Phase-specific mortality risk of serum lactate thresholds in very low birth weight infants with late-onset sepsis: a retrospective cohort study

Background Late-onset sepsis (LOS) in very low birth weight (VLBW) infants confers substantial mortality risk. While lactate monitoring is standard, validated prognostic thresholds for mortality remain unestablished in this population. Methods This retrospective cohort study (2014–2024) analyzed 596 VLBW infants (≤32 weeks; <1,500 g) with LOS at a tertiary NICU. Multivariable regression and piecewise linear modeling identified lactate-mortality thresholds, adjusting for gestational age, respiratory failure severity, vasopressor requirement, and multiorgan dysfunction. Bootstrap validation (1,000 iterations) assessed threshold stability. Results Mortality occurred in 21% (125/596). Nonlinear analysis revealed critical inflection points at 2.2 mmol/L (95%CI: 1.9–2.5) and 4.0 mmol/L (95%CI: 3.7–4.3) ( p < 0.001). Lactate ≤ 2.2 mmol/L demonstrated no mortality association (adjusted odds ratio = 1.84, 95%CI: 0.64–5.34; p = 0.260). Within the 2.2–4.0 mmol/L transition zone, each 1 mmol/L increment conferred a 7.0-fold mortality risk (aOR = 7.0, 95%CI: 2.13–22.78; p < 0.001). Beyond 4.0 mmol/L, the relationship attenuated (aOR = 0.90, 95%CI: 0.52–1.43; p = 0.568). Subgroup analyses indicated amplified risk among epinephrine-exposed infants (aOR = 3.40 vs. 1.78; P interaction = 0.094) and those with moderate-to-severe respiratory failure. Conclusion Lactate reveals phase-specific mortality associations in VLBW infants with LOS. The 7.0-fold mortality risk increase per mmol/L in the 2.2–4.0 mmol/L interval suggests potential metabolic resuscitation targets for precision monitoring in sepsis management.