In this cross-sectional study involving participants ages 45 years and above, we found a lower odd of RLS associated with a higher score of overall ideal CVH metric or its individual metrics in BMI, physical activity, blood pressure and TC. Consistently, a greater proportion of RLS was observed in those who rarely conducted physical activity, who had diabetes, who had higher levels of TG and FBG, who got poor level of total and specific LS7 metrics (e.g., BMI, physical activity, blood pressure, and TC) than non-RLS. The pattern of the association was largely robust to confounder adjustment but was significantly modified by sex and age stratifications.
An increasing number of studies have reported the associations between CVDs and RLS. For instance, a prior cross-sectional study in 3,433 middle-aged and elderly people observed independently association between RLS and CAD (OR = 2.05; 95% CI 1.38–3.04) and CVDs (OR = 2.07; 95% CI 1.43-3.00) [12]. The similar pattern of association of RLS with MI (HR = 1.80; 95%CI 1.07–3.01) and CHD (HR = 1.46; 95% CI 0.97–2.18) was also reported in another cross-sectional study based on 70,977 women in the Nurses’ Health Study [13]. Of note, our study captured the beneficial effect of a favorable CVH metric score against the RLS. However, a study by Dredla BK et al. [24] did not observe a significant association between CVH metric and RLS, being contradict with our findings. Given that their study [24] was basically performed among adults Amerindians aged ≥ 40 years in South America, the variability in racial or ethnic factors across study populations may be responsible for the controversial findings as inheritance has been known to play a potential role in the etiology of RLS[26]. Furthermore, the relatively small sample size (665) in their study [24], might have led to underestimation of their results due to the insufficient testing power, thus contributing to the discrepancies between our study and their report. Although the mechanisms underlying the association between the CVDs and RLS are not fully understood, the periodic limb movement burden during sleep (PLMS) is related to incident CVDs [27] and increased blood pressure [28] due to the sympathetic activation accompanying PLMS. Autonomic dysregulation is a hallmark of RLS, and the presence of PLMS is commonly found in patients with RLS [24], and this means that cardiovascular health might be worse in those had RLS coexisting PLMS [28].
The CVH metric was defined by SL7, so it may be assumed that different prevalence and incidence of RLS could be attributed to different health profiles. Common risk factors (e.g., female sex, smoking, HDL-C, LDL-C, TC) were found to be significantly associated with RLS [14], and our study reinforce the idea that the percentages of those who attaining ideal metrics for the overall CVH metric and specific LS7 (BMI, physical activity, blood pressure and TC) was higher in non-RLS group than in RLS group. These suggest that ideal adherence to these healthy life recommendations might be negatively associated with RLS. Even though there are limited studies directly concerning the relationship of each LS7 profile with RLS, prior studies have found the association between higher scores of ideal LS7 and better brain or neurological health. For instance, a study based on 1,987 subjects from the Washington Heights-Inwood Columbia Aging Project (WHICAP) found that a higher the LS7 components of physical activity was associated with lower risk of dementia among elder population [29]. Another study based on UK-biobank also suggested that adherence to ideal metrices of blood pressure, TC and FBG might offset the risk of dementia [30]. Higher scores of SL7 components (blood pressure, TC and FBG) might alleviate the pathology of AD by reducing pathological biomarkers in cerebrospinal fluid [31]. Therefore, understanding the effect of total CVH metric or LS7 on RLS is of great value for recognizing risk factors or helping patients to improve the prognosis.
It is in line with the studies by Xiang Gao et al., [7] and K De Vito et al., [32] which state that obesity was associated with increased risk of developing RLS, we found that individuals with RLS were more prone to have higher BMI level (presented as poor BMI metric). The increased RLS risk by high BMI could be explained by the reductions in dopamine D2 receptor. On one hand, dopamine deficiency could lead to obesity because dopamine is a neurotransmitter modulating motivation or reward circuits of foods [33]. On the other hand, low doses of dopamine agonists or α2δ ligands are uniquely recommended in clinical therapy of RLS [34] because a variety of cognitive, behavioral, and sensory-motor functions are regulated by the dopaminergic system[35]. In addition, iron deficiency, a common known risk factor of RLS, is also positively associated with obesity or overweight [36]. As for physical activity, one of the common cardiovascular related factors, Philips et al., [37] found a significantly lower prevalence of RLS in subjects exercising more than three hours a month compared with subjects exercising less than three hours a month. Conversely, insufficient physical activity close to bedtime was associated with increased prevalence of RLS [38]. Moreover, undertaking moderate exercises, particularly light physical activity, in the evening could alleviate the symptoms of RLS[39], which further confirm the beneficial effect of physical activity against RLS. There is still a lack of knowledge about the mechanism(s) through which exercise might relieve RLS symptoms. One explanation is the positive effect of physical activity on the β-endorphin system. The β-endorphin is an endogenous opioid that promotes feelings of well-being and pain relief, while a defective opioid system might be part of the pathophysiology of RLS [40]. Besides, aerobic exercise may improve RLS symptoms by increasing blood flow to the brain and HD efficiency [41]
Interestingly, according to the findings of a German study based on two cohort studies (the Dortmund Health Study [n = 1312] and the Study of Health in Pomerania [n = 4308]), hypercholesterolaemia and hypertension have both been known as independent predictors of RLS incidence [42]. Our results reconfirmed these conclusions. Regarding to hypercholesterolaemia, prior studies conducted in US [32], in Israel [14], or in China [8] have demonstrated that RLS patients were more likely to have a disorder of lipid metabolism than those non-RLS. Although there is no consensus on the potential mechanisms causing RLS, it is well known that RLS patients tended to feel uncomfortable sensations and urge to stretch, move their legs and even walk during sleeping, thus contributing to sleep fragmentation, and sleep disorders are associated with hypercholesterolaemia and hypertension[43]. For instance, a Korean study showed that RLS patients were prone to have lower quality of sleep, and RLS patients suffering from insufficient or low quality of sleep tended to have worse serum lipid profile (higher LDL-C and TC) [44]. Furthermore, most prospective studies have reported significant elevations in nocturnal blood pressure in adults with RLS[28, 45]. Among them, the blood pressure and heart rate during sleeping could concomitantly rise after periodic limb movements indicating autonomic activation [43, 45]. Another possible explanation is that RLS symptoms may be attenuated by a wide range of common antihypertensive drugs, including certain alpha-2 agonists and beta-blockers, supporting a possible role of autonomic dysfunction in RLS aetiology [46].
Our stratified analysis pointed out that sex of female exerted significant modification effect on the association between smoking metric and RLS, and female with low score of smoking metric was more likely to develop RLS compared with the males. A prior epidemiologic literature in France also observed a higher prevalence of RLS in women rather than in men (10.8% vs. 5.8%, P < 0.001) [47]. This female-specific vulnerability of RLS might be partly ascribed to the sleep initiation insomnia, which is more prevalent in females than males [48]. In addition, as a not uncommon risk factor of RLS since the physiological bleeding during menstruation in women, dysfunction in iron metabolism [48], as well as other hormonal factors[49], contributes to the pathophysiology of RLS in women. Thus, the relationship between smoking and RLS was aggregated by female sex in our study. Similarly, significant interaction of age was found in the negative association between overall CVH metric and RLS, and the benefit of CVH metric to RLS was more evident in elderly participants aged ≥ 60 years. Although RLS can appear at any age, the vulnerability of RLS increases with age [34]. The prevalence of high CVH was lower at older ages as aging is a significant risk factor in the development of CVDs [50]. Physiological risk factors like blood pressure, cholesterol, and glucose were higher among older adults compared with younger adults, whereas ideal behavioural factors like physical activity and diet were less prevalent among them [50]. Thus it is interesting to consider that maintaining high CVH in elderly adults may result in markedly lower rate of RLS.
Our study also has limitations. First, we enrolled our participants by using stratified cluster random sampling method, and this might result in enrolment and selection bias and limit the generalizability of our findings. Second, the recall bias can not be ruled out in our results because the self-reported data was collected based on memory. Third, our findings could not be directly generalized to other populations. Finally, the nature of cross-sectional study does not enable us to draw any causal conclusions between CVH metric and risk of RLS.