Hip and knee replacements, also known as total hip arthroplasty (THA) and total knee arthroplasty (TKA), are commonly performed surgical procedures intended to relieve pain and restore joint function by replacing damaged or diseased joints. Obesity is a well-established risk factor for complications following these procedures, including surgical site infections, mechanical failure of the implant, and the need for revision surgery.1 As a result, many health systems and insurers require a certain BMI threshold to be eligible for surgery, and patients are frequently advised to lose weight before undergoing joint replacement. However, the evidence for preoperative weight loss as a strategy to reduce surgical complications is mixed. Some prior studies have found that patients who lost weight before joint replacement did not have improved infection or readmission rates compared to those whose weight remained stable,2 and other studies reported that patients who lost weight before hip replacement and maintained that loss actually had a higher likelihood of deep surgical site infection.3 With the growing use of GLP-1 receptor agonists for weight management, more patients might present for surgery after significant weight loss, making it increasingly important to understand how the magnitude of preoperative weight change relates to postoperative outcomes.
To understand how preoperative BMI changes across a range of starting weight categories relate to the risk of surgical site infections and mechanical failures after hip and knee replacement surgery, we studied more than one million adult patients who had one of these procedures between January 2017 and September 2025. We classified patients into five groups based on the percentage change in their BMI between a reading from the year prior to surgery and their BMI reading from within a month of the procedure: 2% or more gain, stable (within 2% change), 2% to less than 10% loss, 10% to less than 20% loss, and 20% or more loss. These weight change groups were further stratified by starting BMI class (overweight, obese class 1, obese class 2, and severely obese), comparing each weight change group to patients in the same class whose weight remained stable. We accounted for demographics, Area Deprivation Index based on most recently documented address, and relevant comorbidities (such as diabetes, obstructive sleep apnea, cardiovascular disease, hypertension, and conditions suggesting immunosuppression).
Compared to patients in the same starting BMI class whose weight remained stable, the likelihood of surgical site infection after knee replacement increased with greater preoperative weight loss, with the steepest gradient among overweight (BMI 25-<30) and class 1 obesity (BMI 30-<35) patients, as seen in Figure 1. Among overweight patients, each level of additional weight loss corresponded to a progressively higher likelihood of infection, reaching more than six times as likely among those who lost 20% or more. Among patients with class 1 obesity, losses of 20% or more were associated with more than four times the likelihood. The gradient was less steep at higher starting BMI classes; among severely obese (BMI 40+) patients, losses of 20% or more were associated with only a 27% higher likelihood. Weight gain was also associated with modestly elevated likelihood across all BMI classes.
Compared to patients in the same BMI class whose weight remained stable, preoperative weight loss was also associated with higher likelihood of surgical site infection after hip replacement among overweight, class 1, and class 2 obesity patients, as seen in Figure 2. Among overweight patients, losses of 20% or more were associated with a 169% higher likelihood, and among class 1 obesity patients, a 152% higher likelihood. Among severely obese patients, however, preoperative weight loss was not associated with a statistically significant change in infection likelihood at any level of loss. Weight gain of 2% or more was associated with a modestly higher likelihood across all BMI classes, ranging from a 31% to 38% increase.
Compared to patients in the same BMI class whose weight remained stable, preoperative weight loss showed associations with mechanical failure after knee replacement that was dependent on starting BMI. Among patients with an overweight BMI, losses of 20% or more were associated with a 128% higher likelihood, and among class 1 obesity patients, a 102% higher likelihood. Among severely obese patients, greater weight loss corresponded to a lower likelihood of mechanical failure, though not statistically significantly. Losses under 10% showed little meaningful association for any BMI class.
Compared to patients in the same BMI class whose weight remained stable, the relationship between preoperative weight change and mechanical failure after hip replacement was the most variable of the four outcomes. Among overweight patients, losses of 10–20% were associated with a 103% higher likelihood and losses of 20% or more with a 138% higher likelihood, as seen in Figure 4. Among class 1 obesity patients, losses of 20% or more were associated with a 169% higher likelihood. Among severely obese patients, weight loss was not associated with higher mechanical failure likelihood at any level.
This study measured the association between preoperative weight change and surgical complications but cannot distinguish whether weight loss itself influenced outcomes or whether it reflects underlying factors, such as illness, frailty, or nutrition status, which might independently affect surgical risk. Additionally, obesity carries well-established risks beyond those studied here.1,3
