A marathon run increases the susceptibility of LDL to oxidation in vitro and modifies plasma antioxidants

MING-LIN LIU, ROBERT BERGHOLM, SARI MA¨ KIMATTILA, SANNI LAHDENPERA¨ , MIIA VALKONEN, HANNELE HILDEN, HANNELE YKI-JA¨ RVINEN, AND MARJA-RIITTA TASKINEN Division of Endocrinology and Diabetology, Department of Medicine, Helsinki University Central Hospital, FIN-00029 HUCH, Helsinki, Finland

Liu, Ming-lin, Robert Bergholm, Sari Ma¨ kimattila, creased production of free radicals. However, if the Sanni Lahdenpera¨ , Miia Valkonen, Hannele Hilden, production of free radicals is excessive, as observed Hannele Yki-Ja¨ rvinen, and Marja-Riitta Taskinen. A during strenuous aerobic exercise (28, 30), or if antioxi- marathon run increases the susceptibility of LDL to oxidation dant defenses are severely hampered, the balance in vitro and modifies plasma antioxidants. Am. J. Physiol. between prooxidants and antioxidants is lost. This may 276 (Endocrinol. Metab. 39): E1083–E1091, 1999.—Physical lead to tissue damage (20). Thus there is an apparent activity increases the production of oxygen free radicals, paradox between the benefits of heavy aerobic exercise which may consume antioxidants and oxidize low-density lipoprotein (LDL). To determine whether this occurs during on cardiovascular risk factors and the potentially delete- strenuous aerobic exercise, we studied 11 well-trained run- rious consequences of free radicals generated during ners who participated in the Helsinki City Marathon. Blood heavy exercise. samples were collected before, immediately after, and 4 days Oxidative modification of low-density lipoprotein after the race to determine its effect on circulating antioxi- (LDL) greatly increases its atherogenicity and is consid- dants and LDL oxidizability in vitro. LDL oxidizability was ered to be a key step in the development of atherosclero- increased as determined from a reduction in the lag time for sis (53, 65). The susceptibility of LDL to oxidation in formation of conjugated dienes both immediately after (180 Ϯ vitro has been reported to be associated with the 7 vs. 152 Ϯ 4 min, P Ͻ 0.001) and 4 days after (155 Ϯ 7 min, severity of atherosclerosis (45). Also, autoantibodies P Ͻ 0.001) the race. No significant changes in lipid-soluble against oxidized LDL seem to predict the progression of antioxidants in LDL or in the peak LDL particle size were carotid atherosclerosis (47). Oxidative modification of observed after the race. Total peroxyl radical trapping antioxi- LDL is induced by oxygen free radicals (39). Generally, dant capacity of plasma (TRAP) and uric acid concentrations were increased after the race, but, except for TRAP, these LDL in the circulation is well protected against active changes disappeared within 4 days. Plasma thiol concentra- oxidation by highly efficient plasma antioxidant de- tions were reduced after the race. No significant changes were fense mechanisms. Total peroxyl radical trapping anti- observed in plasma ascorbic acid, ␣-tocopherol, ␤-carotene, oxidant capacity of plasma (TRAP) reflects the total and retinol concentrations after the marathon race. We combined antioxidant capacity of all individual antioxi- conclude that strenuous aerobic exercise increases the suscep- dants. If, however, oxidative stress exceeds the capacity tibility of LDL to oxidation in vitro for up to 4 days. Although of the antioxidant defense, LDL may be oxidized. the increase in the concentration of plasma TRAP reflects an Marathon running represents an extreme form of increase of plasma antioxidant capacity, it seems insufficient physical exercise and provides a model to study the to prevent the increased susceptibility of LDL to oxidation in effects of exercise-induced oxidative stress. In the pres- vitro, which was still observed 4 days after the race. ent study, we determined the acute and postexercise low-density lipoprotein oxidation; total peroxyl radical trap- effects of a marathon run on the susceptibility of LDL to ping antioxidant potential; lipids; low-density lipoprotein size oxidation in vitro, LDL particle size, antioxidants in LDL and plasma, and TRAP.

SUBJECTS AND METHODS PHYSICAL ACTIVITY is associated with beneficial changes in circulating lipids and lipoproteins (10, 41), body Subjects. Eleven healthy male marathon runners participat- weight, blood pressure, insulin sensitivity (22), and ing in the Helsinki City Marathon were studied. Written coagulation parameters (12, 13). These antiatherogenic informed consent was obtained after explanation of the purpose, nature, and potential risks of this study to the changes could contribute to the reduced prevalence of subjects. The experimental protocol was approved by the cardiovascular disease in active individuals (4, 5, 16). Ethical Committee of the Minerva Foundation for Medical Heavy endurance exercise increases the rate of oxy- Research. Data of marathon runners were compared with gen consumption in humans up to 20-fold, which in- those of an age-, sex-, and weight-matched healthy control duces oxidative stress and generates excess oxygen free group (n ϭ 10) of untrained subjects. Clinical characteristics radicals (6, 52). Under normal circumstances, the body of the study groups are summarized in Table 1. has adequate antioxidant reserves to cope with in- Study design. The subjects ran a full marathon (42.2 km), except for one subject who interrupted the race after 81 min (16 km). The running times varied from 3.13 to 5.52 h. The The costs of publication of this article were defrayed in part by the mean energy consumption during the race was calculated by payment of page charges. The article must therefore be hereby multiplying body weight in kilograms by the MET values marked ‘‘advertisement’’ in accordance with 18 U.S.C. Section 1734 (metabolic equivalent; work metabolic rate divided by resting solely to indicate this fact. metabolic rate; see Ref. 64) and duration of activity in hours

0193-1849/99 $5.00 Copyright ௠ 1999 the American Physiological Society E1083 E1084 STRENUOUS AEROBIC EXERCISE, LDL OXIDIZABILITY, AND PLASMA ANTIOXIDANTS

Table 1. Clinical characteristics of the subjects prepared CuSO4 solution to a final concentration of 10.64 µmol/l to the pure LDL solution. The kinetics of LDL oxida- Marathon Untrained tion were determined by monitoring the change in absorbance Runners Subjects at 234 nm in a motorized six-cuvette cell-equipped Shimadzu Number 11 10 spectrophotometer (Shimadzu UV-1201; Shimadzu, Kyoto, Age, yr 31Ϯ333Ϯ3 Japan) connected to a computer through a RS232 cable. Height, cm 178Ϯ2 179Ϯ2 Absorbance was recorded every 2 min. The change in absor- Weight, kg 73.6Ϯ3.1 80.1Ϯ2.3 bance at 234 nm over time could be divided into three 2 Ϯ Ϯ Body mass index, kg/m 23.3 1.0 24.9 0.6 consecutive phases: lag phase, propagation phase, and decom- Fat, % 11.4Ϯ2.0 Maximal aerobic power, ml·kgϪ1 ·minϪ1 57.2Ϯ1.8 position phase (15). The lag time (in min), the propagation Ϫ1 Ϫ1 Systolic blood pressure, mmHg 119Ϯ4 125Ϯ5 rate (in nmol·mg LDL ·min ), and the diene concentration Diastolic blood pressure, mmHg 73Ϯ377Ϯ4 (in nmol/mg LDL) were used as measures of LDL to oxidation Heart rate, beats/min 56Ϯ3 in vitro. Training history, yr 8.6Ϯ2.7 Measurement of plasma TRAP. TRAP was determined Ϯ Training extent before marathon, km/wk 46.8 8.2 spectrophotometrically using a recently validated method Best marathon time, min 232.8Ϯ15.6 (59). In this assay, we used 2Ј,7Ј-dichlorofluorescein diacetate Data are shown as means Ϯ SE. (DCFH-DA) to follow the formation of free radicals during decomposition of 2,2Ј-diazobis-(2-amidinopropane)dihydro- chloride (AAPH). Free radicals were formed during thermal (1). The subjects had no dietary restrictions during the days decomposition of AAPH in water and followed by measuring before the marathon. The runners were interviewed by a the conversion of DCFH-DA to the highly fluorescent dichloro- dietitian about their diet 24 h before the race and about their fluorescein (DCF). The DCF formation was measured at 504 energy and lipid intake during and immediately after the nm in a Shimadzu spectrophotometer. Plasma was mixed race. The calculated average energy expenditure during the Ϯ with PBS to a final dilution of 1%, followed by addition of marathon was 3,579 323 kcal. Total energy and nutrient DCFH-DA to a final concentration of 14 µmol/l. The reaction intakes 24 h before the marathon and liquid nutrient intakes was started by adding AAPH to a final concentration of 56 during the race were calculated using the Micro-Nutrica 2.0 mmol/l. The AAPH stock solution was stored at Ϫ20°C, PC program (The Social Insurance Institution, Turku, Fin- land; see Ref. 32). Twenty-four hours before the marathon, thawed, and kept in ice until added to the incubation. Trolox, total energy intake averaged 3,074 Ϯ 418 kcal, and comprised 8.4 µmol/l, was used as an internal standard, and it was 66 Ϯ 3% carbohydrates, 21 Ϯ 2% fat, and 13 Ϯ 1% protein. added during the propagation phase when the absorbance During the marathon, total energy intake averaged 249 Ϯ 53 had increased to 0.25–0.45. DCF fluorescence or absorbance kcal, 98 Ϯ 1% from carbohydrates, 1 Ϯ 0.5% from fat, and 1 Ϯ formation contains four phases. The first lag phase is due to 0.5% from protein. the antioxidants in the sample. After their consumption by Venous blood samples were collected from every runner free radicals formed from AAPH, the reaction proceeds to the before, immediately after, and 4 days after the race. Plasma first propagation phase. The second lag phase, which inter- was separated by centrifugation and stored at Ϫ80°C until rupts this propagation, is due to the addition of the internal analyzed. To correct for possible changes in plasma volume standard, Trolox, to the incubation, and, in accordance, the that may occur during the marathon (41), the hematocrit was second propagation of the reaction follows the consumption of measured from fresh whole blood both before and immedi- the Trolox. The reaction gives the measured TRAP (TRAPmea). ately after the race. The percentage change in plasma volume A theoretical calculated TRAP value (TRAPcalc) can be esti- was calculated based on the hematocrit measurements using mated by using the stoichiometric values for antioxidants the formula of Beaumont (3). Immediately after the mara- after the measurement of their concentrations in human thon, hematocrit was significantly higher than before the race samples. In the present study, the TRAPcalc was calculated (43.9 Ϯ 0.7% before vs. 45.1 Ϯ 0.7% immediately after, P ϭ from four main contributors of plasma antioxidant capacity, 0.046). Consequently, the plasma volume decreased by 2.6 Ϯ e.g., ␣-tocopherol, ascorbic acid, thiol, and uric acid. The 1.1% immediately after the race. The parameters measured difference between TRAPmea and TRAPcalc is due to antioxi- in plasma immediately after the race were corrected for by dants, which are still unknown, or to the synergism among taking into account this change in plasma volume. the various antioxidants present in the plasma (TRAPunknown). Determination of LDL oxidation. LDL was isolated by Quantitation of LDL particle size. Nondenaturing PAGE short-run ultracentrifugation using a density gradient. Briefly, was performed on serum samples and was stored at Ϫ80°C plasma (up to 5 ml) was adjusted with solid NaBr to a density using gels casted in our laboratory as previously described in of 1.5 g/ml and layered on the bottom of a centrifuge tube. detail (34, 40). Gels were stained with Sudan Black B lipid This layer was then successively overlaid with 2.5 ml each of stain and were scanned with a computer-assisted laser 1.21 and 1.063 g/ml NaCl solutions and 2.0 ml distilled water. All solutions contained 1 mg/ml of EDTA. The tubes were scanning densitometer (Personal Densitometer; Molecular centrifuged in a Beckman SW 40 Ti rotor (Beckman Instru- Dynamics, Sunnyvale, CA) using a 50-m pixel size and 12-bit ments, Fullerton, CA) in a Beckman L8–70 ultracentrifuge signal resolution. Mean particle diameter of the major LDL (Beckman, Palo Alto, CA) at 40,000 rpm at 4°C for 2.5 h. After peak was determined by comparing the mobility of the centrifugation, the main lipoproteins were well separated sample with the mobility of a calibrated reference LDL from each other, and the distinct LDL band was carefully standard run on each gel. The particle diameters of the separated. Thereafter, EDTA was removed from LDL by using reference LDL preparations were determined by electron small dextran-sulfate affinity columns (Liposorber LA-15; microscopy. The coefficients of variation for intergel and Kaneka, Osaka, Japan), as previously described (60). intragel precisions for the control sample were 0.98 and 1.8%. The oxidation of LDL in vitro was performed by using a The cutoff value for large and small LDL particle diameter modification of the procedure described by Esterbauer et al. was set at 25.5 nm, and the size of the mean particle diameter (15). The LDL oxidation was initiated by adding freshly of the major LDL peak was also measured. STRENUOUS AEROBIC EXERCISE, LDL OXIDIZABILITY, AND PLASMA ANTIOXIDANTS E1085

To avoid interassay variation, the susceptibility of LDL to points were tested by Bonferroni t-test. Statistical compari- in vitro oxidation, TRAP, and quantitation of LDL particle sons between the baseline values before the race in marathon size was analyzed in the samples of each subject from runners and those of sedentary control subjects were per- different time points (before, immediately after, and 4 days formed using the unpaired Student’s t-test. P values Ͻ 0.05 after) at the same run. were considered as statistically significant. Data are shown Measurement of plasma antioxidants. Plasma ␣-tocoph- as means Ϯ SE. erol, ␤-carotene, and retinol were measured by reverse-phase HPLC, as described by Scha¨fer-Elinder and Walldius (50). We RESULTS used a Hewlett-Packard reverse-phase HPLC column (ODS Susceptibility of LDL to oxidation in vitro and the Hypersil 5 mm, 200 mm ϫ 2.1 mm) connected to a Waters HPLC system. The latter consisted of a M600 controller, concentrations of lipid-soluble antioxidants in LDL. M486 tunable ultraviolet (UV)-absorbance detector, M717ϩ The lag time for LDL oxidation was longer in the autosampler, and Millenium 2010 single-system chromatogra- marathon runners before the race than in the un- phy manager (Waters, Milford, MA). The UV detector was set trained subjects (180 Ϯ 7 vs. 156 Ϯ 10 min, P Ͻ 0.05). at 326 nm for retinol, 292 nm for ␣-tocopherol, and 450 nm for The lag time, the propagation rate, and diene concentra- ␤-carotene. Plasma ascorbic acid was measured using the tion before and after the marathon run are shown in spectrophotometric method of Denson and Bowers (8). Plasma Table 2. CuSO4-induced LDL oxidation in vitro was protein-bound thiol (sulfhydryl groups) was determined as accelerated at the early stages of the process, both described by Ellman (11). Plasma uric acid concentrations immediately and 4 days after the race (Fig. 1). As were measured by an enzymatic colorimetric assay (Roche shown in Fig. 2A, the lag time of conjugated diene Unimate 5 UA; Roche). formation was significantly reduced immediately and 4 Measurement of lipid-soluble antioxidants in LDL.LDL was isolated by the same way as for the determination of LDL days after the race in each of the runners. The rate of oxidation. A volume containing ϳ100 µg LDL protein was oxidation measured as the rate of propagation was also used to measure the concentrations of ␣-tocopherol, ␤-caro- significantly increased immediately after the race, but tene, and retinol in LDL. All further steps including HPLC this change was abolished 4 days after the race. There separation were the same as described above for plasma were no significant differences in the concentration of lipid-soluble antioxidants. The ␣-tocopherol, ␤-carotene, and diene formation during LDL oxidation before or after retinol concentrations of LDL samples are expressed in the race. The concentrations of the lipid-soluble antioxi- relation to the total LDL protein content (35) estimated dants in LDL are shown in Table 2. There were no concomitantly in each sample, i.e., nanomole antioxidant per significant changes of ␣-tocopherol, ␤-carotene, or reti- milligram LDL protein. nol concentrations in LDL samples after compared with Concentrations of lipid and lipoproteins. Plasma lipopro- before the marathon race. The mean size of the major teins were isolated as previously described (23, 55) by using sequential ultracentrifugation (L8–70; Beckman) at the fol- LDL peak was similar immediately and 4 days after the lowing densities: very-low-density lipoprotein Ͻ1.006 g/ml, marathon race compared with the baseline value before intermediate-density lipoprotein 1.006–1.019 g/ml, LDL the race (Table 2). 1.019–1.063 g/ml, high-density lipoprotein (HDL) 1.063– 1.210 g/ml. The concentrations of cholesterol and triglycerides in plasma Table 2. Conjugated diene formation, lipid-soluble and lipoprotein subfractions were determined by enzymatic colori- antioxidants in LDL, and LDL particle size metric assays (kit 07 3680 5 for triglycerides and 07 3664 3 for of subjects before, immediately after, cholesterol; Hoffman-La Roche, Basel, Switzerland) in an and 4 days after the marathon autoanalyzer (Cobas Mira; Hoffmann-La Roche). Maximal aerobic power and body composition. Maximal Immediately 4 Days P Value aerobic power (V˙ O2max) was determined by using a work- Before After After (ANOVA) conducted upright exercise test with an electrically braked LDL oxidation cycle ergometer (Bosch ERG 220; Robert Bosch, Berlin, Lag time, min 180Ϯ7 152Ϯ4a 155Ϯ7a Ͻ0.001 Germany) combined with a continuous analysis of expiratory Diene, nmol/mg gases and minute ventilation (EOS-Spint; Erich Jaeger, LDL 475.2Ϯ20.7 477.3Ϯ14.1 487.3Ϯ23.9 NS Wurtzburg, Germany). Exercise was started at a workload of Rate, nmol·mg 50 W and was then increased by 50 W at 3-min intervals until LDLϪ1 ·minϪ1 3.87Ϯ0.24 4.68Ϯ0.23b 4.31Ϯ0.27 Ͻ0.05 perceived exhaustion or until a respiratory quotient of 1.10 Antioxidants in LDL, was reached. The highest V˙ O2 observed during a 30-s period nmol/mg was defined as V˙ O2max. Fat-free mass and the percent body fat were determined LDL protein ␣-Tocopherol using bioelectrical impedance plethysmography (Bio-Electri- in LDL 10.1Ϯ0.9 10.8Ϯ1.0 11.0Ϯ1.1 NS cal Impedance Analyzer System, model BIA-101A; RJL Sys- ␤-Carotene tem, Detroit, MI). in LDL 0.78Ϯ0.10 0.77Ϯ0.07 0.73Ϯ0.10 NS Statistical analyses. Statistical analyses were done using Retinol in LDL 0.06Ϯ0.00 0.06Ϯ0.00 0.063Ϯ0.00 NS the SYSTAT statistical package (SYSTAT, Evanston, IL). LDL peak par- Values are given as means Ϯ SE. The statistical analysis ticle size, nm 27.1Ϯ0.3 27.5Ϯ0.3 27.2Ϯ0.2 NS among the repeated observations measured before, immedi- Values are means Ϯ SE. LDL, low-density lipoprotein; NS, not ately after, and 4 days after the race were tested by one-way significant. Statistical comparisons among repeated measures of repeated-measures ANOVA for multiple comparisons. When subjects were tested by one-way ANOVA for repeated measures the overall model proved statistical significance (P Ͻ 0.05), followed by Bonferroni t-test. a P Ͻ 0.001 and b P Ͻ 0.05, compared the differences between the observations at different time with values before the marathon. E1086 STRENUOUS AEROBIC EXERCISE, LDL OXIDIZABILITY, AND PLASMA ANTIOXIDANTS

creased significantly compared with baseline values before the race. The TRAPmea increased in all subjects (Fig. 2B), and the TRAPcalc increased in all except two subjects (Fig. 2C). After 4 days, TRAPmea had decreased significantly compared with the values immediately after the race, but it was still significantly and 11.7% higher than the baseline value before the race. TRAPunknown was increased significantly both immedi- ately and 4 days after the race (Fig. 3). Immediately after the race, plasma concentrations of uric acid were higher than before the race (Table 3). Plasma thiol concentrations were significantly lower immediately after than before the race (362 Ϯ 15 vs. 315 Ϯ 12 µmol/l, P Ͻ 0.001). Four days after the race, all of these parameters were comparable to the baseline Fig. 1. Examples of continuous measurement of low-density lipopro- values before the race. We observed no significant tein (LDL) oxidation by conjugated diene absorption at 234 nm changes in plasma ascorbic acid, ␣-tocopherol, ␤-caro- (A234nm). Oxidation curves were obtained from one participant before, immediately after, and 4 days after the marathon race. tene, and retinol concentrations immediately or 4 days after the marathon race. The plasma ascorbic acid TRAP and circulating antioxidants. TRAP was higher (63 Ϯ 8 µmol/l), ␣-tocopherol (21.9 Ϯ 1.7 µmol/l), thiols in the marathon runners before the race than in the (331 Ϯ 12 µmol/l), and uric acid (336 Ϯ 12 µmol/l) in the untrained subjects (1,072 Ϯ 60 vs. 864 Ϯ 80 µmol/l, P Ͻ untrained subjects were not significantly different from 0.05). Plasma TRAP and antioxidant data in the run- those in the runners. ners are shown in Table 3. Immediately after the race, The extent of training (km/wk) before the marathon both plasma TRAPmea and TRAPcalc values corrected for correlated positively with the plasma concentration of changes in plasma volume (Table 3 and Fig. 3) in- uric acid (r ϭ 0.848, P ϭ 0.001) but inversely with the

Fig. 2. Effects of marathon race on the lag time (min) of diene formation during LDL oxidation in vitro (A), measured trapping antioxidant capacity of plasma (TRAPmea; B), calculated TRAP (TRAPcalc; C), and plasma thiol concentrations (D). Values before, immediately after, and 4 days after the race of each marathon runner (n ϭ 11) are presented. Presented values of TRAPmea, TRAPcalc, and thiol concentrations immediately after the race are the corrected values according to the plasma volume shifts. STRENUOUS AEROBIC EXERCISE, LDL OXIDIZABILITY, AND PLASMA ANTIOXIDANTS E1087

Table 3. Plasma concentrations of TRAPmea, TRAPcalc, plasma concentration of protein thiols before the mara- ascorbic acid, thiol, uric acid, ␣-tocopherol, thon run (r ϭϪ0.648, P ϭ 0.031). ␤-carotene and retinol of subjects before, Plasma lipids and lipoproteins. Plasma lipids and immediately after, and 4 days after the marathon lipoproteins were comparable in marathon runners before the race and in untrained subjects, except for One-Way HDL cholesterol, which was significantly higher in Repeated- Immediately Measures marathon runners before the race than in untrained Before After 4 Days ANOVA subjects (1.52 Ϯ 0.10 vs. 1.25 Ϯ 0.04 mmol/l, P Ͻ 0.05). (baseline) (corrected) After P Value Plasma lipids and lipoprotein concentrations in mara- a d,e TRAPmea 1,071Ϯ59 1,338Ϯ75 1,197Ϯ52 Ͻ0.001 thon runners before and after the run are shown in Ϯ Ϯ b Ϯ TRAPcalc 724 20 795 29 734 17 0.01 Table 4. Immediately after the race, the lipids and Ascorbic acid 86Ϯ10 103Ϯ10 74Ϯ4e Ͻ0.01 lipoproteins concentrations were not significantly differ- Thiol 362Ϯ15 315Ϯ12a 348Ϯ13e Ͻ0.001 Uric acid 295Ϯ16 341Ϯ16c 324Ϯ15 Ͻ0.01 ent from the baseline value. ␣-Tocopherol 17.5Ϯ1.9 19.2Ϯ2.1 17.4Ϯ2.1 Ͻ0.05 ␤-Carotene 0.53Ϯ0.04 0.57Ϯ0.04 0.51Ϯ0.05 NS DISCUSSION Retinol 2.13Ϯ0.16 2.19Ϯ0.19 2.12Ϯ0.15 NS In the present study, we found a clear reduction in Ϯ Values are means SE. Units are µmol/l. TRAPmea, measured the lag time of LDL oxidation in vitro when measured trapping antioxidative capacity of plasma (TRAP); TRAPcalc, calcu- immediately and 4 days after a marathon race. The lated TRAP. Statistical comparisons for repeated observations were tested by one-way ANOVA for repeated measures followed by Bonfer- propagation rate of conjugated diene formation also roni t-test. Values immediately after the marathon run shown here increased significantly. These data indicate that strenu- are results corrected for plasma volume shifts. P gives statistical ous exercise increases the susceptibility of LDL particle significance for comparison among baseline, immediately after, and 4 to oxidation in vitro, an effect that is potentially a Ͻ b Ͻ c Ͻ days after repeated values. P 0.001, P 0.05, and P 0.01 harmful. We also found an increase of plasma TRAP. before vs. immediately after. d P Ͻ 0.001 before vs. 4 days after. e P Ͻ 0.05 immediately after vs. 4 days after. The increase was explained by an increase in the concentration of the water-soluble antioxidant uric acid and the increase in the unknown component of TRAP, which is thought to reflect synergistic effects of antioxi- dants and the contribution of unidentified antioxidant mechanisms to TRAP (59, 63). LDL oxidizability. Sanchez Quesada et al. (48) re- ported that intensive aerobic exercise increases LDL susceptibility to CuSO4-induced oxidation. Consistent with these data, in the present study, the marathon race increased the susceptibility of LDL particle to oxidation in vitro and decreased the lag time of conju- gated diene formation. This change was not a transient phenomenon but persisted over 4 days. Consistent with

Table 4. Lipid parameters of subjects before, immediately after, and 4 days after the marathon

One-Way Repeated- Immediately Measures Before After 4 Days ANOVA (baseline) (corrected) After P Value Cholesterol, mmol/l Total 3.96Ϯ0.40 4.05Ϯ0.43 4.15Ϯ0.27 NS VLDL 0.30Ϯ0.06 0.31Ϯ0.07 0.22Ϯ0.06 NS LDL 2.46Ϯ0.23 2.49Ϯ0.23 2.38Ϯ0.20 NS HDL 1.52Ϯ0.10 1.56Ϯ0.12 1.42Ϯ0.10a,b Ͻ0.01 Triglycerides, mmol/l Total 0.92Ϯ0.14 1.21Ϯ0.11 0.99Ϯ0.17 NS Fig. 3. Measured (TRAPmea), calculated (TRAPcalc), and unknown VLDL 0.61Ϯ0.12 0.66Ϯ0.09 0.58Ϯ0.15 NS (TRAPunknown) total peroxyl radical trapping capacity and antioxi- dant properties of 4 main circulating antioxidants of the marathon Values are means Ϯ SE. VLDL, very-low-density lipoprotein; HDL, runners before, immediately after (corrected for plasma volume high-density lipoprotein. Statistical comparisons for repeated obser- shifts), and 4 days after the marathon race. Antioxidant capacities of vations were tested by one-way ANOVA for repeated measures each antioxidant were calculated by multiplying its plasma concentra- followed by Bonferroni t-test. Values immediately after the marathon tion by the stoichiometric value (molar amount of free radical trapped run shown here are results corrected for plasma volume shifts. P by mole of each antioxidant). *P Ͻ 0.05, **P Ͻ 0.01, and ***P Ͻ gives statistical significance for comparison among baseline, cor- 0.001 before vs. immediately after. ††P Ͻ 0.01 and †††P Ͻ 0.001, rected immediately after, and 4 days after repeated values. P Ͻ 0.05 before vs. 4 days after. P Ͻ 0.05 immediately after vs. 4 days after. before vs. 4 days after (a) and immediately after vs. 4 days after (b). E1088 STRENUOUS AEROBIC EXERCISE, LDL OXIDIZABILITY, AND PLASMA ANTIOXIDANTS previous cross-sectional data (49), the lag time of LDL However, in the present study, the plasma HDL concen- to oxidation in marathon runners before the marathon trations were similar before and immediately after the run was longer than in sedentary controls. These data marathon run but whether the antioxidant properties suggest that, while an acute single bout of strenuous of HDL were altered cannot be determined after a exercise increased the susceptibility of LDL to oxida- marathon run. tion, other training-associated factors counteract such Plasma TRAP and circulating antioxidants.Inthe changes. The degree of LDL oxidation reflects the net present study, plasma TRAPmea was significantly in- effects of its various prooxidants and endogenous anti- creased both immediately and 4 days after the race. In oxidants, the concentration of its oxidizable substrates, addition, plasma TRAP was higher in marathon run- especially polyunsaturated fatty acids in LDL (44, 56), ners before the race than in untrained controls. Vasan- and LDL particle size (7). Any alterations of these kari et al. (61) have also reported that acute exercise factors will influence the oxidative modification of LDL significantly increases serum TRAP. These data sug- and consequently may influence its atherogenicity. gest that acute exercise activates antioxidant defenses Because oxidation of LDL does not occur until its in the body. This activation can be viewed as an natural endogenous antioxidants have been consumed adaptive defensive mechanism to cope with increased (14), the concentration of natural endogenous antioxi- oxidative stress. Regarding the causes of the increase dants in LDL is an important determinant of the in TRAP, we found an increase in the concentration of susceptibility of LDL to oxidation. In the present study, uric acid (Table 3) and in the unidentified antioxidant the concentrations of lipid-soluble antioxidants (␣- capacity in plasma (TRAPunknown; Fig. 3). tocopherol, ␤-carotene, and retinol) in LDL were compa- During exercise, energy-rich phosphates are utilized, rable before and after the race. These data are consis- resulting in hypoxanthine accumulation in tissues (52). tent with previous cross-sectional data (49) in which Conversion of hypoxanthine to uric acid does not occur the content of lipid-soluble antioxidants in LDL did not until an adequate oxygen supply has been reinstituted. differ between trained and sedentary subjects. Vasan- The latter process is associated with the formation of kari et al. (61) observed no changes in the LDL TRAP toxic oxygen free radicals (52). On the other hand, uric during a 31-km run or during a marathon run. Because acid possesses antioxidant activity (63) and serves as a the concentration of lipid-soluble antioxidants in LDL free radical scavenger in vivo (24). Plasma uric acid can is the major determinant of the LDL TRAP, these data trap peroxyl radicals in the aqueous phase and contrib- suggest that acute exercise does not influence the ute to plasma antioxidant defenses (63). Some previous antioxidant defense capacity of LDL particles. Small studies have suggested that acute exercise increases dense LDL particles are more prone to oxidation than plasma uric acid concentrations (42, 46). This was large buoyant particles and are an independent risk confirmed in our study in which plasma uric acid predictor of cardiovascular disease (7). In the present increased during exercise, but the change was only study, LDL particle peak size was unaffected by the temporary and was not seen 4 days after the race. marathon race. This finding agrees with previous stud- However, we observed a highly significant correlation ies in which acute exercise did not significantly change between the extent of training and the baseline concen- the concentrations of LDL subfractions (2, 48). tration of plasma uric acid before the race. These data Taken together, we cannot explain the observed suggest that exercise increases uric acid concentration. increase in the susceptibility of LDL to oxidation in This increase could reflect enhanced purine oxidation vitro by changes in either the concentrations of lipid- in muscle (46). soluble antioxidants in LDL or LDL particle size. Of the Thiols (sulfhydryl groups) are known to scavenge lipid-soluble antioxidants in LDL, vitamin E seems to aqueous peroxyl radicals (17, 63). In the present study, be a more effective antioxidant than ␤-carotene (18, 25, the plasma thiol levels were significantly reduced imme- 43). Ubiquinol-10 may, however, protect human LDL diately after the marathon run in all subjects (Fig. 2D). even more efficiently against lipid peroxidation than These findings are similar to the data reported by vitamin E (54). Ubiquinol-10 in LDL or plasma was not Inayama et al. (26), who found acute exercise to de- determined in the present study. Recent studies have crease plasma thiol concentrations. Wayner et al. (63) reported that HDL can protect LDL from oxidation both reported that plasma thiols are the first antioxidants in vitro and in vivo (31, 36). Some studies suggest that consumed during 2,2Ј-azobis-(2-amidopropane)hydro- paraoxonase, an HDL-associated enzyme, may be re- chloride-initiated peroxidation of plasma. Thus the sponsible for the antioxidative action of HDL (37). observed decrease of the plasma thiol concentrations However, mechanisms via which HDL can protect LDL could reflect their oxidation by exercise-induced free from oxidation and its contribution to TRAP are still radicals in the aqueous phase of plasma. Physical poorly understood. HDL particles seem to be important training may also chronically deplete plasma thiols, for transport of circulating plasma lipid hydroperox- since we found an inverse correlation between the ides. Mackness et al. (36) reported that inhibition of the extent of training and baseline plasma thiol concentra- formation of LDL lipid peroxides is dependent on the tions. concentration of HDL. It was shown that the increased The effect of acute exercise on plasma ascorbic acid resistance to oxidation of LDL in trained subjects concentrations has been variable. Both increases (46, (compared with sedentary controls) may be due to their 62) and no change (19) in plasma ascorbic acid concen- increased plasma HDL cholesterol concentration (49). trations have been reported in different studies. In the STRENUOUS AEROBIC EXERCISE, LDL OXIDIZABILITY, AND PLASMA ANTIOXIDANTS E1089 present study, the plasma ascorbic acid concentration viewed as an adaptive defensive mechanism against tended to increase after the race, but the change was exercise-induced oxidative stress. The decrease of the not significant. Some (46, 61), but not all (38, 62) plasma thiol concentration immediately after the race previous studies, have suggested that exercise in- may reflect its role as a first-line defense mechanism creases plasma ␣-tocopherol concentrations. The in- and consumption by reactive oxygen species in the crease of plasma ␣-tocopherol induced by exercise may aqueous phase of plasma. The marathon run also be due to a shift in the interorgan distribution where increased the susceptibility of LDL to oxidation in vitro, liver and adipose tissue may be donors and muscles and and this change persisted over 4 days. This increased heart are vitamin E receivers (58). In our study, imme- susceptibility of LDL to oxidation in vitro was not diately after the race, the plasma ␣-tocopherol concen- explained by changes in either endogenous lipid- tration, corrected for the plasma volume shifts, was not soluble antioxidants in LDL or LDL particle size. significantly increased. Several other circulating plasma We are grateful to Helina¨ Perttunen-Nio, Sari Ha¨ma¨la¨inen, Kati antioxidants and antioxidant enzymes (␤-carotene, reti- Tuomola, Ulla Minkkinen, Ritva Marjanen, Kikka Runeberg, and nol, ubiquinol-10, bilirubin, glutathione, glutathione Leena Lehikoinen for skillful technical assistance. We are grateful to peroxidase, catalase, superoxide dismutase, transfer- Dr. Mikko Syva¨nne and Juha Vakkilainen for helpful advice on the statistical analyses. We also thank Maaria Puupponen for help on rin, and ceruloplasmin) also contribute to the plasma this manuscript. antioxidant capacity (21, 27–30, 46, 51, 62, 63). The This work was supported by grants from the Helsinki University relative contribution of these components to TRAP is, Central Hospital EVO Funds, the Academy of Finland (H. Yki- however, trivial (63). In the present study, exercise had Ja¨rvinen, M.-R. Taskinen), Finska La¨karesa¨llskapet (R. Bergholm), no effect on the plasma ␤-carotene and retinol concen- and the Sigrid Juselius Foundation (H. Yki-Ja¨rvinen, M.-R. Taski- nen). trations. These findings are consistent with previous Address for correspondence: M.-R. Taskinen, Div. of Endocrinology studies in which acute exercise did not change plasma and Diabetology, Dept. of Medicine, Haartmaninkatu 4, PO Box 340, ␤-carotene (46) and retinol levels (46, 61). FIN-00029 HUCH, Helsinki, Finland (E-mail: mataskin@helsinki.fi). Plasma antioxidants can also act synergistically in Received 14 July 1998; accepted in final form 19 February 1999. vivo to provide more protection against free radical damage than could be provided by any single antioxi- REFERENCES dant alone (63). In the present study, before the race, 1. Ainsworth, B. E., W. L. Haskell, A. S. Leon, D. R. J. Jacobs, TRAP and TRAP amounted to 68 and 32% of H. J. Montoye, J. F. Sallis, and R. S. J. Paffenbarger. calc unknown Compendium of physical activities: classification of energy costs the TRAPmea (Fig. 3). In contrast, after the marathon of human physical activities. Med. Sci. Sports Exerc. 25: 71–80, run, the increase of TRAPmea was due to a 73% increase 1993. of TRAPunknown, whereas the increases in the concentra- 2. Baumstark, M. W., I. Frey, and A. Berg. Acute and delayed tions of uric acid and other antioxidants accounted for effects of prolonged exercise on serum lipoproteins. II. Concentra- tion and composition of low-density lipoprotein subfractions and 27%. This marked increase of TRAPunknown suggests very low-density lipoproteins. Eur. J. Appl. Physiol. 66: 526–530, that acute exercise may activate unidentified antioxi- 1993. dant defenses as a physiological adaptive response 3. Beaumont, W. V. Evaluation of hemoconcentration from hema- against oxidative stress. Even so, neither of the compo- tocrit measurements. J. Appl. Physiol. 32: 712–713, 1972. 4. Blair, S. N., H. W. Kohl, R. S. J. Paffenbarger, D. G. Clark, nents of TRAP seem sufficient to prevent the increased K. H. Cooper, and L. W. Gibbons. Physical fitness and all- susceptibility of isolated LDL particle to oxidation in cause mortality. A prospective study of healthy men and women. vitro. JAMA 262: 2395–2401, 1989. Lipid and lipoproteins. Previous data (10) have shown 5. Chandrashekhar, Y., and I. S. Anand. Exercise as a coronary that regular physical training is associated with lower- protective factor. Am. Heart J. 122: 1723–1739, 1991. 6. Davies, K. J., A. T. Quintanilha, G. A. Brooks, and L. ing of plasma total triglycerides, total cholesterol and Packer. Free radicals and tissue damage produced by exercise. LDL concentrations, and an increase in the plasma Biochem. Biophys. Res. Commun. 107: 1198–1205, 1982. HDL cholesterol concentration. The effect of a single 7. De Graaf, J., H. L. Hak Lemmers, M. P. Hectors, P. N. bout of exercise on lipid concentrations has, however, Demacker, J. C. Hendriks, and A. F. Stalenhoef. Enhanced susceptibility to in vitro oxidation of the dense low density been highly variable (10, 41). For example, increases lipoprotein subfraction in healthy subjects. Arterioscler. Thromb. (9), decreases (57), or no change (33) in plasma total 11: 298–306, 1991. triglycerides and/or total cholesterol have been re- 8. Denson, K. W., and E. F. Bowers. The determination of ported in different studies. The changes of plasma LDL ascorbic acid in white blood cells: a comparison of ascorbic acid and phenolic acid excretion in elderly patients. Clin. Sci. Mol. cholesterol concentrations have also been variable, Med. 21: 157–162, 1961. whereas the increase in HDL cholesterol has been more 9. Dufaux, B., U. Order, R. Muller, and W. Hollmann. Delayed consistent (9, 33, 57). In the present study, the concen- effects of prolonged exercise on serum lipoproteins. Metabolism trations of total cholesterol, triglycerides, and LDL 35: 105–109, 1986. cholesterol remained unchanged immediately and 4 10. Durstine, J. L., and W. L. Haskell. Effects of exercise training on plasma lipids and lipoproteins. Exerc. Sport Sci. Rev. 22: days after the marathon run. The plasma HDL choles- 477–521, 1994. terol concentration immediately after the race was 11. Ellman, G. L. Tissue sulfhydryl groups. Arch. Biochem. Biophys. slightly, but not significantly, increased compared with 82: 70–77, 1959. the baseline value. 12. El Sayed, M. S. Effects of exercise on blood coagulation, fibrino- lysis and platelet aggregation. Sports Med. 22: 282–298, 1996. Conclusion. A single bout of heavy endurance physi- 13. Ernst, E., and K. L. Resch. Therapeutic interventions to lower cal activity increases TRAP and some (uric acid, plasma fibrinogen concentration. Eur. Heart J. 16, Suppl.A: TRAPunknown) of its components. These increases can be 47–52, 1995. E1090 STRENUOUS AEROBIC EXERCISE, LDL OXIDIZABILITY, AND PLASMA ANTIOXIDANTS

14. Esterbauer, H., G. Jurgens, O. Quehenberger, and E. Kol- androgenic hormones in healthy males. Metabolism 33: 527–531, ler. Autooxidation of human low density lipoprotein: loss of 1984. polyunsaturated fatty acids and vitamin E and generation of 34. Lahdenpera¨ , S., J. Puolakka, T. Pyo¨ra¨la¨ , H. Luotola, and aldehydes. J. Lipid Res. 28: 495–509, 1987. M. R. Taskinen. Effects of postmenopausal estrogen/progestin 15. Esterbauer, H., G. Striegl, H. Puhl, and M. Rotheneder. replacement therapy on LDL particles; comparison of transder- Continuous monitoring of in vitro oxidation of human low mal and oral treatment regimens. Atherosclerosis 122: 153–162, density lipoprotein. Free Radic. Res. Commun. 6: 67–75, 1989. 1996. 16. Fletcher, G. F., G. Balady, S. N. Blair, J. Blumenthal, C. 35. Lowry, O. H., N. J. Rosebrough, A. L. Farr, and J. R. Caspersen, B. Chaitman, S. Epstein, E. S. Sivarajan Froeli- Randall. Protein measurement with the Folin phenol reagent. J. cher, V. F. Froelicher, I. L. Pina, and M. L. Pollock. State- Biol. Chem. 193: 265–275, 1951. ment on exercise: benefits and recommendations for physical 36. Mackness, M., C. Abbott, S. Arrol, and P. Durrington. The activity programs for all Americans. A statement for health role of high-density lipoprotein and lipid-sobuble antioxidant professionals by the Committee on Exercise and Cardiac Rehabili- vitamins in inhibiting low-density lipoprotein oxidation. Bio- tation of the Council on Clinical Cardiology, American Heart chem. J. 294: 829–834, 1993. Association. Circulation 94: 857–862, 1996. 37. Mackness, M. I., and P. N. Durrington. HDL, its enzymes and 17. Frei, B., R. Stocker, and B. N. Ames. Antioxidant defenses and its potential to influence lipid peroxidation. Atherosclerosis 115: lipid peroxidation in human blood plasma. Proc. Natl. Acad. Sci. 243–253, 1995. USA 85: 9748–9752, 1988. 38. Meydani, M., W. J. Evans, G. Handelman, L. Biddle, R. A. 18. Gaziano, J. M., A. Hatta, M. Flynn, E. J. Johnson, N. I. Fielding, S. N. Meydani, J. Burrill, M. A. Fiatarone, J. B. Krinsky, P. M. Ridker, C. H. Hennekens, and B. Frei. Blumberg, and J. G. Cannon. Protective effect of vitamin E on Supplementation with beta-carotene in vivo and in vitro does not exercise-induced oxidative damage in young and older adults. inhibit low density lipoprotein oxidation. Atherosclerosis 112: Am. J. Physiol. 264 (Regulatory Integrative Comp. Physiol. 33): 187–195, 1995. R992–R998, 1993. 19. Ginsburg, G. S., A. Agil, M. O’Toole, E. Rimm, P. S. Douglas, 39. Morel, D. W., J. R. Hessler, and G. M. Chisolm. Low density and N. Rifei. Effects of a single bout of ultraendurance exercise lipoprotein cytotoxicity induced by free radical peroxidation of on lipid levels and susceptibility of lipids to peroxidation in lipid. J. Lipid Res. 24: 1070–1076, 1983. triathletes. JAMA 276: 221–225, 1996. 40. Nichols, A. V., R. M. Krauss, and T. A. Musliner. Nondenatur- 20. Gutteridge, J. M. Lipid peroxidation and antioxidants as ing polyacrylamide gradient gel electrophoresis. Methods Enzy- biomarkers of tissue damage. Clin. Chem. 41: 1819–1828, 1995. mol. 128: 417–431, 1986. 21. Halliwell, B., and J. M. Gutteridge. The antioxidants of 41. Pronk, N. P. Short term effects of exercise on plasma lipids and human extracellular fluids. Arch. Biochem. Biophys. 280: 1–8, lipoproteins in humans. Sports Med. 16: 431–448, 1993. 1990. 42. Rasanen, L. A., P. A. Wiitanen, E. M. Lilius, S. Hyyppa, and 22. Hardin, D. S., B. Azzarelli, J. Edwards, J. Wigglesworth, L. A. R. Poso. Accumulation of uric acid in plasma after repeated Maianu, G. Brechtel, A. Johnson, A. Baron, and W. T. bouts of exercise in the horse. Comp. Biochem. Physiol. B Garvey. Mechanisms of enhanced insulin sensitivity in endur- Biochem. Mol. Biol. 114: 139–144, 1996. ance-trained athletes: effects on blood flow and differential 43. Reaven, P. D., A. Khouw, W. F. Beltz, S. Parthasarathy, and expression of GLUT 4 in skeletal muscles. J. Clin. Endocrinol. J. L. Witztum. Effect of dietary antioxidant combinations in Metab. 80: 2437–2446, 1995. humans. Protection of LDL by vitamin E but not by beta- 23. Havel, J. R., Eder H. A., and Bragdon J. H. Distribution and carotene. Arterioscler. Thromb. 13: 590–600, 1993. chemical composition of ultracentrifugally separated lipopro- 44. Reaven, P. D., and J. L. Witztum. Oxidized low density teins in human serum. J. Clin. Invest. 34: 1345–1353, 1955. lipoproteins in atherogenesis: role of dietary modification. Annu. 24. Hellsten, Y., P. C. Tullson, E. A. Richter, and J. Bangsbo. Rev. Nutr. 16: 51–71, 1996. Oxidation of urate in human skeletal muscle during exercise. 45. Regnstrom, J., J. Nilsson, P. Tornvall, C. Landou, and A. Free Radic. Biol. Med. 22: 169–174, 1997. Hamsten. Susceptibility to low-density lipoprotein oxidation 25. Hennekens, C. H., J. E. Buring, J. E. Manson, M. Stampfer, and coronary atherosclerosis in man. Lancet 339: 1183–1186, B. Rosner, N. R. Cook, C. Belanger, F. LaMotte, J. M. 1992. Gaziano, P. M. Ridker, W. Willett, and R. Peto. Lack of effect 46. Rokitzki, L., E. Logemann, A. N. Sagredos, M. Murphy, W. of long-term supplementation with beta carotene on the inci- Wetzel Roth, and J. Keul. Lipid peroxidation and antioxidative dence of malignant neoplasms and cardiovascular disease. N. vitamins under extreme endurance stress. Acta Physiol. Scand. Engl. J. Med. 334: 1145–1149, 1996. 151: 149–158, 1994. 26. Inayama, T., Y. Kumagai, M. Sakane, M. Saito, and M. 47. Salonen, J. T., S. Yla Herttuala, R. Yamamoto, S. Butler, H. Matsuda. Plasma protein-bound sulfhydryl group oxidation in Korpela, R. Salonen, K. Nyyssonen, W. Palinski, and J. L. humans following a full marathon race. Life Sci. 59: 573–578, Witztum. Autoantibody against oxidised LDL and progression of 1996. carotid atherosclerosis. Lancet 339: 883–887, 1992. 27. Jenkins, R. R. Exercise, oxidative stress, and antioxidants: a 48. Sanchez Quesada, J. L., R. Homs Serradesanferm, J. Ser- review. Int. J. Sport Nutr. 3: 356–375, 1993. rat Serrat, J. R. Serra Grima, F. Gonzalez Sastre, and J. 28. Ji,L.L.Oxidative stress during exercise: implication of antioxi- Ordonez Llanos. Increase of LDL susceptibility to oxidation dant nutrients. Free Radic. Biol. Med. 18: 1079–1086, 1995. occurring after intense, long duration aerobic exercise. Atheroscle- 29. Ji,L.L.Exercise and oxidative stress: role of the cellular rosis 118: 297–305, 1995. antioxidant systems. Exerc. Sport Sci. Rev. 23: 135–166, 1995. 49. Sanchez-Quesada, J. L., H. Ortega, A. Payes-Romero, J. 30. Kanter, M. Free radicals and exercise: Effects of nutritional Serrat-Serrat, F. Gonzalez-Sastre, M. A. Lasuncion, and J. antioxidant supplementation. Exerc. Sport Sci. Rev. 23: 375– Ordonez-Llanos. LDL from aerobically-trained subjects shows 397, 1995. higher resistance to oxidative modification than LDL from 31. Klimov, A., V. Gurevich, A. Nikiforova, L. Shatilina, A. sedentary subjects. Atherosclerosis 132: 207–213, 1997. Kuzmin, S. Plavinsky, and N. Teryukova. Antioxidative 50. Scha¨ fer-Elinder, L., and G. Walldius. Simultaneous measure- activity of high density liproteins in vivo. Atherosclerosis 100: ment of serum probucol and lipid-soluble antioxidants. J. Lipid 13–18, 1993. Res. 33: 131–137, 1992. 32. Knuts, L., M. Rastas, P. Haapala, R. Seppa¨ nen, R. Karvetti, 51. Sen, C. K. Oxidants and antioxidants in exercise. J. Appl. and A. Aro. Nutrica-a computer program for calculation of food Physiol. 79: 675–686, 1995. and nutrient intake (Abstract). In: Proc. Meet. Eurofoods 3rd 52. Sjo¨ din, B., Y. H. Westing, and F. S. Apple. Biochemical Warsaw 1987,p.16. mechanisms for oxygen free radical formation during exercise. 33. Kuusi, T., E. Kostiainen, E. Vartiainen, L. Pitkanen, C. Sports Med. 10: 236–254, 1990. Ehnholm, H. J. Korhonen, A. Nissinen, and P. Puska. Acute 53. Steinberg, D., S. Parthasarathy, T. E. Carew, J. C. Khoo, effects of marathon running on levels of serum lipoproteins and and J. L. Witztum. Beyond cholesterol. Modifications of low- STRENUOUS AEROBIC EXERCISE, LDL OXIDIZABILITY, AND PLASMA ANTIOXIDANTS E1091

density lipoprotein that increase its atherogenicity. N. Engl. J. 60. Valkonen, M., and T. Kuusi. Passive smoking induces athero- Med. 320: 915–924, 1989. genic changes in low-density lipoprotein. Circulation 97: 2012– 54. Stocker, R., V. W. Bowry, and B. Frei. Ubiquinol-10 protects 2016, 1998. human low density lipoprotein more efficiently against lipid 61. Vasankari, T. J., U. M. Kujala, T. M. Vasankari, T. Vuori- peroxidation than does alpha-tocopherol. Proc. Natl. Acad. Sci. maa, and A. Markku. Effects of acute prolonged exercise on USA 88: 1646–1650, 1991. serum and LDL oxidation and antioxidant defences. Free Radic. 55. Taskinen, M. R., T. Kuusi, E. Helve, E. A. Nikkila¨ , and H. Biol. Med. 22: 509–513, 1997. Yki Ja¨ rvinen. Insulin therapy induces antiatherogenic changes 62. Viguie, C. A., B. Frei, M. K. Shigenaga, B. N. Ames, L. of serum lipoproteins in noninsulin-dependent diabetes. Arterio- Packer, and G. A. Brooks. Antioxidant status and indexes of sclerosis 8: 168–177, 1988. oxidative stress during consecutive days of exercise. J. Appl. 56. Thomas, M. J., T. Thornburg, J. Manning, K. Hooper, and Physiol. 75: 566–572, 1993. L. L. Rudel. Fatty acid composition of low-density lipoprotein 63. Wayner, D. D. M., G. W. Burton, K. U. Ingold, L. C. Barclay, influences its susceptibility to autoxidation. Biochemistry 33: and S. J. Locke. The relative concentributions of vitamin E, 1828–1834, 1994. urate, ascorbate and proteins to the total peroxyl radical- 57. Thompson, P. D., E. M. Cullinane, O. Henderson, and P. N. trapping antioxidant activity of human blood plasma. Biochim. Herbert. Acute effects of prolonged exercise on serum lipids. Biophys. Acta 924: 408–419, 1987. Metabolism 29: 662–665, 1980. 64. Wilson, P. W. F., R. S. Paffenbarger, J. N. Morris, and R. J. 58. Tiidus, P. M., and M. E. Houston. Vitamin E status and Havlik. Assessment methods for physical activity and physical response to exercise training. Sports Med. 20: 12–23, 1995. fitness in population studies: report of a NHLBI workshop. Am. 59. Valkonen, M., and T. Kuusi. Spectrophotometric assay for total Heart J. 111: 1177–1192, 1986. peroxyl radical-trapping antioxidant potential in human serum. 65. Witztum, J. L. The oxidation hypothesis of atherosclerosis. J. Lipid Res. 38: 823–833, 1997. Lancet 344: 793–795, 1994.