Association of Diabetic Autonomic Neuropathy With Hemodynamic Instability Following Spinal Anaesthesia in Patients With Type 2 Diabetes Mellitus: A Prospective Observational Study

Background: Diabetic autonomic neuropathy (DAN) blunts the compensatory sympathetic and baroreflex responses that limitthe fall in blood pressure after spinal anaesthesia. We examined whether DAN, diagnosed preoperatively by cardiovascularautonomic reflex tests (CARTs), is associated with hemodynamic instability after subarachnoid block in patients with type 2diabetes mellitus (T2DM). Materials and Methods: In this prospective observational study, 50 adults with T2DM scheduledfor elective infra-umbilical surgery under spinal anaesthesia underwent the COMPASS-31 symptom questionnaire and the fiveEwing CARTs before surgery and were classified as having no, early, definite or severe DAN. Blood pressure and heart ratewere recorded at baseline and at 1, 3, 5, 10, 15, 20, 30, 45 and 60 min after intrathecal hyperbaric bupivacaine. The primaryoutcome was post-spinal hypotension (fall in systolic blood pressure [SBP] >20% from baseline); secondary outcomes weresevere hypotension (>30% fall), bradycardia (heart rate <50 beats/min), vasopressor requirement and the nadir of SBP andmean arterial pressure (MAP). Results: DAN was present in 32 patients (64%; early 10, definite 14, severe 8). Hypotensionoccurred in 23 of 32 patients with DAN versus 4 of 18 without (71.9% vs 22.2%; relative risk 3.23, 95% CI 1.33–7.88; P =0.002), and severe hypotension in 40.6% versus 5.6% (P = 0.020). Patients with DAN had a larger maximum fall in SBP(median 28.4% vs 15.6%; P < 0.001), a lower nadir MAP (median 70.0 vs 85.7 mmHg; P = 0.002), more hypotensive episodesand greater ephedrine and rescue-fluid requirements (all P < 0.01). Every outcome worsened stepwise with DAN severity (allP ≤ 0.002 across categories); all eight patients with severe DAN developed severe hypotension. On multivariable logisticregression adjusting for age, diabetes duration, hypertension and baseline SBP, DAN remained independently associated withhypotension (adjusted odds ratio 7.71, 95% CI 1.08–55.2; P = 0.042). The number of abnormal CARTs, COMPASS-31 scoreand orthostatic SBP fall each discriminated hypotension moderately well (area under the curve 0.79–0.83). Conclusion:Preoperatively detectable DAN is strongly and independently associated with the incidence, severity and treatment burden ofhypotension after spinal anaesthesia in T2DM, with a clear severity gradient. Simple bedside autonomic testing can identifydiabetic patients who warrant intensified hemodynamic vigilance and prophylaxis.