BackgroundWhile morbid obesity has been associated with increased complication risk in primary total knee arthroplasty (TKA), limited evidence is available to attribute decreased surgical complication rates with body mass index (BMI) reduction. MethodsWe retrospectively assessed 464 unilateral TKAs performed in morbidly obese patients, including 158 extremely obese (BMI ≥ 45) and 306 severely obese patients (BMI 40 to 44.9). A detailed medical record review identified concurrent modifiable risk factors and successful preoperative BMI reduction, reaching either a contemporary risk target (BMI < 40) or an institutionally accepted threshold (BMI < 45). Postoperative blood glucose levels and 1-year adverse outcomes (periprosthetic joint infection, wound dehiscence, knee manipulation, periprosthetic fracture) were compared to 557 contemporary control subjects with expected slightly lower (moderate obesity, BMI 35 to 39.9) or sufficiently lower complication risk (overweight, BMI 25 to 29.9). ResultsPeriprosthetic joint infection and postoperative hyperglycemia were identified more frequently among morbidly obese patients in comparison with a moderately obese control group. Extremely obese patients (BMI ≥ 45) whose BMI improved below 45 had no measurable difference in infection risk from the control group (odds ratio [OR] 0.84, 95% confidence interval [CI] 0.04 to 16.88), while those with a nonimproved BMI had a significantly higher risk (OR 7.70, 95% CI 1.89 to 31.41). No significant differences in the risk for infection were observed between severely obese patients (BMI 40 to 44.9) with preoperative BMI improvement (1.5% rate, OR 1.70, 95% CI 0.17 to 16.57) or nonimprovement (1.7% rate, OR 1.87, 95% CI 0.41 to 8.43). ConclusionsPreoperative medical optimization may decrease postoperative TKA complications. The findings of this study support BMI improvement for extremely obese patients (BMI ≥ 45). The assignment of 40 BMI as a threshold for otherwise healthy patients may exclude patients from potential surgical benefits without realizing risk reduction.