Introduction: A total of 0.1–0.8% of AIS patients progress to severe stages without clear mechanisms, and AIS girls are more prone to curve progression than boys. Recent studies suggest that AIS girls have systemic and persistent low bone-mineral density (BMD), which has been shown to be a significant prognostic factor of curve progression in AIS. The present study aimed to (a) investigate the prevalence of low BMD in patients with severe AIS and (b) assess the sexual dimorphism and independent risk factors of low BMD in severe AIS patients. Materials and Methods: A total of 798 patients (140 boys vs. 658 girls) with AIS who reached surgical threshold (Cobb ≥ 40°) were recruited. BMD were assessed using BMD Z-scores from dual-energy X-ray absorptiometry (DXA). Demographic, clinical, and laboratory values of the subjects were collected from their medical records. Logistic regression analysis was performed to identify independent risk factors of low BMD. Results: The overall prevalence of BMD Z-score ≤ −2 and ≤ −1 were 8.1% and 37.5%, respectively. AIS boys had significantly lower BMD Z-scores (−1.2 ± 0.96 vs. −0.57 ± 0.92) and higher prevalence of low BMD (Z-score ≤ −2: 22.1% vs. 5.2%, p < 0.001; Z-score ≤ −1: 59.3% vs. 32.8%, p < 0.001) than girls. Sex, BMI, serum alkaline phosphatase, and potassium were independent factors of low BMD in the severe AIS patients. Conclusions: The present large cohort of surgical AIS patients revealed that low BMD is more prevalent and severe in boys than in girls with severe curves. Low BMD may serve as a more valuable predictive factor for curve progression to the surgical threshold in boys than girls with AIS.

Background X-linked early-onset osteoporosis, caused by mutations in plastin3 (PLS3), is an extremely rare disease characterized by low bone mineral density (BMD) and recurrent osteoporotic fractures. There is limited information on genetic and phenotypic spectrum, as well as genotype–phenotype correlations of the disease. Moreover, whether decreased PLS3 levels were also involved in osteoporosis among subjects without PLS3 pathogenic mutations remains unknown. Methods Whole-exome sequencing and bidirectional Sanger sequencing were performed for screening and validation of pathogenic mutations. Serum biochemical parameters and clinical information of the subjects were retrospectively collected. ELISA and online datasets were utilized to investigate the association between PLS3 expression and BMD. Results We identified a novel splicing mutation (c.892-2A > G) which led to the skipping of exon 9 in a family with X-linked early-onset osteoporosis. Scoliosis represents a potential new phenotype in the patients harboring PLS3 mutations, which may be corrected by brace treatment. Genotype–phenotype analysis reveals that there was no significant difference in BMD z-scores between different types of reported mutations including this study (p = 0.5). There is a marginally significant negative correlation between age and BMD z-score (p = 0.059, r = − 0.30). The conditions of osteoporosis in all patients were improved after bisphosphonates therapy, with mean BMD z-score increased from − 2.9 to − 0.57 (p < 0.0001). Serum PLS3 levels in adolescents and adults without PLS3 pathogenic mutations but representing osteoporosis were also evaluated, while no association was found between bone mineral density and PLS3 levels (p > 0.05). Conclusions Our findings expanded the mutation and phenotype spectrum of the rare disease and highlights the importance of early diagnosis and early treatment with bisphosphonates. More reports of cases with PLS3 mutation and function studies of the gene are warranted to understand genotype–phenotype correlations.