Avaliação da eficácia dos testes para diagnóstico de disautomia cardiovascular

Abstract

Autonomic nervous system dysfunction encompasses a wide spectrum of clinical syndromes that are often underdiagnosed and carry significant implications for morbidity and mortality. Among them, cardiovascular autonomic neuropathy (CAN) stands out, particularly in individuals with diabetes mellitus (DM), and is considered an independent marker of cardiovascular risk. This study aimed to evaluate the performance of the seven-test protocol—comprising four classical cardiovascular autonomic reflex tests (CARTs) and three spectral heart rate variability (HRV) tests—in diagnosing CAN in a population with clinical suspicion of autonomic dysfunction, not exclusively diabetic. A total of 177 patients referred for autonomic assessment between 2017 and 2024 were analyzed. In addition to using statistical tests such as Student’s t-test and Mann-Whitney, we assessed diagnostic accuracy through ROC curves, calculating the area under the curve (AUC), sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV). A p-value < 0.05 was considered statistically significant. The mean age was 53.3 ± 21.5 years, with a predominance of females. The most prevalent autonomic syndrome was neurogenic orthostatic hypotension (58.8%). The prevalence of established CAN (≥3 positive tests) was only 18%. The spectral HRV components outperformed the CARTs. The very low-frequency (VLF) component showed an AUC of 0.854, with 81.2% sensitivity and 89.5% specificity; the high-frequency (HF) component presented an AUC of 0.842, sensitivity of 78.1%, and specificity of 90.2%, with an overall diagnostic accuracy of 88%. In contrast, respiratory-based reflex tests had low sensitivity but maximum specificity (100%). The classic CARTs score also showed low sensitivity (25%), limiting its utility as a standalone screening tool. Age was the only demographic factor significantly associated with CAN. Based on these findings, we propose a two-stage assessment model. The first stage—screening—includes the three spectral HRV components (VLF, LF, and HF), given their high sensitivity. The presence of at least one abnormal component indicates the need for further diagnostic investigation. The second stage—confirmation—is based on the reflex tests with the highest specificity: Valsalva ratio, 30:15 coefficient, and the 3-minute orthostatic blood pressure response. The diagnosis of CAN is supported when at least one confirmatory test is abnormal. This strategy aims to enhance the efficiency and reduce the cost of autonomic evaluation compared to the full seven-test protocol. This study contributes to the ongoing discussion on standardizing CAN diagnosis and proposes a more sensitive, objective, and clinically applicable approach.