Introduction
Hyperglycemia, glucose intolerance and insulin resistance are common features of critically ill patients, especially in patients with sepsis or septic shock and even in those without a history of diabetes [
1‐
3]. Maintenance of normoglycemia (blood glucose levels of 110 mg/dL) with intensive insulin therapy has been shown to improve survival and reduce morbidity in patients with prolonged critically illness after cardiac surgery [
4], whereas its beneficial effect on the outcome of patients in medical intensive care units (ICU) is controversial [
5,
6]. In patients with obesity, metabolic syndrome and type 2 diabetes, several adipocytokines have been identified that mediate agonistic and antagonistic effects on insulin resistance [
7,
8]. In these patients, chronic inflammatory conditions apparently promote adipocytic secretion of mediators of insulin resistance into the circulation, thereby providing a link between adipocytokines, inflammation and systemic insulin resistance [
7,
8].
Retinol binding protein 4 (RBP4), a 21-kDa protein synthesized in the liver and adipose tissue, has recently been described as an adipokine involved in the development of insulin resistance in mice and humans [
9,
10]. In the past, RBP4 was solely recognized for its role as the specific transport protein for vitamin A (retinol) in the circulation, whose function is to deliver hydrophobic retinol from the liver to target tissues [
11]. In animal models of insulin resistance, the expression of RBP4 was strongly induced in adipose tissue, and systemic release of RBP4 appeared to be a crucial signal for the development of systemic insulin resistance [
10,
12]. These results have been translated into the pathogenesis of insulin resistance in humans. Serum RBP4 correlated positively with the presence of insulin resistance in individuals with obesity, impaired glucose tolerance or type 2 diabetes [
13,
14]. Elevated serum RBP4 concentrations were an independent predictive biomarker at early stages of insulin resistance identifying individuals at risk of developing diabetes and were even found in healthy individuals with a strong family history of type 2 diabetes [
13,
14].
Recently, low levels of serum RBP4 have been reported in critical ill patients with sepsis of pulmonary origin compared to nonseptic patients [
15]. However, it remained unclear whether this observation was related to sepsis or whether it could be extrapolated to all critically ill patients, because this study analyzed only patients with a respiratory disease as the main reason for admission to the ICU. Moreover, the pathogenic and/or diagnostic relevance of RBP4 in ICU patients is presently not known [
15]. Various essential therapeutic interventions in the initial phase of intensive care treatment, e.g., fluid challenge, insulin therapy, nutritional support and renal replacement therapies, potentially influence concentrations of biomarkers. At present, there is a lack of studies investigating the regulation of adipocytokines in critical illness at the point of admission before any substantial therapeutic interference.
Our study investigated serum RBP4 concentrations in a large cohort of "untouched," treatment-naive critically ill patients (septic and nonseptic patients) from a medical ICU at the moment of admission to the ICU. We aimed at understanding the potential involvement of RBP4 in the pathogenesis of insulin resistance in critical illness, its regulation in severe systemic inflammation and its potential clinical use as a biomarker in ICU patients. We demonstrate that serum RBP4 levels were significantly reduced in ICU patients as compared to controls, independent of etiology of critical illness or origin of sepsis (pulmonary versus nonpulmonary), and they were closely related to liver and kidney function. The dysregulation of serum RBP4 in ICU patients was not associated with preexisting diabetes or obesity, but might contribute to the insulin-resistant state of the critically ill patients. High RBP4 levels were indicative of a beneficial short-term course of disease, but did not predict the overall survival of the patients in a 3-year follow-up period.
Discussion
RBP4, the circulating transporter for vitamin A, has recently been recognized as an important mediator of insulin resistance in mice and humans, but its potential role in sepsis or critical illness is presently a matter being researched. A recent single-center study reported low circulating levels of RBP4 in patients with acute critical illness of respiratory etiology compared to healthy controls. In the cohort of septic patients with identified pulmonary focus, lower levels of RBP4 were found than in nonseptic patients [
15]. The authors hypothesized that an acute decrease of RBP4 concentrations could be explained by reduced synthesis or increased removal by extravasation due to capillary leakage or increased metabolic clearance [
15].
Our study shows that reduced serum concentrations of RBP4 are a general response in critically ill patients, independent of the origin of the critical illness (sepsis or nonsepsis). The source of circulating RBP4 in critically ill patients appears to be primarily a combination of two factors: hepatic synthesis and renal clearance.
It is well established that the liver is the main source of RBP4 in humans [
11,
12], although adipocyte RBP4 secretion has been linked to obesity, insulin resistance and type 2 diabetes [
12,
13,
25]. Consequently, patients with chronic liver diseases and liver cirrhosis have significantly decreased RBP4 serum concentrations, and serum RBP4 is directly linked to liver function in these patients [
18,
26,
27]. Interestingly, we also observed a close correlation between the hepatic biosynthetic function and serum RBP4 in ICU patients, suggesting that circulating serum RBP4 is a direct consequence of hepatic synthesis comparable to the immediately increased hepatic synthesis of many acute phase proteins in critical illness. Furthermore, the lack of correlation between RBP4 and other adipokines indicates that the adipose tissue does not extensively contribute to serum RBP4 in patients with critical illness.
In addition, it is well established that serum RBP4 is increased in patients with chronic renal failure, making renal clearance a relevant confounding factor for serum RBP4 concentrations in patients with diabetes or renal disease [
28,
29]. As acute renal failure, mainly of prerenal origin during hemodynamic deterioration, is a classical feature of critically ill patients, our study demonstrated a direct correlation between increasing serum RBP4 and decreasing renal function in ICU patients. In multivariate analysis, hepatic and renal biomarkers independently correlated with serum RBP4 (data not shown), indicating that both mechanisms occur in parallel in critically ill patients regulating serum RBP4 levels. However, overall RBP4 serum levels remained significantly lower than in healthy controls.
Our results further indicate that the elevated serum RBP4 might be involved in the development of insulin resistance in ICU patients. Hyperglycemia and insulin resistance are well-known features of critical illness and affect the mortality rate, especially in septic patients [
1,
30,
31]. Prior studies have confirmed excessive endogenous insulin secretion during sepsis, which represented an important pathogenetic and prognostic factor in ICU patients [
30,
32]. We show here that the circulating RBP4 level in ICU patients is not the result of preexisting diabetes or glucose intolerance (as reflected by HbA1c or plasma glucose). However, we did see a correlation with current endogenous insulin secretion, reflected by elevated C-peptide levels, and insulin resistance as calculated by the HOMA index, indicating that circulating RBP4 levels in ICU patients may contribute to the insulin-resistant state frequently observed in critically ill patients. Further studies in appropriate animal models of experimental sepsis are needed to clarify the potential pathogenetic role and metabolic consequences of RBP4 in critical illness.
The reduction of serum RBP4 in critically ill patients compared to controls was also of direct prognostic relevance. Our data revealed an association between low RBP4 and higher short-term mortality on ICU, whereas long-term survival was not affected. The underlying pathomechanisms are currently unclear. One explanation could be that this is a simple epiphenomenon as high(er) RBP4 levels may indicate preserved liver function; along this line, high "classical" biomarkers of liver function are also associated with a beneficial outcome (data not shown), and the predictive value of serum RBP4 for ICU survival was statistically indistinguishable from liver function in multivariate Cox regression analysis.
On the other hand, other explanations should be considered. Interestingly, serum RBP4 levels were negatively correlated with markers of inflammation. This raises the possibility that RBP4 could be an important component of the so-called anti-acute phase proteins [
24], at least in patients with sepsis. In line with this possibility, decreasing serum RBP4 levels have been reported after acute surgery [
33]. Functionally, elevated circulating RBP4 increased blood glucose in mouse models by inhibiting insulin signaling in skeletal muscle and upregulating gluconeogenesis through increased expression of the gluconeogenic enzyme phosphoenolpyruvate carboxykinase (PEPCK) in the liver [
12]. The concomitant reduction in muscular glucose uptake and increase in hepatic glucose output could possibly favor an adequate stress response in the initial phase of critical illness and sepsis, which would explain the positive association between relatively high RBP4 and recovery from acute illness.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
AK, FT and CT designed the study, analyzed data and wrote the manuscript, RW performed RBP4 and adipokine measurements, and ES, HZ, HD and SV collected data and assisted in patient recruitment.