to 62.6 mg/dL. Non-HDL cholesterol, the fraction carrying most of the atherogenic particle count, fell from 151.9 to 134.0 mg/dL. Each shift was significant at p<0.001. Two other lines on the same report are less tidy: total cholesterol did not change, and the fall in LDL cholesterol — the number most lipid guidelines are organised around — lost statistical significance once the authors adjusted for multiple comparisons.

What survived thirteen years

All five composite indices the authors tracked improved at p<0.001: the atherogenic index of plasma, the triglyceride-to-HDL ratio, both Castelli risk indices and the atherogenic coefficient. Those indices are arithmetic performed on the same four or five measured values, so their improvement restates the triglyceride and HDL shifts rather than confirming anything independently.

The design limits how far the result travels. It is a retrospective single-centre cohort with no control arm, built on operations from an era whose technique and follow-up differ from current practice. Patients were their own comparison, which controls for differences between people but not for what else changed over thirteen years: age, menopause, diet, and lipid-lowering prescriptions.

One internal signal stands out. Greater percentage total body weight loss was independently associated with higher HDL cholesterol and lower triglycerides at follow-up, which suggests the lipid signature tracks how much weight came off and stayed off rather than the fact of an operation.

The line the authors declined to defend

The paper’s conclusion is narrower than its table. Improvements in triglycerides, HDL and non-HDL cholesterol “remain robust” beyond a decade, the authors write, then add that the findings “should be interpreted cautiously regarding cardiovascular and atherosclerotic risk.” A lipid panel is a set of surrogates. Nothing in an 86-patient retrospective cohort establishes fewer heart attacks, strokes or cardiovascular deaths, and the authors do not claim it does.

In mice, cholesterol falls without weight loss

A UCLA group publishing in the Journal of Lipid Research on 8 October reached a mechanistic result that sits oddly beside the Greek cohort. In a murine model of vertical sleeve gastrectomy, plasma cholesterol fell markedly compared with sham-operated animals, with no difference in body weight between the groups. The proposed route runs through the liver and the bile: gastric resection lowered the body’s systemic cholesterol burden by promoting hepatobiliary elimination, through upregulation of bile acid synthesis. The authors conclude that gastric resection affects hepatic physiology and cholesterol homeostasis independently of weight loss, and note that human studies report roughly 40% of sleeve patients can discontinue statins within six months.

Read together, the two are measuring different things at different distances. The mouse is a controlled comparison shortly after a specific operation; the humans are thirteen years downstream, with the medication changes and ageing that entails. The Greek cohort recorded lipid-lowering drug use as a secondary outcome, but the published abstract does not report how it changed, so the most obvious explanation for a flat total cholesterol can be neither confirmed nor excluded.

Fat tissue rewrote part of its instructions

A third paper adds a human tissue layer. In International Journal of Obesity, a Turku-based group profiled targeted RNA expression in subcutaneous and visceral fat from 42 women with severe obesity undergoing metabolic surgery, alongside 15 sex-matched controls without obesity. Before surgery, subcutaneous fat in obesity showed up-regulated inflammatory and stress-response genes with reduced de novo lipogenesis and cholesterol synthesis relative to controls, while visceral fat carried the more inflammatory profile. After surgery, subcutaneous fat restored lipogenic and adipogenic gene expression while downregulating inflammatory, cholesterol-biosynthesis and splicing-factor genes. Both underlying studies are registered as NCT00793143 and NCT01373892.

A transcriptome is not a lipid panel. It describes what fat cells were transcribing, in 42 women, and cannot say whether plasma LDL rose or fell.

What the agency page says, and what an index can promise

NIDDK’s patient-facing page on the benefits of weight-loss surgery lists unhealthy cholesterol levels and heart disease among the problems the operation can improve. It was last reviewed in September 2020 and summarises a literature rather than reporting a trial. Both statements hold at once: lipid improvement after surgery is well documented at population level, and a thirteen-year single-centre panel can still show the LDL fraction failing to hold.

Events, not a better index, would settle it — lipid trajectories linked to myocardial infarction, stroke and cardiovascular death in a prospectively followed cohort with repeated measurement. That is the design the LABS study, funded by NIDDK, was built to provide, and this archive’s LABS study pages set out what it measured and what it left open.

The same gap between marker and meaning has appeared before after sleeve surgery. In a smaller biopsy-staged cohort published this month, a liver fibrosis index improved while the biopsy stage did not.

This archive is independent of NIH and NIDDK. It reports what published studies and agency pages state, with the study design named, and gives no advice about whether surgery is right for any individual.