microRNAs (miRNAs) are small, non-coding RNAs that regulate gene expression post-transcriptionally and are highly stable in body fluids, making them promising non-invasive biomarkers and potential therapeutic targets (Condrat et al., 2020; Huber et al., 2023). The regulation of several pathological processes by miRNAs includes bone loss, metastasis, cancer cell proliferation, and osteoblast and osteoclast differentiation (Sharma et al., 2023; Doghish et al., 2023). In AIS several studies have investigated circulating miRNAs in relation to abnormal bone phenotypes and disease progression (Zhang et al., 2018). Preclinical studies have examined miRNAs in AIS as regulators of critical molecular pathways and as potential biomarkers for disease monitoring and therapeutic guidance. Clinical studies have compared circulating miRNAs between age-matched AIS patients and controls, and stratified patients to more accurately define the multifactorial nature of the disease. Nonetheless, the translation of these markers into clinical practice remains limited by methodological variability of the studies, small sample sizes, and the absence of longitudinal validation.

miR-145-5p has been shown to disrupt osteocyte function via the Wnt/β-catenin pathway, correlating negatively with bone markers such as sclerostin, osteopontin, and osteoprotegerin (Zhang et al., 2018), while miR-96-5p and miR-224 are elevated in AIS plasma and regulate osteoblast activity, correlating with bone to certain bone quality parameters and turnover markers (P1NP and CTx). Notably, miR-96-5p shows potential as a biomarker for diagnosing AIS and may contribute to the low bone mass observed in these patients (Chen et al., 2022; Cheng et al., 2020).

Next-generation sequencing identified a six-miRNA signature (miR-122-5p, miR-671-5p, miR-223-5p, miR-1226-5p, miR-27a-5p, miR-1306-3p), with a validated four-miRNA panel achieving AUC 0.95, targeting genes involved in bone metabolism and Wnt signaling (García-Giménez et al., 2018). A composite predictive model combining clinical features (Cobb angle, Risser sign, age, menarche) with biomarkers such as miR-167 and P1NP allowed prediction of severe curve progression (>40° Cobb angle) over long-term follow-up. The model demonstrated improved predictive accuracy compared to individual parameters, achieving a high hazard ratio (HR) further validated in an independent cohort with 72.7% sensitivity and 90% specificity (Zhang et al., 2020). Additionally, circulating miRNAs carried in extracellular vesicles, particularly the miR-30 family, were elevated in severe AIS females and impaired osteoblast differentiation, suggesting an active role in disease progression (Raimondi et al., 2024). These findings allow speculation that circulating miRNAs may reflect AIS severity and curve progression, displaying sex- and age-specific patterns potentially influenced by hormonal factors such as age at menarche. miRNA sequencing identified expression profiles in severe and mild AIS patients and healthy controls. Bioinformatic analysis and validation into a larger cohort—matched for clinical characteristics except for main curve Cobb angle—revealed elevated miR-151a-3p expression with the highest diagnostic accuracy for severe AIS, associated with dysregulated bone metabolism (Wang et al., 2020). Genome-wide plasma miRNA profiling and Random Forest modeling identified a six-miRNA panel (miR-1-3p, miR-19a-3p, miR-19b-3p, miR-133b, miR-143-3p, miR-148b-3p) that accurately predicts scoliosis progression, highlighting their potential as early circulating biomarkers in AIS (Khatami et al., 2025).

Overall, these results support the use of circulating miRNAs as diagnostic biomarkers and lay the groundwork for a predictive model of AIS progression integrating validated biomarkers and clinical features.

cell-free DNA (ccf-DNA, nuclear (n) and mitochondrial (mt)) is considered a promising biomarker due to its biological origin, detectability in blood, and association with pathophysiological processes. Ccf-DNA fragments are released by cells undergoing apoptosis, necrosis, or through active secretion, reflecting cell turnover and tissue damage. Ccf-DNA differs in AIS patients relative to controls and associates with clinical parameters. Plasma n-ccf-DNA was decreased in AIS, while mt-ccf-DNA was largely unchanged. Sex-specific differences included higher n-DNA in male controls versus male AIS and female controls, and elevated mt-DNA in female AIS versus male AIS. Lenke type-specific patterns revealed that Lenke type 1 patients had lower ccf n-DNA levels, whereas Lenke type 5 patients had higher ccf mt-DNA levels compared with those of controls. However, no significant correlations with Cobb angle were observed (Li et al., 2019).

Exosomes impact bone growth by delivering functional molecules, including proteins and miRNAs, to bone marrow cells, thereby regulating the balance between bone formation and resorption (Lyu et al., 2020). Given their role in bone metabolism, exosome-derived miRNAs are emerging as promising biomarkers for AIS, useful for diagnosis and monitoring disease. Results showed that low expression of miR-27a-5p, miR-539-5p, and miR-1246 effectively differentiated AIS patients from healthy controls (Yuan et al., 2024). While these studies yield promising early results, the absence of longitudinal data limits validation.

Proteins represent promising potential blood biomarkers in AIS. A proteomic approach identified a differential protein expression in plasma exosomes of AIS patients, highlighting Cartilage intermediate layer protein 1 (CILP-1) and the transforming growth factor beta 1 (TGF-β1)/Smad pathway as key factors in muscle asymmetry and fibrosis, with potential implications for AIS progression. The study subdivided AIS patients into a “follow-up” group (Cobb angle <40°) and a “surgery” group (Cobb angle ≥40°) to assess progression. Exosome proteomic profiles differentiated early/milder from more advanced AIS, suggesting protein changes correlate with severity or progression. Elevated CILP-1 expression and its muscle-side asymmetry, linked to curve severity, proposed it as a potential progression biomarker. The connection between exosomal protein changes (systemic biomarker) and local muscle pathology (concave-side fibrosis/alteration) highlighted how molecular alterations may drive the structural and biomechanical evolution of the curve (Wang Q. et al., 2025).

A cross-sectional study comparing plasma dipeptidyl peptidase-4 (DPP-4) activity in 113 AIS girls and 62 age-matched controls found reduced DPP-4 levels in AIS, which further decreased with increasing curve severity. Significant differences emerged at key thresholds—becoming apparent at 30° and statistically significant above 50°—suggesting circulating DPP-4 as a potential biomarker for predicting curve progression (Normand et al., 2017). Another study demonstrated that reduced DPP-4 in AIS disrupts insulin-mediated myoblast metabolism, impairing proliferation and myogenesis, which may contribute to musculoskeletal abnormalities linked to AIS severity and progression, even though direct correlations with curve progression were not assessed (Dai et al., 2022). Pointing to an inflammatory component, the study found that specific inflammatory blood proteins—particularly beta-2 and gamma globulins—correlate with AIS severity, suggesting that inflammation may contribute to its progression and could aid in disease monitoring (Bertelè et al., 2024). A mendelian randomization study identified resistin (RETN) as a causal inflammatory cytokine associated with scoliosis, highlighting its potential role in disease pathogenesis and as a therapeutic target (Mardan et al., 2024).

Metabolites and Peptides. A serum metabolic profiling of AIS patients versus healthy controls identified seven differential metabolites—mainly lipids (glycerophospholipids, glycerolipids, a fatty acyl carnitine) and one glucuronic acid derivative—indicating perturbed lipid metabolism in AIS pathogenesis and serve as potential diagnostic biomarkers. Stratification by Cobb angle (mild <40°, moderate 40°–50°, severe >50°) showed no clear association between metabolite levels and curve severity (Sun et al., 2016). A separate metabolomic study identified metabolites (e.g., arginine, N-acetylaspartate, citrate) significantly associated with clinical measures—including Cobb angle—in AIS patients, using Spearman correlation. However, the analysis was cross-sectional (i.e., curve size at time of sampling) without stratifying patients by longitudinal progression. Thus, while certain metabolites correlated with curve magnitude, the study did not provide direct evidence that they predict the rate of curve progression (Xiao et al., 2021).

A recent study investigated whether circulating muscle-derived myokines could predict brace treatment outcomes in treatment-naïve AIS girls aged 10–14 years (Cobb 20°–35°, Risser 0–4). Lower baseline levels of FSTL1, apelin, fractalkine, and musclin were associated with curve progression, with FSTL1 emerging as an independent predictor of brace failure (AUC = 0.729). Given its role in bone metabolism, energy regulation, and muscle activity, reduced FSTL1 may indicate impaired muscle–bone signaling in AIS. These findings highlight skeletal muscle secreted factors as potential biomarkers for predicting bracing outcomes and guiding personalized interventions (Feng et al., 2023).

Hormones. Curvature often begins during periods of hormonal change, with the onset of idiopathic scoliosis frequently aligning with key hormonal shifts. Hormonal imbalances are common in scoliosis and during puberty are related to the onset and progression of AIS also through a direct influence on bone growth. Addressing circulating hormones as biomarkers may be crucial for managing curvature progression in AIS (Leboeuf et al., 2009).

Leptin and ghrelin are two key factors that regulate energy balance, growth and glucose/lipid metabolism (Schorr and Miller, 2017).

Several studies have reported serum leptin and ghrelin level disorders in AIS patients (Liang et al., 2012; Liu et al., 2012; Qiu et al., 2007; Sales de Gauzy et al., 2015; Tam et al., 2014; Wang et al., 2016). Circulating level of leptin were low (Qiu et al., 2007) and the circulating level of ghrelin were high (Sales de Gauzy et al., 2015) in AIS girls. These data have been confirmed in a more recent study (Yu et al., 2018). Interestingly, this study reported ghrelin as new quantitative indicator (levels of >6.48 ng/mL) for predicting curve progression in AIS girls in 18 months of follow-up. Since ghrelin regulates bone remodeling and promotes osteoblast activity through the CREB and Runx2 pathways (Ma et al., 2015), it may play a role in the low bone mineral density commonly observed in girls with AIS (Cheng et al., 2001). A study found a link between another adipokine (adiponectin) and osteopenia in AIS, with genetic variation possibly playing a role (Zhang et al., 2019). The authors showed that adiponectin levels were significantly higher in AIS patients with osteopenia than in those with normal bone mass and in controls. Although patients with normal bone mass showed moderately elevated levels, the difference from controls was not significant. Together with previous studies (Clark et al., 2014), these data suggest that adiponectin may play a specific role in AIS-related osteopenia. Adiponectin also increased RANKL and IL-6 expression in AIS primary cells more than in normal cells, promoting osteoclast activation, differentiation, and survival, thereby contributing to reduced bone mass. Finally, a pilot study investigated serum adipokines (leptin, adiponectin, resistin and visfatin) in a case-control study and suggested that adipokines are implicated in AIS development and/or progression (Normand et al., 2022). Even in this case, girls with AIS exhibited higher adiponectin levels and a lower leptin/adiponectin ratio compared to controls. Additionally, AIS participants with a Cobb angle greater than 25° showed elevated resistin levels relative to controls. Finally, melatonin may also contribute to AIS-related bone alterations. Studies on melatonin suggest that impaired signaling, rather than low serum levels, may influence osteoblast activity and bone growth in AIS, as circulating melatonin shows no clear correlation with Cobb angle (Brodner et al., 2000; Bagnall et al., 1996; Sadat-Ali et al., 2000; Azeddine et al., 2007).

An overview of the selected studies, including the clinical features and the methodological approaches with the main findings, is provided in Supplementary Table S1.