Review Papers Claim by Claim

Concise critique

This pyrosequencing study of peripheral blood DNA (N=63 IS patients, N=30 controls) reports lower LBX1 promoter forward strand methylation in IS (20.77% vs 24.89%, p=0.0037) and higher ESR1 T-DMR2 methylation in IS (54.68% vs 50.44%, p<0.0001), with no robust association between these methylation measures and Cobb angle severity; authors acknowledge limited sample size, age differences, tissue specificity, and lack of gene expression data Primary

Full evidence based review and critical appraisal

Study summary (what was done and main results)

The authors measured site level DNA methylation by pyrosequencing in peripheral blood DNA for selected regions of LBX1 (forward and reverse promoter strands), ESR1 (T-DMR1 and T-DMR2), and ESR2 (promoter 0N and exon 0N) comparing idiopathic scoliosis (IS) patients (N=63; 56 female) and controls (N=30; 27 female). Key reported group differences were lower mean LBX1 promoter forward strand methylation in IS patients (20.77% ± 5.27) versus controls (24.89% ± 6.21; p=0.0037) and higher ESR1 T-DMR2 methylation in IS patients (54.68% ± 5.51) versus controls (50.44% ± 3.87; p < 0.0001); ESR2 regions showed no global mean differences though many individual CpG sites differed. No methylation differences were associated with Cobb angle severity groups ≤75° vs >75° Primary

Strengths

• Targeted CpG resolution data by pyrosequencing provides per CpG quantitation allowing detection of site specific differences and not just array level averages Pyrosequencing
• Authors applied multiple testing correction (Benjamini-Hochberg) and reported per-site and region mean statistics, increasing transparency of the pattern-level results Statistical
• Gene selection is rationale driven: LBX1 has strong genetic association with IS and ESR1/ESR2 are plausible due to sex bias in IS Rationale

Limitations and potential biases (critical)

1. Tissue mismatch and biological interpretation — methylation was measured in peripheral blood whereas the biomechanics and molecular pathophysiology of IS likely involve paraspinal muscle, bone, and neural tissues; the authors acknowledge tissue-specific methylation and that blood is a surrogate biomarker not the target tissue. This weakens causal claims linking promoter methylation to pathogenesis Tissue
2. Confounding by age and composition — controls are older on average (paper notes age difference) and blood cell composition differences can drive apparent methylation differences; authors state mean age difference is small but do not present cell composition adjustment or include cell type deconvolution, a common and important confounder in blood methylation studies AgeConfident concise judgement: promising preliminary epigenetic associations worthy of replication but not yet sufficient to infer causality or clinical value (see improvement steps above).