Practices regarding the COVID-19 outbreak and life satisfaction: A moderated mediation model of psychological distress and fear of Covid-19

Abstract

Increasing attention has been recently paid to the influences of the COVID-19 outbreak on the human psyche due to its potentially detrimental after-effects. However, little is known about the effects of practices introduced to contain the COVID-19 outbreak, such as social isolation and lockdowns, on individuals' psychological problems and well-being, or how a fear of COVID-19 amplifies or reduces these effects. Using a sample of 2680 Vietnamese adults, data were collected between 15 August and 15 November 2021 through an online-based survey. This study adopted a moderated mediation model. Remarkably, the fear of COVID-19 was not only found to significantly exacerbate the adverse effects of psychological distress on life satisfaction, but it also significantly decreased the impact of COVID-19 practices on satisfaction with life. The fear of COVID-19 significantly moderated the mediation effect of psychological distress on the relationship between COVID-19 practices and life satisfaction. This study makes significant and novel contributions to our extant knowledge about the destructive consequences of COVID-19. The findings of our study can benefit policymakers and practitioners and include valuable recommendations on how to avert psychological crises and increase individuals' well-being during or after a pandemic.

1. Introduction

The specter of a new coronavirus disease (COVID-19) pandemic has been looming over the global economy, dramatically changing and disrupting many aspects of our lives. As of December 26, 2022, over 200 countries worldwide had reported cases of and deaths due to COVID-19. The total number of confirmed cases has exceeded 651 million, while the total number of deaths has surpassed 6.6 million (WHO, 2022). The COVID-19 pandemic has thus created challenges worldwide (Duong, 2021; Ghosh, 2022; Yuen et al., 2022), and continues to do so still today (Ortiz-Calvo et al., 2020). The rapid spread of the COVID-19 pandemic, and its deadly consequences, drove governments to impose a variety of strict measures, such as physical distancing, quarantine, school closures, closures of nonessential services and businesses, and even partial or complete nationwide lockdowns, in attempts to curb the outbreak of the COVID-19 pandemic (Kokkinos et al., 2022; Fosu & Ankrah Twumasi, 2021). Although practices related to the COVID-19 outbreak (PRC) helped governments better control the spread of the infectious disease, these measures can significantly influence people's way of life, as well as leading to a body of negative psychological issues (Karataş & Tagay, 2021). This is why the effects of the COVID-19 outbreak on the mental health and quality of life of both infected and uninfected people have been of prime interest to scholars over the past two years (e.g., Akintunde et al., 2021; Duong, 2021; Ortiz-Calvo et al., 2020).

The negative psychological consequences of the COVID-19 outbreak on individuals have been empirically confirmed by prior studies (e.g., Borah et al., 2022; Kokkinos et al., 2022; Wen et al., 2022), and some scholars have emphasized that the prolonged enforcement of strict measures such as social isolation and lockdowns can exacerbate mental health problems such as stress, anxiety or psychological distress (Truong et al., 2021). Bou-Hamad et al. (2021) defined PRC as various types of practices engaged in by the local population, such as social isolation, wearing facemasks, and washing hands, to control the spread of COVID-19. These practices may increase psychological distress (PD) (Taylor et al., 2020), which refers to the unpleasant emotions experienced by an individual when overwhelmed (Kessler et al., 2003). Moreover, PRC can affect individuals' life satisfaction (LS), which is defined as their general judgment of well-being and life quality, and which reflects their emotional and cognitive appraisal of their lives (Diener et al., 1985). However, little attention has been paid to examining the statistical links between PRC, PD and LS (Carvalho & de Sousa Soares, 2020).

Fear is defined as “a basic, intense emotion aroused by the detection of imminent threat, involving an immediate alarm reaction that mobilizes the organism by triggering a set of physiological changes” (American Psychological Association, 2020, p. 413). Thus, fear of COVID-19 (FOC) is identified as an unpleasant psychological state arising unexpectedly in the face of danger from the current pandemic scenario (Kubo et al., 2022). Even though several recent studies have reported that FOC can result in mental health problems and reduce well-being (Deniz, 2021; Giesbrecht et al., 2022), to the best of our knowledge, no prior studies have considered FOC as a moderator or examined its moderation effects on the paths between PRC and PD and LS. This study is thus the first to investigate the moderation effects of FOC; this is necessary to shed more light on how the impacts of PRC on PD and LS will be changed depending on the different levels of FOF. Our study adopted a moderated mediation model to explore the impacts of PRC on PD and LS, to investigate the mediating role of PD, and to examine the moderating influences of FOC on the paths between PRC and LS via PD.

2. Theoretical framework and hypothesis development

2.1. Practices regarding COVID-19, psychological distress, and life satisfaction

COVID-19 spread into almost every country in the world, resulting in a global health crisis, decreasing social and physical interactions, and psychological disorders (Labrague, 2021). According to the World Health Organization (WHO), a variety of austerity practices such as social isolation, wearing facemasks, washing hands, closing schools, and enforcing lockdowns, were a crucial strategy for curbing the spread of the infection, and reducing the number of infected cases and deaths – essential due to the subsequent overloading of healthcare systems around the world (Bou-Hamad et al., 2021). Although prior studies have illustrated that social isolation significantly contributes to the onset of depression, psychological distress, and other mental health problems (Taylor et al., 2018; Taylor et al., 2020), little is known about the psychological consequences of PRC (Al-Wutayd et al., 2021).

Jacobson et al. (2020) argue that mental health problems such as anxiety, fear, sleep disturbance, and negative thoughts have significantly increased after the application of strict measures such as the stay-at-home policy and PRC. A recent study showed that PRC can increase the levels of anxiety, depression, and stress among Pakistani adults (Rizwan et al., 2021). However, Oosterhoff et al. (2020) carried out a study with 683 American youths and reported that engaging in PRC (i.e., social distancing) was not statistically correlated with their mental health, but that their particular motives for engaging in PRC were significantly related to different psychological issues. This study emphasized that individuals' motives to prevent a COVID-19 infection or to avoid judgment from surrounding people increased their anxiety symptoms. Moreover, Melo and Soares (2020) report that PRC, such as social isolation, can increase depressive mood, fear and anxiety. They also suggest that it is necessary to research the psychological effects of these practices, but studies on this relationship are scarce and show biases.

Prior studies discuss how individuals' psychological well-being and LS can be affected by PRC such as social isolation and lockdowns (e.g., Duarte & Jiménez-Molina, 2021; Kokkinos et al., 2022; Simon et al., 2021), yet there is scant empirical evidence to support this, underlining the need for further empirical research (Brooks et al., 2020; Clair et al., 2021). Indeed, in a qualitative study, Williams et al. (2020) revealed that practices such as physical distancing and social isolation related to COVID-19 policies dramatically influenced people's mental health, well-being, or satisfaction with life. Moreover, Preetz et al. (2021) also indicated that COVID-19 practices (e.g., school closures, stay-at-home policies, and social distancing) interrupted adults' emerging social life. Continuously repetitive PRC (e.g., washing hands) can result in unpleasant emotions related to the highly infectious nature of COVID-19 (Truong et al., 2021), while helping and motivating others to follow the governments' recommendations can make individuals happier and increase their well-being (Curry et al., 2018). This can explain why satisfaction with life was reduced and mental health problems were increased during the pandemic, but this study did not empirically examine the link between PRC and LS. In contrast to these discussions, several current studies claim that although measures related to the COVID-19 outbreak significantly increased psychological and mental health issues, PRC also can augment LS (e.g., Bachmann et al., 2021; Duong, 2021). This is because these practices make them realize the value of life and strive for a better work-life balance, and they have time to exercise more regularly and eat more healthily (Bou-Hamad et al., 2021). Consequently, it can be hypothesized that PRC can significantly increase PD and LS.

2.2. Psychological distress as a mediator

Even though PD was determined to have a negative effect on LS (Carranza Esteban et al., 2022; Duong, 2021; Maria-loanna & Patra, 2020), not all people with COVID-19 practices show the same levels of LS. Thus, further investigating the underlying mediation mechanism of PD would broaden our knowledge of how PRC affects LS via PD, considering PD as a detrimental antecedent of well-being (Carranza Esteban et al., 2022; Lam & Zhou, 2020). The current research proposes that PD is a potential mediator of the linkage between PRC and LS. The reasons why PD can mediate the effect of PRC on LS are twofold. First, previous research indicated that PD acted as a mediator of the effects of different antecedents, such as fear and anxiety of COVID-19 (Duong, 2021; Satici et al., 2020), short periods of sleep (Zhi et al., 2016), and procrastination (Maria-loanna & Patra, 2020) on LS. Second, Melo and Soares (2020) believe that maintaining COVID-19 practices such as social distancing can significantly increase PD, and then result in a loss of meaning as well as decreased self-worth. Thus, PRC significantly influences PD, which in turn is positively related to LS. To the best of our understanding, no prior studies have examined whether PD mediates the link between PRC and LS.

2.3. Fear of COVID-19 as a moderator

Theoretically, moderating influences occur when a third variable – a moderator – changes the influence level or direction of the effect of an independent variable on a dependent variable. The introduction of a moderator is considered to be an appropriate method for addressing this inconsistency (Baron & Kenny, 1986). First, according to earlier studies, the correlation between PD and LS appears elusive. Some studies indicate that PD is negatively associated with LS (e.g., Duong, 2021; McCleary-Gaddy & James, 2022). In contrast, Zhi et al. (2016) insist that there is a negative correlation between PD and LS underlying the mediation mechanism of sleep quality. Moreover, previous studies differed widely in their assessments of PD's degree of influence on LS (Carranza Esteban et al., 2022). Second, although no prior studies have provided empirical evidence for the relation between PRC and LS, we believe that PRC might have direct and indirect impacts on LS via PD, but that not all individuals experience this impact equally. These ambiguous correlations between PRC, PD, and LS therefore suggests the existence of a moderator.

FOC is also known as a detrimental factor for LS (Abd-Ellatif et al., 2021) and the main cause of PD (Duong, 2021; Kubo et al., 2022). However, FOC has rarely been scrutinized as a moderator in the paths between different antecedents and LS, even though it is one of the most established and studied constructs in terms of examining the effects of COVID-19 on individuals' mental health problems and well-being (Baerg & Bruchmann, 2022; Jue & Ha, 2022). Our study, therefore, further proposes that FOC can serve as a moderator of the direct effects of PRC and PD on LS, as well as of the indirect effects of PRC on LS via PD. As mentioned above, PRC may be significantly related to PD (Melo & Soares, 2020) and LS (Duarte & Jiménez-Molina, 2021; Kokkinos et al., 2022). The above effects can be increased or decreased by FOC (Karataş & Tagay, 2021; Yeni et al., 2022). For example, when individuals have a high FOC, they tend to experience a larger reduction in LS due to PD (Duong, 2021). In contrast, when individuals have a low FOC, the negative impact of PD on their LS is weaker (Labrague, 2021). Thus, the positive impact of PRC on LS becomes weaker while the negative impact of PD on LS becomes stronger when FOC is high. Also, the indirect influential effect of PRC on LS via PD is increased when FOC is high.

2.4. Proposed conceptual model and hypotheses

A moderated mediation model was employed in the current research to bridge the research gap and explain the effects of PRC on PD and LS and explore the moderation effects of FOC on the paths from PRC and LS via PD (see Fig. 1 ). Thus, the hypotheses are formulated as follows:

H1

PRC is positively correlated with PD.

H2

PRC is positively correlated with LS.

H3

PD is negatively correlated with LS.

H4

PD mediates the link between PRC and LS. Particularly, PRC is positively correlated with PD, which in turn is partially correlated with LS.

H5

FOC significantly moderates the link between PRC and LS. Particularly, the positive impact of PRC on LS is weakened when FOC is high.

H6

FOC significantly moderates the link between PD and LS. Particularly, the negative impact of PD on LS is reinforced when FOC is high.

H7

The mediation impact of PD on the link between PRC and LS is moderated by FOC. Particularly, the positive mediation effect of PD on the link between PRC and LS is strengthened when FOC is high.

Fig. 1.

The proposed moderation mediation conceptual framework.

3. Methods

3.1. Participants

Our research utilized the convenience sampling approach and an online survey to collect the data between 15 August and 15 November 2021 in Vietnam. During this time period, Vietnam was experiencing the fourth wave of the COVID-19 pandemic (Nguyen et al., 2021). Certain restrictions, social distancing measures, and vaccinations had been seriously implemented by the government to limit the spread of COVID-19, especially when almost all the patients in Vietnam were infected with the Delta variant (Nguyen Van et al., 2021). In the same period, a novel variant of SARS-CoV-2 called Omicron (B.1.1.529) had appeared and been declared as a variant of significant concern by the WHO and scholars (Malink et al., 2022). Therefore, it was appropriate to utilize an online survey using Google Forms (Duong, 2021).

The participants in the study were explicitly informed that their participation was entirely voluntary and that they could withdraw from the research at any time. Additionally, they were assured that all the information provided would be kept confidential and used solely for academic purposes. The study was approved by the institutional review board of the authors' affiliated institution. To collect the data, 8000 online questionnaires were administered directly via popular social media platforms in Vietnam, including personal emails, Facebook, Zalo, and Viber, inviting individuals to participate in the survey. Finally, 2680 questionnaires were completed, giving a response rate of 33.5 %. The social-demographic information of the respondents is presented in Table 1 .

Table 1.

Social-demographic profiles of respondents.

Variables		Frequency	%	Population structure (%)	χ2 for difference between sample and population
Gender	Male	1174	43.8	49.8	0.82 (p < 0.001)
	Female	1506	56.2	50.2	0.64 (p < 0.001)
Age	18–28	1632	60.9	29.6	16.09 (p < 0.001)
	29–38	561	20.9	22.1	0.07 (p < 0.001)
	39–48	343	12.8	17.7	1.88 (p < 0.001)
	49–58	102	3.8	14.8	31.84 (p < 0.001)
	Over 59	42	1.6	15.8	126.02 (p < 0.001)
Monthly income	<10 million VND	1551	57.9	N/A	N/A
	From 10 to 20 million VND	655	24.4
	From 20 to 30 million VND	314	11.7
	Over 30 million VND	160	6.0
Educational level	High school	591	22.1	36.1	8.86 (p = 0.087)
	Bachelor's degree	1837	68.5	57.9	1.64 (p = 0.339)
	Master/PhD degree	252	9.4	5.9	1.30 (p = 0.496)
Marital status	Single	1777	66.3	51.9	3.12 (p = 0.685)
	Married	903	33.7	48.1	6.15 (p = 0.959)
Jobs/employments	Worker	92	3.4	N/A	N/A
	Office staff	399	14.9
	Student	1328	49.6
	Techer/lecturer	160	6.0
	Civil servants/public employees	163	6.1
	Freelancer worker	271	10.1
	Others	267	10.0

Notes: N = 2680, 1 USD = 23,585.00 VND (exchange rate on 26 December 2022).

In order to assess whether the sample was representative, the demographic characteristics of the participants were compared with those of the population of Vietnam, as recommended by GSO (2020). First, 43.8 % respondents are females while only 43.8 % are males. This ratio is similar to the overall Vietnamese population structure in 2020. Moreover, 57.9 % respondents held bachelor's degrees, and only 5.9 % had obtained a master's/PhD degree. These proportions are also similar to the actual percentages in the population. Moreover, more than half (66.3 %) of the participants were single, which also adequately reflects the percentage of single people in the whole population. The majority of the respondents were aged 18–28, accounting for 60.9 %, which is significantly different to the percentage of this age group in the Vietnamese population. However, the proportions of the other age groups were similar to those of the whole population. Generally, the research sample met the requirements of representativeness for further analysis.

3.2. Measures and questionnaire development

To serve the research purposes and test the hypotheses, all of the scales were adopted from previous studies (see Table 2 ). Particularly, the seven-item scale measuring PRC was derived from the research of Bou-Hamad et al. (2021) while the scale for FOC, which included seven items, was adopted from Martínez-Lorca et al. (2020). Additionally, the scale for PD developed by Kessler et al. (2003), which consisted of ten items, was used in the present study. Finally, the scale for LS was adapted from Diener et al. (1985) with five items. All the items were scored by the participants from 1 to 7, corresponding with “strongly disagree” to “strongly agree”, respectively.

Table 2.

Cronbach's alpha, exploratory factor analysis, confirmatory factor analysis and descriptive statistics of constructs.

Code	Variables	Mean	SD	Pattern matrix (EFA)				λ (CFA)
				F1	F2	F3	F4
PRC	Practices regarding COVID-19 outbreak (α = 0.933; CR = 0.929; AVE = 0.651)	5.2884	1.27820
PRC1	I wash hands for a min.30 s	5.3420	1.4586		0.799			0.796
PRC2	I follow the government's recommendation related to the pandemic	5.2463	1.49648		0.825			0.827
PRC3	I try not to leave the house	5.1500	1.56789		0.773			0.790
PRC4	I’ am going to exercise regularly	5.1104	1.60246		0.770			0.795
PRC5	I’m going to eat heathier	5.3474	1.48673		0.873			0.835
PRC6	I follow the recommendations for social distancing	5.3313	1.50000		0.845			0.810
PRC7	I motivate others to follow the recommendations related to the pandemic	5.4914	1.46577		0.829			0.792
PD	Psychological distress (α = 0.958; CR = 0.955; AVE = 0.682)	3.5177	1.49934
PD1	I often feel tired out for no good reason	3.6205	1.70491	0.774				0.778
PD2	I often feel nervous	3.7451	1.70861	0.739				0.815
PD3	I often feel so nervous that nothing could calm me down	3.3519	1.72191	0.898				0.874
PD4	I often feel hopeless	3.2183	1.75116	0.874				0.847
PD5	I often feel restless or fidgety	3.5433	1.76493	0.812				0.848
PD6	I often feel so restless that I could not sit alone	3.4388	1.77746	0.844				0.861
PD7	I often feel depressed	3.8034	1.82833	0.744				0.797
PD8	I often feel that everything is not effort	3.5552	1.77556	0.862				0.810
PD9	I often feel so sad that nothing could cheer me up	3.3993	1.74961	0.900				0.836
PD10	I often feel worthless	3.5015	1.82840	0.863				0.785
FOC	Fear of COVID-19 (α = 0.914; CR = 0.915; AVE = 0.682)	3.9551	1.56588
FOC1	I am most afraid of COVID-19	4.2246	1.83967				0.874	0.829
FOC2	It makes me uncomfortable to think about COVID-19	4.0582	1.80528				0.818	0.848
FOC4	I am afraid of losing my life because of COVID-19	4.0754	1.89937				0.862	0.822
FOC5	When watching news and stories about COVID-19 on social media, I become nervous or anxious	3.9093	1.78971				0.789	0.863
FOC6	I cannot sleep because I′ am worrying about getting COVID-19	3.5078	1.73859				0.692	0.770
LS	Life Satisfaction (α = 0.922; CR = 0.916; AVE = 0.687)	4.3933	1.48931
LS1	In most ways my life is close to my ideal	4.1944	1.68715			0.795		0.753
LS2	The conditions of my life are excellent	4.2205	1.65915			0.893		0.840
LS3	I am satisfied with my life	4.4567	1.70380			0.897		0.911
LS4	So far, I have gotten the important things I want in life	4.6396	1.72263			0.776		0.810
LS5	If I could live my life over, I would change almost nothing	4.4552	1.76002			0.820		0.822

Note: N = 2680; α = Cronbach's alpha; EFA = Exploratory factor analysis; λ = Factor loading; CFA = Confirmatory factor analysis.

The final section of the questionnaire survey included questions about the respondents' social-demographic profile, such as gender, age, monthly income, educational level, employment status, and marital status. Since the target respondents were Vietnamese people, the survey was first translated from English into Vietnamese. Then, three language experts translated it back to English independently and compared the versions in order to secure the accuracy of the translation process. Some minor modifications were made to some items so that they were appropriate for the Vietnamese culture and language.

3.3. Statistical analysis

All analyses were carried out utilizing SPSS 24.0 and AMOS 24.0. First, it is essential to examine whether or not the data are distributed normally before testing the hypotheses (Duong, 2023); thus, the skewness and kurtosis of PRC, PD, FOC and LS were calculated. Second, we estimated Cronbach's alpha for each variable and conducted both exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) to confirm the reliability and validity of the scales. Third, after standardizing the variables, hierarchical regression analysis was employed to test direct correlations, controlled by the socio-demographic variables (age, gender, educational degrees, etc.). Last, the PROCESS macro was utilized with Model 4 (Hayes, 2018) to examine the mediation effect of PD, whereas moderated mediation analysis was conducted utilizing Model 15 (PROCESS macro) to examine the moderated mediation effects of FOC. Based on 5000 random samplings, bootstrapping with 95 % confidence intervals was used to estimate the statistical significance of association in Model 4 and Model 15 (Hayes, 2018).

4. Results

4.1. Scale assessment

Table 2 demonstrates the reliability and discriminant validity of the constructs. Cronbach's alphas for PRC, PD, FOC, and LS were 0.933, 0.958, 0.914, and 0.922, respectively, while the correlated item-total correlation was higher than 0.4 for all the items. Therefore, internal consistent reliability was confirmed (Hair et al., 2020). Then, all the items were adjusted for the EFA analysis with the method of principal axis factoring and Promax rotation. Nevertheless, the initial results showed that the factor loadings of FOC3 “My hands become clammy when I think about COVID-19” and FOC7 “My heart races or palpitates when I think about getting COVID-19” only reached 0.425 and 0.455, respectively, which were lower than the threshold value of 0.5. FOC3 and FOC7 were therefore excluded. Afterwards, the EFA was conducted again. The results showed that a total of four factors were loaded, with an extracted variance of 74.098 %. The KMO (Kaiser-Meyer-Olkin) value was 0.942, indicating a high degree of sampling adequacy, and all the items had factor loadings >0.5. The EFA was then reperformed, and four factors were loaded with a total extracted variance of 74.098 %, the KMO (Kaiser-Meyer-Olkin) value equaled 0.942, and the factor loadings of all the items were higher than 0.5.

Then, the remaining items were adjusted to the CFA, and the results revealed large degree of fit indices: χ²(299) = 2593.669; χ²/df = 8.674; p < 0.001; GFI = 0.924 > 0.9; AGFI = 0.904 > 0.9; CFI = 0.963 > 0.9; TLI = 0.957 > 0.9; NFI = 0.959 > 0.9, and RMSEA = 0.054 < 0.08 (Anderson & Gerbing, 1988). As demonstrated in Table 2, the factor loadings (λ) of the items were higher than 0.5. Moreover, the Average Variance Extracted (AVE) and Composite Reliability (CR) of the constructs were above the cut-off values of 0.5 and 0.7, respectively. Hence, the reliability and discriminant validity of the constructs were confirmed (Hair et al., 2020).

To control for common method variance (CMV), a procedurally and statistically approached methodology was implemented in our research. First, in the survey, the items of all the constructs were mixed and presented in a shuffled manner. Our research also tried to take out any signs that could affect the respondents at any point, and to survey a diverse group of adults in different regions of Vietnam. In addition, Harman's one-factor test was conducted using an unrotated factor solution. The results showed an explained variance of 36.067 %, which was significantly lower than the 50 % threshold value, indicating that common method bias was not a significant issue in the data (Podsakoff et al., 2003). CFA analysis was conducted with all the items constrained to one factor. The results showed poor indices of fitness: χ²(324) = 33,147.286; Chi-Square/df = 102.306; GFI = 0.394; AGFI = 0.293; CFI = 0.476; TLI = 0.433; NFI = 0.474; RMSEA = 0.194. Consequently, the absence of common method variance in our study was confirmed (Hair et al., 2020). This result also affirmed the representativeness of the research sample (Kruskal & Mosteller, 1980).

4.2. Preliminary analyses

Table 3 illustrates the hierarchical regression analysis of direct correlations. Noticeably, socio-demographic variables, such as gender, age, monthly income, educational degree, marital status, and employment are controlled as covariates in these regression analyses. First, the results report that PD is significantly correlated with gender (γ = 0.136, p < 0.05), educational level (γ = −0.227, p < 0.001), and marital status (γ = 0.274, p < 0.001), but it is not significantly related to age, monthly income, or employment (p > 0.05). Second, LS is significantly associated to employment (γ = −0.040, p < 0.05), but it is not significantly related to gender, age, monthly income, educational level, or marital status. Third, PRC is positively associated with PD (γ = 0.106, p < 0.001) and LS (γ = 0.331, p < 0.001), but PD is negatively correlated with LS (γ = −0.136, p < 0.001). Thus, H1, H2, and H3 were statistically supported.

Table 3.

Hierarchical regression analysis of the direct correlations, controlling by socio-demographic variables.

Variables	Psychological distress				Life satisfaction				VIF
	Model 1		Model 2		Model 3		Model 4
	γ	t	γ	t	γ	t	γ	t
Constant	3.274⁎⁎⁎	19.519	2.766⁎⁎⁎	13.899	4.207⁎⁎⁎	25.099	2.174⁎⁎⁎	11.113
Gender	0.139⁎	2.367	0.136⁎	2.334	−0.013	−0.213	−0.039	−0.707	1.027
Age	−0.007	0.170	−0.016	−0.337	0.033	0.770	0.006	0.160	1.968
Monthly income	0.028	0.734	0.020	0.507	−0.026	−0.661	−0.057	−1.556	1.492
Educational level	−0.221⁎⁎⁎	−4.036	−0.227⁎⁎⁎	−4.171	0.079	1.443	0.089	1.707	1.087
Marital status	0.278⁎⁎⁎	3.428	0.274⁎⁎	3.388	0.135	1.659	0.083	1.084	1.787
Employment	0.009	0.454	0.008	0.672	−0.036	−1.816	−0.040⁎	−2.090	1.290
Practices regarding COVID-19 outbreak			0.106⁎⁎⁎	4.696			0.331⁎⁎⁎	15.395	1.023
Psychological distress							−0.136⁎⁎⁎	−7.424	1.025
ΔR2	0.016		0.024		0.004		0.109
R2	0.016		0.024		0.004		0.109
ΔF	7.294⁎⁎⁎		9.452⁎⁎⁎		1.729⁎⁎⁎		40.885⁎⁎⁎

Note: N = 2680, VIF: Variance Inflation Factor. P-values are provided in brackets.

^⁎p < 0.05.

^⁎⁎p < 0.01.

^⁎⁎⁎p < 0.001.

4.3. Testing for the mediation effect

As illustrated in Table 4 , the results of mediation analyses using Hayes's PROCESS macro (Model 4) revealed that the PRC was positively correlated with PD (β _PRC-PD = 0.112, p < 0.001; 95 % CI [0.0669, 0.1554]), which in turn was negatively linked with LS (β _PD-LS = −0.2975, p < 0.001; 95 % CI [−4.066, −0.1885]). The bootstrapping results also reported that the indirect effect of PRC on LS via PD (ab = 0.0148, 95 % CI [0.0081, 0.0026]) and the mediation effect of PD accounted for 10.51 % of the incremental variance in LS of the total effect. Furthermore, the direct effect of PRC on LS was statistically significant (β _PRC-LS = 0.5993, p < 0.001; 95 % CI [0.4927, 0.6859]). Hence, the link between PRC and LS was partially mediated by PD, which supported H4.

Table 4.

Moderated mediation analyses.

Predictor	Β (Coeff)	se	t	p	LLCI	ULCI
	Psychological distress (M) (R2 = 0.0090; F = 24.2661⁎⁎⁎)
Constant	2.9299⁎⁎⁎	0.1228	23.8650	0.0000	2.6891	3.1706
Practices regarding COVID-19 outbreak (X)	0.1112⁎⁎⁎	0.0255	34.2072	0.0000	0.0669	0.1554
	Life satisfaction (Y) (R2 = 0.1051; F = 157.2377⁎⁎⁎)
Constant	1.6983⁎⁎⁎	0.2949	5.7579	0.0000	1.1199	2.2766
Practices regarding COVID-19 outbreak (X)	0.5893⁎⁎⁎	0.0493	11.9596	0.0000	0.4927	0.6859
Psychological distress (M)	−0.2975⁎⁎⁎	0.0556	−5.3502	0.0000	−0.4066	−0.1885
Fear of COVID-19 (Z)	0.2790⁎⁎⁎	0.0788	3.5411	0.0004	0.1245	0.4334
Practices regarding COVID-19 outbreak x Fear of COVID-19 (X x Z)	−0.0838⁎⁎⁎	0.0133	−6.3090	0.0000	−0.1099	−0.0578
Psychological distress x Fear of COVID-19 (M x Z)	0.0842⁎⁎⁎	0.0114	7.3624	0.0000	0.0617	0.1066
Conditional effects of the practices regarding COVID-19 outbreak (focal predictor) at the values of fear of COVID-19 (moderator): Z = M ± S.D.	Direct effect	se	t	p	LLCI	ULCI
-1 S.D. (−1.5659)	0.3890⁎⁎⁎	0.0248	15.7134	0.0000	0.3405	0.4376
M (0.0000)	0.2577⁎⁎⁎	0.0227	11.3377	0.0000	0.2132	0.3023
+1 S.D. (1.5659)	0.1265⁎⁎⁎	0.1265	3.5258	0.0004	0.0561	0.1968
Conditional effects of the psychological distress (focal predictor) at the values of fear of COVID-19 (moderator): Z = M ± S.D.	Direct effect	se	t	p	LLCI	ULCI
-1 S.D. (−1.5659)	−0.0965⁎⁎	0.0329	−2.9301	0.0034	−0.1610	−0.0319
M (0.0000)	0.0353	0.0240	1.4708	0.1415	−0.0118	0.0824
+1 S.D. (1.5659)	0.0171⁎⁎⁎	0.0266	6.2712	0.0000	0.1148	0.2193
	Boot indirect effect	BootSE			BootLLCI	BootULCI
Indirect effect of practices regarding COVID−19 outbreak (X) on life satisfaction (Y) via psychological distress (M)	0.0148⁎	0.0037			0.0081	0.0226
Fear of COVID-19 (Z)	Boot indirect effect	BootSE			BootLLCI	BootULCI
Moderated mediation effect	0.0094⁎	0.0024			0.0050	0.0143
Conditional indirect effects of practices regarding COVID-19 outbreak on life satisfaction at the values of fear of COVID-19 (moderator): Z = M ± S.D.	Effect	BootSE			BootSE	BootSE
-1 S.D. (−1.5659)	−0.0107⁎	0.0047			−0.0208	−0.0025
M (0.0000)	0.0039	0.0030			−0.0017	0.0104
+1 S.D. (1.5659)	0.0186⁎	0.0048			0.0100	0.0287
Pairwise contrasts between conditional indirect effects (effect 1 minus effect 2)	Effect 1	Effect 2	Contrast	BootSE	BootLLCI	BootULCI
	0.0039	−0.0107	0.0146	0.0037	0.0080	0.0223
	0.0186	−0.0107	0.0293	0.0073	0.0161	0.0447
	0.0186	0.0039	0.0146	0.0037	0.0090	0.0223

Note: N = 2680.

^⁎⁎⁎p < 0.001.

^⁎⁎p < 0.01.

^⁎p < 0.05.

4.4. Testing for moderated mediation

Table 4 also shows the results of moderated mediation analysis utilizing Hayes' PROCESS macro (Model 15). The results show that the interaction between PRC and FOC is significantly and negatively associated with LS (β _PRC⁎FOC-LS = −0.0838, p < 0.001; 95 % CI [−0.1099, −0.0578]). It accounts for 1.29 % of the incremental variance in LS of the main effects: ΔR² = 0.0129. The result of the simple slope tests illustrates that the relationship between PRC and LS is significant at a low level of FOC (β _simple = 0.3890, p < 0.001; 95 % CI [0.3405, 0.4376]), yet this correlation is much lower at a high level of FOC (β _simple = 0.1265, p < 0.001; 95 % CI [0.0561, 0.1968]). Therefore, H5 is supported.

The interaction between PD and FOC is significantly and positively correlated with LS (β _PD⁎FOC-LS = 0.0842, p < 0.001; 95 % CI [0.0617, 0.1066]). It accounts for 1.75 % of the incremental variance in LS of the main effects: ΔR² = 0.0175. The result of the simple slope tests reveals that the association between PD and LS is significant and negative at a low degree of FOC (β _simple = −0.0965, p < 0.01; 95 % CI [−0.1610, −0.0319]), but that this relation is significant and positive at a high degree of FOC (β _simple = 0.0171, p < 0.01; 95 % CI [0.1148, 0.2193]). It means that when FOC is at a low level (all observations-SD and run in the model), the effect of PD on LS is negative; when the FOC is at a high level (all observations + SD and run in the model), the effect of PD on LS (Life satisfaction) is positive. Thus, this result supports H6.

Having confirmed that the moderation link is supported, the moderated mediation effect is then further analyzed to examine whether or not FOC significantly moderates the mediation impact of PD on the link between PRC and LS. The output of this analysis illustrates the detailed outcomes of the interaction impact by demonstrating one standard deviation (SD) above and below the mean (M). This result enables us to confirm the value of FOC for which the conditional indirect effect is significant at α =0.05. The result shows that the moderated mediation effect is positive and has non-zero probability (β _{moderated mediation} = 0.0094, p < 0.001; 95 % CI [0.0050, 0.0143]). Strikingly, this moderated mediation effect is significant when the level of FOC is low (-1SD) or high (+1SD), but not significant when it is equal to the standardized scale.

Also, as detailed in Table 4, pairwise contrasts between conditionally indirect impacts (Effect 1 minus Effect 2) illustrate that the impacts at the mean FOC are higher than those at a low FOC (Contrast = 0.0146; 95 % CI [0.0080, 0.0223]), while the effects at a high FOC are higher than the same effects at a low FOC (Contrast = 0.0293; 95 % CI [0.0161, 0.0447]), and higher than the same impacts at the mean FOC (Contrast = 0.0146; 95 % CI [0.0090, 0.0223]). The results revealed that the conditionally indirect effects of PRC on LS through PD at various degrees of FOC are significantly different from each other. In other words, the bootstrapped 95 % CIs did not consist of 0 for three pairwise contrasts between conditionally indirect impacts, which therefore supported the hypothesis that the mediation impact of PD on the link between PRC and LS was moderated by FOC. Hence, H7 was supported.

5. Discussion

In the present research, PRC significantly contributed to an increase in PD in the cultural context of Vietnam. This provides empirical support for suggestions that the strict measures of the government, such as stay-at-home policies, social isolation, and lockdown, can result in psychological problems, including PD (Al-Wutayd et al., 2021; Melo & Soares, 2020). However, our study found that PRC was positively interrelated with LS. This finding demonstrates that the strict measures of the government to curb the spread of the COVID-19 pandemic not only have negative impacts, but that these practices also have positive perspectives. Indeed, when individuals act upon the government's recommendations (hand washing, staying at home, social distancing, motivating others to follow these recommendations, doing exercise regularly, and eating healthily), they realize the values and meaning of life better, and are more satisfied with life (Bachmann et al., 2021). In line with some prior studies, our results reported a negative correlation between PD and LS (e.g., Lam & Zhou, 2020; Maria-loanna & Patra, 2020). This means that when individuals perceive high PD, their LS is decreased. In addition, while prior studies have acknowledged that PD served as a mediator of the paths from different antecedents to LS (e.g., Maria-loanna & Patra, 2020; Satici et al., 2020; Zhi et al., 2016), our study documented that PD partially mediated the effect of PRC on LS. As such, practices to restrict the spread of the COVID-19 pandemic first influenced PD, which in turn significantly affected LS.

Consistent with our expectation, the link between PD and LS was moderated by FOC. Even though a body of prior empirical studies has been performed to confirm that FOC can amplify PD and reduce LS (e.g., Abd-Ellatif et al., 2021; Duong, 2021; Kubo et al., 2022), the present research is the first, to the best of our knowledge, to reveal that the impact of PD on LS is negative when FOC is at low level, however, when FOF is high, the effect of PD on LS is positive. Specifically, individuals with a high PD were more likely to be less satisfied with life when they had a high FOC. Moreover, our study also reported that FOC negatively moderated the relation between PRC and LS. That is, the interaction effects of PRC and FOC on LS were significant. This finding illustrates that when individuals conform with the government's recommendations in order to curb the spread of COVID-19, they are likely to be more satisfied with their lives (Jue & Ha, 2022; Melo & Soares, 2020). However, if they have a high FOC, this beneficial effect is weakened. Finally, our study revealed that the mediation effect of PD on the relationship between PRC and LS was also significantly moderated by FOC. This finding, therefore, fortifies our understanding of the adverse effects of COVID-19 on mental health issues and well-being.

5.1. Study strengths and practical implication

This study is novel since it is the first to empirically investigate the effects of PRC on PD and LS, as well as adopt the moderated mediation model to investigate the moderation effects of FOC on the paths from PRC to PD and LS. This research enriches our knowledge about the role of practices/measures to curb the spread of the COVID-19 outbreak and the adverse effects of COVID-19 on adults' mental health. Firstly, this study confirmed that PRC is positively correlated with PD and LS. Secondly, this research found that the link between PRC and LS was negatively moderated by FOC. Thirdly, FOC was found to significantly moderate the adverse effect of PD on LS. In addition, the indirect path from PRC on LS via PD was positively moderated by FOC. Finally, the findings of this study can help policymakers and practitioners to not only realize the importance of practices or measures for restricting the spread of COVID-19, but also to take into account the adverse effects of these practices on individuals' psychological issues, including PD. Moreover, in order to increase individuals' well-being during COVID-19 outbreaks, besides effective measures to curb the spread of the disease, governments should implement appropriate strategies to reduce individuals' FOC and PD.

5.2. Limitations

Although this study provides significant contributions to both theoretical and practical aspects, some following limitations should be considered. First, our sample was recruited by using convenience sampling with an online survey approach. Thus, the representativeness of the sample could be decreased, and future research should adopt random sampling methods to augment confidence in the results. Second, the target participants of this research were adults aged over 18 years old, rather than other groups, such as young people or people with COVID-19; the present findings can therefore not be generalized to other populations. Therefore, in order to deal with this issue, further studies might adopt this model with other samples. Third, a follow-up survey was not carried out in this study to advance our knowledge of the development of or changes in the relations between the constructs in the conceptual model. Future studies could therefore utilize longitudinal designs to provide robust statistical results for the causal correlations that have been conceptualized in this study. Moreover, due to the nature of our online survey, over half of the respondents were aged 18–28 years old, and this can be considered as another limitation of our study. Future studies should use a random sample, which would be more representative of the population. In addition, our results showed that the moderator had a rather minor effect, and researchers could re-test these influences in order to shed a brighter light on this effect. Finally, our study only investigates the links between PRC, PD and LS, and the moderation effect of FOC; therefore, further studies could expand the research model to enrich our understanding of the different influences that aspects of the COVID-19 pandemic have on human society.

Abbreviations

PRC

Practices related to COVID-19

PD

Psychological distress

FOC

Fear of COVID-19

LS

Life satisfaction

VND

Vietnamese Dong

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

Data will be made available on request.