Association of physical activity with vascular aging in a population with intermediate cardiovascular risk, analysis by sex: MARK study

The aim of this study was to analyze the association of physical activity and its intensity with arterial stiffness and vascular aging and differences by sex in a Spanish population with intermediate cardiovascular risk. Cross-sectional study. A total of 2475 individuals aged 35–75 years participated in the study. Brachial–ankle pulse wave velocity (baPWV) was measured using a VaSera VS-1500® device. Based on the age and sex percentile presented by the participants, the latter were classified as follows: those with a percentile above 90 and presenting established cardiovascular disease were classified as early vascular aging (EVA); those with a percentile between 10 and 90 were classified as normal vascular aging (NVA) and those with a percentile below 10 were classified as healthy vascular aging (HVA). Physical activity was analyzed through the short version of the Minnesota Leisure Time Physical Activity Questionnaire (MLTPAQ). The mean age of the participants was 61.34 ± 7.70 years, with 61.60% men. Of the total sample, 86% were sedentary (83% men vs 90% women). The total physical activity showed a negative association with baPWV (β = − 0.045; 95% CI − 0.080 to − 0.009). Intense physical activity showed a negative relationship with baPWV (β = − 0.084; 95% CI − 0.136 to − 0.032). The OR of the total physical activity and the intense physical activity carried out by the subjects classified as NVA with respect to those classified as HVA was OR = 0.946; (95% CI 0.898 to 0.997) and OR = 0.903; (95% CI 0.840 to 0.971), and of those classified as EVA it was OR = 0.916; (95% CI 0.852 to 0.986) and OR = 0.905; (95% CI 0.818 to 1.000). No association was found with moderate- or low-intensity physical activity. The results of this study suggest that, when intense physical activity is performed, the probability of presenting vascular aging is lower. In the analysis by sex, this association is only observed in men. This study is the first to find a negative association between total physical activity and arterial stiffness and vascular aging in adults with intermediate cardiovascular risk. This study also shows that the association of physical activity with arterial stiffness and vascular aging in adults with intermediate cardiovascular risk varies depending on its intensity, with a negative association being found only with intense physical activity. In the analysis by sex, the results found suggest that the association is greater in men. This study is the first to find a negative association between total physical activity and arterial stiffness and vascular aging in adults with intermediate cardiovascular risk. This study also shows that the association of physical activity with arterial stiffness and vascular aging in adults with intermediate cardiovascular risk varies depending on its intensity, with a negative association being found only with intense physical activity. In the analysis by sex, the results found suggest that the association is greater in men.


Background
Physical activity performed regularly at moderate-high intensity reduces the early morbidity and mortality of a large number of cardiometabolic diseases, tumors and mental diseases. Similarly, in both primary and secondary prevention, physical activity reduces or improves the prognosis of numerous medical affectations, increasing the quality of life of the people who practice it, with the corresponding benefits for the healthcare system and society [1][2][3][4][5][6][7]. Thus, most international guides report about the type, amount and intensity of the physical activity necessary to maintain and improve population health [2,8], recommending 150 min/week of intensemoderate physical activity and considering this the minimum required to obtain health benefits. Moreover, recent evidence suggests that including low-intensity activities, at least for currently inactive populations, can be beneficial for health [3,8]. Arterial stiffness, evaluated through brachial-ankle pulse wave velocity (baPWV), increases with age and predicts morbimortality by cardiovascular diseases [9]. Several studies and systematic reviews conducted in different populations indicate that physically active individuals have a lower arterial stiffness than sedentary individuals, suggesting that regular physical activity is an effective index of arterial stiffness [10][11][12]. However, depending on its intensity and type, the effect of physical activity on arterial stiffness is still controversial. Thus, some authors propose that the benefit is only obtained with exercises of moderate-high intensity, and that such benefit can vary depending on the type of exercise performed [13][14][15] and by sex [16]. Arterial stiffness is associated with the appearance of vascular aging [17,18], reflecting the dissociation between chronological age and biological age of the main arteries, with their alteration preceding the appearance of cardiovascular events [17][18][19]. In recent decades, epidemiological studies have been carried out to determine the key factors of vascular aging, which has gained great interest, as it shows a stronger relationship with the morbimortality by cardiovascular diseases than biological aging [17,18]. These studies suggest that vascular aging is healthier in the Spanish population than in other countries. Thus, in a general Spanish population without prior cardiovascular disease, a study showed that vascular age and heart age were − 3.08 ± 10.15 and − 2.98 ± 10.13 years younger, respectively. Likewise, 63.2% (men: 59.3%; women 67.1%) and 66.11% (men: 52.0%; women 71.1%) had vascular and heart ages below their chronological age [20]. These results are more favorable than those published in the review by Groenewegen et al. [21], which found that the difference between vascular age or heart age and chronological age was between 2 and 26 years higher, results consistent with those published in various studies [22][23][24][25]. Along the same lines, a second study carried out on a Spanish population analyzed vascular aging through arterial stiffness measured with the cfPWV and found that one in ten had HVA and one in five EVA. The prevalence of EVA was higher in men, and vascular aging deteriorated with increasing sedentary time and decreasing physical activity, as measured objectively over a week with an accelerometer [26]. These data are generally more favorable than those published in other countries [27][28][29][30][31]. However, it should be noted that the prevalences found in previous studies are not comparable, given that the criteria used to define HVA and EVA, the distribution of the population by age and sex, and their characteristics regarding the presence of risk factors and morbidity vary across the studies analyzed.
However, depending on the intensity and type of physical activity, and sex, its effect on arterial stiffness may vary. Thus, Vandercappellen et al. [13] examined the association of the amount and pattern of intense physical activity with arterial stiffness in the Maastricht study. They reported that while a negative association between intense physical activity and arterial stiffness was found, this was not the case with regular or weekend physical activity. Königstein et al. [14] evaluated the effects of light, moderate and vigorous physical activity independently of cardiorespiratory fitness on arterial stiffness, finding a reduction of 0.32 m/s for every 10-min increment in moderate or vigorous intensity physical activity per day, while physical activity of light intensity had no relevant effect on baPWV. Finally, in a review carried out by Figueroa et al. [15], aiming to examine the effects of resistance training intensity on arterial stiffness and blood pressure (peripheral and central) in young and older adults, it was found that high-intensity resistance training could decrease systemic arterial stiffness in healthy young adults, or not affect arterial stiffness in middle-aged and older adults. On the other hand, several authors have described differences between the sexes. Thus, Seals et al. [32] found that regular aerobic exercise improved endothelial function and vascular aging in men (by reducing oxidative stress and preserving NO bioavailability), but not consistently in estrogen-deficient post-menopausal women [32]. Conversely, Collier et al. [16] found that moderate-intensity physical exercise may be more favorable for women, as a treatment option for hypertension, due to greater reductions in diastolic blood pressure and increase in flow-mediated arterial dilation without increase in arterial stiffness compared to men. For all the above, the aim of this study was to analyze the association of physical activity and its intensity with arterial stiffness and vascular aging in a Spanish population with intermediate cardiovascular risk and explore the existence of differences between sexes.

Study design and population
The "improving intermediate risk management" study (MARK study) is a multicentre cross-sectional study registered in ClinicalTrials.gov: NCT01428934 [33].
The participants were recruited by family doctors and selected by random sampling among individuals visiting primary healthcare centers of three autonomous communities of Spain between July 2011 and June 2013. Inclusion criteria: (1) age between 35 and 75 years; (2) intermediate cardiovascular risk defined with equation A (5-15% coronary risk estimated with Framingham's risk equation adapted to the Spanish population [34]), B (1-5% vascular mortality risk in 10 years estimated with the equation for the European population [35]), or C (moderate risk following the criteria of the European Society of Hypertension for the management of arterial hypertension [36]); and (3) signed the informed consent before being included in the study. Exclusion criteria: terminal disease, hospitalization, or history of atherosclerosis. A total of 2495 individuals were recruited. This manuscript presents an analysis of 2475 individuals recruited in the MARK study, excluding 20 individuals for not having their baPWV measured.
To estimate the power of the sample, we considered the difference in the main variable, total PA, between the HVA group (241 subjects) and the NVA group (1927 subjects) and between the HVA group (241 subjects) and the EVA group (307 subjects). Therefore, for a bilateral contrast and assuming an alpha risk of 0.05, in the first case, with a SD of 2525, the estimated power was 32% and in the second case, with a SD of 2522, the estimated power was 58%.
A detailed description of the procedures followed by the gathering of clinical data, anthropometric measures, cardiovascular risk factors and laboratory tests has previously been published in the protocol of the study [33].

Ethical approval and consent to participate
Before being included in the study, the participants were briefed on all the procedures that they would be subjected to. This study was revised and approved by the Research Ethics Committee of the Jordi Gol Primary Healthcare Research Institute, the Drug Research Ethics Committee of Salamanca, and the Research Ethics Committee of Palma de Mallorca. During the realization of the study, the rules of the Declaration of Helsinki were followed [37].

Variables and measurement instruments
Before starting the study, healthcare professionals were trained in each data-gathering center to record the measurements and complete the questionnaires following a standardized protocol.

Evaluation of physical activity
The physical activity performed was recorded using the short version of the Minnesota Leisure Time Physical Activity Questionnaire (MLTPAQ) [38], which was designed to evaluate the amount and intensity of physical activity performed in leisure time (leisure and housekeeping activities), in the year prior to questionnaire completion, in middle-aged American men [39]. Elosua et al. validated the version of this questionnaire in Spanish men and women aged 18-61 years [40,41]. This original questionnaire consists of 67 items, whereas the short version of the questionnaire used in this work [38] is made up of six items, which gather information about different types of physical activity: strolling, walking, farming/gardening, trekking, climbing stairs and practicing sports/exercising. The questionnaire evaluates the intensity of physical activity as low (strolling), moderate (walking, farming/gardening) and intense (trekking, climbing stairs and practicing sports/exercising).
The physical activity questionnaire was completed through a personal interview with qualified interviewers, who collected detailed information about the physical activity of the participants in the year before the study, the number of times they performed physical activity, and the average duration of each physical activity in each period. From the amount, type, frequency and duration of each session of physical activity, the baseline metabolic equivalents of task (MET) were calculated (MET) using the proposition of Ainsworth et al. [42] as a reference. One MET equals 1 kcal/kg of body weight per hour, and a consumption of 3.5 ml of oxygen/kg of body weight per min. The consumption of MET-min/14 days was estimated by multiplying the METs of each physical activity by its duration (in min), its accumulated frequency in the month prior to the interview (or habitual month), and by the months per year that the activity was performed, dividing by 365 days/year and multiplying by 14 days.
Following the recommendations of the American Heart Association, the participants were classified as sedentary if the moderate physical activity performed was < 675 METs min/week or the intense physical activity was < 420 METs min/week [43].

Brachial-ankle pulse wave velocity (baPWV)
The baPWV was measured using a VaSera VS-1500 ® device (FukudaDenshi), which employs an oscillometric method. The participants were requested to refrain from smoking or consuming caffeine one hour before the test and to attend the latter with comfortable clothing, remaining at rest for at least 10 min before conducting the measurements. The cuffs were adapted to the circumference of the arms and legs of both limbs. The electrodes were connected to the right and left arms and ankles of the participants, fixing a heart sound microphone to the chest, over the sternum, in the second intercostal space. The baPWV was calculated with the following equation: baPWV = (0.5934 × height (cm) + 14.4724)/tba (tba is the time interval between the arm and ankle waves) [44].

Definition of vascular aging
In the selected sample, subjects with cardiovascular disease were identified. Therefore, in the first step, 20 individuals who presented cardiovascular disease (kidney disease, peripheral arteriopathy or heart failure), based on the criteria established in the 2013 clinical practice guide of the European Associations of Hypertension and Cardiology for the treatment of arterial hypertension [36], were classified as early vascular aging (EVA). Secondly, we classified the individuals using the 10th and 90th percentiles of baPWV by age group and sex. The individuals with baPWV values above the 90th percentile were considered to have EVA, those between the 10th and 90th percentiles were classified as having normal vascular aging (NVA), and those with values below the 10th percentile were considered to have HVA. The 10th and 90th percentile values of the ba-PWV by age and sex are shown in Additional file 2: Table S1.

Statistical analysis
The continuous variables are shown as mean ± standard deviation, and the categorical variables are presented as number and percentage. The comparison of means between the two independent groups was carried out with Student's t-test, and between more than two groups using a one-way analysis of variance. The post hoc analyses to explore the differences between more than two groups were conducted with the least significant difference test (LSD). In the comparison of categorical variables, the χ 2 test and Fisher's exact test were used. To analyze the association of physical activity with baPWV, two multiple linear regression analysis models were applied. Model 1, used baPWV as a dependent variable and total physical activity (measured in METs/min/ week) as independent variable. Model 2 used baPWV as a dependent variable and intense physical activity, moderate physical activity and low physical activity (all included in the model and measured in METs/min/week) as independent variables. In both models, the adjustment variables we used were age (years), sex (woman = 1, man = 0), years of smoking, grams of alcohol per week, adherence to the Mediterranean diet, systolic blood pressure (mmHg), glycosylated hemoglobin (%), body mass index (kg/m 2 ), antihypertensive drugs, and lipid-lowering and hypoglycemic drugs (Yes = 1 and Non = 0). To analyze the association of physical activity with vascular aging, we performed two multinomial logistic regression models (Model Fitting Criteria -2 Log Likelihood of Reduced Model). In Model 1, the degree of vascular aging was used as a dependent variable, coded as HVA = 1, NVA = 2 and EVA = 3, using HVA as the reference value and total physical activity (measured in METs/min/week) as independent variable. In Model 2, the degree of vascular aging was used as a dependent variable, coded as HVA = 1, NVA = 2 and EVA = 3, with HVA as the reference value and intense physical activity, moderate physical activity and light physical activity (all included in the model and measured in METs/min/week) as independent variables. Both models were controlled for age (years) and sex (woman = 1, man = 0), with years of smoking, grams of alcohol per week, adherence to the Mediterranean diet, type 2 diabetes mellitus, arterial hypertension, obesity, antihypertensive drugs, lipid-lowering drugs and hypoglycemic drugs (Yes = 1 and Non = 0) used as adjustment variables. The analyses were conducted globally and by sex. In the hypothesis test, an alpha risk of 0.05 was established as the limit of statistical significance. All the analyses were performed using the SPSS v.25 software for Windows (IBM Corp, Armonk, NY, USA).

Clinical characteristics and vascular aging
The cardiovascular risk factors and diseases, lifestyles, total physical activity, intensity of the physical activity and vascular function of the participants are shown in Table 1 globally and by sex. The men performed more total physical activity, of high-and moderate-intensity, and consumed more alcohol and tobacco, whereas the women showed greater adherence to the Mediterranean diet (MD). Table 2 shows the total physical activity and its different levels of intensity as a function of the degree of vascular aging in the global sample and by sex. The individuals classified as HVA globally and in men performed more total physical activity and of high intensity compared to the individuals classified as NVA and EVA. No differences were found in the women.
The prevalence of individuals classified as HVA, NVA and EVA globally and by sex is shown in Fig. 1, finding differences between sex in the proportion of subjects in each category (p = 0.031). Subjects performing ≥ 2500 METs/m/W made up 36.4% of the sample, with 42.1% ≥ 625 and < 2500 METs/m/W and 21.5% < 625 METs/m/W. The distribution according to the degree of vascular aging is shown in Fig. 2, with no differences found between groups (p = 0.146).

Association of arterial stiffness and vascular aging with physical activity and its intensity
In the multiple regression analysis, after controlling for possible confounding factors, the total physical activity showed a negative association with baPWV (β = − 0.045; 95% CI − 0.080 to − 0.009). Intense physical activity showed a negative relationship with baPWV (β = − 0.084; 95% CI − 0.136 to − 0.032), as can be observed in Table 3.
The results of the multinomial logistic regression analysis are presented in Table 4 and in Additional file 1: Fig.  S1, globally and by sex. The OR of the total physical activity and the intense physical activity carried out by the subjects classified as NVA with respect to those classified as HVA was OR = 0.946; (95% CI 0.898 to 0.997) and OR = 0.903; (95% CI 0.840 to 0.971) and of those classified as EVA was OR = 0.916; (95% CI 0.852 to 0.986) and OR = 0.905; (95% CI 0.818 to 1.000). No association was found with moderate-or low-intensity physical activity. In the analysis by sex, there is no interaction of sex and physical activity with arterial stiffness or vascular aging.

Discussion
This study analyzed the association of physical activity and its intensity (measured with the short version of MLTPAQ) [38] with arterial stiffness and vascular aging in a Spanish population with intermediate cardiovascular risk. The results show that total physical activity and high intensity were negatively associated with arterial stiffness and improved the degree of vascular aging in the total sample. Along these lines, previous studies have shown that regular physical activity and physical training delay the increase of arterial stiffness related to aging. Moreover, it has been observed that physical activity improves many other aspects of human health in different age groups, in healthy individuals and in those with cardiovascular risk factors, measured subjectively (through questionnaires) or objectively (using accelerometers), suggesting that arterial stiffness is negatively associated with regular physical activity [10][11][12]26]. In the present study, this association was only found in men and with high-intensity physical activity, but results vary across different studies. Thus, Kozakova et al. [45] detected a reverse relationship between regular physical activity measured with an accelerometer and arterial stiffness in healthy adults in both men and women [45], although other authors have not found this association in women [46]. Similarly, and in agreement with the results of this study, the benefit of physical activity on arterial stiffness was only achieved when its intensity was high [14,15,47,48]. Similarly, the recent study by Vandercappellen et al. [13] reported that practicing intense physical activity was associated with lower arterial stiffness, detecting no differences among different exercise patterns; thus, they concluded that, from the perspective of arterial stiffness, performing intense physical activity, regardless of the weekly pattern conducted, can be an important strategy to reduce the risk of cardiovascular disease. Moreover, the effects on arterial stiffness vary depending on the modality of physical activity performed. It has been reported that aerobic exercise has a beneficial effect on arterial stiffness, whereas strength exercise has different effects on arterial stiffness depending on the type and intensity of the exercise [49]. Likewise, a meta-analysis that reviewed 38 clinical trials, with a total of 2089 patients, concluded that aerobic and strength exercise reduced arterial stiffness [50]. In agreement with these results, previous studies have reported that greater physical activity (measured with an accelerometer) is associated with lower vascular stiffness [51]. Therefore, the current results suggest that regular physical activity is an effective index of arterial stiffness, and the effect of physical activity can be improved by increasing its intensity [10].
Similarly, in the present study, the beneficial effects of physical activity on arterial stiffness were confirmed on vascular aging. It is known that the regular and moderate-intense practice of physical activity in the general population improves vascular aging, as it reduces   oxidative stress and preserves the bioavailability of NO [50,52,53]. In line with the results of this study, several authors have reported that healthy vascular aging is associated with greater physical activity [24,53,54]. Furthermore, regular aerobic exercise also improves the endothelial function with age in men, although this is not constant in post-menopausal women with estrogen deficiency. According to these findings, a certain, currently unknown estrogen blood level (greater than that found in post-menopausal women who are not being treated with estrogens) may be required for the physiological stimulus created by regular aerobic exercise to improve the endothelial function in post-menopausal women [32,55,56]. Furthermore, in women, reductions in circulating estrogens associated with menopause interact with other cellular aging processes to influence vascular function through mechanisms that, although not fully understood, also appear to involve oxidative stress and low-grade inflammation [32,[57][58][59]. However, the association of vascular aging with physical activity was not found in the Framingham study [28]. Therefore, considering the numerous benefits induced by regular physical activity, it seems logical that the harmful effects of aging on vascular function can be counteracted with the regular practice of physical activity [10]. The discrepancies found among the studies that analyzed the effect of physical activity and its intensity on arterial stiffness and aging may be caused by different factors: (1) samples of individuals of different ages, (2) physical activity measured with different methods (subjectively through questionnaires or objectively with accelerometers), and (3) arterial stiffness measured with different methods and using different devices.
This study has some limitations that must be highlighted. Firstly, the cross-sectional analysis does not allow causality to be inferred. Secondly, the results of this study only refer to the Spanish population of intermediate vascular risk, and they cannot be generalized to other groups or ethnicities. Thirdly, the physical activity was self-reported by the participants through validated questionnaires, that is, subjectively, since people tend to Table 3 Multiple regression analysis of arterial stiffness with physical activity Model 1. Multiple regression analysis using baPWV m/s as dependent variables, total physical activity independent variable, and age (years), sex (woman = 1 and man = 2), years of smoker, alcohol consumption in g/week, Mediterranean diet adherence score, systolic blood pressure in mmHg; glycosylated hemoglobin in %, body mass index, antihypertensive drugs, lipid-lowering drugs and hypoglycaemic drugs (Yes = 1 and Non = 0) as adjustment variables Model 2. Multiple regression analysis using baPWV m/s as dependent variables, high-intensity physical activity, moderate-intensity physical activity and low-intensity physical activity independent variable, and age (years), sex (woman = 1 and man = 2), years of smoker in years, alcohol consumption in g/week, Mediterranean diet adherence score, systolic blood pressure in mmHg; glycosylated hemoglobin in %, body mass index, antihypertensive drugs, lipid-lowering drugs and hypoglycaemic drugs (Yes = 1 and Non = 0) as adjustment variables remember moderate and intense physical activity more accurately than low physical activity.

Perspectives and significance
In clinical practice, health professionals may ask about the level and intensity of patients' physical activity performed in order to obtain more information that can help them better understand vascular health. In this study, the importance in vascular aging of the intensity of the physical activity performed is also described. For this purpose, more large-scale prospective cohort studies are needed to explore the process of vascular aging according to the amount and intensity of physical activity performed in adult patients and to analyze whether there are differences between the sexes, as well as the possible mechanisms involved in them.

Table 4 Association of physical activity with vascular aging
Multinomial logistic regression analysis Model 1. Multiple logistic regression analysis using vascular aging (HVA = 1; NVA = 2 and EVA = 3) as dependent variables (HVA reference), total physical as independent variable, as adjustment variables in the global analysis: age in years, consumption of alcohol in grams/week, years of smoking, sex (woman = 1 and man = 2), adherence to the Mediterranean diet, type 2 diabetes mellitus, arterial hypertension, obesity, antihypertensive drugs, lipid-lowering drugs and hypoglycaemic drugs (Yes = 1 and Non = 0) as adjustment variables). Model Fitting Criteria-2 Log Likelihood of Reduced Model Model 2. Multiple logistic regression analysis using vascular aging (HVA = 1; NVA = 2 and EVA = 3) as dependent variables (HVA reference), high-intensity physical activity, moderate-intensity physical activity and low-intensity physical activity independent variable, as adjustment variables age in years, consumption of alcohol in grams/week,, years of smoking, sex (woman = 1 and man = 2), adherence to the Mediterranean diet, type 2 diabetes mellitus, arterial hypertension, obesity, antihypertensive drugs, lipid-lowering drugs and hypoglycaemic drugs (Yes = 1 and Non = 0) as adjustment variables.