Objective: To compare the overnight variation trends in the duration of obstructive apnea events, and to explore the adaptive capacity to the pathophysiological consequences of periodic sleep disordered-breathing and its underlying mechanism in patients with obstructive sleep apnea hypopnea syndrome (OSAHS). Methods: A retrospective analysis were performed of the polysomnographic (PSG) results of 89 snoring patients including 10 non-OSAHS, 15 mild, 29 moderate and 35 severe OSAHS. The total record time was divided into four equal phases, and the variation trends of the mean apnea duration (MAD) and the longest apnea duration (LAD) were compared with the progression of sleep phases in different groups. Correlation analysis was conducted with demographic indicators, pulse oxygen saturation (SpO2) and sleep related indicators. In addition, the number of apneas-time variability curve was plotted for fitting analysis. Results: In patients with severe OSAHS, both MAD [26.1(20.9, 31.4) s] and LAD [56.5(46.5, 82.0) s] were significantly higher than those of non-OSAHS, mild and moderate OSAHS (P<0.001). In addition, the MAD in the third and fourth quartiles were significantly higher than that in the first quartile [(28.4±9.0) s, (27.3±9.8) s, (22.3±9.9) s, respectively, P=0.046], and the LAD in the third quartile was significantly higher than that in the first quartile [56.5(38.5, 71.0) s, 41.0(28.0, 53.0) s, respectively, P=0.018]. In all subjects, the MAD and LAD in the third and fourth quartiles were significantly higher than those in the first quartile [MAD: 20.3(10.3, 29.2) s, 18.5(11.3, 24.2) s, 12.9(0.0, 21.8) s, respectively, P<0.001; LAD: 28.0(10.3, 50.5) s, 28.0(12.0, 44.5) s, 14.5(0.0, 32.3) s, respectively, P<0.001]. There was no statistical difference in the lowest SpO2 (LSpO2), the mean SpO2 (MSpO2), and the percent of sleep time oxygen saturation below 90% (T90%) of all subjects in different sleep phases (P>0.05). The LAD was positively correlated with obstructive apnea index (OAI, OR=1.660, P=0.025), but no correlation was observed with other indicators (P>0.05). The MAD increased 0.22 s per episode at the onset of sleep (1-31 apnea events), then dropped to 0.04 s of increase per episode, with a dynamics change of 5.5-fold slower. Conclusions: The MAD and LAD show a gradual prolongation trend with the progression of sleep phases, and the prolongation trend is the most obvious in patients with severe OSAHS, while the dynamic change trend of SpO2 is not obvious. There may be multiple adaptation mechanisms for recurrent hypoxic episodes, and the adaptation occurr in stages, with a rapid increase in MAD at the onset of sleep, follow by a markedly slower increase. Patients with severe OSAHS express the most complete pattern, suggesting the most severe pathophysiological outcomes.