Table 4.

Evidence hierarchy and critical appraisal of key pathogenetic and clinico-pathological associations in PDB

Association	Evidence Level	Study Designs Supporting Association	Supporting Rationale	Major Confounders / Limitations	Key References
Genetic factors
Severe, polyostotic PDB in SQSTM1 mutation carriers	Level 2: Strong Mechanistic	Family-based linkage studies; case-control genetic association studies; in vitro functional assays (osteoclast cultures, autophagy flux); genetically engineered mouse models (SQSTM1 knock-in/knockout)	Well-defined molecular cascade (NF-κB/autophagy dysregulation) leading to the aggressive bone phenotype; orbital involvement is a direct secondary consequence	Incomplete penetrance & significant phenotypic heterogeneity within families limit clinical predictive value. Majority of mechanistic data are derived from non-human models or cellular systems; direct validation in human bone in vivo remains partial. Potential ascertainment bias in studies focused on severe familial cases	Gennari et al., 2022; Shaik et al., 2021
AS in TNFRSF11B-related JPD	Level 2: Strong Mechanistic	Case series of genetically confirmed JPD patients with detailed ophthalmological phenotyping; genetic sequencing studies; mechanistic extrapolation from OPG’s known physiological role	OPG’s known role in inhibiting vascular calcification provides a direct, plausible pathway from genetic defect to Bruch’s membrane pathology	Evidence is based almost exclusively on the rare, monogenic JPD population; its direct applicability to the pathogenesis of AS in classic, late-onset PDB is less certain. Small sample sizes inherent to ultra-rare disorders	Kerr et al., 2010
Environmental factors
Rural residence/cattle density and PDB risk	Level 3/4: Primarily Correlative	Ecological (population-level) studies; retrospective analyses of large primary care or administrative health databases	Ecological association with no specified molecular mechanism; highly susceptible to confounding	High risk of confounding: Association may be attributable to unmeasured factors correlated with rural residence (specific occupational exposures, diet, socioeconomic status). Ecological fallacy. Dependent on accuracy and consistency of diagnostic coding in electronic records	Rebollo-Najera et al., 2025
Vit D deficiency and PDB.	Level 4: Purely Correlative	Cross-sectional observational studies comparing serum 25-hydroxyvitamin D levels between PDB patients and healthy or disease controls	Observational association only; high likelihood of reverse causality (reduced mobility in PDB) or confounding	High probability of reverse causality: PDB-related skeletal pain and disability likely reduce outdoor activity and sun exposure. Confounding by age, season, latitude, body mass index, and chronic kidney disease. No prospective or interventional data to support a causal role	Rendina et al., 2019
SQSTM1 mutation and pollutant (Cd, smoke) interaction	Level 3: Correlative with Plausible Mechanism	In vitro experiments using osteoclasts differentiated from PBMCs of SQSTM1 mutation carriers and controls, exposed to heavy metals or cigarette smoke condensate	In vitro data shows altered osteoclast biology; proposed oxidative stress mechanism is plausible but lacks direct in vivo causal proof in humans	Evidence is preclinical (in vitro only). Doses and exposure durations may not reflect real-world human environmental exposure Small patient-derived cell sample sizes. Lacks in vivo validation in animal or human studies to confirm this interaction causes clinical disease	Numan et al., 2019
Historical coal exposure and PDB emergence	Level 3: Correlative with Plausible Mechanism	Historical ecological analyses examining spatial-temporal disease patterns alongside fuel use; migration studies comparing disease prevalence in migrants from high-risk areas to native populations	Strong spatiotemporal correlation; coal-derived toxins (heavy metals, fluorides) are biologically plausible triggers, but direct evidence is historical/epidemiological	Evidence is circumstantial and ecological; cannot establish causality at the individual level. Relies on the accuracy of historical disease and exposure data. The precise pathogenic constituent(s) of coal smoke remain unidentified	Cundy, 2024
Direct structural/compressive ocular lesions
Ptosis due to PDB-related orbital GCT	Level 1: Causal	Case reports and small surgical series with direct clinico-radiological-pathological correlation (imaging confirming mass, histopathology confirming GCT, surgery relieving symptoms)	Direct mass effect confirmed clinically and radiologically; resection relieves symptoms	Evidence limited to level IV/V case reports/series, subject to publication bias (only remarkable or surgically managed cases are reported). No population-based data on the incidence of this rare complication	Pecorella et al., 2000
Compressive optic neuropathy from pagetic OCS	Level 1: Causal	Case series and individual case reports with direct correlation between high-resolution CT/MRI evidence of bony canal narrowing and corresponding clinical/functional optic nerve deficits	Direct imaging evidence of bony compression at the exact site of neurological deficit	Evidence primarily from severe, often syndromic (JPD) cases; may not represent the typical course in adult PDB. Small sample sizes and lack of prospective longitudinal data on progression	Kerr et al., 2010; Eretto et al., 1984
Vascular/metabolic ocular lesions
Vascular optic neuropathy from pagetic hypervascularity (“steal” phenomenon)	Level 1/2: Causal / Strong Mechanistic	Level 1: Case reports demonstrating visual improvement following intervention (BP infusion, corticosteroids, or vascular embolization) Level 2: Case reports with angiographic (e.g., DSA, MRA) confirmation of aberrant, high-flow vasculature adjacent to the ischemic nerve	Level 1: Intervention (BPs, embolization) reverses ischemia/edema. Level 2: Angiographic confirmation of abnormal, high-flow vasculature adjacent to the nerve	Evidence is entirely from level IV/V case reports. The diagnosis of “vascular steal” is often inferential, based on a characteristic treatment response rather than direct, quantitative hemodynamic measurement. No controlled studies exist	Isasi et al., 2002; Yuan et al., 2010

Abbreviations: Paget’s disease of bone(PDB), Sequestosome1 (SQSTM1), Nuclear Factor Kappa B (NF-κB), Angioid Streaks (AS), Tumor Necrosis Factor Receptor Superfamily Member 11B (TNFRSF11B), Juvenile Paget’s Disease (JPD), Osteoprotegerin (OPG), Vitamin D (Vit D), Cadmium (CD), Peripheral Blood Mononuclear Cells (PBMCs), Giant Cell Tumor (GGT), optic canal stenosis (OCS), Computed Tomography (CT), Magnetic Resonance Imaging (MRI), Bisphosphonate (BP), bisphosphonates (BPs), Digital Subtraction Angiography (DSA), Magnetic Resonance Angiography (MRA)