With interest we read the article by Adebiyi et al. about 80 patients with epilepsy who were investigated for cardiac autonomic neuropathy (CAN) by means of 5 tests, the heart rate response to deep breathing, to the Valsalva manoeuvre, and standing, and blood pressure variation during isometric handgrip and after 3min of standing.[1] It was found that 52.5% had CAN, of which 69% had pure parasympathetic dysfunction and only 3% pure sympathetic dysfunction.[1] Patients with >4 seizures per month and those receiving antiepileptic drugs (AEDs) for <10y had higher odds to develop CAN.[1] We have the following comments and concerns.

The main shortcoming of the study is that neuropathy was not confirmed by other means than the five applied tests. Most patients with autonomic neuropathy also develop neuropathy of motor and sensory fibers, which can be documented on nerve conduction studies (NCSs). Thus, it is crucial that patients with suspected CAN undergo NCSs. CAN can also be documented by ¹²³I-metaiodobenzylguanidine (MIBG) single photon emission computed tomography (SPECT).[2] Thus, we should know how many of the included patients had CAN also on NCSs and on MIBG SPECT. 

A further shortcoming is that the type of epilepsy was not specified. Thus, we should know how many of the 80 patients had structural, genetic, infectious, metabolic, immunologic, or kryptogenic epilepsy. Knowing how many had genetic epilepsy is crucial since some of the genetic epilepsies are associated with neuropathy.[3] Genetic epilepsies may also manifest in the heart with ventricular arrhythmias or conduction defects.[4] Additionally, mutations in certain genes, such as KIF1A, may manifest with epilepsy and autonomic neuropathy.[5] 

Autonomic neuropathy usually not only manifests in the autonomic fibers supplying the heart but also in parasympathetic or sympathetic fibers supplying the pupillary muscles, the lacrimal glands, the salivary glands, bronchial smooth muscles, the intestines and pancreas, the liver, the suprarenal gland, the urinary bladder, and sexual organs. Thus, we should know if the 42 patients with CAN, had pupillary dysfunction, hypo/hypersecretion of lacrimal fluid or saliva, bronchospasm, dyspepsia, gastrointestinal dysfunction, urinary voiding dysfunction, or sexual dysfunction. 

Since CAN may be associated with systolic or diastolic dysfunction and heart failure, we should be informed about the results of echocardiography and the pro-brain natriuretic peptide (BNP) values in the 42 patients. We should also know if infectious, immunologic and metabolic causes of epilepsy and autonomic neuropathy were excluded. 

Interesting would be follow-up data to see if any of the 42 patients experienced sudden unexpected death in epilepsy (SUDEP). Since CAN may contribute to the occurrence of SUDEP, it is conceivable that some of those with CAN experienced near-SUDEP or SUDEP during the disease course.

Overall, the appealing study has some shortcomings which should be addressed before drawing conclusion as those presented. The high frequency of CAN in selected epilepsy patients may depend on the methods applied to measure CAN and on the inclusion criteria. Epilepsy should be classified according to etiology to see if there is a correlation between the cause of epilepsy and CAN.