Heart rate variability in the assessment of cardiac autonomic function in type 2 diabetes: from time and frequency domains to nonlinear analysis

Abstract

Cardiac autonomic neuropathy (CAN) is a common complication of type 2 diabetes. It is often underdiagnosed until it presents with cardiovascular complications such as silent myocardial ischaemia, orthostatic hypotension or arrhythmia. Cardiovascular autonomic reflex tests (CARTs) remain the diagnostic reference, but they need time, trained staff and good patient cooperation, which limits their use in busy clinics. Analysis of heart rate variability (HRV) from a five-minute resting recording is quick, non-invasive and inexpensive. This narrative review analyses HRV across three domains: time, frequency and nonlinear. For each domain, we explain the physiological basis of the main indices and their limits. We discuss the position of HRV relative to CARTs, summarise studies of its clinical use and its prognostic value in type 2 diabetes, and propose a standard recording protocol and a minimum reporting checklist. Based on current evidence, we suggest that HRV may be most useful for screening and follow-up, while the diagnosis of CAN still relies on CARTs; this framework has not yet been validated. Time- and frequency-domain indices measure the amplitude of heart rate fluctuations, whereas nonlinear indices, such as sample entropy and the short-term scaling exponent of detrended fluctuation analysis (DFA-α1), describe how these fluctuations are organised. We hypothesise that nonlinear indices may retain some discriminating ability in advanced disease, when amplitude-based indices cluster near the lower limit of their range, but this has not been tested directly. Validated cut-offs are lacking for most indices, and no reference values exist for the Vietnamese population. Therefore, standard recording and complete reporting are essential if HRV is to become a reliable clinical tool.