Abstract Background In adolescent idiopathic scoliosis (AIS), low bone mineral density (BMD) and impaired bone qualities have been demonstrated to be an important risk factor for curve progression. This prospective cohort study evaluated associations of repeated serum 25OHD measurements with BMD and volumetric bone qualities from early adolescence to early adulthood in AIS patients with low BMD. Methods A total of 110 girls with AIS (aged 11–14 years, femoral neck BMD Z-score ≤ 0) were recruited (baseline) and followed up at 1, 2, and 6 years thereafter. Longitudinal data of bone parameters and serum 25OHD were analyzed using linear mixed-effect models. Patients were categorized into persistent severe Vitamin D (Vit-D) deficiency (≥ 2 measurements of 25OHD < 30 nmol/L), persistent Vit-D deficiency (≥ 2 measurements of 25OHD < 50 nmol/L), or normal groups. Difference between Vit-D status was compared using ANCOVA. Results Patients were followed from a mean age of 12.98 ± 0.85 to 19.25 ± 0.90 years. During the peripubertal period (from baseline to 2-year follow-up), higher 25OHD levels were shown to be associated with higher total volumetric BMD (0.22 mg/cm3 per unit of 25OHD (adjusted), p = 0.043). At early adulthood (6-year follow-up), persistent severe Vit-D deficiency group exhibited lower total volumetric BMD (353.55 ± 45.67 vs 398.36 ± 66.65 mg/cm3, p = 0.018), lower cortical volumetric BMD (904.59 ± 31.05 vs 922.371 ± 21.29 mg/cm3, p = 0.033), and less cortical thickness (0.92 ± 0.13 vs 1.02 ± 0.12 mm, p = 0.024). Conclusion Low 25OHD is associated with suboptimal BMD and impaired volumetric bone qualities during the peripubertal period in adolescents with idiopathic scoliosis. Persistent severe Vit-D deficiency predicts poor early adulthood bone outcomes, supporting future evaluation of interventional strategies.
Abstract Context There is limited data of the associations between sex hormones and muscle mass and physical function in men over 80 years old. Objective This study evaluated associations of sex hormone concentrations with muscle mass and physical performance in older men. Methods We studied 477 community dwelling men, ages 78-97 years, who participated in the Osteoporotic Fractures in Men (MrOS) study. Follow-up measures were obtained in 247 men after 6 years. Testosterone, estradiol, estrone and dihydrotestosterone (DHT) were measured at the first visit by liquid chromatography/mass spectrometry, free testosterone by equilibrium dialysis, and sex hormone binding globulin (SHBG) by radioimmunoassay. Muscle mass was measured via D3-Cr dilution, and lean body mass via DXA. Grip strength, chair stands, 6m walk speed and Physical Activity Scale for the Elderly (PASE) were measured using standard methods. Results In a cross-sectional multivariable joint model including total testosterone, estradiol and SHBG, higher total testosterone was associated with higher D3Cr muscle mass and higher DXA appendicular lean mass while lower SHBG was associated with higher D3Cr muscle mass. In joint mixed effects models including total testosterone, estradiol and SHBG simultaneously, higher total testosterone was associated with less decline in grip strength, walking speed, and chair stand performance during follow-up. Conclusions In both cross-sectional and longitudinal analysis in older men, joint modeling of sex hormones revealed positive associations between total testosterone and various measures of muscle mass, physical performance and strength.
CONTEXT Repeated metabolic-bariatric surgery (MBS, r-BS) represents 10-25% of all MBS and is commonly-performed for weight regain after prior significant weight-loss. It remains unclear how adipose tissue adapts to surgically-induced weight-loss-regain in obesity. OBJECTIVE We hypothesized that compared to primary bariatric surgery (p-BS), women undergoing r-BS exhibit a distinct adipose tissue signature. DESIGN Subcutaneous/visceral adipose tissues (SAT/VAT, respectively) from female MBS patients were assessed histologically/molecularly/functionally (activation of human microglia cells by SAT-secretome). SETTING Elective MBS in a public university hospital. PATIENTS Women undergoing either p-BS, or r-BS with documented >15% weight-loss after prior MBS. INTERVENTION MBS. MAIN OUTCOME MEASUREMENT Inflammation in SAT. RESULTS Women undergoing r-BS (n=21) trended to be older than those undergoing p-BS (n=35)(47.2 vs. 40.5y, p=0.06), with a lower BMI (42.3 vs. 45.6 kg/m2, respectively, p=0.103), and a trend for improved cardiometabolic risk parameters (fasting insulin, CRP and HDL-c). Adipocyte size, adipose tissue fibrosis, macrophage and crown-like structures abundance were similar, while mast-cells were slightly more prevalent in r-BS. Single-nucleus RNA-seq-based deconvolution algorithm applied unto bulk RNA-seq confirmed the absence of a major shift in adipose tissue cell-type composition. Yet, RNA-seq revealed unique transcriptome of r-BS SAT, with activation of inflammatory pathways. Consistently, SAT explants from r-BS secreted higher protein concentrations of NFκB-regulated cytokines IL6/IL8. Functional impact of the more inflammatory-secretome was suggested by its increased ability to inflammatory-activate human microglia cells. CONCLUSIONS Prior BS-induced weight-loss and subsequent regain in women is associated with inflammatory transcriptome and secretome of SAT that can activate cultured human microglia cells.
Abstract Background Large language models (LLMs) have shown promise for clinical decision support (CDS), but evaluation under clinical equipoise—scenarios lacking a consensus correct answer—remains rare. We evaluated commercially available LLM performance for thyroid nodule biopsy recommendations under clinical equipoise. Methods One hundred clinical vignettes describing patients with incidentally discovered thyroid nodules of varying sizes and risk features were presented to 94 clinicians, recruited by convenience sampling from professional networks, and to multiple LLMs (GPT-5, Claude Sonnet 3.5/4.0/4.5, Llama 3.2-90B, DeepSeek R1, and OpenEvidence). Responses were dichotomized as biopsy vs no biopsy. Agreement was assessed using Krippendorff alpha, and a deterministic algorithm served as a provisional benchmark. Selected LLM responses were reviewed by experts. Results Clinicians recommended biopsy in 58% of responses (95% CI 50%, 66%), with unanimous responses in only 50 vignettes. Large language model biopsy rates ranged from 46% (Claude 3.5) to 74% (Llama 3.2-90B). Large language model–clinician agreement was moderate (α = .372-.453). Agreement with the deterministic algorithm ranged from high (α = .954, Detailed Thyroid Imaging and Data Reporting System GPT-5) to low (α = .265, Claude Sonnet 4.5). Llama 3.2-90B recommended biopsy in 8 of 8 cases where clinicians unanimously recommended against it. Conclusion Large language models differed meaningfully in biopsy recommendation patterns, with potential population-level consequences. Evaluating and aligning LLMs for CDS under equipoise requires stakeholders and processes analogous to clinical guideline committees. Because LLMs used as CDS could powerfully disseminate different practice patterns, this creates new responsibilities for guideline committees and other stakeholders.
