Impact of Cardiovascular Risk Factors on Arterial Stiffness in a Countryside Area of Switzerland: Insights from the Swiss Longitudinal Cohort Study

The results of this study, collected in two countryside villages in Southern Switzerland, confirm the correlation between aortic pulse wave velocity and age, arterial hypertension, hyperlipidemia, diabetes mellitus, and obesity, and further emphasize a cumulative impact of a series of cardiovascular risk factors on wall properties of large arteries. Importantly, the study demonstrates that traditional cardiovascular risk factors are important modulators of arterial stiffness [8‐11, 13] also in rural surroundings [15]. The investigated countryside area, known to have a favorable cardiovascular health profile [19], is characterized by a low migration rate, therefore offering ideal candidates to explore the prognostic importance of the observed associations over a longitudinal follow-up, as foreseen in the SWICOS project.

The novelty of this study is the focus on a countryside area and the differential attention paid to male and female participants in different age groups. Indeed, in this analysis, traditional cardiovascular risk factors were more prevalent among men as compared to women. Interestingly, however, the aortic pulse wave velocity was more often pathologically altered in women over 40 years of age than in men of the same age. These data are basically in line with data obtained in urban areas. In women, the arterial stiffness is lower than in age-matched men from puberty to menopause [22]. However, this trend reverses during and after menopause, with a steeper increase of aortic pulse wave velocity in women compared to men [23]. The mechanisms underlying this increase in aortic pulse wave velocity in postmenopausal women are unknown. It has been speculated that this evolution might reflect hormonal changes occurring after menopause, given that estrogens exert a beneficial effect on cardiovascular health. However, estrogen supplementation does not decelerate progression of arterial stiffness in postmenopausal women [22]. Alternatively, it has been suggested that increased arterial stiffness in older women results from an elevation in proinflammatory cytokines, which follows the decline in ovarian function with menopause [24]. Interestingly, some previous works concluded that no sex difference in pulse wave velocity exists after correcting for other cardiovascular risk factors [25]. The finding of a significant difference between sexes in our study may therefore appear, at first look, surprising. However, looking at the age subgroups (Fig. 2, Table 2), it becomes clear that age has a major impact in shaping this difference. Although previous works found such an association exclusively in older, peri- and postmenopausal women [26], our data detected a significant difference already in the group 40–60 years old. We are unable to state whether this is because of a different region and population (genetic background, lifestyle factors, environmental influences, proportion of postmenopausal women in the age group 40–60 years of age, etc.) or because of a secular trend.

Arterial wall properties are the result of complex biological interactions. Thus, the value of pulse wave velocity needs to be carefully addressed, and this is why multivariable analysis, accounting for traditional cardiovascular risk factors, is so important in interpreting our results. There is still some controversy regarding whether systolic or mean blood pressure should be used in such models, however. Unsurprisingly, in both correlation analysis and simple linear regression, pulse wave velocity was significantly associated with both systolic and mean blood pressures (Supplementary Table S1). From a biological perspective, while systemic vascular resistance is mainly determined by the peripheral arteries and it is translated into the mean arterial pressure, pressure fluctuations in the circulation depend on arterial compliance, mainly determined by aorta and large arteries [27, 28]. It is therefore usual to appreciate arterial compliance changes while assessing pulse pressure and systolic blood pressure, but not necessarily mean arterial pressure [29]. Thus, although several studies had detected a mathematical relationship between pulse wave velocity and mean blood pressure, we felt that biologically it made more sense to correct pulse wave velocity (which is a measure of arterial stiffness) by systolic rather than by mean blood pressure (Table 5 and Supplementary Table S3). As this issue is still controversial, sensitivity analyses correcting for mean blood pressure were performed (Supplementary Tables S2 and S4), mainly confirming the same results and therefore supporting the robustness of our analysis.

This study has strengths and limitations, partly related to the recruitment strategy, which was based on voluntary participation. Both general practitioners and regional authorities were informed about the project, and participation was offered free of charge to local residents. The voluntary participation might be seen either as a limitation (volunteering bias) or a strength (high population coverage) of this study. Indeed, the bias effect due to the voluntary, non-randomized inclusion of the subjects in this long-term cohort project may have been substantially tempered by the high proportion of participants, when compared to the whole population. The similar sex distribution in participants and the whole population further supports this assumption. The failure to assess the participants’ menopausal status and the cross-sectional design of the current analysis are relevant limitations of this analysis.

The methods used to measure brachial blood pressure and central hemodynamic parameters can be seen both as a strength and as a limitation. The standardized measurement of peripheral and central hemodynamic factors with an operator-independent, non-invasive device is a notable strength of this study [10‐13]. Moreover, patients were allowed to sit for 5 min and at least two blood pressure readings were taken, as recommended at the time of study performance [30]. However, only the second blood pressure value was retained (and not the mean of the last two readings as recommended by the newer guidelines [6]). Furthermore, for pulse wave velocity we tolerated a beat-to-beat variation up to 1.0 m/s, while some authorities recommend to repeat a third measurement when the first two differ by more than 0.5 m/s [21]. Although most studies on pulse wave velocity used tonometric devices, we used an oscillometric device, which is user-friendly and easy to integrate into routine clinical care. While this represents a potential strength, Arteriograph^® comes with its own limitations. First, it is unknown whether the cutoff value of 10 m/s traditionally used for tonometric devices can be applied also to the Arteriograph^®. Second, it is virtually impossible to determine the exact location of the artery reflection wave [31] and its dependency from heart rate might have influenced the magnitude of the significant association we detected between heart rate and pulse wave velocity. However, it is important to note that this device has been validated both with respect to oscillometric devices [32] and invasive measures [12].