To the best of our knowledge, this is the first study to compare CA in participants from different geographical regions with marked environmental and socio-economic differences. These pooled analyses suggest no geographical differences in key, commonly measured cerebral haemodynamic parameters, including CBFv, CrCP, RAP and ARI. Overall, and in keeping with previous studies, impairment of cerebral haemodynamic parameters was reported in stroke compared to control participants, particularly in the affected hemisphere. Importantly, no geographical regional differences were found.
Recognizing similarities and differences in the cerebral haemodynamics in healthy subjects and stroke has public health and clinical implications for several cardiovascular outcomes. Haemodynamic abnormalities may be useful early clinical markers in the development of age-related vascular neurodegenerative disorders. Describing cerebral haemodynamic changes and their regulation in different geographical populations may help understand differences across age and ethnicities, as well as the potential generalizability of clinical studies results.
Impairment of CBFv and CA has a direct and major impact on secondary brain injury and clinical outcomes [3, 4], as well as on planning effective therapeutic strategies that consider BP management and early mobilization protocols. In line with previous studies, our results showed a deterioration of important haemodynamic parameters in the acute phase of ischaemic stroke, particularly reduced phase and ARI values compared to controls [3, 20,21,22,23]. Moreover, no significant alteration in gain between groups was found [6, 20]. Though TFA became a popular approach for assessment of dynamic CA, very few acute stroke studies have reported the use of coherence estimates [3, 22]. These previous studies failed to find raised coherence values in acute stroke, but an analogous parameter (the squared correlation coefficient) was found to be increased bilaterally after the first 48 h of stroke onset, suggesting worsening of CA .
The importance and novelty of our results, however, stem from the similarities of the haemodynamic responses from research centres in both UK and Brazil, irrespective of the participant group (stroke or healthy controls). Despite the potential differences in dietary habits and lifestyles, and particularly in SP participants, higher prevalence of uncontrolled hypertension and widespread use of over-the-counter medications, cerebral haemodynamic parameters were not significantly different in our study. The authors believe that the outcome of this paper will strengthen the argument in favour of multicentre and multinational collaborative studies of the impact of cerebral haemodynamics in stroke and other conditions, such as sepsis and traumatic brain injury.
Environment effects on cerebral haemodynamics
Since differences in CBF and its control mechanisms between European and South American populations have not been previously reported, it is difficult to assess consistency with other studies in the literature. A previous TCD study of healthy young participants (n = 20) in Germany and Hong Kong described no CBFv differences in the posterior cerebral arteries at rest and during cerebral activation . Nevertheless, slower initial haemodynamic responses to visual activation paradigms were described in the Asian group that may be related to deficits in the nitric oxide system. Previous studies have also compared cerebral haemodynamic responses between South Asian and Caucasian participants, but the results were inconsistent. In a United Kingdom-based study, resting MCA CBFv and cerebrovascular resistance parameters (Pourcelot’s resistive index and Gosling’s pulsatility index) were significantly higher, and CA impaired in South Asian participants . By contrast, no difference in the same parameters of CBFv and cerebrovascular resistance was found in another recent study of Canadian-based South Asians and Caucasians .
Disparities in cardiovascular and cerebrovascular health and mortality among populations have been well documented, but poorly understood [27, 28]. Studies have shown that the prevalence and mortality from hypertensive heart disease, stroke, and renal disease are higher among individuals of African compared to Caucasian descent . Moreover, intracranial atherosclerosis is highly prevalent among patients with Asian, Hispanic, and African ancestry .
The present study did not demonstrate significant geographical differences in cerebral haemodynamic parameters between UK and Brazilian populations, with exception of gain at high frequency band. Immink and colleagues (2005) found an increase in gain only at higher frequencies in MCA-only stroke group. This may be the reason for the difference between regions, as SP comprised MCA infarcts only, whereas seven lacunar strokes were included in the LEI group (Table 2). Though recommended by the Cerebral Autoregulation Network (CARNet) , there is limited information in the literature on CA status in the complete frequency dependence of coherence, gain and phase in the range 0.02–0.40 Hz. While the TFA parameters changes in VLF and LF ranges are familiar, very little is known about the behaviour of the BP/CBFv system at frequencies higher than 0.25 Hz. In the TFA model, phase and gain are considered two different aspects of a high-pass filter that acts primarily in the VLF and LF ranges. At high frequencies, CA is considered less relevant and CBFv changes are associated with heart stroke volume beat-to-beat variability. More clinical studies are necessary to investigate the clinical importance of investigating CA in the higher frequency band, taking advantage of the work that has already been done. A further finding of the present study related to the responses of cerebrovascular resistance mechanisms, derived from a two-parameter model: RAP-CrCP. Levene’s test revealed a marginal difference in CrCP between geographical regions, with a tendency to higher values in LEI participants. Further research is needed to assess the clinical value of this finding.
The experimental design of previous cerebral haemodynamic studies has been inconsistent, making it difficult to compare results directly [32, 33]. Furthermore, there have been no previous CA studies in either older or stroke populations in Brazil. In contrast, in the UK, some studies have previously described the haemodynamic responses to spontaneous BP fluctuations, carbon dioxide modulations and brain activation paradigms in both populations [12, 13, 34, 35]. More recently, the Department of Cardiovascular Sciences at the University of Leicester (Leicester, UK) has constructed a large database incorporating recordings from a series of separate studies performed in the same laboratory, using similar protocols, operator training and equipment . They have presented normative values for cerebral haemodynamic parameters in a large healthy population indicating parameters that may help distinguish between normal and abnormal CA.
The present study was only possible due to the development of standardized data collection and analysis that provided a robust approach for the systematic evaluation of CBFv and its main regulatory mechanisms . Similar to the Leicester normative study , all data collected in this study were acquired with similar study protocols and laboratory set-up, albeit with minor differences in the equipment used, and most importantly, without observer variability since all recordings were performed by only one researcher (AS). This avoids inter-observer variability in the study protocol (particularly concerning the TCD data), and it consequently increases the reproducibility of study reports.
This study has limitations, including the use of non-invasive measurements of CBF and BP. Another limitation to consider is stroke patients received medication according to local guidelines for secondary prevention at the time of assessment. Though patients were on vasoactive therapy at time of assessment, no significant difference between populations was found. Although TFA and the ARI index can be regarded as the most widely used approach for assessment of dynamic CA, it is important to note that neither can be regarded as a ‘gold standard’. Future studies are needed to replicate our findings using alternative approaches such as time-domain analysis or the Mx index . Finally, the sample size is relatively small, and is unlikely to be representative enough to ensure robust conclusions. Therefore, the authors strongly advocate a large multicentre validation study with larger sample size to explore further the possibility of regional geographical influences on cerebral haemodynamics, and possible mechanisms to support any differences.