Exercise training and artery function in humans: nonresponse and its relationship to cardiovascular risk factors
Daniel J. Green, Thijs Eijsvogels, Yvette M. Bouts, Andrew J. Maiorana, Louise H. Naylor, Ralph R. Scholten, Marc E. A. Spaanderman, Christopher J. A. Pugh, Victoria S. Sprung, Tim Schreuder, Helen Jones, Tim Cable, Maria T. E. Hopman and Dick H. J. Thijssen J Appl Physiol 117:345-352, 2014. First published 19 June 2014; doi:10.1152/japplphysiol.00354.2014 You might find this additional info useful... This article cites 38 articles, 16 of which can be accessed free at: /content/117/4/345.full.html#ref-list-1 Updated information and services including high resolution figures, can be found at: /content/117/4/345.full.html Additional material and information about Journal of Applied Physiology can be found at: http://www.the-aps.org/publications/jappl
Journal of Applied Physiology publishes original papers that deal with diverse areas of research in applied physiology, especially those papers emphasizing adaptive and integrative mechanisms. It is published 12 times a year (monthly) by the American Physiological Society, 9650 Rockville Pike, Bethesda MD 20814-3991. Copyright © 2014 by the American Physiological Society. ISSN: 0363-6143, ESSN: 1522-1563. Visit our website at http://www.the-aps.org/.
Downloaded from on August 18, 2014
This information is current as of August 18, 2014.
J Appl Physiol 117: 345–352, 2014. First published June 19, 2014; doi:10.1152/japplphysiol.00354.2014.
Exercise training and artery function in humans: nonresponse and its relationship to cardiovascular risk factors Daniel J. Green,1,2 Thijs Eijsvogels,3,4 Yvette M. Bouts,1,3 Andrew J. Maiorana,5,6 Louise H. Naylor,1 Ralph R. Scholten,3 Marc E. A. Spaanderman,3 Christopher J. A. Pugh,2 Victoria S. Sprung,2 Tim Schreuder,3 Helen Jones,2 Tim Cable,2 Maria T. E. Hopman,3 and Dick H. J. Thijssen2,3 1
School of Sports Science, Exercise and Health, The University of Western Australia, Nedlands, Australia; 2Research Institute for Sport and Exercise Sciences, Liverpool John Moores University, Liverpool, United Kingdom; 3Department of Physiology, Radboud University Medical Center, Nijmegen, The Netherlands; 4Henry Low Heart Center, Department of Cardiology, Hartford Hospital, Hartford, Connecticut; 5School of Physiotherapy and Exercise Science, Curtin University, Perth, Australia; and 6Advanced Heart Failure and Cardiac Transplant Service, Royal Perth Hospital, Perth, Australia Submitted 21 April 2014; accepted in final form 17 June 2014
cardiovascular risk; nitric oxide; flow-mediated dilation; endothelial function; physical activity CARDIOVASCULAR DISEASE (CVD) remains the world’s leading cause of mortality. Physical activity and/or exercise training represent potent strategies to reduce the risk of future
Address for reprint requests and other correspondence: D. J. Green, School of Sports Science, Exercise and Health, The Univ. of Western Australia, Nedlands, Australia (e-mail: [email protected]). http://www.jappl.org
cardiovascular events in asymptomatic subjects and those with preexisting disease (20, 21, 24). While changes in traditional cardiovascular risk factors explain some of the cardioprotective effects of training (24), a substantial proportion of the benefit remains unexplained, and direct effects of exercise on the arterial wall may contribute to this “risk factor gap” (12, 17). Despite the overall health benefits of exercise training, recent studies have described heterogeneous adaptations to training (28). In subjects who undertook similar exercise training interventions, some demonstrated large improvements in parameters such as cardiopulmonary fitness, blood pressure, and cholesterol, while others exhibited smaller increases or even “adverse” responses (4, 5). Such findings are consistent with anecdotal clinical observations that some individuals only respond modestly to exercise. No previous study has explored the heterogeneity of changes in artery function or health in responses to exercise training, or tried to identify predictors of training-mediated adaptations in arterial function. A widely accepted index of artery function and health is flow-mediated dilation (FMD%), the vasodilator response to an imposed shear stress stimulus following cuff-induced increases in blood flow. Recent studies indicate that brachial artery FMD% is a surrogate for assessment of coronary artery endothelial function (30) which predicts cardiovascular outcomes in humans (11, 15, 26). The first aim of this study was therefore to explore the proportion of “responders” and “nonresponders,” in terms of change in artery function (FMD%), following exercise training in humans. Second, we examined whether a priori characteristics or physiological measures can predict “response” vs. “nonresponse.” Finally, we examined whether changes in endothelial function as a result of exercise training correlate with changes in other (traditional) cardiovascular risk factors. For this purpose, we pooled data from 13 studies, involving 182 subjects who undertook exercise training in our laboratories. These studies all involved individualized, center-based exercise prescription, were supervised by exercise scientists, and utilized identical vascular function analysis methodology. We hypothesized that exercise training would induce a heterogeneous effect on FMD%, which might partly be explained by subject and exercise characteristics, and that training-induced changes in FMD% would not correlate with changes in traditional cardiovascular risk factors.
8750-7587/14 Copyright © 2014 the American Physiological Society
345
Downloaded from on August 18, 2014
Green DJ, Eijsvogels T, Bouts YM, Maiorana AJ, Naylor LH, Scholten RR, Spaanderman ME, Pugh CJ, Sprung VS, Schreuder T, Jones H, Cable T, Hopman MT, Thijssen DHJ. Exercise training and artery function in humans: nonresponse and its relationship to cardiovascular risk factors. J Appl Physiol 117: 345–352, 2014. First published June 19, 2014; doi:10.1152/japplphysiol.00354.2014.—The objectives of our study were to examine 1) the proportion of responders and nonresponders to exercise training in terms of vascular function; 2) a priori factors related to exercise training-induced changes in conduit artery function, and 3) the contribution of traditional cardiovascular risk factors to exercise-induced changes in artery function. We pooled data from our laboratories involving 182 subjects who underwent supervised, large-muscle group, endurance-type exercise training interventions with pre-/posttraining measures of flowmediated dilation (FMD%) to assess artery function. All studies adopted an identical FMD protocol (5-min ischemia, distal cuff inflation), contemporary echo-Doppler methodology, and observerindependent automated analysis. Linear regression analysis was used to identify factors contributing to changes in FMD%. We found that cardiopulmonary fitness improved, and weight, body mass index (BMI), cholesterol, and mean arterial pressure (MAP) decreased after training, while FMD% increased in 76% of subjects (P ⬍ 0.001). Training-induced increase in FMD% was predicted by lower body weight ( ⫽ ⫺0.212), lower baseline FMD% ( ⫽ ⫺0.469), lower training frequency ( ⫽ ⫺0.256), and longer training duration ( ⫽ 0.367) (combined: P ⬍ 0.001, r ⫽ 0.63). With the exception of a modest correlation with total cholesterol (r ⫽ ⫺0.243, P ⬍ 0.01), changes in traditional cardiovascular risk factors were not significantly related to changes in FMD% (P ⬎ 0.05). In conclusion, we found that, while some subjects do not demonstrate increases following exercise training, improvement in FMD% is present in those with lower pretraining body weight and endothelial function. Moreover, exercise training-induced change in FMD% did not correlate with changes in traditional cardiovascular risk factors, indicating that some cardioprotective effects of exercise training are independent of improvement in risk factors.
346
Effect of Training on Artery Function
•
Green DJ et al.
Experimental Procedures
Subjects
Subject characteristics. Body mass and stature were recorded using standard methods. Blood pressure was measured after a ⱖ5-min resting period in the supine position (as this measure was typically performed before assessment of the brachial artery FMD) using a manual sphygmomanometer or an automated device (Dinamap, Tampa, FL) by a well-trained researcher. Independent of the method used, blood pressure was measured twice and averaged. MAP was calculated using the diastolic and systolic blood pressures. In all studies we examined levels of fasting blood glucose and total cholesterol. Although methods for these analyses differed between studies, all laboratory tests were undertaken in clinical centers in which robust validation and reproducibility testing is performed. Furthermore, for any given individual, lab tests were performed using identical approaches. These blood samples were all obtained from an antecubital vein. Cardiopulmonary fitness. To obtain sport-specific cardiopulmonary ˙ O2 peak) was measured using fitness data, peak oxygen consumption (V a treadmill (n ⫽ 77), cycle ergometer (n ⫽ 86), or a rowing ergometer (n ⫽ 19), according to the training undertaken by the subjects. Repeated measures were always collected using the same modality and protocols. In all cases, exercise was performed until volitional fatigue, according to standard indications for performing a successful exercise test (23). Gas analyzers and flow probes were calibrated before each test. Oxygen consumption was expressed relative to body weight (ml·kg⫺1·min⫺1). Peak oxygen consumption was calculated from data in the 15– 60 s before volitional exhaustion. Brachial artery endothelial function. FLOW-MEDIATED DILATION. Flow-mediated dilation (FMD%) examines brachial artery endothelium-dependent dilation. All FMD-procedures were performed in line with recent guidelines (31) and all studies used the same, observerindependent edge detection and wall tracking analysis software to analyze arterial diameter (38). In all studies, subjects were asked to fast for ⬎4 h prior to each visit and to avoid caffeine, alcohol, and exercise for ⬎12 h. We ensured that pre- and posttraining measures in premenopausal women were performed during the first phase of the menstrual cycle, given the impact of the menstrual cycle on menstrual cycle on vascular function (36). All studies were conducted in a quiet, temperature-controlled environment. Each visit for a given subject
Exercise training studies performed in our laboratories in the past 20 years were eligible for inclusion. Criteria for inclusion were 1) study performed in our laboratory using comparable methodological approaches (31), including operator-independent analysis with edge detection and wall tracking software; 2) availability of pretraining subject characteristics; 3) successful completion of exercise training, involving supervised, center-based, endurance-type exercise of the large muscle groups; 4) exercise training performed ⱖ2 times/wk for ⱖ8 wk and ⱕ18 wk; 5) brachial artery FMD% measured before and after exercise training; 6) brachial artery FMD% assessed according to best practice recommendations (31); 7) cardiopulmonary fitness measured using gas exchange analysis as the peak oxygen consumption during a sport-specific maximal exercise test; 8) exercise training performed in able-bodied adult subjects; 9) study procedures approved by the local ethics committee and adherent to the Declaration of Helsinki; and 10) pre- and posttraining subject characteristics available [age, body weight, cardiopulmonary fitness, body mass index (BMI)] including traditional cardiovascular risk factors [BMI, mean arterial pressure (MAP), cholesterol, blood pressure, glucose]. Data collected using nonoptimal analysis methods, as well as those with incomplete data sets for our primary outcome parameter (i.e., brachial artery FMD%) or any of the potential moderating variables, were excluded. None of our experiments involved smokers. A flow chart (Fig. 1) provides insight into the inclusion and exclusion of our previous exercise training studies. This resulted in the collection of data from 13 exercise training studies, involving 182 subjects. Study Design In all studies, subject characteristics, cardiopulmonary fitness, and brachial artery FMD% were measured before and after exercise training. Posttraining measurements were performed within 3– 4 days of the last exercise bout, but not within 24 h of the last exercise bout. A number of studies included in our pooled analysis performed repeated measurements collected every 2 wk across an 8-wk exercise training intervention to describe time-dependent changes in brachial artery FMD%. For these studies, we only included the pre- and posttraining values.
Fig. 1. Flow chart on all exercise training studies performed in the last 15 years (1998 –2013) in our laboratories and the reason for exclusion of individual studies. FMD, flow-mediated dilation.
J Appl Physiol • doi:10.1152/japplphysiol.00354.2014 • www.jappl.org
Downloaded from on August 18, 2014
METHODS
Effect of Training on Artery Function
Statistical Analysis All analyses were performed using the Statistical Package for the Social Sciences (IBM SPSS Statistics for Windows, Version 20.0. Armonk, NY). Data are presented as means ⫾ SD. Normality distribution was examined using a Kolmogorov-Smirnov test. In case of non-Gaussian distribution, log-transformation was performed and the data were reexamined for normality distribution. The effect of exercise training was determined with a paired Student’s t-test. Change in endothelial function was defined as the difference in FMD% between pre- and postexercise training measurements. FMD% is the most frequently adopted method to present changes in conduit artery function and health in humans. Based on the 25th, 50th, and 75th percentile cut-points, four quartiles were created to compare differences between nonresponders (Q1), low responders (Q2), moderate responders (Q3), and high responders (Q4). One-way analysis of variance was used to assess differences across quartiles for continuous variables, while a chi-squared test was used for categorical variables (e.g., sex, cardiovascular disease status, medication use). Bonferroni correction was applied to post hoc testing to identify differences across quartiles. A backward linear regression analysis was performed to identify predictors of training-induced changes in FMD%. The pre-/posttraining change in FMD% was entered as the dependent (continuous) variable. Parameters that demonstrated a difference between the quartiles with a P value ⬍ 0.10 were included in our model as ˙ O2 peak, training independent variables: sex, height, weight, MAP, V mode, training duration, training frequency, training intensity, pretraining FMD%, and GTN%. To validate potential predictors of
Green DJ et al.
347
changes in FMD%, we performed a binary logistic regression analysis with responder level [including nonresponders (Q1) and high responders (Q4) only, n ⫽ 90] as our dependent variable. Similar parameters were included in this secondary model. Finally, the effects of exercise training on (traditional) cardiovascular risk factors were determined with a two-way repeated measurements analysis of variance to examine whether the effect of training (i.e., main effect “time”) differed between the quartiles (i.e., main effect “Q”). Post hoc paired Student’s t-tests were performed per quartile in case of an interaction effect (Q ⫻ time). The relationship between changes in endothelial function and changes in cardiovascular risk factors was assessed using Pearson’s correlation coefficient. RESULTS
All subjects (n ⫽ 182) successfully completed the exercise training program. On average, subjects demonstrated a significant decrease in weight (86.5 ⫾ 18.1 to 85.9 ⫾ 17.9 kg, P ⬍ 0.001), BMI (27.9 ⫾ 4.9 to 27.7 ⫾ 4.8 kg/m2, P ⬍ 0.001), MAP (91 ⫾ 12 to 88 ⫾ 12 mmHg, P ⬍ 0.001), and total cholesterol levels (4.9 ⫾ 1.2 to 4.7 ⫾ 1.0 mmol/l, P ⫽ 0.014), ˙ O2 peak (31.6 ⫾ 12.8 to 34.3 ⫾ 12.9 ml·kg⫺1·min⫺1, while V P ⬍ 0.001) and brachial artery FMD% (5.0 ⫾ 3.2 to 7.0 ⫾ 3.8%, P ⬍ 0.001) increased. We observed no significant change in blood glucose levels (6.0 ⫾ 2.5 to 5.8 ⫾ 2.0 mmol/l, P ⫽ 0.107) or resting brachial artery diameter (3.9 ⫾ 0.8 to 3.9 ⫾ 0.8, P ⫽ 0.241). Responders vs. Nonresponders Exercise training resulted in an average increase in brachial artery FMD% of 1.8 ⫾ 3.1% (P ⬍ 0.001). Interestingly, 76% of our study population exhibited improved FMD, whereas 24% of the subjects demonstrated a decrease in brachial artery FMD% (Fig. 2). Quartile 1 consisted of subjects who demonstrated no change or a decrease in FMD% after training only (i.e., Q1; nonresponders), while the other three quartiles consisted of subjects with an increase in FMD% after training of 0.1–1.4% (Q2; low responders), 1.5–3.5% (Q3; moderate responders), and ⬎3.6%, (Q4; high responders), respectively. Differences in subject characteristics of the quartiles are presented in Table 1. Predictors of Changes in FMD% The backward regression analysis excluded sex, height, ˙ O2 peak, training mode, training intensity, and GTN% MAP, V from the final model as these parameters did not significantly contribute to changes in FMD%. For the remaining parameters, we found that an increase in FMD% was predicted by lower pretraining FMD% (P ⬍ 0.001,  ⫽ ⫺0.469), weight (P ⫽ 0.003,  ⫽ ⫺0.212), and training frequency (P ⫽ 0.001,  ⫽ ⫺0.256), and longer training duration (P ⬍ 0.001,  ⫽ 0.367) (r ⫽ 0.63, P ⬍ 0.001). Backward binary logistic regression analysis confirmed these findings and identified similar predictors for identifying high responders (P ⬍ 0.001, Nagelkerke r2 ⫽ 0.68). Relationship Between Changes in FMD% and Changes in Cardiovascular Risk Factors When comparing the adaptations in the traditional cardiovascular risk factors after exercise training between the four quartiles of FMD% change, we found similar adaptations in ˙ O2 peak across groups weight, BMI, MAP, cholesterol, and V
J Appl Physiol • doi:10.1152/japplphysiol.00354.2014 • www.jappl.org
Downloaded from on August 18, 2014
was performed at the same time of day when measures were repeated posttraining (between 8 AM and 4 PM), to avoid the potential impact of diurnal variation on FMD% (16). Furthermore, posttraining assessment of FMD% was performed ⬎24 h after the last training session to prevent potential influence of the last training session. To examine brachial artery FMD, the arm was extended and positioned at an angle of ⬃80° from the torso. A rapid inflation and deflation pneumatic cuff (D.E. Hokanson, Bellevue, WA) was positioned on the forearm, immediately distal to the olecranon process to provide a stimulus to forearm ischemia. Multifrequency (7.5–10 MHz) array probes, attached to high-resolution ultrasound machines, were used to image the brachial artery in the distal one-third of the upper arm. When an optimal image was obtained, the probe was held stable and the ultrasound parameters were set to optimize the longitudinal, B-mode image of the lumen-arterial wall interface. Continuous Doppler velocity assessments were also obtained using the ultrasound and were collected using the lowest possible insonation angle (always ⬍60°). Following baseline assessments, a forearm cuff was inflated (⬎200 mmHg) for 5 min. Diameter and flow recordings resumed 30 s prior to cuff deflation and continued for 3 min thereafter, in accordance with recent technical specifications (3, 38). An experienced sonographer performed the measurements, with the same sonographer performing both pre- and posttraining measurements within each subject. GLYCERYL TRINITRATE. Using a sublingual dose of glyceryl trinitrate (GTN), we recorded the diameter response for ⬎5 min to examine endothelium-independent dilation of the brachial artery. This test assesses whether training-induced changes in FMD% can be attributed to changes in endothelium-independent mechanisms. BRACHIAL ARTERY DIAMETER AND BLOOD FLOW ANALYSIS. In all studies included in this pooled analysis, we performed analysis of brachial artery diameter using edge-detection and wall-tracking software, which is largely independent of investigator bias. Previous papers contain detailed descriptions of our analysis approach (3, 38). This automated software provides more reproducible and valid results than manual methods, reduces observer error significantly, and possesses an intra-observer CV of 6.7% (38).
•
348
Effect of Training on Artery Function
•
Green DJ et al.
(Table 2). Interestingly, the direction of change in baseline arterial diameter significantly differed between quartiles (Q ⫻ time: P ⫽ 0.015), with post hoc analysis revealing an increase in diameter after training in Q1, but no changes in Q2–Q4 (Table 2). As a consequence of using the change in FMD% to characterize the quartiles, the direction and magnitude of change in FMD% differed between groups (Table 2). Pearson’s correlation coefficients between changes in FMD% and changes in cardiovascular parameters are presented in Table 3. We found a significant, inverse correlation between the change in FMD% and the changes in diameter and cholesterol (Table 3), while changes in the other (traditional) cardiovascular risk factors did not relate to change in FMD%. DISCUSSION
J Appl Physiol • doi:10.1152/japplphysiol.00354.2014 • www.jappl.org
Downloaded from on August 18, 2014
Fig. 2. Data of all individual subjects (n ⫽ 182) on the change in brachial ˙ O2 peak, artery flow-mediated dilation (FMD%, A), cardiopulmonary fitness (V B), and cholesterol (in mmol/l, C) after supervised, endurance-type exercise training. Individual data are organized based on the magnitude of change, with the largest deterioration (i.e., data in red) presented on the left and the largest improvement (i.e., data in green) on the right end of the spectrum. Based on the 25th, 50th, and 75th percentile cut points, 4 quartiles were created to compare differences between nonresponders (Q1 red), low responders (Q2 orange), moderate responders (Q3 gold), and high responders (Q4 green). These color-based FMD% quartiles for change are maintained in the bottom panels, visually illustrating the lack of consistency between change in the respective variables.
Our grouped analysis reveals substantial heterogeneity in the impact of exercise training on FMD%, with 76% of the subjects studied showing some improvement. Larger improvements in FMD% were linked to longer exercise training interventions and lower cardiopulmonary fitness and baseline FMD%. These factors explained ⬃39% of the variation in adaptation of artery function. Finally, the exercise traininginduced change in FMD% did not correlate with changes in traditional cardiovascular risk factors, except for a modest correlation with change in total cholesterol. These findings confirm our previous finding in a much smaller subgroup (13) and suggest that some component of the cardioprotective effect of exercise training occurs independently of change in traditional cardiovascular risk factors and may be related to improvements in artery function. We found that ⬃24% of the subjects demonstrated no change in endothelial function with training. Although no previous study has reported the heterogeneity in arterial adaptation to exercise training, recent publications by Bouchard and colleagues have raised the general concept of a lack of responsiveness to exercise training (4 – 6). The authors suggested that training may have an “adverse” impact on some risk factors (e.g., blood pressure, triglycerides, HDL), in certain individuals (4, 5). Whether such adverse effects translate into poor clinical outcomes is unknown, especially since CVD risk assessment is typically undertaken by examining a cluster of biomarker and/or risk factors. Nonetheless, these observations demonstrate that the effects of training are not simply unidirectional, an important message when evaluating the impact of training programs at individual or group levels. Despite the strong predictive capacity of FMD% (11, 26), we do not interpret our observation of decline in FMD% in 24% of the subjects as evidence that exercise training increased cardiovascular risk. This group of nonresponders also demonstrated increases in resting brachial artery diameter after training, an effect that was not apparent in the other FMD% quartiles. This raises the possibility that exercise training induced arterial adaptation in all subjects, but of distinct forms. It is now known that exercise training induces changes in both artery function and structure and that these occur according to a distinct time course (19). We (2, 33, 34) and others (14) have demonstrated that initial improvements in function are superseded by compensatory changes in artery size. It is therefore possible that the FMD% nonresponders in the present study are in fact “early structural responders” (27), and that change in
Effect of Training on Artery Function
•
349
Green DJ et al.
Table 1. Subject baseline characteristics presented per quartile Q1 (⌬FMD ⱕ0.1%)
Q2 (0.1% ⬍ ⌬FMD ⱕ1.4%)
Q3 (1.4% ⬍ ⌬FMD ⱕ3.5%)
Q4 (⌬FMD ⬎3.5%)
P Value
45 822,4 41 ⫾ 17 1.80 ⫾ 0.092,3,4 94.7 ⫾ 22.33,4 28.6 ⫾ 6.3
Anthropometrics 45 531 42 ⫾ 15 1.73 ⫾ 0.091 86.0 ⫾ 19.1 28.1 ⫾ 5.3
47 60 42 ⫾ 14 1.74 ⫾ 0.101 84.9 ⫾ 14.41 28.0 ⫾ 3.8
45 441 45 ⫾ 15 1.73 ⫾ 0.091 80.6 ⫾ 12.71 27.0 ⫾ 3.9
0.002 0.60 0.001 0.002 0.45
89 ⫾ 12 5.9 ⫾ 2.5 4.7 ⫾ 1.1
92 ⫾ 11 6.4 ⫾ 3.1 5.1 ⫾ 1.1
89 ⫾ 11 5.5 ⫾ 1.5 5.1 ⫾ 1.5
0.09 0.20 0.19
45 22 33 38
40 28 32 36
33 40 27 40
0.18
Subjects, n Sex, % male Age, yr Height, m Weight, kg BMI, kg/m2
General health 94 ⫾ 12 6.7 ⫾ 3.2 4.6 ⫾ 1.3
MAP, mmHg Glucose, mmol/l* Cholesterol, mmol/l† CVD risk status Healthy, % Increased risk, % Manifest CVD disease, % Medication use, %
47 13 40 42
0.94
Exercise characteristics 38.0 ⫾ 17.62,3,4 82:182,3,4 9 ⫾ 24
29.6 ⫾ 9.31 53:471 10 ⫾ 3
30.2 ⫾ 11.21 53:471 11 ⫾ 3
28.6 ⫾ 9.61 49:511 11 ⫾ 31
0.001 0.004 0.036
67 33 0 5.7 ⫾ 4.52,3,4 78 ⫾ 54
58 33 9 3.3 ⫾ 1.51 73 ⫾ 7
51 34 15 3.9 ⫾ 2.81 73 ⫾ 11
49 29 22 3.6 ⫾ 2.11 71 ⫾ 101
0.06
Diameter, mm FMD, % GTN, %‡
4.1 ⫾ 0.7 6.1 ⫾ 2.54 11.1 ⫾ 5.64
3.8 ⫾ 0.8 3.7 ⫾ 2.71,2 16.2 ⫾ 6.21
0.14 0.003 0.006
0.001 0.006
Vascular characteristics 3.8 ⫾ 0.9 5.7 ⫾ 3.84 13.1 ⫾ 5.4
3.9 ⫾ 0.9 4.7 ⫾ 3.4 14.5 ⫾ 8.1
˙ O2peak, Values are means ⫾ SD. *n ⫽ 148, †n ⫽ 149, ‡n ⫽ 141. BMI, body mass index; MAP, mean arterial pressure; CVD, cardiovascular disease; V maximum oxygen uptake, CARE, combined aerobic and resistance exercise; HRmax, maximum heart rate; FMD, flow-mediated dilation; GTN, glyceryl trinitrate. P value refers to a one-way ANOVA between the 4 quartiles. 1,2,3,4Statistically significant differences with Q1, Q2, Q3, and Q4 respectively.
FMD% occurred before posttraining measures were applied. This does not satisfactorily explain the absence of structural adaptation in the upper quartiles, an observation that is further complicated by the finding that the highest FMD responders trained for longer duration. These findings highlight the need for separate and specific future investigation around the nature of response profiles and timing in humans.
An associated issue is the possibility that the increase in baseline diameter in the lower quartile of FMD responders limited the FMD change in this group by virtue of an arithmetic effect, since both parameters are included in the FMD equation (1). We think this unlikely, however, as the change in baseline diameter was modest and similar to that in the other quartiles. Moreover, GTN% responses, which
Table 2. The effect of exercise training, presented per quartile Q1 (⌬FMD ⱕ0.1%) Pre
Weight, kg BMI, kg/m2 MAP, mmHg Glucose, mmol/l* Cholesterol, mmol/l† ˙ O2peak, V ml·kg⫺1·min⫺1 Diameter, mm FMD, % GTN, %‡
Q2 (0.1% ⬍ ⌬FMD ⱕ1.4%) Q3 (1.4% ⬍ ⌬FMD ⱕ3.5%)
Post
Pre
Post
Pre
Post
Q4 (⌬FMD ⬎3.5%) Pre
Post
2-Way ANOVA Time
94.7 ⫾ 22.3 94.5 ⫾ 21.9 86.0 ⫾ 19.1 85.2 ⫾ 18.7 84.9 ⫾ 14.4 83.9 ⫾ 14.4 80.6 ⫾ 12.7 79.9 ⫾ 12.4 ⬍0.001 28.6 ⫾ 6.3 28.5 ⫾ 6.1 28.1 ⫾ 5.3 27.8 ⫾ 5.2 28.0 ⫾ 3.8 27.7 ⫾ 3.8 27.0 ⫾ 3.9 26.7 ⫾ 3.9 ⬍0.001 94 ⫾ 12 89 ⫾ 11 89 ⫾ 12 87 ⫾ 14 92 ⫾ 11 89 ⫾ 12 89 ⫾ 11 86 ⫾ 10 ⬍0.001 6.7 ⫾ 3.2 6.2 ⫾ 1.5 5.9 ⫾ 2.5 5.5 ⫾ 1.9 6.4 ⫾ 3.1 6.3 ⫾ 2.7 5.5 ⫾ 1.5 5.4 ⫾ 1.5 0.131 4.6 ⫾ 1.3
4.7 ⫾ 1.2
4.7 ⫾ 1.1
38.0 ⫾ 17.6 40.1 ⫾ 17.3 29.6 ⫾ 9.3 4.1 ⫾ 0.7 4.3 ⫾ 0.7* 3.8 ⫾ 0.9 6.1 ⫾ 2.5 4.5 ⫾ 2.4§ 5.7 ⫾ 3.8 11.1 ⫾ 5.6 9.2 ⫾ 6.8 13.1 ⫾ 5.4
4.6 ⫾ 0.9
5.1 ⫾ 1.1
4.9 ⫾ 1.1
5.1 ⫾ 1.5
32.4 ⫾ 9.2 30.2 ⫾ 11.2 33.4 ⫾ 12.2 28.6 ⫾ 9.6 3.9 ⫾ 1.0 3.9 ⫾ 0.9 3.9 ⫾ 0.9 3.8 ⫾ 0.8 6.5 ⫾ 3.8§ 4.7 ⫾ 3.4 7.1 ⫾ 3.3§ 3.7 ⫾ 2.7 12.9 ⫾ 6.0 14.5 ⫾ 8.1 13.4 ⫾ 5.8 16.2 ⫾ 6.2
Q
Time ⫻ Q
0.001 0.412 0.230 0.237
0.415 0.522 0.104 0.776
0.329
0.081
31.1 ⫾ 9.4 ⬍0.001 0.002 3.7 ⫾ 0.7 0.231 0.029 9.8 ⫾ 3.4§ ⬍0.001 0.152 15.3 ⫾ 5.9 0.043 ⬍0.001
0.610 0.015 ⬍0.001 0.676
4.7 ⫾ 1.0
0.032
Values are means ⫾ SD. *n ⫽ 148, †n ⫽ 149, ‡n ⫽ 140. §Post hoc significantly different from pretraining at P ⬍ 0.05 (only for diameter and FMD, given the significant “Time ⫻ Q” interaction). P value refers to 2-way ANOVA on whether the impact of exercise training (“Time”) differs between the 4 quartiles (“Q”). J Appl Physiol • doi:10.1152/japplphysiol.00354.2014 • www.jappl.org
Downloaded from on August 18, 2014
˙ O2peak, ml·kg⫺1·min⫺1 V Training mode, %CARE vs. %aerobic Training duration, wk Training duration 8 wk, % 12 wk, % 16 wk, % Training frequency, sessions/wk Training intensity, %HRmax
350
Effect of Training on Artery Function
•
Green DJ et al.
Table 3. Bivariate correlations between changes in FMD% and changes in cardiovascular parameters
⌬FMD (%) ⌬GTN (%) ⌬D (mm) ⌬Weight (kg) ⌬BMI (kg/m2) ⌬MAP (mmHg) ⌬Glucose (mmol/l) ⌬Cholesterol (mmol/l) ˙ O2peak (ml ⌬V ·kg⫺1·min⫺1)
⌬FMD (%)
⌬GTN (%)
⌬D (mm)
⌬Weight (kg)
⌬BMI (kg/m2)
⌬MAP (mmHg)
⌬Glucose (mmol/l)
0.074 ⫺0.268† ⫺0.041 ⫺0.027 0.087 0.052 ⫺0.243†
0.066 ⫺0.016 ⫺0.015 0.040 0.227* 0.026
0.004 ⫺0.015 ⫺0.018 0.139 0.058
0.991† ⫺0.003 ⫺0.064 ⫺0.006
0.008 ⫺0.065 ⫺0.032
⫺0.001 0.013
0.032
0.066
⫺0.067
0.018
⫺0.210†
⫺0.195†
⫺0.019
⫺0.031
⌬Cholesterol (mmol/l)
˙ O2peak (ml ·kg⫺1·min⫺1) ⌬V
⫺0.103
*P ⬍ 0.05, †P ⬍ 0.01.
Limitations A potential limitation is that we included a heterogeneous group of subjects who all participated in an exercise training intervention, rather than a randomized controlled design in which participants were randomized to exercise training or a control intervention. Another limitation is that we did not control for diet and other lifestyle-related factors. Furthermore, we cannot extrapolate our findings to subjects who participate in modalities or intensities of training which are known to trigger different physiological adaptations in the heart (29) and vessels (22, 37). Finally, although FMD% is largely NOmediated (10) and endothelium-dependent (9), we did not report data using other forms of vascular assessment such as plethysmography. We did, however, include GTN% responses. None of the quartiles showed a change in GTN%, while the GTN% response was not identified by the statistical tests as a significant predictor. This suggests that the training-induced changes in FMD% are not related to changes in endotheliumindependent mechanisms, an important internal control measure. Conclusions We found that some subjects do not demonstrate improvement in endothelial function following exercise training. This implies the need to personalize exercise interventions to optimize outcomes and to critically appraise the current “one size fits all” approaches to exercise promotion. In identifying factors that contribute to a larger improvement in arterial function following exercise training (i.e., longer exercise training, lower cardiopulmonary fitness, lower baseline FMD%), our paper provides an initial step toward personalizing exercise interventions to improve vascular health. We also found that changes in endothelial function after training were largely independent of changes in traditional cardiovascular risk factors, reinforcing the diversity of physiological benefits that accrue in response to physical training. These data also support the notion of the “risk factor gap,” the implication of which is that the cardiovascular health benefits of exercise exceed those apparent in terms of traditional risk factor modification. GRANTS D. J. Green is funded by the Australian Research Council (DP 130103793) and the National Health and Medical Research Council (APP1045204). D. H. J. Thijssen is recipient of the E. Dekker Stipend (Netherlands Heart
J Appl Physiol • doi:10.1152/japplphysiol.00354.2014 • www.jappl.org
Downloaded from on August 18, 2014
also include baseline diameter, showed no change after training by quartile or predictive capacity. This suggests that our observation regarding the absence of an increase in FMD after exercise training in Q1 is not simply due to changes in baseline diameter. Our study also provides some insight into factors that contribute to change in artery function. In our pooled analysis, we found that training interventions of longer duration were associated with larger improvements in endothelial function. This finding reinforces the “dose-response” relation between physical activity level and protection against cardiovascular disease (20). We also observed that low initial cardiopulmonary fitness and FMD% were associated with larger training-induced improvements in FMD%. This may reflect a “reserve for improvement,” that is, those with lower a priori fitness and/or endothelial function may have more to gain from training. This is an encouraging observation, as it suggests that previously sedentary individuals as well as those with already established cardiovascular risk and/or disease have the greatest potential for training benefit. Finally, lower pretraining body weight was associated with larger response in terms of FMD%. It is interesting that this relationship was not apparent for pretraining BMI and that change in body weight and BMI were unrelated to those in FMD%. It is unlikely, in our view, that a causal relationship exists between change in FMD and change in body weight within subjects. In agreement with previous meta-analyses (7, 8, 18, 35), our pooled analysis revealed relatively modest improvements in traditional cardiovascular risk factors, such as body weight (⬃0.6 kg), mean arterial pressure (⬃3 mmHg), and total cholesterol (⬃0.2 mmol/l). Furthermore, these changes in traditional risk factors did not relate to changes in endothelial function after training. Our observations reinforce the presence of a “risk factor gap” (12, 17): the concept that the relatively modest changes in traditional risk factors induced by exercise training (32) cannot fully account for the large clinical benefit of a physically active lifestyle (21, 25). Assessment of endothelial function has powerful predictive capacity for future cardiovascular disease, independent of other risk factors (26). The improvement in endothelial function after training is largely independent of change in risk factors. Clinically, this highlights that the cardioprotective benefits of training should not simply be viewed as those associated with easily measured, traditional cardiovascular risk factors.
Effect of Training on Artery Function Foundation, 2009T064). T. Eijsvogels is financially supported by the Netherlands Organization for Scientific Research (Rubicon Grant 825.12.016). L. H. Naylor is supported by BrightSpark Foundation. DISCLOSURES No conflicts of interest, financial or otherwise, are declared by the author(s). AUTHOR CONTRIBUTIONS Author contributions: D.J.G., T.C., M.T.H., and D.H.T. conception and design of research; D.J.G., A.J.M., L.H.N., R.R.S., M.E.S., C.J.P., V.S.S., T.H.S., H.J., T.C., M.T.H., and D.H.T. performed experiments; D.J.G., T.E., Y.M.B., A.J.M., L.H.N., R.R.S., M.E.S., C.J.P., V.S.S., T.H.S., H.J., M.T.H., and D.H.T. analyzed data; D.J.G., T.E., Y.M.B., and D.H.T. interpreted results of experiments; D.J.G., T.E., Y.M.B., and D.H.T. prepared figures; D.J.G., T.E., and D.H.T. drafted manuscript; D.J.G., T.E., Y.M.B., A.J.M., L.H.N., R.R.S., M.E.S., C.J.P., V.S.S., T.H.S., H.J., T.C., M.T.H., and D.H.T. edited and revised manuscript; D.J.G., T.E., Y.M.B., A.J.M., L.H.N., R.R.S., M.E.S., C.J.P., V.S.S., T.H.S., H.J., T.C., M.T.H., and D.H.T. approved final version of manuscript. REFERENCES
Green DJ et al.
351
16. Jones H, Green DJ, George K, Atkinson G. Intermittent exercise abolishes the diurnal variation in endothelial-dependent flow-mediated dilation in humans. Am J Physiol Regul Integr Comp Physiol 298: R427–R432, 2010. 17. Joyner MJ, Green DJ. Exercise protects the cardiovascular system: effects beyond traditional risk factors. J Physiol 587: 5551–5558, 2009. 18. Kodama S, Tanaka S, Saito K, Shu M, Sone Y, Onitake F, Suzuki E, Shimano H, Yamamoto S, Kondo K, Ohashi Y, Yamada N, Sone H. Effect of aerobic exercise training on serum levels of high-density lipoprotein cholesterol: a meta-analysis. Arch Intern Med 167: 999 –1008, 2007. 19. Laughlin MH. Endothelium-mediated control of coronary vascular tone after chronic exercise training. Med Sci Sports Exerc 27: 1135–1144, 1995. 20. Lee IM, Sesso HD, Oguma Y, Paffenbarger RS Jr. Relative intensity of physical activity and risk of coronary heart disease. Circulation 107: 1110 –1116, 2003. 21. Lee IM, Shiroma EJ, Lobelo F, Puska P, Blair SN, Katzmarzyk PT. Effect of physical inactivity on major non-communicable diseases worldwide: an analysis of burden of disease and life expectancy. Lancet 380: 219 –229, 2012. 22. Maiorana AJ, Naylor LH, Exterkate A, Swart A, Thijssen DH, Lam K, O’Driscoll G, Green DJ. The impact of exercise training on conduit artery wall thickness and remodeling in chronic heart failure patients. Hypertension 57: 56 –62, 2011. 23. Midgley AW, McNaughton LR, Polman R, Marchant D. Criteria for determination of maximal oxygen uptake: a brief critique and recommendations for future research. Sports Med 37: 1019 –1028, 2007. 24. Mora S, Cook N, Buring JE, Ridker PM, Lee IM. Physical activity and reduced risk of cardiovascular events: potential mediating mechanisms. Circulation 116: 2110 –2118, 2007. 25. Morris JN, Heady JA, Raffle PA, Roberts CG, Parks JW. Coronary heart-disease and physical activity of work. Lancet 265: 1111–1120; concl, 1953. 26. Ras RT, Streppel MT, Draijer R, Zock PL. Flow-mediated dilation and cardiovascular risk prediction: A systematic review with meta-analysis. Int J Cardiol 168: 344 –351, 2013. 27. Rowley NJ, Dawson EA, Birk GK, Cable NT, George K, Whyte G, Thijssen DH, Green DJ. Exercise and arterial adaptation in humans: uncoupling localized and systemic effects. J Appl Physiol 110: 1190 – 1195, 2011. 28. Skinner JS, Jaskolski A, Jaskolska A, Krasnoff J, Gagnon J, Leon AS, Rao DC, Wilmore JH, Bouchard C. Age, sex, race, initial fitness, and response to training: the HERITAGE Family Study. J Appl Physiol 90: 1770 –1776, 2001. 29. Spence AL, Naylor LH, Carter HH, Buck CL, Dembo L, Murray CP, Watson P, Oxborough D, George KP, Green DJ. A prospective randomised longitudinal MRI study of left ventricular adaptation to endurance and resistance exercise training in humans. J Physiol 589: 5443– 5452, 2011. 30. Takase B, Uehata A, Akima T, Nagai T, Nishioka T, Hamabe A, Satomura K, Ohsuzu F, Kurita A. Endothelium-dependent flow-mediated vasodilation in coronary and brachial arteries in suspected coronary artery disease. Am J Cardiol 82: 1535–1539, A1537–A1538, 1998. 31. Thijssen DH, Black MA, Pyke KE, Padilla J, Atkinson G, Harris RA, Parker B, Widlansky ME, Tschakovsky ME, Green DJ. Assessment of flow-mediated dilation in humans: a methodological and physiological guideline. Am J Physiol Heart Circ Physiol 300: H2–H12, 2011. 32. Thompson PD, Lim V. Physical activity in the prevention of atherosclerotic coronary heart disease. Curr Treat Options Cardiovasc Med 5: 279 –285, 2003. 33. Tinken TM, Thijssen DH, Black MA, Cable NT, Green DJ. Time course of change in vasodilator function and capacity in response to exercise training in humans. J Physiol 586: 5003–5012, 2008. 34. Tinken TM, Thijssen DH, Hopkins N, Dawson EA, Cable NT, Green DJ. Shear stress mediates endothelial adaptations to exercise training in humans. Hypertension 55: 312–318, 2010. 35. Weinheimer EM, Sands LP, Campbell WW. A systematic review of the separate and combined effects of energy restriction and exercise on fat-free mass in middle-aged and older adults: implications for sarcopenic obesity. Nutr Rev 68: 375–388, 2010. 36. Williams MR, Westerman RA, Kingwell BA, Paige J, Blombery PA, Sudhir K, Komesaroff PA. Variations in endothelial function and arterial
J Appl Physiol • doi:10.1152/japplphysiol.00354.2014 • www.jappl.org
Downloaded from on August 18, 2014
1. Atkinson G, Batterham AM, Thijssen DH, Green DJ. A new approach to improve the specificity of flow-mediated dilation for indicating endothelial function in cardiovascular research. J Hypertens 31: 287–291, 2013. 2. Birk GK, Dawson EA, Atkinson C, Haynes A, Cable NT, Thijssen DH, Green DJ. Brachial artery adaptation to lower limb exercise training: role of shear stress. J Appl Physiol 112: 1653–1658, 2012. 3. Black MA, Cable NT, Thijssen DH, Green DJ. Importance of measuring the time course of flow-mediated dilatation in humans. Hypertension 51: 203–210, 2008. 4. Bouchard C, Blair SN, Church TS, Earnest CP, Hagberg JM, Hakkinen K, Jenkins NT, Karavirta L, Kraus WE, Leon AS, Rao DC, Sarzynski MA, Skinner JS, Slentz CA, Rankinen T. Adverse metabolic response to regular exercise: is it a rare or common occurrence? PLoS One 7: e37887, 2012. 5. Bouchard C, Rankinen T. Individual differences in response to regular physical activity. Med Sci Sports Exerc 33: S446 –S451; discussion S452– S443, 2001. 6. Bouchard C, Sarzynski MA, Rice TK, Kraus WE, Church TS, Sung YJ, Rao DC, Rankinen T. Genomic predictors of the maximal O2 uptake response to standardized exercise training programs. J Appl Physiol 110: 1160 –1170, 2011. 7. Cornelissen VA, Buys R, Smart NA. Endurance exercise beneficially affects ambulatory blood pressure: a systematic review and meta-analysis. J Hypertens 31: 639 –648, 2013. 8. Cornelissen VA, Fagard RH, Coeckelberghs E, Vanhees L. Impact of resistance training on blood pressure and other cardiovascular risk factors: a meta-analysis of randomized, controlled trials. Hypertension 58: 950 – 958, 2011. 9. Dawson EA, Rathore S, Cable NT, Wright DJ, Morris JL, Green DJ. Impact of catheter insertion using the radial approach on vasodilatation in humans. Clin Sci (Lond) 118: 633–640, 2010. 10. Green DJ, Dawson EA, Groenewoud HM, Jones H, Thijssen DH. Is flow-mediated dilation nitric oxide mediated?: A meta-analysis. Hypertension 63: 376 –382, 2014. 11. Green DJ, Jones H, Thijssen D, Cable NT, Atkinson G. Flow-mediated dilation and cardiovascular event prediction: does nitric oxide matter? Hypertension 57: 363–369, 2011. 12. Green DJ, O’Driscoll G, Joyner MJ, Cable NT. Exercise and cardiovascular risk reduction: time to update the rationale for exercise? J Appl Physiol 105: 766 –768, 2008. 13. Green DJ, Walsh JH, Maiorana A, Best MJ, Taylor RR, O’Driscoll JG. Exercise-induced improvement in endothelial dysfunction is not mediated by changes in CV risk factors: pooled analysis of diverse patient populations. Am J Physiol Heart Circ Physiol 285: H2679 –H2687, 2003. 14. Hunt JE, Galea D, Tufft G, Bunce D, Ferguson RA. Time course of regional vascular adaptations to low load resistance training with blood flow restriction. J Appl Physiol 115: 403–411, 2013. 15. Inaba Y, Chen JA, Bergmann SR. Prediction of future cardiovascular outcomes by flow-mediated vasodilatation of brachial artery: a metaanalysis. Int J Cardiovasc Imaging 26: 631–640, 2010.
•
352
Effect of Training on Artery Function
compliance during the menstrual cycle. J Clin Endocrinol Metab 86: 5389 –5395, 2001. 37. Wisloff U, Stoylen A, Loennechen JP, Bruvold M, Rognmo O, Haram PM, Tjonna AE, Helgerud J, Slordahl SA, Lee SJ, Videm V, Bye A, Smith GL, Najjar SM, Ellingsen O, Skjaerpe T. Superior cardiovascular effect of aerobic interval training versus moderate continuous training
•
Green DJ et al.
in heart failure patients: a randomized study. Circulation 115: 3086 –3094, 2007. 38. Woodman RJ, Playford DA, Watts GF, Cheetham C, Reed C, Taylor RR, Puddey IB, Beilin LJ, Burke V, Mori TA, Green D. Improved analysis of brachial artery ultrasound using a novel edge-detection software system. J Appl Physiol 91: 929 –937, 2001.
Downloaded from on August 18, 2014
J Appl Physiol • doi:10.1152/japplphysiol.00354.2014 • www.jappl.org
Daniel J. Green, Thijs Eijsvogels, Yvette M. Bouts, Andrew J. Maiorana, Louise H. Naylor, Ralph R. Scholten, Marc E. A. Spaanderman, Christopher J. A. Pugh, Victoria S. Sprung, Tim Schreuder, Helen Jones, Tim Cable, Maria T. E. Hopman and Dick H. J. Thijssen J Appl Physiol 117:345-352, 2014. First published 19 June 2014; doi:10.1152/japplphysiol.00354.2014 You might find this additional info useful... This article cites 38 articles, 16 of which can be accessed free at: /content/117/4/345.full.html#ref-list-1 Updated information and services including high resolution figures, can be found at: /content/117/4/345.full.html Additional material and information about Journal of Applied Physiology can be found at: http://www.the-aps.org/publications/jappl
Journal of Applied Physiology publishes original papers that deal with diverse areas of research in applied physiology, especially those papers emphasizing adaptive and integrative mechanisms. It is published 12 times a year (monthly) by the American Physiological Society, 9650 Rockville Pike, Bethesda MD 20814-3991. Copyright © 2014 by the American Physiological Society. ISSN: 0363-6143, ESSN: 1522-1563. Visit our website at http://www.the-aps.org/.
Downloaded from on August 18, 2014
This information is current as of August 18, 2014.
J Appl Physiol 117: 345–352, 2014. First published June 19, 2014; doi:10.1152/japplphysiol.00354.2014.
Exercise training and artery function in humans: nonresponse and its relationship to cardiovascular risk factors Daniel J. Green,1,2 Thijs Eijsvogels,3,4 Yvette M. Bouts,1,3 Andrew J. Maiorana,5,6 Louise H. Naylor,1 Ralph R. Scholten,3 Marc E. A. Spaanderman,3 Christopher J. A. Pugh,2 Victoria S. Sprung,2 Tim Schreuder,3 Helen Jones,2 Tim Cable,2 Maria T. E. Hopman,3 and Dick H. J. Thijssen2,3 1
School of Sports Science, Exercise and Health, The University of Western Australia, Nedlands, Australia; 2Research Institute for Sport and Exercise Sciences, Liverpool John Moores University, Liverpool, United Kingdom; 3Department of Physiology, Radboud University Medical Center, Nijmegen, The Netherlands; 4Henry Low Heart Center, Department of Cardiology, Hartford Hospital, Hartford, Connecticut; 5School of Physiotherapy and Exercise Science, Curtin University, Perth, Australia; and 6Advanced Heart Failure and Cardiac Transplant Service, Royal Perth Hospital, Perth, Australia Submitted 21 April 2014; accepted in final form 17 June 2014
cardiovascular risk; nitric oxide; flow-mediated dilation; endothelial function; physical activity CARDIOVASCULAR DISEASE (CVD) remains the world’s leading cause of mortality. Physical activity and/or exercise training represent potent strategies to reduce the risk of future
Address for reprint requests and other correspondence: D. J. Green, School of Sports Science, Exercise and Health, The Univ. of Western Australia, Nedlands, Australia (e-mail: [email protected]). http://www.jappl.org
cardiovascular events in asymptomatic subjects and those with preexisting disease (20, 21, 24). While changes in traditional cardiovascular risk factors explain some of the cardioprotective effects of training (24), a substantial proportion of the benefit remains unexplained, and direct effects of exercise on the arterial wall may contribute to this “risk factor gap” (12, 17). Despite the overall health benefits of exercise training, recent studies have described heterogeneous adaptations to training (28). In subjects who undertook similar exercise training interventions, some demonstrated large improvements in parameters such as cardiopulmonary fitness, blood pressure, and cholesterol, while others exhibited smaller increases or even “adverse” responses (4, 5). Such findings are consistent with anecdotal clinical observations that some individuals only respond modestly to exercise. No previous study has explored the heterogeneity of changes in artery function or health in responses to exercise training, or tried to identify predictors of training-mediated adaptations in arterial function. A widely accepted index of artery function and health is flow-mediated dilation (FMD%), the vasodilator response to an imposed shear stress stimulus following cuff-induced increases in blood flow. Recent studies indicate that brachial artery FMD% is a surrogate for assessment of coronary artery endothelial function (30) which predicts cardiovascular outcomes in humans (11, 15, 26). The first aim of this study was therefore to explore the proportion of “responders” and “nonresponders,” in terms of change in artery function (FMD%), following exercise training in humans. Second, we examined whether a priori characteristics or physiological measures can predict “response” vs. “nonresponse.” Finally, we examined whether changes in endothelial function as a result of exercise training correlate with changes in other (traditional) cardiovascular risk factors. For this purpose, we pooled data from 13 studies, involving 182 subjects who undertook exercise training in our laboratories. These studies all involved individualized, center-based exercise prescription, were supervised by exercise scientists, and utilized identical vascular function analysis methodology. We hypothesized that exercise training would induce a heterogeneous effect on FMD%, which might partly be explained by subject and exercise characteristics, and that training-induced changes in FMD% would not correlate with changes in traditional cardiovascular risk factors.
8750-7587/14 Copyright © 2014 the American Physiological Society
345
Downloaded from on August 18, 2014
Green DJ, Eijsvogels T, Bouts YM, Maiorana AJ, Naylor LH, Scholten RR, Spaanderman ME, Pugh CJ, Sprung VS, Schreuder T, Jones H, Cable T, Hopman MT, Thijssen DHJ. Exercise training and artery function in humans: nonresponse and its relationship to cardiovascular risk factors. J Appl Physiol 117: 345–352, 2014. First published June 19, 2014; doi:10.1152/japplphysiol.00354.2014.—The objectives of our study were to examine 1) the proportion of responders and nonresponders to exercise training in terms of vascular function; 2) a priori factors related to exercise training-induced changes in conduit artery function, and 3) the contribution of traditional cardiovascular risk factors to exercise-induced changes in artery function. We pooled data from our laboratories involving 182 subjects who underwent supervised, large-muscle group, endurance-type exercise training interventions with pre-/posttraining measures of flowmediated dilation (FMD%) to assess artery function. All studies adopted an identical FMD protocol (5-min ischemia, distal cuff inflation), contemporary echo-Doppler methodology, and observerindependent automated analysis. Linear regression analysis was used to identify factors contributing to changes in FMD%. We found that cardiopulmonary fitness improved, and weight, body mass index (BMI), cholesterol, and mean arterial pressure (MAP) decreased after training, while FMD% increased in 76% of subjects (P ⬍ 0.001). Training-induced increase in FMD% was predicted by lower body weight ( ⫽ ⫺0.212), lower baseline FMD% ( ⫽ ⫺0.469), lower training frequency ( ⫽ ⫺0.256), and longer training duration ( ⫽ 0.367) (combined: P ⬍ 0.001, r ⫽ 0.63). With the exception of a modest correlation with total cholesterol (r ⫽ ⫺0.243, P ⬍ 0.01), changes in traditional cardiovascular risk factors were not significantly related to changes in FMD% (P ⬎ 0.05). In conclusion, we found that, while some subjects do not demonstrate increases following exercise training, improvement in FMD% is present in those with lower pretraining body weight and endothelial function. Moreover, exercise training-induced change in FMD% did not correlate with changes in traditional cardiovascular risk factors, indicating that some cardioprotective effects of exercise training are independent of improvement in risk factors.
346
Effect of Training on Artery Function
•
Green DJ et al.
Experimental Procedures
Subjects
Subject characteristics. Body mass and stature were recorded using standard methods. Blood pressure was measured after a ⱖ5-min resting period in the supine position (as this measure was typically performed before assessment of the brachial artery FMD) using a manual sphygmomanometer or an automated device (Dinamap, Tampa, FL) by a well-trained researcher. Independent of the method used, blood pressure was measured twice and averaged. MAP was calculated using the diastolic and systolic blood pressures. In all studies we examined levels of fasting blood glucose and total cholesterol. Although methods for these analyses differed between studies, all laboratory tests were undertaken in clinical centers in which robust validation and reproducibility testing is performed. Furthermore, for any given individual, lab tests were performed using identical approaches. These blood samples were all obtained from an antecubital vein. Cardiopulmonary fitness. To obtain sport-specific cardiopulmonary ˙ O2 peak) was measured using fitness data, peak oxygen consumption (V a treadmill (n ⫽ 77), cycle ergometer (n ⫽ 86), or a rowing ergometer (n ⫽ 19), according to the training undertaken by the subjects. Repeated measures were always collected using the same modality and protocols. In all cases, exercise was performed until volitional fatigue, according to standard indications for performing a successful exercise test (23). Gas analyzers and flow probes were calibrated before each test. Oxygen consumption was expressed relative to body weight (ml·kg⫺1·min⫺1). Peak oxygen consumption was calculated from data in the 15– 60 s before volitional exhaustion. Brachial artery endothelial function. FLOW-MEDIATED DILATION. Flow-mediated dilation (FMD%) examines brachial artery endothelium-dependent dilation. All FMD-procedures were performed in line with recent guidelines (31) and all studies used the same, observerindependent edge detection and wall tracking analysis software to analyze arterial diameter (38). In all studies, subjects were asked to fast for ⬎4 h prior to each visit and to avoid caffeine, alcohol, and exercise for ⬎12 h. We ensured that pre- and posttraining measures in premenopausal women were performed during the first phase of the menstrual cycle, given the impact of the menstrual cycle on menstrual cycle on vascular function (36). All studies were conducted in a quiet, temperature-controlled environment. Each visit for a given subject
Exercise training studies performed in our laboratories in the past 20 years were eligible for inclusion. Criteria for inclusion were 1) study performed in our laboratory using comparable methodological approaches (31), including operator-independent analysis with edge detection and wall tracking software; 2) availability of pretraining subject characteristics; 3) successful completion of exercise training, involving supervised, center-based, endurance-type exercise of the large muscle groups; 4) exercise training performed ⱖ2 times/wk for ⱖ8 wk and ⱕ18 wk; 5) brachial artery FMD% measured before and after exercise training; 6) brachial artery FMD% assessed according to best practice recommendations (31); 7) cardiopulmonary fitness measured using gas exchange analysis as the peak oxygen consumption during a sport-specific maximal exercise test; 8) exercise training performed in able-bodied adult subjects; 9) study procedures approved by the local ethics committee and adherent to the Declaration of Helsinki; and 10) pre- and posttraining subject characteristics available [age, body weight, cardiopulmonary fitness, body mass index (BMI)] including traditional cardiovascular risk factors [BMI, mean arterial pressure (MAP), cholesterol, blood pressure, glucose]. Data collected using nonoptimal analysis methods, as well as those with incomplete data sets for our primary outcome parameter (i.e., brachial artery FMD%) or any of the potential moderating variables, were excluded. None of our experiments involved smokers. A flow chart (Fig. 1) provides insight into the inclusion and exclusion of our previous exercise training studies. This resulted in the collection of data from 13 exercise training studies, involving 182 subjects. Study Design In all studies, subject characteristics, cardiopulmonary fitness, and brachial artery FMD% were measured before and after exercise training. Posttraining measurements were performed within 3– 4 days of the last exercise bout, but not within 24 h of the last exercise bout. A number of studies included in our pooled analysis performed repeated measurements collected every 2 wk across an 8-wk exercise training intervention to describe time-dependent changes in brachial artery FMD%. For these studies, we only included the pre- and posttraining values.
Fig. 1. Flow chart on all exercise training studies performed in the last 15 years (1998 –2013) in our laboratories and the reason for exclusion of individual studies. FMD, flow-mediated dilation.
J Appl Physiol • doi:10.1152/japplphysiol.00354.2014 • www.jappl.org
Downloaded from on August 18, 2014
METHODS
Effect of Training on Artery Function
Statistical Analysis All analyses were performed using the Statistical Package for the Social Sciences (IBM SPSS Statistics for Windows, Version 20.0. Armonk, NY). Data are presented as means ⫾ SD. Normality distribution was examined using a Kolmogorov-Smirnov test. In case of non-Gaussian distribution, log-transformation was performed and the data were reexamined for normality distribution. The effect of exercise training was determined with a paired Student’s t-test. Change in endothelial function was defined as the difference in FMD% between pre- and postexercise training measurements. FMD% is the most frequently adopted method to present changes in conduit artery function and health in humans. Based on the 25th, 50th, and 75th percentile cut-points, four quartiles were created to compare differences between nonresponders (Q1), low responders (Q2), moderate responders (Q3), and high responders (Q4). One-way analysis of variance was used to assess differences across quartiles for continuous variables, while a chi-squared test was used for categorical variables (e.g., sex, cardiovascular disease status, medication use). Bonferroni correction was applied to post hoc testing to identify differences across quartiles. A backward linear regression analysis was performed to identify predictors of training-induced changes in FMD%. The pre-/posttraining change in FMD% was entered as the dependent (continuous) variable. Parameters that demonstrated a difference between the quartiles with a P value ⬍ 0.10 were included in our model as ˙ O2 peak, training independent variables: sex, height, weight, MAP, V mode, training duration, training frequency, training intensity, pretraining FMD%, and GTN%. To validate potential predictors of
Green DJ et al.
347
changes in FMD%, we performed a binary logistic regression analysis with responder level [including nonresponders (Q1) and high responders (Q4) only, n ⫽ 90] as our dependent variable. Similar parameters were included in this secondary model. Finally, the effects of exercise training on (traditional) cardiovascular risk factors were determined with a two-way repeated measurements analysis of variance to examine whether the effect of training (i.e., main effect “time”) differed between the quartiles (i.e., main effect “Q”). Post hoc paired Student’s t-tests were performed per quartile in case of an interaction effect (Q ⫻ time). The relationship between changes in endothelial function and changes in cardiovascular risk factors was assessed using Pearson’s correlation coefficient. RESULTS
All subjects (n ⫽ 182) successfully completed the exercise training program. On average, subjects demonstrated a significant decrease in weight (86.5 ⫾ 18.1 to 85.9 ⫾ 17.9 kg, P ⬍ 0.001), BMI (27.9 ⫾ 4.9 to 27.7 ⫾ 4.8 kg/m2, P ⬍ 0.001), MAP (91 ⫾ 12 to 88 ⫾ 12 mmHg, P ⬍ 0.001), and total cholesterol levels (4.9 ⫾ 1.2 to 4.7 ⫾ 1.0 mmol/l, P ⫽ 0.014), ˙ O2 peak (31.6 ⫾ 12.8 to 34.3 ⫾ 12.9 ml·kg⫺1·min⫺1, while V P ⬍ 0.001) and brachial artery FMD% (5.0 ⫾ 3.2 to 7.0 ⫾ 3.8%, P ⬍ 0.001) increased. We observed no significant change in blood glucose levels (6.0 ⫾ 2.5 to 5.8 ⫾ 2.0 mmol/l, P ⫽ 0.107) or resting brachial artery diameter (3.9 ⫾ 0.8 to 3.9 ⫾ 0.8, P ⫽ 0.241). Responders vs. Nonresponders Exercise training resulted in an average increase in brachial artery FMD% of 1.8 ⫾ 3.1% (P ⬍ 0.001). Interestingly, 76% of our study population exhibited improved FMD, whereas 24% of the subjects demonstrated a decrease in brachial artery FMD% (Fig. 2). Quartile 1 consisted of subjects who demonstrated no change or a decrease in FMD% after training only (i.e., Q1; nonresponders), while the other three quartiles consisted of subjects with an increase in FMD% after training of 0.1–1.4% (Q2; low responders), 1.5–3.5% (Q3; moderate responders), and ⬎3.6%, (Q4; high responders), respectively. Differences in subject characteristics of the quartiles are presented in Table 1. Predictors of Changes in FMD% The backward regression analysis excluded sex, height, ˙ O2 peak, training mode, training intensity, and GTN% MAP, V from the final model as these parameters did not significantly contribute to changes in FMD%. For the remaining parameters, we found that an increase in FMD% was predicted by lower pretraining FMD% (P ⬍ 0.001,  ⫽ ⫺0.469), weight (P ⫽ 0.003,  ⫽ ⫺0.212), and training frequency (P ⫽ 0.001,  ⫽ ⫺0.256), and longer training duration (P ⬍ 0.001,  ⫽ 0.367) (r ⫽ 0.63, P ⬍ 0.001). Backward binary logistic regression analysis confirmed these findings and identified similar predictors for identifying high responders (P ⬍ 0.001, Nagelkerke r2 ⫽ 0.68). Relationship Between Changes in FMD% and Changes in Cardiovascular Risk Factors When comparing the adaptations in the traditional cardiovascular risk factors after exercise training between the four quartiles of FMD% change, we found similar adaptations in ˙ O2 peak across groups weight, BMI, MAP, cholesterol, and V
J Appl Physiol • doi:10.1152/japplphysiol.00354.2014 • www.jappl.org
Downloaded from on August 18, 2014
was performed at the same time of day when measures were repeated posttraining (between 8 AM and 4 PM), to avoid the potential impact of diurnal variation on FMD% (16). Furthermore, posttraining assessment of FMD% was performed ⬎24 h after the last training session to prevent potential influence of the last training session. To examine brachial artery FMD, the arm was extended and positioned at an angle of ⬃80° from the torso. A rapid inflation and deflation pneumatic cuff (D.E. Hokanson, Bellevue, WA) was positioned on the forearm, immediately distal to the olecranon process to provide a stimulus to forearm ischemia. Multifrequency (7.5–10 MHz) array probes, attached to high-resolution ultrasound machines, were used to image the brachial artery in the distal one-third of the upper arm. When an optimal image was obtained, the probe was held stable and the ultrasound parameters were set to optimize the longitudinal, B-mode image of the lumen-arterial wall interface. Continuous Doppler velocity assessments were also obtained using the ultrasound and were collected using the lowest possible insonation angle (always ⬍60°). Following baseline assessments, a forearm cuff was inflated (⬎200 mmHg) for 5 min. Diameter and flow recordings resumed 30 s prior to cuff deflation and continued for 3 min thereafter, in accordance with recent technical specifications (3, 38). An experienced sonographer performed the measurements, with the same sonographer performing both pre- and posttraining measurements within each subject. GLYCERYL TRINITRATE. Using a sublingual dose of glyceryl trinitrate (GTN), we recorded the diameter response for ⬎5 min to examine endothelium-independent dilation of the brachial artery. This test assesses whether training-induced changes in FMD% can be attributed to changes in endothelium-independent mechanisms. BRACHIAL ARTERY DIAMETER AND BLOOD FLOW ANALYSIS. In all studies included in this pooled analysis, we performed analysis of brachial artery diameter using edge-detection and wall-tracking software, which is largely independent of investigator bias. Previous papers contain detailed descriptions of our analysis approach (3, 38). This automated software provides more reproducible and valid results than manual methods, reduces observer error significantly, and possesses an intra-observer CV of 6.7% (38).
•
348
Effect of Training on Artery Function
•
Green DJ et al.
(Table 2). Interestingly, the direction of change in baseline arterial diameter significantly differed between quartiles (Q ⫻ time: P ⫽ 0.015), with post hoc analysis revealing an increase in diameter after training in Q1, but no changes in Q2–Q4 (Table 2). As a consequence of using the change in FMD% to characterize the quartiles, the direction and magnitude of change in FMD% differed between groups (Table 2). Pearson’s correlation coefficients between changes in FMD% and changes in cardiovascular parameters are presented in Table 3. We found a significant, inverse correlation between the change in FMD% and the changes in diameter and cholesterol (Table 3), while changes in the other (traditional) cardiovascular risk factors did not relate to change in FMD%. DISCUSSION
J Appl Physiol • doi:10.1152/japplphysiol.00354.2014 • www.jappl.org
Downloaded from on August 18, 2014
Fig. 2. Data of all individual subjects (n ⫽ 182) on the change in brachial ˙ O2 peak, artery flow-mediated dilation (FMD%, A), cardiopulmonary fitness (V B), and cholesterol (in mmol/l, C) after supervised, endurance-type exercise training. Individual data are organized based on the magnitude of change, with the largest deterioration (i.e., data in red) presented on the left and the largest improvement (i.e., data in green) on the right end of the spectrum. Based on the 25th, 50th, and 75th percentile cut points, 4 quartiles were created to compare differences between nonresponders (Q1 red), low responders (Q2 orange), moderate responders (Q3 gold), and high responders (Q4 green). These color-based FMD% quartiles for change are maintained in the bottom panels, visually illustrating the lack of consistency between change in the respective variables.
Our grouped analysis reveals substantial heterogeneity in the impact of exercise training on FMD%, with 76% of the subjects studied showing some improvement. Larger improvements in FMD% were linked to longer exercise training interventions and lower cardiopulmonary fitness and baseline FMD%. These factors explained ⬃39% of the variation in adaptation of artery function. Finally, the exercise traininginduced change in FMD% did not correlate with changes in traditional cardiovascular risk factors, except for a modest correlation with change in total cholesterol. These findings confirm our previous finding in a much smaller subgroup (13) and suggest that some component of the cardioprotective effect of exercise training occurs independently of change in traditional cardiovascular risk factors and may be related to improvements in artery function. We found that ⬃24% of the subjects demonstrated no change in endothelial function with training. Although no previous study has reported the heterogeneity in arterial adaptation to exercise training, recent publications by Bouchard and colleagues have raised the general concept of a lack of responsiveness to exercise training (4 – 6). The authors suggested that training may have an “adverse” impact on some risk factors (e.g., blood pressure, triglycerides, HDL), in certain individuals (4, 5). Whether such adverse effects translate into poor clinical outcomes is unknown, especially since CVD risk assessment is typically undertaken by examining a cluster of biomarker and/or risk factors. Nonetheless, these observations demonstrate that the effects of training are not simply unidirectional, an important message when evaluating the impact of training programs at individual or group levels. Despite the strong predictive capacity of FMD% (11, 26), we do not interpret our observation of decline in FMD% in 24% of the subjects as evidence that exercise training increased cardiovascular risk. This group of nonresponders also demonstrated increases in resting brachial artery diameter after training, an effect that was not apparent in the other FMD% quartiles. This raises the possibility that exercise training induced arterial adaptation in all subjects, but of distinct forms. It is now known that exercise training induces changes in both artery function and structure and that these occur according to a distinct time course (19). We (2, 33, 34) and others (14) have demonstrated that initial improvements in function are superseded by compensatory changes in artery size. It is therefore possible that the FMD% nonresponders in the present study are in fact “early structural responders” (27), and that change in
Effect of Training on Artery Function
•
349
Green DJ et al.
Table 1. Subject baseline characteristics presented per quartile Q1 (⌬FMD ⱕ0.1%)
Q2 (0.1% ⬍ ⌬FMD ⱕ1.4%)
Q3 (1.4% ⬍ ⌬FMD ⱕ3.5%)
Q4 (⌬FMD ⬎3.5%)
P Value
45 822,4 41 ⫾ 17 1.80 ⫾ 0.092,3,4 94.7 ⫾ 22.33,4 28.6 ⫾ 6.3
Anthropometrics 45 531 42 ⫾ 15 1.73 ⫾ 0.091 86.0 ⫾ 19.1 28.1 ⫾ 5.3
47 60 42 ⫾ 14 1.74 ⫾ 0.101 84.9 ⫾ 14.41 28.0 ⫾ 3.8
45 441 45 ⫾ 15 1.73 ⫾ 0.091 80.6 ⫾ 12.71 27.0 ⫾ 3.9
0.002 0.60 0.001 0.002 0.45
89 ⫾ 12 5.9 ⫾ 2.5 4.7 ⫾ 1.1
92 ⫾ 11 6.4 ⫾ 3.1 5.1 ⫾ 1.1
89 ⫾ 11 5.5 ⫾ 1.5 5.1 ⫾ 1.5
0.09 0.20 0.19
45 22 33 38
40 28 32 36
33 40 27 40
0.18
Subjects, n Sex, % male Age, yr Height, m Weight, kg BMI, kg/m2
General health 94 ⫾ 12 6.7 ⫾ 3.2 4.6 ⫾ 1.3
MAP, mmHg Glucose, mmol/l* Cholesterol, mmol/l† CVD risk status Healthy, % Increased risk, % Manifest CVD disease, % Medication use, %
47 13 40 42
0.94
Exercise characteristics 38.0 ⫾ 17.62,3,4 82:182,3,4 9 ⫾ 24
29.6 ⫾ 9.31 53:471 10 ⫾ 3
30.2 ⫾ 11.21 53:471 11 ⫾ 3
28.6 ⫾ 9.61 49:511 11 ⫾ 31
0.001 0.004 0.036
67 33 0 5.7 ⫾ 4.52,3,4 78 ⫾ 54
58 33 9 3.3 ⫾ 1.51 73 ⫾ 7
51 34 15 3.9 ⫾ 2.81 73 ⫾ 11
49 29 22 3.6 ⫾ 2.11 71 ⫾ 101
0.06
Diameter, mm FMD, % GTN, %‡
4.1 ⫾ 0.7 6.1 ⫾ 2.54 11.1 ⫾ 5.64
3.8 ⫾ 0.8 3.7 ⫾ 2.71,2 16.2 ⫾ 6.21
0.14 0.003 0.006
0.001 0.006
Vascular characteristics 3.8 ⫾ 0.9 5.7 ⫾ 3.84 13.1 ⫾ 5.4
3.9 ⫾ 0.9 4.7 ⫾ 3.4 14.5 ⫾ 8.1
˙ O2peak, Values are means ⫾ SD. *n ⫽ 148, †n ⫽ 149, ‡n ⫽ 141. BMI, body mass index; MAP, mean arterial pressure; CVD, cardiovascular disease; V maximum oxygen uptake, CARE, combined aerobic and resistance exercise; HRmax, maximum heart rate; FMD, flow-mediated dilation; GTN, glyceryl trinitrate. P value refers to a one-way ANOVA between the 4 quartiles. 1,2,3,4Statistically significant differences with Q1, Q2, Q3, and Q4 respectively.
FMD% occurred before posttraining measures were applied. This does not satisfactorily explain the absence of structural adaptation in the upper quartiles, an observation that is further complicated by the finding that the highest FMD responders trained for longer duration. These findings highlight the need for separate and specific future investigation around the nature of response profiles and timing in humans.
An associated issue is the possibility that the increase in baseline diameter in the lower quartile of FMD responders limited the FMD change in this group by virtue of an arithmetic effect, since both parameters are included in the FMD equation (1). We think this unlikely, however, as the change in baseline diameter was modest and similar to that in the other quartiles. Moreover, GTN% responses, which
Table 2. The effect of exercise training, presented per quartile Q1 (⌬FMD ⱕ0.1%) Pre
Weight, kg BMI, kg/m2 MAP, mmHg Glucose, mmol/l* Cholesterol, mmol/l† ˙ O2peak, V ml·kg⫺1·min⫺1 Diameter, mm FMD, % GTN, %‡
Q2 (0.1% ⬍ ⌬FMD ⱕ1.4%) Q3 (1.4% ⬍ ⌬FMD ⱕ3.5%)
Post
Pre
Post
Pre
Post
Q4 (⌬FMD ⬎3.5%) Pre
Post
2-Way ANOVA Time
94.7 ⫾ 22.3 94.5 ⫾ 21.9 86.0 ⫾ 19.1 85.2 ⫾ 18.7 84.9 ⫾ 14.4 83.9 ⫾ 14.4 80.6 ⫾ 12.7 79.9 ⫾ 12.4 ⬍0.001 28.6 ⫾ 6.3 28.5 ⫾ 6.1 28.1 ⫾ 5.3 27.8 ⫾ 5.2 28.0 ⫾ 3.8 27.7 ⫾ 3.8 27.0 ⫾ 3.9 26.7 ⫾ 3.9 ⬍0.001 94 ⫾ 12 89 ⫾ 11 89 ⫾ 12 87 ⫾ 14 92 ⫾ 11 89 ⫾ 12 89 ⫾ 11 86 ⫾ 10 ⬍0.001 6.7 ⫾ 3.2 6.2 ⫾ 1.5 5.9 ⫾ 2.5 5.5 ⫾ 1.9 6.4 ⫾ 3.1 6.3 ⫾ 2.7 5.5 ⫾ 1.5 5.4 ⫾ 1.5 0.131 4.6 ⫾ 1.3
4.7 ⫾ 1.2
4.7 ⫾ 1.1
38.0 ⫾ 17.6 40.1 ⫾ 17.3 29.6 ⫾ 9.3 4.1 ⫾ 0.7 4.3 ⫾ 0.7* 3.8 ⫾ 0.9 6.1 ⫾ 2.5 4.5 ⫾ 2.4§ 5.7 ⫾ 3.8 11.1 ⫾ 5.6 9.2 ⫾ 6.8 13.1 ⫾ 5.4
4.6 ⫾ 0.9
5.1 ⫾ 1.1
4.9 ⫾ 1.1
5.1 ⫾ 1.5
32.4 ⫾ 9.2 30.2 ⫾ 11.2 33.4 ⫾ 12.2 28.6 ⫾ 9.6 3.9 ⫾ 1.0 3.9 ⫾ 0.9 3.9 ⫾ 0.9 3.8 ⫾ 0.8 6.5 ⫾ 3.8§ 4.7 ⫾ 3.4 7.1 ⫾ 3.3§ 3.7 ⫾ 2.7 12.9 ⫾ 6.0 14.5 ⫾ 8.1 13.4 ⫾ 5.8 16.2 ⫾ 6.2
Q
Time ⫻ Q
0.001 0.412 0.230 0.237
0.415 0.522 0.104 0.776
0.329
0.081
31.1 ⫾ 9.4 ⬍0.001 0.002 3.7 ⫾ 0.7 0.231 0.029 9.8 ⫾ 3.4§ ⬍0.001 0.152 15.3 ⫾ 5.9 0.043 ⬍0.001
0.610 0.015 ⬍0.001 0.676
4.7 ⫾ 1.0
0.032
Values are means ⫾ SD. *n ⫽ 148, †n ⫽ 149, ‡n ⫽ 140. §Post hoc significantly different from pretraining at P ⬍ 0.05 (only for diameter and FMD, given the significant “Time ⫻ Q” interaction). P value refers to 2-way ANOVA on whether the impact of exercise training (“Time”) differs between the 4 quartiles (“Q”). J Appl Physiol • doi:10.1152/japplphysiol.00354.2014 • www.jappl.org
Downloaded from on August 18, 2014
˙ O2peak, ml·kg⫺1·min⫺1 V Training mode, %CARE vs. %aerobic Training duration, wk Training duration 8 wk, % 12 wk, % 16 wk, % Training frequency, sessions/wk Training intensity, %HRmax
350
Effect of Training on Artery Function
•
Green DJ et al.
Table 3. Bivariate correlations between changes in FMD% and changes in cardiovascular parameters
⌬FMD (%) ⌬GTN (%) ⌬D (mm) ⌬Weight (kg) ⌬BMI (kg/m2) ⌬MAP (mmHg) ⌬Glucose (mmol/l) ⌬Cholesterol (mmol/l) ˙ O2peak (ml ⌬V ·kg⫺1·min⫺1)
⌬FMD (%)
⌬GTN (%)
⌬D (mm)
⌬Weight (kg)
⌬BMI (kg/m2)
⌬MAP (mmHg)
⌬Glucose (mmol/l)
0.074 ⫺0.268† ⫺0.041 ⫺0.027 0.087 0.052 ⫺0.243†
0.066 ⫺0.016 ⫺0.015 0.040 0.227* 0.026
0.004 ⫺0.015 ⫺0.018 0.139 0.058
0.991† ⫺0.003 ⫺0.064 ⫺0.006
0.008 ⫺0.065 ⫺0.032
⫺0.001 0.013
0.032
0.066
⫺0.067
0.018
⫺0.210†
⫺0.195†
⫺0.019
⫺0.031
⌬Cholesterol (mmol/l)
˙ O2peak (ml ·kg⫺1·min⫺1) ⌬V
⫺0.103
*P ⬍ 0.05, †P ⬍ 0.01.
Limitations A potential limitation is that we included a heterogeneous group of subjects who all participated in an exercise training intervention, rather than a randomized controlled design in which participants were randomized to exercise training or a control intervention. Another limitation is that we did not control for diet and other lifestyle-related factors. Furthermore, we cannot extrapolate our findings to subjects who participate in modalities or intensities of training which are known to trigger different physiological adaptations in the heart (29) and vessels (22, 37). Finally, although FMD% is largely NOmediated (10) and endothelium-dependent (9), we did not report data using other forms of vascular assessment such as plethysmography. We did, however, include GTN% responses. None of the quartiles showed a change in GTN%, while the GTN% response was not identified by the statistical tests as a significant predictor. This suggests that the training-induced changes in FMD% are not related to changes in endotheliumindependent mechanisms, an important internal control measure. Conclusions We found that some subjects do not demonstrate improvement in endothelial function following exercise training. This implies the need to personalize exercise interventions to optimize outcomes and to critically appraise the current “one size fits all” approaches to exercise promotion. In identifying factors that contribute to a larger improvement in arterial function following exercise training (i.e., longer exercise training, lower cardiopulmonary fitness, lower baseline FMD%), our paper provides an initial step toward personalizing exercise interventions to improve vascular health. We also found that changes in endothelial function after training were largely independent of changes in traditional cardiovascular risk factors, reinforcing the diversity of physiological benefits that accrue in response to physical training. These data also support the notion of the “risk factor gap,” the implication of which is that the cardiovascular health benefits of exercise exceed those apparent in terms of traditional risk factor modification. GRANTS D. J. Green is funded by the Australian Research Council (DP 130103793) and the National Health and Medical Research Council (APP1045204). D. H. J. Thijssen is recipient of the E. Dekker Stipend (Netherlands Heart
J Appl Physiol • doi:10.1152/japplphysiol.00354.2014 • www.jappl.org
Downloaded from on August 18, 2014
also include baseline diameter, showed no change after training by quartile or predictive capacity. This suggests that our observation regarding the absence of an increase in FMD after exercise training in Q1 is not simply due to changes in baseline diameter. Our study also provides some insight into factors that contribute to change in artery function. In our pooled analysis, we found that training interventions of longer duration were associated with larger improvements in endothelial function. This finding reinforces the “dose-response” relation between physical activity level and protection against cardiovascular disease (20). We also observed that low initial cardiopulmonary fitness and FMD% were associated with larger training-induced improvements in FMD%. This may reflect a “reserve for improvement,” that is, those with lower a priori fitness and/or endothelial function may have more to gain from training. This is an encouraging observation, as it suggests that previously sedentary individuals as well as those with already established cardiovascular risk and/or disease have the greatest potential for training benefit. Finally, lower pretraining body weight was associated with larger response in terms of FMD%. It is interesting that this relationship was not apparent for pretraining BMI and that change in body weight and BMI were unrelated to those in FMD%. It is unlikely, in our view, that a causal relationship exists between change in FMD and change in body weight within subjects. In agreement with previous meta-analyses (7, 8, 18, 35), our pooled analysis revealed relatively modest improvements in traditional cardiovascular risk factors, such as body weight (⬃0.6 kg), mean arterial pressure (⬃3 mmHg), and total cholesterol (⬃0.2 mmol/l). Furthermore, these changes in traditional risk factors did not relate to changes in endothelial function after training. Our observations reinforce the presence of a “risk factor gap” (12, 17): the concept that the relatively modest changes in traditional risk factors induced by exercise training (32) cannot fully account for the large clinical benefit of a physically active lifestyle (21, 25). Assessment of endothelial function has powerful predictive capacity for future cardiovascular disease, independent of other risk factors (26). The improvement in endothelial function after training is largely independent of change in risk factors. Clinically, this highlights that the cardioprotective benefits of training should not simply be viewed as those associated with easily measured, traditional cardiovascular risk factors.
Effect of Training on Artery Function Foundation, 2009T064). T. Eijsvogels is financially supported by the Netherlands Organization for Scientific Research (Rubicon Grant 825.12.016). L. H. Naylor is supported by BrightSpark Foundation. DISCLOSURES No conflicts of interest, financial or otherwise, are declared by the author(s). AUTHOR CONTRIBUTIONS Author contributions: D.J.G., T.C., M.T.H., and D.H.T. conception and design of research; D.J.G., A.J.M., L.H.N., R.R.S., M.E.S., C.J.P., V.S.S., T.H.S., H.J., T.C., M.T.H., and D.H.T. performed experiments; D.J.G., T.E., Y.M.B., A.J.M., L.H.N., R.R.S., M.E.S., C.J.P., V.S.S., T.H.S., H.J., M.T.H., and D.H.T. analyzed data; D.J.G., T.E., Y.M.B., and D.H.T. interpreted results of experiments; D.J.G., T.E., Y.M.B., and D.H.T. prepared figures; D.J.G., T.E., and D.H.T. drafted manuscript; D.J.G., T.E., Y.M.B., A.J.M., L.H.N., R.R.S., M.E.S., C.J.P., V.S.S., T.H.S., H.J., T.C., M.T.H., and D.H.T. edited and revised manuscript; D.J.G., T.E., Y.M.B., A.J.M., L.H.N., R.R.S., M.E.S., C.J.P., V.S.S., T.H.S., H.J., T.C., M.T.H., and D.H.T. approved final version of manuscript. REFERENCES
Green DJ et al.
351
16. Jones H, Green DJ, George K, Atkinson G. Intermittent exercise abolishes the diurnal variation in endothelial-dependent flow-mediated dilation in humans. Am J Physiol Regul Integr Comp Physiol 298: R427–R432, 2010. 17. Joyner MJ, Green DJ. Exercise protects the cardiovascular system: effects beyond traditional risk factors. J Physiol 587: 5551–5558, 2009. 18. Kodama S, Tanaka S, Saito K, Shu M, Sone Y, Onitake F, Suzuki E, Shimano H, Yamamoto S, Kondo K, Ohashi Y, Yamada N, Sone H. Effect of aerobic exercise training on serum levels of high-density lipoprotein cholesterol: a meta-analysis. Arch Intern Med 167: 999 –1008, 2007. 19. Laughlin MH. Endothelium-mediated control of coronary vascular tone after chronic exercise training. Med Sci Sports Exerc 27: 1135–1144, 1995. 20. Lee IM, Sesso HD, Oguma Y, Paffenbarger RS Jr. Relative intensity of physical activity and risk of coronary heart disease. Circulation 107: 1110 –1116, 2003. 21. Lee IM, Shiroma EJ, Lobelo F, Puska P, Blair SN, Katzmarzyk PT. Effect of physical inactivity on major non-communicable diseases worldwide: an analysis of burden of disease and life expectancy. Lancet 380: 219 –229, 2012. 22. Maiorana AJ, Naylor LH, Exterkate A, Swart A, Thijssen DH, Lam K, O’Driscoll G, Green DJ. The impact of exercise training on conduit artery wall thickness and remodeling in chronic heart failure patients. Hypertension 57: 56 –62, 2011. 23. Midgley AW, McNaughton LR, Polman R, Marchant D. Criteria for determination of maximal oxygen uptake: a brief critique and recommendations for future research. Sports Med 37: 1019 –1028, 2007. 24. Mora S, Cook N, Buring JE, Ridker PM, Lee IM. Physical activity and reduced risk of cardiovascular events: potential mediating mechanisms. Circulation 116: 2110 –2118, 2007. 25. Morris JN, Heady JA, Raffle PA, Roberts CG, Parks JW. Coronary heart-disease and physical activity of work. Lancet 265: 1111–1120; concl, 1953. 26. Ras RT, Streppel MT, Draijer R, Zock PL. Flow-mediated dilation and cardiovascular risk prediction: A systematic review with meta-analysis. Int J Cardiol 168: 344 –351, 2013. 27. Rowley NJ, Dawson EA, Birk GK, Cable NT, George K, Whyte G, Thijssen DH, Green DJ. Exercise and arterial adaptation in humans: uncoupling localized and systemic effects. J Appl Physiol 110: 1190 – 1195, 2011. 28. Skinner JS, Jaskolski A, Jaskolska A, Krasnoff J, Gagnon J, Leon AS, Rao DC, Wilmore JH, Bouchard C. Age, sex, race, initial fitness, and response to training: the HERITAGE Family Study. J Appl Physiol 90: 1770 –1776, 2001. 29. Spence AL, Naylor LH, Carter HH, Buck CL, Dembo L, Murray CP, Watson P, Oxborough D, George KP, Green DJ. A prospective randomised longitudinal MRI study of left ventricular adaptation to endurance and resistance exercise training in humans. J Physiol 589: 5443– 5452, 2011. 30. Takase B, Uehata A, Akima T, Nagai T, Nishioka T, Hamabe A, Satomura K, Ohsuzu F, Kurita A. Endothelium-dependent flow-mediated vasodilation in coronary and brachial arteries in suspected coronary artery disease. Am J Cardiol 82: 1535–1539, A1537–A1538, 1998. 31. Thijssen DH, Black MA, Pyke KE, Padilla J, Atkinson G, Harris RA, Parker B, Widlansky ME, Tschakovsky ME, Green DJ. Assessment of flow-mediated dilation in humans: a methodological and physiological guideline. Am J Physiol Heart Circ Physiol 300: H2–H12, 2011. 32. Thompson PD, Lim V. Physical activity in the prevention of atherosclerotic coronary heart disease. Curr Treat Options Cardiovasc Med 5: 279 –285, 2003. 33. Tinken TM, Thijssen DH, Black MA, Cable NT, Green DJ. Time course of change in vasodilator function and capacity in response to exercise training in humans. J Physiol 586: 5003–5012, 2008. 34. Tinken TM, Thijssen DH, Hopkins N, Dawson EA, Cable NT, Green DJ. Shear stress mediates endothelial adaptations to exercise training in humans. Hypertension 55: 312–318, 2010. 35. Weinheimer EM, Sands LP, Campbell WW. A systematic review of the separate and combined effects of energy restriction and exercise on fat-free mass in middle-aged and older adults: implications for sarcopenic obesity. Nutr Rev 68: 375–388, 2010. 36. Williams MR, Westerman RA, Kingwell BA, Paige J, Blombery PA, Sudhir K, Komesaroff PA. Variations in endothelial function and arterial
J Appl Physiol • doi:10.1152/japplphysiol.00354.2014 • www.jappl.org
Downloaded from on August 18, 2014
1. Atkinson G, Batterham AM, Thijssen DH, Green DJ. A new approach to improve the specificity of flow-mediated dilation for indicating endothelial function in cardiovascular research. J Hypertens 31: 287–291, 2013. 2. Birk GK, Dawson EA, Atkinson C, Haynes A, Cable NT, Thijssen DH, Green DJ. Brachial artery adaptation to lower limb exercise training: role of shear stress. J Appl Physiol 112: 1653–1658, 2012. 3. Black MA, Cable NT, Thijssen DH, Green DJ. Importance of measuring the time course of flow-mediated dilatation in humans. Hypertension 51: 203–210, 2008. 4. Bouchard C, Blair SN, Church TS, Earnest CP, Hagberg JM, Hakkinen K, Jenkins NT, Karavirta L, Kraus WE, Leon AS, Rao DC, Sarzynski MA, Skinner JS, Slentz CA, Rankinen T. Adverse metabolic response to regular exercise: is it a rare or common occurrence? PLoS One 7: e37887, 2012. 5. Bouchard C, Rankinen T. Individual differences in response to regular physical activity. Med Sci Sports Exerc 33: S446 –S451; discussion S452– S443, 2001. 6. Bouchard C, Sarzynski MA, Rice TK, Kraus WE, Church TS, Sung YJ, Rao DC, Rankinen T. Genomic predictors of the maximal O2 uptake response to standardized exercise training programs. J Appl Physiol 110: 1160 –1170, 2011. 7. Cornelissen VA, Buys R, Smart NA. Endurance exercise beneficially affects ambulatory blood pressure: a systematic review and meta-analysis. J Hypertens 31: 639 –648, 2013. 8. Cornelissen VA, Fagard RH, Coeckelberghs E, Vanhees L. Impact of resistance training on blood pressure and other cardiovascular risk factors: a meta-analysis of randomized, controlled trials. Hypertension 58: 950 – 958, 2011. 9. Dawson EA, Rathore S, Cable NT, Wright DJ, Morris JL, Green DJ. Impact of catheter insertion using the radial approach on vasodilatation in humans. Clin Sci (Lond) 118: 633–640, 2010. 10. Green DJ, Dawson EA, Groenewoud HM, Jones H, Thijssen DH. Is flow-mediated dilation nitric oxide mediated?: A meta-analysis. Hypertension 63: 376 –382, 2014. 11. Green DJ, Jones H, Thijssen D, Cable NT, Atkinson G. Flow-mediated dilation and cardiovascular event prediction: does nitric oxide matter? Hypertension 57: 363–369, 2011. 12. Green DJ, O’Driscoll G, Joyner MJ, Cable NT. Exercise and cardiovascular risk reduction: time to update the rationale for exercise? J Appl Physiol 105: 766 –768, 2008. 13. Green DJ, Walsh JH, Maiorana A, Best MJ, Taylor RR, O’Driscoll JG. Exercise-induced improvement in endothelial dysfunction is not mediated by changes in CV risk factors: pooled analysis of diverse patient populations. Am J Physiol Heart Circ Physiol 285: H2679 –H2687, 2003. 14. Hunt JE, Galea D, Tufft G, Bunce D, Ferguson RA. Time course of regional vascular adaptations to low load resistance training with blood flow restriction. J Appl Physiol 115: 403–411, 2013. 15. Inaba Y, Chen JA, Bergmann SR. Prediction of future cardiovascular outcomes by flow-mediated vasodilatation of brachial artery: a metaanalysis. Int J Cardiovasc Imaging 26: 631–640, 2010.
•
352
Effect of Training on Artery Function
compliance during the menstrual cycle. J Clin Endocrinol Metab 86: 5389 –5395, 2001. 37. Wisloff U, Stoylen A, Loennechen JP, Bruvold M, Rognmo O, Haram PM, Tjonna AE, Helgerud J, Slordahl SA, Lee SJ, Videm V, Bye A, Smith GL, Najjar SM, Ellingsen O, Skjaerpe T. Superior cardiovascular effect of aerobic interval training versus moderate continuous training
•
Green DJ et al.
in heart failure patients: a randomized study. Circulation 115: 3086 –3094, 2007. 38. Woodman RJ, Playford DA, Watts GF, Cheetham C, Reed C, Taylor RR, Puddey IB, Beilin LJ, Burke V, Mori TA, Green D. Improved analysis of brachial artery ultrasound using a novel edge-detection software system. J Appl Physiol 91: 929 –937, 2001.
Downloaded from on August 18, 2014
J Appl Physiol • doi:10.1152/japplphysiol.00354.2014 • www.jappl.org
