Acute kidney injury (AKI) is defined as an abrupt decrease in kidney function, typically diagnosed by an increase in serum creatinine or a decrease in urine output, according to the Kidney Disease: Improving Global Outcomes (KDIGO) criteria. AKI is a common complication in critically ill patients, affecting up to 25% of children in the intensive care unit (ICU) [1,2]. Importantly, AKI is a leading cause of in-hospital morbidity and mortality. A recent meta-analysis of 133,876 hospitalized children with AKI reported a pooled in-hospital mortality rate of 18.27%, with mortality increasing progressively from 8.19% in stage 1 to 27.78% in stage 3 [3]. Beyond its impact on patient outcomes, AKI imposes a substantial economic burden on healthcare systems [4,5]. These findings underscore the urgent need for early identification and risk stratification of high-risk pediatric AKI patients to improve clinical outcomes and reduce healthcare expenditures. Although various critical illness scoring tools and single biomarkers are currently used to assess the prognosis of children with AKI, they have notable limitations in clinical practice. First, most scoring systems are based on data collected within the first 24 h of admission and cannot dynamically reflect real time changes in a patient's condition. Second, traditional indicators such as creatinine and urine output respond slowly to renal injury, making early warning difficult. Therefore, there is an urgent need for novel prognostic markers that can capture dynamic disease progression and enable continuous bedside monitoring, thereby facilitating early identification and timely intervention in high-risk pediatric AKI patients.

Lactate is a metabolite derived from pyruvate under hypoxic or anaerobic conditions and serves as a sensitive marker of tissue hypoperfusion and cellular hypoxia. It has been widely used for risk assessment and prognosis prediction in critically ill patients [6]. In populations with sepsis, pneumonia, trauma, or post-cardiac surgery, both admission lactate levels and lactate clearance have been shown to be closely associated with mortality [7–11]. In recent years, the prognostic value of lactate in patients with AKI has gained increasing attention. Multiple studies have demonstrated that elevated admission lactate is significantly associated with increased in-hospital mortality in AKI patients [12–15]. However, most existing studies have relied on lactate measurements at a single time point. In fact, the clinical course of critically ill patients is continuously evolving, and therapeutic interventions constantly influence lactate production and clearance. Consequently, a single lactate value may not adequately capture disease progression or response to treatment, potentially leading to underestimation or overestimation of mortality risk. Trajectory analysis has been applied in adult populations with AKI and other critical conditions, revealing distinct lactate patterns that are strongly associated with prognosis. However, evidence regarding the association between dynamic lactate trajectories and short-term outcomes in children with AKI remains lacking. Therefore, this study aims to identify distinct lactate trajectories based on serial measurements over the first four days of PICU admission using latent growth mixture model and to examine their association with 30-day mortality in pediatric AKI patients. The findings may provide a novel dynamic marker for early risk stratification and targeted intervention in this vulnerable population.

The present study investigated for the first time the association between both admission lactate levels and early lactate trajectories with 30-day all-cause mortality in pediatric patients with AKI. The main findings are as follows: First, admission lactate levels were monotonically associated with an increased risk of 30-day mortality. Second, using LGMM based on daily serum lactate measurements collected during the first four days of PICU admission, three distinct serum lactate trajectories were identified: Class 1 (high decreasing group), Class 2 (low stable group), and Class 3 (persistent high group). Notably, these trajectories demonstrated significant prognostic differences, with Class 3 showing the poorest clinical outcomes.

Lactate is a commonly used biomarker reflecting tissue perfusion, oxidative metabolism, and cellular dysfunction. Its prognostic value is particularly notable in patients with AKI, as AKI impairs lactate clearance, while systemic inflammation and microcirculatory dysfunction further exacerbate tissue hypoxia [17–19]. Numerous studies have demonstrated that elevated blood lactate levels are significantly associated with short-term mortality in AKI patients [12,13]. However, most existing studies have relied on lactate measurements at a single time point, which limits their ability to capture the dynamic evolution of lactate levels in response to disease progression and therapeutic interventions. In critically ill patients, the clinical course is continuously evolving, and treatments such as fluid resuscitation, vasoactive agents, and renal replacement therapy can all influence lactate levels. In recent years, dynamic lactate trajectory analysis has emerged as a valuable approach in critical care research. To date, relevant studies have been limited to adult populations. Children differ substantially from adults in terms of metabolic rate, hepatic and renal functional reserve, etiology, and treatment response, and findings from adult studies cannot be directly extrapolated to pediatric populations. Moreover, no study has systematically investigated the association between dynamic lactate trajectories and short-term outcomes in children with AKI.

The findings of this study confirm previous observations that elevated admission lactate levels are closely associated with adverse clinical outcomes. It should be emphasized that elevated lactate itself is not a direct cause of death, but rather an important biomarker of critical illness. The association between hyperlactatemia and increased mortality involves multiple pathophysiological mechanisms. First, tissue hypoperfusion and hypoxia represent the core mechanisms driving lactate elevation. Under hypoxic conditions, cells shift from aerobic ATP production to anaerobic glycolysis, leading to the conversion of pyruvate to lactate [20]. This disruption of cellular energy metabolism impairs cell function, affecting myocardial contractility and neuronal activity, thereby increasing the risk of multiple organ dysfunction [21,22]. Second, mitochondrial dysfunction plays a critical role in lactate accumulation. Impaired entry of pyruvate into the mitochondria for aerobic oxidation results in lactate buildup [23]. Moreover, mitochondrial dysfunction can induce apoptosis and pyroptosis, exacerbating endothelial injury and organ failure [24]. Third, elevated lactate induces metabolic acidosis. Increased hydrogen ion concentration inhibits various enzymes and interferes with cellular metabolism and ion channel function, leading to altered myocardial excitability and reduced vascular responsiveness to vasoactive agents, thereby worsening circulatory failure [25]. In addition, impaired lactate clearance contributes to persistent hyperlactatemia. The liver and kidneys are the primary organs responsible for lactate metabolism. Hepatic or renal dysfunction, as well as impaired cellular lactate uptake associated with sepsis, can reduce clearance capacity. Lactate accumulation further aggravates acidosis and cellular dysfunction, establishing a vicious cycle. Furthermore, hyperlactatemia activates inflammatory signaling pathways, promoting the release of pro-inflammatory cytokines such as IL-6 and TNF-α. This can induce or exacerbate systemic inflammation, leading to immune dysregulation, increased risk of infection, and aggravated organ damage [26,27]. Collectively, these interconnected mechanisms form a vicious cycle that ultimately contributes to a significantly increased risk of death.

Utilizing latent growth mixture model, the present study identified three distinct lactate trajectories. Class 2 maintained lactate levels within the near-normal range throughout PICU admission and exhibited the lowest mortality. In contrast, both Class 1 and Class 3 were associated with significantly higher 30-day mortality than Class 2, with Class 3 showing the poorest survival, indicating that a persistently elevated lactate trajectory represents the subgroup with the worst prognosis among pediatric AKI patients. Notably, although Class 1 and Class 3 both presented with elevated lactate levels, they differed substantially in clinical characteristics, underlying pathophysiology, and outcomes. As shown in Table 1, patients in Class 1 had higher admission lactate levels, more severe acidosis, and worse hepatic, renal, and coagulation parameters, suggesting a more critical condition at baseline with severe tissue hypoperfusion and multi-organ impairment. In Class 1, despite markedly elevated admission lactate, lactate levels declined rapidly during subsequent treatment, indicating a favorable response to early resuscitative measures, with prompt improvement in tissue perfusion and restoration of lactate clearance capacity. Thus, Class 1 represents reversible hyperlactatemia. By contrast, patients in Class 3 did not have the highest admission lactate levels, but their lactate levels continued to rise during the PICU stay without effective decline. This trajectory pattern suggests treatment-resistant hyperlactatemia, in which standard therapeutic interventions are insufficient to reverse the underlying pathophysiology, ultimately leading to the poorest outcomes. From a clinical perspective, these findings have important implications for risk stratification and management. For Class 1 (high decreasing group), despite critical illness at admission, early aggressive intervention may effectively reverse the clinical course. Thus, adequate resuscitative efforts and close monitoring of lactate clearance are warranted to assess treatment response. For Class 3 (persistent high group), conventional therapy may prove inadequate, and escalation of treatment, such as earlier initiation of renal replacement therapy or even extracorporeal membrane oxygenation, should be considered. For Class 2 (low stable group), a less intensive monitoring strategy may be appropriate to avoid unnecessary interventions.

The present study has several strengths. First, it represents the first investigation to identify dynamic lactate trajectories in pediatric patients with AKI using latent growth mixture model based on daily lactate measurements over the first four days of PICU admission, thereby addressing a significant gap in the literature for this population. Second, in contrast to traditional single-point lactate measurements, dynamic trajectory analysis distinguishes reversible hyperlactatemia from treatment-resistant hyperlactatemia, providing insights into treatment responsiveness and offering more refined information for clinical risk stratification. Nevertheless, several limitations should be acknowledged. First, the data were derived from a single center, and the generalizability of the findings requires validation through multicenter prospective studies. Second, Class 3 (persistent high group) accounted for only 2.84% of the total cohort, and the relatively small sample size limits statistical power. Third, detailed information regarding the timing and dosage of therapeutic interventions, such as fluid resuscitation and renal replacement therapy, was not recorded, precluding analysis of their direct impact on lactate trajectories. Fourth, patients who died within the first four days of PICU admission were not included, which may introduce survival bias and affect the completeness of trajectory classification. Fifth, urine output data in this retrospective database had inherent limitations, including potential inaccuracies in infants using diapers and interference from diuretic therapy, which may have introduced bias in AKI staging.

The present study demonstrates that admission lactate levels in pediatric patients with AKI are significantly and positively associated with 30-day mortality. Compared with traditional single-point lactate measurement, lactate trajectory analysis not only identifies patient subgroups with distinct prognoses but also provides insights into treatment responsiveness.

Conceptualization, methodology, data curation and writing-original draft were performed by Xiaodong Jiang. Formal analysis and writing-original draft were performed by Yu Zhang. Methodology and writing-review and editing was performed by Hua Lu.

This study involves human participants and this project has been approved by the Institutional Review Committee of the Children’s Hospital of Zhejiang University School of Medicine (ethics reference number: 2019-IRB-052), and all procedures were conducted in accordance with the ethical standards of the Declaration of Helsinki and relevant regulations. Given the retrospective nature of the study and the use of anonymized data, the requirement for informed consent was waived.

No generative AI tools were used in the preparation of this manuscript.

The author(s) declare that no financial support was received for the research, authorship, and/or publication of this article.

The datasets presented in this study can be found in online repository: https://www.physionet.org/content/picdb/1.1.0/.