Effect of insomnia on anxiety and depression: Mediation of cognitive failures and moderated mediation of neuroticism

Abstract

BACKGROUND

The coronavirus disease 2019 pandemic has had an ongoing impact on the public’s mental health that requires long-term attention. Exploring the relationship between mental health indicators would aid in identifying solutions to improve public mental health.

AIM

To investigate the prevalence of anxiety, depression, and insomnia and explore the relationship among neuroticism, cognitive failures, and mental health.

METHODS

A cross-sectional survey was conducted in December 2023 using an online platform to recruit adult participants. The neuroticism, subjective cognitive function, and mental health of the participants were assessed using the neuroticism subscale of the Chinese Big Five Personality Inventory Brief Version, cognitive failures questionnaire, generalized anxiety disorder-7, patient health questionnaire-9, and insomnia severity index. Pearson's correlation analysis, independent samples t-tests, one-way analysis of variance, and structural equation model were used to examine the relationship between mental health indicators.

RESULTS

A total of 1011 valid questionnaires were collected, of which 343 were completed by male (33.93%) respondents and 668 were completed by female (66.07%) respondents. The rates of anxiety, depression, and insomnia were 41.3%, 44.6%, and 36.3%, respectively, most cases of which were mild. Among the mental health indicators, there were significant differences by age and between those with siblings and those who were only children. Neuroticism and cognitive failures were significantly positively correlated with mental health indicators. Further moderated mediation analysis showed that cognitive failures mediated the relationship between insomnia and anxiety and between insomnia and depression, with neuroticism moderating the first half of this pathway and the effect being greater in the low-neuroticism group.

CONCLUSION

Cognitive failures and neuroticism play important roles in mental health. Therefore, enhancing subjective cognitive function and regulating emotional stability may contribute to the improvement of mental health.

Key Words: Insomnia; Anxiety; Depression; Neuroticism; Cognitive failures

Core Tip: The coronavirus disease 2019 pandemic has had a significant and ongoing impact on public mental health. A cross-sectional survey involving 1011 participants revealed that approximately 40% reported symptoms of anxiety, depression, and insomnia, with the majority being mild. Cognitive failures mediated the relationship between insomnia and anxiety, as well as between insomnia and depression, and neuroticism moderated the first half of the pathway. Consequently, interventions aimed at enhancing subjective cognitive function and regulating emotional stability in individuals could significantly contribute to the improvement of mental health.

INTRODUCTION

Since the onset of the coronavirus disease 2019 (COVID-19) pandemic, individuals have faced significant challenges to their physical and mental health. Both the general public and COVID-19 survivors have been more susceptible to distressing symptoms, including physical fatigue, diminished sleep quality, and bodily pain, compared with the pre-pandemic era[1,2]. Although approximately 4 years have passed since the outbreak began, the public has continued experiencing lingering effects from this unprecedented event. Several studies have indicated an ongoing mental health crisis with the persistent presence of the aforementioned symptoms over an extended period[3]. Studies have shown that approximately 10%–20% of individuals who contracted COVID-19 developed post-COVID-19 symptoms, together commonly referred to as "long-term COVID". These symptoms often include physical and psychological complications, such as depression, anxiety, and reduced quality of life[4]. Therefore, it is imperative for us to prioritize public mental health.

In the existing research, anxiety, depression, and insomnia are widely recognized as crucial indicators of mental health that have a profound impact on individual emotional well-being and that significantly affect daily life and productivity[5]. Sleep quality plays a particularly crucial role in maintaining mental health. Numerous studies have demonstrated a close relationship among sleep quality, emotional state, and overall well-being. Poor sleep has been associated with increased irritability and negative emotions[6]. Extensive research has confirmed the link among sleep problem, health-risk behaviors, and the worse self-rated mental health[7].

A comprehensive literature review reveals a significant interaction among sleep quality, depression, and anxiety, with sleep playing a pivotal role in regulating depressive mood[8]. Declining sleep quality has been associated with greater symptoms of anxiety and depression[9-11]. Previous studies employing polysomnography to monitor the sleep patterns of patients with depression identified various abnormalities, such as decreased chronic sleep waves, shortened rapid eye movement latency, and increased rapid eye movement density[12]. Considering the intricate relationship between sleep and emotion, investigating the association among insomnia, anxiety, and depression may provide valuable insights for improving public mental health.

A noteworthy finding is that the relationship among insomnia, anxiety, and depression can be influenced by various factors, including personality traits, cognitive inflexibility, dysfunctional beliefs, and other biological, psychological, and social mechanisms[13]. Cognitive failures have been used as indicators to assess individual subjective cognitive functions encompassing perceptual, memory, and motor slips encountered in everyday activities. These functions include a multitude of subjective dimensions, such as impaired decision-making, reduced attention span, and memory lapses[14,15]. Previous research has found that insomnia severity predicts overall cognitive failures and corresponding dimensions[16,17] such that individuals who experience higher levels of cognitive failures in their daily lives are more likely to encounter increased stress, anxiety, and depression[18]. These findings underscore the significant role of cognitive failures in the relationship among insomnia, anxiety, and depression.

Currently, two common theories are proposed to explain cognitive failures in everyday life. The overload theory, which concerns psychological fatigue and the depletion of psychological resources, asserts that human psychological resources are finite and possess dual characteristics. On the one hand, these resources could be utilized to meet the demands of individual activities; on the other hand, they could be replenished through resting and recovery processes to ensure their restoration[19]. Insomnia typically results in inadequate psychological resources for individuals, making them susceptible to fatigue. In a fatigued state, individual motivation diminishes, impacting normal behavioral performance and elevating the risk of experiencing cognitive failures[20].

In contrast, the complaint hypothesis is based on the personality trait level of cognitive failures. This theory proposes that self-awareness with neurotic tendencies may increase individual self-reported cognitive failure. Because individuals with neurotic personalities are susceptible to impairments in self-awareness or lack confidence, they experience unwarranted anxiety that resembles self-directed complaints directed toward objectively occurring cognitive irrationalities. Such self-directed complaints could subsequently increase individual recall of memories linked to cognitive failures[21]. Based on existing research, we hypothesized that insomnia can influence anxiety and depression both directly and indirectly through cognitive failures, and that neuroticism may moderate the association between insomnia and cognitive failures.

Overall, the relationship among neuroticism, cognitive failures, and mental health remains unclear. Therefore, the present study aimed to achieve three main objectives: (1) To assess the mental health status of the recruited participants; (2) To explore the correlation between different mental health indicators; and (3) To investigate whether cognitive failures mediate the relationship between insomnia and anxiety and/or between insomnia and depression and whether neuroticism moderates the first half of the pathway (Figure 1).

Figure 1  The hypothetical model of this study.

MATERIALS AND METHODS

Study design and participants

A cross-sectional study was conducted in December 2023 utilizing convenience sampling to gather data through an online survey of adult participants. The electronic questionnaire was developed and implemented using the Wenjuanxing platform, and participants accessed it by scanning a provided QR code. Standardized questionnaire guidelines were implemented throughout the study. Participants completed the questionnaire voluntarily and anonymously after providing informed consent, with technical measures ensuring single submissions per IP address. A total of 1028 questionnaires were received, of which 17 were excluded following comprehensive validation processes, including integrity verification, logical consistency checks, and response time analysis. Eventually, 1011 valid questionnaires (98.34%) were included in the analysis.

This study received approval from the First Hospital of Shanxi Medical University (No. KYLL-2023-269). This study adhered to the reporting guidelines established by the Strengthening the Reporting of Observational Studies in Epidemiology statement.

Survey instrument

The insomnia severity index (ISI)[22] was employed as the primary instrument for assessing insomnia, comprises seven items that assess sleep symptom severity, sleep pattern satisfaction, functional impairment due to insomnia, perceived impact on quality of life, and associated distress. Each item is evaluated using a 5-point Likert scale (0 = not at all, 4 = very severe). A total score ranging from 0 to 7 indicated insomnia without clinical relevance, 8 to 14 points signified mild insomnia, 15 to 21 points represented clinical insomnia of moderate severity, and 22 to 28 points designated clinical insomnia of severe severity. The internal consistency reliability of the ISI, as measured by Cronbach's alpha, was 0.91, demonstrating excellent reliability.

The generalized anxiety disorder-7 (GAD-7)[23] was utilized to evaluate anxiety levels. The GAD-7 consists of 7 items that assess mental and emotional symptoms experienced during the preceding two-week period. Individuals scored each item with a 4-point Likert scale (0 = not at all, 3 = almost every day), yielding a total score ranging from 0 to 21. Thresholds of 5, 10, and 15 points were adopted to classify anxiety levels as mild, moderate, and severe, respectively. The Cronbach's alpha coefficient of GAD-7 attained a value of 0.93 in our measurement, which underscored its superior reliability.

The patient health questionnaire-9 (PHQ-9)[24] was employed for the assessment of depressive symptoms. It consisted of nine items, each of which was scored on a 4-point Likert scale (0 = not at all, 3 = almost every day). Scores on the PHQ-9 ranged from 0 to 27, with different ranges indicating varying levels of depression severity. Specifically, total scores between 0 and 4 indicated no depression, scores from 5 to 9 suggested the possibility of mild depression, scores from 10 to 14 indicated possible moderate depression, and scores from 15 to 27 signified possible severe depression. In this survey, the Cronbach's alpha coefficient of the PHQ-9 was 0.90, thereby demonstrating excellent reliability.

The Chinese Big Five Personality Inventory Brief Version (CBF-PI-B)[25] comprises 40 items, including five dimensions of neuroticism, conscientiousness, agreeableness, openness and extraversion. Each dimension contains 8 items, assessed on a 6-point Likert scale. Scores range from 1 (highly inconsistent) to 6 (highly consistent). Higher scores on each dimension reflected a stronger manifestation of the corresponding personality trait. In this study, the neuroticism subscale of CBF-PI-B was used to evaluate neuroticism level of the subjects. The Cronbach's alpha coefficient of the neuroticism subscale was 0.82, with good reliability.

The cognitive failures questionnaire (CFQ)[26] was employed to evaluate the frequency of behavioral errors in an individual's daily lives. The CFQ comprises 18 items that were categorized into three dimensions: (1) Attention failure; (2) Memory failure; and (3) Action failure. Participants responded to each question using a 5-point Likert scale ranging from 0 to 4, with a score of 0 indicating never and a score of 4 indicating always. Higher score indicated greater frequency of cognitive failures experienced by the individuals. In our study, the Cronbach's alpha coefficient of the CFQ attained a value of 0.95, indicating a remarkable level of reliability.

Statistical analysis

Statistical analysis was conducted using IBM Statistical Package for the Social Sciences Statistics (SPSS) 22.0. A P value < 0.05 was deemed statistically significant. To elaborate on the demographic characteristics and mental health status of the participants, descriptive statistics were employed. Categorical data were presented as the frequency and percentage, and continuous data were presented as the mean and standard deviation. To examine variations in mental health status across different demographic groups, independent sample t-tests and one-way analysis of variance were performed, with anxiety, depression, and insomnia considered mental health outcomes. Pearson's correlation analysis was performed to investigate the associations among neuroticism, cognitive failures, and mental health.

To investigate the correlation among neuroticism, cognitive failures, and mental health, the bootstrapping method in the SPSS PROCESS macro (models 4 and 7) was used to examine the effects of insomnia on anxiety and depression[27]. The deviation-corrected bootstrap method was used to test the significance of the regression coefficient, with 5000 repeated samples set with a confidence interval (CI) of 95%. If the 95%CI did not encompass zero, the result was considered significant.

RESULTS

Common method deviation test

To examine the potential impact of common method deviation in this study, the unmeasured latent method construct approach was employed[28]. Initially, a first-order baseline model (M1) was constructed using confirmatory factor analysis, incorporating seven variables. These variables were neuroticism, the GAD-7 score, the PHQ-9 score, the ISI, and the three dimensions of the CFQ. Subsequently, a bi-factor model (M2) was constructed by incorporating a method factor before the fit indices of both models were compared (M2-M1). The analysis showed minimal changes in the fit indices (ΔRFI = 0.02, ΔTLI = 0.022, ΔRMSEA = -0.05, ΔSRMR = -0.004), indicating no substantial improvement in the model fit. These results suggested that the model with the added method factor did not significantly outperform the baseline model, indicating the absence of significant common method deviation in the present study.

Prevalence of anxiety, depression, and insomnia

The survey results revealed that the total detection rates for anxiety, depression, and insomnia were 41.3%, 44.6%, and 36.3%, respectively, with the majority being mild symptoms, as depicted in Table 1.

Table 1 Mental health status of participants (n = 1011), n (%).

Degree	Anxiety	Depression	Insomnia
No	593 (58.7)	560 (55.4)	644 (63.7)
Mild	265 (26.2)	273 (27.0)	282 (27.9)
Moderate	105 (10.4)	100 (9.9)	73 (7.2)
Severe	48 (4.7)	78 (7.7)	12 (1.2)

The results indicated that the extent of anxiety (F = 4.162, P = 0.016) and depression (F = 5.005, P = 0.007) significantly differed among the age groups. Specifically, the anxiety scores of those aged 26–50 years were significantly higher than the scores of those aged over 50 years [mean deviation (MD) = 1.508, P = 0.005]. The depression scores of those aged 18–25 years (MD = 1.574, P = 0.013) and those aged 26–50 years (MD = 1.911, P = 0.002) were significantly higher than the scores of those aged over 50 years. However, there were no significant differences in anxiety and depression between those aged 18–25 years and those aged 26–50 years. In addition, the anxiety (t = 2.772, P = 0.006), depression (t = 2.139, P = 0.033), and insomnia (t = 2.295, P = 0.022) scores of only children were significantly higher than the scores of those with siblings. These symptoms did not significantly differ by sex, ethnicity, or education (Table 2).

Table 2 Comparison of mental health among participants with different demographic characteristics, n (%).

Variable	Category		Generalized anxiety disorder-7	Patient health questionaire-9	Insomnia severity index
Sex	Male	343 (33.9)	4.59 ± 4.80	5.14 ± 5.15	6.14 ± 5.41
	Female	668 (66.1)	4.43 ± 4.93	5.33 ± 5.60	6.24 ± 5.41
	t value		0.491	-0.507	-0.273
Ethnicity	Han	996 (98.5)	4.48 ± 4.89	5.25 ± 5.44	6.20 ± 5.42
	Other	15 (1.5)	5.00 ± 4.28	6.40 ± 6.14	6.33 ± 5.19
	t value		0.680	0.146	0.926
Age (years)	18-25	354 (35.0)	4.32 ± 4.72	5.22 ± 5.45	6.28 ± 5.46
	26-50	563 (55.7)	4.79 ± 5.01	5.56 ± 5.49	6.17 ± 5.32
	> 50	94 (9.3)	3.28 ± 4.47	3.65 ± 5.00	6.12 ± 5.78
	F value		4.162a	5.005b	0.053
Education	High school or below	77 (7.6)	5.79 ± 5.33	6.70 ± 6.60	7.40 ± 6.35
	College degree	164 (16.2)	4.41 ± 4.82	5.45 ± 5.56	6.55 ± 5.92
	Bachelor degree	440 (43.5)	4.41 ± 4.85	5.15 ± 5.25	5.93 ± 5.17
	Master degree or above	330 (32.6)	4.31 ± 4.82	4.99 ± 5.33	6.11 ± 5.20
	F value		2.033	2.194	1.892
Only child	Yes	281 (27.8)	5.17 ± 4.98	5.85 ± 5.45	6.83 ± 5.47
	No	730 (72.2)	4.22 ± 4.82	5.04 ± 5.44	5.96 ± 5.37
	t value		2.772b	2.139a	2.295a

^aP < 0.05.

^bP < 0.01.

Correlation analysis

Correlation analysis was conducted to examine the relationships between neuroticism, cognitive failures, and mental health among all participants. The scores of neuroticism were significantly associated with the CFQ (r = 0.454, P < 0.001), GAD-7 (r = 0.608, P < 0.001), PHQ-9 (r = 0.579, P < 0.001), and ISI (r = 0.410, P < 0.001). Furthermore, the findings indicated that the scores on the CFQ, GAD-7, PHQ-9, and ISI were significantly and positively correlated with one another, as detailed in Table 3.

Table 3 Correlation between neuroticism, cognitive failures, and mental health.

	Neuroticism	CFQ	GAD-7	PHQ-9	ISI
Neuroticism	1.000
CFQ	0.454b	1.000
GAD-7	0.608b	0.430b	1.000
PHQ-9	0.579b	0.504b	0.786b	1.000
ISI	0.410b	0.454b	0.518b	0.611b	1.000

^bP < 0.01.

CFQ: Cognitive failures questionnaire; GAD-7: Generalized anxiety disorder-7; PHQ-9: Patient health questionaire-9; ISI: Insomnia severity index.

Moderated mediation effects

Based on the findings of recent studies[29], mediation analysis of model 4 in PROCESS was conducted to test the mediation effect of cognitive failures. The independent variable was insomnia; the dependent variable was anxiety; and the control variables were sex, age, ethnicity, education, and being an only child. Mediation analysis revealed that the total effect between ISI score and GAD-7 score was significant (95%CI: 0.414-0.510), and further analysis revealed that both the direct effect (95%CI: 0.306-0.410) and the indirect effect (95%CI: 0.076-0.133) of CFQ score were significant. These results demonstrate that the CFQ score plays a mediating role between the ISI score and the GAD score (Table 4).

Table 4 The total effect, indirect effect, and direct effect between insomnia severity index and generalized anxiety disorder-7.

	Effect	Boot SE	Bootstrapped 95%CI		Relative effects
			Lower limit	Upper limit
Total effect	0.462	0.024	0.414	0.510
Mediation effect of cognitive failures questionnaire	0.104	0.015	0.076	0.133	22.51%
Direct effect	0.358	0.026	0.306	0.410	77.49%

The Boot SE and Bootstrapped 95%CI respectively refer to the standard error and the 95%CI of the indirect effects estimated by the percentile Bootstrap method of deviation correction. All values are rounded to the nearest three decimal places.

When the same approach was used to examine the mediating role of cognitive failures between insomnia and depression, the results revealed that the total effect between the ISI score and the PHQ-9 score was significant (95%CI: 0.561-0.660). Upon further analysis, both the direct effect (95%CI: 0.426-0.531) and the indirect effect (95%CI: 0.104-0.163) of the CFQ score emerged as statistically significant. These findings suggest that the CFQ score serves as a mediating factor in the relationship between the ISI score and the PHQ-9 score (Table 5).

Table 5 The total effect, indirect effect, and direct effect between insomnia severity index and patient health questionaire-9.

	Effect	Boot SE	Bootstrapped 95%CI		Relative effects
			Lower limit	Upper limit
Total effect	0.610	0.025	0.561	0.660
Mediation effect of cognitive failures questionnaire	0.132	0.015	0.104	0.163	21.64%
Direct effect	0.478	0.027	0.426	0.531	78.36%

The Boot SE and Bootstrapped 95%CI respectively refer to the standard error and the 95%CI of the indirect effects estimated by the percentile Bootstrap method of deviation correction. All values are rounded to the nearest three decimal places.

Model 7 in PROCESS was then used to test the moderated effect of neuroticism. The independent variable was insomnia; the dependent variable was anxiety; the moderated variable was neuroticism; and the controlling variables were sex, ethnicity, age, education, and being an only child. As shown in Table 6, two regression equations were established. In equation 1, the product term of insomnia and neuroticism had a significant predictive effect on cognitive failure (β = -0.025, SE = 0.008, P < 0.01, 95%CI: -0.040 to -0.010), and neuroticism had a significant moderating effect. Using the same procedure and replacing only the dependent variable with depression, the moderating effect of neuroticism remained significant (Table 7). These findings were demonstrated in Figure 2.

Figure 2 The moderated mediation model testing the indirect effect of insomnia on anxiety and depression through cognitive failures by neuroticism. ^cP < 0.001.

Table 6 Bootstrap confidence intervals for the moderated mediation effects with dependent variable of anxiety.

Predictors	Cognitive failures (M)				Anxiety (dependent variable)
	Coeff	SE	t value	95%CI	Coeff	SE	t value	95%CI
Insomnia (X)	0.849c	0.072	11.742	0.707-0.991	0.358c	0.026	13.574	0.307-0.410
Cognitive failures (M)					0.096c	0.011	8.614	0.074-0.118
Neuroticism (W)	0.605c	0.055	11.094	0.498-0.712
X × W	-0.025b	0.008	-3.309	-0.040 to 0.010
Constant	21.748c	2.659	8.178	16.529-26.966	4.506c	1.025	4.398	2.495-6.517
		R2 = 0.311				R2 = 0.325
		F = 56.525, P < 0.001				F = 69.098, P < 0.001

^bP < 0.001.

^cP < 0.001.

Coeff: Coefficient; M: Mediator; W: Moderator; X: Independent variable.

Table 7 Bootstrap confidence intervals for the moderated mediation effects with dependent variable of depression.

Predictors	Cognitive failures (M)				Depression (dependent variable)
	Coeff	SE	t value	95%CI	Coeff	SE	t value	95%CI
Insomnia (X)	0.849c	0.072	11.742	0.707-0.991	0.479c	0.027	17.814	0.427-0.532
Cognitive failures (M)					0.122c	0.011	10.783	0.099-0.144
Neuroticism (W)	0.605c	0.055	11.094	0.498-0.712
X × W	-0.025b	0.008	-3.309	-0.040 to -0.010
Constant	21.748c	2.659	8.178	16.529-26.966	4.117c	1.038	3.965	2.079-6.155
		R2 = 0.311				R2 = 0.444
		F = 56.525, P < 0.001				F = 114.639, P < 0.001

^bP < 0.001.

^cP < 0.001.

Coeff: Coefficient; M: Mediator; W: Moderator; X: Independent variable.

To further explain the moderating effect, the participants were divided into high-neuroticism and low-neuroticism groups by adding or subdividing the mean ± SD, and a simple slope test was performed (Figure 3). The results revealed that insomnia was a significant predictor of cognitive failures in both the high-neuroticism and low-neuroticism groups, with the predictive effect of insomnia on cognitive failures greater in the low-neuroticism group (β = 0.102, P < 0.001) than in the high-neuroticism group (β = 0.076, P < 0.001). Additionally, at the three levels of neuroticism, the mediating effect of cognitive failures on the relationship between insomnia and anxiety, as well as insomnia and depression, was diminishing (Tables 8 and 9).

Figure 3  Interaction effect of insomnia and neuroticism on cognitive failures.

Table 8 The mediating effect of cognitive failures on insomnia and anxiety at different neurotic levels.

Neuroticism	Effect	Boot SE	Bootstrapped 95%CI
			Lower limit	Upper limit
-6.941 (M-1SD)	0.098	0.013	0.072	0.126
0 (M)	0.081	0.011	0.06	0.103
6.941 (M+1SD)	0.064	0.011	0.043	0.088

The Boot SE and Bootstrapped 95%CI respectively refer to the standard error and the 95%CI of the indirect effects estimated by the percentile Bootstrap method of deviation correction. All values are rounded to the nearest three decimal places. M: Mediator.

Table 9 The mediating effect of cognitive failures on insomnia and depression at different neurotic levels.

Neuroticism	Effect	Boot SE	Bootstrapped 95%CI
			Lower limit	Upper limit
-6.941 (M-1SD)	0.125	0.016	0.096	0.158
0 (M)	0.103	0.013	0.080	0.130
6.941 (M+1SD)	0.082	0.013	0.057	0.110

The Boot SE and Bootstrapped 95%CI respectively refer to the standard error and the 95%CI of the indirect effects estimated by the percentile Bootstrap method of deviation correction. All values are rounded to the nearest three decimal places. M: Mediator.

DISCUSSION

In this study, we investigated the relationship among neuroticism, cognitive failures, and mental health. The findings revealed that a considerable proportion of participants reported experiencing varying degrees of anxiety, depression, and insomnia, with most cases being classified as mild, and that neuroticism was closely related to cognitive failures and mental health. Moreover, the results of moderated mediation analysis suggested that cognitive failures mediated the relationship between insomnia and anxiety as well as between insomnia and depression and that neuroticism moderated the first half of the pathway.

This survey found that the percentages of participants who reported experiencing symptoms of anxiety, depression, and insomnia were 41.3%, 44.6%, and 36.3%, respectively. These figures aligned closely with the results of a cross-sectional survey conducted in 2020 among medical staff treating COVID-19 patients in China, which reported rates of 44.6%, 50.4%, and 34.0% for anxiety, depression, and insomnia symptoms, respectively[30]. A comparative analysis of these data revealed a striking similarity between the prevalence of mental health symptoms among participants and frontline healthcare workers during the pandemic.

Another survey of medical staff and community residents showed that the percentages of hospital anxiety and depression scale and ISI scores exceeding the normal cutoff value among all participants were 51.94% and 31.11%, respectively[31]. Additionally, a study investigating anxiety levels among college students in the post-pandemic era reported that approximately 40% exhibited mild to severe anxiety symptoms[32]. Collectively, these investigations underscore the continued importance of addressing public mental health in the post-epidemic era, emphasizing the need for ongoing attention and intervention.

This study revealed higher levels of anxiety and depression among participants aged 25–50 years compared with those aged 50 years and above. Studies consistently indicate that coping strategies for stressful life events are associated with mental health, such as depression, anxiety, and stressful experiences[33,34]. Older individuals generally possess greater life experience and better coping skills, which enabled them to adopt more adaptive strategies when they were confronted with the challenges posed by COVID-19. Conversely, young people may have faced additional stress related to schooling or work disruptions caused by the pandemic[35,36].

In addition, we observed that compared with individuals with siblings, only children experienced higher levels of anxiety, depression, and insomnia. The term “only children” refers to individuals who do not have any siblings and are the sole offspring of their parents. Research on the mental health of only children is mixed. On the one hand, some studies have shown that only children are more likely to experience depression, pessimism, and serious psychological problems[37,38]. Researchers posit that only children lack the natural support of siblings and often miss out on important opportunities to develop psychosocial skills, obtain emotional support, and take advantage of learning opportunities that could be provided by siblings[39]. On the other hand, other studies have revealed that adolescents from families with more than one child may face mental health burdens due to the responsibility of maintaining close sibling bonds and report poorer psychological adjustment ability[40,41].

When individuals face stressful life events, receiving instrumental support (i.e., financial aid) is one of the most effective ways to alleviate stress and reduce the negative impact of these events[42]. In our survey sample, participants with siblings may have been more likely to have received greater support when dealing with the pressure induced by COVID-19. We observed no sex differences in mental health indicators despite studies showing that females experienced more anxiety and depression symptoms than males[43-45]. However, since our survey included approximately twice as many women as men, it is possible that the sex ratio imbalance contributed to the inconsistent results. The male and female samples were analyzed separately, and the results were presented in the Supplementary Tables 1 and 2.

In addition, we observed that insomnia was strongly associated with anxiety and depression, which is consistent with existing research. Experimental studies on sleep deprivation demonstrated that negative stimuli intensify the physiological response of emotions, highlighting the significant link between sleep and maintaining adaptive emotional regulation and reactivity[46]. Based on previous research and the results of our study, we could reasonably speculate that good sleep reduces the formation of negative emotions. The exacerbation of negative emotions may be closely tied to the intrinsic mechanisms of the sleep process, further supporting the notion that insomnia could contribute to heightened levels of depression and anxiety[6]. Therefore, optimizing sleep quality may play a proactive role in alleviating depressive and anxious moods.

Interestingly, cognitive failures played a partial mediating role in the relationship between insomnia and anxiety and between insomnia and depression. Different aspects of sleep, including its duration and quality, are crucial for cognitive function and emotion[47]. Researchers have found that sleep duration is correlated with distractibility in cognitive failures[48]. Lack of sleep correlates with a decline in cognitive functions such as vigilance, attention, and decision-making[49]. Additionally, increased cognitive failure has been associated with higher levels of anxiety[50]. These findings indicate that sleep not only directly affects emotion but also influences anxiety and depression by impacting cognitive failures.

More importantly, we observed that neuroticism moderated the first half of the mediated pathway. Specifically, at low levels of neuroticism, insomnia had a significant positive predictive effect on cognitive failures. At high levels of neuroticism, the positive predictive effect of insomnia on cognitive failures was reduced, suggesting that neuroticism mitigated the risk effect of insomnia on cognitive failures. This was an interesting result. Previous research has positively associated both insomnia and the symptoms of insomnia with neuroticism[51,52]. Based on existing research, it seems easy to speculate that neuroticism might be a risk factor for insomnia affecting cognitive failures.

However, research investigating the relationship between neuroticism and health has yielded mixed results. Despite the overwhelming evidence linking neuroticism to adverse health outcomes, several studies have nevertheless identified neuroticism as a potential protective factor for personal health[53,54]. The theory of healthy neuroticism posits that neuroticism can manifest as either adaptive or maladaptive behaviors, depending on other characteristics of the individual[55]. A tendency towards healthy neuroticism might be associated with a greater propensity for seeking treatment, which could potentially lead to earlier detection of medical issues and enhanced recovery due to adherence to medication. This offers new insights into the role of neuroticism in mental health. It would be prudent to say that a larger sample size is needed to further validate the results of this study in the future.

Despite our noteworthy findings, our study has several limitations. First, the sample size of this study was limited, and a larger scope of investigation should be conducted in the future to gain a dynamic understanding of the state of public mental health in the post-epidemic era. Second, as a cross-sectional study, it was challenging to establish a causal relationship among neuroticism, cognitive failures, and mental health indicators. Therefore, longitudinal studies should be designed to examine the relationship among these variables in greater depth. Third, this study only conducted an online survey at the behavioral level. In the future, electrophysiology, neuroimaging, and other technologies could be combined to further explore the neural mechanism among neuroticism, cognitive failures, and mental health to further reveal the neurobiological basis of mental health.

CONCLUSION

This study showed that the mental health of the public should not be neglected. Our findings indicate that cognitive failures mediate the relationship between insomnia and anxiety, as well as the relationship between insomnia and depression, and neuroticism moderates the first half of the pathway. Consequently, it is imperative to maintain a long-term focus on public mental health. Enhancing subjective cognitive function and regulating emotional stability in individuals may prove beneficial in ameliorating overall mental well-being within the general population.