BACKGROUND: Low back pain and depression are common among total knee arthroplasty (TKA) candidates, and each predicts worse outcomes. Whether their co-occurrence adds beyond either condition alone has not been characterized at registry scale and mean-based analyses may obscure effects confined to the lowest-improving patients.
METHODS: We analyzed 4,173 patients who underwent primary TKA in a national arthroplasty registry. Patients were grouped as Neither (n = 2,261), low back pain only (LBP; n = 1,138), depression only (Dep; n = 394), and Both (n = 380). The primary outcome was covariate-adjusted six-month change in Oxford Knee Score (change in OKS). The secondary analyses included logistic regressions for the minimum clinically important difference (MCID; change in OKS ≥ 8), ordinal regressions for satisfaction, quantile regressions, and additive-interaction tests.
RESULTS: All three groups had lower adjusted mean change in OKS than Neither: LBP β = -1.51 (95% confidence interval (CI) -2.13 to -0.90), Dep β = -1.12 (-2.06 to -0.19), and Both β = -2.08 (-3.13 to -1.04), all below the eight-point MCID. Both had lower odds of achieving MCID (odds ratio [OR] = 0.47, 95% CI 0.32 to 0.68), as did LBP (OR = 0.54). Quantile regression localized the disadvantage to the lowest quartile (Both 25th-percentile β = -3.11). There was no interaction on either scale, indicating additive rather than synergistic effects. Low back pain independently reduced satisfaction (OR = 0.78, 95% CI 0.67 to 0.90). Readmissions and reoperations did not differ.
CONCLUSION: Co-occurring low back pain and depression were associated with small, but significant reductions in mean functional improvement and clinically meaningful reductions in MCID achievement concentrated in the lowest-improving patients. Effects appeared additive rather than synergistic. The findings support preoperative screening for both conditions and distributional, not only mean-based, outcome analysis in registry research.
Read the article: PubMed · Publisher (DOI)