Every clinical claim BoneBot makes is drawn from a reviewed source in this register, kept in sync with docs/EVIDENCE.md. The deterministic model predicts; the LLM only explains from these sources — it does not introduce new clinical claims.
| Evidence role | Source |
|---|---|
| Current clinical guideline formally defining T-score/Z-score calculation and WHO diagnostic thresholds (normal / osteopenia / osteoporosis). | LeBoff et al., 2022, "The clinician's guide to prevention and treatment of osteoporosis" (Osteoporos Int) |
| Current review confirming DXA as the gold-standard method for osteoporosis risk stratification and diagnosis, and its role alongside FRAX/trabecular bone score in fracture-risk assessment. | Shahane, Lim & Bolster, 2025, "Updates on the Role of DXA in the Evaluation and Monitoring of Osteoporosis" (Curr Rheumatol Rep, PMC) |
| UK patient explanation of DXA, its limits, and clinical risk factors (including prior fracture, smoking, family history, low BMI and glucocorticoids). | NHS: Bone density scan—when it is used |
| UK guideline for fracture-risk assessment, DXA, lifestyle and clinical management. | NOGG 2024 guideline |
| Screening context: all women 65+ and postmenopausal women under 65 at increased risk; not a recommendation for people with known fragility fracture or secondary osteoporosis. | USPSTF clinical summary |
| NIH patient guidance on calcium, vitamin D, activity, smoking and alcohol. | NIAMS osteoporosis guidance |
| Multi-ethnic longitudinal menopause-transition cohort: BMD loss was greatest around the final menstrual period. | Finkelstein et al., 2012 (PubMed) |
| Randomised WHI evidence that hormone therapy affects BMD/fracture outcomes; supports context, never a treatment recommendation. | Cauley et al., 2003 (PubMed) |
| Prospective Study of Osteoporotic Fractures analysis, supporting the importance of fracture history in later bone-risk assessment. | Crandall et al., 2018 (PubMed) |
| Systematic review/meta-analysis of exercise training and BMD in postmenopausal women. | Shojaa et al., 2023 (PMC) |
| Public DXA provenance for the training-data workflow; records measurement and variables. | CDC NHANES DXA documentation |
| Public wrist physical-activity-monitor provenance for the activity feature workflow. | CDC NHANES PAM documentation |
| Government-funded cohort data reference for heel ultrasound bone measures; not interchangeable with clinical DXA. | UK Biobank bone-densitometry category |
| NHS pathology context: total ALP can be elevated in bone and liver disease. | Eastern Pathology Alliance: ALP |
| Cross-sectional postmenopausal cohort (n=168): higher ALP was associated with lower T-score only in the osteopenia subgroup; it explained about 7% of T-score variance and was not a predictor in the osteoporosis subgroup. | Tariq et al., 2019 (PMC) |
| Mixed-sex, single-centre patient/control study plus mouse models. It found no difference in RBC, haemoglobin, or haematocrit between clinical osteoporosis cases and controls, demonstrating that CBC associations are not consistent enough for screening. | Li et al., 2022 |
| Evidence review of clinical and experimental CBC–bone literature; it concludes that clinical utility and integration into risk assessment need further validation. | Ha et al., 2025 (PMC) |
| Prospective MrOS cohort of older men: faster hip BMD loss was associated with anaemia, high neutrophils, and low lymphocytes. It does not validate isolated RBC count in postmenopausal women. | Valderrábano et al., 2017 (PMC) |
| Government health explanation of full blood count measures, including RBC and differential white-cell counts. | MedlinePlus: Complete Blood Count |
| Small cross-sectional postmenopausal study reporting an ALC/BMD association; hypothesis-generating only. | Di Monaco et al., 2004 (PubMed) |
| Cross-sectional study of peripheral blood-cell counts and BMD in healthy postmenopausal women; association, not prediction. | Kim et al., 2011 (PubMed) |
| Population-based postmenopausal cohort with two assessments; differential counts did not consistently predict BMD or microarchitecture. | Biver et al., 2022 (PubMed) |
| Individual-person meta-analysis of 29 prospective cohorts (~2M participants) commissioned to update FRAX: RA raised fracture risk (HR 1.49 any clinical fracture, HR 2.23 hip fracture) independent of glucocorticoid exposure and femoral-neck BMD. | Kanis et al., 2025 (PubMed) |
| Meta-analysis of 13 studies: pooled fracture risk ratio 2.25 [1.76–2.87] in RA vs non-RA, 1.99 [1.58–2.50] in the female subgroup. | Xue et al., 2017 (PMC) |
| Review of thyroid hormone disease mechanisms and effects on bone remodelling and fracture risk. | Thyroid Hormone Diseases and Osteoporosis (PMC) |
| Meta-analysis: endogenous hyperthyroidism is associated with increased osteoporosis and fracture risk via increased bone resorption. | Vestergaard & Mosekilde, 2003 (Thyroid) |
| Review: bone mineral density is reduced in treated adult coeliac disease, including in postmenopausal patients. | Osteoporosis in treated adult coeliac disease (PMC) |
| Prospective cohort of newly diagnosed adult coeliac patients: bone mineral density alterations in 60.3%, with osteoporosis in roughly half of affected patients. | Risk factors associated with osteoporosis in a cohort of prospectively diagnosed adult coeliac patients (PMC) |
| Review: fracture risk is markedly elevated in chronic kidney disease, particularly stages G3–G5D, and is an overlooked complication. | Fracture Risk in Chronic Kidney Disease (PMC) |
| Systematic review: 18–32% of CKD patients also have osteoporosis, with fracture risk over 2.5x that of people without CKD. | Osteoporosis in Patients with Chronic Kidney Diseases (PMC) |
| Global review: the large majority of postmenopausal women received no osteoporosis treatment in the period following a fragility fracture (GLOW and related cohorts). | Osteoporosis in 2022: care gaps to screening and personalised medicine (PMC) |
| European primary-care cohort: about 75% of elderly women at high fracture risk received no osteoporosis treatment. | Osteoporosis treatment gap in European primary care, Osteoporos Int 2020 |
| Cross-national review describing the fragility-fracture treatment gap as a persistent international phenomenon. | Fragility fractures and the osteoporosis care gap: an international phenomenon |
| Comparative review reporting FRAX sensitivity of only ~33% for DXA-defined osteoporosis in women aged 50–64 at a 10-year major-fracture threshold of 9.3%. | Risk assessment tools for screening, Curr Osteoporos Rep 2015 |
| Network meta-analysis comparing OST, ORAI, SCORE, OSIRIS and FRAX, grouping SCORE/ORAI/OST as higher-sensitivity and FRAX/OSIRIS as higher-specificity. | Comparative accuracy of screening tools for osteoporosis, Int J Nurs Stud 2025 |
| Performance review of the age + weight-only OST screening tool. | OST performance review (PMC) |
| Systematic review and meta-analysis of clinical risk-assessment instruments, reporting pooled OST sensitivity ~89% / specificity ~41%. | Systematic review & meta-analysis of clinical risk-assessment instruments, Osteoporos Int 2015 |
| NHANES machine-learning model for low bone density (12,108 adults ≥50); sex-stratified analysis shows age+BMI dominate in women, and menopause status and physical activity were excluded for missing data. | Karaismailoglu & Karaismailoglu, Risk prediction of low bone density with ML, Balkan Med J 2025 (PMC) |
| NHANES 2005–2006 accelerometer study linking objectively measured activity to higher bone density and trabecular bone score in older adults. | Accelerometry, BMD and trabecular bone score, NHANES 2005-2006, Arch Osteoporos 2019 |
| NHANES 2007–2018 analysis: ≥38 MET-hours/week of activity was linked to lower osteoporosis risk in post-menopausal women. | Physical activity and spine BMD in post-menopausal women, NHANES 2007-2018, J Orthop Surg Res 2023 |
| UK Biobank study: even brief bouts of higher-intensity activity predicted bone health in pre- and post-menopausal women. | Brief high-intensity activity and bone health, UK Biobank, Int J Epidemiol 2017 |
| Demographic/socioeconomic analysis of Fitbit ownership showing wearable data skews toward younger, more active, more affluent people. | Demographic/socioeconomic factors in Fitbit ownership, IJERPH 2025 |
| UK analysis of physical-activity surveillance via apps/wearables and how representative that data is of the general population. | Physical activity surveillance via apps/wearables: representativeness in the UK, JMIR 2019 (PMC) |
| Commentary on wearable-device access becoming a social determinant of health. | Access to wearables as a social determinant of health, Healthcare IT News |
| Review of health literacy and health-IT/digital-health adoption (the "digital divide"). | Health literacy and health-IT adoption (PMC) |
| Machine-learning comparison of accelerometric versus non-accelerometric factors for predicting fall risk in older adults. | Predicting fall risk in older adults: ML comparison, Digit Health 2025 (PMC) |
| Wearable-sensor gait study reporting ~81.6% accuracy classifying future fallers in community-dwelling older adults. | Prediction of fall risk using a wearable system (PMC) |
| Source of the NHANES III young-adult proximal-femur BMD reference data used to derive T-scores. | Looker et al., "Updated data on proximal femur bone mineral levels of US adults", Osteoporos Int 1998. |
| Review of osteoporosis treatment outcomes: real-world medication adherence ~60%, ~70% discontinue within a year, and physicians overestimate adherence (perceived 69% vs actual <49%). | A New Look at Osteoporosis Outcomes, Mayo Clin Proc |
| Review of the economic case for improving osteoporosis medication adherence. | The Economics of Improving Medication Adherence in Osteoporosis (PMC) |
| Community-based cohort of sedentary women 70+: only 17% reached recommended activity levels in a walking programme. | Predictors of Long-term Exercise Adherence in Older Women (PMC) |
| Cohort of mobile resistance-training app users: only 18.1% of beginners remained adherent at 6 months. | Predictors of long-term resistance exercise adherence among beginners (SportRxiv) |
| US healthcare-cost projection: osteoporotic fracture costs rising from ~$57B (2018) to over $95B/year by 2040 as fracture counts rise from 1.9M to 3.2M/year. | Healthcare Policy Changes in Osteoporosis Can Improve Outcomes and Reduce Costs, JBMR Plus (PMC) |
| US postmenopausal cohort: a single fracture is linked to roughly $30,000 in all-cause healthcare costs in the following year. | Long-term economic burden of osteoporotic fracture in US postmenopausal women, Osteoporos Int 2020 |
54 sources. See docs/EVIDENCE.md for curation rules, per-topic appraisals, and how each card is scoped and limited.