Introduction Patients with sleep apnea hypopnea syndrome (SAHS) show an increased risk of suffering from cardiovascular diseases. Although this is a multifactorial relationship, sympathetic activation seems to carry out an essential role. The analysis of heart rate variability (HRV) from ECG, i.e. the RR interval (time between consecutive R peaks) time series, has been widely used as a measure of autonomous control. HRV recordings have been mainly analyzed in the frequency domain. However, as other physiological signals, heart rate is not strictly periodic. Therefore, the use of nonlinear techniques could provide additional and essential information about HRV dynamics in SAHS patients. This study is aimed at quantifying different nonlinear metrics from HRV recordings of SAHS patients in order to characterize their dependence with the severity of the disease. Materials and methods A total of 240 subjects derived to the sleep-related breathing disorders unit were involved in the study. Standard in-hospital polysomnography (PSG) were carried out in order to diagnose SAHS. Subjects with an apnea–hypopnea index (AHI), 10 events per hour (e/h) from PSG were diagnosed as suffering from SAHS. According to PSG, 160 out of 240 patients suffered from SAHS. ECG was recorded at a sampling rate of 200 Hz. A QRS detection algorithm based on the Hilbert transform was applied to obtain the HRV signal. Three nonlinear methods were subsequently applied: Sample Entropy (SampEn), which quantifies irregularity; Lempel–Ziv complexity (LZC), which measures complexity; and Central Tendency Measure (CTM), which is a variability measure. Patients were divided into 4 groups according to their AHI: Non-SAHS subjects (0 30). Results Population under study had mean age of 52.2 years, mean body mass index (BMI) of 29.6 kg/m2, and 77.5% of patients were male. SAHS positive patients showed a mean AHI of 33.2 e/h. Regarding nonlinear analysis, HRV recordings from SAHS positive patients showed lower irregularity (0.370 vs. 0.432), lower complexity (0.333 vs. 0.363) and lower variability (0.946 vs. 0.880) than non-SAHS subjects. A significant correlation was found between AHI and nonlinear measures: −0.296 (SampEn), −0.268 (LZC), and 0.256 (CTM). These correlations remain significant when adjusted for age. In stratified analyses by sex, in female patients only lZC showed correlation between AIH and HRV. In male patients CTM, SampEn and LZC show a significant correlation between AIH and HRV. Conclusion Fluctuations of HRV decreases as SAHS severity increases(higher AHI index). Therefore, HRV recordings from severe SAHS patients (AHI > 30) show significantly lower nonlinear irregularity, complexity and variability. Acknowledgements Acknowledgements to the Sleep Group of Hospital Universitario Rio Hortega and Biomedical Engineering Group (GIB).