Low aerobic fitness is a significant and prevalent risk factor for future cardiovascular disease (CVD) 12. An increase in aerobic fitness through exercise training reduces CVD risk 34. In keeping with this concept, increased fitness has been shown to improve functional measures of cardiovascular health, such as aortic stiffness measured using aortic pulse wave velocity (aPWV) 5. It also results in femoral arterial remodeling as determined from Doppler ultrasound imaging of arterial diameter 678. In addition, elevated resting shear stress in the femoral artery results from disuse of the lower extremities 910. Considering that disuse is also associated with increased CVD (reviewed in 11), elevated resting shear stress may also be associated with increased CVD risk.

Elevated adiposity also increases the risk for CVD and has increased in prevalence over the last several decades 121314. Therefore, reducing the prevalence of obesity is a major public health initiative. Elevated adiposity is also associated with low physical activity and weight reduction typically involves an increase in physical activity. It is not presently clear whether elevated adiposity is an independent risk factor for CVD or merely a proxy for low physical activity, low fitness, and poor diet. Prospective studies have shown that aerobic fitness reduces the negative effect of adipose tissue on CVD mortality 15. Additionally, weight loss programs have been associated with repetitive weight loss and weight gain, net increases in weight and increased CVD mortality 161718. This suggests that programs designed to improve cardiovascular health may do better to focus on improved aerobic fitness rather than weight loss.

The circulating concentrations of several proteins that may originate from, or influence the development of, atherosclerotic plaques are currently under investigation as adjuncts to cholesterol/lipoprotein measurements in assessment of risk for cardiovascular disease. Among those reported to be independent predictors of cardiovascular disease risk are C-reactive protein (CRP) 19, soluble intercellular adhesion molecule-1 (sICAM-1) 20, and leptin 21.

Elevated levels of CRP and sICAM-1 have been associated with CVD 22, as well as with adiposity 2324. Adiposity is also associated with elevations in leptin, an adipocyte-derived, 16 kDa hormone/cytokine that influences energy balance, thermoregulation and reproductive function 25. Leptin receptors have been identified on vascular cells 26, and several in vitro and animal studies have indicated that leptin may influence atherosclerotic plaque development (reviewed in 27). A reduction in these biological proteins might confer protection against the progression of atherosclerosis.

Previous studies that have evaluated the interaction between aerobic fitness and adiposity have commonly used body mass index (BMI) as an index of adiposity. BMI is not a measure of the amount of adipose tissue an individual has and its relationship to adiposity, or percent body fat, may vary considerably between individuals 28. Furthermore, without an accurate measure of adiposity, maximal oxygen consumption as an index of aerobic fitness is compromised. For example, total body weight is typically used to normalize oxygen consumption. Therefore, increased adiposity will automatically lower aerobic fitness irrespective of the maximal capacity of the individual's cardiorespiratory system and muscles to deliver and utilize oxygen. Dual energy x-ray absorptiometry (DXA) yields accurate estimates of fat-free mass 29 which can be used to normalize maximal oxygen consumption and help minimize this limitation 30. The purpose of this study was to determine whether a 14-week aerobic exercise program would 1) improve aerobic fitness without changing adiposity, 2) improve functional measures of vascular health, and 3) reduce serum concentrations of proteins associated with risk for cardiovascular disease. A secondary purpose was to determine whether changes in functional measures of vascular health and serum concentrations of proteins associated with risk for cardiovascular disease were dependent upon changes in aerobic fitness.

The physical characteristics of the subjects are shown in Table 1. The training program did not result in a change in body weight or BMI. There was a trend for small increases in fat-free mass and decreases in fat mass, but these changes were not statistically significant. Training resulted in a significant decrease in resting heart rate (4.6%), although systolic and diastolic arterial blood pressure did not change.

VO₂peak increased 6% with training (44.5 ± 6.8 vs. 47.1 ± 5.7 ml·kgFFM^-1·min^-1, P = 0.03). In addition, at the highest work level of the maximal exercise test that all subjects were able to perform before and after training (115 W), there was a decrease in heart rate after training that approached statistical significance (155 vs 150 bpm, P = 0.07), and a significant reduction in the rate pressure product (2.5·10⁴ vs 2.2·10⁴ mmHg*bpm, P = 0.02). These changes are consistent with an improvement in aerobic fitness.

Training resulted in a significant increase in femoral artery diameter from 5.1 ± 0.5 to 5.7 ± 0.5 mm (P < 0.001), whereas femoral artery blood flow did not change (167 ± 89 vs. 160 ± 68, P = 0.15) (Figure 2). In line with these two findings, there was a significant decrease in calculated shear rate from 24.9 ± 9.5 to 17.6 ± 6.1 (P = 0.007). aPWV did not change with training (886 ± 196 vs. 903 ± 227, P = 0.30).

No significant changes in CRP, sICAM or leptin were observed with the training program (Table 2). Leptin values did correlate with adiposity (r = 0.81 pre, and 0.82 post). No significant differences in serum estradiol or progesterone were observed in the pre- versus post-training samples.

The key finding of this study was that femoral artery diameter increased with an aerobic training program that resulted in an increase in fitness without a decrease in adiposity in middle aged women. The exercise program used in this study was similar to commonly used aerobic training programs in that the exercise intensity was within the guidelines suggested by the American College of Sports Medicine 37. What was slightly different was our use of varying exercise intensities within each training session in order to provide variety and to reduce boredom for participants. The participants overwhelmingly viewed the training paradigm favorably and indicated that the program had achieved this goal. VO₂peak values increased 10% when normalized to body weight. When normalized to lean tissue, the increase in VO₂peak was 6%, which is slightly less than reported in a comparable study that also normalized VO₂peak to lean body mass 38. This may be due to a shorter training program in the present study, as well as subtle differences in the composition of exercise sessions between the two studies. We had similar rates of adherence to the training program. The improvements in aerobic fitness, in combination with the reduction in rate-pressure product during maximal exercise testing, suggest a significant overall physiological adaptation.

The training program implemented in this study resulted in femoral artery remodeling that was independent of adiposity. Physical activity 639 and aerobic fitness 40 have previously been found to be associated with larger femoral artery diameters. Although previous studies have found that aging and/or long-term atherosclerotic plaque accumulation also result in increased artery size as well as central arterial stiffness 41, the current study evaluated subjects over a period of only 16 weeks and the study population was homogeneous in age and free of cardiovascular disease. Our findings of increased aerobic fitness and decreased heart rate also suggest that sympathetic drive was reduced. One might expect that a reduction in sympathetic outflow to the legs would result in increased dilation and consequently an increase in artery diameter. However, previous data suggest that endurance-training results in either no or slightly elevated sympathetic drive to the leg at rest 42, which would have resulted in an opposite change in diameter. Therefore, it is more likely that the increase in femoral artery diameter found in this cohort is a sign of remodeling than of arterial deterioration or increased dilation.

The increase in arterial diameter was associated with a decrease in resting femoral artery shear rate. Low resting shear rates represent a positive adaptation to exercise training as elevated shear rates during exercise induce the positive arterial adaptations 434445. These adaptations result in larger arteries that subsequently have reduced shear rates at rest. Adults with spinal cord injury are unable to increase shear rates in their affected limbs and subsequently have smaller arteries and elevated resting shear rates compared to able-bodied control groups 910. This suggests that increased resting shear rate is at least associated with vascular deterioration since spinal cord injury also results in accelerated CVD progression 46. Therefore, our results can be cautiously interpreted to suggest that the training program improved vascular health of the study participants.

Exercise training did not decrease arterial stiffness (i.e., aPWV) in this study. Although cross-sectional studies suggest that higher fitness is associated with lower aPWV, age may help explain our results. Aortic PWV was found to increase with age in a group of 480 Chinese citizens (ages 3 to 89) in whom atherosclerosis is known to be rare and who did not have CVD 47. Also, cross-sectional studies of endurance trained older adults do not find the age-related increases in aPWV found in sedentary older adults 4849. These data suggest that age is a significant and independent predictor of central arterial stiffness and that its effects are lessened when accompanied by years of aerobic exercise. Therefore, central arterial stiffness may have not been sufficiently elevated in this study for exercise and increased fitness to evoke a measurable reduction in this study. Alternatively, the morphological aspects of the aorta that aPWV reflects, namely atherosclerotic plaque accumulation, loss of elastin, and collagen build-up, may not be significantly reversible in the time frame of this study 50.

Aerobic training did not reduce serum sICAM-1 concentrations. This result was contrary to our hypothesis that training-induced reductions in shear stress would reduce sICAM-1 levels. This prediction was based on evidence that shear stress increases ICAM-1 expression and release of sICAM-1 from isolated human saphenous vein endothelial cells 51. However other factors, including paracrine influences of IL-1β and TNFα can also increase ICAM-1 expression and shedding by endothelial cells 52. A year-long study of weight reduction through exercise and diet reported a significant 26% reduction in serum ICAM-1 24, whereas a 12-week aerobic training program with weight maintenance reported no changes in serum ICAM-1 53.

Aerobic training did not affect serum CRP concentrations. In fact, the intrasubject coefficient of variability in the present study (36%) was less than that reported in other studies which had no intervention of any kind (42–61%, reviewed in 54). A previous study reported reductions in serum CRP in middle-aged women after a 2-year weight loss program that coupled reduced caloric intake with increased physical activity 55. In contrast, serum CRP did not change in 9–15 year old girls after a 12-week aerobic training program involving a weight-maintaining diet 53. Other studies have found a positive relationship between CRP and adiposity 2456. Taken together, these studies indicate that CRP covaries with adiposity, but not aerobic fitness.

Consistent with previous studies 5758, we also found that leptin levels correlated with adiposity. While one study involving a very low energy diet and weight loss detected an independent effect of exercise training on leptin concentrations by partial regression analysis 58, our results were in agreement with a study reporting that exercise without fat loss has little effect on leptin 57. We found no evidence that exercise training and increased aerobic fitness lowered leptin concentrations independent of adiposity.

This study had some limitations that may limit interpretation of the results. For example, a potential criticism is the lack of a control group. Traditional cross-sectional studies use both experimental and control groups to examine the effect of the intervention or exposure. However, this study used a longitudinal, repeated measures design where the pre-treatment measures from the experimental group serve as the control. In the repeated measures design all the participants serve as their own control because they are involved in the experimental and control groups. We did not employ a time-matched control group that had measurements taken twelve weeks apart. Instead, each subject's pre-treatment data served as control data since the pre and post data represent measures of the same physiological variable, except without a preceding training period. This design is ideal when a limited number of participants are available because it increases the statistical power relative to sample size, and therefore the efficiency and sensitivity of the experiment 5960. In addition, we tested a small cohort without signs of cardiovascular disease. Therefore, we feel that the limitations of this study potentially result from the ability to generalize to diseased populations rather than from the lack of a control group. Although the results of this study should be interpreted with caution, it has provided salient findings using precise measurements that will assist in the design of future studies that are larger in breadth.

In conclusion, this study found that 14 weeks of aerobic exercise increased aerobic fitness and favorably altered resting femoral artery diameter and shear rate in middle-aged women without changing adiposity. Future studies seeking to exercise-train sedentary populations may do well to use variable intensity programs to increase subjects' interest and potentially increase adherence and retention. The interpretations of this study are inherently limited because long-term follow-up would be required to ultimately determine how these findings relate to progression of CVD. However, these results suggest that aerobic training positively influences functional measures of arterial health without a concomitant loss of body fat but that circulating levels of biological proteins associated with CVD risk may be more strongly linked to adiposity than to fitness and/or require longer exercise interventions.

The authors declare that they have no competing interests.

MJS conceived of the study, conducted and coordinated the experiments, performed data analysis, and drafted the manuscript. EHS participated in preparation of the study design, conducted experiments and exercise training, and participated in writing the manuscript. RL participated in the preparation of the study design and in writing the manuscript. GS maintained regulatory compliance and performed immunoassay of serum samples, and performed data analysis. JC maintained regulatory compliance, administered the budget, and participated in data analysis and writing sections of the manuscript. KM conceived of the study and participated in its design and coordination, and in writing the manuscript. All authors read and approved the final manuscript.