Genomic Claims You Can Trace

 Quick Analysis Plan Copied

 Long Analysis Plan

Verdict: the GHRL genotype-biomarker literature in this evidence set is hypothesis-generating, not clinical-grade. No study here has been replicated; all human association studies are small (n = 76–215), single-population, and uncorrected for multiple testing.

1. What the reported data actually show

The most quantitatively detailed result comes from the bariatric cohort (n = 76 obese patients): GHRL rs696217 T-allele carriers after laparoscopic sleeve gastrectomy achieved 12-month TWL of 41.35% vs 27.09% for G-allele carriers (p < 0.001), with matching ghrelin reductions at 6 months (−46.75%, p < 0.001); gastric plication showed no genotype effect, and GAE showed smaller genotype differences Bariatric GHRL genotype outcomes. Two critical caveats: the procedure heterogeneity (only LSG showed the large effect) and the absence of correction for multiple subgroup comparisons weaken the claimed predictive value of the T allele.

Notably, rs696217 points in opposite directions across contexts: the T allele predicts better weight loss post-LSG rs696217 T allele bariatric benefit, yet the Met72 (T) allele is more frequent in obese Turkish-Cypriots (p = 0.0005; OR = 0.459) and linked to lower HDL-C Leu72Met obesity association. This internal tension is a strong signal of context-dependence, confounding, or chance findings rather than a stable biomarker effect.

Reported values from the bariatric cohort study (n = 76 obese patients); LGCP bar shown as zero because no significant genotype effect was reported.

2. Evidence quality scorecard

Scores are the papers' own assessed quality metrics from the evidence records — they capture internal rigor but not replication status, which is the decisive missing element for a biomarker claim.

Higher quality scores belong to non-genotype studies (comparative genomics, mouse proteomics); the genotype-association papers cluster at 4–7, the range where false-positive risk is highest.

3. Battle-test: where the evidence breaks

• Sample-size fragility: The psoriasis study's headline p ≈ 0.000 for rs34911341 GHRL psoriasis association rests on 80 cases / 70 controls, and the Algerian prostate-cancer study explicitly notes HWE deviations in patients for rs696217 and rs2922126 — a classic genotyping-quality red flag GHRL/GHSR prostate cancer null study.
• Overfitting risk in interaction models: The prostate-cancer MDR model reports near-perfect metrics (PR-AUC = 1.00, F1 = 0.953) — values this extreme in n = 215 with 10-fold CV almost certainly reflect overfitting despite permutation testing; the authors themselves flag this MDR overfitting caveat.
• No replication anywhere in the set: The Leu72Met obesity association Turkish-Cypriot obesity study and the rs34911341 psoriasis association are single-population findings with acknowledged lack of functional validation.
• Expression-based GHRL measures are not genotype biomarkers: The schizophrenia pilot found no GHRL expression difference overall (2.70% vs 2.25% MFI-normalized), and the single significant male FTO result (p = 0.041) is explicitly flagged by authors as preliminary with confounding by age and antipsychotic exposure GHRL expression schizophrenia pilot.
• Protein-level biomarker claims lack genotyping: The Giardia study reports serum ghrelin falling from 82.52 to 28.89 ng/mL in infection Giardia ghrelin biomarker study but has no genotype data at all, and its own limitations note inconsistent sample-size reporting — it should not be counted as genotype-biomarker evidence.

4. Strongest counter-evidence in the set

Paradoxically, the highest-quality papers here argue against GHRL as a universal biomarker: snakes have independently lost GHRL and MBOAT4 entirely across 32 species while maintaining energy homeostasis via intermittent feeding Serpentes GHRL gene loss study, and GHRL plasma levels are sex- and strain-differential in mice (higher in BALB/c females, conserved in humans) Sex-strain GHRL proteomics study. Together these imply any ghrelin-based biomarker must be stratified by sex and species/strain context — a confounder no genotype study here controlled.

5. What would change this conclusion

Large multi-center replication of rs696217 T-allele benefit post-LSG (n > 500, FDR-corrected, ideally with serum ghrelin measured) would upgrade this from hypothesis-generating to usable biomarker evidence. Conversely, a single well-powered null replication would falsify the strongest current claim, as the bariatric study's own falsification criteria state Falsification criteria bariatric study.