1 Introduction

Celiac disease (CD) is a chronic autoimmune enteropathy characterized by an abnormal immune response to dietary gluten, primarily affecting the small intestine.¹ With a global prevalence estimated between 0.7 % and 2.9 %,^2,3 CD has gained increasing public and clinical attention over the past few decades.⁴ While CD predominantly manifests as gastrointestinal symptoms such as diarrhea, malabsorption, abdominal pain, bloating, and vomiting, it can also present with extraintestinal manifestations, including dermatitis, chronic fatigue, iron deficiency anemia, osteoporosis, arthritis, and neurological symptoms.^5–7 The disease exhibits a bimodal age distribution, with peaks of onset occurring around two years of age and again between 20 and 30 years.^8,9 Individuals with a first-degree relative diagnosed with CD, those with type 1 diabetes mellitus (T1DM), or those with other autoimmune disorders are at a heightened risk of developing the disease.¹⁰

As a lifelong condition with no definitive cure, the primary management of CD is strict adherence to a gluten-free diet, which is generally effective in alleviating symptoms and preventing disease-related complications.¹¹ However, while dietary modifications can improve bone mineral density (BMD),¹² complete normalization is not always achievable. This can result in adverse consequences, including reduced dietary fiber intake, micronutrient deficiencies, and potential exposure to heavy metals.¹³ Notably, up to 75 % of patients with CD, and 40 % of those diagnosed in adulthood, present with low BMD, predisposing them to compromised bone microarchitecture, increased fracture risk, and impaired absorption of vitamin D and calcium.^14–17 While a gluten-free diet has been shown to enhance BMD in 9–47 % of affected individuals,¹⁶ those who do not experience improvement may suffer from persistent nutritional deficiencies. Furthermore, inadequate preoperative nutritional status has been associated with an increased risk of complications following total joint arthroplasty, including surgical site infections, periprosthetic infections, and hospital readmission.^16,18,19

Total knee arthroplasty (TKA) is a widely performed surgical intervention aimed at alleviating pain and restoring function in patients with end-stage knee osteoarthritis.²⁰ Recognized as one of the most effective and cost-efficient orthopedic procedures,²¹ the demand for TKA has risen considerably. Projections indicate an 85 % increase in TKA procedures in the United States from 2014 to 2030, reaching an estimated 1.26 million cases.²² Additionally, Singh et al. forecasted that, compared to 2014, the number of TKAs in the U.S. would rise by 56 % by 2020, 110 % by 2025, 182 % by 2030, and 401 % by 2040.²³ Notably, individuals undergoing TKA between the ages of 70 and 80 years have demonstrated similar revision and mortality rates, reinforcing the safety and feasibility of the procedure in elderly populations.^24,25 As life expectancy continues to rise in the United States,²⁶ the annual volume of TKAs is expected to increase correspondingly. Although TKA is generally considered safe, the procedure carries inherent risks, including infection, pulmonary embolism, and postoperative complications,^27,28 all of which must be considered when counseling patients.

Despite the established impact of CD on bone and joint health, its influence on arthroplasty outcomes remains poorly understood. A review of the current literature yields only one study examining the relationship between CD and total joint arthroplasty, specifically total hip arthroplasty (THA) and TKA.¹⁶ Given this gap in knowledge, the present matched cohort study aims to assess the impact of CD on postoperative outcomes in patients undergoing TKA. We hypothesize that individuals with CD will exhibit higher rates of complications within the 90-day postoperative period compared to matched controls. The findings of this study will provide critical insights for orthopedic surgeons and healthcare providers when counseling CD patients who are candidates for total joint replacement.

2 Methods

Database: This study utilized the PearlDiver database (PearlDiver Inc., Fort Wayne, IN, USA), a national insurance claims repository affiliated with Humana Health Insurance. The database facilitates large-scale retrospective analyses and allows patient identification using International Classification of Diseases, Ninth and Tenth Revision (ICD-9 and ICD-10) codes, as well as Current Procedural Terminology (CPT) codes.

Data Collection: A retrospective cohort study was conducted to identify patients aged 18 years and older who underwent TKA between January 1, 2010, and December 31, 2020. Patients were required to have a minimum of three months of continuous follow-up within the database following their procedure; those not meeting this criterion were excluded.

Patients were stratified into disease and control cohorts based on a preoperative diagnosis of CD. The final study population included 11,562 patients, of whom 5,781 (50 %) had a documented diagnosis of CD prior to undergoing TKA. A 1:1 matching protocol based on age and gender was employed to ensure comparability between groups, resulting in equal numbers of patients in both cohorts. Baseline demographic characteristics, including age (>50 years) and sex, were comparable between groups. The cohorts were further analyzed for differences in comorbid conditions, including heart disease, diabetes, tobacco use, obesity, rheumatoid arthritis, osteoarthritis, chronic kidney disease, and hypertension.

The primary outcomes of interest included postoperative complications occurring within 90 days of the index procedure, encompassing infection, emergency department (ED) visits, acute kidney injury, pneumonia, urinary tract infection, sepsis, and deep venous thrombosis. Secondary outcomes included rates of revision arthroplasty following the initial procedure.

Data Analysis: Comparisons of postoperative complications and revision rates between the CD and control groups were conducted using chi-square tests, with statistical significance defined at an α level of 0.05. Univariate analyses were performed to evaluate associations between the primary and secondary outcomes and potential confounding variables, including surgical year, geographic region, and medical comorbidities. Variables that demonstrated statistical significance (p < 0.05) in univariate analysis were subsequently incorporated into a multivariate logistic regression model to adjust for confounding factors. Results from the multivariate analysis were reported as odds ratios (OR) with 95 % confidence intervals (CI). All statistical analyses were conducted using R statistical software (R Foundation for Statistical Computing, Vienna, Austria).

3 Results

Overall, 11,562 patients who underwent knee arthroplasties from 2010 to 2020 were included in the study. Of them, 5,781 (50 %) had a preoperative diagnosis of celiac disease. The matched cohorts of disease and control patients contained 5,781 individuals each. Baseline demographics of age (>50 years) and gender were similar between cohorts. Preoperative comorbidities differed significantly with respect to obesity, Elixhauser Comorbidity Indices (ECI), asthma, COPD, chronic kidney disease (CKD), heart failure, coronary artery disease (CAD), ischemic heart disease (IHD), osteoarthritis, pulmonary heart disease (PHD), and rheumatoid arthritis (RA) (Table 1).

Patients in the celiac disease group had significantly higher rates of 90-day complications including infection (OR 1.9, 95 % CI 1.5–2.3, p < 0.0001), ED visits (OR 2.8, 95 % CI 2.5–3.1, p < 0.001), acute kidney disease (AKD) (OR 2.2, 95 % CI 1.7–2.9, p < 0.0001), pneumonia (OR 3.1, 95 % CI 2.3–4.1, p < 0.001), urinary tract infection (UTI) (OR 3.9, 95 % CI 3.2–4.6, p < 0.001), sepsis (OR 2.6, 95 % CI 1.7–4.0, p < 0.0001), deep venous thrombosis (DVT) (OR 2.8, 95 % CI 1.7–4.5, p < 0.0001). In a multivariate analysis controlling for age, gender, and Charlson comorbidity index, CD was a significant predictor of all-cause complications within 90 days from arthroplasty (OR 2.2, 95 % CI 2.0–2.4, p < 0.0001) (Table 2). Patients in the celiac disease group (3.8 %) had lower rates of subsequent revision arthroplasty procedures following knee arthroplasty compared to control patients (4.2 %), although this was not significant. Of the 459 patients that received revision arthroplasty, 219 were in the disease cohort, and 240 were in the control cohort (OR 0.9, 95 % CI 0.8–1.1, p < 0.341) (Table 3).

4 Discussion

Our analysis specific to the association between celiac disease and postoperative knee arthroplasty complications is one of the first of its kind. There is limited available literature that investigates total joint replacements and compares postoperative complications in CD patients to healthy controls. In one study conducted by Cole et al., within 90 days following primary TKA, CD patients exhibited higher yet statistically comparable rates for all medical complications queried (DVT, pulmonary embolism, AKI, myocardial infarction, transfusion, inpatient readmission) compared to controls.¹⁶ Likewise, we also found that individuals with celiac disease experienced increased 90-day postoperative complications, including infection, ED visits, AKD, pneumonia, UTI, sepsis, and DVT (Table 1). In contrast to previous studies, our findings were statistically significant and demonstrated that individuals with celiac disease experienced increased 90-day postoperative complications, including infection, emergency department (ED) visits, AKI, pneumonia, urinary tract infection (UTI), sepsis, and DVT (Table 1). However, the rates of revision arthroplasty were not significantly different between the two groups. These results underscore the systemic challenges associated with CD and its influence on perioperative outcomes.^4,16,29

Celiac disease is independently associated with reduced BMD in children and adults 18 years or older.^29,30 Low BMD induces bone demineralization, making bones susceptible to osteoporosis (OP).³¹ Osteoporosis is a skeletal disease that promotes bone fragility and enhances the risk for bone fractures.^32,33 Characterized by bone mass depletion and bone structure deterioration,³⁴ OP development increases the risk of postoperative complications, including embolism, stroke, UTI, pneumonia, and rates of revision surgeries.^35,36 These findings may explain the increased 90-day postoperative complications observed in CD patients compared to healthy controls. Furthermore, in a study evaluating the long-term outcomes of osteoporosis in patients undergoing TKA (11,288 patients), Chee et al. found OP to be independently associated with any overall surgical and medical complications in TKA patients at two years postoperatively (OR ≥ 1.42, p < 0.001).³⁷

We also found that CD patients had significantly higher rates of UTI and AKI within the postoperative three-month index compared to healthy controls. This aligns with studies demonstrating the increased risk of concomitant occurrence of UTI in individuals with CD.^38,39 Also, patients in the disease cohort had significantly higher rates of acute kidney disease postoperatively, aligning with the findings of Nurmi et al., which identified CD individuals as having an increased risk for kidney disease.⁴⁰ In a retrospective analysis of over 15,000 patients hospitalized with celiac disease, Jung et al. concluded that CD was associated with both pneumococcus speciated infections (OR 2.16, 95 % CI 1.38–3.38) and infections commonly caused by S. pneumoniae (OR, 1.78, 95 % CI 1.61–1.96), including pneumonia (OR, 1.70, 95 % CI 1.53–1.89).⁴¹ Furthermore, their study revealed that patients with CD had a significantly longer length of stay (p < 0.001) and a higher cost of hospitalization (p < 0.001) with pneumococcus-associated infections.⁴¹The observed increase in complications among CD patients may be attributed to underlying factors such as reduced bone mineral density, chronic inflammation, and malnutrition, all of which are known to influence healing and recovery.

The observed increase in complications among CD patients may be attributed to underlying factors such as reduced bone mineral density, chronic inflammation, and malnutrition, which are known to influence healing and recovery. These systemic effects of CD may also account for the elevated risks of infection and thromboembolic events observed in our study. Future research should explore targeted perioperative management strategies to mitigate these risks in CD patients undergoing TKA.

Interpreting these findings through the lens of clinically significant outcomes (CSO) is essential. Minimal clinically important differences (MCID) represent the smallest measurable change that patients perceive as beneficial, while substantial clinical benefit (SCB) indicates more profound improvements that meaningfully enhance quality of life. Similarly, the patient acceptable symptom state (PASS) reflects the threshold of symptoms that patients consider tolerable. While the statistically significant differences observed in this study are noteworthy, it remains unclear whether these findings meet established thresholds for MCID, SCB, or PASS in the TKA population with CD. The large sample size in this study increases statistical power, potentially detecting differences that may not translate into meaningful clinical impacts. This distinction underscores the need for future studies to establish CD-specific MCID and SCB thresholds for arthroplasty outcomes. The elevated complication rates may represent clinically relevant risks, but further investigation is needed to establish their true significance in terms of patient-reported outcomes and functional recovery.^44–46Incorporating PASS as an endpoint in future research can provide a more patient-centered understanding of how these complications affect satisfaction and recovery trajectories^5,11,28.

5 Limitations

Our study faces limitations relating to its methodology. First, by retrospectively assessing patients within a 90-day postoperative period, this analysis is restricted to only short-term outcomes. Evaluating patients over a longer timeframe should be conducted to supplement our retrospective study. Furthermore, prospectively assessing individuals who receive knee arthroplasties and have celiac disease may offer a better association with postoperative complications.

In addition, our study populated its disease and control groups from the insurance claims database, PearlDiver. Due to the deidentified nature of its patient data, PearlDiver may not be an ideal resource for investigating long-term outcome studies.⁴² Additionally, PearlDiver may not account for subgrouping specific cohorts and conditions, which can limit in-depth analyses.⁴³ With our study, we could not further categorize our disease group patients based on the severity of their condition. We encourage future studies to determine whether differences in CD severity significantly impact postoperative complications in total joint arthroplasties. Lastly, PearlDiver's patient populations are mainly over 65 years of age, limiting the ability to generalize to national demographics.⁴⁴ Regardless, the accessibility to an extensive database afforded by PearlDiver that evaluates specific elective procedures with common comorbidities makes it a valuable resource for clinical research.

6 Conclusion

Patients with celiac disease, defined as a malabsorptive autoimmune disorder, who undergo knee arthroplasty are more likely to experience postoperative complications within 90 days compared to healthy control patients. Specific complications include infection (OR 1.9), ED visits (OR 2.8), AKD (OR 2.2), pneumonia (OR 3.1), UTI (OR 3.9), sepsis (OR 2.6), and DVT (OR 2.8). Additionally, CD patients may be more likely than healthy patients to require a revision arthroplasty procedure (OR 0.9). These findings can help orthopaedic surgeons make more informed decisions when evaluating celiac disease patients who may be knee arthroplasty candidates.

CRediT authorship contribution statement

Camden Bohn: Writing – review & editing. Catherine Hand: Writing – review & editing. Chase Gornbein: Conceptualization, Methodology. Daanish Khazi-Syed: Software, Validation. Josh Chang: Software, Validation, Methodology. Brian Forsythe: Supervision, Project administration.

Guardian/Patients consent

The study utilized the PearlDiver Database, a national insurance claims database run through Humana Health Insurance (PearlDiver Inc, Fort Wayne, IN, USA). All data available in the database is anonymized and de-identified and was queried only using ICD-9, ICD-10, and Current Procedural Terminology (CPT) codes. Thus, no human subjects were directly involved nor any protected health information collected and no guardian/patient consent was necessary.

Ethical Statement for Solid State Ionics

Hereby, I Brian Forsythe consciously assure that for the manuscript “Outcomes and Complications of Total Knee Arthroplasty in Patients with Celiac Disease: A Large Insurance Claims Matched Cohort Analysis,” The following is fulfilled:1)This material is the authors' own original work, which has not been previously published elsewhere.2)The paper is not currently being considered for publication elsewhere.3)The paper reflects the authors' own research and analysis in a truthful and complete manner.4)The paper properly credits the meaningful contributions of co-authors and co-researchers.5)The results are appropriately placed in the context of prior and existing research.6)All sources used are properly disclosed (correct citation). Literally copying of text must be indicated as such by using quotation marks and giving proper reference.7)All authors have been personally and actively involved in substantial work leading to the paper, and will take public responsibility for its content.

The violation of the Ethical Statement rules may result in severe consequences.

To verify originality, your article may be checked by the originality detection software iThenticate. See also http://www.elsevier.com/editors/plagdetect.

I agree with the above statements and declare that this submission follows the policies of Solid State Ionics as outlined in the Guide for Authors and in the Ethical Statement.

Funding source declaration

This research was supported by internal department funds. No external funding was received for this project.