Details on Skaraborg population study have been described in detail previously [16]. In brief, it is a longitudinal observational study based on representative participants from Vara and Skövde municipalities in south-western Sweden. Between 2002 and 2005, a total of 2816 randomly selected individuals were examined (Men = 1400) for early signs off cardiovascular disease and risk factors for the development of CVD. After 10 years (9.7 years), and between 2012 and 2014, a representative sample of 1327 people were re-examined according to the same protocol as at visit 1 (Fig. 1). Two hundred eighty-eight participants do not have data on (Leisure Time Physical Activity) LTPA or arterial elasticity at visit 1, the corresponding number is 425 at visit 2. The number of participants with known hypertension was 104 at visit 1, and 141 at visit 2, respectively. Hypertensive medication can interfere with the elasticity indices significantly thus we chose to exclude participants with known hypertension at any of the visits. As such, in the cross-sectional analysis, 935 participants were included, while 761 participants were included in the longitudinal analysis at both visit 1 and 2 (Fig. 1).

Detailed information on the medical history including heart failure, atrial fibrillation, myocardial infarction, stroke, undergoing coronary heart surgery, as well as medication psychosocial and lifestyle was collected using validated questionnaires. All participants were asked alcohol [17] and smoking habits using validated questionnaires.

Body height and weight were measured in light clothes and no shoes to the nearest 1 cm and 0.1 kg, while waist- and hip circumferences were measured between the lowest rib margin and the iliac crest and at the largest circumference between waist and thighs, respectively. The blood pressure was measured to the closest 2 mmHg in the right arm after a five minutes rest in a supine position [16, 18]. Hypertension was defined based on Blood Pressure Management JNC-7 [19].

All participants underwent an oral glucose tolerance test (OGTT), whereby fasting and 2 h glucose values were obtained. Diabetes diagnosis was based on the 1999 WHO recommendation [20]. Fasting plasma insulin was analyzed using an enzyme-linked immunosorbent assay (ELISA) with < 0.3% cross-reactivity for proinsulin. Insulin resistance by HOMA-IR was calculated with the formula below [21]: fasting insulin (micro IU/L) x fasting glucose (nmol/L)/22.5. Assays for LDL cholesterol, triglyceride, C-reactive protein (CRP) and creatinine were conducted.

The level of physical activity was estimated using a question about Leisure Time Physical Activity (LTPA) which has been validated in several previous studies [22‐24]. All subjects were asked for their LTPA in the last 2 weeks. This scale has four levels based on four answer alternatives to the question “How much physical effort do you put yourself through in your leisure time?”.

Arterial compliance was measured based on diastolic pulse contour analysis of arterial waveform using the HDI/Pulse WaveTM CR2000 (Hypertension Diagnostics, Inc., Eagan, NY, USA). The tonometer was centered over the radial artery. The optimal radial artery waveform was obtained by non-invasive recording by applanation tonometry. This technique is based on a modified Windkessel model that gives an evaluation of the elasticity of the large conduit arteries (Large Arterial Elasticity Index, LAEI, C1) and the small microcirculatory arteries (Small Arterial Elasticity Index, SAEI, C2) [25, 26].

In this study, we showed a strong and independent association between leisure-time physical activities and small artery elasticity index independent of possible confounders both at baseline and approximately 10 years later. The association between SAEI and LTPA was strong and significant even after adjusting in the full model. These results indicate potential salutary effects of exercise on SAEI that may be persistent longitudinally.

Saladini et al. observed the effect of physical activity in 151 subjects with hypertension during a period of 6 years [14]. In line with our findings, they also found that physical activity predicted better SAEI. Moreover, in the Whitehall II study [15], the follow-up time was 5 years and found that moderate-to-vigorous, but not mild, activity was associated with slower progression of aortic stiffness over time. Notably, we were able to include HOMA-IR, diabetes mellitus, cardiovascular morbidity, LDL cholesterol and alcohol consumption as covariates in the models. Our study confirmed these findings and expanded them even in the non-hypertensive population. To our knowledge this is the first study to investigate the association between reported leisure time physical activities and small artery elasticity index in a middle-aged representative cohort of men and women without hypertension.

Physical activity has many benefits both with traditional cardiovascular risk factors and with mortality [28, 29]. In cross-sectional observations, higher artery elasticity is associated with being physically active [10, 12, 13, 30]. In line with these observations, our study found that all indices were significantly better with higher LTPA level at both visits before adjustments.

Small artery elasticity index (SAEI) seems to be strongly associated with changes in the artery wall associated with hypertension [14], diabetes mellitus, and atherosclerosis compared to LAEI [6]. Similarly, it is strongly associated with chronic heart disease, stroke and heart failure [4]. In our study, difference in LTPA levels seems to have a stronger effect on SAEI than LAEI. The stronger association between LTPA levels and SAEI compared with LAEI might depend on that the development of the atherosclerotic disease changes in the endothelial function before structural changes [5] including increased arterial stiffness, particularly in the smaller arteries [6]. In fact, physical activity reduces the plasma endothelin concentration the endothelin mediated vascular tone and thus reduces arterial stiffness [31]. Moreover, physical activity improves endothelial function by up-regulating endothelial nitric oxide synthase (eNOS) protein expression [32]]. Regular physical activity has long-term anti-inflammatory effects and improves metabolic control. It is suggested this occurs by skeletal muscle serving as an endocrine organ through a production and secretion of myokines in response to exercise. These mechanisms affect many organs and metabolic pathways, including endothelial cell function [33].

Arterial elasticity indices change over time because of structural modifications in the blood vessels wall like the degeneration of elastin material and an increase in collagenous material and in-ground substance [7]. The factors which can affect the stiffening processes are many, but want to underline ageing, blood pressure, blood glucose, lipids, diseases (hypertension, diabetes mellitus type 1 and 2, coronary artery diseases, heart failure, stroke), medication, lifestyle (smoking, alcohol consumption, physical activity). We have adjusted for these possible confounders in our statistical analysis and found that high level of leisure-time physical activity was still associated with small artery elasticity index suggesting an independent association with these variables.

All measurements were computed by the same trained staff using a strict protocol providing a strong internal validity in the measurements. The large variety of information and the variables provided by the study permitted the testing of different models. In this study, the inclusion of both men and women made it possible to investigate possible gender differences in the associations and interaction analyses. Even if these differences in SAEI and LAEI between groups with low respective high LTPA are small numerically, these differences are significant at group level and it is consistent with the expected effect of physical activity. A high participation rate at both visits gave higher power to our analyses. A weakness of the study is the use of self-reported physical activity, only providing a subjective estimation of the leisure time physical activities. This scale is however validated and seems to reflect with good approximation the physical activities. Our study was observational without randomization, thereby there is a risk for residual confounding.

Cardiovascular prevention is one of the main tasks in primary care. Early introduction of lifestyle change especially physical activity might be important to reduce cardiovascular risk through positive effects on vascular elasticity. Public health strategies introducing leisure-time physical activities early and helping the population to maintain it during life should be effective in maintaining good vascular health, reducing the incidence of hypertension and cardiovascular disease.