Effects of exercise on emerging and traditional cardiovascular risk factors

Preventive Medicine 39 (2004) 894 – 902 www.elsevier.com/locate/ypmed

Effects of exercise on emerging and traditional cardiovascular risk factors Glen E. Duncan, Ph.D., RCEPSM, a,* Michael G. Perri, Ph.D., b Stephen D. Anton, M.S., b Marian C. Limacher, M.D., c A. Daniel Martin, Ph.D., P.T., d David T. Lowenthal, M.D., Ph.D., c,e Erland Arning, Ph.D., f Teodoro Bottiglieri, Ph.D., f and Peter W. Stacpoole, Ph.D., M.D. c,g a Department of Epidemiology, Nutritional Sciences Program, University of Washington, Seattle, WA 98195, USA Department of Clinical and Health Psychology, University of Florida Health Science Center, Gainesville, FL 32610, USA c Department of Medicine, University of Florida Health Science Center, Gainesville, FL 32610, USA d Department of Physical Therapy, University of Florida Health Science Center, Gainesville, FL 32610, USA e GRECC, VA Medical Center, Gainesville, FL 32610, USA f Baylor Institute of Metabolic Disease, Dallas, TX 75226, USA g Department of Biochemistry and Molecular Biology, University of Florida Health Science Center, Gainesville, FL 32610, USA b

Available online 30 April 2004

Abstract Background. Common cardiovascular disease risk factors (e.g., insulin and aerobic fitness) are improved with exercise; however, few studies have addressed the potential for training to modify emerging cardiovascular disease risk factors such as homocysteine and highsensitivity C-reactive protein. Methods. Sedentary adults (n = 324, 48.9 F 8.4 years) were randomized to four groups differing in training intensity (moderate = 45 – 55% or high = 65 – 75% of heart rate reserve) and frequency (low = 3 – 4, 30-min sessions/week or high = 5 – 7, 30 min-sessions/week). Results. Within-group changes in homocysteine, insulin, and aerobic fitness were significant (all P < 0.0125). Furthermore, homocysteine increased in the high-intensity – low-frequency (0.98 F 2.32 Amol/L) and high-intensity – high-frequency (0.93 F 2.56 Amol/L) groups, while aerobic fitness increased in the moderate-intensity – high-frequency (0.99 F 2.01 mL min
1 kg
1) and high-intensity – high-frequency (1.77 F 2.97 mL min
1 kg
1) groups (all P < 0.003). The change in aerobic fitness was greater in the high-intensity – high-frequency compared to the moderate-intensity – low-frequency group (1.77 F 2.97 vs. 0.36 F 2.10 mL min
1 kg
1, P = 0.0014) (effect size estimate = 0.60 mL min
1 kg
1). The main effects for intensity, with respect to the change in insulin (effect size estimate = 0.46 AU/mL), and frequency, with respect to the change in aerobic fitness (effect size estimate = 0.38 mL min
1 kg
1), were significant ( P < 0.0125). Conclusion. Although frequent bouts of higher intensity exercise were particularly effective in reducing fasting insulin and improving fitness, they resulted in slightly increased homocysteine levels. D 2004 The Institute For Cancer Prevention and Elsevier Inc. All rights reserved. Keywords: Homocysteine; C-reactive protein; Insulin; Oxygen consumption; Aerobic exercise

Introduction Increased physical activity in previously sedentary individuals modifies known risk factors for cardiovascular disease (CVD), including decreases in blood pressure [1], circulating levels of serum cholesterol [1] and plasma

* Corresponding author. Department of Epidemiology, Nutritional Sciences Program, University of Washington, 305 Raitt Hall, Box 353410, Seattle, WA 98195. Fax: +1-206-685-1696. E-mail address: [email protected] (G.E. Duncan).

insulin (INS) [2], and increases in insulin action [3] and cardiorespiratory fitness (VO2max) [1,4]. In turn, these alterations are associated with reductions in CVD morbidity and mortality [5]. While the effects of exercise on these traditional risk factors are well documented, few studies have addressed the potential for exercise to modify more recently recognized CVD risk factors, including circulating levels of homocysteine (Hcy) and high-sensitivity C-reactive protein (CRP). Previous reports demonstrate that total Hcy concentration measured in plasma or serum is an independent risk factor for CVD [6 –9]. Purported mechanisms for this

0091-7435/$ - see front matter D 2004 The Institute For Cancer Prevention and Elsevier Inc. All rights reserved. doi:10.1016/j.ypmed.2004.03.012

G.E. Duncan et al. / Preventive Medicine 39 (2004) 894–902

association include the ability of Hcy to promote endothelial cell injury, stimulate vascular smooth muscle cell growth, increase platelet adhesiveness, enhance LDL oxidation and deposition in the arterial wall, and directly activate the coagulation cascade [10]. Although habitual physical activity level was inversely related to Hcy in an epidemiological study [11], equivocal results have been obtained from studies examining the effects of acute and chronic exercise on Hcy [12 –16]. A relationship between CRP, an acute phase reactant synthesized by the liver [17], and atherosclerotic disease has been observed in epidemiological investigations [18,19]. However, whether CRP is a marker for disease progression or whether it is directly involved in the pathogenesis of atherosclerosis remains unclear. CRP is associated with several features of the metabolic syndrome, including increased fasting INS, body mass index (BMI), and waist circumference [20]. Increased levels of CRP that are common in obesity may be due to interleukin-6. This cytokine is produced by adipocytes in proportion to fat cell mass [21] and stimulates hepatic synthesis of CRP. In line with this observation, weight loss due to caloric restriction reduced CRP levels in obese women [22,23]. Trained athletes have lower CRP concentrations compared to sedentary controls [24], and increased physical activity reduced fasting CRP in a healthy elderly population [25]. Furthermore, CRP levels were reduced following 9 months of increased running mileage in a group of athletes preparing for a marathon [26] and after 3 months of supervised exercise training in patients with intermittent claudication [27]. However, the effect of exercise on CRP has not been examined prospectively in previously sedentary, but otherwise healthy, adults. Current guidelines promulgated by the CDC/American College of Sports Medicine [28] and the U.S. Surgeon General [29] recommend that all adults accumulate 30 or more minutes of moderate-intensity physical activity on most, if not all, days of the week to achieve health benefits. However, these reports also describe a dose –response relationship between physical activity and health outcomes and further suggest that greater benefits can be derived from engaging in frequent bouts of high-intensity exercise [28,29]. No study has validated this recommendation, however, and the exact dose –response relationship between exercise training and CVD risk factors is unknown. Accordingly, we hypothesized that both moderate- and higher intensity aerobic exercise training would significantly decrease circulating levels of Hcy, CRP, and INS, and improve VO2max (within-group analyses). Furthermore, we hypothesized that frequent bouts of higher intensity exercise would result in significantly greater changes in the primary outcome measures, compared to less frequent and less intense bouts of exercise (between-group analyses).

895

Methods Subjects Sedentary adults were recruited (primarily through newspaper advertisement) to take part in a 2-year study of the effects of varying formats of aerobic training on several health and fitness measures. The present study examined the dose –response of exercise on Hcy, CRP, INS, and VO2max over the initial 6 months of training. Through an initial telephone interview, we excluded individuals who had known chronic disease (e.g., coronary heart disease, congestive heart failure, renal disease, and diabetes) or who engaged in structured physical activity on more than two occasions per week, each lasting 30 or more minutes, over the previous 12 months. Participants who passed this screen were invited to attend an information session where they completed demographic, health history, and two physical activity questionnaires [30,31] to further establish that they were sedentary. Individuals interested in continuing their participation were scheduled for a series of evaluations that included a detailed medical history, exercise testing, and laboratory measures of endocrine, hematological, and metabolic function. Based on these evaluations, individuals with abnormal findings were excluded from further participation. Prescription medication use was documented at baseline and at each subsequent evaluation period, and none of the participants were taking prescription medications that are known to influence any of the outcome measures. Study measurements were taken in the General Clinical Research Center (GCRC), Shands Hospital at the University of Florida. Before testing, all participants provided written informed consent according to the standards established by the Health Science Center Institutional Review Board. All study-related visits requiring blood sampling were conducted in the GCRC in the morning following an overnight fast. Participants were instructed to keep physical activity to a minimum on the day preceding all baseline testing. For all 6-month assessments, participants were studied 24 – 48 h after their last training bout. Measures Aerobic capacity Graded treadmill exercise (Bruce protocol) was performed to volitional fatigue before and after training (see below) to measure VO2max and maximal heart rate. Subjects were required to meet two of three standard criteria for having achieved VO2max (heart rate z age-predicted maximum heart rate, respiratory exchange ratio z 1.10, rating of perceived exertion z 19). Pulmonary gas exchange variables were measured continuously using a metabolic cart (TrueMax 2400, ParvoMedics, Inc., Sandy, UT). Pulmonary ventilation (VE
l min
1) was measured by a pneumotach that was calibrated daily and fractions of O2 and CO2 via analyzers that were calibrated with gases of known concen-

896

G.E. Duncan et al. / Preventive Medicine 39 (2004) 894–902

tration before each test. Heart rate was measured by continuous 12-lead electrocardiography. Resting heart rate was the average of three seated measurements, performed and averaged over 2 different days.

interassay coefficient of variation for insulin is <5% in this laboratory.

Anthropometrics Body mass index (BMI = kg/m2) was calculated as an index of total body mass. For all BMI measurements, height was measured using a wall-mounted stadiometer and weight on a balance beam scale; both with shoes removed. The waist-to-hip circumference ratio was calculated as an index of body fat distribution. Waist circumference was measured at the narrowest part of the torso, between the xiphoid process and the umbilicus, and hip circumference was measured at the maximal circumference of the buttocks, above the gluteal fold, using a spring-retractable tape measure.

Following baseline assessments, participants were assigned an individualized exercise prescription. The training program consisted of walking, prescribed at an intensity of either 45 – 55% (moderate intensity) or 65– 75% (high intensity) of individual heart rate reserve [33] and a frequency of either 3– 4 (low frequency) or 5– 7 (high frequency) days per week. By manipulating intensity and frequency, we formed four training groups (highintensity – high-frequency, high-intensity – low-frequency, moderate-intensity – high-frequency, and moderate-intensity –low-frequency) to which participants were randomly assigned. To minimize potential confounding influences on changes in the outcome measures, subjects were stratified by age (30 –49 years and 50– 69 years) and sex. Women were further stratified based on menopausal status and use of hormone replacement therapy. To account for possible seasonal and laboratory measurement variations in the outcome measures, participants began in one of five cohorts (with approximately 20 subjects in each of the four training groups) whose initial training sessions were separated by 3-month intervals during the initial year of the study. Participants were required to walk within the individualized target heart rate training zone for 30 min per day (either in a continuous bout or in up to three bouts, each of at least 10-min duration), for the prescribed number of days per week. Training could occur at home, at a work site, or both. While participants were given flexibility in scheduling their exercise routines, the importance of achieving 30 min of exercise in the assigned target heart rate training zone was strongly emphasized. Participants wore a heart rate monitor (Polar Beat, Polar Electro, Inc., Port Washington, NY) to gauge exercise intensity and were instructed to record the most frequently observed heart rate during the exercise session in an exercise log. Participants were instructed to use daily training logs for self-monitoring of their exercise, including the duration (i.e., number of minutes) and the intensity (i.e., average heart rate) of all bouts of walking of at least 10-min duration. Because each bout of walking was recorded separately, the frequency of walking was also recorded. Staff members collected the training logs at the intervention sessions (see below). A participant who missed a session was contacted by a member of the research team to obtain the exercise information and to encourage future attendance. Two major indicators of exercise adherence were calculated, based on the data recorded in the daily training logs over a 6-month period: (a) percentage of prescribed minutes of walking within the prescribed target heart rate training zone and (b) total minutes of walking.

C-reactive protein Serum CRP was measured with commercially available kits (N High Sensitivity CRP Reagent, Dade-Behring Diagnostics, Newark, DE) designed for use with a BN-100 nephalometric detector (Dade-Behring Diagnostics). Polystyrene particles that contained specific antibodies to human CRP were agglutinated when mixed with samples that contained human CRP. The intensity of the scattered light in the nephalometer was proportional to the CRP content of the sample. The CRP concentration was quantitated by comparison to dilutions of a standard of known concentration. Samples for CRP were analyzed by the Baylor Institute of Metabolic Disease. The within-day and between-day coefficients of variation of the CRP assay in this laboratory are 5.5% and 5.3%, respectively. Homocysteine Serum total Hcy was measured by reverse-phase HPLC, coupled with colometric-electrochemical detection [32]. Each sample was reduced with Tris-2-carboxyethyl-phosphine (TCEP), followed by precipitation of serum proteins with trichloroacetic acid (TCA). The TCA precipitates were injected into an integrated HPLC system (ESA, Inc., Chelmsford, MA) and separated on a reverse-phase column (Prodigy 5 ODS, Phenomenex, Inc., Torrence, CA). Detection of total Hcy was performed by oxidation at +900 mV on the electrode surface. Quantitation of total Hcy was performed with an internal calibrator, penicillamine, with D,LHcy as the external calibration standard. Samples for Hcy were analyzed by the Baylor Institute of Metabolic Disease. The within-day and between-day coefficients of variation of the Hcy assay are 6.2% and 12.6%, respectively, in this laboratory. Insulin Fasting plasma INS levels were analyzed using a chemiluminescent, double-antibody technique by the Diagnostic Reference Laboratory at the University of Florida. The

Exercise training

G.E. Duncan et al. / Preventive Medicine 39 (2004) 894–902

Exercise intervention After randomization, all participants received a standard exercise intervention [34]. The intervention included 11 group sessions over 6 months. Sessions were held weekly during Month 1, biweekly during Months 2– 3, and monthly during Months 4– 6. They were conducted by counselors with graduate training in exercise science and/or behavioral science who followed a structured protocol as outlined below. Each group session was conducted in three segments. Segment One included a review of participants’ progress in implementing the strategies recommended for increasing their activity levels since the previous session. Segment Two included didactic instruction on various exercise and fitness topics, such as proper walking techniques and behavioral self-management skills. Segment Three included a 30-min walking bout supervised by group leaders. Total caloric intake and the percentage of calories from fat, carbohydrate, and protein were estimated from food records. Each participant was instructed to record the types and amounts of foods and beverages consumed on 2 different days during the week and weekend. Individuals taking over the counter supplements were asked to record the type, dose, and duration of supplement used. The nutrition data were analyzed using the Nutrition Data System, Version 4.01 (Food and Nutrient Database 29, Nutrition Coordinating Center, University of Minnesota, Minneapolis, MN), and the average values from the 4-day food records were used for data analyses. We were particularly concerned with changes in folate, B6, and B12, because these nutrients are known to modulate Hcy metabolism [10]. Statistical analyses The primary analysis was performed using a two-way ANOVA, testing the hypothesis that the change in each outcome (6 months
baseline = change) for all four treatment groups combined was not significantly different from zero. Statistical significance was established at P < 0.05/4 = 0.0125. Paired differences in the change in each outcome for each individual group were then determined, and statistical significance was established at P < 0.05/16 = 0.003. Paired differences between the extremes of training (moderate-intensity – low-frequency vs. high-intensity – high-frequency) with respect to the change in each outcome were also analyzed ( P < 0.05/4 = 0.0125). Finally, a 2  2 factorial ANOVA was used to examine intensity as a factor (adjusting for frequency) and frequency as a factor (adjusting for intensity) ( P < 0.05/4 = 0.0125). Effect size estimates were calculated as the difference between the mean values for the change in a given variable for the two comparison groups, divided by the common standard deviation.

897

Two-tailed t tests were used to evaluate significant within-group changes in the anthropometric, dietary, and exercise adherence variables, and the significance level was adjusted for the number of pairwise comparisons. We also used Pearson product– moment correlation coefficient to examine significant relationships (a = 0.05) among variables. Data are presented as the mean F standard deviation. SAS Version 8 (Cary, NC) and SPSS Version 10.1 (Chicago, IL) were used to perform the stratified randomization and statistical analyses.

Results Table 1 summarizes the descriptive and anthropometric characteristics of the study participants. A total of 397 individuals were randomized to one of four training groups. Pre- and posttraining data were available for 324 participants (approximately 18% dropout rate). Samples for plasma INS were drawn from a pool of participants who agreed to allow us to collect and store samples for future studies (233 samples equally distributed across groups or approximately 72% of the total posttraining sample). The study groups were similar with respect to age, sex distribution ratio, and anthropometric variables (BMI and waist-to-hip ratio). Furthermore, neither BMI nor the waist-to-hip ratio changed within any group during the study. The racial make-up of the final sample was approximately 77% White, 14% African American, 5% Hispanic, 3% Asian American, and 1% other. Exercise adherence, defined as the percentage of prescribed minutes of walking within the prescribed target heart rate training zone, differed significantly ( P < 0.0125) among the groups. Adherence was significantly greater ( P < 0.001) in the moderate-intensity – low-frequency (77%) compared to the high-intensity – low-frequency (48%) and high-intensity – high-frequency groups (45%), but was not different than the moderate-intensity – highfrequency (72%) group. We also calculated total minutes of walking as another indicator of exercise adherence, which also differed ( P < 0.0125) among groups. Total minutes of Table 1 Descriptive and anthropometric characteristics of the study participants

N Sex (F/M %) Age (year) BMI (kg/m2) W:H

HH

HL

MH

ML

83 59/41

77 68/32

92 65/35

72 56/44

49.4 F 8.4

49.4 F 8.7

48.4 F 8.2

48.5 F 8.3

28.5 F 5.6

27.7 F 4.6

27.6 F 4.6

28.4 F 4.8

0.84 F 0.10

0.82 F 0.09

0.83 F 0.10

0.85 F 0.09

BMI indicates body mass index; W:H, waist-to-hip circumference ratio; HH, high intensity – high frequency; HL, high intensity – low frequency; MH, moderate intensity – high frequency; and ML, moderate intensity – low frequency. Data are mean F standard deviation.

898

G.E. Duncan et al. / Preventive Medicine 39 (2004) 894–902

walking was greater ( P < 0.001) in the moderate-intensity – high-frequency group (approximately 116 min/week) than in the moderate-intensity – low-frequency (approximately 73 min/week) and high-intensity – low-frequency (approximately 60 min/week) groups. Similarly, total minutes of walking in the high-intensity –high-frequency group (approximately 105 min/week) was significantly greater ( P < 0.001) than in both the moderate-intensity – low-frequency and high-intensity – low-frequency groups, but was not different from the moderate-intensity – highfrequency group. Values for baseline, 6-month postexercise training and the change in each primary outcome measure are listed in Table 2. Overall, combining all four treatment groups together, the average changes in Hcy (0.64 F 2.44 Amol/ L), INS (
0.26 F 3.93 AU/mL), and VO2max (0.93 F 2.33 mL min
1 kg
1) were all significant ( P < 0.0125), whereas CRP did not change ( P = 0.16) with training. Specific within-group comparisons revealed that Hcy increased significantly in both higher intensity groups, whereas VO2max increased significantly in both higher frequency groups (all P < 0.003). Next, we examined the difference in the change in each outcome measure between the two extremes of training (i.e., moderate-intensity –low-frequency compared to high-intensity – high-frequency). There was a significant ( P < 0.0125) difference between these groups with respect to the change in VO2max (0.36 F 2.10 mL min
1 kg
1 and 1.77 F 2.97 mL min
1 kg
1 for moderate-intensity – low-frequency and Table 2 Metabolic parameters before and after 6 months of exercise training Variable

Group Baseline value

Hcy (Amol/L)y HH HL MH ML CRP (mg/L) HH HL MH ML INS (AU/mL)y HH HL MH ML VO2max HH (mL min
1 HL kg
1)y MH ML

7.73 7.20 7.62 8.27 3.55 3.87 3.25 3.42 7.69 6.92 5.99 6.47 25.35 25.37 25.41 26.88

F F F F F F F F F F F F F F F F

6-Month value 2.02 1.83 3.04 4.25 5.13 5.50 3.36 3.80 6.95 5.56 4.34 4.57 6.77 5.54 5.45 5.68

8.66 8.19 8.09 8.45 3.14 2.84 3.28 3.16 6.48 5.81 6.98 6.75 27.12 25.98 26.40 27.24

F F F F F F F F F F F F F F F F

Change 2.79 2.79 3.68 3.77 3.32 2.78 3.73 3.53 5.42 4.81 6.06 5.78 7.63 5.64 5.64 5.99

0.93 0.98 0.47 0.17
0.41
1.03 0.04
0.27
1.22
1.10 0.99 0.28 1.77 0.61 0.99 0.36

F F F F F F F F F F F F F F F F

2.56* 2.32* 2.74 2.12 3.89 4.80 3.53 2.47 4.63 3.49 4.02 3.58 2.97*,z 2.27 2.01* 2.10z

Hcy indicates homocysteine; CRP, C-reactive protein; INS, insulin; VO2max, maximal aerobic capacity; HH, high intensity – high frequency; HL, high intensity – low frequency; MH, moderate intensity – high frequency; and ML, moderate intensity – low frequency. Data are mean F standard deviation. * Significant within-group changes for individual groups ( P < 0.003). y Significant within-group changes for all four training groups combined ( P < 0.0125). z Significant between-group difference for HH compared to ML ( P < 0.0125).

Table 3 Dietary variables before and after 6 months of exercise training Variable

Group Baseline value 6-Month value Change

A. Select macronutrient variables before and after 6 months of exercise training Energy (Kcal) HH 2071 F 607 1955 F 527
117 F 456 HL 1978 F 614 1878 F 571
100 F 572 MH 1879 F 514 1768 F 543
110 F 407 ML 1948 F 399 2001 F 673 53 F 563 % Fat HH 34.1 F 6.5 36.8 F 8.7 2.7 F 8.3 HL 32.9 F 6.5 34.1 F 6.2 1.2 F 6.7 MH 35.1 F 6.9 35.2 F 7.2 0.1 F 7.8 ML 34.0 F 6.3 33.6 F 7.6
0.4 F 6.9 % Carbohydrates HH 50.5 F 9.2 48.8 F 12.4
1.8 F 9.8 HL 50.2 F 7.5 50.3 F 8.7
1.7 F 7.3 MH 49.0 F 9.2 49.5 F 8.5 0.5 F 8.7 ML 50.0 F 9.2 50.4 F 11.3 0.4 F 7.1 % Protein HH 14.8 F 2.7 14.8 F 3.9 0.0 F 3.0 HL 14.9 F 2.7 16.0 F 3.5 1.0 F 4.0 MH 15.6 F 3.6 15.0 F 2.8
0.6 F 3.7 ML 15.2 F 3.0 15.6 F 3.3 0.4 F 2.4 B. Select micronutrient variables before and after 6 months of exercise training Folate (Ag/d) HH 427 F 232 328 F 185
99 F 210* HL 367 F 201 373 F 231 6 F 221 MH 351 F 185 275 F 135
76 F 152* ML 387 F 205 351 F 215
35 F 172 B6 (mg/d) HH 2.7 F 1.7 2.3 F 1.7
0.4 F 1.2 HL 2.5 F 1.5 3.9 F 8.8 1.4 F 9.0 MH 2.9 F 4.8 1.9 F 1.0
1.1 F 5.0 ML 3.7 F 6.8 2.4 F 1.4
1.3 F 6.4 B12 (Ag/d) HH 7.5 F 6.1 6.7 F 7.5
0.8 F 5.0 HL 6.6 F 6.0 7.4 F 9.5 0.8 F 8.6 MH 6.6 F 5.4 5.4 F 5.8
1.13 F 8.1 ML 7.9 F 8.9 7.8 F 10.1
0.1 F 3.71 HH indicates high intensity – high frequency; HL, high intensity – low frequency; MH, moderate intensity – high frequency; and ML, moderate intensity – low frequency. Data are mean F standard deviation. * Significant within-group changes for individual groups ( P < 0.004).

high-intensity – high-frequency, respectively; effect size estimate = 0.60 mL min
1 kg
1). However, there were no other between-group differences for the remaining outcome measures. Finally, we examined the effect of frequency (while adjusting for intensity) and intensity (while adjusting for frequency) on the change in each outcome measure. There was a significant ( P < 0.0125) main effect for intensity (high-intensity – low-frequency + high-intensity – high-frequency vs. moderate-intensity– low frequency + moderate-intensity– high-frequency) with respect to the change in INS (effect size estimate = 0.46 AU/mL for higher intensity vs. moderate-intensity groups combined). There was also a significant ( P < 0.0125) main effect for frequency (moderate-intensity – high-frequency + high-intensity – high-frequency vs. moderate-intensity – low-frequency + high-intensity – low-frequency) with respect to the change in VO2max (effect size estimate = 0.38 mL min
1 kg
1 for higher frequency vs. lower frequency groups combined).

G.E. Duncan et al. / Preventive Medicine 39 (2004) 894–902

Complete baseline and 6-month dietary records were available for 153 participants (approximately 47% of the total posttraining sample, equally distributed across groups). Select macronutrient (Table 3A) and micronutrient (Table 3B) variables were examined. Self-reported dietary folate intake decreased ( P < 0.004) in both the moderate-intensity – high-frequency and high-intensity – high-frequency groups. There were no other significant changes in any of the other variables.

Discussion Contrary to our hypothesis, we found that 6 months of exercise training increased serum total Hcy in all exercise groups combined. Furthermore, Hcy increased in both higher intensity training groups (high-intensity – low-frequency and high-intensity – high-frequency). The change in Hcy was also positive, but not significant, in both moderateintensity training groups (moderate-intensity – low-frequency and moderate-intensity– high frequency). However, mean Hcy levels were within the normal range (5 –15 Amol/L) before and after training (see Table 2). The clinical significance of small changes in Hcy (1 Amol/L or less), particularly when the levels remain within the normal range, is unknown. Boushey et al. [6] reported that the odds ratio for coronary heart disease was 1.8 and 1.6 in women and men for each 5-Amol/L increase in Hcy above the median, which is a much greater increase than that experienced in our study. The finding that circulating Hcy increased significantly with higher intensity training was unexpected, but is not unprecedented. For example, men randomized to train while inspiring either a normobaric, normoxic, or a normobaric, hypoxic gas mixture had a mean 11% decrease in Hcy after vigorous hypoxic training but a 10% increase in Hcy following vigorous normoxic training [13]. Nevertheless, individuals in the vigorous normoxic group had normal fasting Hcy levels before and after training (6.9 F 1.2 vs. 7.6 F 1.7 Amol/L, respectively) [13]. In contrast, 6 months of brisk walking lowered plasma total Hcy levels (from 10.1 F 3.2 to 7.4 F 2.0 Amol/L) and the waist-to-hip ratio, and increased VO2max, in a group of young, overweight women with polycystic ovary syndrome [31]. Total body weight did not change in that study. Thus, the pre- and posttraining values in both high-intensity training groups in our study (see Table 2), and the average 11% increase in Hcy following training in these groups, are consistent with the findings reported by Bailey et al. [13], but not Randeva et al. [15] who found that exercise training decreased Hcy concentrations. An inverse relationship between mean Hcy levels and the extent of self-reported leisure time activity has been reported in middle-aged to older adults [11]. However, that study only graded individuals into four categories based on self-reported activity. Recently, no relationship between

899

plasma Hcy and VO2max was found in young healthy men following acute exercise [16]. This suggests that selfreported activity may not adequately assess the relationship between Hcy and exercise. Similarly, Hcy and VO2max were not correlated at baseline or at 6 months in our study (data not shown). Together, these data indicate that fasting serum Hcy concentrations and aerobic fitness may not be related. Our participants were not taking medications known to influence Hcy metabolism, and all were screened for CVD and diabetes before study entry. The training groups were similar with respect to age, sex distribution, BMI, and the waist-to-hip ratio (see Table 1); all of which are reported to affect Hcy metabolism. However, neither BMI nor the waist-to-hip ratio was altered in this study. This may have influenced the measured changes in Hcy, because our participants remained overweight throughout the study. Although the overall self-reported intake of folate did not change during this study, folate levels did decrease significantly in both higher frequency training groups (see Table 3B). However, changes in Hcy were not related to changes in folate or either of the B vitamins at baseline or at 6 months (data not shown). Furthermore, using baseline folate and change in folate as covariates, we found no association between these measures and the observed change in Hcy. Because we were not able to directly study changes in specific amino acids or enzymatic activity related to Hcy metabolism, the mechanisms responsible for the increased Hcy concentrations following exercise, particularly with high-intensity training, in the present study are unknown and warrant further investigation. We also found that exercise training failed to alter CRP levels in any of the treatment groups. Fasting serum CRP was highly correlated with BMI before and after a weight loss intervention in obese women [22]. Similarly, we found significant ( P < 0.001) correlations between BMI and CRP both before (r = 0.30) and after (r = 0.27) training. However, in contrast to the study of Heilbronn et al. [22], BMI did not change over the course of the intervention in our participants. Thus, the finding that CRP was not altered with training may be due to the fact that participants remained overweight throughout the duration of the study and that changes in CRP may be at least partially dependent on changes in total body mass, as reported by others [21,25]. None of our participants were taking medications known to effect CRP (e.g., statins), and subgroup analyses for men and women separately with respect to the change in CRP resulted in no significant effect for sex (data not shown). Although increased levels of self-reported physical activity were associated with reduced CRP concentrations in a healthy elderly population [25], using multivariate regression models, this association was at least partially explained by BMI. We also found that increased levels of VO2max were associated ( P < 0.001) with reduced CRP levels both before (r =
0.32) and after (r =
0.35) training. Serum

900

G.E. Duncan et al. / Preventive Medicine 39 (2004) 894–902

CRP levels were reported to decrease in 10 of 12 moderately trained runners following 9 months of increased running mileage [26]. This finding is consistent with the observation that serum CRP levels decreased following 3 months of supervised exercise training in patients with intermittent claudication [27] and that athletes have lower CRP concentrations than controls [24]. However, to our knowledge, ours is the first study to prospectively determine the effect of exercise on CRP levels in a large group of previously sedentary adults. Thus, exercise training without weight loss or dietary modification does not appear to be effective in reducing CRP levels in overweight adults. When all training groups were combined, our findings support the hypothesis that both moderate- and higher intensity aerobic exercise significantly decreases fasting plasma INS levels. However, there were no significant changes in INS within any individual treatment group. On the other hand, there was a significant main effect for intensity with respect to the change in INS. The average difference for the change in INS in the combined higher intensity compared to the moderate-intensity groups was
1.80 AU/mL, supporting the notion that higher intensity training is more effective than moderate-intensity training in reducing fasting plasma INS concentrations. Fasting plasma INS levels decreased from approximately 14 to 12 AU/mL after 6 months of a combination aerobic training program in 31 premenopausal, obese women [2]. In contrast, fasting INS levels did not change in 49 obese, middle-aged, and older men after 9 months of moderate- to high-intensity training [35] or in six young men after 1 year of mild jogging [3]. Importantly, mean BMI levels did not change in any of these studies. The modest decreases in fasting plasma INS noted in our study are thus reasonable, based on findings that neither BMI nor the waist-to-hip ratio changed with training and consistent with the findings of the studies noted above [2,3,35]. Changes in fasting plasma INS would have likely been much greater had there also been a concomitant weight loss, as suggested by others [2,35]. Nonetheless, our results related to changes in fasting plasma INS suggest that aerobic training in the absence of weight loss does have beneficial effects on markers of carbohydrate metabolism, consistent with recent findings by our group [36]. The overall change in VO2max was significant when all four treatment groups were combined. We also found significant within-group changes in VO2max in both higher frequency groups. These findings support our hypothesis that both moderate- and higher intensity aerobic training significantly improves VO2max. Similarly, there was a significant main effect for frequency with respect to the change in VO2max, indicating that frequent exercise is more effective than less frequent exercise in improving VO2max. However, between-group differences for VO2max reached significance only in the high-intensity – high-frequency compared to the moderate-intensity –high-frequency groups, representing the two extremes of training used in our study.

These findings support our hypothesis that frequent bouts of higher intensity training result in greater improvements in VO2max compared to less frequent bouts of moderateintensity training and also support the dose –response concept in that frequent bouts of higher intensity exercise appear to be the ‘‘optimal’’ training format for improving fitness. There was a nearly fivefold greater increase in fitness in the high-intensity– high-frequency compared to the moderate-intensity – low-frequency group. This study has several potential limitations. For example, our measure of exercise adherence was derived from self-reported exercise logs. Exercise adherence, defined as the percentage of prescribed minutes of walking within the prescribed target heart rate training zone, was particularly poor in the two high-intensity groups. This is consistent with the notion that walking prescribed at a moderate intensity results in better adherence than walking prescribed at a higher intensity in previously sedentary adults [37]. The poorer adherence in the higher intensity groups may have also been due in part to the higher rates of selfreported exercise-related injuries in these groups, as reported by us previously [37]. Furthermore, the overall levels of exercise completed, in terms of total minutes of walking, fell below the amounts prescribed. However, an intent-to-treat approach was used in the analyses of data from the exercise logs. Missing data were treated in a conservative fashion with the assumption that if a participant did not record an exercise bout in his or her log, the exercise did not occur. Thus, it is possible that we underestimated the actual levels of exercise and thus adherence in our participants. Another potential limitation is that our sample was 78% Caucasian, and the overall level of education was relatively high (mean = 15.9 years completed). Thus, our findings may not be generalizable to individuals of other ethnic or racial backgrounds and education levels. This study also has several strengths including (1) the use of a randomized prospective study design, (2) large sample size of sedentary and overweight adults, (3) the manipulation of both exercise intensity and frequency, (4) the use of individualized target heart rate training zone exercise prescriptions based on maximal exercise testing, (5) the use of heart rate monitors to gauge exercise intensity, and (6) a conservative approach to the treatment of missing data. In conclusion, our data indicate that higher intensity exercise is more effective than moderate-intensity exercise in reducing fasting INS, while frequent exercise is more effective than less frequent exercise in improving VO2max, in a representative sample of previously sedentary, middleaged adults. Exercise did not alter CRP levels, and higher intensity exercise elevated serum Hcy concentrations. Thus, although frequent bouts of higher intensity aerobic exercise were particularly effective in reducing fasting plasma INS levels and improving aerobic fitness, they may result in increased serum Hcy levels. Together, our data indicate a

G.E. Duncan et al. / Preventive Medicine 39 (2004) 894–902

potential difference in the ability of exercise to modulate traditional CVD risk factors (e.g., fasting INS and cholesterol concentrations, blood pressure, etc.) compared to relatively emerging ones (e.g., serum total Hcy and CRP levels). A final caveat is that our conclusions are based on the exercise prescriptions that the participants were assigned, when in reality, the measured responses are the result of the exercise that was actually performed. Thus, it is possible that individuals who adhere closely to a prescription of frequent bouts of higher intensity walking exercise will derive even greater benefits than were measured in the present study.

Acknowledgments We are deeply grateful to our late colleague Michael L. Pollock for his inspiration and guidance in the development of this work. We are also grateful for the assistance of several colleagues who contributed to the completion of this large clinical study: Abbie Beacham, Nichole E. Berlant, Natalie Blevens, Valerie Bonk, Antoinette Chiara, Joyce Corsica, Diego de Hoyos, Kimberly Dempsey, Patricia E. Durning, Gretchen Forbes, Abdel Gabr, Cheryl Griswold, Aisha Kazi, William F. Kanasky, Jr., Timothy U. Ketterson, Mark Lui, Kimberly McKessey, Anna Moore, Robert L. Newton, Jr., Sheila Rauch, Sumner J. Sydeman, Robyn Wallace, Gretchen Wolff, and Weing Zhao. The statistical analyses were performed by Jon Shuster, Ph.D., and Douglas Theriaque, M.S., with the Biostatistical Core of the GCRC. This study was supported by NIH R01 HL58873 (MGP), American Heart Association/Florida – Puerto Rico Affiliate Postdoctoral Fellowship (9920174V) (GED), American College of Sports Medicine Foundation/Polar Research grant (GED), and General Clinical Research Center grant RR00082.

References [1] Dunn AL, Marcus BH, Kampert JB, Garcia ME, Kohl III HW, Blair SN. Reduction in cardiovascular disease risk factors: 6-month results from project active. Prev Med 1997;26:883 – 92. [2] Lamarche B, Despres JP, Pouliot MC, et al. Is body fat loss a determinant factor in the improvement of carbohydrate and lipid metabolism following aerobic exercise training in obese women? Metabolism 1992;41:1249 – 56. [3] Oshida Y, Yamanouchi K, Hayamizu S, Sato Y. Long-term mild jogging increases insulin action despite no influence on body mass index or VO2max. J Appl Physiol 1992;66:2206 – 10. [4] Blair SN, Kohl HW, Barlow CE, Paffenbarger Jr RS, Gibbons LW, Macera CA. Changes in physical fitness and all-cause mortality: a prospective study of healthy and unhealthy men. JAMA 1995;273: 1093 – 8. [5] Paffenbarger Jr RS, Hyde RT, Wing AL, Lee I, Jung DL, Kampert JB. The association of changes in physical-activity level and other lifestyle characteristics with mortality among men. N Engl J Med 1993;328:538 – 45.

901

[6] Boushey CJ, Beresford SAA, Omenn GS, Motulsky AG. A quantitative assessment of plasma homocysteine as a risk factor for vascular disease. Probable benefits of increasing folic acid intakes. JAMA 1995;274:1049 – 57. [7] Nygard O, Nordrehaug JE, Refsum H, Ueland PM, Farstad M, Vollset SE. Plasma homocysteine levels and mortality in patients with coronary artery disease. N Engl J Med 1997;337:230 – 6. [8] Perry IJ, Refsum H, Morris RW, Ebrahim SB, Ueland PM, Shaper AG. Prospective study of serum total homocysteine concentration and risk of stroke in middle-aged British men. Lancet 1995;346:1395 – 8. [9] The Homocysteine Studies Collaboration. Homocysteine and risk of ischemic heart disease and stroke: a meta-analysis. JAMA 2002; 288:2015 – 22. [10] Fonseca V, Guba SC, Fink LM. Hyperhomocysteinemia and the endocrine system: implications for atherosclerosis and thrombosis. Endocr Rev 1999;20:738 – 59. [11] Nygard O, Vollset SE, Refsum H, et al. Total plasma homocysteine and cardiovascular risk profile. The Hordaland homocysteine study. JAMA 1995;274:1526 – 33. [12] Ali A, Mehra MR, Lavie CJ, et al. Modulatory impact of cardiac rehabilitation on hyperhomocysteinemia in patients with coronary artery disease and ‘‘normal’’ lipid levels. Am J Cardiol 1998;82:1543 – 5. [13] Bailey DM, Davies B, Baker J. Training in hypoxia: modulation of metabolic and cardiovascular risk factors in men. Med Sci Sports Exercise 2000;32:1058 – 66. [14] DeCree C, Malinow MR, van Kranenburg GP, Geurten PG, Longford NT, Keizer HA. Influence of exercise and menstrual cycle phase on plasma homocyst(e)ine levels in young women—A prospective study. Scand J Med Sci Sports 1999;9:272 – 8. [15] Randeva HS, Lewandowski KC, Drzewoski J, et al. Exercise decreases plasma total homocysteine in overweight young women with polycystic ovary syndrome. J Clin Endocrinol Metab 2002; 87:4496 – 501. [16] Wright M, Francis K, Cornwell P. Effect of acute exercise on plasma homocysteine. J Sports Med Phys Fitness 1998;38:262 – 5. [17] Pepys MB. C-reactive protein fifty years on. Lancet 1981;1:653 – 7. [18] Koenig W, Sund M, Frohlich M, et al. C-reactive protein, a sensitive marker of inflammation, predicts future risk of coronary heart disease in initially healthy middle-aged men: results from the MONICA (Monitoring Trends and Determinants in Cardiovascular Disease) Augsburg Cohort Study, 1984 – 1992. Circulation 1999;99: 237 – 42. [19] Kuller LH, Tracy RP, Shaten J, Meilahn EN. Relation of C-reactive protein and coronary heart disease in the MRFIT nested case-control study: Multiple Risk Factor Intervention Trial. Am J Epidemiol 1996; 144:537 – 47. [20] Festa A, D’Agostino Jr R, Howard G, Mykkanen L, Tracy RP, Haffner SM. Chronic subclinical inflammation as part of the insulin resistance syndrome: the Insulin Resistance Atherosclerosis Study (IRAS). Circulation 2000;102:42 – 7. [21] Mohamed-Ali V, Goodrick S, Rawesh A, et al. Subcutaneous adipose tissue releases interleukin-6, but not tumor necrosis factor-alpha, in vivo. J Clin Endocrinol Metab 1997;82:4196 – 200. [22] Heilbronn LK, Noakes M, Clifton PM. Energy restriction and weight loss on very-low-fat diets reduce C-reactive protein concentrations in obese, healthy women. Arterioscler, Thromb, Vasc Biol 2001;21: 968 – 70. [23] Tchernof A, Nolan A, Sites CK, Ades PA, Poehlman ET. Weight loss reduces C-reactive protein levels in obese postmenopausal women. Circulation 2002;105:564 – 9. [24] Dufaux B, Order U, Greyer H, Hollmann W. C-reactive protein serum concentrations in well-trained athletes. Int J Sports Med 1984;5: 102 – 6. [25] Geffken D, Cushman M, Burke G, Polak J, Sakkinen P, Tracey R. Association between physical activity and markers of inflammation in a healthy elderly population. Am J Epidemiol 2001;153:242 – 50. [26] Mattusch F, Dufaux B, Heine O, Mertens I, Rost R. Reduction of the

902

[27]

[28]

[29]

[30]

[31]

G.E. Duncan et al. / Preventive Medicine 39 (2004) 894–902 plasma concentration of C-reactive protein following nine months of endurance training. Int J Sports Med 2000;21:21 – 4. Tisi PV, Hulse M, Chulakadabba A, Gosling P, Shearman CP. Exercise training for intermittent claudication: does it adversely affect biochemical markers of the exercise-induced inflammatory response? Eur J Vasc Endovasc Surg 1997;14:344 – 50. Pate RR, Pratt M, Blair SN, et al. Physical activity and public health. A recommendation from the Centers for Disease Control and Prevention and the American College of Sports Medicine. JAMA 1995; 273:402 – 7. U.S. Department of Health and Human Services. Physical Activity and Health: a report of the Surgeon General. Atlanta: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Centers for Chronic Disease Prevention and Health Promotion; 1996. p. 28. Kohl III HW, Blair SN, Paffenbarger Jr RS, Macera CA, Kronenfeld JJ. A mail survey of physical activity habits as related to measured physical fitness. Am J Epidemiol 1988;127:1228 – 39. Sallis JF, Haskell WL, Wood PD, et al. Physical activity assessment methodology in the Stanford five-city project. Am J Epidemiol 1985;121:91 – 106.

[32] Martin SC, Hilton AC, Bartlett WA, Jones AF. Plasma total homocysteine measurement by ion-paired reversed-phase HPLC with electrochemical detection. Biomed Chromatogr 1999;13:81 – 2. [33] American College of Sports Medicine. ACSM’s Guidelines for Exercise Testing and Prescription. fifth ed. Baltimore: Williams and Wilkins; 1995. p. 158 – 60. [34] Kohl III HW, Dunn AL, Marcus BH, Blair SN. A randomized trial of physical activity interventions: design and baseline data from project Active. Med Sci Sports Exerc 1998;30:275 – 83. [35] Katzel LI, Bleecker ER, Colman EG, Rogus EM, Sorkin JD, Goldberg AP. Effects of weight loss vs. aerobic exercise training on risk factors for coronary disease in healthy, obese, middle-aged and older men. JAMA 1995;274:1915 – 21. [36] Duncan GE, Hutson AD, Perri MG, Eckel RH, Stacpoole PW. Exercise training, without weight loss, increases insulin sensitivity and postheparin plasma lipase activity in previously sedentary adults. Diabetes Care 2003;26:557 – 62. [37] Perri MG, Anton SD, Durning PE, et al. Adherence to exercise prescriptions: effects of prescribing moderate versus higher levels of intensity and frequency. Health Psychol 2002;21: 452 – 8.