Diabetic cardiac autonomic neuropathy (DCAN) is a serious complication of diabetes and is independently associated with cardiovascular events, incidence and mortality. As the global prevalence of diabetes continues to rise, the research focus on DCAN has also increased significantly in recent years [1]. Epidemiological studies have shown that the prevalence of DCAN varies significantly between patients with type 1 diabetes (T1DM) and those with type 2 diabetes (T2DM), influenced mainly by various clinical features and demographic factors. In T1DM patients, chronic hyperglycemia is the principal risk factor driving the development of DCAN. In contrast, the pathogenesis of DCAN in T2DM patients is complicated by the cumulative effects of concomitant metabolic derangements—namely obesity, hypertension, and dyslipidemia. It is worth noting that insulin resistance, which is the underlying cause of T2DM and metabolic syndrome, plays a direct role in the onset of DCAN [1, 2].

DCAN often presents as asymptomatic, or with indistinct symptoms in the early stage, which greatly limits the effectiveness of treatment. Early diagnosis of DCAN is therefore crucial, especially when intervention is performed during the reversible stage of the disease [2]. The decrease in heart rate variability (HRV) is the earliest clinical indicator of subclinical DCAN. Diagnostic methods for DCAN include the cardiovascular autonomic reflex test, analysis of HRV, 24-h monitoring of blood pressure, baroreflex sensitivity test, and cardiac sympathetic nerve imaging. However, these methods are mostly limited to the research environment, with their wider application in clinical practice still facing many challenges [2, 3, 4]. In-depth studies on the epidemiological characteristics, pathogenesis, diagnostic methods, and treatment strategy of DCAN therefore have major clinical significance. Early diagnosis and intervention can improve the prognosis of patients, reduce the incidence of cardiovascular events and the mortality rate, and improve the patients’ quality of life [2, 5]. Intensive glycemic control is the main strategy used to prevent DCAN in patients with T1DM. To delay disease progression and improve the quality of life for patients with T2DM, more comprehensive management strategies are required, including improvement of microcirculation, neurotrophic support, drug therapy, and lifestyle intervention [2]. Although progress has recently been made in understanding the pathogenesis of DCAN, there is still no clear treatment that specifically targets this condition. Future studies should aim to develop novel therapeutic drugs that can alter the natural progression of the disease, and even reverse its course. In addition, further research is needed into the role of social determinants in the etiology of DCAN so that more targeted interventions can be developed [2, 5].

As one of the important complications of diabetes mellitus, the pathophysiological mechanisms underlying DCAN involve multiple complex molecular and cellular processes. Although the exact mechanisms have yet to be fully elucidated, the main mechanisms can be divided into the categories described below, based on results from existing studies.

DCAN mainly affects functions of the heart, blood vessels, gastrointestinal tract, genitourinary system, and pupil that are regulated by the autonomic nervous system. Its clinical manifestations are diverse, with cardiovascular symptoms being particularly prominent, including resting tachycardia, exercise intolerance, as well as symptoms related to orthostatic hypotension, such as dizziness, blurred vision, neck pain, and syncope [1, 27].

The Toronto Neuropathy Experts have emphasized the importance of the Cardiovascular Autonomic Reflex Test (CART) and regard it as the “gold standard” for diagnosing DCAN [5]. CART evaluates the functions of parasympathetic and sympathetic nerves through a series of tests, including deep breathing, the Valsalva maneuver, HRV during supine and standing positions, and the supine-to-standard blood pressure test (see Table 1). These tests provide important information about cardiac autonomic nerve function, thus helping doctors to diagnose DCAN more accurately [27].

Although CART is widely regarded as the “gold standard” for diagnosing DCAN, this method also has some limitations and drawbacks. It includes multiple steps and tests, such as deep breathing, the Valsalva maneuver, supine-to-standing HRV, and supine-to-standard blood pressure test, all of which require operation and interpretation by professionals. This complex and time-consuming process is unlikely to be suitable for rapid screening or for the assessment of large populations. Moreover, the implementation of CART requires professional equipment such as cardiac autonomic function testing systems, which may limit its application in medical institutions with limited resources or insufficient equipment. Therefore, in recent years the clinical community has been searching for simpler, faster, and more accurate methods to optimize the diagnostic process.

HRV analysis is an important non-invasive tool discovered in recent years for evaluating DCAN, with its value confirmed by multi-dimensional studies. Genetic studies suggest that HRV is regulated by specific genetic loci. Nolte et al. [28] identified multiple gene loci associated with HRV through genome-wide association analysis. These loci are also associated with the risk of cardiac disorder, indicating that HRV is not only an indicator of autonomic nerve function, but may also have a genetic basis and value for cardiovascular prognosis [28]. Zhang et al. [29] showed that combining the traditional symptom scoring tool Composite Autonomic Symptom Score 31 (COMPASS 31) with HRV parameters could significantly improve the diagnostic accuracy for DCAN in T2DM patients, reflecting the superiority of integrating multiple indicators. Recent advances in analytical methods have led to the successful application of machine learning (ML) technology for in-depth mining of HRV data. Alkhodari et al. [30] extracted HRV traits from 24-h dynamic ECG data and constructed a diagnostic model by combining with an ML algorithm. This could efficiently screen DCAN in diabetes patients with microvascular complications, thus providing an automated solution for large-scale population screening [30]. However, attention should be paid to the reliability of HRV measurements. Besson et al. [31] found that although strict control of the test environment and positioning was associated with high clinical reliability of short-term HRV measurement, the results could be affected by test conditions. This suggests that standardized detection procedures are crucial for ensuring the accuracy of screening results.

The scope of applications for HRV is continuously expanding. In addition to its use in evaluating autonomic nerve function, HRV is also associated with mental and psychological disorders such as anxiety. Tomasi et al. [32] explored its potential as a biomarker for anxiety disorder. These authors suggested that attention should be paid to the potential impact of psychological factors on HRV results in diabetes patients [32]. Besides HRV, other ECG indicators such as the QT interval have also been used in DCAN assessment. Vasheghani et al. [33] found the QT interval index is significantly correlated with DCAN and can serve as an effective supplementary diagnostic indicator in addition to HRV.

Beyond HRV, additional physiological indices can also play a pivotal role in the diagnosis of DCAN. Alkhodari et al. [30] demonstrated the potential of ML by using demographic, clinical, and laboratory data to screen for T2DM microvascular complications, further demonstrating the AI-led development of diabetes management. Sudoscan is an innovative medical device that assesses sweat gland function in a non-invasive manner, thereby providing a new technical approach for the evaluation of diabetic autonomic neuropathy [34].

In the field of imaging-assisted diagnosis, the Silesia Diabetes Heart Study by Nabrdalik et al. [35] demonstrated that AI-based classification of DCAN from retinal fundus photographs offers a novel imaging biomarker and diagnostic paradigm for this condition. Moreover, Jaiswal et al. [36] found that impaired cardiovascular autonomic nerve function in T1DM patients was significantly associated with the occurrence of severe hypoglycemia events. This suggests that autonomic nerve regulation disorders may weaken the body’s normal compensatory response to hypoglycemia, thereby increasing the risk [36]. Glycemic variability (GV) itself is also considered an important driving factor for the progression of neuropathy. Zhang et al. [37] systematically reviewed GV and noted that it can promote nerve injury by mechanisms such as exacerbation of oxidative stress and inflammatory reactions. Therefore, the analysis of GV parameters derived from continuous glucose monitoring (CGM) is not only helpful for the refinement of blood glucose management, but may also be an additional reference for the diagnosis and risk stratification of DCAN [37]. In conclusion, the current diagnosis of DCAN integrates multi-dimensional information that spans traditional functional assessment to novel AI image recognition, and autonomic nerve-specific detection to blood glucose system evaluation. This is a reflection of the general trend toward multimodality and AI.

DCAN is a complex, multi-system disease. Although significant progress has recently been made in understanding the underlying pathological mechanism and improving diagnosis and treatment, several remaining key challenges and knowledge gaps urgently need to be addressed by future research. First, current diagnosis is still highly dependent on functional investigations such as HRV and the cardiovascular reflex test. The operational complexity of these tests and their dependence on standardized procedures have greatly limited their clinical application. Therefore, future research should focus on developing and validating novel biomarkers that are non-invasive, highly sensitive, and easily standardized. For example, body fluid detection systems based on collagen metabolism markers (e.g., PRO-C6, C3M) [5, 15], inflammatory factor profiles (e.g., TNF-$\alpha$, IL-1$\beta$) or purinergic receptors (e.g., P2X3, P2Y12) [21, 22, 23, 24, 25, 26], combined with AI algorithms for the integrated analysis of multi-modal data (e.g., ECG ambulatory, CGM, retina images) [38] should enable the construction of more objective and reproducible models for early diagnosis. Moreover, these models should assist with DCAN typing, and stratified diagnosis and treatment systems.

At the treatment level, although new hypoglycemic drugs such as SGLT-2 inhibitor and GLP-1RA have shown cardiovascular protective potential, their specific intervention effects in DCAN and the mechanisms involved have yet to be elucidated. Randomized controlled trials with DCAN will be required in the future to determine whether these drugs can delay or even reverse the progression of DCAN by improving blood glucose fluctuation, inhibiting inflammatory reactions, or regulating autonomic nerve remodeling. In addition, precise treatment strategies that target specific mechanisms should also be investigated, such as targeted inhibition of harmful purinergic signaling pathways [26], the use of neurotrophic factors or antioxidants such as vitamin C [45] to improve neuron survival and function, and regulation of autonomic nerve balance through exercise intervention [43]. In summary, interdisciplinary integration and the convergence of innovative technologies should enable research encompassing “mechanism exploration to biomarker identification to individualized treatment”. This will be a key goal for achieving breakthroughs in the DCAN field, and ultimately for improving the long-term prognosis of diabetes patients.

NW was responsible for writing all of themanuscript, systematically collecting and integrating relevant literature, synthesizing key evidence, and providing novel insights for future research. JZ, as the corresponding author, provided comprehensive oversight of the research design and data analysis and played a leading role in drafting and critically reviewing the manuscript. Both authors edited and revised themanuscript to ensure accuracy and completeness of content. Both authors have read and approved the final manuscript. Each author has fully participated in the work and agreed to be accountable for all aspects of the work, ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Not applicable.

Not applicable.

The authors acknowledge all the participants in this study. This study supported by Zhu Xiushan talents program (Grant No. 2023HMJQ15), 2023 Ningbo International Cooperation Program (Grant No. 2023H024).

The authors declare no conflict of interest.

During the preparation of this work the authors used ChatGpt-3.5 in order to check spell and grammar. After using this tool, the authors reviewed and edited the content as needed and takes full responsibility for the content of the publication.