Background/Purpose: Symptomatic Hand Osteoarthritis (SxHOA) is a painful, destructive and deforming polyarticular disorder that is highly prevalent. The epidemiology of incident SxHOA and its association with metabolic processes is unclear with limited studies and conflicting results. We aimed to characterize individuals who develop symptomatic interphalangel, thumb-based and erosive hand osteoarthritis and compare them to those who did not.

Methods: We evaluated 3604 participants in the Osteoarthritis Initiative (OAI) with complete data for baseline and 48-month radiographic hand OA. We defined interphalangeal OA as a hand with at least two joints with KL grade ≥2 on at least two fingers (digits 2 to 5), thumb-base OA using same KL criteria of either the CMC or STT joints, and erosive hand OA if one erosion was also present on any interphalangel joint.

SxHOA required both presence of radiographic OA and symptoms (pain, aching, stiffness) on most days for the past 30 days. A metabolic syndrome score (MetS) was defined by summating presence of abdominal obesity (defined by gender-specific waist circumference), hypertension, diabetes, and dyslipidemia (range 0-4). Logistic regression was performed adjusting for age, sex and race for all models and other confounders (knee OA, smoking) depending on the outcome examined using a backward selection process to define the most parsimonious model.

Results: The overall 4-year incidence of symptomatic interphalangeal OA (SxIPOA), thumb-based OA(SxTBOA) and erosive hand OA(SxeHOA) were: SxIPOA n=536 (14.9%) ; SxTBOA n=551(15.3%); SxeHOA n=150 (4.2%). Older age, female sex, and knee OA were associated with SxIPOA, SxTBOA, while SxeHOA was not associated with knee OA. Black race compared to whites was inversely associated with all three symptomatic hand OA sub-types. (Table 1) A monotonically increasing risk of SxIPOA was found related to MetS compared to no metabolic risk factors (1 component OR=1.36; 2 components OR=1.50; 3 or 4 components OR=1.59, P for trend <0.01). Of the MetS risk factors dyslipidemia was significantly associated an increased risk of SxIPOA (OR=1.34, 95% CI 1.09,1.65 and SxTBOA(OR=1.34, 95% CI 1.09,1.64) while hypertension was associated with incident SeHOA(OR=1.52, 95% CI 1.06,2.18). MetS was not associated with sxTBOA or SxeHOA nor with incident radiographic defined IPOA, TBOA and eHOA unrelated to symptoms.

Conclusion: Metabolic syndrome risk factors demonstrate a monotonically increased risk for SxIPOA but not with radiographically defined IPOA. Further research into the metabolic pathways, sex and race differences in the progression of symptomatic hand OA appear warranted.