Background/Purpose: The prevalence of hand osteoarthritis (OA) has doubled in the United States over the past half century – a finding that cannot be solely explained by increases in longevity, raising the possibility that environmental factors play a role. In cross-sectional studies, exposure to per- and polyfluoroalkyl substances (PFAS; “forever chemicals”), synthetically produced water-, grease-, or stain-resistant chemicals, has been associated with self-reported OA diagnosis. Given the ongoing use of PFAS and the scale of contamination, there is an urgent need for longitudinal studies to clarify the relationship between PFAS and specific phenotypes of OA (e.g., symptomatic hand OA) to inform public health policies and clinical monitoring. Hence, we explored whether specific PFAS and a mixture of PFAS – reflecting the joint effects and correlations among each PFAS – relate to incident hand OA and progression.

Methods: We leveraged a case-cohort sample (n = 1,878) from the Osteoarthritis Initiative among participants with hand radiographs at the baseline and 48-month visits. We assayed baseline serum samples for 12 PFAS via online solid-phase extraction liquid chromatography tandem mass spectrometry. For analyses, we a priori included only 8 PFAS for which 60% of samples were above the limits of detection. Our 2 primary outcomes were incident symptomatic hand OA and an increased number of joints with radiographic OA (Kellgren-Lawrence≥2; yes/no). We classified a participant as having incident symptomatic hand OA if at the 48-month visit, but not at baseline, they had ≥2 finger joints (distal interphalangeal, proximal interphalangeal, or metacarpophalangeal joints) on at least 2 separate digits with at least a small osteophyte or mild joint space narrowing (modified Kellgren-Lawrence grade≥2) on radiograph and reported on a questionnaire hand/finger pain, aching, or stiffness on more than half of the 30 days prior to the radiograph. We used covariate-adjusted weighted logistic regression models to assess single PFAS (continuous and quartiles) and quantile-based g-computation to assess the PFAS mixture in relation to our primary outcomes.

Results: rticipants were mostly female (58%), White (83%), and on average 61 years of age and overweight (mean BMI = 28.6 kg/m2). Participants with higher serum PFDA and PFNA had greater odds of incident symptomatic hand OA [OR (95%CI) per interquartile-range increment: 1.12 (1.05, 1.20) for PFDA; 1.07 (1.00, 1.13) for PFNA], but the associations were not monotonic when represented in quartiles (Table 1). Participants with higher PFHxS had lower odds of hand OA [e.g., OR (95%CI): 0.93 (0.86, 1.00) per interquartile-range increment]. Participants in the highest quartile of PFHpS (vs. quartile 1) had lower odds of hand OA. We observed no other monotonic associations between individual PFAS or the PFAS mixture with either primary outcome (Tables 1 and 2).

Conclusion: We observed possible associations of PFDA and PFNA with symptomatic hand OA incidence that warrant further study in cohorts with more nuanced measures of hand OA symptoms, but we otherwise found no consistent evidence that greater serum PFAS concentrations relate to a greater chance of developing hand OA incidence or progression.

Table 1. Associations [OR (95%CI)] of baseline serum PFAS concentration with symptomatic hand OA incidence among 1,729 participants without baseline hand osteoarthritis

Table 2. Associations [OR (95%CI)] of baseline serum PFAS concentration with having an increase (vs. not) in total number of joints with a KL grade ≥2 between baseline and the 48-month follow-up visit (N&#3f1,878)

« Back to ACR Convergence 2025

ACR Meeting Abstracts - https://acrabstracts.org/abstract/per-and-polyfluoroalkyl-substances-and-hand-osteoarthritis-data-from-the-osteoarthritis-initiative/