Background/Purpose: It is well recognized that factors beyond structural features contribute to the pain experience in people with knee osteoarthritis (OA). Independent of structural pathology, characteristics such as psychological factors, sleep, and sensitization may increase the risk of developing pain. We examined the relation of pain susceptibility phenotypes to incident pain in people free of pain, but with or at risk of knee OA using indicators of psychological and neurophysiological aspects of pain.

Methods: We used data from the Multicenter Osteoarthritis (MOST) Study, a NIH-funded longitudinal prospective cohort of 3026 older adults with or at risk of knee OA. We identified subjects who were free of frequent knee pain (FKP) (pain on most days during the past month) at both the clinic visit and a telephone screen ~1 month before the clinic visit. We excluded subjects who had a total knee replacement or possible peripheral neuropathy. We used latent class analysis to determine groupings (phenotypes) of baseline psychological and neurophysiological characteristics likely to influence pain (widespread pain, sleep, pain catastrophizing, positive affect, depressive symptoms and quantitative sensory testing (pressure pain thresholds (PPT), temporal summation (TS)). We assessed the relation of these phenotypes to incident consistent FKP (CFKP) (i.e., pain at both the clinic visit and telephone screen ~30 days prior) 2 years later using logistic regression, and examined predictors of class membership (age, sex, education, race, BMI, comorbidities, radiographic knee OA).

Results: 852 participants met inclusion criteria (mean age; 67.1; mean BMI 29.5 kg/m², 55% women); 87 (11%) developed incident CFKP over 2 years. We identified 3 classes (phenotypes) that we labeled as “low”, “moderate” and “high” risk phenotypes based on prevalence of pain risk factors. PPT was a distinguishing characteristic of the high-risk group, while psychological factors were not (Figure 1A). Those in the moderate and low risk groups had significantly lower incidence of CFKP over 2 years OR (95% CI) 0.53 (0.21, 0.86) and OR 0.62 (0.27, 0.96) respectively, compared with the high-risk group. Women were more likely to be in the moderate and high risk groups compared to the low, while those with older age were more likely to be in the high-risk group only (Figure 1B).

Conclusion: In this cohort free of frequent knee pain with or at risk of knee OA, 3 pain susceptibility phenotypes were identified based on psychological and neurophysiological indicators that were associated with differential risk of developing knee pain, irrespective of structural disease. The high-risk pain susceptibility group was predominated by neurophysiological evidence of sensitization. Women were more likely to have higher risk pain susceptibility phenotypes suggesting that sex specific pain phenotypes may be an avenue for future study.