Introduction

The COVID-19 pandemic is a significant public health issue that caused adverse psychological implications (Alzahrani et al. 2022; Aylie et al. 2020; Saladino et al. 2020; Shi et al. 2020; Sundarasen et al. 2020). Hand-in-hand with the physical threat of the loss of life is the psychological sequel from the drastic changes in lifestyle due to the implementation of lockdown measures, and impacts on economic stability such as financial losses and unemployment as a result of business and economic collapse (Ho et al. 2020). With the spread of COVID-19 to Malaysia, the movement control order (MCO) was initiated since March 18, 2020, resulting in mandatory stay-at-home policies such as complete or partial closure of schools and working from home (Prime Minister’s Department 2020). 

Due to the uncertainties faced by worldwide populations, anxiety levels were expected to increase. However, some studies revealed a mixed or a wide range of prevalence of mental health among the general population. A systematic review by Xiong et al. (2020) encompassing eight countries revealed rates of anxiety between 6.33% and 50.90%, depression between 14.60% and 48.30% and psychological distress between 34.43% and 38.00%. Another systematic review revealed prevalence of anxiety of 31.90% and 33.70% of depression was reported in general population (Salari et al. 2020). A survey among university students in France indicated that two thirds of the studied sample perceived that their anxiety had increased (Husky et al. 2020). A nationally representative study in the United Kingdom (UK) found that the level of anxiety had also risen compared to pre-pandemic estimates, but not drastically (Shevlin et al. 2020). A meta-analysis further revealed that after the peak of COVID-19 epidemic in China, the prevalence of anxiety among the general population was 44% (Deng et al. 2021).

Specific groups of people were at increased odds of being affected with mental health due to the COVID-19 pandemic. Consistently, women and younger age groups were more affected by the COVID-19 pandemic (El-Zoghby et al. 2020; Hyland et al. 2020; Moghanibashi-Mansourieh 2020; Sundarasen et al. 2020). Other risk factors identified were staying alone (Sundarasen et al. 2020), suffering from a loss of income (Hyland et al. 2020), low social support (Petzold et al. 2020), having children and being married or cohabitation also increased the risk of experiencing anxiety (Malesza & Kaczmarek et al. 2020).

During the COVID-19 pandemic, a study among 983 Malaysian public and private university students indicated that 29.80% had at least minimal anxiety (Sundarasen et al. 2020). Another study covering 16 universities indicated a higher prevalence, whereby 87.70% of the study participants reported at least mild anxiety based on the Generalised Anxiety Disorder Scale 7-item (GAD-7) (Irfan et al. 2021). On the other hand, a multi-center study on six Southeast Asian countries on healthcare workers, revealed that Malaysians had higher prevalence of depression (14.30%) and anxiety (14.90%) (Chew et al. 2020). However, studies on anxiety during the COVID-19 pandemic in Malaysia had disproportionately focused on university students and healthcare workers. This is a disadvantage as diverse demographic groups for the general population may yield a deeper understanding on at-risk groups who may require targeted interventions. 

During the COVID-19 pandemic, lockdowns and the preference to stay at home may have given rise to loneliness (Luchetti et al. 2020). A study among UK adults showed that between March to April 2020, the prevalence of loneliness was 27%, and those with higher difficulties in regulating their emotions experienced greater loneliness (Groarke et al. 2020). Another study on Polish young adults showed that lonelier individuals, worried about financial difficulties and social isolation had greater anxiety and depression (Okruszek et al. 2020). In Malaysia, Sundaresen et al. (2020) found that university students who stayed alone had two times more of anxiety; this may be related to higher loneliness among those who stayed alone. 

Apart from loneliness, boredom is also an important topic to be addressed during the pandemic. Bozdağ’s (2020) study among Turkish adults found that staying at home resulted in boredom. Boredom has been found to be a factor associated with anxiety, even before the pandemic (Lee & Zelman 2019). Similarly, a study in China during the pandemic found that those with higher boredom reported higher anxiety (Chao et al. 2020). However, there has been a lack of studies investigating the relationship between boredom and anxiety in Malaysia. 

Based on the literature reviewed above, this study aimed to examine the prevalence of anxiety among Malaysians during the MCO and its associated risk and protective factors.

Materials and Methods

Study Design and Participants

A cross-sectional study was undertaken. This study was a part of the International Citizen Project (ICPCOVID) involving 25 countries in the Global South. Questionnaires were self-administered via online mode using the platform provided by Seven Days Digital (https://www.icpCOVID.com/en-my/form/pfcc). Only data from Malaysians were extracted and analysed for this paper. Inclusion criteria were being a Malaysian citizen and aged 18 years and above. Participants were excluded if they did not provide informed consent.

Measures

The following measures were self-administered in English or Malay language.

Socio-demographic Data

Information on socio-demography such as age, gender, marital status, education, economic status, area of domicile, and number of children was collected.

Anxiety Symptoms

The GAD-7 scale was used to screen for anxiety symptoms (Spitzer et al. 2006), measured on a four-point Likert scale ranging from 0 which indicates “Not at all” to 3 which indicates “Nearly every day” to ascertain their anxiety level for the past two weeks. Higher scores indicated higher levels of anxiety. Besides, cut-offs of 0 to 4 indicated normal, 5 to 9 indicated mild, 10 to 14 indicated moderate and 15 and above indicated severe anxiety (Spitzer et al. 2006). The validated Malay version of the GAD-7 was used in this study (Sidik et al. 2012). The internal consistency reliability of the scale score in this study was α = .93.

Well-being

The World Health Organisation (WHO) Well-Being Index was used to measure overall positive mental health (Löwe et al. 2004). The measure consists of five items and has been employed to measure well-being during the COVID-19 pandemic (White & Van Der Boor 2020). Participants answered on a six-point Likert scale ranging from 0 representing “At no time” to 5 representing “All of the time”.  Higher scores indicated better well-being. This questionnaire was translated by the research team members into the Malay language. The internal consistency reliability of the scale score in this study was α = 0.91.

Boredom

The Short Boredom Proneness Scale (Struk et al. 2017) was used to measure boredom during the COVID-19 pandemic. The scale consisted of eight items, which were scored on a seven-point Likert scale ranging from 1 indicated “Very strongly disagree” to 7 indicated “Very strongly agree”. The higher the score, the higher the boredom level in an individual. This questionnaire was translated by the research team members into the Malay language. The internal consistency reliability of the scale score in this study was α = 0.92.

Loneliness

The Short Scale for Measuring Loneliness in Large Surveys (Hughes et al. 2004) consists of three items to measure loneliness. In this study, we added a question pertinent to the COVID-19 situation, which was “At home, do you feel trapped with too many people in the house?”. Participants answered on a three-point Likert scale ranging from 1 indicated “Hardly ever” to 3 indicated “Often”. Higher scores denoted greater levels of loneliness. This questionnaire was translated by the research team members into the Malay language. The internal consistency reliability of the scale score in this study was α = 0.84. 

Coping

A single-item question was included to measure perception of coping, “In your opinion, do you think you are coping well during the COVID-19 situation?” Participants answered on a five-point Likert scale ranging from 1 indicated “Not well at all” to 5 indicated “Very well”. This item was translated by the research team members into the Malay language.

Ethics Considerations 

Ethics approval was obtained from the University of Malaya Research Ethics Committee (UMREC) (Approval No.: UM.TNC2/UMREC-950). Informed consent was obtained from all participants, prior to data collection.

Procedures

Questionnaires that were only available in the English language were translated to Malay by the researchers. The questionnaire was distributed by the researchers through online platforms such as email and social media platforms such as Facebook and WhatsApp using the convenience sampling method. Participants could choose to answer either the English or Malay version of the questionnaire. The participant information sheet provided information about the study, including the research objectives, risk and benefits of participating in this research, confidentiality and organisations involved. Participants were also given a consent form. After participants gave their consent to participate in the study, they completed the self-administered online survey. A psychoeducation feedback sheet was given to participants to understand their level of anxiety, and advice was provided on ways to manage their anxiety, such as getting enough sleep.

Data Analysis

All analyses were performed using SPSS version 27 (IBM Corp 2020, Armonk, NY). Descriptive statistics such as frequency and percentage were used to report the demographic information of participants. Pearson correlation test was done to test the associations between well-being, boredom, loneliness and coping with anxiety. Multiple logistic regression model was constructed to determine the influence of the relevant variables on anxiety (moderate-to-severe anxiety vs. no/mild anxiety) as the criterion. A cut-off score of ≥10 was used to categorise (i) normal/mildly anxious and (ii) moderate to severely anxious groups. The predictors were chosen based on a zero-order correlation conducted prior to running the regression models. Variables that were non-significant in the correlation analysis were not entered into the multiple logistic regression model. The model was adjusted for age, gender, marital status, education, socio-economic status, having children, and area of domicile. Missing data were deleted list wise during analysis.

Results

A total of 543 Malaysian participants responded to the survey. Using G*Power (version 3.1.9.2) post-hoc calculation with a medium effect size (0.15), a 5% significance level, thirteen predictors, and a sample size of 543 resulted in the power of the study to be about 99.9%. Majority of the participants were between 18 to 24 years old (27.8%), female (72.8%), had postgraduate education (35.0%), hail from the lower to middle income category (47.5%), had no children (60.4%) and lived in the urban area (84.5%). More than half (59.3%) reported mild (25.2%), moderate (13.6%) to severe (19.5%) anxiety symptoms (Table 1). Meanwhile, with regards to mental health and well-being indicators, the mean score for anxiety was 7.68 + 6.69, well-being (13.17 + 5.28), boredom (25.67 + 10.42), loneliness (4.78 + 1.89) and coping was 3.60 + 0.83.

Pearson’s correlation analysis results showed that well-being, boredom, loneliness, and coping were significantly correlated with anxiety scores (Table 2). Multiple logistic regression was conducted to examine factors associated with moderate to severe anxiety and showed significant, χ2(21) = 94.83, p<0.001, explaining 24.3% (Nagelkerke R2) of the variance in anxiety and 70.6% of the cases were classified correctly. The Hosmer and Lemeshow test indicated that the model was of good fit (p = 0.243). The results indicated that higher odds of moderate to severe anxiety was reported among those divorced/separated/widowed (aOR 6.778, 95% CI [1.200, 38.296]), adults aged 25 to 34 years (aOR 2.038, 95% CI [1.049, 3.958]), 35 to 44 years (aOR 2.279, 95% CI [1.064, 4.882]), and with higher boredom levels (aOR 1.063, 95% CI [1.035, 1.092]). Those within the 55 to 64 age bracket had lower odds of reporting moderate to severe anxiety (aOR 0.073, 95% CI [0.016, 0.330]) (Table 3).

Discussion

Since this study aimed to examine the prevalence of anxiety among Malaysians from all age groups and factors associated with moderate to severe anxiety, this findings revealed that one in two (55.2%) Malaysians surveyed reported at least mild anxiety since the COVID-19 pandemic and 33.6% reported experiencing moderate to severe anxiety. The proportion of individuals reported to experience anxiety was higher in this study, compared to a pre-COVID-19 study conducted among the general population in Selangor despite utilising a lower cut-off of 8 and above on the GAD-7 questionnaire, with a prevalence of 8.2% (Kader Maideen et al. 2015). However, the prevalence of anxiety in this study was lower compared to another Malaysian study involving university students during the COVID-19 pandemic, whereby 62.2% of the students experiencing moderate to severe anxiety (Irfan et al. 2021). Compared to other countries, the prevalence of anxiety symptoms in our study is higher than the pooled anxiety prevalence of 17 studies reviewed in a meta-analysis (31.9%, N = 63,439)(Salari et al. 2020), and specific countries such as Hong Kong (14.0%, N = 774)(Choi et al. 2020) and Ireland (20.0%, N = 1,041)(Hyland et al. 2020). However, the prevalence was lower than in China (35.1%, N = 7,236) (Huang & Zhao 2020), Taiwan (52.1%, N = 3555)(Wong et al. 2020) and Turkey (45.1%, N = 343) (Ozdin & Bayrak Ozdin 2020). The variability of anxiety prevalence could be due to cohort and methodological differences, and therefore should be interpreted with caution.

The findings revealed that socio-demographic factors such as being separated/divorced/widowed and adults between 25 and 44 years old were at higher odds of developing anxiety, with the highest odds among the separated/divorced/widowed group, while those between 55 and 64 years old had lower odds of anxiety. Those who were separated/divorced/widowed had higher odds of anxiety, and this was also observed in past studies during pre-pandemic phase. A systematic review revealed that the estimated pooled prevalence of anxiety among widowed individuals was 27.0% (Kristiansen et al. 2019), and the risk of heightened anxiety remained two years after the spouses’ death (Feng et al. 2019). Another study indicated that divorced individuals had higher level of anxiety than the general population before the COVID-19 pandemic (Hald et al. 2020) and during the COVID-19 pandemic (Lei et al. 2020; Zhao et al. 2020). These populations may be vulnerable due to pre-existing lack of emotional support or grief from the loss of a partner (Hald et al. 2020; Zhao et al. 2020).

Adults aged between 25 and 45 years old had higher levels of anxiety in this study compared to university students, upon controlling for employment and marital status. Past studies on anxiety during the COVID-19 pandemic generally found that people from younger age group were more affected (Shevlin et al. 2020). However, in this study, further findings showed that there was a difference between the youngest age group (18 to 24 years old), which had significantly lower odds of developing anxiety compared to the age group associated with the workforce (i.e., 25 to 44 years old). This was associated with heavy responsibilities towards family and workplace among adults from this age group, thus engendering higher levels of anxiety when faced with the challenges caused by the pandemic, compared to those who is still as students in university. As mentioned by Zhao et al. (2020) adults in the working age range are the “backbones” of the family in Asian culture, and therefore face heavier burdens, with regards the concept of filial piety is widely practiced by Asian culture, thus accounting for higher prevalence of anxiety in this studied population. Finally, this study is also consistent with Shevlin et al. (2020) suggesting that older adults during the COVID-19 pandemic had lower odds of generalised anxiety and the general study also found that older adults had less worried compared to younger age cohorts (Gonçalves & Byrne 2013).

Among the psychological variables that were investigated in this study (well-being, boredom, loneliness, and coping), only boredom was significantly associated with higher odds of anxiety. This was consistent with another study conducted in China, whereby boredom was linked with higher levels of anxiety during the COVID-19 pandemic (Chao et al. 2020). This study provided an insight in comparing other important psychological factors such as loneliness and coping, boredom may be most prominent in the context of a pandemic which required movement control measures, as the participants were constrained to limit their activities and movement (Brooks et al. 2020). Therefore, we need to take seriously the impact of boredom towards anxiety in the new normal of lockdowns and quarantine situations. 

Several limitations were identified in this study. As this is an online survey, we could only reach participants who have internet access. Due to the urgency of the research and limitations imposed by the lockdown, we had used a non-probability sampling method through an online survey. Henceforth, we could not ensure a diverse representation of our participants. As a result, this study has a larger representation of highly educated individuals, which may skew the findings, even though the influence of education attainment was controlled in the multivariate model. The cross-sectional nature of this study prevented us from making causal inferences on the study results. In addition, as this questionnaire was not administered face-to-face, we could not verify that the individuals who responded were Malaysian citizens. As this was an online survey, we did not control the possibility of multiple responses by the same individual, as we did not collect their IP address so that full participant anonymity was assured. Several important confounding factors were not considered in this paper, such as COVID-19 related factors, and the history of pre-existing mental health conditions. Future studies should attempt to obtain a representative sample while working within the constraints of the COVID-19 pandemic. 

Conclusion

This study showed that socio-demographic features such as being divorced/separated/widowed and within working age were associated with higher odds of moderate to severe anxiety symptoms. Among the psychological factors measured, boredom emerged as the significant factor which increased the odds of anxiety symptoms, compared to loneliness, well-being and coping. Therefore, there is a need to address the risk factors above through public health messages in the identified populations, including coping skills that increase the resilience of the family unit and alleviating boredom in the context of an infectious disease pandemic.

Acknowledgement

This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.