Received: 5 June 2002Accepted: 20 October 2002Published online: 29 November 2002© Springer-Verlag 2002

Abstract Background: Cardiopul-monary bypass induces a generalizedinflammatory reaction accompaniedby free radical generation. Depletionof antioxidants could result and is re-ported for vitamin E and C. We in-vestigated the effect of cardiopulmo-nary bypass on plasma concentra-tions of α-tocopherol, retinol, andbiochemical variables (e.g., triacyl-glycerol, cholesterol, and C-reactiveprotein). Patients and methods: Plas-ma levels of all parameters were investigated by serial sampling in ten men undergoing elective coro-nary artery bypass grafting. Sampleswere taken before, during, and up to 48 h after bypass to obtain timeprofiles of the laboratory indices.Results: α-Tocopherol and retinoldecreased during cardiopulmonarybypass when not adjusted for con-founders. After adjustment for he-modilution and lipids, no significantchange was noted during bypass.

However, a reduction in retinol was observed 48 h postoperatively.Conclusions: These data indicatethat vitamin E and A analysis to as-certain links to their consumption viathe production of free radicals underconditions of cardiopulmonary by-pass may be inappropriate. Specifi-cally, during bypass a reduction insystemic vitamin E and A seems tobe a response to changes in bloodvolume and liver function.

Keywords Vitamin E · Vitamin A ·Antioxidant vitamins · Cardiopulmonary bypass

Langenbecks Arch Surg (2003) 387:372–378DOI 10.1007/s00423-002-0336-4 O R I G I N A L A RT I C L E

Rainer SchindlerSuliko BerndtPeter SchroederOskar OsterGerhard RaveHans-Hinrich Sievers

Plasma vitamin E and A changes during cardiopulmonary bypassand in the postoperative course

Introduction

Numerous experimental and clinical studies have dem-onstrated that cardiopulmonary bypass (CPB) is associat-ed with oxidative stress [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]. Thegeneration of oxygen-derived free radicals induced byCPB is considered to be responsible for complicationsfollowing cardiac surgery, such as myocardial stunning(a prolonged reduction in cardiac performance). To pre-vent disintegration of host tissues the body has a com-plex antioxidant protective defense system. Vitamin E(mainly α-tocopherol, α-T) and vitamin A (retinol and

its derivatives) are components of this defense network[11, 12, 13]. However, it is still controversial whether in-creases in free radicals have any adverse effects on plas-ma vitamin E status during extracorporeal circulation [1,3, 6, 8, 9]. For example, Sisto et al. [8] reported a signifi-cant decline in its systemic level even after pretreatmentwith vitamin E. Ballmer and colleagues [1] observed thatonly plasma vitamin C levels decreased; vitamin E con-centrations did not change. In a prospective study, Cava-rocchi and coworkers [3] noted that plasma vitamin Elevels fell in proportion to increasing hydrogen peroxidelevels. An inverse association is also observed for sys-

R. Schindler (✉)Department of Human Nutrition and Food Science, University of Kiel,Düsternbrooker Weg 17, Kiel, Germanye-mail: rschindler@nutrfoodsc.uni-kiel.deTel.: +49-431-8805675Fax: +49-431-8805679

S. Berndt · H.-H. SieversDepartment of Cardiac Surgery, Medical University of Lübeck, Lübeck,Germany

P. SchroederDepartment of General and Vascular Surgery, Gilead Hospitals,Bielefeld, Germany

O. OsterDepartment of Pediatrics, University of Kiel, Kiel, Germany

G. RaveDepartment of Variationsstatistik, University of Kiel, Kiel, Germany

temic vitamin E and plasmatic myeloperoxidase [6]. Incontrast, a more recent investigation by Tangney and as-sociates [9] reported an increase in vitamin E during re-perfusion. Finally, myocardial tissue vitamin E concen-trations are reduced with ischemia and/or reperfusionduring coronary artery bypass graft, which also impliesthat at least part of this is the result of disturbances inblood-borne vitamin E [2, 5, 10]. The contrary observa-tions made on levels of antioxidant vitamins probably re-sult from differences in study design; some trials take theeffect of hemodilution into account [1, 9].

Vitamin E, released from the liver in association withvery low-density lipoproteins, circulates in the blood-stream mainly bound to low-density lipoproteins [14].Systemic α-T concentration and its delivery to extrahe-patic target tissues appear to be regulated by factors thatcontrol both total lipid content and low-density lipopro-teins carrier capacity of the plasma [15]. Unlike α-T, vi-tamin A is released into plasma at a steady rate to meetmetabolic demands from the liver, where over 90% ofthe body’s retinoid reserves are stored [16]. In a highlyregulated process, retinol is secreted into the circulationbound to retinol-binding protein [16]. Depletion of trans-port capacity secondary to catabolic stress therefore alsoaffects the plasma levels of vitamin E and/or vitamin A.

This study was designed to investigate the course ofchanges in both α-T and retinol levels during CPB andthe immediate postoperative phase, taking into consider-ation hemodilution and lipid binding. Additionally, otherplasma biochemical parameters were monitored to pro-vide some interesting data as to whether – apart from ox-idative stress – metabolic alterations are also responsiblefor these changes. In contrast to several other studies, thetime-integrated measurements are reported only for pa-tients who did not receive blood transfusions.

Subjects and methods

Patients

The study was designed as a case series including ten men under-going elective coronary artery bypass grafting. Their mean agewas 62±9.5 years. Myocardial infarction within 6 months prior tosurgery was an exclusion criterion. All patients had an ejectionfraction exceeding 45% and a left ventricular end-diastolic pres-sure less than 15 mmHg. Patients were admitted to the trial if thefollowing criteria were met: arterial hypertension, hyperlipidemia,and overweight (10–25 kg). Patients with diabetes mellitus, hepat-ic or renal dysfunction, cancer, or excessive alcohol consumptionwere excluded from the study. This was verified via thorough his-tory taking, physical investigation, ultrasound examination, andappropriate laboratory testing. Patients smoking in the past had tohave refrained from smoking for at least 6 months.

Anesthesia and operation

Anesthetic effect was induced with etomidate or hypnomidate, suf-entanil, and midazolam. Muscle relaxation was provided by pancu-

ronium and analgesia by sufentanil. A triple-lumen central venouscatheter and a radial artery cannula were placed. Anesthesia wasmaintained by isoflurane and sufentanil. The pump prime volumethat resulted in expansion of the circulating volume was 1500 ml.Pump flow ranged between 2.0 and 2.4 l min–1 m–2 body surfacearea. The heart was exposed by a median sternotomy. Sodium hep-arin was administered intravenously (300 U/kg body weight) afterpreparation of the internal mammary artery, aiming for an activatedclotting time longer than 450 s. After aortic and venous cannulationthe patient was systemically cooled to 28°C. Cardiac arrest wasachieved with a mean of 600 ml St. Thomas cardioplegic solutioninto the aortic root at 4°C. During cross-clamping cardioplegia wasadministered every 20–30 min. Topical hypothermia was inducedwith crushed ice saline. Mean arterial pressure was maintained at50–70 mmHg by nitroglycerin and nifedipine.

Study protocol

After approval by the local ethics committee and the patients’written informed consent, arterial blood samples were obtained atten different times: at anesthesia induction (t1), after heparinizat-ion/before instituting CPB (t2), 10 min after commencing CPB(t3), 20 min after clamping the aorta (t4), 5 min (t5) and 20 min af-ter cross-clamp removal (t6), 10 min after discontinuation of CPB(t7), 6 h (t8), 24 h (t9), and 48 h postoperatively (t10). Plasma ob-tained from arterial blood from all patients was stored at –72°Cprior to analysis.

Extraction, separation, and quantification of vitamins E and A

Analyses of the antioxidants α-T and retinol were performed ondeproteinized, hexane-extracted samples [17]. The concentrationsof α-T (reference: 18–29 µmol/l [18]) and retinol (reference:0.35–1.75 µmol/l [18]) in plasma were determined fluorimetrically(α-T: excitation at 295 nm and emission at 390 nm; retinol: excita-tion at 325 nm, emission at 480 nm) by reverse-phase high-pres-sure liquid chromatography using a Kontron spherisorb ODS(5 mm inner diameter ×25 cm) column eluted with methanol at aflow rate of 1.8 ml/min. In pure standard solutions the fat-solublevitamins were quantified using the following extinction coeffi-cients (E 1%/cm): for α-T the extinction coefficient was 75.8 at292 nm in ethanol, for retinol 1830 at 325 nm in petroleum ether[19, 20]. Analyses were performed at room temperature.

Additional blood parameters investigated

Total cholesterol, triacylglycerol, albumin, total protein, glutamic-oxaloacetic transaminase (SGOT), glutamic-pyruvic transaminase(SGPT), γ-glutamyltranspeptidase (γ-GT), and C-reactive protein(CRP) were also measured in blood samples. Simultaneously a fullblood count for hematocrit and hemoglobin was taken each time tocontrol the above substance plasma levels for hemodilution. He-matocrit was determined by centrifugation; all other parameters byusing a Hitachi 911-E automatic analyzer. Baseline blood concen-trations of hemoglobin, hematocrit, total cholesterol, triacylglycer-ols, albumin, total protein, SGOT, SGPT, γ-GT, and CRP werenormal in all subjects. All plasma concentrations of the above pa-rameters were corrected for hemodilution.

Calculations

To account for hemodilution as a consequence of volume shiftsduring CPB and postoperative volume reconstitution, antioxidantplasma concentrations were corrected according to the formula:

373

374

Ccorr is defined as the corrected plasma concentration, and Cact isthe actual plasma concentration. The dilution factor is referred toas df, calculated as the ratio of baseline hematocrit and current he-matocrit.

Experimental evidence has revealed that systemic levels of vi-tamin E and vitamin A are highly correlated with total plasma lip-ids [15]. Consequently their concentrations were expressed as lip-id-standardized values. Lipid adjustment was conducted accordingto the following formula:

The regression plane with estimated slopes B1 for cholesterol andB2 for triacylglycerols is fitted through x1, x2, and y. x1 and x2 arethe cholesterol and triacylglycerol levels measured. x1,0 and x2,0are standard lipid values (5.18 mmol/l for cholesterol and1.25 mmol/l for triacylglycerols) to which adjustment is per-formed. y is the measured uncorrected vitamin plasma level andyADJ the lipid-adjusted value [15].

Statistical analysis

All analyses were performed in duplicate. Data are expressed asleast squares means and their standard error. Estimation of thetime profile and comparisons between time points were made us-ing repeated-measures analysis of variance. Statistical analysiswas carried out with Procedure Mixed of the SAS AnalysisSystem (SAS Institute, 1996, Cary, N.C., USA). Goodness-of-fitcriteria provided in the Procedure Mixed were used to choose thecovariance structure of the repeated measures: the first-order auto-regressive structure was best suited to determine random variationand autocorrelation. Dunnett’s test for many-to-one comparisonswas used to compare differences between time points. PlasmaCRP data were log transformed to achieve acceptable homogene-ity of variance. Association between the variables was determinedvia linear regression analysis. Pearson’s correlation coefficientwas calculated for selected variables. Differences were consideredto be significant at a level of P<0.05.

Results

At presentation systemic vitamin E concentration was25.6±1.6 µmol/l and therefore were within normal limitsexcept for two patients. One of them had an abnormallyhigh initial vitamin E value (60.4 µmol/l); the other had abaseline level (11.2 µmol/l) lower than normal (cutoff,18.0 µmol/l). With a mean of 1.5±0.07 µmol/l, preopera-tive retinol level was in the reference range in all patients.The admission levels of these antioxidant micronutrientsindicate that vitamin E and A nutrition was sufficient.

Prior to controlling for hemodilution, a reproducibleand substantial fall in vitamin E and A plasma concen-trations was noted during and after CPB. Uncorrectedconcentrations of plasma vitamin E, i.e., true mean val-ues for all patients, decreased by 50% (P<0.01) in theischemic phase (Fig. 1). Uncorrected concentrations ofvitamin A decreased by 55.3% (P<0.001) into the rangeof marginal deficiency (0.65±0.07 µmol/l) during isch-emia (Fig. 2). After correction for hemodilution (whichmasks any effects the higher plasma volume has on theconcentrations of blood constituents) this decrease in vi-tamin E and vitamin A levels was less marked, and theeffect of volume shifting became obvious (Figs. 1, 2).

Significant correlations were noted between the plas-ma concentrations of some of the determined parameters(Table 1). Alterations in cholesterol were paralleled bychanges in albumin concentration. Triacylglycerol, andto a lesser extent cholesterol (as a proxy of lipoproteins)and albumin showed the strongest correlation (P<0.001)with the fat-soluble vitamins (Table 1). Because of this,vitamin E and A values were corrected for both triacyl-glycerol and cholesterol (which is masking any effectsthe confounders have on the concentrations of bloodconstituents). Expressing α-T and retinol plasma concen-

Fig. 1 Time course of changesin the level of plasma α-toco-pherol before, during, and afterCPB (i.e., 6, 24, and 48 h post-operatively). Values representleast squares means determinedfor ten patients. Circles Beforecontrolling for hemodilutionand lipid dependence; squaresafter adjustment for hemodilu-tion and prior controlling forlipid dependence; triangles lip-id-standardized α-tocopherolafter correction for hemodilu-tion; filled symbols differentfrom t1 (preoperative value),P<0.01; right-hand symbolswith error bars estimates of theSE indicating the variance inthe data

375

trations as lipid-adjusted values revealed that neither de-creased significantly after connection of the extracorpo-real circuit for CPB (Figs. 1, 2). At the end of the reper-fusion period (t6), only vitamin E declined insignificant-ly, by 10% compared to preoperative levels (Fig. 1).

When albumin instead of triacylglycerols and cholesterolwas accounted for – in form of a vitamin E/albumin ratio(concentration in µmol/l were divided by g albumin/l togive µmol/g albumin) or a retinol/albumin ratio – similarfindings resulted (data not shown).

Fig. 2 Time course of changesin the level of plasma retinolbefore, during, and after CPB(i.e., 6, 24, and 48 h postopera-tively). Values represent leastsquares means determined forten patients. Circles Beforecontrolling for hemodilutionand lipid dependence; squaresafter adjustment for hemodilu-tion and before controlling forlipid dependence; triangles lip-id-standardized retinol aftercorrection for hemodilution;filled symbols different from t1(preoperative value), P<0.05;right-hand symbols with errorbars estimates of the SE indi-cating the variance in the data

Fig. 3 Effect of CPB on sys-temic retinol (y1 axis), C-reac-tive protein (log of the meanCRP concentration, y2 axis),and glutamic-oxaloacetic trans-aminase (SGOT, y3 axis) in tenpatients before, during, and af-ter (i.e., 6, 24, and 48 h postop-eratively) surgical treatment.Results are presented as leastsquares means for the numberof patients reported. SquaresCRP; circles SGOT; trianglesretinol; filled symbols differentfrom t1 (preoperative value),P<0.05; right-hand symbolswith error bars estimates of theSE indicating the variance inthe data

Table 1 Correlations (Pearson’s r)between systemic concentrationsof α-tocopherol, retinol, and sev-eral other plasma constituents inpatients undergoing CPB (con-centrations corrected for hemodi-lution) (parenthesis partial corre-lation coefficients after control-ling for plasma lipids)

Analyte α-Tocopherol Retinol

r P r P

Triacylglycerol 0.50 (0.02) 0.001 (0.408) 0.63 (–0.03) 0.001 (0.370)Cholesterol 0.47 (0.01) 0.001 (0.470) 0.51 (–0.15) 0.001 (0.067)Albumin 0.50 (0.04) 0.001 (0.355) 0.49 (–0.14) 0.001 (0.091)Total protein 0.28 (–0.05) 0.005 (0.325) 0.36 (–0.32) 0.001 (0.001)α-Tocopherol – – 0.63 (0.32) 0.001 (0.001)

The courses of plasma vitamin E and A in the postop-erative period differed. After surgery the plasma concen-trations of α-T remained at the reduced level throughout48 h (Fig. 1). In contrast to α-T, a decrease in plasma ret-inol levels was noted 6 h postoperatively (t8). With0.88±0.09 µmol/l, meaning 37.6% of preoperative levels,this decline was significant 48 h after the operation(Fig. 2). It was inversely correlated (r=–0.42, P<0.005)with the postoperative increase in the acute-phase mark-er CRP (Fig. 3). Coincident with the fall in plasma reti-nol a short-lived rise in SGOT activity to 26.5±2.13 Uwas noted 24 h postoperatively (Fig. 3). This rise is incontrast to SGPT and plasma γ-GT activities (not shown)which were normal, but in congruence with open-heartsurgery and electric cardioversion. Compared to the risein SGOT activities, the increase in CRP concentrationstarted somewhat later and had not reached its maximumwithin 48 h postsurgery (Fig. 3).

The plasma concentrations data also show that retinolfell below the critical lower limit of 0.35 µmol/l in twopatients. Both had postoperative wound healing distur-bances. The plasma vitamin A levels of six patients ap-proached the borderline of deficiency. The postoperativedecrease in vitamin A was not correlated with cross-clamping time or duration of CPB.

Discussion

In the literature it has frequently been reported that plas-ma concentrations of vitamin E decrease as a result ofextracorporeal circulation [3, 6, 8]. The tocopherol-low-ering effect of CPB may be brought about via differentmechanisms in various operative stages. In this context ithas been suggested that the increased production of freeradicals occurring during CPB, which increase thebody’s requirements for antioxidant micronutrients, suchas vitamin E and A, may be one of these mechanisms [1,2, 3, 4, 5, 6, 7, 8, 9, 10, 21]. Because substantial evi-dence has been adduced for this hypothesis, the present-ed study was not designed either to support or to refuteit. Other possible mechanisms for the tocopherol-lower-ing effect may also be operating under these conditions,for example, hemodilution and low levels of carrier mol-ecules, factors only partly focused on to date. Identifica-tion of the mechanism(s) underlying the adverse effectsof CPB on plasma antioxidant status should suggest sup-plementation strategies for improving the free radical de-fense system potential that would benefit patients under-going CPB. Against this background, a clinical studywas undertaken with a view to understanding the rea-son(s) for the depletion of plasma vitamin E and A in re-sponse to cardiac operations.

As anticipated from previous clinical evidence [1, 9],uncorrected plasma levels of both vitamins declined sig-nificantly during and after CPB. Our hematocrit data

show that hemodilution is most likely a major mecha-nism by which CPB exerts its vitamin E and A loweringeffect. Therefore by using a complementary approach(that focuses not only on vitamin E but also on vitaminA) these findings confirm and extend the observation(based on vitamin E values only) made by others [1, 3,9] as to the fall in the antioxidant content of blood rela-tive to its volume, the consequence of which is that pe-ripheral tissues may have smaller exposure to plasmaconstituents, with resulting reduction in the intracellulardeposition of antioxidant vitamins (for example see [2,5, 10]).

The strong correlation between both vitamins andplasma lipids or proteins suggests that the carrier-medi-ated transport of these essential nutrients in the circula-tion is another important factor in the time-dependentchanges in their plasma concentrations. One proposedmechanism underlying the initial drop of plasma vitaminE and A at t3 after correction for hemodilution is an im-paired synthesis and/or secretion of carrier (apolipo)pro-tein from the liver. Possible causes include liver cell in-jury and dysfunction. In the present study SGPT and γ-GT as sensitive indicators of liver cell injury, however,did not exceed normal limits. Thus liver damage is un-likely to be the causative factor in the fall in concentra-tion of the two vitamins reported here. An alternative ex-planation for the lower levels is that they are a conse-quence of disturbances in the supply and handling of thecarrier particles to and in the plasma. This assumption isfavored by the observation that the course of albuminparallels the changes in the two vitamins during CPBand up to 6 h postoperatively. This may be interpreted tomean that the initial short-lived drop in plasma levels ofvitamin E and A is indeed related to impaired secretionof the appropriate hydrophobic molecule transporters(i.e., very low-density lipoproteins, retinol-binding pro-tein) from the liver in patients undergoing CPB. Thispoor mobilization might be the expression of decreasedmetabolic activity as a result of both hypothermia and in-sufficient perfusion of the liver.

When adjusted for confounding factors (blood vol-ume and plasma lipids), however, the changes in plasmaconcentration of vitamin E and A have somewhat differ-ent courses. The major differences between them wereseen within 48 h postoperatively. Retinol but not α-T de-clined progressively following cardiac surgery. As withother antioxidants, one can assume that vitamin A isconsumed in protecting tissue against oxidative stress.Although the study was not designed to answer whetheroxidative stress contributes to this decline, one wouldthen expect vitamin E to be affected as well. The longduration since reperfusion of ischemic tissue also arguesagainst this hypothesis. However, the significant de-crease in vitamin A is more likely to be related to surgi-cal catabolism. This explanation is based on the observa-tion that the decrease in plasma vitamin A was preceded

376

377

by elevations in SGOT activity and paralleled by in-creases in CRP. This pattern of changes strongly arguesfor an effect of trauma on various biochemical processes.For example, it is known that inflammatory and traumat-ic stress is accompanied by marked alterations in proteinmetabolism [22]. The free amino acids resulting fromwhole-body protein breakdown can serve as substrate forthe biosynthesis of positive acute-phase proteins at theexpense of other proteins. Since retinol-binding proteinhas been shown to be depressed during the acute-phaseresponse [23], it is possible that the marked decline inretinol 48 h after surgery is caused by a diminished syn-thesis rate of its carrier protein [23].

In summary the findings of this study do not argue infavor of vitamin E and A consumption only, although ex-cessive utilization of these vitamins due to oxidativestress cannot be excluded. Our observations in their en-tirety suggest that a diminished mobilization of the lipid-soluble vitamins is likely to be part of the problem oftheir lower concentration during CPB, especially theirdepletion in the postoperative course. Regardless of thecause the depletion of vitamin E and A hints at an imbal-

ance in antioxidant defense. Since plasma vitamin E is inequilibrium with body stores, low circulating α-T levelsmay indicate that heart reserves of vitamin E are begin-ning to decrease [24]. If this is the case, cardiac musclein CPB-treated patients may be more vulnerable to oxi-dant injury. Therefore it appears that these patientswould benefit from receiving perioperative vitamin Ewhen undergoing cardiac surgery. Because the intraoper-ative therapeutic use of vitamin E is limited by low lev-els of circulating carrier molecules, vitamin E supple-mentation should be given before the surgical session indoses large enough to saturate at least the circulating li-poprotein particles. For vitamin A, retinol-binding pro-tein transport capacity could be circumvented by admin-istering retinoic acid which is carried on albumin [25].However, further trials are necessary to evaluate the clin-ical significance of antioxidant treatment prior to cardiacbypass surgery.

Acknowledgements We are pleased to acknowledge the assis-tance of Prof. Paul A. Lear, Southmead Hospital, Bristol, UK, forreading the manuscript and for suggestions. In addition, specialthanks go to Olaf Lange for his expert technical assistance.

References

1. Ballmer PE, Reinhart WH, Jordan P,Bühler E, Moser UK, Gey KF (1994)Depletion of plasma vitamin C but notof vitamin E in response to cardiac op-erations. J Thorac Cardiovasc Surg108:311–320

2. Barsacchi R, Pelosi G, Maffei S, Baroni M, Salvatore L, Ursini F, Verunelli F, Biagini A (1992) Myocardial vitamin E is consumedduring cardiopulmonary bypass: indi-rect evidence of free radical generationin human ischemic heart. Int J Cardiol37:339–342

3. Cavarocchi NC, England MD, O’BrienJF, Solis E, Russo P, Schaff HV, Orszulak TA, Pluth JR, Kaye MP(1986) Superoxide generation duringcardiopulmonary bypass: is there a rolefor vitamin E? J Surg Res 40:519–527

4. Mezzetti A, Lapenna D, PierdomenicoSD, Di Giammarco G, Bosco G, Di IlioC, Santarelli P, Calafiore AM, Cuccurullo F (1993) Myocardial anti-oxidant defenses during cardiopulmo-nary bypass. J Cardiovasc Surg8:167–171

5. Mickle DAG, Weisel RD, Burton GW,Ingold KU (1991) Effect of orally ad-ministered alpha tocopherol acetate onhuman myocardial alpha tocopherollevels. Cardiovasc Drugs Ther5:309–312

6. Pincemail J, Faymonville ML, Deby-Dupont G, Thirion A, Deby C,Lamy M (1989) Neutrophil activationevidenced by plasmatic myeloperoxi-dase release during cardiopulmonarybypass in humans. Ann NY Acad Sci570:501–502

7. Pyles LA, Fortney JE, Kudlak JJ, Gustafson RA, Einzig S (1995) Plasmaantioxidant depletion after cardiopul-monary bypass in operations for con-genital heart disease. J Thorac Cardio-vasc Surg 110:165–171

8. Sisto T, Paajanen H, Metsä-Ketelä T,Harmoinen A, Nordback I, Tarkka M(1995) Pretreatment with antioxidantsand allopurinol diminishes cardiac onsetevents in coronary artery bypass graft-ing. Ann Thorac Surg 59:1519–1523

9. Tangney CC, Hankins JS, MurtaughMA, Piccione W (1998) Plasma vita-mins E and C concentrations of adultpatients during cardiopulmonary by-pass. J Am Coll Nutr 17:162–170

10. Yau TM, Weisel RD, Mickle DA, Burton GW, Ingold KU, Ivanov J, Mohabeer MK, Tumiati L, Carson S(1994) Vitamin E for coronary bypassoperations. A prospective, double-blind, randomized trial. J Thorac Car-diovasc Surg 108:302–310

11. Burton GW, Ingold KU (1981) Autoxi-dation of biological molecules. 1. Theantioxidant activity of vitamin E andrelated chain-breaking phenolic antiox-idants in vitro. J Am Chem Soc21:6473–6477

12. Ciaccioa M, Valenza M, Tesoriere L,Bongiorno A, Albiero R, Livrea MA(1993) Vitamin A inhibits doxorubicin-induced membrane lipid peroxidationin rat tissues in vivo. Arch BiochemBiophys 302:103–106

13. Monaghan BR, Schmitt FO (1932) Theeffects of carotene and of vitamin A onthe oxidation of linoleic acid. J BiolChem 96:387–395

14. Behrens WA, Thompson JN, Madere R(1982) Distribution of α-tocopherol inhuman plasma lipoproteins. Am J ClinNutr 35:691–696

15. Jordan P, Brubacher D, Moser U, Stähelin HB, Gey KF (1995) Vitamin Eand vitamin A concentrations in plas-ma adjusted for cholesterol and tri-glycerides by multiple regression. ClinChem 41:924–927 (see also Erratum,Clin Chem 41:1547)

16. Goodman DS (1984) Plasma retinol-binding protein. In: Sporn MB, RobertsAB, Goodman DS (eds) The retinoids,vol 2. Academic Press, Orlando,pp 41–88

17. Schindler R, Klopp A, Gorny C, Feldheim W (1985) Comparison be-tween three fluorometric micromethodsfor determination of vitamin A in se-rum. Int J Vitam Nutr Res 55:25–34

378

18. Young DS (1987) Implementation of SIunits for clinical laboratory data. AnnIntern Med 106:114–129

19. Epler KS, Ziegler RG, Craft NE (1993)Liquid chromatographic method for thedetermination of carotenoids, retinoidsand tocopherols in human serum and infood. J Chromatogr 619:37–48

20. Thompson JN, Erdody P, Brien R,Murray TK (1971) Fluorometric deter-mination of vitamin A in human bloodand liver. Biochem Med 5:67–89

21. Ferreira RF, Milei J, Llesuy S, FlechaBG, Hourquebie H, Molteni L, de Palma C, Paganini A, Scervino L,Boveris A (1991) Antioxidant action ofvitamins A and E in patients submittedto coronary artery bypass surgery. VascSurg 25:191–195

22. Wannemacher RW (1977) Key role ofvarious individual amino acids in hostresponse to infections. Am J Clin Nutr30:1269–1280

23. Louw JA, Werbeck A, Louw MEJ, Kotze TJ, Cooper R, Labadarios D(1992) Blood vitamin concentrationsduring the acute-phase response. CritCare Med 20:934–941

24. Bieri JG, Evarts RP (1975) Effect ofplasma lipid levels and obesity on tis-sue stores of α-tocopherol. Proc SocExp Biol Med 149:500–502

25. Smith JE, Milch PO, Muto Y, Goodman DS (1973) The plasma trans-port and metabolism of retinoic acid inthe rat. Biochem J 132:821–827