1 Introduction

Over the past 5 decades, obesity has increased in prevalence, with recent studies showing that around 40% of adults in the United States are obese.^1,2 This trend represents a major concern for public health, as obesity is linked with non-communicable comorbidities, such as diabetes, cardiovascular disease, and malignancy.³ Despite strong efforts in preventing and treating obesity, the effectiveness of lifestyle modification and pharmacologic interventions remain limited.⁴ As a result, obesity leads to a large economic burden, with medical care costs increasing progressively across weight classes. Indeed, individuals with class III obesity may pay up to 240% more in medical costs than those with normal weight.^5,6

However, the relationship between obesity and bone fractures is complex. Many prior studies suggest that obesity plays a protective role against fractures, potentially due to increased bone mass and favorable microarchitecture.⁷ Other studies suggest that the prevalence of fractures in obese individuals is dependent on the site of injury, demonstrating that the impact of obesity varies by location.⁸ Though prior works have explored obesity and fracture incidence, less is known about how obesity associates with the anatomic distribution of fractures along the length of long bones. A prior study has reported a positive association between increasing BMI and distal femur fractures, in contrast to the more common proximal femur fracture.⁹ However, this relationship has not been explored in all bones of the lower extremity.

This study uses a large, nationally representative database to examine the association between BMI and fracture location along the femur, tibia, fibula, and malleoli. We hypothesize that increasing BMI is associated with a shift toward less common, more distal fracture locations.

2 Methods

This study used data from the National Inpatient Sample (NIS) database which included information from 2016 to 2019. This database utilizes inpatient and disposition data from hospitals across 47 states in addition to the District of Columbia. The NIS represents about 20% of United States hospitals yet is estimated to be representative of 97% of the country's population. Patient demographics, baseline comorbidities, diagnoses, mortality, post-operative complications, discharge disposition, and multiple other factors are available in this database. The study was approved by the local institutional review board and was conducted according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.

The data from the NIS was collected for all patients that included a BMI diagnosis as defined by ICD-10 diagnosis code. The most up-to-date coding was manually collected to include BMI ICD-10 codes of five different categories; underweight (<19.9 kg/m²), normal weight (20.0-24.9 kg/m²), overweight (25.0-29.9 kg/m²), class I/II obesity (30.0-39.9 kg/m²), and class III obesity (>40.0 kg/m²). The baseline demographic data collected was age, sex, race, income level, and Charlson comorbidity index. The primary outcome was fracture location among patients with lower extremity long bone fractures. The three analyzed long bones (femur, tibia, fibula) were each subdivided by location into proximal (upper) end, shaft, and distal (lower) end. In addition, other subdivisions included the tibial pilon, bimalleolar and trimalleolar ankle fractures, and subtrochanteric and pertrochanteric fractures of the femur. Multivariable logistic regression analysis was performed to evaluate the association between BMI category and fracture location. Records were excluded if no BMI diagnosis code was present on the same discharge record, if multiple BMI diagnosis codes were present, or if multiple concomitant fracture location codes were present. The full list of ICD-10 Codes, including BMI categories, fracture location, and comorbidities, that were utilized for this study can be found in Supplemental Digital Content 1.

The data was analyzed with descriptive statistics using SAS Proc SurveyFreq version 9.4. Multivariate logistic regression models were produced for each fracture location to determine odds ratios (ORs) with 95% confidence intervals (CIs), and p-values for the independent influence of the BMI. Statistical significance was determined as p-value <0.05. Figures were generated using R (version 4.3.2).

3 Results

Between 2016 and 2019, a total of 1,949,605 patients sustained an isolated long bone fracture of the lower extremity. After excluding those without a BMI diagnosis code, 262,535 patients were included in the dataset. Among these, 33,475 (12.8%) were tibia fractures, 11,975 (4.6%) were fibula fractures, 33,420 (12.7%) were malleolar fractures, and 183,665 (70.0%) were femoral fractures. Among the tibia, 18,595 (55.6%) were proximal tibia fractures, 4290 (12.8%) were tibia shaft fractures, 7530 (22.5%) were distal tibia fractures, and 3060 (9.1%) were tibial pilon fractures. Among the fibula, 560 (4.7%) were proximal fibula fractures, 4380 (36.5%) were fibula shaft fractures, and 7035 (58.8%) were distal fibula fractures. Among the malleoli, 16,475 (49.3%) were bimalleolar fractures and 16,945 (50.7%) were trimalleolar fractures. Among the femur, 67,240 (36.5%) were proximal femur fractures, 11,430 (6.2%) were femoral shaft fractures, 22,815 (12.4%) were distal femur fractures, 74,145 (40.4%) were pertrochanteric fractures, and 7855 (4.28%) were subtrochanteric fractures. Throughout all lower extremity fractures, BMI codes were entered inconsistently, ranging from only 10.8% of all proximal femoral fractures possessing a BMI code to 22.9% of proximal fibular fractures possessing a BMI code.

Progression of BMI category was assessed within each specific bone fracture. As BMI category increased across cohorts (underweight, normal weight, overweight, obese class I/II, and obese class III), the progression of subtrochanteric, shaft, and distal femoral fractures increased in a stepwise manner. Pertrochanteric and proximal femoral fractures decreased in a stepwise manner. Distal fibular fractures increased with BMI until obese class III, while proximal fibular fractures remained uncommon across all classes. The proportion of tibial pilon fractures increased with BMI, followed by a plateau after obese class I. There were no other stepwise trends noted in the data. The number of fractures associated with each long bone based on BMI category is visualized in Tables 1–4 and Figs. 1–4.

Close

Fig. 1 Distribution of femoral fracture location by body mass index category. Grouped bar graph demonstrating the distribution of femoral fracture locations across body mass index (BMI) categories. Bars represent the percentage of fractures occurring at each anatomic location within each BMI group. Femoral fracture locations include subtrochanteric, pertrochanteric, proximal, shaft, and distal regions.

Export to PPT

Close

Fig. 2 Distribution of tibial fracture location by body mass index category. Grouped bar graph showing the proportion of tibial fractures by anatomic location across BMI categories. Percentages reflect the distribution of fracture location within each BMI group. Tibial fracture locations include proximal, shaft, and distal, and tibial pilon fractures.

Export to PPT

Close

Fig. 3 Distribution of fibular fracture location by body mass index category. Grouped bar graph illustrating the distribution of fibular fracture locations stratified by BMI category. Bars represent the percentage of fractures within each BMI group occurring at the proximal, shaft, and distal fibula.

Export to PPT

Close

Fig. 4 Distribution of malleolar fracture patterns by body mass index category. Grouped bar graph depicting the distribution of malleolar fracture patterns across BMI categories. Percentages represent the proportion of fractures within each BMI group classified as bimalleolar or trimalleolar.

Export to PPT

Multivariate regression analysis demonstrated that increasing age was significantly associated with higher odds of proximal femur (OR = 1.04, P < 0.0001), pertrochanteric (OR = 1.04, P < 0.0001), and subtrochanteric (OR = 1.01, P < 0.0001) fractures, with significantly decreased odds of all other lower extremity fractures. Charlson comorbidity score was associated with significantly increased odds of proximal femur (OR = 1.01, P < 0.0001), distal femur (OR = 1.04, P < 0.0001), pertrochanteric (OR = 1.03, P < 0.0001), fibula shaft (OR = 1.08, P < 0.0001), and distal fibula (OR = 1.05, P < 0.0001) fractures. Tobacco use was associated with significantly increased odds of proximal femur (OR = 1.10, P < 0.05), pertrochanteric (OR = 1.23, P < 0.0001), and all tibial fractures. Female sex was associated with significantly increased odds of distal femur (OR = 1.58, P < 0.0001), bimalleolar (OR = 1.43, P < 0.0001), and trimalleolar (OR = 2.15, P < 0.0001) fractures.

BMI was found to have a significant effect on the location of all lower extremity long bone fractures. Underweight BMI significantly increased odds of proximal femur (OR = 1.23, P < 0.0001) and pertrochanteric (OR = 1.78, P < 0.0001) fractures. Underweight BMI was associated with significantly decreased odds of all other fractures of the lower extremity. Individuals in the normal weight category had increased odds of proximal femur (OR = 1.14, P < 0.0001) and pertrochanteric (OR = 1.28, P < 0.0001) fractures, and had significantly decreased odds of fibular shaft (OR = 0.68, P < 0.05), distal fibula (OR = 0.68, P < 0.05), bimalleolar (OR = 0.37, P < 0.0001), trimalleolar (OR = 0.34, P < 0.0001), and all tibial fractures. Overweight BMI had significantly increased odds of distal femur (OR = 1.12, P < 0.05), distal tibia (OR = 1.22, P < 0.05), tibial pilon (OR = 1.39, P < 0.05), distal fibula (OR = 1.31, P < 0.05), bimalleolar (OR = 1.35, P < 0.0001), and trimalleolar fractures (OR = 1.23, P < 0.05), and significantly decreased odds of proximal femur fractures (OR = 0.92, P < 0.05). The Obese I/II category was associated with significantly decreased odds of proximal femur (OR = 0.55, P < 0.0001) and pertrochanteric (OR = 0.73, P < 0.0001) fractures but had significantly increased odds of all other lower extremity fractures. Lastly, the Obese III category followed the same trends as the obese I/II category, however it did not significantly correlate with subtrochanteric fractures. The full multivariate regression results can be seen in Table 5.

4 Discussion

In this nationally representative cohort of patients with lower extremity fractures, BMI was associated with fracture location. Increasing BMI was associated with a relative shift toward distal fractures of the femur, tibia, and fibula, as well as higher odds of ankle fractures, whereas underweight BMI was associated with proximal femoral fracture patterns. These findings add to prior work on BMI and fracture risk by showing that BMI is associated with fracture location among patients with lower extremity long bone fractures. However, prior studies are generally limited to single bones and do not evaluate whether fracture distributions extend across multiple lower extremity long bones.

The femoral fracture findings of this study closely parallel prior reports demonstrating that increasing BMI is associated with a shift away from proximal femoral fractures and towards more distal areas.^10,11 In this study, obesity was associated with distal femoral, subtrochanteric, and shaft fractures, while inversely associated with more proximal femur fractures such as intertrochanteric fractures and fractures of the femoral neck. The reduction in proximal femur fractures in obese individuals may be due to increased padding around the hip, altering the force distribution. Consistent with this pattern, previous studies have shown that obese patients more frequently sustain subtrochanteric and femoral shaft fractures, which are associated with greater operative complexity and longer operative times.^12,13 Additionally, despite an overall increased fracture burden, obese patients experience fewer classic hip fractures, further supporting a BMI-associated shift away from proximal femoral injury.¹⁴

Outside of the femur, there were observed trends toward distal tibial, tibial pilon, distal fibular, and bi- or trimalleolar ankle fractures as BMI increased. These findings align with a prior study showing that central obesity is associated with distal lower limb fractures.¹⁵ Similarly, pediatric studies have demonstrated increased distal tibial fractures with higher BMI. Outside of the long bones, smaller-bone studies have additionally shown shifts in fracture zones based on BMI, including increased fractures in weight-bearing bones such as the fifth metatarsal, supporting the idea that BMI may influence where fractures occur.^16,17 These distal fractures may be due to increased axial loading at the ankle due to the high BMI. Studies that look at fracture risk in obese populations show that obesity is associated with more complex fracture morphology and increased morbidity in lower extremity injuries.^18,19 Although the focus was on long bone fractures only, the findings support the idea that obesity itself may alter injury mechanisms in the lower extremity.

The relationship between BMI and bone architecture is complex, and prior studies have shown that higher BMI is associated with increased bone mineral density (BMD) and this is thought to be due to greater mechanical loading and the endocrine impact of adipose tissue ^20–24. Nonetheless, even though these patients have higher BMD, some studies show that obese individuals experience a higher rate of fractures at specific skeletal sites ^25–27. Since these patients have a higher BMD but differing risks in fractures, the body composition or mass distribution, as opposed to BMD alone, may play a more important role in determining fracture risk. Indeed, studies that look at abdominal obesity have found that fat distribution is a stronger predictor of fracture risk than BMI. For example, in prior studies, waist circumference was associated with distal lower limb fractures, even in those with a normal BMI.^15,28

In this cohort, underweight individuals were more likely to sustain proximal femoral and pertrochanteric fractures. This aligns with prior studies demonstrating that low body mass index is associated with a disproportionately increased risk of hip and proximal femoral fractures ^29–32. It has consistently been shown that low BMI can lead to frailty, sarcopenia, and a reduced BMD, ultimately increasing fracture risk ^33–35. Other studies demonstrate a U-shaped association between BMI and fracture risk, with both underweight and obese individuals experiencing the highest rates of fractures.^36,37 Underweight patients often have low muscle mass, which increases fall risk, and the reduced bone density and soft tissue protection could lead to vulnerability of the proximal femur to fracture. In addition, studies show that underweight patients with fractures experience longer hospital stays and higher readmission rates.³⁸

Taken together, prior studies evaluating BMI and fracture location have been limited to single bones, specific fracture zones, or specific populations, and this study is the first to demonstrate a BMI-dependent shift in fracture location along the length of multiple lower extremity long bones in a nationally representative cohort. The findings of this study have several clinical implications. First, BMI should not only be considered in fracture risk, but also in predicting location. Surgeons should expect a higher likelihood of distal and shaft fractures in high BMI patients, which may require different operative planning. Next, fracture prevention could be based around BMI category, as obesity increases risk for distal fractures while underweight patients are at risk for proximal femoral injuries. However, these findings still highlight a central issue with relying on BMD or BMI when assessing fracture risk, as body composition and mass distribution also play important roles.

This study has several limitations. BMI coding within the NIS is inconsistent and was preferentially recorded at the extremes of body mass. In this cohort, underweight and obese BMI categories were documented substantially more frequently than normal weight and overweight. As a result, fracture proportions among normal weight and overweight individuals may be underestimated relative to their true rates. Next, BMI may not reflect true body composition and does not account for fat distribution, which may alter fracture patterns. Third, the NIS lacks data on the mechanism of injury and outpatient-managed fractures, limiting the ability to account for fracture energy or outcomes outside of the inpatient setting, further reducing the cohort. Finally, the present study could not account for confounders such as bone mineral density, sarcopenia, muscle mass, and muscle strength which may influence fracture patterns.

Future studies should involve a prospective approach with detailed injury mechanisms and imaging to provide a better idea of outcomes associated with fractures in patients of varying BMI classes. Secondly, investigating waist circumference, visceral adiposity, and muscle mass will help improve the understanding of characteristics associated with fracture location. Finally, evaluation of whether fracture patterns related to BMI translate into differences in surgical technique, complication rates, union, and outcomes could contribute to prevention and treatment planning.

5 Conclusions

In this large, nationally representative cohort, BMI was a strong independent correlate of lower extremity fracture location. Increasing BMI was associated with a shift toward more distal fracture patterns of the femur, tibia, fibula, as well as higher odds of ankle fractures, and underweight BMI was associated with proximal femur fractures. These findings demonstrate that BMI is associated with the anatomic distribution of lower extremity fractures. BMI has important implications for injury mechanism, supporting its incorporation into clinical risk assessment and decision-making.

Patient consent

Patient consent was waived due to the retrospective nature of the study and the use of de-identified data.

Ethical statement

This study was reviewed and approved by the Southern Illinois University School of Medicine Institutional Review Board (IRB #23-275). The study was conducted in accordance with the ethical standards of the Declaration of Helsinki (2013). Given the retrospective nature of the study, the requirement for informed consent was waived.

CRediT author statement

Noah B. Drewes – Conceptualization, Methodology, Investigation, Writing – Original Draft, Visualization.

Kyle Struck – Investigation, Writing – Review and Editing.

Anthony Sleiman - Investigation, Writing – Review and Editing.

Jeffrey Baker - Investigation, Writing – Review and Editing.

Jake Walters - Investigation, Writing – Review and Editing.

Carson Lewis - Investigation, Writing – Review and Editing.

Shaddy Khalafallah - Investigation, Writing – Review and Editing.

Steve Scaife – Validation, Formal Analysis, Investigation, Data curation,

Sowmyanarayanan Thuppal – Conceptualization, Supervision, Project Administration,

Funding sources

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.