Karaphillis Erica, Goldstein Ryan, Murphy Sharif, Qayyum Rehan
aDepartment of Internal Medicine, University of Tennessee College of Medicine at Chattanooga, Chattanooga, Tennessee bJohns Hopkins School of Medicine, Division of General Internal Medicine, Baltimore, Maryland, USA.
Eur J Gastroenterol Hepatol. 2017 Mar;29(3):284-288. doi: 10.1097/MEG.0000000000000778.
Studies have examined the relationship between serum alanine aminotransferase (ALT) and mortality with inconsistent results. Our aims were to examine the association of normal range serum ALT with mortality, to explore a nonlinear relationship between ALT and mortality, and to investigate whether age modifies this relationship.
We used the continuous National Health and Nutrition Examination Survey (NHANES) from 1999 to 2010. Vital status of the participants was obtained by probabilistic matching between NHANES and the National Death Index. Cox proportional models were used to examine the relationship with and without adjustment for age, sex, race, BMI, hypertension, diabetes, alcohol use, serum triglycerides, prescription drug use, and glomerular filtration rate, and accounting for the sampling methodology of NHANES. Nonlinear relationship was examined using spline (single knot at 17 U/l) regression. Interaction terms were used to examine effect modification by age.
Higher serum ALT was associated with lower all-cause mortality [adjusted hazard ratio (HR)/ALT increment=0.98, 95% confidence interval (CI)=0.97-0.99]; however, this relationship was nonlinear and present only until 17 U/l (adjusted HR/ALT increment=0.93, 95% CI=0.91-0.95) and not thereafter. Age modified the relationship between ALT and mortality; elderly patients (>64 years) had a 6% lower adjusted mortality risk than younger (<35 years) participants (HR/ALT increment=0.94, 95% CI=0.91-0.96; interaction P<0.001).
Increase in serum ALT within the normal range is initially associated with lower mortality, but has no effect after 17 U/l. The elderly show a significantly larger decrease in mortality with an increase in ALT than younger individuals. The mechanisms underlying this relationship need further exploration.
多项研究探讨了血清丙氨酸氨基转移酶（ALT）与死亡率之间的关系，但结果并不一致。我们的目的是研究正常范围血清ALT与死亡率之间的关联，探索ALT与死亡率之间的非线性关系，并研究年龄是否会改变这种关系。
我们使用了1999年至2010年的连续全国健康与营养检查调查（NHANES）。通过NHANES与国家死亡指数之间的概率匹配获得参与者的生命状态。使用Cox比例模型来研究在调整和未调整年龄、性别、种族、体重指数、高血压、糖尿病、饮酒、血清甘油三酯、处方药使用和肾小球滤过率的情况下的关系，并考虑NHANES的抽样方法。使用样条（在17 U/l处有一个节点）回归来检验非线性关系。使用交互项来检验年龄的效应修正。
较高的血清ALT与较低的全因死亡率相关[调整后的风险比（HR）/ALT增量=0.98，95%置信区间（CI）=0.97-0.99]；然而，这种关系是非线性的，仅在17 U/l之前存在（调整后的HR/ALT增量=0.93，95%CI=0.91-0.95），之后则不存在。年龄改变了ALT与死亡率之间的关系；老年患者（>64岁）的调整后死亡风险比年轻患者（<35岁）低6%（HR/ALT增量=0.94，95%CI=0.91-0.96；交互作用P<0.001）。
正常范围内血清ALT的升高最初与较低的死亡率相关，但在17 U/l之后则无影响。老年人ALT升高时死亡率的下降幅度明显大于年轻人。这种关系背后的机制需要进一步探索。
