AUTHOR=Chen Mengyuan , Wang Shu-an , Yang Jiayao , Bai Jiawang , Gu Jingyue , Luo Haolong , Zhang Xudong , Han Yan , Shao Jihong , Xu Yan , Guo Shuyan , Ren Xiangmei TITLE=Association of systemic immune-inflammation index with malnutrition among Chinese hospitalized patients: a nationwide, multicenter, cross-sectional study JOURNAL=Frontiers in Nutrition VOLUME=11 YEAR=2024 URL=https://www.frontiersin.org/journals/nutrition/articles/10.3389/fnut.2024.1375053 DOI=10.3389/fnut.2024.1375053 ISSN=2296-861X ABSTRACT=Background

Systemic immune-inflammation index (SII) is associated with increased risk in a wide range of illnesses. However, few studies have explored the associations between SII and the risk of malnutrition. Therefore, this study aimed to investigate the association between SII and malnutrition in a nationwide, multicenter, cross-sectional study involving Chinese hospitalized patients.

Design

From August 2020 to August 2021, a total of 40,379 hospitalized patients met the inclusion and exclusion criteria. Detailed demographic data, diagnoses, as well as physical and laboratory examination results were recorded. The diagnosis of malnutrition was used with two distinct methods: the Malnutrition Screening Tool 2002 (NRS 2002) + Global Leaders Initiative on Malnutrition (GLIM) criteria and the controlling nutritional status (CONUT) score. The risk factors for malnutrition were analyzed using binary logistic regression and multiple logistic regression to obtain odds ratios (OR) and 95% confidence intervals (CI). Restricted cubic spline (RCS), linear spline, and receiver operating characteristic (ROC) analysis were also used.

Results

The prevalence of malnutrition diagnosed by the two methods was 13.4% and 14.9%, respectively. In the NRS 2002 + GLIM diagnostic model, lnSII showed statistical significance between the malnutrition and non-malnutrition group (6.28 ± 0.78 vs. 6.63 ± 0.97, p < 0.001). A positive association was observed between higher SII and the risk of malnutrition in both before and after adjustment models compared to the first quartile (Q3 vs. Q1, OR = 1.27, 95%CI: 1.15–1.40; Q4 vs. Q1, OR = 1.83, 95%CI: 1.67–2.00). However, a significant reduction in prevalence was observed when SII was in the second quartile (Q2 vs. Q1, OR < 1), as indicated by a restricted cubic spline with a U trend (p for nonlinear <0.001). According to the CONUT score, the prevalence of individuals with normal nutritional status decreased with increasing SII, while the occurrence of three different degrees of malnutrition generally increased. The Kappa value between the two diagnostic methods was 0.23, and the merged data observed an area under the ROC curve of 0.73 (95%CI: 0.714–0.742).

Conclusion

The U-shaped association between SII and the prevalence of malnutrition was observed. Both lower and higher SII levels (either continuous or categorical variable) were significantly associated with an increased risk of malnutrition.