Decreased phase shift between neck suction induced blood pressure and cerebral blood flow velocity oscillations indicates impaired cerebral autoregulation in type II diabetics

Introduction: Compromised cerebral autoregulation (CA) contributes to cerebrovascular complications in diabetic patients. In patients with diabetic cardiac autonomic neuropathy (CAN), standard methods of CA evaluation often fail to objectify the quality of CA. 0.1Hz sinusoidal neck suction (NS) is a refined technique to quantify CA using transfer function analysis and calculation of the gain and phase shift between NS induced blood pressure (BP) and cerebral blood flow velocity (CBFV) oscillations. Objective: To determine whether NS is suited to objectify CA in diabetic patients with CAN. Methods: In 9 type II diabetics with pre-existing CAN (57±7 years) and 11 age-matched controls (59±11 years), we continuously recorded RR-intervals (RRI), mean BP and mean middle cerebral artery CBFV at rest and during 0.1Hz sinusoidal NS (-30mmHg). We determined BP and CBFV responses to NS from spectral powers of BP and CBFV oscillations at 0.1Hz at rest and during NS and calculated the 0.1Hz transfer function gain and phase shift between BP and CBFV oscillations as parameters of CA for coherence >0.5. We compared parameters before and during NS using the Wilcoxon-test, and patient and control data by the U-test. Significance was set at p<0.05. Results: In both groups, NS did not significantly influence RRI, BP and CBFV. NS significantly increased 0.1Hz powers of BP (0.6±0.4 vs. 1.3±0.9mmHg²) and CBFV (0.6±0.5 vs. 1.5±1.0cm²/s²/mmHg²) in the controls, but not in the patients. The gain between 0.1Hz BP and CBFV oscillations remained stable in both groups. Phase shift remained stable in controls but decreased in patients (-1.1±0.7 rad vs. -1.7±0.5 rad). Conclusion: In our patients, the lack of 0.1Hz power increase of BP and CBFV oscillations was due to CAN. Thus, gain analysis failed to show altered CA. In contrast, phase shift analysis revealed a decrease in patients and identified altered CA in the diabetics. To summarize, 0.1Hz NS and analysis of phase shift between 0.1Hz BP and CBFV oscillations is suited to identify subtle impairment of CA in diabetics in the presence of coexisting CAN, even if gain analysis fails to identify abnormal CA.