Low circulatory levels of total cholesterol, HDL-C and LDL-C are associated with death of patients with sepsis and critical illness: systematic review, meta-analysis, and perspective of observational studies

Abstract

Background: Mechanistic studies have established a biological role of sterol metabolism in infection and immunity with clinical data linking deranged cholesterol metabolism during sepsis with poorer outcomes. In this systematic review we assess the relationship between biomarkers of cholesterol homeostasis and mortality in critical illness.

Methods: We identified articles by searching a total of seven electronic databases from inception to October 2023. Prospective observational cohort studies included those subjects who had systemic cholesterol (Total Cholesterol (TC), HDL-C or LDL-C) levels assessed on the first day of ICU admission and short-term mortality recorded. Meta-analysis and meta-regression were used to evaluate overall mean differences in serum cholesterol levels between survivors and non-survivors. Study quality was assessed using the Newcastle-Ottawa Scale.

Findings: From 6469 studies identified by searches, 24 studies with 2542 participants were included in meta-analysis. Non-survivors had distinctly lower HDL-C at ICU admission -7.06 mg/dL (95% CI -9.21 to -4.91, p < 0.0001) in comparison with survivors. Corresponding differences were also seen less robustly for TC -21.86 mg/dL (95% CI -31.23 to -12.49, p < 0.0001) and LDL-C -8.79 mg/dL (95% CI, -13.74 to -3.83, p = 0.0005).

Interpretation: Systemic cholesterol levels (TC, HDL-C and LDL-C) on admission to critical care are inversely related to mortality. This finding is consistent with the notion that inflammatory and metabolic setpoints are coupled, such that the maladaptive-setpoint changes of cholesterol in critical illness are related to underlying inflammatory processes. We highlight the potential of HDL-biomarkers as early predictors of severity of illness and emphasise that future research should consider the metabolic and functional heterogeneity of HDLs.

Funding: EU-ERDF-Welsh Government Ser Cymru programme, BBSRC, and EU-FP7 ClouDx-i project (PG).

Keywords: Cholesterol; Critical illness; Immunometabolism; Lipids; Sepsis.

Fig. 1

Study selection flow-diagram showing included and excluded studies at each stage.

Fig. 2

Unadjusted meta-analysis for total cholesterol showing pooled mean difference in serum total cholesterol and 95% CI between non-survivors and survivors (control) across studies (mg/dL). Heterogeneity shown as I².

Fig. 3

Unadjusted meta-analysis for HDL cholesterol showing pooled mean difference in serum HDL-C and 95% CI between non-survivors and survivors (control) across studies (mg/dL). Heterogeneity shown as I².

Fig. 4

Unadjusted meta-analysis for LDL cholesterol showing pooled mean difference in serum LDL-C and 95% CI between non-survivors and survivors (control) across studies (mg/dL). Heterogeneity shown as I².

Fig. 5

Schematic highlighting normal and inflammatory associated HDL and LDL particle features and macrophage cholesterol immunometabolism. Panel a top left represents normal baseline conditions with healthy cholesterol efflux, anti-inflammatory and cytoprotective functions of HDL particles and associated markers: Apolipoprotein A1 (ApoA1), Apolipoprotein E (ApoE), Paraoxonase 1 (PON1) and LDL with Apolipoprotein B (ApoB). Top right panels illustrates changes that occur in systemic inflammatory states where lower paraoxonase 1 (PON1) levels decrease the capacity of HDL to prevent oxidation of LDL and the exchange of APOA1 with the acute phase Serum Amyloid A (SAA) protein can lead to reduced cholesterol efflux functions; lower levels of cholesterol esters (CE) and phospholipids (PL); number of small HDL particles (sHDLp) and potential development of ‘pro-inflammatory’ subclass of HDL (dysHDL) with associated marker of SAA and increased proportion of oxidised LDL (oxLDL) and phosphotidylcholine (oxPC). Arrows indicate causal drivers of IL-6 signalling in reducing small HDL particle numbers. ApoIII = Apolipoprotein C-III, SAA1 = serum amyloid A1, Lp-PLA2 = Lipoprotein-Associated Phospholipase A2. Panel b illustrates the immunometabolism of cholesterol biosynthesis and the delivery of cholesterol via LDL binding its cognate receptor (LDLR) and in inflammatory states, where LDLR is downregulated, oxLDL has preferential uptake via the Scavenger Receptor (SR-B1). The efflux of cholesterol to HDL in macrophage uses primarily the cholesterol exporter (ABCG1) that in inflammatory conditions is upregulated in conjunction with an increased proportion of dysfunctional HDL particles (dysHDLp) that have lost their anti-inflammatory features. The linkage to immunometabolism is indicated by the change in flux of cholesterol biosynthesis and its association with macrophage immunophenotype switching between pro and anti-inflammatory states.