BGPT: Paper Review: Alterations in glucocorticoid homeostasis following sleeve gastrectomy

Explore by Goal

Quick Answer Copied

Core claim (what the paper is trying to show)

In a diet-induced obesity mouse model, sleeve gastrectomy (SG) improves glucose tolerance and shifts the systemic active/inactive glucocorticoid balance (rodent B/A ratio = corticosterone / 11-dehydrocorticosterone) toward lower active glucocorticoid exposure, with evidence pointing most strongly to increased renal HSD11B2 activity rather than hepatic HSD11B1 activity.

Long Answer

Paper Review (skeptical, evidence-focused): Alterations in glucocorticoid homeostasis following sleeve gastrectomy

Preprint dated Sep 05, 2025 (doi: 10.1101/2025.09.01.673462)

Mechanistic logic tree (what should causally move if the hypothesis is correct)

SG should improve glucose tolerance versus sham on HFD, measured by OGTT and/or AUC.
Active/inactive glucocorticoid balance should shift: circulating B/A (corticosterone/11-dehydrocorticosterone) should decrease after SG compared with sham (at least trending) and show correlation with OGTT metrics.
Renal HSD11B2 should explain systemic B/A: SG should increase renal HSD11B2 activity and align with intrarenal B/A shifts; hepatic HSD11B1 activity should not be the primary driver (in their data, intrahepatic B/A is described as unchanged and hepatic activity as not substantially changing in the intended direction).

Study design & population (from the manuscript text you provided)

Group	Diet	Procedure	N (mice)	Timepoint assays
LC	Standard chow	Sham-operated	8	Sacrifice at ~5 weeks post-op; OGTT before sacrifice
Sham	HFD	Sham surgery	8	Sacrifice at ~5 weeks post-op; OGTT at 4 weeks post-op
SG	HFD	Sleeve gastrectomy	10	Sacrifice at ~5 weeks post-op; OGTT at 4 weeks post-op

Source: provided manuscript text and figure legends for group sizes (LC n=8; Sham n=8; SG n=10).

What was measured where (so you can check mechanistic specificity)

Systemic: serum glucocorticoids (corticosterone B; 11-dehydrocorticosterone A) and their ratio B/A, plus mineralocorticoids/other steroids.

Renal: intrarenal B and A, renal Hsd11b2 mRNA, and renal HSD11B2 activity; also renal cytokine panel (NF-κB-linked and proinflammatory markers).

Hepatic: intrahepatic B/A, hepatic Hsd11b1 mRNA/protein, hepatic HSD11B1 activity; bile-acid ratios used as biomarkers for HSD11B1 inhibition status.

Adipose (epididymal WAT): E-WAT mass and gene/protein markers including Hsd11b1 and H6pd, as well as transcription factors and inflammatory markers.

This “triangulation” approach is a strength for causal plausibility, but it still cannot fully disentangle whether systemic B/A changes are driven solely by renal enzyme activity versus multiple upstream changes (stress physiology, diet composition, feeding/fasting effects, circadian timing, etc.).

Evidence consistency checklist (based on reported statistics, not raw units)

OGTT improvement: SG reverses impaired glucose tolerance seen in sham, with reported group-level AUC differences (p<0.0001 for AUC vs lean).
B/A ratio direction: sham shows elevated serum B/A vs lean; SG “tends” to reverse with p=0.0567 (borderline).
Correlation (key for linking mechanism to phenotype): serum B/A positively correlates with glucose at 120 min (p=0.046) and with OGTT AUC (p=0.0269).
Renal enzyme activity: renal HSD11B2 activity is higher after SG compared to LC (p=0.0252).
Hepatic contribution (claimed limited): intrahepatic B/A does not correlate with OGTT outcomes and SG does not change intrahepatic B/A (as described), supporting the paper’s stance that renal HSD11B2 is the dominant systemic driver in their model.

p-value categories (reported) for the central mechanistic link

This plot uses only the reported p-value categories included in the provided full text; it does not assume effect sizes.

Skeptical critique (strengths, limits, and what could mislead)

Strength: multi-tissue enzyme/activity triangulation

The paper measures serum steroids plus tissue concentrations, tissue enzyme mRNA/protein, and direct activity assays for HSD11B1 and HSD11B2, then also profiles transcription factors and cytokines that are plausibly upstream regulators. That breadth reduces the risk that the central correlation is a single-measurement artifact.

Red flag: mechanistic certainty is constrained by borderline group-level effects

The SG effect on circulating serum B/A is described as a trend (p=0.0567) rather than a clear statistical separation. The mechanistic link is therefore driven substantially by the correlation between B/A and OGTT outcomes rather than a robust group-level shift.

This is not “wrong,” but it means the claim is more sensitive to sample size, multiple testing, outliers (they use Grubbs’ test), and circadian/fasting alignment across groups.

Confounding risk: LC vs HFD comparison is not an isocaloric “enzyme-only” perturbation

LC is fed standard chow while Sham and SG are fed HFD, so differences reflect both metabolic state (obesity model) and diet composition. The manuscript discusses dietary differences (e.g., potassium in the liver/adrenal steroid discussion), but the overall interpretation still requires caution about attributing causality exclusively to SG-induced changes in renal HSD11B2.

What would most likely disprove/undermine the model?

If SG did not increase renal HSD11B2 activity (or if HSD11B2 activity increased without shifting B/A), the renal mechanism claim would weaken.
If intrahepatic B/A were strongly predictive of OGTT outcomes and changed with SG, their claim that hepatic HSD11B1 activity is not a substantial driver would be undermined.
If correlation between B/A and OGTT were not reproducible under matched fasting/circadian conditions or with larger N, the inference connecting enzyme biology to metabolic phenotype would be less persuasive.

Translational context check (human sleeve gastrectomy outcomes—only as broad anchors)

Human sleeve gastrectomy studies report improvements in glucose-insulin homeostasis and incretin responses at 1 and 6 months in cohorts with impaired glucose tolerance and type 2 diabetes (not measuring glucocorticoid B/A directly, but consistent with metabolic benefit plausibility). However, because the mechanistic axis here is renal HSD11B2 and glucocorticoid B/A (a rodent active/inactive ratio), direct extrapolation to human CKD/hypertension risk would require appropriately designed human studies measuring glucocorticoid metabolites/ratios with timing control.

Author review links (BGPT)

Feedback:

Updated: March 24, 2026