From Hip and Knee to Shoulder: Is the Obesity Paradox Becoming a Surgical Reality?

Obesity remains one of the most pressing global health challenges; it is strongly associated with a wide spectrum of adverse outcomes, including hypertension, type-2 diabetes mellitus, coronary heart disease, heart failure, atrial fibrillation, several cancers, and premature mortality1. In surgical settings, obesity often has been viewed as an independent risk factor for complications2, driving assumptions that higher body mass increases perioperative risk. Yet, over the past 2 decades, a large body of primarily observational evidence has revealed a surprising pattern: in several chronic and acute disease states, overweight individuals or those with moderate obesity have consistently demonstrated better survival than their normal-weight counterparts. This phenomenon, known as the obesity paradox3, has been repeatedly observed in patients with acute coronary syndromes, established coronary heart disease, atrial fibrillation, end-stage renal disease, chronic obstructive pulmonary disease, and pulmonary embolism4,5. Importantly, these observations largely arise from observational studies and therefore do not establish causality. However, they challenge conventional expectations and have prompted vigorous scientific debate about their underlying mechanisms and clinical implications. In the musculoskeletal field, emerging evidence suggests that the obesity paradox may also extend to patients undergoing joint replacement surgery. A recent article on the obesity paradox in joint replacement for osteoarthritis reviewed growing data from large national registries and showed that overweight and obese individuals often experience lower postoperative mortality following hip or knee replacement compared with normal-weight patients, while underweight individuals consistently show markedly higher risk5. That work highlighted that the paradox might be more widespread than previously appreciated and raised important questions about whether body mass index (BMI) should be used to restrict eligibility for joint replacement. Indeed, as pressures increase on health-care systems, some regions and health systems have implemented or considered policies that restrict access to hip and knee arthroplasty with the use of BMI thresholds (often ≥35 or ≥40 kg/m2)6,7, despite inconsistent evidence that higher BMI independently worsens surgical outcomes8. Shoulder arthroplasty, however, differs from hip and knee replacement in several important respects, including older average patient age, a higher prevalence of comorbidities, and distinct biomechanical and functional demands related to upper-limb use rather than weight-bearing9. For shoulder arthroplasty, such restrictions are not formally embedded in policy, but the existing literature has been limited by small sample sizes, inconsistent methods, inadequate control for confounding, and conflicting findings10. These differences limit the generalizability of evidence from lower-limb arthroplasty and underscore the need for shoulder-specific evaluation. Given the rising global demand for shoulder replacement, the absence of high-quality evidence, and concern that BMI-based rationing may exacerbate inequities in care, a robust population-level investigation of BMI and shoulder arthroplasty outcomes is urgently needed. A recent study by Valsamis et al. directly addressed this gap with data from 2 national health-care systems11. Drawing on >20,000 elective primary shoulder replacement procedures from the United Kingdom (2018 to 2022) and Denmark (2006 to 2021), the study examined the association between BMI and several key clinical outcomes, including 90-day and 1-year mortality, 90-day serious adverse events, and revision surgery up to 4.5 years11. Serious adverse events were defined as clinically important medical complications necessitating hospital admission, including conditions such as pulmonary embolism, myocardial infarction, lower respiratory tract infection, acute kidney injury, urinary tract infection, cerebrovascular events, and death from any cause. The findings were remarkably consistent across both countries. Importantly, increasing BMI was not significantly associated with increased risk of 90-day mortality, 90-day serious adverse events, and revision surgery. In fact, increasing BMI was linked to lower mortality at 365 days, strongly echoing the obesity paradox observed in patients who undergo hip and knee arthroplasty and in multiple other clinical settings. By contrast, underweight patients (BMI < 18.5 kg/m2) had substantially worse outcomes, including higher 90-day mortality, higher 90-day serious adverse events, and a greater likelihood of revision surgery11. These findings remained robust across sensitivity analyses, suggesting that BMI-based restrictions for shoulder replacement lack supporting evidence. The study further underscored that underweight patients warrant careful preoperative optimization since their risks appear to be meaningfully elevated. However, these results should be considered alongside prior studies that have reported contrasting findings. A limited number of registry-based and observational analyses, largely from earlier periods, smaller cohorts, and predominantly U.S.-based data sets, have reported heterogeneous associations between obesity and mortality following shoulder arthroplasty, ranging from no association to lower or higher risk12–17. Differences in study design, sample size, era of practice, outcome definitions, adjustment for confounding, and surgical technique may partially explain these discrepancies. To contextualize the findings of Valsamis et al.11 within this heterogeneous literature, key prior studies and their principal findings are summarized in Table 1. Although limited by missing BMI data and the absence of detailed information on comorbidities such as cardiovascular disease (CVD), which may introduce selection bias despite sensitivity analyses, the study by Valsamis et al.11 benefited from large, nationally representative cohorts, validated outcome definitions, harmonized analytic approaches across 2 countries, and contemporary data. As with all observational studies, residual confounding cannot be fully excluded, and causal inferences cannot be drawn. Overall, the totality of evidence suggests that higher BMI is not consistently associated with excess risk following shoulder arthroplasty, even though outcomes among underweight patients appear uniformly poor, highlighting the need for cautious interpretation and policy translation. Table 1 - Observational Studies Assessing the Association Between Body Mass Index and Mortality Following Primary Shoulder Replacement* Publication Location Population Source Baseline Age (yr) Follow-up Duration No. of Joints/Patients Major Indication for Surgery Outcome Findings Singh et al. (2011)12 U.S. Total joint registry 65 (range, 18-67) 90 days 3,480 OA 90-day mortality BMI ≥ 30 kg/m2 was associated with lower risk of 90-day mortality (OR: 0.25; 95% CI: 0.08-0.78) Griffin et al. (2014)16 U.S. Nationwide Inpatient Sample 68.8 In-hospital stay 31,924 NR In-hospital mortality Morbidly obese patients (BMI ≥ 40 kg/m2) had an increased in-hospital mortality rate (0.2%) compared with both obese and non-obese patients (0.1%), but the difference was not significant Anakwenze et al. (2017)17 U.S. Registry 70.1 (range, 45-85) 2.6 yr 4,630 OA 1-year mortality Higher BMI was not associated with higher risk of 1-year mortality Katayama et al. (2024)13 U.S. Institutional database 61.9-72.4 10 yr 466 Degenerative joint disease or OA Survival rate No significant difference in survival between obese patients and underweight/normal weight patients Capotosto et al. (2024)14 U.S. NSQIP database 18-≥75 30 days 22,132 NR Mortality Underweight patients had an increased risk of mortality (OR: 6.18; p = 0.008); no significant difference in mortality risk between patients with BMI < 30 kg/m2 or BMI ≥ 30 kg/m2 Glenn et al. (2025)15 U.S. NSQIP database ≥18 30 days 34,953 OA 30-day mortality Patients who died had lower BMI values *OA = osteoarthritis, BMI = body mass index, OR = odds ratio, CI = confidence interval, NR = not reported, and NSQIP = National Surgical Quality Improvement Program. The persistence of the obesity paradox across cardiovascular, renal, respiratory, and now multiple orthopaedic contexts raises important mechanistic questions. Several plausible explanations have been proposed (Fig. 1). The mechanisms described below should be regarded as hypotheses rather than established explanations in shoulder arthroplasty populations. Higher BMI often correlates with greater absolute lean mass, which confers metabolic and functional advantages, including improved immune responses, greater physiological reserve during acute stress, and enhanced capacity for postoperative recovery18. Cardiorespiratory fitness (CRF) also plays a central role3. Many individuals with obesity maintain adequate or high CRF, and strong evidence demonstrates that CRF substantially improves survival across all levels of adiposity. This “fat but fit” phenomenon may offset the harmful effects that are traditionally attributed to obesity19. In the context of shoulder arthroplasty, where functional recovery depends heavily on upper-limb muscle strength, neuromuscular control, and adherence to postoperative rehabilitation rather than weight-bearing tolerance, preserved muscle mass and physical fitness may be particularly relevant. Additionally, adipose tissue may provide metabolic and anti-inflammatory reserves in the context of acute illness or surgical stress. Conversely, underweight individuals are more likely to exhibit frailty, sarcopenia, and unrecognized chronic illness20, which significantly increase perioperative risk. Residual confounding and selection biases, such as collider bias and confounding by smoking, may contribute, but they are unlikely to fully explain the consistent, cross-condition patterns that have been observed5. The reproducibility of the paradox across different populations and disease states suggests that real physiological mechanisms may be involved, although definitive mechanistic evidence in shoulder arthroplasty remains limited.Fig. 1: Potential mechanisms of the “obesity paradox.” CRF = cardiorespiratory fitness.The clinical and policy implications of these findings are profound. First, they challenge the rationale for using BMI thresholds to restrict access to elective joint replacement. BMI is a crude measure that does not discriminate between fat and lean mass, nor does it capture physical fitness, functional capacity, or frailty21—factors that are much more relevant to surgical outcomes. Evidence from the study by Valsamis et al.11, together with prior research in hip and knee replacement5,22, strongly suggests that BMI should not be used in isolation as a predictor of adverse events; it should be considered alongside other measures within a multimodal risk assessment framework rather than serving as a gatekeeping tool. Patients should be assessed holistically, with emphasis on modifiable risk factors that truly matter, such as physical fitness, muscle strength, metabolic control, and comorbidity management. Second, the findings highlight underweight status as a marker of vulnerability that deserves focused attention. Nutritional optimization, frailty assessment, and multidisciplinary prehabilitation may be particularly important in this group. Third, the shoulder arthroplasty evidence reinforces the potential value of preoperative fitness interventions. Even without substantial changes in BMI, improving CRF may meaningfully reduce postoperative risk and improve long-term outcomes. Looking ahead, additional research is needed to better understand the mechanistic pathways underlying the obesity paradox in shoulder arthroplasty and beyond. Importantly, the existing literature is characterized by substantial heterogeneity and conflicting findings, with some studies suggesting excess risk among patients with obesity but others demonstrating no increased risk or paradoxical associations. These contradictions underscore the need for higher-quality evidence, including well-designed prospective studies with rigorous confounder control, standardized outcome definitions, and sufficient power to assess clinically meaningful end points. Integrating body composition assessments, direct measures of CRF, inflammatory biomarkers, and granular comorbidity data will help to disentangle the contributions of adiposity, muscle mass, fitness, and frailty. Prospective studies evaluating targeted prehabilitation strategies across BMI categories would clarify whether improving fitness can mitigate risk, regardless of weight. Additionally, understanding whether the paradox is modified by age, sex, comorbid CVD, or underlying musculoskeletal pathology will have practical relevance for personalized care. Overall, the new evidence by Valsamis et al.11 offers a timely and compelling contribution to ongoing debates about the role of BMI in surgical decision-making. While this large, well-conducted population-based study adds important contemporary data to the literature, it should be interpreted alongside a broader body of observational and registry-based studies that have reported heterogeneous and sometimes opposing findings. The findings did not provide definitive resolution, but rather highlighted the complexity of the relationship between BMI and outcomes following shoulder replacement. Accordingly, the existing evidence base remains insufficient to support rigid BMI-based rationing policies in isolation, and underscores the value of focusing on fitness, frailty, and holistic patient optimization. As the field continues to evolve, progress will depend on higher-quality, methodologically rigorous studies that reconcile conflicting results and better inform equitable, evidence-based surgical decision-making.