Research On The Mechanism Of Occurrence Of Chronic Insomnia Based On Sleep Reactivity And Its Clinical Application

BackgroundStudies have shown that about 30% to 40% of adults have varying degrees of insomnia,and the prevalence of chronic insomnia（CI）is about 10%.CI can have a huge impact on cardiovascular,cerebrovascular and metabolic diseases,and mental health.Although medications and cognitive-behavioral therapy have some efficacy in CI,many patients still have problems with relapse and drug resistance.The reasons for this is that we still do not fully understand the complex neurobiological mechanisms of the occurrence and development of CI,so that the target of existing therapeutic methods is unclear or only symptomatic.Only a few prospective studies have shown that sleep reactivity（SR）levels are associated with the occurrence and severity of CI,that is,SR can predict the risk of future insomnia.However,neurobiological and applied research on SR is still lacking.Therefore,it is of great significance to determine the underlying mechanism of the interconnectedness of SR and the development of insomnia for the development of clinical prevention,diagnosis and treatment strategies for CI.ObjectiveIn order to verify that the level of SR is related to the occurrence of CI,and to further clarify the relationship between it and the neural biomarkers of CI,this project takes healthy people and patients with CI as the object,and intends to explore:（1）Whether the SR level of well-sleep individuals is related to different cortisol responses and changes in cortical EEG;（2）Whether the SR level of insomniac individuals is related to more severe sleep macrostructure and cortical EEG changes;（3）Whether the SR level of insomnia individuals is related to the changes of serological markers of more severe insomnia and its relationship with subjective and objective sleep quality.Methods（1）Fifty-six volunteers with good sleep were randomly divided into stress（n = 33）or without treatment（n =23,control group）conditions.According to the the Ford Insomnia Response to Stress Test Chinese Version（FIRST-C）score,the stressed subjects were respectively divided into high sleep reactivity（H-SR,n = 17）and low sleep reactivity（L-SR,n = 16）groups.Insomnia Severity Index scale（ISI）,17-item Hamilton Depression scale（HAMD-17）and 14-item Hamilton Anxiety scale（HAMA-14）were used to evaluate the sleep quality and affection of the subjects.A total of two nights were arranged in the experiment.For the first night,the subjects were allowed to adapt to the laboratory environment to eliminate the first-night effect.For the second night,all subjects completed polysomnography（PSG）,while the stressed individuals received Trier Social Stress test（TSST）within 2 hours before PSG monitoring.The saliva samples were collected at 20 min before,1 min,15 min and 30 min after the start of the stress experiment.PSG parameters mainly include: total sleep time（TST）,sleep efficiency（SE）,sleep latency（SL）,time in rapid eye movement sleep（REM）,awakening time after falling asleep（WASO）,time in non-REM sleep periods 1,2 and 3（N1,N2,N3）and their percentages of total sleep time（N1%,N2%,N3%）,REM percentage of total sleep time（REM%）.Fast Fourier transform（FFT）was used to analyze the spectral density of EEG in different sleep stages.（2）A total of 81 patients with CI were included in the sleep disorder clinic.According to the intensity of SR,they were divided into H-SR group（n = 50）and L-SR group（n = 31）.At the same time,40 healthy subjects were taken as the control group（17 males / 23 females,44.66 ± 6.53 years old）.ISI,Pittsburgh Sleep quality Index（PSQI）,HAMD-17,HAMA-14 and pre-sleep arousal scale（PSAS）were used to evaluate the sleep quality and emotion.At the experiment,the first night was assigned the subjects to adapt to the laboratory environment,and the second night was assigned to complete PSG.FFT was used to analyze the spectral density of EEG in different sleep stages.Fasting venous blood was drawn from 7 am to 8:00 on the day after PSG monitoring,and CI serum markers,such as astrocyte damage markers（GFAP,BDNF）and HPA axis hormones（CRH,cortisol）,were detected with ELISA.Results1.Cortisol reactivity,sleep structures and cortical electroencephalogram after subjected to stress in good-sleep individuals with different SR levels:（1）PSG sleep parameters: Compared to the control group,the H-SR group had decreased TST,SE,N2,REM and REM%,and increased SL and WSAO,while the LSR group only had decreased SE.Compared to the L-SR group,the H-SR group had significant severer changes in TST,SL,SE,N2,REM and REM%.（2）EEG spectrum: Compared to the control group,the H-SR group had significant decrement of relative power of δ but increment of relative power of β and γ in the N1 stage,and significant increment of relative σ and γ powers in N2 stage,and significant increment of relative γ power in N3 stage and REM sleep（Ps < 0.05）.however,the LSR group had no significant changes of frequency-related powers.Compared to the LSR group,relative powers significantly decreased in δ but increased in β and γ in N1 stage in the H-SR group（Ps < 0.05）.（3）Cortisol reactivity: Compared with the L-SR group,the salivary cortisol level in the H-SR group significantly increased after stress（P < 0.001）,and the response to cortisol had significantly increased（Cohen’s d = 0.872）.2.Sleep structures and cortical electroencephalogram changes in CI individuals with different SR levels:（1）PSG sleep parameters: Compared with the control group,the H-SR group and L-SR group had decreased TST,SE,N2,N3,REM,N2%,N3% and REM%,and increased WASO,SL,N1 and N1%（Ps < 0.05）.Compared with the L-SR group,the HSR group had decreased TST,SE,N2,N3,REM and N2% and increased WASO（Ps <0.05）.（2）EEG spectrum analysis: Compared with the control group,the H-SR group had significant increment of the relative power of β and γ in N1 stage,significant decrement of the δ relative power and significant increment of the β relative power in N2 and N3 stage.In addition,the H-SR group had significant decrement of θ relative power and increment relative power of α,β and γ in REM stage（Ps < 0 05）.Compared with the control group,the L-SR group had significant increment of the relative power of α in REM stage（Ps < 0.05）.Compared with the L-SR group,β relative power in N1 and N2 stage had significantly increased in the H-SR group,θ relative power in REM stage increased had significantly decreased,and β and γ relative power had significantly increased in the H-SR group（Ps < 0.05）.3.The changes of serum markers in CI patients with different SR levels and the correlations with the subjective and objective sleep quality:（1）Serum markers: Compared with the control group,the serum levels of GFAP,CRH and COR had significantly increased in the H-SR group and the L-SR group（Ps <0.001）,while the serum BDNF level had significantly decreased（Ps < 0.001）.Compared with the L-SR group,the serum levels of CRH and COR had significantly increased in the H-SR group（Ps < 0.001）.（2）The correlations between SR levels and the serum markers and subjective sleep quality: Spearman correlation analysis showed that FIRST-C scores were negatively correlated with the serum BDNF level,positively correlated with the serum CRH level and COR level and positively correlated with ISI and PSQI scores.Serum GFAP,CRH,COR levels were positively correlated with ISI and PSQI scores,serum BDNF level were negatively correlated with ISI and PSQI scores.After controlling sex,age and HAMD scores,partial correlation analysis showed that FIRST-C scores were positively correlated with serum GFAP,CRH,COR levels and negatively correlated with BDNF levels,serum GFAP and CRH levels were positively correlated with ISI and PSQI scores,serum BDNF levels were negatively correlated with ISI and PSQI scores,and serum COR levels were positively correlated with PSQI scores.（3）The correlations between SR levels and the serum marker levels and objective sleep parameters: After controlling sex,age and HAMD scores,partial correlation analysis showed that the FIRST-C scores were negatively correlated with TST,SE,N3,N3,REM and REM%,and positively correlated with WASO.The levels of serum GFAP were negatively correlated with TST,SE,N2,N3 and N3%.BDNF levels were positively correlated with TST,SE,N3 and N3%,and negatively correlated with WASO,N1 and N1%.CRH levels were negatively correlated with TST,SE,N2,N3 and N3%,and positively correlated with SL,WASO,N1 and N1%.The levels of COR were negatively correlated with TST,SE,N2,N3 and N3%,and positively correlated with WASO.（4）Principal factor analysis: FIRST-C,GFAP,BDNF,CRH,ISI,PSQI,TST,WASO,SE,N1%,N3,REM and REM% had high load coefficient in principal factor 1,while TST,N2,N3,N3 and REM% had high load coefficient in principal factor 2.The load of WASO and CRH was slightly higher in the main factor 3.SummaryThe sleep structures of the H-SR individuals with good sleepers after acute stress were broken,the excitability of cerebral cortex had increased at the night,and the cortisol reactivity increased,which constituted its characteristic vulnerability.H-SR in CI individuals predicted more severe sleep fragmentation,nocturnal cortical excitement,excessive cognitive and emotional arousal,and more significant changes in serum marker levels.The levels of SR and serum markers were associated with subjective and objective sleep quality in patients with CI.