| BackgroundInsomnia is one of the common complaints in the patients from neurological settings. Because of the adverse consequences of the individual and society, insomnia has become a major clinical problem. Cognitive dysfunction is one of the adverse consequences of insomnia due to the changed neuroendocrine-immunology, physiology, and structure and function of brain, and also the main reason why patients come to the clinic for treatment. However, besides the small sample size, the differences of sample source, diversity of tested tools and lack of objective sleep assessment, just the defect that the studies were not enrolled insomnia disorder (ID) patients can make an inconclusive results that insomnia damages memory. Until now, the pathogenesis of insomnia and the mechanism of changed cognitive function caused by insomnia are still unknown.ObjectiveTo explore the changes of cognitive function (memory), the state of objective sleep and blood parameters of neuroendocrine-cytokine in the patients with ID, and to investigate the association among them.MethodsThe patients with ID and insomnia comorbid-depression (ICD) were studied with health controls. The Pittsburgh Sleep Quality Index (PSQI), Hamilton Depression Rating Scale-17 Items (HAMD-17) and Montreal Cognitive Assessment (MoCA), were used to evaluate sleep quality, depression state and general cognitive function in all subjects, respectively. The partial patients were monitored for sleep with polysomnography (PSG). Each subject was then asked to rate his or her memory on a scale from 0 (bad) to 4 (good), and these scores were recorded as self-assessment memory (SAM) scores. A modified Nine-Boxes Maze Test protocol was used to examine the spatial working memory (SWM), object working memory (OWM), spatial reference memory (SRM), object reference memory (ORM) and object recognition memory (ORcM). We sum up the different memory errors as the performance of reference memory (RM, SRM+ORM), working memory (WM, SWM+OWM), object memory (OM, OWM+ORM), spatial memory (SM, SWM+OWM) and total error number (TEN), respectively. The serum concentrations of corticotrophin-releasing hormone (CRH), thyrotrophin-releasing hormone (TRH), gonadotropin releasing hormone (GnRH), adenocorticotropic hormone (ACTH), thyroid stimulating hormone (TSH), cortisol, total triiodothyronine (TT3), total thyroxine (TT4), interleukin (IL)-1β and tumor necrosis factor (TNF)-a were detected.Results1. Cognitive function (memory):Compared to the normal control group (n=30), the ID patients (n=53) had poorer performance of TEN, SWM, ORcM, WM and SM; the ICD patients (n=50) had poorer performance of all tested memory measures; between the two insomnia groups, the performance of SWM, WM, SM and TEN was better in the ID patients than that in the ICD patients. According to the clinical manifestations, the ID patients were classified into the different subgroups consisted of the patients with difficulty initiating sleep (n=11), early morning awakening (n=22), difficulty maintaining sleep and mixed sleep difficulty (n=51). There were significant differences of SWM, SM and WM among the four subgroups. The patients in difficulty maintaining sleep group had the best performance among them, and the patients in early morning awakening and mixed sleep difficulty had the worst performance. The results of the partial correlation analysis (controlling for sex, age, educated level) showed that:the score of SAM was negatively associated with the scores of HAMD-17 and PSQI, and the errors of SRM, SWM, ORcM, TEN, RM, WM and SM were positively associated with the HAMD-17 and PSQI scores. The score of MoCA was negatively associated with HAMD-17 scores.2. The changes of sleep parameters and cognitive function:1) The sleep evaluation by PSG:Compared to the normal reference value, insomnia patients had decreased total sleep time and sleep efficiency, prolonged sleep latency and WASO, declined REM% and N3%, elevated N1% and N2%, increased numbers of arousals and shift times of waking. There was no significant difference of common sleep parameters between the ID patients (n=96) and the ICD patients (n=48).2) PSG results in the ID patients with different subtypes:Among the four subtypes (difficulty initiating sleep, early morning awakening, difficulty maintaining sleep and mixed sleep difficulty), there were significant differences of sleep efficiency and wake after sleep onset (WASO). Compared to the difficulty maintaining sleep (n=22) the patients in early morning awakening (n=22) and mixed sleep difficulty (n=51) had lower sleep efficiency, longer WASO, shorter REM and more N1%; in addition, the patients in early morning awakening group had fewer REM density, the patients in mixed sleep difficulty had longer awake time, shorter REM, and the patients in difficulty initiating sleep group (n=11) had more N1%.3) Correlation analysis between cognitive function and sleep parameters:The results of ordinal regression analysis showed that the higher PSQI score associated with the more error numbers of TEN, WM, OM, SM and SWM. For all the insomnia patients, the results of the partial correlation analysis (controlling for sex, age, educated level and HAMD-17) showed that the different memories as measured by error number were correlated with the different sleep parameters respectively:ORM positively with the N2 latency, and negatively with the N2, N3, N2% and N3%; SRM positively with the latency of N1; OWM positively with waking times; SWM positively with the N2 latency; ORcM and SM positively with the N3 latency; TEN positively with the latencies of N2 and N3, N3, and negatively with N3%; RM negatively with N3 and N3%; OM negatively with N2, N3, N2% and N3%. For the ID patients, the results of partial correlation analysis (controlling for sex, age, educated level and HAMD-17) showed that MoCA score was negatively correlated with the N2 latency and the time in bed; the error number of ORM was negatively correlated with N2 and N2%; the errors of SRM was negatively correlated with the REM density; the TEN and the errors of SWM, WM and SM were negatively correlated with the REM and REM%, and were positively correlated with the N2 latency. The results of the canonical correlation analysis showed that the canonical variable V1 reflected SWM, W1 reflected REM, REM%, N2 latency, N2, N2% and N3%. And the error number of SWM was negatively correlated with REM, REM%, N2% and N3%, and was positively with the N2 latency and N2.3. The changes of cognitive function and the serum levels of hormones and cytokines:1) Hormones and cytokines:Except for the ACTH and TSH in the ID patients (n=30), the serum levels of all the hormones and cytokines in the insomnia groups were higher than those in normal group (n=30). Between the two insomnia groups, the ID patients had higher levels of CRH, cortisol, TT3 and TT4, and lower levels of ACTH, TRH, GnRH, IL-1β and TNF-α than the ICD patients (n=30).2) Correlations between insomnia severity and serum hormones and cytokines:The results of the partial correlation analysis (controlling for sex, age and educated level) showed that the PSQI score was positively correlated with the HAMD-17 score, the serum levels of CRH, TRH, TRH, TSH, cortisol, TT3, TT4, IL-1β and TNF-α. After controlling for HAMD-17 score simultaneously, the items those correlated with the PSQI changed, e.g. the PSQI score was positively correlated with the serum levels of CRH, GnRH, cortisol, TT3 and TT4, and was negatively correlated with the serum levels of ACTH and IL-1β. The results of linear regression analysis showed that in addition to the HAMD-17, the PSQI score was closely correlated with CRH and IL-1β independently.3) Correlations analysis between cognitive function and serum hormones and cytokine:The partial correlation analysis (controlling for sex, age and educated level) results showed that the MoCA score, which was represented as the general cognitive function, and the SAM score, which was a subjective memory test, were negatively correlated with morning serum levels of TRH, GnRH, interleukin-1β and TNF-α. The correlation between objective memory and hormones and cytokines correlated was complex. Considering the sleep quality and depression degree were also affects hormones and cytokines, we used the partial correlation analysis controlling for the scores of PSQI and HAMD-17. The results showed that MoCA score was negatively correlated with serum levels of IL-1β, TNF-α and GnRH, the ORcM error was negatively correlated with the cortisol serum level, and the OMW error was positively correlated with TSH. The results of the canonical correlation analysis showed that the canonical variable V1 reflected the error numbers of SWM and ORcM, W1 reflected the levels of IL-1β, CRH, cortisol, TT3 and GnRH. And the error numbers of SWM and ORcM were negatively correlated with the cortisol level, and were positively with the levels of IL-1β, CRH, cortisol, TT3 and GnRH.4. ROC curve analysis:For the discrimination of ID from normal and ICD, all the AUCs of cortisol, GnRH, TRH, TT3, TT4, IL-1β, TNF-α were larger than 0.9. Besides, on the distinguish ID from normal, the AUC of CRH was also larger than 0.9.ConclusionID patients did have impairment of memory forms, mainly on the SWM and ORcM. The substypes affected the severity of memory impairment, with the ID patients with early morning awakening and mixed sleep difficulty had the worst memory. The dysfunction of SWM and ORcM in the ID patients was more associated with insomnia and mood disorder than the abnormal of hormones levels (such as CRH, cortisol, TRH, TT3, TT4, GnRH) and cytokines levels (such as IL-1β, TNF-α), especially the sleep structure (such as REM and N2) and depressive mood. The result of PSQI in ID patients was consistent with the sleep parameters recorded by PSG It was possible that there was a close link between the occurrence and development of insomnia and the abnormal serum levels of CRH and IL-1β independently. Cortisol, TT3, TT4, GnRH, CRH, IL-1β and TNF-α may provide information for the diagnosis and differential diagnosis of insomnia. |
PDF Full Text Request