BackgroundThe American Academy of Orthopaedic Surgeons (AAOS) defined the acceptable threshold for elective safe surgery as a body mass index (BMI) under 40 due to the increased risk of complications. A consequence of this recommendation has been a hard cutoff based on BMI, which restricts access to care for an increasingly large and diverse population. There is an improved understanding that excess adipose tissue confers additional risk for postoperative complications, including infection, through mechanical and physiologic mechanisms. But it is unclear if BMI is an accurate indicator of adiposity in total joint arthroplasty (TJA) patients and, thus, whether BMI is capturing clinically relevant information in obese patients. Our objective was to determine the relationship between peri-incisional adiposity (PIA) and BMI in a consecutive series of diverse primary TJA patients. MethodsA consecutive series of patients indicated for primary TJA were preoperatively evaluated. For each patient, the following variables were collected: BMI, and measures of peri-incisional adiposity (PIA) on radiographs and ultrasounds. ResultsIn THA patients (N = 99), Pearson’s correlation coefficient (r) = 0.436, which indicates a moderate correlation between BMI and adiposity. In TKA patients (N = 271), r = 0.395 for femoral PIA (FPIA) and r = 0.249 for tibial PIA (TPIA), which indicates a weak correlation between BMI and adiposity measured on radiography. In TKA patients, r = 0.560 for FPIA and r = 0.544 for TPIA, which indicates a moderate correlation between BMI and adiposity measured on ultrasound. ConclusionQuantification of obesity has become common practice in the preoperative evaluation of primary TJA patients. The intent is to determine the magnitude of adipose tissue, which is one of the main drivers of postoperative complications in obesity. The BMI is ubiquitously used as a proxy for obesity due to its simplicity of attainment and calculation. We report that BMI has a weak to moderate association with peri-incisional adiposity in this population. These findings indicate that BMI may not accurately represent the condition of peri-incisional adipose tissue and, thus, is not capturing the relevant obesity data for preoperative risk stratification in primary total joint arthroplasty patients.