The study retrospectively analysed 1070 patients undergoing total hip or knee arthroplasty between January 2019 and July 2022 at third-level referral institutions. The observational study followed the STROBE guidelines (von Elm et al., 2007). The study adhered to the Declaration of Helsinki principles. No ethical approval was required according to the Local Ethics Committees, due to the retrospective and observational nature of the study.

Inclusion criteria were (i) patients diagnosed with primary osteoarthritis (OA); (ii) undergoing THA or TKA; and (iii) minimum 12 months follow-up.

Exclusion criteria were (i) THA or TKA implanted for secondary OA, femoral neck fracture or oncological lesions; (ii) lacking nutritional assessment; with (iii) incomplete data.

PJI diagnosis was made according to the 2018 Philadelphia International Consensus Meeting (ICM) criteria (Balato et al., 2019). The following data were recorded: demographic data; anthropometric measurements, such as BMI and the Charlson comorbidity index (CCI); clinical history, and surgical history.

According to the Nutritional Risk Screening (NRS) score, patients were classified in two different groups: patients with NRS ≥ 3 were included in the Malnutrition group, and patients with NRS < 3 in the Control group.

Before the surgery, patients' nutritional status were assessed through the Nutritional Risk Screening 2002 (NRS-2002) (Kondrup et al., 2003) and Controlling Nutritional Status (CONUT) scores (Ignacio de Ulíbarri et al., 2005). NRS-2002 is calculated by assessing nutritional status, disease severity, and patient age. Nutritional status is determined by evaluating weight loss over the last 3 months (0 points for less than 5 % loss, 1 point for 5 %–10 % loss, and 2 points for more than 10 % loss), a reduction in food intake compared to usual requirements (0 points for less than 25 % reduction, 1 point for 25 %–50 % reduction, and 2 points for more than 50 % reduction), and BMI (0 points for a BMI ≥ 20.5, 1 point for a BMI between 18.5 and 20.5, and 2 points for a BMI less than 18.5). Disease severity is evaluated by assigning 1 point for mild clinical conditions and 2 points for moderate. A total score of 3 or more indicates a nutritional risk. The CONUT score, instead, assesses nutritional status using serum albumin, total cholesterol, and total lymphocyte count. Points are assigned based on thresholds: serum albumin (> 3.5 g dL⁻¹ = 0, 3.0–3.49 = 2, 2.5–2.99 = 4, < 2.5 = 6), total cholesterol (> 180 mg dL⁻¹ = 0, 140–180 = 1, 100–139 = 2, < 100 = 3), and lymphocyte count (> 1600 mm⁻³ = 0, 1200–1599 = 1, 800–1199 = 2, < 800 = 3). The total score ranges from 0 to 12, with higher scores indicating worse nutritional status. A last follow-up was additionally provided a quality-of-life assessment through EuroQol 5 Dimensions (EQ5-5D) (Giesinger et al., 2015). It consists of five different subscales, assessing pain, mobility, daily activities, self-care, and mental health. A higher value corresponds to a worse outcome.

The data are reported as the mean and standard deviation for continuous variables and frequency distributions in percentages for categorical variables. Continuous variables underwent comparison through either the t student test or the Mann-Whitney U test, depending on the distribution of the data. Categorical variables were represented as proportions and subjected to comparison utilizing either the Fisher exact test or the chi-squared test. Logistic univariate regression, adjusted for age and sex, was performed to assess the role of malnutrition as a risk factor in terms of outcomes and quality of life. The significance was established for a p value < 0.05. Data analysis was donee with the SPSS 29.0 software program (SPSS, Inc., Chicago, IL, USA).

In terms of outcomes, there were no statistically significant differences in overall complication rates between the two groups (p=0.271). However, the Malnutrition group demonstrated a higher rate of infectious complications at 3.3 %, compared to 0.5 % in the Control group (p=0.009) (Fig. 2). Other complications, including the total number of dislocations and re-interventions, as well as mortality rates, did not significantly differ between the groups (p > 0.05) (Table 1).

Quality of life in patients in the Malnutrition group was slightly but significantly worse than the Control group in all the subscales, except the pain subscale (Table 1).

As further analysis, a univariate regression, adjusted for age and sex, was performed, in order to deeper assess the role played by malnutrition in clinical outcomes and quality of life, analysing both NRS and CONUT scores as risk factors. The results are shown in Tables 2 and 3.

These factors can decrease the benefits of the surgery, delaying mobility improvements and pain relief that are expected from the procedure (Ihle et al., 2018; Roche et al., 2018). In addition, the significant findings from this study, recognizing malnutrition as a risk factor for increased infectious complications, can highlight the compromised immune function in malnourished individuals, increasing their susceptibility to infections (De Mauro et al., 2024b; Gu et al., 2019; Pes et al., 2023; Yi et al., 2015). Malnourished patients often have weakened and reduced muscle mass, and impaired wound healing, all of which can lead to longer recovery times, increased complications, and diminished functional outcomes post-surgery (Eminovic et al., 2021; Wilson et al., 2020).

This study demonstrated that malnutrition, assessed using both the CONUT score and NRS-2002, also significantly impacts the quality of life (EQ5) of patients undergoing total joint arthroplasty, affecting all subscales of the quality-of-life measure, such as mobility, self-care, usual activities, pain/discomfort, and depression. This reduced quality of life directly influences the perceived benefits of TJA. This emphasizes the need for early detection and management of malnutrition to improve patient outcomes and reduce complication risks. Integrating comprehensive nutritional assessments into patient care is thus essential for addressing malnutrition and enhancing overall well-being (De Mauro et al., 2024a; Schroer et al., 2019).

The literature shows a wide range of malnutrition prevalence among patients undergoing primary prosthesis surgery, from 3.7 % to about 56.0 %, influenced by differences in study populations and assessment methods (Gu et al., 2019). In contrast, our study, involving 682 patients, found a malnutrition prevalence of 15.6 %. This rate is lower than some higher estimates and shows a more moderate value. Most studies in the literature demonstrate significant correlations between serum marker values and morbidity, complications, hospitalization, and clinical outcomes, addressing these findings as malnutrition (Bohl et al., 2016; De Mauro et al., 2023, 2025; Fu et al., 2017; Roche et al., 2018; Rudasill et al., 2018; Walls et al., 2015). However, these serum markers alone do not provide direct and conclusive evidence of malnutrition. Alterations in markers such as albumin, prealbumin, and total leucocyte count, while commonly associated with adverse clinical outcomes, can be influenced by a range of factors including inflammation, infection, or other non-nutritional conditions (Jin et al., 2023). Eminovic et al. (2021) demonstrated that alterations in serum values, such as albumin levels and total leucocyte count, are observed in both malnourished and well-nourished patients. Our study's restrictive definition of malnutrition minimizes the inclusion of patients in malnutrition spectrum compared with non-specific serum values, reducing false positives (Blevins et al., 2018; Ellsworth and Kamath, 2016; Eminovic et al., 2021). This finding suggests that relying only on serum biomarkers may not provide a fully accurate assessment of nutritional status, as these changes can occur due to various factors unrelated to malnutrition. By combining serum-based assessments like the CONUT score with clinical and anthropometric evaluations from the NRS-2002, a more accurate identification of malnourished individuals was achieved. This rigorous methodology ensures that these findings offer a reliable reflection of malnutrition prevalence and its impact on quality of life and complications.

Evaluating and addressing nutritional deficiencies before surgery can improve the surgical effectiveness, reduce the complication rate, and enhance recovery (Schroer et al., 2019). This approach not only benefits individual patients but also has positive implications for public health and the healthcare system by reducing overall healthcare costs (Bala et al., 2020; Rudasill et al., 2018). Therefore, integrating thorough nutritional assessments into pre-operative care is a valuable strategy for improving patient outcomes and promoting public health.

The strengths of this study include the robustness of the malnutrition assessment criteria, the homogeneity of the patient cohort, and the sample size (which is consistent with other similar studies on this topic in the literature). By using comprehensive and well-defined criteria for evaluating malnutrition, and by studying a relatively large and uniform patient population, our findings provide a reliable and relevant contribution to the field. However, the study also has limitations. First, its retrospective nature means that it relies on existing data, which can limit the depth of analysis and the ability to establish causality. In fact, potentially, several factors such as comorbidity burden, systemic inflammatory status, frailty, and socioeconomic status were not recorded in the current study, and they may act as confounder factors influencing the increased risk of infections in malnourished patients. Second, approximately 36 % of the initially selected cohort was excluded, primarily due to missing nutritional assessments or incomplete clinical data. Although this was necessary to ensure data completeness and reliability, this exclusion may introduce selection bias and limit the generalizability of the results. Third, while malnutrition is often considered a potentially modifiable risk factor, not all forms or causes of malnutrition are equally valuable to correction. The severity, etiology, and time course of malnutrition, as well as the feasibility and effectiveness of targeted nutritional interventions, may all influence post-operative outcomes. In addition, the absence of a longitudinal analysis of the impact of nutritional interventions over time means that we cannot fully assess how changes in nutritional status might affect long-term outcomes. These limitations suggest areas for future research, such as prospective studies with repeated assessments of nutritional status to better understand the dynamic relationship between nutrition and patient outcomes.