1 INTRODUCTION

Stroke is defined as a sudden loss of focal cerebral function. This process lasts for 24 hr or more and is thought to be caused by an inadequate blood supply to some parts of the brain (ischemic stroke), or spontaneous hemorrhage of brain substance (primary intracerebral hemorrhage), where brain imaging was normal or showed evidence of recent ischemia or hemorrhage. Conversely, focal arterial ischemia with transient symptoms (lasting <24 hr) and without any evidence of infarction according to pathological or imaging examinations should be considered a TIA (transient ischemic attack; Sacco et al., 2013). Stroke is the second most common cause of death worldwide as well as the leading cause of long-term disability (Tsai, Thomas, & Sudlow, 2013). Mood and emotional disturbances are frequent symptoms in stroke survivors (Hackett, Köhler, O'Brien, & Mead, 2014), including post-stroke depression (PSD), post-stroke anxiety (PSA), post-stroke emotional incontinence (PSEI), post-stroke anger proneness (PSAP), and post-stroke fatigue (PSF; Kim, 2016). Previous studies have found that PSA is closely associated with PSD (Campbell Burton et al., 2013), and three-quarters of the anxious patients had comorbid major or minor depression (Castillo, Schultz, & Robinson, 1995). The core symptoms of PSA are excessive anxiousness or worry, and difficulty in controlling worries. In addition to these symptoms, a diagnosis of anxiety requires three or more of the following: restlessness, decreased energy, poor concentration, irritation, nervous tension, and insomnia (American Psychiatric Association, 2013). Patients with PSA do not have a previous history of anxiety disorder prior to the onset of the stroke and can be considered an emotional complication after stroke. The combined rate of anxiety by time after stroke was as follows: 20% within 1 month after stroke onset; 23% 1–5 months after stroke onset; and 24% six or more months after stroke onset (Campbell Burton et al., 2013). Despite the high prevalence of anxiety after stroke, the understanding of PSA is still limited (Cumming, Blomstrand, Skoog, & Linden, 2016; Kim, 2016; Liu, Cai, Zhang, Zhu, & He, 2018).

Anxiety-related neural circuits span a wide range of brain structures, including subcortical white matter and the limbic system (Allsop, Vander Weele, Wichmann, & Tye, 2014; Westlye, Bjørnebekk, Grydeland, Fjell, & Walhovd, 2011). However, the pathophysiological mechanisms underlying the development of PSA remain unclear. Currently, some studies have found that PSA is associated with greater dependence in daily activities, higher mortality and poorer overall quality of life (Broomfield, Quinn, Abdul-Rahim, Walters, & Evans, 2014; Campbell Burton et al., 2013; Li et al., 2019). Therefore, identifying the risk factors for PSA is urgent and necessary. Previous studies have shown that age, gender, inability to work, depression, smoking, stroke severity, low levels of serum 25-hydroxyvitamin D [25(OH)D ≤ 38.48 nmol/L], stroke areas, previous history of insomnia, and low socioeconomic status are associated with anxiety disorders (Ayerbe, Ayis, Crichton, Wolfe, & Rudd, 2014; Bushnell et al., 2014; Kuchcinski et al., 2017; Leppävuori, Pohjasvaara, Vataja, Kaste, & Erkinjuntti, 2003; Li et al., 2019; Thayabaranathan et al., 2018; Wu et al., 2016). However, no consensus has been reached regarding the risk factors of anxiety in stroke patients.

Anxiety is one of the most common mental disorders, and an epidemiological study has shown that insomnia and anxiety disorders are closely related (Ohayon, 2002). A systematic review concluded that the relationship between insomnia and anxiety is bidirectional (Alvaro, Roberts, & Harris, 2013). Insomnia has been found to be a risk factor of anxiety in the general population as well as in pre- and post-menopausal women and adolescents (Friedman, Brooks, Bliwise, & Yesavage, 1993; Johnson, Roth, & Breslau, 2006; Neckelmann, Mykletun, & Dahl, 2007; Osnes, Roaldset, Follestad, & Eberhard-Gran, 2019; Swanson, Pickett, Flynn, & Armitage, 2011; Terauchi et al., 2012). Similarly, epidemiological studies have shown a relationship between sleep quality and anxiety. Previous studies have shown that sleep disorders, difficulty in falling asleep, and frequent sleep deficits increase the risk of anxiety. Poor sleep quality is associated with anxiety, whether in children and adolescents, college students, pregnant women, or the elderly (Adams & Kisler, 2013; Brown et al., 2018; Volkovich, Tikotzky, & Manber, 2016; Xiong et al., 2019). A study of 3,987 Chinese elderly, aged 60 years or more, published in 2020, suggests that compared with those with good sleep quality, the OR (95% CI) of anxiety for those with poor sleep quality was 5.12 (3.88–6.77) (Shi et al., 2020). Post-stroke insomnia (PSI) has a significant incidence in the acute phase of stroke (Sterr et al., 2018). Up to 70% of the patients with acute stroke have sleep disorders including excessive daytime sleepiness, insomnia, hypersomnia, and fatigue (Pasic, Smajlovic, Dostovic, Kojic, & Selmanovic, 2011). Nevertheless, the association between sleep quality before stroke and PSA in patients with acute ischemic stroke (AIS) has not been elucidated.

The present study was conducted to explore the relationship between sleep quality at admission and PSA in Chinese patients 1 month after the stroke.

2 MATERIALS AND METHODS

2.1 Subjects

The patients' data from September 2015 to June 2016 were collected from the stroke unit of the First Affiliated Hospital of Wenzhou Medical University. The inclusion criteria were as follows: (a) Chinese ethnicity, (b) 18–80 years of age, (c) suffered acute stroke within 3 days before admission, (d) CT and/or MRI supported diagnosis of AIS, and (e) informed consent from the patient. A total of 412 AIS patients were screened. From them, 173 patients were excluded (36 for TIA, 47 for history of nervous system disease, 35 for history of tumor or trauma, eight for a history of depression, six for a history of anxiety, 15 for severe aphasia, and 26 for decline). Thus, 349 patients were enrolled in the present study. One month after AIS onset, follow-up could not be done with 22 patients. The final analysis included 327 patients (Figure 1).

This study was conducted following the Helsinki Declaration (World Medical Association, 2013). Participation in the study was voluntary, and written informed consent was obtained. The participants could withdraw their consent without any explanation at any time during the study. This study was approved by the ethical committee of The First Affiliated Hospital of Wenzhou Medical University (Wenzhou, China).

3 RESULTS

4 DISCUSSION

As far as we know, this study is the first one to explore the association between sleep quality and PSA. Our results demonstrate that poor sleep quality before AIS is a significant risk factor for anxiety in AIS patients 1 month after the stroke onset, regardless of whether or not there was a history of sleep disorders before the stroke.

Sleep is of vital importance for health (Luyster, Strollo, Zee, & Walsh, 2012). Sleep deprivation can exert harmful effects on mental health (Roberts, Roberts, & Duong, 2009). As a response to mild stressors, sleep deprivation could create escalated negative effects on people (Minkel et al., 2012). Individuals who suffer from lack of sleep are more prone to develop anxiety and depression (Choueiry et al., 2016; Luyster et al., 2012). Numerous studies have confirmed the bidirectional association between insomnia and anxiety (Alvaro et al., 2013). Sleeping problems are a vital manifestation of anxiety and depression (Kokras et al., 2011). Previous studies have shown that insomnia or poor sleep quality is an important precursor and indicator of anxiety (Neckelmann et al., 2007; Sørengaard et al., 2019; Vedaa et al., 2016). These studies have highlighted the correlation between insomnia or poor sleep quality and anxiety. The results of the present study indicate that poor sleep quality is a significant risk factor for anxiety in AIS patients 1 month after the stroke onset. As a common and long-lasting complication, early recognition and treatment are particularly important, but the mechanism of PSA induced by poor sleep quality remains unclear. Currently, some previous studies have shown that sleep deprivation leads to impaired executive functioning (Drummond, Gillin, & Brown, 2001; Durmer & Dinges, 2005; Nilsson et al., 2005), including the inhibition of attention and memory functions (Drummond et al., 2001; Durmer & Dinges, 2005). Reduced functional connectivity with the prefrontal cortex may reduce connections to brain areas associated with executive functions, thus reducing the brain's capacity to regulate and inhibit anxiety (Cox & Olatunji, 2016; Ma, Dinges, Basner, & Rao, 2015; Verweij et al., 2014; Wright et al., 2015) Disorder of the HPA axis is also associated with anxiety-related disorders. Sleep deprivation increases the secretion of cortisol in human body (Wright et al., 2015). Specifically, the decrease of sleep time is related to the gradual decline of cortisol (Van Lenten & Doane, 2016) and the increase of cortisol secretion at night (Abell, Shipley, Ferrie, Kivimäki, & Kumari, 2016). The HPA axis abnormality is also evident in anxiety disorders. For example, cortisol output is reduced in patients with PTSD (Morris, Compas, & Garber, 2012), and in women with anxiety disorders, cortisol responses are insensitive when facing acute stressors (Zorn et al., 2017). Over time, chronically increases in cortisol or an inadequate response to acute stress may influence the occurrence of anxiety disorders (Cox & Olatunji, 2020). Moreover, most sleep regulation models involved the monoamine and cholinergic systems, and the inhibitory GABA (γ-aminobutyric acid) mechanisms in sleep regulation (Mignot, Taheri, & Nishino, 2002). Because the dysfunction of these neurotransmitter systems is related to anxiety (Kent, Mathew, & Gorman, 2002), the changes of GABA caused by sleep disorders may mediate the occurrence of anxiety disorders (Staner, 2003).

A study from the South London Stroke Register indicated that in long-term observations, PSA was a common problem, with the 10-year prevalence rate ranging from 17% to 24% and a cumulative incidence of 57% (Ayerbe et al., 2014). A meta-analysis of 44 studies comprising 5,760 stroke patients reported a pooled PSA prevalence of 20% 1 month after the stroke onset (Campbell Burton et al., 2013), which is similar to our results. We found that the patients with higher NIHSS scores were more likely to develop PSA than those with lower NIHSS scores. Stroke severity and physical disability have been reported to be predictors of PSA (Castillo et al., 1995), consistent with the findings of our study. Using multiple stepwise logistic regression analysis, we found that lower serum 25-hydroxyvitamin D [25(OH)D] level (≤38.48 nmol/L) was also a risk factor of PSA, which is in accordance with the results of our previous study (Wu et al., 2016). The present study demonstrated females and patients with poor functional outcome tended to have a higher prevalence of PSA. Nevertheless, in multiple stepwise logistic regression analysis, no association was found between these two factors and PSA, which may be attributed to the relatively small number of females and patients with poor functional outcome enrolled in this study. Furthermore, we found moderate and low incomes were correlated with PSA. Few previous studies have focused on the relationship between annual income and PSA. A study of anxiety in gynecologic cancer patients showed that low household income was associated with anxiety (Corrales et al., 2018). A cross-sectional observational study in China pointed out that the factors associated with poor concordance rate included the patient's annual household income, and clinically significant self-reported symptoms of anxiety and hypochondriasis (Wang, Murray, et al., 2019). The MMSE score and educational level were not associated with PSA, which needs to be validated through further studies.

This study has several limitations: First, the relatively small sample size weakened the statistical strength of the study. Thus, a multicenter study with a larger sample size is needed. Second, a 1-month (rather than 3 or 6 months) period was used in PSA assessment. It could be more useful to evaluate “true” anxiety rather than “reactive” anxiety. Third, the short follow-up precluded us from exploring the effects of lengthy institutionalization on PSA. Moreover, severe aphasia patients were not included in the study due to their inability to complete the assessment, which might have weakened the generalization of the present study. Fourth, baseline levels of anxiety were not controlled. Fifth, although the PSQI is a measure of sleep over the past month, retrospective bias may increase the reporting of sleep disturbance following stroke. Finally, the data of depression symptoms were not available.

5 CONCLUSIONS

This study demonstrates that poor sleep quality before stroke is associated with PSA and may be an independent risk factor of PSA 1 month after the stroke onset. Hopefully, our findings would contribute to the prevention and treatment of PSA.

Our findings suggest that for patients with AIS, attention should be paid to the screening of sleep quality level at admission, and for patients with poor sleep quality, the assessment of anxiety level should be strengthened for early intervention. Further, multicenter and larger prospective studies are needed to confirm this association.

CONFLICT OF INTEREST

None of the authors has any conflict of interest to disclose.

AUTHORS' CONTRIBUTIONS

Xiao MJ and He JC designed the study. Xiao MJ and Huang GQ interpreted data. Feng L, Luan XQ, Wang QZ, and Ren WW prepared figures. Huang GQ, Xiao MJ, and Feng L did the statistical analyses. Huang GQ, Luan XQ, Wang QZ, and Chen SY screened and extracted data. He JC supervised study. All authors have made an intellectual contribution to the manuscript and approved the submission.