Zhang Sheng, Su Zhongzhou, Wen Xianqiang
Department of Neurosurgery, Huzhou Central Hospital, Huzhou, China.
Front Endocrinol (Lausanne). 2025 Jan 8;15:1509501. doi: 10.3389/fendo.2024.1509501. eCollection 2024.
Abnormal thyroid hormone levels may occur in critical illness, which may have an interactive relationship with inflammatory reaction. At present, the relationship between triiodothyronine (T3)/thyroxine (T4) ratio and inflammatory indicators and all-cause mortality of stroke survivors is still unclear.
We obtained the relevant data of the respondents from 2007 to 2012 through the National Health and Nutrition Examination Survey (NHANES) database for statistical analysis. The ratio of T3/T4, a continuous variable, is transformed into three groups of classified variables, namely Q1, Q2 and Q3. The relationship between T3/T4 ratio and mortality was analyzed by Log-Rank test and K-M survival curve. Pearson correlation analysis was used to analyze the correlation between T3/T4 ratio and white blood cell (WBC), Neutrophil-to-Lymphocyte Ratio (NLR), systemic immune-inflammation index (SII) and neutrophil percentage-to-albumin ratio (NPAR). Cox univariate and multivariate analysis was used to identify independent risk factors for all-cause mortality in stroke survivors and a nomogram was drawn. Restricted cubic spline (RCS) curve was drawn to determine whether there was a linear relationship between T3/T4 ratio and mortality and the best cut-off value. Subgroup analysis showed the difference between the T3/T4 ratio and all-cause mortality among subgroups and a forest plot was drawn. The mediation effect analysis was used to analyze whether the ratio of T3/T4 could mediate the survival time through inflammatory indicators.
According to the inclusion and exclusion criteria, a total of 267 people were included in the study, with a mortality rate of 49.06% (131/267), an average survival time of 111.22 ± 3.19 months, and a median survival time of 130 ± 11.27 months. The Log-Rank test and K-M survival curve showed that there were statistical differences among the Q1, Q2, and Q3 groups of the T3/T4 ratio ( = 16.32, <0.001), and the lower the T3/T4 level, the shorter the survival time. Pearson correlation analysis showed that the T3/T4 ratio had a linear relationship with NLR, SII, and NPAR, and only had a weak correlation with NPAR ( = -0.31, <0.001). Cox univariate analysis showed that age, marital status, race, cancer, T3/T4 ratio, NPAR and all-cause mortality were related. Multivariate regression analysis showed that age ≥ 60 years, race of non-Hispanic black, low T3/T4 ratio ( = 0.014, = 0.92, 95% = 0.870.98) and high NPAR ( = 0.009, = 2.50, 95% = 1.264.99) were independent risk factors for all-cause mortality. The RCS curve shows that the ratio of T3/T4 is linearly correlated with mortality, and the optimal cutoff value of T3/T4 is 12.97. Subgroup analysis showed that T3/T4 ratio is more likely to affect the survival of stroke survivors with BMI 18.5~28. Mediation effect analysis showed that there was a mediation effect between T3/T4 ratio, NPAR and survival time. The effect size of T3/T4 directly affecting survival time is 78.45%, and the effect size of T3/T4 indirectly affecting survival time through NPAR is 21.55%.
T3/T4 ratio is an independent risk factor for all-cause mortality in stroke survivors, especially in the people with BMI 18.5~28. T3/T4 ratio may mediate the survival time through NPAR level. Therefore, monitoring thyroid function is beneficial to the management of stroke survivors.
危重症患者可能出现甲状腺激素水平异常，且可能与炎症反应存在交互关系。目前，三碘甲状腺原氨酸（T3）/甲状腺素（T4）比值与炎症指标及卒中幸存者全因死亡率之间的关系仍不明确。
通过美国国家健康与营养检查调查（NHANES）数据库获取2007年至2012年受访者的相关数据进行统计分析。将连续变量T3/T4比值转化为Q1、Q2和Q3三组分类变量。采用Log-Rank检验和K-M生存曲线分析T3/T4比值与死亡率的关系。采用Pearson相关分析分析T3/T4比值与白细胞（WBC）、中性粒细胞与淋巴细胞比值（NLR）、全身免疫炎症指数（SII）及中性粒细胞百分比与白蛋白比值（NPAR）之间的相关性。采用Cox单因素和多因素分析确定卒中幸存者全因死亡率的独立危险因素，并绘制列线图。绘制限制立方样条（RCS）曲线以确定T3/T4比值与死亡率之间是否存在线性关系及最佳截断值。亚组分析显示亚组间T3/T4比值与全因死亡率的差异，并绘制森林图。采用中介效应分析分析T3/T4比值是否可通过炎症指标介导生存时间。
根据纳入和排除标准，共纳入267人进行研究，死亡率为49.06%（131/267），平均生存时间为111.22±3.19个月，中位生存时间为130±11.27个月。Log-Rank检验和K-M生存曲线显示，T3/T4比值的Q1、Q2和Q3组间存在统计学差异（ = 16.32，<0.001），T3/T4水平越低，生存时间越短。Pearson相关分析显示，T3/T4比值与NLR、SII和NPAR呈线性关系，与NPAR仅呈弱相关（ = -0.31，<0.001）。Cox单因素分析显示，年龄、婚姻状况、种族、癌症、T3/T4比值、NPAR与全因死亡率相关。多因素回归分析显示，年龄≥60岁、非西班牙裔黑人种族、低T3/T4比值（ = 0.014， = 0.92，95% = 0.870.98）和高NPAR（ = 0.009， = 2.50，95% = 1.264.99）是全因死亡率的独立危险因素。RCS曲线显示，T3/T4比值与死亡率呈线性相关，T3/T4的最佳截断值为12.97。亚组分析显示，T3/T4比值更易影响BMI为18.5~28的卒中幸存者的生存。中介效应分析显示，T3/T4比值、NPAR与生存时间之间存在中介效应。T3/T4直接影响生存时间的效应大小为78.45%，T3/T4通过NPAR间接影响生存时间的效应大小为21.55%。
T3/T4比值是卒中幸存者全因死亡率的独立危险因素，尤其是BMI为18.5~28的人群。T3/T4比值可能通过NPAR水平介导生存时间。因此，监测甲状腺功能有利于卒中幸存者的管理。
