BackgroundDepression often accompanies knee osteoarthritis (OA), exacerbates pain severity, and may negatively affect analgesic treatment outcomes.ObjectivesTo determine whether depression moderates the effect of analgesics on pain severity in persons with or at-risk for symptomatic knee OA.MethodsParticipants (n=2059) were from the Osteoarthritis Initiative, with or at-risk for symptomatic knee OA, and had complete data on selected measures at baseline and four annual follow-up visits. Analgesic initiation (acetaminophen, non-steroidal anti-inflammatory drugs, opioids) was assessed at three annual follow-up visits in those who were not analgesic users at baseline. Depression was evaluated concurrent to assessment of analgesic use with the Center for Epidemiological Studies Depression (CES-D) scale using the corresponding CES-D screening threshold (CES-D score ≥ 16). Pain severity at the fourth annual follow-up visit was the outcome and measured with the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) pain scale (rescaled range = 0-100). Time-invariant confounders measured at baseline were: age, gender, race, education, marital status, employment, health insurance, smoking, alcohol consumption, Charlson comorbidity index, and symptomatic knee OA status. Time-varying confounders measured at three annual follow-up visits were: WOMAC pain score, Kellgren-Lawrence grade, body mass index, physical performance, knee injuries, and knee injections. Structural nested mean models appropriate for evaluating time-varying effect moderation of dynamic treatments by evolving effect modifiers were implemented using an inverse-probability-of-treatment-weighting regression-with-residuals approach.ResultsIn non-depressed participants, analgesic treatment effect at years one, two, and three on pain severity at year four was minimal (Figure 1). Time-specific associations in the non-depressed ranged from -1.18 (95% CI: -2.94, 0.59) to 0.81 (95% CI: -0.99, 2.61) and were not statistically significant. The persistent use of analgesics at years one and two (β = -2.17; 95% CI: -4.62, 0.27) or years one, two, and three (β = -1.36; 95% CI: -4.22, 1.50) did not increase the magnitude of the treatment effect in non-depressed participants. In contrast, time-specific associations in depressed participants increased during follow-up from 0.61 (95% CI: -9.17, 10.38) to -9.64 (95% CI: -17.96, -1.33), and the treatment effect of year three analgesic use on year four pain severity was statistically significant. The magnitude of the treatment effect increased from year one with persistent analgesic use at years one and two (β = -5.75; 95% CI: -20.65, 9.15) and years one, two, and three (β = -15.39; 95% CI: -33.99, 3.21). However, effect moderation was only significant concerning year three analgesic use, where the subsequent difference in treatment effect between depressed and non-depressed participants was -10.46 (95% CI: -18.97, -1.94).Abstract THU0449 – Figure 1ConclusionFindings indicate a significantly greater one-year trea...