Title : Glycaemic variability and all-cause mortality in adult patients with sepsis: A systematic review and critical appraisal
Abstract:
Background: Sepsis accounts for an estimated 49 million cases and 11 million deaths globally each year, representing approximately one in five of all global fatalities. Identifying modifiable risk factors associated with mortality in septic patients is therefore a clinical and public health priority. Glycaemic variability (GV), referring to acute fluctuations in blood glucose levels within or between days, has emerged as a potentially important prognostic marker in critically ill patients. Experimental evidence suggests GV may exacerbate oxidative stress and cellular damage more profoundly than sustained hyperglycaemia through overproduction of reactive oxygen species and activation of pro-inflammatory cascades. Despite this, current Surviving Sepsis Campaign guidelines do not address GV as a therapeutic target. This review systematically examined and critically appraised the existing evidence to determine whether high GV correlates with increased all-cause mortality in adult patients with sepsis.
Methods: A systematic review and critical appraisal of all available clinical evidence examining acute GV and all-cause mortality in adult septic patients was conducted. The existing literature was critically evaluated with individual studies appraised using the Newcastle-Ottawa Scale to assess methodological quality. An independent updated literature search was subsequently performed on 1 June 2024 across six databases (PubMed, Embase, Ovid Medline, Web of Science, Cochrane and Scopus) using standardised search terms combining variations of glycaemic variability, sepsis and mortality. Studies with significant methodological concerns or uncertain sepsis classification were excluded following critical appraisal. A forest plot was independently constructed to synthesise findings.
Results: Eleven cohort studies involving over 11,000 adult septic patients were critically appraised. One study was excluded from quantitative synthesis due to significant methodological flaws and uncertainty regarding sepsis classification. Across the remaining studies, a consistent association was identified between high acute GV and increased all-cause mortality, regardless of GV measurement parameter: standard deviation of blood glucose, coefficient of variation, mean amplitude of glycaemic excursion and glycaemic lability index. The updated forest plot, incorporating the largest available cohort to date (n=7,104), yielded a statistically significant pooled effect (OR 1.05, 95% CI 1.02-1.08). Notably, this effect was attenuated compared to earlier smaller studies, suggesting previous estimates may have overstated the true association. A consistent finding across the majority of studies was that non-diabetic patients demonstrated greater susceptibility to GV-associated mortality than diabetic patients, possibly reflecting adaptation to chronic glucose fluctuations.
Conclusion: This systematic review and critical appraisal confirms a consistent and biologically plausible association between high acute glycaemic variability and increased all-cause mortality in adult septic patients. However, the evidence base remains dominated by retrospective observational studies with significant methodological heterogeneity and effect size attenuates with larger more rigorous studies, warranting cautious interpretation. Large-scale multicentre randomised controlled trials using continuous glucose monitoring and standardised GV protocols are needed to establish causation. If confirmed, GV represents a clinically important and potentially modifiable therapeutic target that should be considered in future sepsis management guidelines, with particular focus on non-diabetic patient subgroups.
Keywords: Glycaemic variability, Sepsis, Mortality, Critical care, Intensive care, Blood glucose, Dysglycaemia, Systematic review
