The association between stress hyperglycemia and poor outcome in critically ill children is modulated by hyperlactatemia

Abstract

Background: The available evidence on tight glycemic control is conflicting, while the interaction between glucose and lactate in critically ill children remains unclear.

Objective: To explore the potential role of hyperlactatemia (HL) in modulating the relationship between stress hyperglycemia (SHG) and poor outcomes, aiming to establish tailored glucose targets in critically ill children.

Methods: This was a secondary analysis of a prospective observational cohort study conducted in five Pediatric Intensive Care Units (PICU) in southwestern China (ChiCTR2000030846). The interaction effect between glucose and lactate metrics concerning outcomes and subsequent subgroup regression analysis was conducted. SHG was defined as glucose > 150 mg/dL(8.3mmol/L) and HL as lactate > 2 mmol/L.

Results: A cohort of 433 pediatric patients with 4885 arterial blood gas measurements were finally enrolled. 90 (20.8%) cases died within 28 days of PICU admission. Significant interaction effects between SHG and HL on outcomes were observed (p < 0.05). In the non-HL group, SHG was not an independent predictor of 28-day mortality (p = 0.656) and was not correlated with either 28-day ventilator-free days (p = 0.916) or 28-day ICU-free days (p = 0.914). In contrast, in the HL group, SHG was independently associated with 28-day mortality (OR 3.55, 95% CI 1.62~7.80, p = 0.002) and correlated with a reduction of 5.04 28-day ventilator-free days (p = 0.003) and 4.10 28-day ICU-free days (p = 0.004).

Conclusions: HL potentially modulates the correlation between SHG and poor outcomes in pediatric critically ill patients. Combined SHG and HL are associated with poor outcomes, whereas SHG without HL is not.

Keywords: children; critically ill; hyperlactatemia; outcome; stress hyperglycemia.

Figure 1

Flowchart illustrating participant selection for the study.

Figure 2

Three-dimensional column diagram of lactate and glucose quartiles related to 28-day mortality. (A) The combined highest Lac_mean and Glu_mean quartiles (upper-right column) indicated the highest 28-day mortality, whereas relatively low Lac_mean even with the highest Glu_mean didn’t correlate with high 28-day mortality (upper-left columns); (B) For the combination of Lac_twmean and Glu_twmean, their highest quartiles (upper-right column) corresponded to peak 28-day mortality. However, lower Lac_twmean remained unrelated to high mortality, even with the highest Glu_twmean (upper-left columns); (C) The combined highest Lac_mean and Glu_max quartiles (upper-right column) corresponded to the highest 28-day mortality, while relatively low Lac_mean, even at the highest Glu_mean quartiles, showed no association with elevated mortality (upper-left columns). Lac_mean, mean lactate concentration; Lac_twmean, time-weighted mean lactate; Lac_max, maximum lactate concentration. Glu_mean, mean glucose concentration; Glu_twmean, time-weighted mean glucose; Glu_max, maximum glucose concentration.

Figure 3

Comparison of 28-day Mortality between SHG and non-SHG group across HL and non-HL Subgroups. ^† p, the p-value from the Breslow-Day test (= 0.007) of Cochran-Mantel- Haenszel test, indicates a significant difference in the odds ratios (ORs) of the association between SHG and 28-day mortality between the HL and non-HL groups. *, the difference between groups was significant.