Awad 2009bagus coy

13
Review Short-term starvation and mitochondrial dysfunction – A possible mechanism leading to postoperative insulin resistance Sherif Awad a , Dumitru Constantin-Teodosiu b , Ian A. Macdonald b , Dileep N. Lobo a, * a Division of Gastrointestinal Surgery, Nottingham Digestive Diseases Centre Biomedical Research Unit, Nottingham University Hospitals, Queen’s Medical Centre, Nottingham, UK b Centre for Integrated Systems Biology and Medicine, School of Biomedical Sciences, University of Nottingham, Nottingham, UK article info Article history: Received 22 January 2009 Accepted 10 April 2009 Keywords: Starvation Fasting Metabolic Stress Perioperative Feeding Carbohydrate Surgery Insulin resistance Insulin sensitivity Mitochondria summary Background: Preoperative starvation results in the development of insulin resistance. Measures to attenuate the development of insulin resistance, such as preoperative carbohydrate loading, lead to clinical benefits. However, the mechanisms that underlie the development of insulin resistance during starvation and its attenuation by preoperative carbohydrate loading remain to be defined. Insulin resistance associated with type 2 diabetes and ageing has been linked to mitochondrial dysfunction. The metabolic consequences of preoperative starvation and carbohydrate loading and mechanisms linking insulin resistance to impaired mitochondrial function are discussed. Methods: Searches of the Medline and Science Citation Index databases were performed using various key words in combinations with the Boolean operators AND, OR and NOT. Key journals, nutrition and metabolism textbooks and the reference lists of key articles were also hand searched. Results: Animal studies have shown that short-term energy deprivation decreases mitochondrial ATP synthesis capacity and complex activity, and increases oxidative injury. Furthermore, evidence from human studies suggests that the development of insulin resistance during starvation may be linked to impaired mitochondrial function. Conclusions: There is evidence from animal studies that short-term starvation causes mitochondrial dysfunction. Future studies should investigate whether mitochondrial dysfunction underlies the devel- opment of insulin resistance in patients undergoing elective surgery. Ó 2009 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved. 1. Introduction Recent studies have demonstrated that preoperative starvation induces metabolic stress and leads to perioperative insulin resis- tance which results in hyperglycaemia. The latter may cause increased infective complications, morbidity and mortality. Measures aimed at decreasing perioperative insulin resistance, such as the avoidance of preoperative starvation by giving patients carbohydrate drinks up to 2 h preoperatively, may lead to clinical benefits. However, the mechanisms that underlie the development of perioperative insulin resistance during starvation and its atten- uation by preoperative carbohydrate drinks are yet to be defined. Understanding these mechanisms would allow the optimisation and improvement of interventions designed to reduce insulin resistance. This review discusses the adverse effects of short-term (up to 36 h) starvation, the development of insulin resistance and its clinical significance, the evidence base that relates to preoper- ative carbohydrate loading and the mechanisms that may link mitochondrial dysfunction to the development of insulin resistance. 2. Search strategy Searches of the Medline (Ovid, PubMed, Embase) and Science Citation Index databases, and the GoogleÔ search engine were performed using the key words metabolic, stress, metabolism, hormones, insulin, insulin resistance, insulin sensitivity, starvation, fast, preoperative, postoperative, surgery, anaesthesia, outcome, complication, carbohydrate, feed, load, mitochondria, oxidative stress and reactive oxygen species in various combinations with the Boolean operators AND, OR and NOT. Animal and human studies published in the last 30 years and key earlier articles were included. * Correspondence to: Mr. D. N. Lobo, Division of Gastrointestinal Surgery, Nottingham Digestive Diseases Centre Biomedical Research Unit, E Floor, West Block, Nottingham University Hospitals, Queen’s Medical Centre, Nottingham NG7 2UH, UK. Tel.: þ44 115 8231149; fax: þ44 115 8231160. E-mail address: [email protected] (D.N. Lobo). Contents lists available at ScienceDirect Clinical Nutrition journal homepage: http://www.elsevier.com/locate/clnu 0261-5614/$ – see front matter Ó 2009 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved. doi:10.1016/j.clnu.2009.04.014 Clinical Nutrition 28 (2009) 497–509

description

bgus bgt

Transcript of Awad 2009bagus coy

lable at ScienceDirect

Clinical Nutrition 28 (2009) 497–509

Contents lists avai

Clinical Nutrition

journal homepage: ht tp: / /www.elsevier .com/locate/c lnu

Review

Short-term starvation and mitochondrial dysfunction – A possible mechanismleading to postoperative insulin resistance

Sherif Awad a, Dumitru Constantin-Teodosiu b, Ian A. Macdonald b, Dileep N. Lobo a,*

a Division of Gastrointestinal Surgery, Nottingham Digestive Diseases Centre Biomedical Research Unit, Nottingham University Hospitals, Queen’s Medical Centre, Nottingham, UKb Centre for Integrated Systems Biology and Medicine, School of Biomedical Sciences, University of Nottingham, Nottingham, UK

a r t i c l e i n f o

Article history:Received 22 January 2009Accepted 10 April 2009

Keywords:StarvationFastingMetabolicStressPerioperativeFeedingCarbohydrateSurgeryInsulin resistanceInsulin sensitivityMitochondria

* Correspondence to: Mr. D. N. Lobo, DivisionNottingham Digestive Diseases Centre Biomedical RBlock, Nottingham University Hospitals, Queen’s Med2UH, UK. Tel.: þ44 115 8231149; fax: þ44 115 823116

E-mail address: [email protected] (D.N

0261-5614/$ – see front matter � 2009 Elsevier Ltd adoi:10.1016/j.clnu.2009.04.014

s u m m a r y

Background: Preoperative starvation results in the development of insulin resistance. Measures toattenuate the development of insulin resistance, such as preoperative carbohydrate loading, lead toclinical benefits. However, the mechanisms that underlie the development of insulin resistance duringstarvation and its attenuation by preoperative carbohydrate loading remain to be defined. Insulinresistance associated with type 2 diabetes and ageing has been linked to mitochondrial dysfunction. Themetabolic consequences of preoperative starvation and carbohydrate loading and mechanisms linkinginsulin resistance to impaired mitochondrial function are discussed.

Methods: Searches of the Medline and Science Citation Index databases were performed using variouskey words in combinations with the Boolean operators AND, OR and NOT. Key journals, nutrition andmetabolism textbooks and the reference lists of key articles were also hand searched.

Results: Animal studies have shown that short-term energy deprivation decreases mitochondrial ATPsynthesis capacity and complex activity, and increases oxidative injury. Furthermore, evidence fromhuman studies suggests that the development of insulin resistance during starvation may be linked toimpaired mitochondrial function.

Conclusions: There is evidence from animal studies that short-term starvation causes mitochondrialdysfunction. Future studies should investigate whether mitochondrial dysfunction underlies the devel-opment of insulin resistance in patients undergoing elective surgery.

� 2009 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.

1. Introduction

Recent studies have demonstrated that preoperative starvationinduces metabolic stress and leads to perioperative insulin resis-tance which results in hyperglycaemia. The latter may causeincreased infective complications, morbidity and mortality.Measures aimed at decreasing perioperative insulin resistance,such as the avoidance of preoperative starvation by giving patientscarbohydrate drinks up to 2 h preoperatively, may lead to clinicalbenefits. However, the mechanisms that underlie the developmentof perioperative insulin resistance during starvation and its atten-uation by preoperative carbohydrate drinks are yet to be defined.Understanding these mechanisms would allow the optimisation

of Gastrointestinal Surgery,esearch Unit, E Floor, Westical Centre, Nottingham NG70.. Lobo).

nd European Society for Clinical N

and improvement of interventions designed to reduce insulinresistance. This review discusses the adverse effects of short-term(up to 36 h) starvation, the development of insulin resistance andits clinical significance, the evidence base that relates to preoper-ative carbohydrate loading and the mechanisms that may linkmitochondrial dysfunction to the development of insulinresistance.

2. Search strategy

Searches of the Medline (Ovid, PubMed, Embase) and ScienceCitation Index databases, and the Google� search engine wereperformed using the key words metabolic, stress, metabolism,hormones, insulin, insulin resistance, insulin sensitivity, starvation,fast, preoperative, postoperative, surgery, anaesthesia, outcome,complication, carbohydrate, feed, load, mitochondria, oxidativestress and reactive oxygen species in various combinations with theBoolean operators AND, OR and NOT. Animal and human studiespublished in the last 30 years and key earlier articles were included.

utrition and Metabolism. All rights reserved.

S. Awad et al. / Clinical Nutrition 28 (2009) 497–509498

Articles published in languages other than English, those publishedonly in abstract form and case reports were excluded. Key journals,textbooks on nutrition and metabolism, and the reference lists ofkey articles were also hand searched.

Fig. 1. Signal transduction chain for the metabolic regulation by insulin. Insulinbinds to its receptor leading to auto-phosphorylation of tyrosine residues in thereceptor protein. This leads to interaction with a family of proteins known as insulinreceptor substrates (IRS), which themselves become phosphorylated and then interactwith the enzyme phosphatidylinositol-3-kinase (PI3K). PI3K generates phosphatidy-linositol (30 ,40 ,50)-trisphosphate (PIP3) in the inner surface of the membrane, whichacts through the enzyme 30-phosphotidylinositol dependent kinase-1 (PDK1) tophosphorylate (and active) protein kinase B (PKB). Activated PKB leads to severalcellular responses to insulin including inhibition of lipolysis, increased glucose trans-port, effects on DNA transcription, and phosphorylation and inactivation of glycogensynthase kinase 3 (GSK3). Inactivation of GSK3 also leads to multiple cellular effectsincluding stimulation of glycogen synthesis, effects on gene expression and protein

3. Metabolic regulation and the development of insulinresistance during starvation and surgery

The regulation of the body’s energy reserves is crucial to survivaland is brought about by metabolic pathways controlled bya number of hormones, the most important being insulin andglucagon (Table 1). Plasma glucose concentrations are primarilycontrolled by insulin and glucagon. On binding to its cell membranereceptor (Fig. 1), insulin initiates a signal that permits the storage ofenergy within liver, muscle and adipose tissue.1,2 The facilitativeglucose transporter (GLUT-4) is of importance in the regulation ofthe body’s energy reserves by permitting the movement of glucosedown a concentration gradient across cell membranes. Followingthe binding of insulin to its receptor, the subsequent intracellularsignal leads to the translocation of intracellular GLUT-4-rich vesi-cles to the cell membrane.1 This increases the amount of GLUT-4transporters available for glucose transport into the cell. Bycontrast, glucagon acts mainly on the liver to increase plasmaglucose concentration. A fall in plasma glucose concentration will,therefore, lead to a decreased ratio of plasma insulin to glucagon,the net effect being elevation of blood glucose concentration.

chain initiation (i.e. mRNA translation). Copyright� 2003 Blackwell. From Metabolicregulation: a human perspective, Frayn KN, 2nd edition, 2003. Modified withpermission from Blackwell publishing.

3.1. Body fuel reserves

Carbohydrate is stored as glycogen with only skeletal muscleand the liver having sufficient glycogen reserves to fulfill bodilyneeds. Skeletal muscle comprises 40% of body weight and containsaround 350–400 g of glycogen (15 g/kg muscle).3 As muscle lacksthe enzyme glucose-6-phosphatase, this store of glycogen cannotbe released into the circulation as glucose. Instead, muscle utilizesglycogen to release glucose precursors (lactate, pyruvate and/oralanine) that are used by the liver for gluconeogenesis.3 Liverglycogen reserves are more readily available in the form of glucoseand play the major role of ‘buffering’ glucose concentrations. Theliver glycogen content varies and is dependent on factors such asdiet and exercise, but is typically around 50–120 g (50–80 g/kgliver).

Table 1Hormones involved in the metabolic regulation of the body’s energy reserves.

Hormone Metabolic effects

Insulin Carbohydrate metabolism: inhibits glycogenolysis, stimulates glcellular glucose uptakeProtein metabolism: decreases protein catabolism, increases amFat metabolism: stimulates lipogenesis, stimulates uptake of fattylipolysis (inhibition of hormone-sensitive lipase), inhibits fatty acGene expression: control of genes involved in glucose metabolis

Glucagon Carbohydrate metabolism: stimulates glycogenolysis, stimulateProtein metabolism: stimulates amino acid catabolismFat metabolism: stimulates fatty acid oxidation, stimulates keto

Glucocorticoids Carbohydrate metabolism: stimulate gluconeogenesis, inhibit mProtein metabolism: stimulate protein catabolismFat metabolism: stimulate lipolysis

Noradrenaline/Adrenaline

Carbohydrate metabolism: stimulate glycogenolysisProtein metabolism: stimulate amino acid catabolismFat metabolism: stimulate lipolysis (via hormone-sensitive lipas

Thyroid hormones Modulate level of response of body to other hormones, in particuProtein metabolism: stimulate protein catabolism

Growth hormone Carbohydrate metabolism: stimulates hepatic gluconeogenesisProtein metabolism: stimulates protein synthesisFat metabolism: stimulates lipolysis

3.2. Energy metabolism interactions

The post-absorptive state is typically represented by the situa-tion after an overnight fast. Plasma glucose concentration is justunder 5 mmol/l and the concentration of insulin is typically around60 pmol/l. At this time, plasma concentrations of glucose andinsulin are at the nadir of the 24-h cycle and glucose enters theblood almost exclusively from the liver following both glycogen-olysis and gluconeogenesis. Gluconeogenic substrates consist oflactate (from sources including muscle, erythrocytes and the renalmedulla), alanine (from muscle) and glycerol (from adipose tissue).Much of the glucose thus produced is taken up by the brain. The

ycogenesis, inhibits gluconeogenesis, stimulates glycolysis, stimulates muscle and fat

ino acid uptakeacids from plasma triacylglycerol into adipose tissue (via lipoprotein lipase), inhibitsid oxidation, inhibits ketogenesism and de novo lipogenesiss gluconeogenesis, inhibits glycolysis

genesisuscle glucose uptake

e)lar, regulate sensitivity of metabolic processes to catecholamines

and glycogenolysis

S. Awad et al. / Clinical Nutrition 28 (2009) 497–509 499

low concentrations of glucose and insulin result in little uptake ofglucose by non-neuronal tissue, net breakdown of protein byskeletal muscle, liberation of fatty acids from adipose tissue (lack ofrestraint of the insulin-sensitive hormone lipase) and ketone bodyformation (provides fuel for muscle, brain and adipose tissue). Theliberated fatty acids become the preferred fuel for muscle, thussparing any plasma glucose for use by brain, erythrocytes and therenal medulla (i.e. obligatory glucose-utilising tissues).2,3

Food intake, digestion and absorption stimulate pancreaticrelease of insulin. The resultant increase in the insulin:glucagonratio switches hepatic glycogen metabolism from breakdown tosynthesis, reduces the release of fatty acids from adipose tissue andincreases glucose uptake by skeletal muscle. The decrease inplasma fatty acid concentration and increase in glucose uptakereduce the drive for muscle to oxidize fatty acids. There follows anincrease in glucose oxidation with increased production of lactateand pyruvate (due to increased glycolysis), increase in muscleglycogenesis and net protein synthesis.2,3 The increased substratesupply also stimulates hepatic gluconeogenesis and glycogenesis.The increase in insulin concentration also drives esterification andstorage of fatty acids as triacylglycerol in adipocytes.

3.3. Metabolic effects of short-term starvation

As liver glycogen stores are virtually depleted within 24 h,3,4

gluconeogenesis supplies the requirements of the brain and otherglucose-requiring tissues. The low insulin:glucagon ratio and theincreased supply of gluconeogenic substrates stimulate gluconeo-genesis. Falling insulin concentrations lead to both net proteolysisin muscle, with release of alanine and glutamine, and lipolysis inadipose tissue, with release of glycerol and non-esterified fattyacids (NEFAs). The increased availability of NEFAs directly stimu-lates muscle oxidation of fat rather than glucose. Gluconeogenesisat this stage proceeds at the expense of muscle protein but giventhat the brain requires around 100–120 g of glucose per day, therate of muscle protein breakdown could be rapid (up to 210 gprotein per day – as not all amino acids can be converted to glucose,around 1.75 g of muscle protein must be broken down to provide1 g of glucose).2,3,5

3.4. Perioperative reduction in insulin sensitivity

Impaired insulin sensitivity or ‘insulin resistance’ signifiesa state of reduced peripheral and hepatic responsiveness to thebiological actions of insulin.2 Approximately 25% of normal indi-viduals and up to 85% of type 2 diabetic populations are insulinresistant.6 Insulin resistance is regarded as the major metabolicanomaly underlying the group of diseases that comprise themetabolic syndrome1,6 (type 2 diabetes, obesity, dyslipidaemia,hypertension, hypercoagulability and non-alcoholic steatohepati-tis) and it also occurs transiently after starvation, trauma andsurgery.1,7–9

There are two main sites of insulin resistance: peripheral tissue(mainly skeletal muscle) and the liver. The former is responsible for70–90% of glucose disposal following a carbohydrate load.10 Inpatients with type 2 diabetes, the development of insulin resistancein muscle has been attributed to decreased insulin-stimulatedmuscle glycogen synthesis, which appears to be at least partly dueto defects in glucose uptake via GLUT-4 transporters.11 Both raisedplasma fatty acid concentrations and defects in mitochondrialfunction (see later) are associated with intramyocellular accumu-lation of lipid metabolites. These lead to defective GLUT-4 activityby abolishing insulin activation of IRS (insulin receptor substrate)-1-associated phosphatidylinositol-3-kinase activity, thus inter-fering with insulin-mediated activation of GLUT-4.12

In healthy volunteers a marked reduction in insulin sensitivityoccurs after short-term (1–3 days) starvation.9,13–15 Although thecellular mechanisms underlying the development of insulin resis-tance following short-term starvation remain to be elucidated, theincreased levels of plasma fatty acids13 and reduction in insulin-stimulated glucose uptake16,17 suggest that such impairment inglucose metabolism may also result from defective GLUT-4 activity.

A reduction in insulin sensitivity also occurs as part of themetabolic response to stress such as trauma,7 burn injury8 andsepsis.18,19 More recently, a number of studies have demonstrateda reduction of up to 50% in insulin sensitivity20–27 followinguncomplicated elective surgery in healthy non-diabetic patients.Nordenstrom et al.26 studied 7 healthy patients before and 24 hafter elective open cholecystectomy. Compared with a controlgroup of 5 patients undergoing elective inguinal hernia repair, butsubjected to an otherwise almost identical perioperative careprotocol, the cholecystectomy group had significantly increasedpostoperative plasma concentrations of glucose (15%) and insulin(50%). These were associated with significant reductions in glucosetransport (35%) and insulin-stimulated lipogenesis (50%) in isolatedfat cells.26 Brandi et al.25 studied 7 patients with a normal glucose-tolerance test before and after uncomplicated elective left colonicresection. Postoperatively, patients exhibited the hallmarks ofsurgery-induced hypercatabolism (increased protein oxidation andenergy expenditure), associated with increased plasma concen-trations of counter-regulatory hormones (cortisol, glucagons,prolactin and growth hormones) and urinary output of catechol-amines. These changes were associated with an eight-fold increasein insulin needed to maintain euglycaemia during 24 h of paren-teral nutrition compared to preoperative requirements. Further-more, 24 h of insulin supplementation during parenteral nutritionnormalised glucose oxidation, restrained lipolysis and preservedprotein stores. However, in a study of 10 patients undergoingelective open cholecystectomy (a moderate surgical stress), therewas a 54% reduction in insulin sensitivity on the first postoperativeday, as determined by hyperinsulinaemic-normoglycaemicclamps.20 A further study of 16 patients undergoing elective opencholecystectomy,28 using the same anaesthetic and surgicalprotocol as the aforementioned study,20 reported no increase in thelevels of counter-regulatory hormones (catecholamines, growthhormone and cortisol) postoperatively. Thus the role played bythese hormones in the aetiology of postoperative insulin resistanceremains unclear.

Hepatic and muscle insulin resistance and increased reliance onfat oxidation have been observed in burn8 and septic18,19 patients. Astudy that examined the impact of surgical stress on intermediarymetabolism in 9 patients undergoing major abdominal surgery29

found similar changes following uncomplicated elective surgery.This insulin resistance was associated with both a reduction inperipheral glucose uptake30 and non-oxidative glucose disposal(mainly glycogen synthesis).29,30 The contribution of reducedenergy intake and bed-rest to the development of perioperativeinsulin resistance was examined in another study that compared 7patients undergoing moderate to major abdominal surgery with 6healthy volunteers who had a similar period (24 h) of rest andreduced energy intake.16 During insulin infusion, 20–30% reduc-tions in insulin-stimulated glucose uptake were found asa response to 24 h of bed-rest and reduced energy intake in healthycontrols. Another study later reported that it was reduced energyintake but not bed-rest that resulted in a decline in peripheralinsulin sensitivity.13 Reduced insulin sensitivity was related to themagnitude of the operation performed31 and persisted for up to 3weeks postoperatively in another study of 10 patients undergoingelective open cholecystectomy.20 The degree of postoperativeinsulin resistance was shown to correlate with length of

S. Awad et al. / Clinical Nutrition 28 (2009) 497–509500

postoperative hospital stay in a retrospective analysis of datapooled from a number of Swedish studies over a period of 6 years(r2¼ 0.28, p¼ 0.0001, n¼ 60).32 Multiple regression analysis ofthese data found that type of surgery (major or minor operation),perioperative blood loss and postoperative insulin resistance wereindependent predictors of length of hospital stay. The overallpredictive value of this regression model was 71%. Furthermore, thepresence of postoperative hyperglycaemia, a consequence ofinsulin resistance, has been found to increase postoperativemortality and morbidity significantly.33,34 In a study of intensiveinsulin therapy in 1548 patients (87% of whom did not havea history of diabetes) admitted to a Belgian ICU (63% of theadmissions followed cardiac surgery), the presence of hyper-glycaemia (mean morning blood glucose level of 8.5 mmol/l inpatients treated in the conventional therapy group) resulted insignificantly increased ICU mortality (8% versus 4.6%), increased in-hospital mortality (10.9% versus 7.2%), increased septic complica-tions (7.8% versus 4.2%) and prolonged mechanical ventilation(median 12 versus 10 days), when compared to patients who wererandomised to an intensive insulin regimen to maintain normo-glycaemia (mean morning blood glucose level of 5.7 mmol/l).34 Theimprovement associated with intensive insulin therapy isthought35 to result from the prevention of glucose-induced toxicityto the mitochondria,36 endothelium37 and immune cells,38 and notthe insulin dose per se.35

4. Preoperative starvation and carbohydrate loading

Current elective surgical practice is to starve patients for 6 h forsolids and 2 h for clear liquids preoperatively, which is based onhistorical concerns of increased risks of aspiration of stomachcontents during anaesthesia.39 However, delays and changes inoperating schedules result in patients being starved for longerperiods – even up to 18 h.40 The dogma of preoperative starvationhas been challenged recently and both animal41–50 andhuman4,13,14,16,21,45,51–56 studies have demonstrated adverse meta-bolic effects caused by preoperative starvation.57,58

4.1. Preoperative starvation: history and current guidelines

The history behind the dogma of preoperative fasting has beenreviewed comprehensively by Maltby.39 It appears that the adop-tion and blanket imposition of the ‘nil per os (NPO) from midnight’guideline for healthy patients undergoing elective surgery, withoutdue distinction between solids and liquids, arose following transferof principles of emergency anaesthesia to elective practice, incor-rect deductions from the results of animal experiments and theperceived ease39 with which such guidelines could be followed inclinical practice.

In the last two decades a number of clinical studies have chal-lenged the traditional belief that all healthy patients undergoingelective surgery should be NPO from midnight. A meta-analysis ofthese studies concluded that intake of oral liquids was safe until 2 hbefore general anaesthesia.59 A Cochrane review of 22 trials with2270 participants similarly found that amongst healthy, non-pregnant, adult patients undergoing elective surgery, there were noreported cases of aspiration/regurgitation and no evidence ofa difference in residual gastric volume or pH of gastric contentswhen a shortened fluid fast (90–180 min preoperatively) wascompared with a standard fast (NPO from midnight).60 Further-more, the volume of fluid intake did not have an impact onparticipants’ gastric volume or pH when compared to a standardfast. Even in patients with comorbid conditions (American Societyof Anaesthesiologists grade IV/V) undergoing emergency surgery,

aspiration occurred infrequently and mortality was low(1:71,829).61

Current guidelines from various national anaesthetic societiespermit a light meal (dry toast and clear liquid) not less than 6 hbefore surgery and unrestricted clear liquids (water, fruit juicewithout pulp, carbonated drinks, clear tea and black coffee) until2 h before surgery.62–64 The intake of these fluids is, however,unlikely to have a major effect on the metabolic state of the bodywhich remains in the starved state. Recently, several authors9,24,65

have raised concerns about performing surgery in the ‘metaboli-cally-stressed’ state after overnight starvation, especially as surgeryitself causes severe catabolic stress.66 Thence followed a number ofexperiments that compared the effects and outcomes of under-going surgery in the starved and ‘fed’ states.

4.2. Animal studies on the metabolic effects of preoperativestarvation

Evidence that short-term starvation has adverse effects on themetabolic and haemodynamic responses to stress was clear fromrat experiments dating back to 1945.40 Compared to starved rats,fed rats were found to lose less nitrogen and have a better hae-modynamic response following experimental haemorrhage.Numerous investigators further studied the effects of starvationand feeding prior to exposure to various stresses on mortality,41–43

liver glycogen metabolism,42,44 endocrine responses,44 skeletalmuscle function67 and enteric bacterial translocation46,47

(Table 241–43,46,49,50,67).These studies proved that the metabolic state (fed versus

starved) of animals prior to exposure to stress was related tomortality and other outcomes following the stress. Although theunderlying mechanisms are yet to be elucidated, a number oftheories have been proposed. The preservation of liver glycogenstores42,44 in fed animals permits the rapid release of glucose intothe bloodstream41,42,46 following stresses such as experimentalhaemorrhage. The resultant hyperosmolar state leads to beneficialeffects on fluid homeostasis by increasing plasma refill, improvingheart function and increasing peripheral blood flow.44,48 Further-more, preservation of liver glycogen is thought to reduce the needfor mobilisation of muscle glycogen stores40 (to enable gluconeo-genesis) which in turn preserves muscle strength both before andafter exposure to stress.40,67 Recent data also suggest that liverglycogen may be of importance in maintaining normal levels ofantioxidant enzymes.49,50 An attenuated endocrine stress responsein fed animals48 reduced the ensuing catabolic response.44 Finally,the preservation of intestinal energy stores49 and, therefore,integrity of the mucosal barrier, may protect against entericbacterial translocation under conditions of stress.40,49,50

4.3. Human studies on the metabolic effects of preoperativestarvation

Few human studies have examined the effects of preoperativestarvation on morbidity and mortality.40 However, the metaboliceffects of short-term starvation in healthy humans have been wellcharacterized. A decrease in insulin sensitivity of peripheraltissues,13,14,51 regarded as a marker of perioperative metabolicstress,32 leads to a decrease in muscle glucose uptake17,51–53 whichis accompanied by a reduction in oxidative glucose disposal16,17,53

and either a decrease16,21 or no change17,53 in non-oxidative glucosedisposal. An increase in resting energy expenditure occurs due tothe metabolically more expensive processes of gluconeogenesisand ketogenesis.54 Elevations in plasma fatty acid concentra-tions13,51,55 are accompanied by decreases in the anti-lipolyticeffects of insulin,56 although this latter finding has not been

Table 2Animal studies investigating the effects of starvation prior to surgical stress. Reprinted and updated by permission of SAGE Publications from Diks et al.40 copyright 2005 (SAGEPublications). Abbreviations: ADMA, asymmetrical dimethylarginine; GSH, glutathione; IL-6, interleukin-6; IR, ischaemia-reperfusion; MDA, malondialdehyde; MPO, myeloperoxidase.

Study Study groups Stress Variable(s) examined Results

Ljungqvist et al.41 Fed versus 24 h starvation,then stress

Haemorrhage Glucose level, 7-day survival Hyperglycaemia developed in fed but not in starvedrats. All fed rats survived, whereas all starved rats died.

Esahili et al.43 Fed versus 24 h starvation,then stress

Endotoxinchallenge(intravenous orintraperitoneal)

7-day survival Starvation associated with 210–240% higher mortalityin intravenously-treated group and 190–200% highermortality in intraperitoneally-treated group.

Alibegovic andLjungqvist42

24 h starvation followed byeither 30% glucose or 0.9%saline infusion, then stress

Haemorrhage 7-day survival, liver glycogen content All starved animals died within 3 h post-haemorrhage,but all fed animals (given glucose infusion pre-haemorrhage) recovered. Liver glycogen content was600% higher in the fed group.

Friberg et al.67 Fed versus 24 h starvation,then stress

Haemorrhage Muscle function After 24 h starvation there was loss of muscle strength,even before exposure to haemorrhagic shock. Afterhaemorrhage muscle strength lower in starved group.

Bark et al.46 Fed versus 24 h starvation,then stress

Haemorrhage Presence of enteric bacteria in mesentericlymph nodes

Incidence of mesenteric lymph nodes with entericbacteria was higher in starved rats (p< 0.05) and thenumbers of bacteria were greater (p< 0.01).

Nettelbladt et al.47 Fed versus 24 h and 48 hstarvation

1. Fasting only2. Haemorrhage

Number of coliform bacteria in caecum,bacterial adherence to epithelium

24 and 48 h starvation increased number of coliformbacteria in caecum by a factor of 25 and 100,respectively. Increase in bacterial adherence tointestinal epithelium by factor of 3000.

van Hoorn et al.50 Fed versus 16 h starvation Intestinalischaemia-reperfusion model

Liver glycogen concentration,myeloperoxidase activity, reduced andoxidised tissue glutathione, ADMAconcentration, IL-6 concentration

Liver glycogen concentration significantly lower(48.2%� 7%) in fasted rats than in fed rats. Lung mye-loperoxidase activity significantly lower in fed groupthan fasted group. Lung GSH concentration significantlyhigher in fed group. Fed GSH concentration almostretained at level of sham fasted animals.Fed group had significantly lower ADMA and IL-6concentrations compared to fasted animals.

van Hoorn et al.49 Fed versus 13 h starvation Intestinalischaemia-reperfusion model

Organ function and vitality, severity ofoxidative stress, energy status of liver andintestine

Heart performance after intestinal IR worse in fastedgroup than in fed group who maintained normal values.Markers of oxidative stress (MDA concentration) higherin intestine and lungs of fasted animals. Lower ATP/ADPconcentrations in liver and intestine of fasted animals.

S. Awad et al. / Clinical Nutrition 28 (2009) 497–509 501

reproduced in other studies.14,53 Whole-body protein catabolism isincreased,45 but in contrast to perturbations in fat metabolism,muscle remains sensitive to the anti-proteolytic effect of insulin.45

Finally, depletion of liver glycogen stores occurs after as little as24 h of starvation.3,4,28 The resulting metabolic state where easilyutilisable energy is unavailable, especially during times of increasedmetabolic demand, is thought to have a detrimental effect onclinical outcome.23,68

4.4. Preoperative carbohydrate loading versus starvation

The aforementioned studies demonstrated that an insulinresistant, ‘metabolically-stressed’69 state can result from even shortperiods of starvation. Furthermore, animal studies showed a clearbenefit from being in a fed as opposed to a fasted state at the onsetof stress.70 Therefore, a number of human studies explored thepossibility that preoperative feeding, instead of starvation, enableda more favourable metabolic response to stress such as surgery(Table 321,23,24,28,68,71–80).

The main objective of these studies on preoperative carbohy-drate loading was to produce a change in metabolism similar to thatoccurring after breakfast, whereby the endogenous release ofinsulin ‘turns off’ the overnight fasted state of metabolism.81 Earlystudies achieved this through the use of intravenous glucose infu-sions (5 mg/kg/min) and demonstrated a 50% reduction in thedevelopment of postoperative insulin resistance in the carbohy-drate-treated groups.24 The high dose of glucose in these infusionswas necessary to induce a sufficiently high endogenous insulinresponse to change metabolism in the desired way27 but sucha dose carried the risk of causing thrombophlebitis.40 An iso-osmolar drink was developed with sufficient carbohydrate toinduce an insulin response similar to that after a meal27 and an

osmolality that permitted rapid emptying from the stomach.27,40,71

These drinks were shown to empty from the stomach within 2 h,71

with no instances of pulmonary aspiration or other drink-relatedcomplications23,68,71–76,78,79 reported following over 6000 patientepisodes.27 Furthermore, preoperative carbohydrate loading wasshown to attenuate the postoperative decrease in peripheral21,23,72

and hepatic75,80 sensitivity to insulin, maintain hepatic glycogenreserves,28 maintain muscle glycogen synthesis,74,77 blunt theendocrine response to surgery,21,72 attenuate the deterioration inpostoperative whole-body protein balance68,80 and muscle func-tion,74 and prevent surgery-induced immunosuppression.78 Finally,some studies have shown that preoperative carbohydrate loadingwas associated with improved patient well-being with decreasedpreoperative thirst,73,78,79 hunger73 and anxiety,73 decreasedpostoperative nausea and vomiting82 and a 20% reduction in lengthof hospital stay,27,32 although other studies have failed to demon-strate similar effects.68,74–76,78,79

The majority of the aforementioned studies were undertaken onsmall numbers of patients and although preoperative feeding wasassociated with beneficial physiological effects, such as reduction inperioperative insulin resistance, it is yet to be proved conclusivelythat preoperative feeding is associated with improved clinicaloutcomes such as reduction in mortality and morbidity.40 None-theless, preoperative carbohydrate feeding is now an accepted partof the evidence-based Enhanced Recovery After Surgery (ERAS)program which aims to allow patients to recover more quickly frommajor surgery, avoid the sequelae of traditional postoperative care(e.g. decline in nutritional status and fatigue) and reduce healthcarecosts by reducing hospital stay.57

Although studies on carbohydrate loading have demonstratedbeneficial physiological effects, the precise mechanism of action ofthese drinks is unclear.27,40 The effects on depleted glycogen stores,

Table 3Studies investigating the effects of carbohydrate loading versus starvation prior to surgery. Abbreviations: CHO, carbohydrate; EGP, endogenous glucose production; FFA,free fatty acids; GIR, glucose infusion rate; ICU, intensive care unit; IR, insulin resistance; LOS, length of stay; NEFAs, non-esterified fatty acids; Preop, preoperative; Postop,postoperative; WGD, whole-body glucose disposal.

Study Study groups Variable(s) examined Results

Ljungqvist et al.24 Patients undergoing elective opencholecystectomy receive either glucose infusion(glucose 200 mg/ml, 5 mg/kg/min) (n¼ 6) or noinfusion (n¼ 6) during preoperative fasting.Insulin sensitivity measured before surgery andon first postop day.

Glucose, insulin and stress hormone levels,insulin sensitivity

Postop glucose levels elevated in both groupsbut insulin levels elevated only in control group.No differences on postop stress hormone levelsbetween two groups. Postop insulin sensitivityreduced by 55� 3% in control group versus32� 4% in glucose group (p< 0.01).

Nygren et al.71 Patients undergoing elective laparoscopiccholecystectomy (n¼ 11) and parathyroidsurgery (n¼ 1). Randomised to 400 ml CHO richdrink (285 mOsm/kg, 12% CHO) (n¼ 6) or400 ml water (n¼ 6) 4 h before induction ofanaesthesia.

Gastric emptying rate, insulin and glucose levels Despite presence of anxiety and hunger onmorning of surgery, oral CHO drink leavesstomach within 90 min of ingestion. In CHO-drink group elevated insulin levels 40 min afteringestion (mean 67� 10 mU/ml) mimickedthose produced by intravenous administeredglucose infusions.24

Thorell et al.28 Patients undergoing elective opencholecystectomy receive either glucose infusion(glucose 200 mg/ml, 5 mg/kg/min) (n¼ 8) or noinfusion (n¼ 8) preoperatively.

Hormonal and FFA response to surgery, liverglycogen content

Preoperative FAA levels lower in glucose group.CHO loading had little effect on stress hormoneresponse to surgery but maintained liverglycogen content (65% higher than in controlgroup).

Nygren et al.21 Patients undergoing elective hip surgery.Randomised to undergo surgery with (n¼ 7) orwithout (n¼ 6) hyperinsulinaemic euglycaemicclamp. Insulin sensitivity assessed before,during and immediately after surgery.

Body metabolism, insulin sensitivity andendocrine responses

Plasma FAA higher in control group duringsurgery and postop. Cortisol levels decreased by65% in insulin group postop but not in controlgroup. Comparing preop and postop levels, nochange seen in GIR or WGD in insulin groupcompared to decreases in GIR and WGD incontrol group. Glucose oxidation lower and fatoxidation higher during and after surgery incontrol group.

Nygren et al.72 Patients undergoing elective colorectal (n¼ 14)and hip surgery (n¼ 16). Patients wererandomised to receive CHO (800 ml of 12.5%iso-osmolar drink on evening before operationand 400 ml 2 h before anaesthesia) or placebodrink preoperatively. In hip group insulinsensitivity measured 1 week preop andimmediately postop. In colorectal group insulinsensitivity measured the day before surgery and24 h postop.

Insulin sensitivity, glucose kinetics In patients undergoing hip surgery, 37%� 8%reduction (p< 0.05) in insulin sensitivitypostoperatively in fasted group whereas noreduction seen in CHO group. In patientsundergoing colorectal surgery, 24.3% greaterreduction in insulin sensitivity postoperativelyin fasted group compared to CHO group(adjusted for confounding variables). Relativereduction in WGD after colorectal surgerygreater in fasted compared to CHO group(�49%� 6% versus �26%� 8%, p< 0.05).

Soop et al.23 Patients undergoing elective hip surgery(n¼ 15). Randomisation (double-blinded) topreop CHO drink (800 ml of 12.5% iso-osmolardrink on evening before operation and 400 ml2 h before anaesthesia) (n¼ 8) or placebo drink(n¼ 7). Insulin sensitivity measured 1 weekpreop and immediately after surgery to avoidconfounding effects of reduced calorie nutritionand bed-rest.

Glucose kinetics and substrate utilization,insulin sensitivity

CHO group had lower concentrations of glyceroland NEFAs preoperatively after ingestion ofdrink. Postop decreases in GIR and WGD less inCHO group then placebo group (�18%� 6%versus �43%� 9%, p< 0.05) and (�19%� 5%versus �37%� 7%, p< 0.05), respectively. CHOgroup had significantly increased glucoseoxidation rates that persisted into postopperiod. Non-oxidative glucose disposal failed toincrease postop in both groups in response toinsulin infusion.

Hausel et al.73 Consecutive patients undergoing electivelaparoscopic cholecystectomy (n¼ 174) andmajor colorectal surgery (n¼ 78). Randomisedto 3 groups: preop CHO drink, preop placebodrink and fasted from midnight. CHO andplacebo groups were double-blinded. CHOgroup received 800 ml of 12.5% iso-osmolardrink on evening before operation and 400 ml2 h before anaesthesia.

CHO-drink-related complications, residualgastric volume, gastric acidity, preop discomfort

No cases of pulmonary aspiration or drink-related complications. Median residual gastricfluid volumes similar in all 3 groups (CHO 20 ml,placebo 20 ml and fasted 22 ml). Gastric pHsimilar in all 3 groups. CHO group had increasedpreop well-being compared to placebo andfasted groups. CHO drink relieved preop thirst,hunger, anxiety and malaise.

Henriksen et al.74 Patients undergoing elective colorectal surgery(n¼ 48) were randomised to 3 groups: preopCHO drink, preop CHO-peptide drink and preopfasting (allowed to drink water). Patients inintervention groups given 800 ml of theintervention drink the evening before and400 ml 3 h before anaesthesia. CHO drinkcontained 12.5 g CHO/100 ml. CHO-peptidedrink contained 12.5 g CHO/100 ml and 3.5 ghydrolyzed soy protein/100 ml.

Residual gastric volumes (n¼ 29), muscleglycogen content and glycogen synthaseactivity, voluntary strength, nutritional intake,ambulation, fatigue, anxiety, discomfort,endocrine response

No differences in residual gastric volumebetween groups. Although no differences weredetectable in muscle glycogen content betweenthe groups, postop glycogen synthase activitywas significantly decreased in the controlcompared to the intervention groups. Voluntaryquadriceps muscle strength did not differbetween the groups when analysed per se butpooled analysis of results from the 2intervention groups showed significantly bettermuscle strength 1 month postop compared tocontrol group. No changes in thirst, hunger,anxiety and overall well-being.

S. Awad et al. / Clinical Nutrition 28 (2009) 497–509502

Table 3 (continued )

Study Study groups Variable(s) examined Results

Soop et al.75 Patients undergoing elective hip surgery.Randomisation (double-blinded) to preop CHOdrink (800 ml of 12.5% iso-osmolar drink onevening before operation and 400 ml 3 h beforesurgery) (n¼ 8) or placebo drink (n¼ 6). Insulinsensitivity assessed 1 week preop and day 3postop. Cumulative nitrogen balance calculatedfrom end of surgery to day 3 postop.

Insulin sensitivity 3 days postop, nitrogenbalance in first 3 days postop, ambulation

EGP significantly lower and attenuated aftersurgery in CHO group compared to placebogroup. Relative reduction in GIR on day 3 did notdiffer between groups.Nitrogen looses lower in CHO group thanplacebo group (difference of 25 mg/kg/day) butoverall no difference in nitrogen balance. Nodifferences in postop ambulation.

Bisgaard et al.76 Patients undergoing elective laparoscopiccholecystectomy. Randomisation (double-blinded) to preop CHO drink (800 ml of 12.5%iso-osmolar drink on evening before operationand 400 ml 2 h before surgery) (n¼ 43) orplacebo drink (n¼ 43).

Well-being, appetite, fatigue, pain, nausea andvomiting on day 1 postop, sleep quality andphysical activity

No cases of apparent or suspected pulmonaryaspiration or other drink-related complications.No inter-group differences in scores of well-being, fatigue, appetite, nausea, vomiting,analgesic and antiemetic requirements, sleepquality or activity levels. Pain scores remainedsignificantly raised for 4 days postop in CHOgroup.

Yuill et al.68 Patients undergoing elective major uppergastrointestinal surgery. Randomisation(double-blinded) to preop CHO drink (800 ml of12.6% iso-osmolar drink on evening beforeoperation and 400 ml 2–3 h before anaesthesia)(n¼ 31) or placebo drink (n¼ 34).

Tolerance and effects of preop CHO loading,effects on body composition, length of hospitalstay

No instances of perioperative aspiration. Athospital discharge no differences in changes inendogenous fat reserves but loss of muscle masssignificantly greater in control group. Trendtowards reduced LOS in CHO group.

Svanfeldt et al.77 Healthy volunteers (n¼ 6) underwent 4protocols in a randomised (unblinded)crossover manner. Protocols designed to mimicperioperative situation: control group wasfasted, one CHO group given 800 ml of 12.6%iso-osmolar CHO drink on the evening beforethe day of the clamp study, second CHO groupgiven 400 ml of 12.6% iso-osmolar CHO drink onthe morning of the clamp study, last CHO groupgiven 800 ml of 12.6% iso-osmolar CHO drink onthe evening before and another 400 ml on themorning of the clamp study.

Insulin sensitivity following different CHOregimens

GIR significantly higher when CHO drink wasgiven in morning compared to protocolswithout a morning dose. GIR not affected by thedrink being given in the evening prior to theclamp study. Non-oxidative glucose disposalhigher when drink had been ingested inmorning. Insulin action enhanced by 50% 3 hafter ingestion of a morning dose of the CHOdrink.

Melis et al.78 Patients undergoing elective orthopaedicsurgery. Randomisation (blinded) to 3 groups:control group (n¼ 10) was fasted from midnightand 2 groups (n¼ 10 in each) were given 2different CHO beverages consumed 4 h beforesurgery.

Cellular immune function, fluid homeostasis,thirst, hunger, nausea, anxiety and weakness

No incidents of pulmonary aspiration occurred.HLA-DR expression decreased significantly aftersurgery in control group whereas no changewas observed in CHO groups. Fasted patientsexperienced more thirst preop than CHOgroups. No difference in other measures of well-being.

Breuer et al.79 ASA III–IV patients undergoing elective cardiacsurgery (including type 2 diabetic patients).Randomised to 3 groups: preop CHO drink(n¼ 56) versus placebo drink (n¼ 60) (double-blind) versus open-labelled control group(n¼ 44) fasted from midnight. 800 ml of CHOdrink (12.5% CHO, iso-osmolar) or placebo takenevening before surgery and 400 ml taken 2 hbefore surgery. Insulin requirement chosen assurrogate marker to estimate postop insulinresistance. Morbidity was measured by organdysfunction.

Postop insulin requirement, residual gastricvolume, preop discomfort, drink-relatedcomplications, morbidity

No differences in insulin requirements orresidual gastric volumes between the differentgroups. No drink-related complications. CHOgroup experienced less thirst than control groupbut no differences in hunger, nausea, anxietyand dryness of mouth. CHO and placebo groupsdid not differ in thirst. CHO group required lessintraoperative inotropes after initiation of CPBweaning. No differences in severity of illnessscores, incidence of postop complications anddurations of hospital/intensive care stay.

Svanfeldt et al.80 Patients undergoing elective colorectal surgery.Randomisation to receive a preop drink witheither high (125 mg/ml, n¼ 6) or low (25 mg/ml, n¼ 6) CHO load. 800 ml were ingestedevening before surgery and 600–800 ml weretaken until 2 h before the estimated time forpre-medication. Insulin sensitivity assessed 5days preop and on day 1 postop.

Postop whole-body protein and glucose kinetics Postop whole-body protein breakdown higherat baseline and during insulin stimulation inlow CHO group. Surgery evoked more negativeprotein balance in low CHO group at baselineand during insulin stimulation. No effect of CHOon postop insulin sensitivity seen. Postopsuppression of EGR less effective in low CHOgroup. Positive correlation between EGR andwhole-body protein breakdown seen aftersurgery (r2¼ 0.432, p¼ 0.02).

S. Awad et al. / Clinical Nutrition 28 (2009) 497–509 503

insulin sensitivity and oxidative stress have already been describedbut the cellular mechanisms that link these processes remain to beelucidated.40 The reduction in perioperative insulin sensitivity innon-diabetic subjects is similar to that seen in patients with type 2diabetes.27 Numerous studies on patients with type 2 diabetes havesuggested a link between impaired mitochondrial function and thedevelopment of insulin resistance.12,83 Thus, it may be possible thatpreoperative carbohydrate loading attenuates the decrease ininsulin sensitivity by preventing mitochondrial dysfunction.

5. Mitochondrial function

Mitochondria produce ATP, serve as biosensors for oxidativestress, and through apoptosis, are effector organelles for celldeath.84 About 98% of inspired oxygen is consumed by mitochon-dria85 and the supply of substrates for mitochondrial oxidation isthe primary aim of ordered food intake, digestion and processing.85

Disorders in mitochondrial function have been implicated in thepathogenesis of a number of disease states such as insulin

S. Awad et al. / Clinical Nutrition 28 (2009) 497–509504

resistance which in turn leads to a number of cardiovasculardiseases.36,83 Furthermore, mitochondrial dysfunction may lead toincreased oxidative stress thus placing many surgical patients atincreased perioperative risk.84,86

5.1. Mitochondrial function in health

Mitochondria produce the ATP needed for normal cellularfunction and metabolic homeostasis by oxidative phosphoryla-tion.87 This process is conducted by a series of five enzymecomplexes located on the inner mitochondrial membrane (Fig. 2).Four of these complexes comprise the mitochondrial electrontransport chain (ETC) and function as a biochemical ‘‘conveyer belt’’for electrons. These electrons are derived from oxidation of glucoseor fatty acids and are transferred through mitochondrial membranecomplexes (MMCs) I–IV via mobile electron carriers.

At complexes I, III and IV protons are pumped out of the mito-chondrial matrix into the intermembrane space. This action resultsin the generation of an electrochemical proton gradient which isused by the fifth enzyme complex (ATP synthase) to drive ATPsynthesis. The ATP produced serves as the ‘‘currency’’ needed formost energy-requiring biological transactions.

5.2. Oxidative stress

Mitochondrial oxidative phosphorylation is the major intracel-lular source of reactive oxygen species (ROS) or ‘free radicals’ suchas superoxide, peroxide or hydroxyl radicals. These ROS aregenerated as by-products of the interaction between free electronsand oxygen84 and are an unavoidable consequence of aerobicmetabolism. Free radicals are highly reactive molecules that candegrade or destroy mitochondrial enzyme complexes, membranesand structural components of cellular microarchitecture, either bydirect contact or through lipid peroxidation.84,88 Intrinsic defensesystems that include superoxide dismutase, catalase, glutathioneperoxidase, copper, zinc superoxide dismutase and manganesesuperoxide dismutase protect against ROS-induced damage byconverting free radicals into oxygen and water.84,88 Although theseendogenous antioxidant defense systems effectively suppress ROS

Fig. 2. Schematic representation of the components needed for mitochondrialoxidative phosphorylation. The electron transport chain is located within the innermitochondrial membrane and is comprised of the oxidase complexes I–IV, coenzyme Q(Co Q) and cytochrome c (Cyt C). Red arrows are indicative of the pathway of electronflow. Complexes I, III and IV pump hydrogen ions (dotted red arrows) into the inter-membrane space and generate the electrochemical gradient that powers the phos-phorylation of adenosine diphosphate (ADP) to adenosine triphosphate (ATP) by ATPsynthase. H2O, water; NAD, nicotinamide adenine dinucleotide; NADH, reduced nico-tinamide adenine dinucleotide; O2, oxygen. (For interpretation of the references tocolour in this figure legend, the reader is referred to the web version of this article.)

concentrations within the cell and mitochondria in health, thesemechanisms become inadequate in many disease states. Rapid oroverwhelming increases in ROS produce oxidative stress which canbe cytotoxic via initiation of apoptosis or disruption of intracellularcalcium regulation.84

5.3. Mitochondrial dysfunction and insulin resistance

The cellular and molecular mechanisms of insulin resistancehave been studied extensively with the aim of providing newtherapeutic targets for the treatment and prevention of type 2diabetes.89 Studies on patients with type 2 diabetes have demon-strated glucose transport as the rate-limiting step for insulin-stimulated muscle glycogen synthesis. A reduction in the latter wasthe major factor responsible for insulin resistance in thesepatients.89 These findings were also demonstrated in healthyinsulin-resistant offspring of parents with type 2 diabetes sug-gesting that reduced insulin-stimulated glucose transport was anearly event in the pathogenesis of the disease.89 Subsequently,intramyocellular lipid concentrations, assessed by 1H magneticresonance spectroscopy (MRS), were found to correlate inverselywith muscle insulin sensitivity.90 Accumulation of intramyocellularlipid metabolites such as fatty acyl CoAs and diacylglycerol resultedfrom increased delivery of fatty acids from plasma and/or reducedmitochondrial b-oxidation. These metabolites activate serine/threonine kinases such as protein kinase C which phosphorylatethe serine residues of IRS-1 leading to defects in insulin signalling,failure of GLUT-4 translocation and reduced insulin-stimulatedglucose uptake.89 Activation of protein kinase C in hepatocytessimilarly gives rise to fat-induced defects in insulin signallingwhich result in reduced insulin stimulation of glycogen synthesisand increased hepatic gluconeogenesis.89

Further studies have linked defective mitochondrial functionwith intramyocellular accumulation of lipid metabolites andsubsequent insulin resistance. Petersen et al.83 matched healthylean elderly volunteers for body mass index and activity withyounger subjects. Compared to younger controls, elderly subjectshad higher concentrations of plasma glucose and insulin, a trendtowards increased plasma fatty acid concentrations and a 40%lower rate of insulin-stimulated peripheral glucose uptake. Thiswas associated with a 45% and 225% increase in triglyceride contentof muscle and liver, respectively, as assessed by 1H MRS. In addition,elderly subjects had a 40% reduction in the rates of muscle mito-chondrial oxidative and phosphorylation activity, assessed by 13Cand 31P MRS, respectively. These findings suggest that acquired lossof mitochondrial function associated with ageing may predisposeto intramyocellular lipid accumulation, which results in insulinresistance through the mechanisms described earlier.89 Similarfindings were noted in young lean insulin-resistant offspring ofparents with type 2 diabetes where severe defects in insulin-stimulated muscle glucose metabolism were associated with an80% increase in intramyocellular lipid content and a 30% reductionin rates of mitochondrial ATP production.91

However, other investigators have suggested that mitochondrialabnormalities represent the end result of insulin resistance ratherthan its cause.92 In this context, reduced insulin action in states ofinsulin resistance is thought to cause impaired mitochondrialprotein synthesis and defective mitochondrial biogenesis (theprocess by which mitochondria increase their ability to make ATPby synthesizing additional enzyme complexes).92,93 Inducing aninsulin deficient state in seven type 1 diabetic patients resulted insignificantly reduced muscle mitochondrial ATP productioncapacity which was accompanied by an increase in whole-bodyoxygen consumption and alterations in transcript levels of genesinvolved in oxidative phosphorylation.94 Thus reduced insulin

S. Awad et al. / Clinical Nutrition 28 (2009) 497–509 505

action and the associated metabolic changes can down regulatemuscle oxidative phosphorylation. Another theory is that mito-chondrial dysfunction in insulin resistance, ageing and type 2 dia-betes is a reflection of cumulative oxidative stress that involvesdamage to mitochondrial DNA.92,95 Finally, a study on the effects ofintensive insulin therapy on mitochondrial integrity and function incritically ill surgical patients demonstrated that hyperglycaemiahad a toxic effect on liver, but not muscle, mitochondria.36 Althougha limitation of the study was that these analyses were performedonly in mitochondria of non-survivors, abnormalities in mito-chondrial ultrastructure were seen in 9% of patients that receivedintensive insulin therapy versus 78% of patients that receivedconventional therapy. Furthermore, the intensive insulin therapygroup had 89% and 40% higher median activity of MMCs I and IV,respectively. The discrepancy in effects on liver and muscle mito-chondria was hypothesized to be due to different mechanisms ofglucose uptake in these organs and suggested that it was hyper-glycaemia, rather than a direct effect of insulin, that led to theaforementioned effects.36

5.4. Mitochondrial function in starvation

Although a number of investigators have examined the effectsof starvation on animal mitochondrial function, this has beenwithin the context of organ transplantation and the associatedischaemia-reperfusion injury. Using a murine model, Jung andHenke96 demonstrated impaired liver mitochondrial respirationfollowing 4 days of starvation although this was not associatedwith decreased energy production (quantified by mitochondrialcontent of adenine nucleotide).

An elegant study on the effects of nutritional status on rat livermitochondria demonstrated interesting findings in the controlgroup.97 Mitochondria isolated from rats with normal livers thatwere fasted for 18 h had greater levels of oxidised lipids and lowercontent of the ATP synthase complex than their fed counterparts.These changes suggested that starvation directly caused oxidativeinjury and decreased mitochondrial ATP synthetic capacity. Adecrease in mitochondrial ATPase activity, associated with a trendtowards lower ATP content, was noted in another study of ortho-topic pig liver transplants following 5 days of starvation.98

However, in this small study (n¼ 5 in each group) the control groupwas starved for 24 h thus the effects of feeding, as opposed tostarvation, on mitochondrial function remain unclear.

Domenicali et al.99 studied the effects of starvation on oxidativebalance in mitochondria isolated from rat livers in a model ofpartial hepatic ischaemia-reperfusion injury. Under baselineconditions, 18 and 36 h periods of starvation progressively depletedmitochondrial antioxidant stores (glutathione) leading to increasedlipid peroxidation. As glutathione stores function as cysteinereservoirs, the former are depleted during food deprivation.100

Starvation also exacerbated mitochondrial oxidative damageassociated with warm ischaemia-reperfusion injury and this effectwas dependent on the duration of food deprivation.99 Finally,transmission electron microscopy demonstrated mitochondrialswelling and ultrastructural abnormalities during post-ischaemicreperfusion in starved rats thus providing indirect evidence ofmembrane injury and mitochondrial dysfunction.99 A studydesigned to examine the effects of starvation on oxidative stress inrat liver mitochondria reported similar findings.101 Compared withovernight starvation in the control group, mitochondria of ratsstarved for 72 h showed significantly increased oxidative andlipoxidative protein damage (levels of N-malondialdehyde lysine,aminoadipic semialdehyde and glutamic semialdehyde inmitochondrial proteins of rat liver). Starvation also modifiedthe fatty acid composition of mitochondrial membranes. The

aforementioned findings suggest that starvation per se adverselyaffects mitochondrial function through increased oxidative stressand generation of ROS. These in turn alter the lipid and proteincomposition of mitochondria leading to structural99 and functionalabnormalities.101

Briet and Jeejeebhoy102 studied the effects of reduced energyintake on the activities of muscle MMCs I–IV and peripheral bloodmononuclear cells (PBMC) MMC I in rats fed enterally. Comparedwith normally fed rats, one week of protein-energy deprivationsignificantly depressed the activities of complexes I (�73%), II(�68%) and III (�92%) in the mitochondria of soleus muscle andcomplex I (�74%) in the mitochondria of PBMC significantly. Animportant finding in this study was that the activity of MMC I inPBMCs was found to correlate with the activity of MMC I in muscle(r2¼ 0.66). The reduction in mitochondrial enzyme activitiesduring protein-energy restriction was hypothesized102 to explainfindings of decreased oxygen uptake, substrate oxidation and ADPphosphorylation noted in previous studies of the effects of reducedenergy intake on oxidative phosphorylation by liver and musclemitochondria.103,104 This hypothesis was subsequently proven ina study of the effects of reduced energy intake on mitochondrialfunction, where decreased activity of complexes I (approximately�57%) and III (w�51%) correlated with a reduction (w�57%) inoxidative phosphorylation rate in rat muscle mitochondria.105

Studies of human PBMC MMC activity similarly demonstrateddepressed complex I activity in malnourished patients compared tocontrols.106,107 The effects of starvation on mitochondrial oxidativephosphorylation remain unclear, however, as the aforementionedstudies were performed in animals and malnourished patients whowere receiving some energy, although not enough to meet dailyrequirements.

There are no reported human studies on the effects ofpreoperative starvation on perioperative mitochondrial function,the majority being studies on mitochondrial function within thecontext of insulin-resistant states such as type 2 diabetes andobesity. However, given that starvation induces a state of insulinresistance, valuable insights into the effects of starvation may begained from these studies. Starvation3,51 and other states ofinsulin resistance108 are associated with elevation in the plasmaconcentration of NEFAs. Elevated NEFA concentrations result inincreased fatty acid flux through skeletal muscle mitochondriawhich results in reduced mitochondrial oxidative phosphoryla-tion,109 increased ROS generation88 and subsequent mitochondrialdysfunction.109 In a study that examined the effects of eatinga high fat diet in humans and mice, the increased flux of fattyacids through muscle was found to down regulate genes thatencoded mitochondrial proteins and transcription factors involvedin mitochondrial biogenesis.109 Interestingly, a similar pattern ofgene expression to that seen after intake of a high fat diet wasinduced by short-term fasting in another study.109,110 Further-more, these changes became more pronounced with increasinglength of food deprivation. These results may explain the signif-icant starvation-induced decrease in mitochondrial proteincontent noted in earlier studies.111

Recent studies have also characterized the effects of insulin onskeletal muscle mitochondrial function.112 Insulin was found tostimulate mitochondrial protein synthesis,113 activate mitochon-drial enzyme activity113 and stimulate oxidative phosphorylationby increasing mitochondrial mRNA transcript expression, proteinsynthesis, activities of cytochrome C oxidase and citrate synthase,and ATP production.93 Given these findings, it may be hypothesizedthat a reduction in insulin concentration and an increase in NEFAconcentration, as seen in starvation,51 lead to reduced mitochon-drial function and oxidative capacity (Fig. 3). This in turn maycontribute to a diminished ability to oxidize fatty acids114 leading to

Fig. 3. Potential mechanisms that may link starvation and mitochondrial function to the development of insulin resistance. FA – fatty acid, NEFAs – non-esterified fatty acids,GLUT-4 – the facilitative glucose transporter 4.

S. Awad et al. / Clinical Nutrition 28 (2009) 497–509506

intramyocellular fat accumulation115 which, as previously dis-cussed, has been linked to the development of skeletal muscleinsulin resistance.112

5.5. Perioperative mitochondrial dysfunction

In a study that examined the effects of surgery and anaesthesiaon mitochondrial function in lymphocytes from 16 patientsundergoing major abdominal surgery, Delogu et al.86 demonstrateda significant increase in mitochondrial reactive oxygen species(ROS) production coupled with a significant depletion (�24%) ofmitochondrial antioxidant (glutathione) stores. These changes inmitochondrial function, although transient, were associated witha significant increase in the rate of apoptosis amongst CD4þ andCD8þ lymphocytes. Similar findings were noted by the same groupin a study on perioperative polymorphonuclear neutrophil mito-chondrial function.116 These studies suggest that an increase inmitochondrial oxidative stress during the perioperative period mayplay a role in mediating the immune suppression that is seen aftersurgery.86 Although mitochondrial dysfunction in this setting mayincrease the risks of infective complications,86 the contribution ofvarious confounding factors such as starvation, insulin resistance95

and anaesthesia117 to this impairment in function was, unfortu-nately, not studied.

5.6. Effects of carbohydrate loading on mitochondrial function

The effects of preoperative carbohydrate loading using an oralbeverage on mitochondrial function have not been studied. Hay-akawa et al.118 examined the effects of intraoperative glucoseinfusions on hepatic mitochondrial energy status (redox state) in 26patients undergoing elective total gastrectomy for cancer. Althoughtheir study was weakened because the control group also receivedintraoperative glucose infusion and the authors used an indirectmeasure of hepatic mitochondrial energy status (the arterial ketone

body ratio), they appeared to show that low mitochondrial energystatus accompanied preoperative starvation and that this may beimproved with intraoperative glucose infusion. Animal102 andhuman106,107 studies on the effects of energy deprivation andmalnutrition on MMC activity have demonstrated that short-termrefeeding (1 day in rats, 7 days in humans) restored mitochondrialactivity. Although a 7-day period of refeeding increased humanPBMC MMC I activity,106 normal levels were only attained after onemonth of refeeding.107 The heterogeneity of conditions causingmalnutrition in these studies makes it difficult, however, to deriveany conclusions regarding the possible protective effects ofpreoperative carbohydrate loading in healthy patients undergoingelective surgery.

6. Conclusions

The mechanisms that underlie the development of perioper-ative insulin resistance during starvation and its attenuation bypreoperative carbohydrate drinks are yet to be defined. Animalstudies have shown that energy deprivation has adverse effects onmitochondrial function by decreasing mitochondrial ATP synthesiscapacity and complex activity, and increasing oxidative injury.Furthermore, evidence from human studies suggests that thedevelopment of insulin resistance during starvation may be linkedto impaired mitochondrial function. Future studies should inves-tigate whether mitochondrial dysfunction underlies the develop-ment of insulin resistance in healthy patients undergoing electivesurgery.

Conflict of Interest

SA has received research funding from Fresenius Kabi andeducational support from Nutricia Clinical Care. DNL has receivedresearch funding, speaker’s honoraria and travel expenses fromFresenius Kabi and Nutricia Clinical Care. DC-T and IAM have noconflicts of interests to declare.

S. Awad et al. / Clinical Nutrition 28 (2009) 497–509 507

Acknowledgements

Funding: SA was supported by the Enhanced Recovery AfterSurgery Group (via an unrestricted educational grant from Frese-nius Kabi, Bad Homburg, Germany), a surgical research fellowshipfrom the Royal College of Surgeons of England and funding from theDoris Mary Sheppard and Rosetrees Trusts.

Author contributions

SA: Structure of review, literature search, review of literature,writing of manuscript, final approval.

DC-T: Structure of review, critical appraisal of literature, writingof manuscript, final approval.

IAM: Structure of review, critical appraisal of literature, criticalrevision, final approval.

DNL: Structure of review, critical appraisal of literature, criticalrevision, final approval.

References

1. Krentz AJ. Insulin resistance: a clinical handbook. Oxford: Blackwell Science;2002.

2. Allison SP, Go VLW. Metabolic issues of clinical nutrition. In: Nestle nutritionworkshop series clinical & performance program, vol. 9. Basel: Karger/NestleNutrition; 2004.

3. Frayn KN. Metabolic regulation: a human perspective. Oxford: BlackwellScience; 2003.

4. Rothman DL, Magnusson I, Katz LD, Shulman RG, Shulman GI. Quantitation ofhepatic glycogenolysis and gluconeogenesis in fasting humans with 13C NMR.Science 1991;254:573–6.

5. Hill GL. Disorders of nutrition and metabolism in clinical surgery: understandingand management. Edinburgh: Churchill Livingstone; 1992.

6. Lebovitz HE. Clinician’s manual on insulin resistance. London: Science Press;2002.

7. Black PR, Brooks DC, Bessey PQ, Wolfe RR, Wilmore DW. Mechanisms ofinsulin resistance following injury. Ann Surg 1982;196:420–35.

8. Wolfe RR, Durkot MJ, Allsop JR, Burke JF. Glucose metabolism in severelyburned patients. Metab Clin Exp 1979;28:1031–9.

9. Nygren J. The metabolic effects of fasting and surgery. Best Pract Res ClinAnaesthesiol 2006;20:429–38.

10. DeFronzo RA. Pathogenesis of type 2 diabetes: metabolic and molecularimplications for identifying diabetes genes. Diabetes Rev 1997;5:177–269.

11. Shulman GI. Cellular mechanisms of insulin resistance in humans. Am J Cardiol1999;84:3J–10J.

12. Petersen KF, Shulman GI. Etiology of insulin resistance. Am J Med2006;119:S10–6.

13. Nygren J, Thorell A, Brismar K, Karpe F, Ljungqvist O. Short-term hypocaloricnutrition but not bed rest decrease insulin sensitivity and IGF-I bioavailabilityin healthy subjects: the importance of glucagon. Nutrition 1997;13:945–51.

14. Newman WP, Brodows RG. Insulin action during acute starvation: evidence forselective insulin resistance in normal man. Metab Clin Exp 1983;32:590–6.

15. Svanfeldt M, Thorell A, Brismar K, Nygren J, Ljungqvist O. Effects of 3 days of‘‘postoperative’’ low caloric feeding with or without bed rest on insulinsensitivity in healthy subjects. Clin Nutr 2003;22:31–8.

16. Nygren J, Thorell A, Efendic S, Nair KS, Ljungqvist O. Site of insulin resistanceafter surgery: the contribution of hypocaloric nutrition and bed rest. Clin Sci(Lond) 1997;93:137–46.

17. Mansell PI, Macdonald IA. The effect of starvation on insulin-induced glucosedisposal and thermogenesis in humans. Metabolism 1990;39:502–10.

18. Little RA, Henderson A, Frayn KN, Galasko CS, White RH. The disposal ofintravenous glucose studied using glucose and insulin clamp techniques insepsis and trauma in man. Acta Anaesthesiol Belg 1987;38:275–9.

19. Carlson GL. Hunterian lecture: insulin resistance in human sepsis: implica-tions for the nutritional and metabolic care of the critically ill surgical patient.Ann R Coll Surg Engl 2004;86:75–81.

20. Thorell A, Efendic S, Gutniak M, Haggmark M, Ljungqvist O. Insulin resistanceafter abdominal surgery. Br J Surg 1994;81:59–63.

21. Nygren JO, Thorell A, Soop M, Efendic S, Brismar K, Karpe F, et al. Perioperativeinsulin and glucose infusion maintains normal insulin sensitivity after surgery.Am J Physiol Endocrinol Metab 1998;275:E140–8.

22. Nygren J, Soop M, Thorell A, Efendic S, Nair KS, Ljungqvist O. Preoperative oralcarbohydrate administration reduces postoperative insulin resistance. ClinNutr 1998;17:65–71.

23. Soop M, Nygren J, Myrenfors P, Thorell A, Ljungqvist O. Preoperative oralcarbohydrate treatment attenuates immediate postoperative insulin resis-tance. Am J Physiol Endocrinol Metab 2001;280:E576–83.

24. Ljungqvist O, Thorell A, Gutniak M, Haggmark T, Efendic S. Glucose infusioninstead of preoperative fasting reduces postoperative insulin resistance. J AmColl Surg 1994;178:329–36.

25. Brandi LS, Frediani M, Oleggini M, Mosca F, Cerri M, Boni C, et al. Insulinresistance after surgery: normalization by insulin treatment. Clin Sci (Lond)1990;79:443–50.

26. Nordenstrom J, Sonnenfeld T, Arner P. Characterization of insulin resistanceafter surgery. Surgery 1989;105:28–35.

27. Ljungqvist O, Nygren J, Thorell A. Modulation of post-operative insulin resis-tance by pre-operative carbohydrate loading. Proc Nutr Soc 2002;61:329–36.

28. Thorell A, Alston-Smith J, Ljungqvist O. The effect of preoperative carbohy-drate loading on hormonal changes, hepatic glycogen, and glucoregulatoryenzymes during abdominal surgery. Nutrition 1996;12:690–5.

29. Brandi LS, Santoro D, Natali A, Altomonte F, Baldi S, Frascerra S, et al. Insulinresistance of stress: sites and mechanisms. Clin Sci (Lond) 1993;85:525–35.

30. Thorell A, Nygren J, Hirshman MF, Hayashi T, Nair KS, Horton ES, et al. Surgery-induced insulin resistance in human patients: relation to glucose transportand utilization. Am J Physiol Endocrinol Metab 1999;276:E754–61.

31. Thorell A, Efendic S, Gutniak M, Haggmark T, Ljungqvist O. Development ofpostoperative insulin resistance is associated with the magnitude of operation.Eur J Surg 1993;159:593–9.

32. Thorell A, Nygren J, Ljungqvist O. Insulin resistance: a marker of surgicalstress. Curr Opin Clin Nutr Metab Care 1999;2:69–78.

33. Van den Berghe G, Wouters PJ, Bouillon R, Weekers F, Verwaest C, Schetz M,et al. Outcome benefit of intensive insulin therapy in the critically ill: insulindose versus glycemic control. Crit Care Med 2003;31:359–66.

34. van den Berghe G, Wouters P, Weekers F, Verwaest C, Bruyninckx F, Schetz M,et al. Intensive insulin therapy in the critically ill patients. N Engl J Med2001;345:1359–67.

35. Vanhorebeek I, Langouche L, Van den Berghe G. Tight blood glucose control:what is the evidence? Crit Care Med 2007;35(9 Suppl.):S496–502.

36. Vanhorebeek I, De Vos R, Mesotten D, Wouters PJ, De Wolf-Peeters C, Van denBerghe G. Protection of hepatocyte mitochondrial ultrastructure and functionby strict blood glucose control with insulin in critically ill patients. Lancet2005;365:53–9.

37. Langouche L, Vanhorebeek I, Vlasselaers D, Vander Perre S, Wouters PJ,Skogstrand K, et al. Intensive insulin therapy protects the endothelium of criti-cally ill patients. J Clin Invest 2005;115:2277–86.

38. Weekers F, Giulietti AP, Michalaki M, Coopmans W, Van Herck E, Mathieu C,et al. Metabolic, endocrine, and immune effects of stress hyperglycemia ina rabbit model of prolonged critical illness. Endocrinology 2003;144:5329–38.

39. Maltby JR. Fasting from midnight – the history behind the dogma. Best PractRes Clin Anaesthesiol 2006;20:363–78.

40. Diks J, van Hoorn DE, Nijveldt RJ, Boelens PG, Hofman Z, Bouritius H, et al.Preoperative fasting: an outdated concept? J Parenter Enteral Nutr2005;29:298–304.

41. Ljungqvist O, Jansson E, Ware J. Effect of food deprivation on survival afterhemorrhage in the rat. Circ Shock 1987;22:251–60.

42. Alibegovic A, Ljungqvist O. Pretreatment with glucose infusion prevents fataloutcome after hemorrhage in food deprived rats. Circ Shock 1993;39:1–6.

43. Esahili AH, Boija PO, Ljungqvist O, Rubio C, Ware J. Twenty-four hour fastingincreases endotoxin lethality in the rat. Eur J Surg 1991;157:89–95.

44. Ljungqvist O, Boija PO, Esahili H, Larsson M, Ware J. Food deprivation altersliver glycogen metabolism and endocrine responses to hemorrhage. Am JPhysiol 1990;259:E692–8.

45. Fryburg DA, Barrett EJ, Louard RJ, Gelfand RA. Effect of starvation on humanmuscle protein metabolism and its response to insulin. Am J Physiol1990;259:E477–82.

46. Bark T, Katouli M, Svenberg T, Ljungqvist O. Food deprivation increases bacterialtranslocation after non-lethal haemorrhage in rats. Eur J Surg 1995;161:67–71.

47. Nettelbladt CG, Katouli M, Volpe A, Bark T, Muratov V, Svenberg T, et al.Starvation increases the number of coliform bacteria in the caecum andinduces bacterial adherence to caecal epithelium in rats. Eur J Surg1997;163:135–42.

48. Ljungqvist O, Efendic S, Eneroth P, Hamberger B, Nylander G, Ware J. Nutri-tional status and endocrine response to hemorrhage. Can J Physiol Pharmacol1986;64:1185–8.

49. van Hoorn DEC, Boelens PG, van Middelaar-Voskuilen MC, Nijveldt RJ, Prins H,Bouritius H, et al. Preoperative feeding preserves heart function and decreasesoxidative injury in rats. Nutrition 2005;21:859–66.

50. van Hoorn EC, van Middelaar-Voskuilen MC, van Limpt CJP, Lamb KJ,Bouritius H, Vriesema AJM, et al. Preoperative supplementation witha carbohydrate mixture decreases organ dysfunction-associated risk factors.Clin Nutr 2005;24:114–23.

51. Tsintzas K, Jewell K, Kamran M, Laithwaite D, Boonsong T, Littlewood J, et al.Differential regulation of metabolic genes in skeletal muscle during starvationand refeeding in humans. J Physiol 2006;575:291–303.

52. Gallen IW, Macdonald IA, Mansell PI. The effect of a 48 h fast on the physio-logical responses to food ingestion in normal-weight women. Br J Nutr1990;63:53–64.

53. Webber J, Taylor J, Greathead H, Dawson J, Buttery PJ, Macdonald IA. Effects offasting on fatty acid kinetics and on the cardiovascular, thermogenic andmetabolic responses to the glucose clamp. Clin Sci (Lond) 1994;87:697–706.

54. Webber J, Macdonald IA. The cardiovascular, metabolic and hormonal changesaccompanying acute starvation in men and women. Br J Nutr 1994;71:437–47.

S. Awad et al. / Clinical Nutrition 28 (2009) 497–509508

55. Samra JS, Clark ML, Humphreys SM, Macdonald IA, Frayn KN. Regulation oflipid metabolism in adipose tissue during early starvation. Am J Physiol1996;271:E541–6.

56. Jensen MD, Haymond MW, Gerich JE, Cryer PE, Miles JM. Lipolysis duringfasting. Decreased suppression by insulin and increased stimulation byepinephrine. J Clin Invest 1987;79:207–13.

57. Fearon KC, Ljungqvist O, Von Meyenfeldt M, Revhaug A, Dejong CH, Lassen K,et al. Enhanced recovery after surgery: a consensus review of clinical care forpatients undergoing colonic resection. Clin Nutr 2005;24:466–77.

58. Ljungqvist O. Preoperative carbohydrates to prepare metabolism for surgery.Acta Anaesthesiol Belg 2004;55(Suppl.):61–3.

59. Soreide E, Stromskag KE, Steen PA. Statistical aspects in studies ofpreoperative fluid intake and gastric content. Acta Anaesthesiol Scand1995;39:738–43.

60. Brady M, Kinn S, Stuart P. Preoperative fasting for adults to prevent peri-operative complications. Cochrane Database Syst Rev 2003;(4):CD004423.

61. Warner MA, Warner ME, Weber JG. Clinical significance of pulmonary aspi-ration during the perioperative period. Anesthesiology 1993;78:56–62.

62. Soreide E, Fasting S, Raeder J. New preoperative fasting guidelines in Norway.Acta Anaesthesiol Scand 1997;41:799.

63. Practice guidelines for preoperative fasting and the use of pharmacologic agentsto reduce the risk of pulmonary aspiration: application to healthy patientsundergoing elective procedures: a report by the American Society of Anesthe-siologist Task Force on Preoperative Fasting. Anesthesiology 1999;90:896–905.

64. Smith AF. Preoperative fasting in adults. In: Hilditch G, editor. Raising thestandard: a compendium of audit recipes: section 1, preoperative care. London:Royal College of Anaesthetists. p. 30–1, http://www.rcoa.ac.uk/docs/ARB-section1.pdf; 2006 [accessed 27.03.09].

65. Ljungqvist O. To fast or not to fast? Metabolic preparation for elective surgery.Scand J Nutr 2004;48:77–82.

66. Moore FD. Metabolic care of the surgical patient. Philadelphia: W.B. Saunders; 1959.67. Friberg B, Heilborn B, Haggmark T, Ljungqvist O. Food deprivation prior to

stress reduces stimulated muscle force both before and after hemorrhage inthe rat. Surg Res Comm 1994;16:131–7.

68. Yuill KA, Richardson RA, Davidson HIM, Garden OJ, Parks RW. The adminis-tration of an oral carbohydrate-containing fluid prior to major elective upper-gastrointestinal surgery preserves skeletal muscle mass postoperatively –a randomised clinical trial. Clin Nutr 2005;24:32–7.

69. Ljungqvist O, Nygren J, Thorell A, Brodin U, Efendic S. Preoperative nutrition –elective surgery in the fed or the overnight fasted state. Clin Nutr2001;20:167–71.

70. Nygren J, Thorell A, Ljungqvist O. Preoperative oral carbohydrate nutrition: anupdate. Curr Opin Clin Nutr Metab Care 2001;4:255–9.

71. Nygren J, Thorell A, Jacobsson H, Larsson S, Schnell PO, Hylen L, et al. Preop-erative gastric emptying. Effects of anxiety and oral carbohydrate adminis-tration. Ann Surg 1995;222:728–34.

72. Nygren J, Soop M, Thorell A, Sree Nair K, Ljungqvist O. Preoperative oralcarbohydrates and postoperative insulin resistance. Clin Nutr 1999;18:117–20.

73. Hausel J, Nygren J, Lagerkranser M, Hellstrom PM, Hammarqvist F,Almstrom C, et al. A carbohydrate-rich drink reduces preoperative discomfortin elective surgery patients. Anesth Analg 2001;93:1344–50.

74. Henriksen MG, Hessov I, Dela F, Vind H, Haraldsted V, Rodt SO. Effects ofpreoperative oral carbohydrates and peptides on postoperative endocrineresponse, mobilization, nutrition and muscle function in abdominal surgery.Acta Anaesthesiol Scand 2003;47:191–9.

75. Soop M, Nygren J, Thorell A, Weidenhielm L, Lundberg M, Hammarqvist F,et al. Preoperative oral carbohydrate treatment attenuates endogenousglucose release 3 days after surgery. Clin Nutr 2004;23:733–41.

76. Bisgaard T, Kristiansen VB, Hjortso NC, Jacobsen LS, Rosenberg J, Kehlet H.Randomized clinical trial comparing an oral carbohydrate beverage withplacebo before laparoscopic cholecystectomy. Br J Surg 2004;91:151–8.

77. Svanfeldt M, Thorell A, Hausel J, Soop M, Nygren J, Ljungqvist O. Effect of‘‘preoperative’’ oral carbohydrate treatment on insulin action – a randomisedcross-over unblinded study in healthy subjects. Clin Nutr 2005;24:815–21.

78. Melis GC, van Leeuwen PAM, von Blomberg-van der Flier BME,Goedhart-Hiddinga AC, Uitdehaag BMJ, Strack van Schijndel RJM, et al. Acarbohydrate-rich beverage prior to surgery prevents surgery-inducedimmunodepression: a randomized, controlled, clinical trial. J Parenter EnteralNutr 2006;30:21–6.

79. Breuer J-P, von Dossow V, von Heymann C, Griesbach M, von Schickfus M,Mackh E, et al. Preoperative oral carbohydrate administration to ASA III–IVpatients undergoing elective cardiac surgery. Anesth Analg 2006;103:1099–108.

80. Svanfeldt M, Thorell A, Hausel J, Soop M, Rooyackers O, Nygren J, et al.Randomized clinical trial of the effect of preoperative oral carbohydratetreatment on postoperative whole-body protein and glucose kinetics. Br J Surg2007;94:1342–50.

81. Ljungqvist O, Soreide E. Preoperative fasting. Br J Surg 2003;90:400–6.82. Hausel J, Nygren J, Thorell A, Lagerkranser M, Ljungqvist O. Randomized

clinical trial of the effects of oral preoperative carbohydrates on postoperativenausea and vomiting after laparoscopic cholecystectomy. Br J Surg2005;92:415–21.

83. Petersen KF, Befroy D, Dufour S, Dziura J, Ariyan C, Rothman DL, et al. Mito-chondrial dysfunction in the elderly: possible role in insulin resistance.Science 2003;300:1140–2.

84. Muravchick S, Levy RJ. Clinical implications of mitochondrial dysfunction.Anesthesiology 2006;105:819–37.

85. Duchen MR. Roles of mitochondria in health and disease. Diabetes2004;53:S96–102.

86. Delogu G, Moretti S, Famularo G, Marcellini S, Santini G, Antonucci A, et al.Mitochondrial perturbations and oxidant stress in lymphocytes from patientsundergoing surgery and general anesthesia. Arch Surg 2001;136:1190–6.

87. Nelson DL, Lehninger AL, Cox MM. Lehninger principles of biochemistry. NewYork: W.H. Freeman; 2008.

88. Jackson MJ, Papa S, Bolanos J, Bruckdorfer R, Carlsen H, Elliott RM, et al.Antioxidants, reactive oxygen and nitrogen species, gene induction andmitochondrial function. Mol Aspects Med 2002;23:209–85.

89. Morino K, Petersen KF, Shulman GI. Molecular mechanisms of insulin resis-tance in humans and their potential links with mitochondrial dysfunction.Diabetes 2006;55:S9–15.

90. Krssak M, Falk Petersen K, Dresner A, DiPietro L, Vogel SM, Rothman DL, et al.Intramyocellular lipid concentrations are correlated with insulin sensitivity inhumans: a 1H NMR spectroscopy study. Diabetologia 1999;42:113–6.

91. Petersen KF, Dufour S, Befroy D, Garcia R, Shulman GI. Impaired mitochondrialactivity in the insulin-resistant offspring of patients with type 2 diabetes. NEngl J Med 2004;350:664–71.

92. Toledo FGS, Kelley DE. Mitochondrial dysfunction in the pathogenesis ofinsulin resistance associated with obesity, diabetes, and aging. Curr OpinEndocrinol Diabetes Obes 2005;12:157–62.

93. Stump CS, Short KR, Bigelow ML, Schimke JM, Nair KS. Effect of insulin onhuman skeletal muscle mitochondrial ATP production, protein synthesis, andmRNA transcripts. Proc Natl Acad Sci U S A 2003;100:7996–8001.

94. Karakelides H, Asmann YW, Bigelow ML, Short KR, Dhatariya K, Coenen-Schimke J, et al. Effect of insulin deprivation on muscle mitochondrial ATPproduction and gene transcript levels in type 1 diabetic subjects. Diabetes2007;56:2683–9.

95. Ritz P, Berrut G. Mitochondrial function, energy expenditure, aging and insulinresistance. Diabetes Metab 2005;31(Spec no. 2). 5S67–5S73.

96. Jung K, Henke W. The effect of starvation on mitochondrial functions in the ratliver and kidney. Res Surg 1993;5:103–5.

97. Vendemiale G, Grattagliano I, Caraceni P, Caraccio G, Domenicali M,Dall’Agata M, et al. Mitochondrial oxidative injury and energy metabolismalteration in rat fatty liver: effect of the nutritional status. Hepatology2001;33:808–15.

98. Fukumori T, Ohkohchi N, Tsukamoto S, Asakura T, Oikawa K, Takayama J, et al.Long-term fasting of donors deteriorates mitochondrial adenosine triphos-phate synthesis in liver grafts during cold preservation. Transplant Proc1997;29:3360–1.

99. Domenicali M, Caraceni P, Vendemiale G, Grattagliano I, Nardo B,Dall’Agata M, et al. Food deprivation exacerbates mitochondrial oxidativestress in rat liver exposed to ischemia-reperfusion injury. J Nutr2001;131:105–10.

100. Cho ES, Sahyoun N, Stegink LD. Tissue glutathione as a cysteine reservoirduring fasting and refeeding of rats. J Nutr 1981;111:914–22.

101. Sorensen M, Sanz A, Gomez J, Pamplona R, Portero-Otin M, Gredilla R, et al.Effects of fasting on oxidative stress in rat liver mitochondria. Free Radic Res2006;40:339–47.

102. Briet F, Jeejeebhoy KN. Effect of hypoenergetic feeding and refeeding onmuscle and mononuclear cell activities of mitochondrial complexes I–IV inenterally fed rats. Am J Clin Nutr 2001;73:975–83.

103. Ardawi MS, Majzoub MF, Masoud IM, Newsholme EA. Enzymic and metabolicadaptations in the gastrocnemius, plantaris and soleus muscles of hypocaloricrats. Biochem J 1989;261:219–25.

104. Ferreira J, Gil L. Nutritional effects on mitochondrial bioenergetics.Alterations in oxidative phosphorylation by rat liver mitochondria. Biochem J1984;218:61–7.

105. Madapallimattam AG, Law L, Jeejeebhoy KN. Effect of hypoenergetic feedingon muscle oxidative phosphorylation and mitochondrial complex I–IV activ-ities in rats. Am J Clin Nutr 2002;76:1031–9.

106. Briet F, Twomey C, Jeejeebhoy KN. Effect of malnutrition and short-termrefeeding on peripheral blood mononuclear cell mitochondrial complex Iactivity in humans. Am J Clin Nutr 2003;77:1304–11.

107. Briet F, Twomey C, Jeejeebhoy KN. Effect of feeding malnourished patients for1 mo on mitochondrial complex I activity and nutritional assessmentmeasurements. Am J Clin Nutr 2004;79:787–94.

108. McGarry JD. Banting lecture 2001: dysregulation of fatty acid metabolism inthe etiology of type 2 diabetes. Diabetes 2002;51:7–18.

109. Sparks LM, Xie H, Koza RA, Mynatt R, Hulver MW, Bray GA, et al. A high-fatdiet coordinately downregulates genes required for mitochondrial oxidativephosphorylation in skeletal muscle. Diabetes 2005;54:1926–33.

110. Jagoe RT, Lecker SH, Gomes M, Goldberg AL, Jagoe RT, Lecker SH, et al. Patternsof gene expression in atrophying skeletal muscles: response to food depri-vation. FASEB J 2002;16:1697–712.

111. Brady LJ, Hoppel CL. Effect of diet and starvation on hepatic mitochondrialfunction in the rat. J Nutr 1983;113:2129–37.

112. Boirie Y. Insulin regulation of mitochondrial proteins and oxidative phos-phorylation in human muscle. Trends Endocrinol Metab 2003;14:393–4.

113. Boirie Y, Short KR, Ahlman B, Charlton M, Nair KS. Tissue-specific regulation ofmitochondrial and cytoplasmic protein synthesis rates by insulin. Diabetes2001;50:2652–8.

S. Awad et al. / Clinical Nutrition 28 (2009) 497–509 509

114. Blaak EE, Hul G, Verdich C, Stich V, Martinez A, Petersen M, et al. Fat oxidationbefore and after a high fat load in the obese insulin-resistant state. J Clin Endo-crinol Metab 2006;91:1462–9.

115. Johnson NA, Stannard SR, Rowlands DS, Chapman PG, Thompson CH,O’Connor H, et al. Effect of short-term starvation versus high-fat diet onintramyocellular triglyceride accumulation and insulin resistance in physicallyfit men. Exp Physiol 2006;91:693–703.

116. Delogu G, Moretti S, Famularo G, Antonucci A, Signore L, Marcellini S, et al.Circulating neutrophils exhibit enhanced apoptosis associated with

mitochondrial dysfunctions after surgery under general anaesthesia. ActaAnaesthesiol Scand 2001;45:87–94.

117. Miro O, Barrientos A, Alonso JR, Casademont J, Jarreta D, Urbano-Marquez A,et al. Effects of general anaesthetic procedures on mitochondrial function ofhuman skeletal muscle. Eur J Clin Pharmacol 1999;55:35–41.

118. Hayakawa J, Motohashi H, Sairenji M, Kobayashi O, Takahashi M, Usuda Y. Theeffects of intraoperative glucose infusion on portal blood insulin concentrationand hepatic mitochondrial redox state during surgery: comparison of short-term and continuous infusions. Surgery 2000;30:228–34.