The COVID-19 pandemic led to the implementation of extraordinary health measures, example of which was mandatory social isolation or lockdown. This event consisted of prolonged periods of time without leaving the home and it has been shown to have repercussions on mental health. Specifically, people with no history of clinical disorders presented symptoms of distress and those with a pre-existing diagnosis showed a worsening of their symptomatology (1). In this regard, Rossi et al. (2) in a large-scale study (N = 18,147) showed that at the beginning of the pandemic there were severe symptoms of depression (20.8%), anxiety (7.3%) and perceived stress (22.9%). In fact, depressive (27.6 %) and anxious (32.6 %) symptoms were higher than in previous years where they barely reached 4% (3, 4). A similar situation occurred during the Severe Acute Respiratory Syndrome (SARS) epidemic when lockdown produced changes in people's anxiety and stress, increasing depression and suicide rates (5). The impact was such that 1 year after the epidemic people still showed symptoms of distress and stress (6).

Stress is defined as an individual's actual or anticipated experience of an adverse, unpredictable and uncontrollable situation (7). In this sense, the cognitive appraisal theory indicates that stress is an individual's evaluative response to a situation greater than the personal resources (8). According to its persistence over time, it can be categorized as acute if the stressor stimulus is transitory, and chronic if the difficulties perceived by the individual are maintained for a prolonged period of time (9). Stress can appear in different ways and have adverse consequences on health. Thus, during COVID-19 lockdown, there is evidence that perceived stress was directly related to insomnia, loneliness and anxiety (10–12).

Anxiety is conceived as the set of physical and psychological manifestations, which can be diffuse, prolonged or intense, that are attributed to a stimulus of real or imaginary danger (13). It is strongly related to fear since they are both experienced in a dangerous situation (14). At low levels, it can be considered as an adaptive response to potential threats (15), and at high levels, it de-structures the individual's personality (15, 16). Thus, anxiety symptoms during lockdown showed a strong relationship with other clinical conditions (12) and its presence in the pandemic gave rise to a new construct such as Worry about COVID-19 contagion.

Worry about COVID-19 contagion can be understood as thoughts about the probability of becoming infected by coronavirus, which in turn affect mood and the ability to perform activities and cause a problem for the individual (17). Studies associated with infectious diseases show that the assessment of the risk of becoming infected is greater in early periods of the pandemic (18). Some studies have indicated that people more worried about the pandemic presented a greater probability of adopting preventive behaviors (19). Furthermore, they experienced a deterioration in their mental health and life satisfaction and an increase in generalized anxiety, stress (20) and depression (21–23).

Depression is a mood disorder that consists of a set of physical and mental symptoms, such as sleep difficulties, reduced appetite, low mood, reduced interest in pleasurable activities, among others (24). It is one of the considerable factors in understanding the risk of suicide, which has increased in recent years (25, 26). Likewise, during health emergency, these situations have also had relevance (5). These data manifest a possible increase in depression figures during the COVID-19 pandemic. They are estimated to reach a prevalence of 27% in the general population (11) and its implications are greater when associated with feelings of loneliness (27).

Loneliness, from a cognitive perspective, is understood as a negative feeling that is the result of unmet needs for bonding and social activities (28). It is usually experienced as a feeling of not belonging and misunderstanding or lack of companionship (29, 30). This causes psychological ailments, expressed in boredom or emotional emptiness (31), and generates fear of finding oneself alone. Fear of loneliness can be defined as worry responses and avoidance behavior to the experience of being alone (32). During lockdown, people who were isolated showed a significant increase in loneliness (27); specifically, young people and adults (33, 34). In these populations, loneliness impacted interpersonal relationships, acquisition of addictive behaviors, depressive symptoms, and decreased subjective wellbeing (27, 28, 35).

The above proves that anxiety, depression, stress, worry about COVID-19 contagion and fear of loneliness may be relevant variables to be examined during COVID-19 lockdown in Peru. In fact, specialized literature states that the onset of a depressive episode is usually preceded by stressful events that alter the individual's wellbeing (36, 37) and provoke anxiety responses (38). A tentative explanation is that stress comprises physiological and behavioral responses that lead the individual to the evaluation of real or perceived threats (39). At this point, stress proves its relationship with anxiety by sharing cognitive or evaluative factors of a dangerous event. These responses increased during the pandemic; especially, in people who worried about their health (40).

Under this idea, Endler (41) states that there is an interconnection between anxiety and stress by personal variables (vulnerability, heredity, cognitive style, trait anxiety) and interaction with stressful situations (discomfort, pain, trauma) or biological ones indicating that these clinical disorders share neural circuits (39). In this sense, during lockdown, stress and depression were symptoms that presented greater connection and strength of relationship (42–44). Likewise, both symptoms were associated with loneliness during confinement (27); especially in people who were alone, between 18 to 30 years old and with low family income (33). At the same time, anxiety proves to be a clinically relevant disorder during lockdown, but unstable to resample procedures in network analysis (12). Perhaps, this is because individuals present a specific anxiety trait such as worry about being infected with COVID-19 that is associated with stress, anxiety (20) and depression (21–23).

The study has strong implications for public health because the strategies of total confinement implemented worldwide occurred in countries where health systems had serious structural problems and may have collapsed (45). This is the case of Peru, where confinement measures were strict. At the Latin American level, confinement revealed the lack of confidence of governments to contain the virus (46). For instance, in countries where focused confinement was implemented, such as Chile, or only social distancing measures were established, such as Uruguay, there was a better localization of the virus, more trust in citizens and better response of the health system to a health emergency (47). Other variables that influenced the implementation of confinement were governments led by extremist political parties and the absence of scientific criteria in those countries (48). At this point, it was not considered that confinement could cause changes in the behavior of the individual. For example, recent research conducted during confinement is showing that the perception of social isolation is a risk factor for the individual's emotional alteration (49) that may predispose them to depression and feelings of loneliness (50, 51). The latter, especially, has a greater impact on stress-related cognitions and behaviors during pandemic (50).

In this regard, this study aims to: (a) identify symptoms of anxiety, depression, stress, worry about COVID-19 Contagion and fear of loneliness in a sample of people during the time of lockdown decreed by the Peruvian government; (b) identify bridging nodes; and (c) compare symptoms according to staying or leaving home during lockdown.

This study is the first to examine the relationships between anxiety, depression, stress, worry about COVID-19 and fear of loneliness during COVID-19 lockdown in Peru, by a network analysis. The construction of the network of partial correlations allowed examining the connections between symptoms. In this sense, the first objective was to identify the central symptoms in the network. The four nodes with the highest centrality in the network were PHQ2 (“You have felt down, depressed or hopeless”), PHQ7 (“You have had difficulty to concentrate on certain activities, such as reading the newspaper or watching television”); PSS3 (“You have felt nervous or stressed”) and GAD5 (“You have felt so restless that you have not been able to sit still”). The first two belong to the depression scale, while the others belong to the stress and anxiety scales, respectively. These results indicate that depressed mood, concentration difficulties, perception of stress and restlessness are central symptoms in the lockdown experience. Symptoms of loneliness fear were not relevant in the network possibly because, during the lockdown, people were with their families and the fear of loneliness was not active (32). A similar situation could have occurred with Worry about COVID-19 Contagion, that by having a greater number of participants in the sample who did not leave home even once, there is a greater willingness in the participants to develop safety behaviors against contagion (20).

The findings demonstrate that there is an interconnection between depressive feelings and stress (PHQ2, PHQ7, PSS3) in Peru during COVID-19 confinement. These results are consistent with previous studies that establish that depression and stress play an important role in the comparison of two temporal networks during COVID-19 (42). In fact, other studies conducted during confinement have pointed out the relationship between social isolation and affective symptoms. It was demonstrated that the level of perceived isolation became more distressing during confinement when there was an absence of physical activity in the individual (49). This may occur when the individual perceives that their personal resources are being overwhelmed by the stressful event (e.g., confinement). As a result, they make a set of unfavorable evaluations (8) that causes a decrease in the individual's well-being and predisposes them toward depressive symptoms (36, 37). On the other hand, detecting a symptom of anxiety (GAD5) is not strange because Zavlis et al. (12) pointed out that, despite the lack of good stability in resampling procedures, anxiety is a central symptom during the COVID-19 pandemic;. In this study, stability of anxiety symptoms was obtained. The explanation for this discrepancy can be found when applying anxiety tests in non-clinical populations. In this scenario, there is a tendency to low scores and anxiety becomes an adaptive factor (15) that does not destructure the personality of the individual (16). It leads, as well, to consider other factors, such as cognitive styles and psychological vulnerability of the individual (41).

A second objective was to identify bridging nodes; those that connect symptom clusters across clinical disorders. The strongest symptoms were PHQ2 (“You have felt down, depressed or hopeless”) and PHQ7 (“You had had difficulty to concentrate on certain activities, such as reading the newspaper or watching television”), both within the depression scale. These results are related to a previous study that compared depression before and after the peak of the COVID-19 outbreak, where it is highlighted that depressive affect is the central node (51). The results indicate that depressed mood and concentration difficulties are important aspects of depressive disorders. It is evidenced as well that in other health emergencies they have had relevance (5) and their implications are greater when associated with feelings of loneliness. Interestingly, there is a study where concentration problems were observed to emerge as a bridging symptom in seven of thirteen networks while depressive mood emerged in six (71). This suggests that certain symptoms may become transdiagnostic symptoms in clinical disorders (72). Also, it is one of the possible explanations for the increased prevalence of depression during the COVID-19 pandemic (11).

A third objective was to compare symptoms according to staying at home or going out during the lockdown. The results indicate that there are no global differences between the networks and that the adjacent matrices are similar. Despite this, differences are evident in the centrality of the symptom BSFL2 (“The idea of being alone worries me”), where the EI value was higher in the group that did not go out during lockdown. These results are consistent with a previous study which showed that loneliness had a greater impact on stress-related cognitions and behaviors during the pandemic (50). These results could be reasonable because lockdown signified a change in human interactions and a need in people for affiliation or belonging (27, 32). Another symptom with strong variance in centrality was GAD4 (“You have had difficulty to relax”) in favor of those who left home. A tentative explanation is that a possible COVID-19 infection caused difficulties to relax and a situation of constant tension in those who left home. This is due to the excessive care taken to avoid infection, exemplified by measures such as physical distancing and the use of alcohol gel and masks.

The results have important implications for public health because the implementation of a total confinement measure was subject to the avoidance of the collapse of health systems, the lack of government confidence to contain COVID-19 (45–47), governments with extremist parties in office and the lack of science in some countries (48). Thus, this research provides scientific evidence regarding the psychological effects that confinement had on the Peruvian population. In particular, the study demonstrates that the most central symptoms of the network are related to depression. These findings support models that emphasize the role of depression prevention in the framework of health emergencies, because of its implication on suicide rates (5). Also, its symptoms may be maintained even up to a year after the epidemic event occurred (6). In addition, the study suggests to give relevance to the interaction between depression, stress and anxiety through public policies that encourage the promotion of healthy habits. Physical activity, for instance, became a protective factor during confinement (49). In this way, remote health services (e.g., online or smartphone therapy) should be designed to encourage support networks such as the search for institutions, friends or family members that provide empathy and solidarity to the person. This may be consistent when observing that depression was substantially increased by the COVID-19 pandemic (11) and that it may be higher if the individual has feelings of loneliness (27). Secondly, it is worth noting that the role of perceived stress during lockdown has been shown to be directly related to loneliness, anxiety and insomnia (10–12). Also, the experience of being isolated affected the individual's mental health as demonstrated in a study at the beginning of the pandemic where depression, anxiety and perceived stress are relevant (2). In fact, these clinical disorders proved a worsening of people with a preexisting clinical diagnosis or the presence of distress in people without history (1).

Despite the interesting findings, the study has some limitations. Even though the sample size was sufficient under the formula P(P-1)/2, other estimation methods based on simulation (52) might suggest a larger number of participants. It should be noted, though, that collection during the months of the study was quite restricted to non-governmental institutions. In addition, the selection of participants was purposive, which reduces the ability to generalize. However, it was the only way to collect data as the collection was conducted during the lockdown and random sampling was complex. Another limitation is related to the difficulty of being able to remotely follow up with respondents or identify them; aspects that are typical of online data collection (58).

In conclusion, the results indicate that some symptoms of depression, anxiety and stress demonstrate greater interconnectedness in the network. This suggests to focus on depressive symptoms to improve the mental health of people who went through COVID-19 lockdown. The strongest bridging nodes in the network were also identified to be those related to depression In addition, there are no substantial differences between those who went out and did not go out during lockdown, except for node BSFL2 (“The idea of being alone worries me.”). These symptoms may be important to understand how the experience of COVID-19 confinement affected the mental health of Peruvians. According to previous studies, this effect is known to be maintained even up to a year after the health emergency (6). These aspects will be important during the coming years to understand the psychological sequelae of confinement and the COVID-19 pandemic, as they try to forewarn of the responses that should be taken by public health entities in the face of an upcoming pandemic.

The original contributions presented in the study are included in the article/supplementary materials, further inquiries can be directed to the corresponding author.

The studies involving human participants were reviewed and approved by Universidad Privada del Norte Ethics Committee. The patients/participants provided their written informed consent to participate in this study.

JV-L, RL-J, EP, IL-M, and SC-B contributed to conception and design of the study. JV-L organized the database, performed the statistical analysis, and critical review of the latest version of the manuscript prior to submission. RL-J, EP, IL-M, and SC-B wrote the first draft of the manuscript. All authors contributed to manuscript revision, read, and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.