Postoperative mortality is the third largest contributor to global death, following ischemic heart disease and stroke (1, 2). Complications such as major adverse cardiac events, stroke, and acute kidney injury are considered as the main reasons for perioperative death (3–5). Among these complications, acute kidney injury (AKI) presents a higher incidence and has a significantly detrimental effect on clinical outcomes.

The overall incidence of AKI varies from 3 to 13% in patients who underwent non-cardiac surgeries and increases up to 20% in older adults (5–7). The underlying etiologies and mechanisms of AKI are multifactorial. Common risk factors include poor renal function (i.e., chronic renal disease), surgery-related stress response, nephrotoxic drugs (i.e., anesthetics and antibiotics), and hemodynamic variability (i.e., hypotension and hypovolemia) (6, 8, 9).

Acute kidney injury is associated with poor clinical outcomes in patients. Several studies indicated that patients with AKI suffer three times the risk of cardiovascular complications and prolonged in-hospital stay (9–11). In a population-based study, patients with AKI experienced an 8-fold increased risk of 30-day mortality (12). Patients with AKI, after vascular surgery, spent an additional amount of US$ 9,000–19,000 on confounding comorbidities and other postoperative complications (13).

Unlike the supportive data on AKI and postoperative short-term (i.e., within 30 days) mortality, it is still uncertain whether AKI is associated with long-term mortality in patients after non-cardiac surgery. After an abdominal surgery, the risk of 1-year mortality increased by 2-fold in patients with AKI (14). This result is consistent with a meta-analysis of 82 cohort studies which showed that patients with AKI were at an increased risk of long-term mortality (13.19 vs. 7.26 deaths per 100 person-years) (15). However, a recent observational study did not find the impact of AKI on 1-year mortality in patients after colorectal cancer surgery (16). A prospective study of critically ill patients also found that AKI was not an independent risk factor of 3-year mortality (17). Evidence from previous studies was mainly based on retrospective designs with inherent flaws, such as missing data and bias (14–16). Strong evidence mapping the relationship between AKI and 3-year mortality in elderly patients after non-cardiac surgery is still lacking (14–18). We hypothesized that patients with AKI may have a higher 3-year mortality compared to patients without AKI after non-cardiac surgery.

The present study was designed to investigate whether AKI was associated with 3-year mortality in elderly patients after non-cardiac surgery.

This analysis was based on a 3-year follow-up dataset of two randomized controlled trials (RCTs) (19, 20). The study protocol was approved by the Clinical Research Ethics Committee of Peking University First Hospital (2020-115). Informed consents were waived because of no intervention in patients.

The present study found that AKI was not associated with 3-year mortality in elderly patients after non-cardiac surgery. One strength of the present study is the relatively large sample size of older adults. Another strength is that the 3-year follow-up was planned at the beginning of the two trials and all data were prospectively collected.

In the present study, the incidence of AKI was ~15.5%, and 85% of them were at stage 1. This result was consistent with previous studies (5–7). Similar to most other studies, we only employed creatinine to diagnose AKI, and this might underestimate the incidence and the severity of AKI (5, 6).

Many studies have reported the positive relationship between AKI and short-term mortality in patients who underwent non-cardiac surgery, but the association between AKI and long-term mortality is still uncertain (10, 23). A retrospective study analyzed 1,869 patients who underwent non-cardiac surgery and found that AKI was associated with an increased risk of mortality between 8 and 365 days (5). In 6,590 patients who underwent colorectal cancer surgery, AKI was associated with an increased risk of mortality between 8 and 30 days and between 31 and 90 days, but it was not related to mortality between 91 and 360 days (16).

The present study was designed to investigate whether AKI was associated with 3-year mortality in elderly patients after non-cardiac surgery, and our results showed that it was not. This result might be attributed to several underlying factors. First, the impaired kidney function of patients with AKI may recover or improve with time, and this potentially reduces its impact on long-term mortality. A long-term follow-up of 1,893 patients after partial nephrectomy showed that the incidence of AKI within a 7-day postoperative period was about 20% (388/1,893), and the renal function of most of these patients recovered to 90% of baseline function within a year after surgery (24). A propensity score matching study showed that the recovery of the renal function was associated with a decreased risk of 1-year mortality in surgical patients (25). Second, about 80% of these patients received surgery for malignant tumor, and 73.8% of mortalities were related to cancer. The recurrence of tumor and metastasis might be the major reason for long-term mortality and surpasses the potential effect of AKI. Third, the incidence and severity of AKI might be underestimated because we only considered serum creatinine levels to diagnose AKI without urine output.

The stage and the onset time of AKI may also exert different effects on long-term mortality. The relationship between the AKI stage and long-term mortality had been reported in many studies (5, 16, 26). However, we did not find the statistical significance between the AKI stage and 3-year mortality. This might be attributed to the lower occurrence of severe AKI (only 31 patients experienced stage 2 and stage 3 AKI) in the present study. The sample size in our study was not enough to explore the relationship between AKI stage and 3-year mortality. A retrospective study showed that both early and late onset of AKI were related to an increased risk of 1-year mortality in critically ill patients after a major non-cardiac surgery (22). We noticed that late onset AKI, but not early onset AKI, was related to an increased risk of 3-year mortality in the present study. However, only 10 patients were diagnosed with late onset AKI in our study, and this small sample size yielded insufficient statistical power. Thus, this result should be considered as an exploratory analysis and interpreted with caution.

The etiologies of overall mortality are multifactorial. Aging was the robust predictor of mortality in many studies (27). In the validation study of a national risk prediction model for perioperative mortality in non-cardiac surgery, male patients were associated with an increased risk of mortality and were considered as one of the key factors to improve the predictive performance of the risk model (27). Lower BMI was an important manifestation of malnutrition and was highly associated with an increased risk of complications and mortality (28). Preoperative renal function impairment and the ASA physical status classification were reflections of disease severity and, therefore, were validated as confounders of mortality in many studies (29, 30). Major surgery was defined as a surgery duration of more than 2 h, and the prolonged duration of surgery was considered as a risk factor of mortality in patients after the multivariate analysis (14). The role of allogenic blood transfusion in long-term mortality received more attention, but the conclusions remain controversial (31). Although several studies reported that perioperative application of dexmedetomidine might decrease the risk of long-term mortality in surgical patients, the results were inconclusive (32, 33).

The present study has several limitations. First, we only adopted serum creatinine levels for the diagnosis of AKI, and this might underestimate its incidence. Second, this was a prospective follow-up of two RCTs, which excluded some high-risk patients with comorbidities (i.e., severe hepatic or renal diseases). This might limit the generality of our results. Third, we did not get the serum creatinine levels during the 3-year postoperative period or at follow-up, which impeded the assessment of renal function at the 3-year postoperative period. Fourth, only 11 patients were diagnosed as AKI stage 3 in the present study. Our result showed that the incidence of death increased from 27.2 to 45.5% in patients with AKI stages 1 and 3. However, the low incidence of severe AKI might underestimate its underlying association with long-term mortality.

The present study found that AKI was common in patients after non-cardiac surgery; however, it was not associated with 3-year mortality. The incidence of severe AKI (i.e., stage 3) was very low, and this might underestimate the underlying association between AKI and long-term mortality. Further studies with a large sample size are needed to verify our result.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

The studies involving human participants were reviewed and approved by the Clinical Research Ethics Committee of Peking University First Hospital. The Ethics Committee waived the requirement of written informed consent for participation.

Q-FW helped in data acquisition, data analysis, and manuscript drafting. M-WX, W-JH, XS, and D-FZ helped in data acquisition. D-LM helped with conceptual idea and design, data analysis, data interpretation, manuscript drafting, critical revision of the manuscript for the important intellectual content, and supervision. D-XW helped with conceptual idea and design, administrative or material support, and critical revision of the manuscript for the important intellectual content. All authors contributed to the article and approved the submitted version.

This trial was supported by the National Key R&D Program of China (Grant No. 2018YFC2001800).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.