Introduction

The novel coronavirus disease (COVID-19), which originated in China, was declared a public health emergency by the World Health Organization (WHO) on January 30th, 2020 [1]. After a steep global increase in the number of infected persons, different countries took various stringent measures to curb its spread, a nationwide “lockdown” was one such step. The Government of India called for a nationwide lockdown from March 25th, 2020 [2]. Citizens were mandated to stay at home, and all major offices, malls, factories, and schools were shut down for 21 days [2]. The lockdown was further extended until May 3rd, with conditional relaxations [3]. While the lockdown was intended to curb the spread of the virus, it had psychological ramifications on the citizens [4–7]. The self-isolation and restrictions on physical mobility due to the lockdown caused major disruptions to routines in life and hindered the ability to meet regular responsibilities—potentially affecting the physical and mental health of individuals. Recent studies have reported higher levels of stress [8, 9], anxiety [8, 10, 11], depression [10, 12, 13], and poor quality of life [9, 10] during the COVID-19 crisis in different populations. However, the repeated extensions in the lockdown period in India led to longer restrictions on physical mobility and prolonged self-isolation measures. This could have increased the intensity of negative psychological outcomes among Indians, leading to a poorer quality of life not just during the lockdown but also after the crisis. Previous studies have shown that prolonged periods of isolation and limited mobility significantly impact mental well-being [14, 15] during crises. Further, a prolonged experience of negative mental health outcomes could have adverse effects on physical health outcomes such as sleep disorders [16] and health-related quality of life [17].

While mental health effects due to the lockdown are likely to be seen among a majority of Indian adults, the impact of a lockdown extended over longer periods might differ across vulnerable groups [18]. For instance, the stress experienced by persons with limited adaptive capacity, fewer financial resources, lower social support, and pre-existing mental health conditions might be higher than among those who do not share these characteristics. With longer lockdown periods, financially weaker individuals might face more challenges in meeting the basic needs of their families. Moreover, the continued restrictions on physical mobility could place a higher burden on the social networks of vulnerable individuals, thus reducing their access to social support over time and impacting their adaptability. Additionally, in a patriarchal society such as India, with a high prevalence of domestic abuse [19], the lockdown [20] (especially limited mobility) can potentially increase the risk of experiencing domestic abuse. Prolonged exposure to the threat of domestic violence could worsen the mental health of women during this crisis.

Despite these risks, several resources that help with coping could be available to individuals. Previous studies have highlighted the role of social support in reducing anxiety and stress [21, 22]. Recent studies, focused on COVID-19 also support this [11, 23]. In addition to social support, there could also be several individual-level resources such as resilience that could help individuals face adversity [24]. Resilience helps to strengthen mental health and reduce the possibility of developing psychiatric morbidities, especially during the COVID-19 pandemic [25, 26]. There is scant research on the mental health effects of such protective factors during the extended lockdown in India.

Reliably measuring the impact of lockdowns that extend over a long period of time requires a longitudinal study design. While several cross-sectional studies [5–7] have focused on psychological well-being during COVID-19 in India, we could not locate any study investigating the change in such psychological outcomes during the lockdown period. A longitudinal investigation helps establish temporal sequence and document trends while exploring how the adverse psychological outcomes change (if at all) over time during the lockdown. To address this gap in the literature, we conducted a longitudinal study in four phases to investigate the changes in three psychological outcomes viz., anxiety, stress, and depressive symptoms, during the lockdown in India. The following were the aims of our study and our hypotheses:

1. The major aim of our study was to investigate how the levels of anxiety, stress, and depressive symptoms changed during the lockdown among Indian adults, independent of their age, gender, income, educational qualification, place of residence, and history of mental health. We hypothesized that as the period of lockdown increases, the levels of anxiety, stress, and depressive symptoms would also increase over time independent of the covariates.
2. The second aim was to find out whether the changes, if any, in the levels of anxiety, stress, and depressive symptoms, differed across gender. Based on the large body of research highlighting gender disparities in the risk of anxiety, stress, and depression [27], we hypothesized that in a patriarchal society such as India, compared to men, women will have a greater increase in levels of anxiety, stress, and depressive symptoms during the lockdown.
3. Thirdly, we assessed whether the protective factors, social support and resilience, were related to the participants’ anxiety, stress, and depressive symptoms during the lockdown. We hypothesized that greater social support and higher individual resilience will be related to lower levels of anxiety, stress, and depressive symptoms, independent of all covariates.

Methods

Results

Preliminary results

We collected data on psychological outcomes from 159 Indian adults across a period of two months during the lockdown. Our sample comprised relatively young participants (mean age = 27.44 years, SD = 9.17 years). About 65% of the sample were men and the remaining were women. The annual income of nearly half of the participants was below 3,00,000 Indian Rupees (a cut-off representing an income allowing decent living in a one-bedroom apartment for a couple in most urban areas of India). About 55% of the participants were at least college-educated, while only about 11% reported having an educational qualification less than high school. The distribution of the psychological outcomes across these groups is presented in Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. The distribution (mean, standard deviation) of anxiety, stress, and depressive symptoms at times T1, T2, and T4 across gender, annual income, education, and place of residence (N = 159).

https://doi.org/10.1371/journal.pone.0240650.t001

Following the cut-offs suggested by Spitzer et al. [28], 70.8% of the participants reported being mildly anxious, 18.8% stated being moderately anxious, and 10.4% reported severe anxiety symptoms at T1 (early days of the lockdown). At T4 (after eight weeks of lockdown), the prevalence of moderate anxiety and severe anxiety increased to 26% and 12.7%, respectively. Similarly, the prevalence of intermediate stress and high stress [36] increased from 20.1% to 27% and 10.7% to 24%, respectively. Our results also showed that while 14.8% of the participants reported being depressed [32, 37] at T2, the prevalence increased to 26.1% during T4.

Trends in anxiety, stress, and depressive symptoms during the lockdown

Our multilevel models, adjusted for sociodemographic variables showed an increase in anxiety (β = 0.81, CI: 0.03, 1.60) (Table 2) and stress scores (β = 0.51, CI: 0.32, 0.70) (Table 3) during the two months (T1 to T4) of follow-up. We also found an increase in depressive symptoms (β = 0.37, CI: 0.13, 0.60) between T2 and T4, independent of the covariates (Table 4). Anxiety (β = 1.62, CI: 0.44, 2.81) and stress scores (β = 0.65, CI: 0.37, 0.94) were found to be higher among women, versus men, even after accounting for time and sociodemographic factors. However, we could not find statistically significant associations of the other sociodemographic variables (age, annual income, educational qualification, and place of residence) with the psychological outcomes.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Changes (regression coefficients, 95% CI) in anxiety between week 1 and week 8 of the lockdown in India (N = 159).

https://doi.org/10.1371/journal.pone.0240650.t002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Changes (regression coefficients, 95% CI) in stress between week 1 and week 8 of the lockdown in India (N = 159).

https://doi.org/10.1371/journal.pone.0240650.t003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Changes (regression coefficients, 95% CI) in depressive symptoms between week 3 and week 8 of the lockdown in India (N = 159).

https://doi.org/10.1371/journal.pone.0240650.t004

Discussion

Using repeated measures of psychological outcomes from 159 Indian adults during two months of the COVID-19 lockdown, we found that there were statistically significant increases in stress, anxiety, and depressive symptoms over this period. Moreover, this increase in adverse psychological outcomes was found to be greater among women compared to men. We also found that a higher level of an individual’s resilience was related to lower levels of anxiety, stress, and depressive symptoms.

Our findings suggest that anxiety, stress, and depressive symptoms increased during the lockdown among adults in India. While depressive symptoms increased in both the genders, the effect size was modest. Nevertheless, the increase in the adverse psychological outcomes could be because of several reasons. First, the nationwide lockdown disrupted the professional, personal, and social lives of individuals, which potentially impacted their psychological well-being. Moreover, the periodic extensions of the lockdown over a considerably long period, accompanied by a steep increase in the number of COVID-19 cases in the country, and even worldwide, probably worsened their anxiety, stress, and depressive symptoms over time. Each announcement of the extension of the lockdown might have elevated the levels of anxiety by engendering a perception of unpredictability. Further, the initial shocks of the lockdown followed by the social isolation maintained for a prolonged time, combined with the emotional and financial losses incurred during the lockdown, might have created a synergistic psychological impact. Notably, the Government of India announced a relaxation of restrictions on certain activities even while extending the lockdown. These relaxations included the resumption of trains, opening of small shops, and inter-state mobility. However, these graded “unlocks” (which potentially allowed increasing physical mobility as well) were not accompanied by reports of a reduction in the number of new COVID-19 cases in India. On the contrary, the case numbers shot up just as the “unlocking” began, which could have reduced the confidence of the citizens, leading to a higher perceived risk of contracting the disease and further increasing their stress, anxiety, and depressive symptoms. However, these findings of our study are contrary to those found by Wang et al. [38] in China, where stress and anxiety were found to be stable across 4 weeks of lockdown. The explanation provided by Wang et al. highlights that China recorded substantial improvements in curbing the spread of the virus due to their rapid decisive measures and the greater number of recovered patients. This might have instilled greater confidence in their public health measures among the Chinese, thus avoiding a worsening of the psychological toll of a prolonged lockdown [38].

As per the findings of our study, compared to men, women had a greater increase in stress, anxiety, and depressive symptoms during the lockdown. Notably, the increases in anxiety and stress in our sample were primarily due to this evidence of a greater increase in anxiety and stress observed in women. There are two potential explanations for this finding. First, it was found that women in the Indian context have significantly more household responsibilities during the lockdown primarily because of the skewed gendered division of household labor in India [39, 40]. For instance, due to the closure of schools and offices, all family members could be staying indoors, leading to an increase in household burden for women who are expected to shoulder most of the childcare, cooking, cleaning, and other household management. This would leave them with very limited time for themselves. Corroborating this, Viglione [41] reported lower publication rates among female academicians in North America compared to their male counterparts during this pandemic, across all disciplines. Our results from India, a society with stronger patriarchy, supports this narrative. We found that women reported a greater increase in their responsibilities during the lockdown compared to men. These added responsibilities, combined with the lack of time for themselves, could increase their stress, anxiety, and depression levels much more than the increase among men. Second, social isolation and the restrictions on physical mobility might increase the exposure of women to hostility at home, especially among women who were already vulnerable to domestic violence. Previous studies have reported an increase in the risk of women across the world experiencing hostility during the lockdown period [42–44]. The high prevalence (~30%) of domestic violence [19] in India is a reflection of the vulnerability of Indian women to domestic violence (ranging from emotional abuse from family members to physical/sexual abuse from intimate partners). Any risk of such hostility could worsen during the lockdown. Prolonged exposure to any risk of domestic hostility (emotional, physical, or sexual) could lead to an increase in stress, anxiety, and depression among Indian women during the lockdown.

We found that the greater increase in stress and anxiety among women versus men persisted even after accounting for social support and resilience. This suggests that this gendered pattern was strong enough to persist despite any protective effect exerted by these buffering factors. Even though the interaction of depressive symptoms with gender was not statistically significant, it suggested that the rate of increase in depressive symptoms was higher among women than men, thus fitting the pattern observed with anxiety and stress by Ozamiz-Etxebarria et al. [27] in Northern Spain. They found higher levels of anxiety, stress, and depression among men compared to women. This contrast with our results could be because of the cultural differences between the two countries with regard to the gendered division of household labor [45, 46].

We also found that persons with a history of mental health issues were likely to report an increase in anxiety, stress, and depressive symptoms during the period of lockdown. Previous research has highlighted that situations of social avoidance could cause a relapse of trauma and depressive events [47]. The social isolation, added responsibilities, and any lack of perceived social support (due to physical mobility restrictions) could trigger those with past depressive episodes. Studies have shown higher stress and anxiety during the lockdown in India [48].

It is important to implement preventive and therapeutic interventions to address the adverse effects of the pandemic, especially the lockdown. Interventions such as cognitive behavior therapy (CBT) and mindfulness-based therapies (MBTs) could be useful in treating psychiatric symptoms during COVID-19 [25]. A Chinese study highlighted a lower prevalence of psychiatric symptoms after the lockdown because of the psychoneuroimmunity prevention measures administered by the Chinese government, which helped instill a sense of confidence in the people [49]. Many such measures suggested by Tan et al. [49] were also adopted in India. For instance, the Ministry of Family Health and Welfare [50] has widely spread public service messages advising all to wear masks, adopt hygiene practices, and social distancing while also promoting precautionary measures at workplaces.

Conclusions

The COVID-19 crisis and the accompanying lockdown have undoubtedly affected every individual in one way or the other. While the lockdown may help in effectively addressing this pandemic, the state and society at large need to be sensitive to the mental health impacts of a long-drawn-out lockdown. Vulnerable populations such as women and the marginalized deserve immediate attention. However, it behooves society to also address the root causes driving the unequal distribution of psychological distress during such crises.

Acknowledgments

This project was completed with the help and support of a lot of people and we wish to acknowledge their contributions. AG conveys special thanks to Vardhaman Gopal for his timely suggestion, which set the ball rolling, and for being a pillar of support. AG is immensely grateful to her parents for being super helpful and patient during the entire process, particularly for their unwavering support throughout the extended lockdown. AG thanks all her friends and family who not only helped in the circulation of the survey but were also a constant support. AG thanks Sitesh Kumar for his help with numerical data representation. AG, AJS, and MAS thank Nilesh Thube for his help in putting together the tables in the results section. AG, AJS, and MAS are very grateful to Dr. Madhuban Gopal as well as the Writing Studio at IIT Gandhinagar for carefully reviewing the manuscript and suggesting language edits. Finally, we cannot thank the participants enough for their continuous contribution though diligently filling up the forms longitudinally.