Background/Purpose: Incident symptomatic OA is a common outcome of interest in epidemiologic studies of risk factors for OA.  We postulate that evaluating this outcome in those who do have radiographic OA (ROA) at baseline but do not have frequent knee symptoms will be more informative than the other at-risk groups.  Crepitus is the complaint of hearing         grating, cracking or popping sounds in and/or around a joint.  This is a common symptom in the clinical setting and is possibly predictive of incident knee OA.  Our purpose was to evaluate whether the association between crepitus and incident knee symptomatic osteoarthritis (SOA) is better detected in a group with ROA but without frequent knee symptoms than other at-risk groups.

Methods: This was a person-based longitudinal study using data from the Osteoarthritis Initiative (OAI).  We evaluated the right knee only and specifically focused on people at baseline without SOA.  Crepitus was assessed using the Knee injury and Osteoarthritis Outcome Score (KOOS) questionnaire at baseline.  PA semi-flexed knee radiographs and frequent knee pain (“During the past 12 months, have you had pain, aching, or stiffness in or around your right knee on most days for at least one month?  By most days, we mean more than half the days of a month” were assessed at baseline and 48-month visits.) were assessed at baseline and 48-month visits. ROA was defined as Kellgren and Lawrence (KL) grade ≥ 2.  We performed logistic regression with crepitus as the predictor and incident SOA by the 48 month visit as the outcome.  Subgroup analyses were performed by three groupings: (1) +ROA, –Sx:  those with ROA, but no frequent symptoms at baseline, (2) –ROA, +Sx: with frequent symptoms but no ROA at baseline, and (3) -ROA, –Sx without ROA or frequent symptoms at baseline.

Results: 2936 people each contributed one observation with a mean age of 61.1 (9.3) years and mean BMI of 28.2 (4.7) kg/m².  42% were male.  Incident SOA was least common in knees that never had crepitus and most common in those who always had crepitus (p for trend <0.0001) (Table 1). More than 75% of incident SOA cases originated from the +ROA, –Sx subgroup, despite it constituting about a third of the total number of observations.

Conclusion: Subjective knee crepitus strongly predicts incident SOA over 4 years.  Crepitus evaluated as a subjective assessment, therefore offers utility for identification of at-risk individuals, predictive modeling, and future research.  Selecting those with ROA but without frequent knee symptoms, may optimally power studies to evaluate incident SOA.  TABLES