Effects of changes in physical and sedentary behaviors on mental health and life satisfaction during the COVID-19 pandemic: Evidence from China

Background While restriction measures are critical in containing the COVID-19 outbreak, limited studies have investigated the behavioral and psychological impact of these measures. This study aimed to investigate the effects of physical and sedentary behavioral changes and online behavior during the COVID-19 pandemic on mental health and life satisfaction among the Chinese population. Methods The data were obtained from a cross-sectional survey of 2145 residents aged between 18 and 80 in Hubei province, China between March 23, 2020, and April 9, 2020. Results Participants who had high frequencies of physical activities before or during the COVID-19 outbreak exhibited higher levels of life satisfaction. Participants who increased their sitting time during the pandemic or kept sitting for more than eight hours before and during the pandemic reported worse mental health than those who maintained less sedentary behavior. Besides, participants who used the Internet for information seeking, communication, and entertainment more frequently reported better mental health and life satisfaction. In contrast, there was a positive association between commercial use of the Internet and symptoms of mental disorders. Conclusion Given the link between physical and sedentary behavioral changes with worse mental wellbeing, strategies to reduce sedentariness and increase physical activity during the COVID-19 pandemic are needed.

study the associations between changes in sitting time and PA during the lockdown period of COVID-19 and online behavior with potential mental disorders." as "we fit multiple logistic regression models to study the associations between changes in sitting time and PA during the lockdown period of COVID-19 and online behavior with symptoms of potential mental disorders." 5. What is meant by depression, page 10 as numbered? I don't see a measure of depression. Is depression used as a synonym for "mental disorders"? Please harmonize the language. If the self-report measure is a measure of mental disorders, then the term mental disorders should be used to refer to findings from this measure.
Response: Thanks for the comment. We used depression as a synonym for mental disorders in the original manuscript. We have harmonized the language in the revised manuscript, particularly in the results and discussion parts. For example, we rewrote the sentence, "kept sitting for more than 8 hours a day at both time points were more likely to develop depressive symptoms" as "kept sitting for more than 8 hours a day at both time points were more likely to experience symptoms of mental disorders" (p.1q).
We rewrote the sentence, "However, changes in the frequency of PA were not associated with depression" as "However, changes in the frequency of PA were not associated with symptoms of mental disorders" (p.11). Besides, we revised the sentence, "Previous studies also found that online shopping may be a negative coping strategy for stress and was associated with increased depression" as "Previous studies also found that online shopping may be a negative coping strategy for stress and was associated with increased symptoms of mental disorders." (p. 13) 6. Please make sure all the dates and details given in the text about the pandemic are correct. It is beyond the scope of the review process to verify this information.
Response: We have double-checked the dates and details about the pandemic and confirmed that the details are correct. 7. Please make sure that your literature review is comprehensive and includes recent research on physical activity during the pandemic, such as this systematic review: Wolf, S., Seiffer, B., Zeibig, J., Welkerling, J., Brokmeier, L., Atrott, B., Ehring, T., & Schuch, F. B. (2021). Is physical activity associated with less depression and anxiety during the COVID-19 pandemic? A rapid systematic review. Sports Medicine (Auckland), 51 (8), 1771-1783. https://doi.org/10.1007/s40279-021-01468-z Response: Thanks for the comment. We have conducted a more comprehensive literature review and added the reference suggested by the Editor and some more recent studies in China in the revised manuscript. The details of the changes are shown on pp. 4-5. "A review of 21 studies about the effect of PA on mental health during the COVID-19 pandemic suggested that people who performed PA regularly with high volume and frequency and kept the PA routines stable would experience fewer symptoms of anxiety and depression [15]." (p. 4) "Although some studies have found that PA participation was positively associated with mental health among the Chinese adults during the COVID-19 pandemic [27,28], they did not specifically examine how the changes in PA and sitting time before and during the COVID-19 pandemic may affect mental health." (p.5) Comments to the Author Reviewer #1: PONE-D-21-27357 1.Abstract The aim of the study is to examine physical and sedentary behaviour changes and online behavior on mental health and life satisfaction -but isn't this regarding a specific time frame of lockdown during COVID (based on your title -"during the covid-19 lockdown")? The abstract does not state that this study is specifically looking at participants behavioural and psychological outcomes during lockdown. Please clarify.
(can't leave their homes) vs non-lockdown periods since people can leave their homes. Please specific what time frame you are looking at and why this time frame (lockdown only or both lockdown and non-lockdown periods, etc.). Also, make sure you clearly state this time frame throughout the manuscript. Also, this seems to be looking at the first lockdown which is also important to state.
Response: Thanks for the comment. We would like to clarify the timeline of the lockdown of Hubei province. Chinese authorities imposed a lockdown on Wuhan, the capital city of Hubei province, on January 23, 2020, and enforced similar measures in 16 neighboring cities in Hubei province, affecting approximately 57 million people (Xiong, 2020). The authorities relaxed the Hubei lockdown on March 23, 2020, and officially lifted the Wuhan lockdown on April 8, 2020, after no new deaths transpired for the first time (He, 2020). The study was conducted between March 23, 2020, and April 9, 2020, when the lockdowns in Hubei were relaxed and eventually lifted. We asked the participants to report their physical activity participation and sitting time in the past two weeks and the past month, respectively, which may be during the lockdown or non-lockdown periods. Thus, we thought it would be better to use "COVID-19 pandemic" instead of "COVID-19 lockdown". We have used such a term throughout the manuscript. 6."the data collection was carried out between March 23 and April 9, 2020, while Hubei's lockdowns were gradually eased. Such data creates a unique opportunity for us to investigate Hubei residents' behavioral changes and psychological wellbeing during the lockdown period." -based on this sentence, my impression is you collected data during a non-lockdown period for participants to retrospectively report their behavioural and psychological outcomes during a lockdown (from Jan-Mar?), is this correct? You are specifically looking at lockdown effects. Why not collect data during the lockdown if its online anyways? Why wait until it eased when you are interested in lockdown effects? Please clarify.
Response: In the revised manuscript, we did not emphasize that we are looking at the lockdown period of the pandemic. We provided the reason in response to the reviewer's comment #5. 7.Page 6 -Please specify the specific "online behaviors" you plan to examine in this study.
Response: Thanks for the suggestion. We have specified the "online behaviors" in the revised manuscript (p.6). The corresponding changes are: "Besides, we examined the impact of different types of Internet use, including information seeking, work/study, buying products, entertainment, reading/online learning, communication with family and friends, and investment/financial management."

8.Methods
How did the participants get recruited? Any incentives for them to complete the survey?
Response: We have provided more information about subject recruitment in the revised manuscript (p.7).
"The online survey was carried out on a professional survey platform in China. The platform sent a notification to respondents in their sample bank with a link to allow access to the questionnaire. Only Hubei residents (with IP address locations in Hubei) can answer the questionnaire. No incentives were provided to the participants." 9.Page 8 -for online behavior, what do you mean communication (talking with family members via zoom?). Please provide an example. Also, you looked at 7 internet uses which mostly seem like positive examples (based on your introduction) -you mention not having a hypothesis, but you only looked at internet uses that could enhance mental health; thus I would hypothesis internet use will have positive effects, correct? That said, why not include the negative internet uses you mentioned in the introduction such as gambling, social media use, shopping? Also, im assuming "buying products" is referring to "shop online that regularizes their routines". Please provide an example of buying products.
Response: Thanks for the questions. To clarify, communication in this study was measured by asking the respondents whether they communicated with family and friends via online communication tools (e.g., WeChat). WeChat is the most widely used communication app in China, which acted as critical infrastructure in fulfilling people's practical, emotional, and medical needs during the COVID lockdown in China (Qian & Hanser, 2020). We have clarified this in the revised manuscript as below.
As the reviewer suggested, we have included an example for buying products in the revised manuscript as "buying products (e.g., groceries)" (p. 9).
10.Page 8 -Did you explain what "moderate to vigorous physical activity" means to the participant in the survey (e.g., moderate to vigorous physical activity means any activity that your heart rate increases and you are sweating) or leave it to the participant to interpret it? For example, someone may interpret physical activity as anything active (e.g., doing daily chores) within a day while someone else may interpret specific working out activities (e.g., running for an hour). As you can imagine, a daily chore may not be as "active" as someone who runs for an hour. You mention in the discussion that future studies should look at type of physical activity, which I agree is very important, but curious how much information was given to the participants regarding this question.
Response: We asked the participants about the frequency of doing moderate-tovigorous physical activity (MVPA), followed by a brief explanation and some examples of MVPA. In the survey, MVPA refers to exercises such as running, playing ball, swimming, dancing, brisk walking, etc. for ≥10 minutes. Typically, physical activity needs to continue for at least 10 minutes to be considered a session of exercise. We have clarified the measure of "moderate to vigorous physical activity" in the revised manuscript as below: "Participants reported the frequency of doing moderate-to-vigorous physical activity (MVPA) in 2019 and in the past month, followed by a brief explanation and some examples of MVPA. MVPA refers to exercises such as running, playing ball, swimming, dancing, brisk walking, etc., for ≥10 minutes, which cause breathing and heart rate to increase." (p.8) 11.Page 8 -"information seeking" -is this specific COVID information seeking? Or other types of information seeking? How did you word this question?
Response: Thanks for the question. The information seeking was measured by general information seeking behavior, not specific COVID-19 information seeking. We have clarified this in the revised manuscript (p.9). "Seven main and conceptually different Internet uses were accessed, including general information seeking;"

12.Results
You have most of the information from Table 1 and 2 in text (e.g., "half of the respondents were female (51.28%)."). Thus, Tables 1 and 2 are redundant. I would put  all the information in text and remove Tables 1 and 2. Please include the sample size for each variable (e.g., half of the respondents were female (n = X; 51.28%)). Or keep Table 1 and 2 but remove the text because it is redundant. If you keep the tables -add the sample size for each variable and separate the Mean (SD) data into one column and the % data into another column.
Response: Thanks for the suggestions. To be consistent with the reporting style in most studies, we decided to keep Tables 1 and 2. As the reviewer suggested, we have removed the statistics and only kept a very brief description of PA, sitting time and online behavior in the revised manuscript.
13.Page 10 -The "M =" are missing for online class, work and study, buying products, investment and financial management.
Response: Thanks for pointing this out. However, as the reviewer suggested in Comment #12, we have deleted these statistics as they are redundant. -since you have table 3 with all the test statistics, you don't need to have the statistics in text. You can remove the test statistic information and just explain the findings, as you have done (i.e., remove the numbers from this section that are in Table 3).

14.Regression results
Response: Thanks for the suggestion. We have removed the statistics in the text. 15.Most participants completed the survey alone online, but some participants completed via phone survey. Any differences between the style of data collection (e.g., participants could ask for clarification during a phone survey from the research assistant vs participants who completed the survey alone on a computer). Were research assistants asked to not provide any additional information but simply read the questions to the participants over the phone? How many completed the survey via phone vs online?
Response: Thanks for the questions. We have clarified the two modes of data collection in the revised manuscript as below: "There are minimal differences between the two modes of data collection. Our research staff will be asked not to provide any additional information but simply read the questions to the participants." (p.7) We also included the number of participants who completed the survey online or via telephone in the revised manuscript.
"After deleting cases with at least one missing value, the final sample included 2145 respondents, of which 1944 (90.63%) completed the online survey, and 201 (9.37%) completed the phone survey." (p.7) 16.Discussion Page 13 -I would not assume communication referring to use of social media, unless the question was framed this way. I would assume communication meaning communicating with others via phone or zoom/skype etc. When you clarify what "communication" means in the methods, this may make sense though.
Response: Thanks for the question. We have clarified this measurement in the revised manuscript.
"communication with family and friends via online communication tools (e.g., WeChat)" 17.I believe it is important to state that this study is looking at a lockdown period which is very different from normal day to day life (pre-and post-COVID). Something that was good during a lockdown (e.g., can't work so they play video games all day for social interaction and something to do), may not be good when the lockdown is lifted (e.g., playing video games but should be at work). Thus, these findings are only relevant when a city/town are under lockdown.
Response: Thanks again for the comment. As we responded to the reviewer earlier, we would not emphasize that we are looking at the lockdown effects. We use "during the COVID-19 pandemic" throughout the manuscript.
18.Page 13 -please clarify what shopping means? Shopping for non-essentials (e.g., equipment) vs essentials (e.g., food)? Or does online shopping have a negative effect even when shopping for essentials (i.e., any type of shopping)? Though, I thought online shopping for essentials could give the impression of regularize daily routines which would be good for mental health? Please clarify.
Response: Thanks for the comment. We did not specify types of shopping in the survey. It could be shopping for essentials or non-essentials. 19.Page 14 -another potential limitation -participants had to retrospectively report their physical activity and sitting time prior to COVID (especially 2019 MVPA rates) which could be inaccurate since it was based on memory (especially among older adults).
Response: Thanks for pointing this out. We have added the reporting bias in the limitation of the revised manuscript. The corresponding revisions are shown below: "Moreover, we asked participants to retrospectively report their physical activity and sitting time prior to the COVID-19 pandemic, which may be subject to recall bias due to inaccurate memory. Future studies may use objective or direct measures of physical activity to increase precision and validate the self-report measures." (p. 15 Dear Editor and reviewer, Thank you very much for your valuable comments on our original manuscript. We have addressed all the comments raised by the reviewer and the Editor. Accordingly, we made extensive revisions to the manuscript based on the comments. We have submitted the revised manuscript highlighting changes in yellow and a clean version. We responded to each comment. Please feel free to contact us if you have further comments.
Thank you for your time!

Introduction Mental health during the COVID-19 pandemic among the general population in China
The COVID-19 pandemic has heavily impacted many societies and economies. The total confirmed infection cases had reached over 483 million, and the death toll hit more than 6.1 million by . [1] To control the spreading of COVID-19, governments worldwide have implemented various restriction measures ranging from social distancing to societal shutdown. [2] While these measures are critical in containing the outbreak, they may compromise individuals' mental health and health activities.

Mental health consequences of physical activity, sedentary behavior, and online activities
Numerous studies have shown that the COVID-19-related restriction measures disrupted daily routines and impacted lifestyle activities, including participation in sports and physical activity (PA) [9] [10]. The decline in PA was accompanied by increased sedentary (sitting) behavior [11]. Given the consistent benefits of regular PA for mental health [12] [13] [14], reductions in PA are likely to exacerbate COVID-19-fueled psychological distress. A review of 21 studies about the effect of PA on mental health during the CCOVID-19 pandemic suggested that people who performed PA regularly with high volume and frequency and kept the PA routines stable would experience fewer symptoms of anxiety and depression [15].
Besides, accumulating evidence has revealed that, independent of PA levels, sedentary behavior is associated with mental health [16] [17]. Several longitudinal investigations have shown that selfreported sedentary time predicts future mental health and wellbeing [18] [19]. A longitudinal study of populations in France and Switzerland found that sedentary behavior increased during the lockdown period of COVID-19, which was associated with poorer physical health, mental health, and subjective vitality [20]. Increasing sitting time during the COVID-19 pandemic may be a risk factor for mental health and life satisfaction.
Moreover, COVID-19-related measures such as closing down schools, workplaces, and entertainment outlets coupled with self-quarantine have led to an unprecedented level of Internet and digital technology use. People move almost every aspect of their daily activities to cyberspace at an accelerating speed, as a "new normal" in the COVID-19 era. However, it remains unclear whether and how Internet usage would exacerbate or alleviate COVID-19-fueled psychological distress. On the one hand, the Internet could enhance mental health by providing access to the latest information and enabling people to work, study, and shop online that regularizes their routines during quarantines. On the other hand, excessive engagement in specific online activities such as gambling, gaming, social media use, and shopping may result in severe problems and elevate the risk of disordered or addictive use of the Internet [21] [22]. Given the mixed evidence regarding the effects of Internet usage on mental health, more studies are warranted to investigate the impact of Internet usage on mental wellbeing during the COVID-19 pandemic.
To date, limited studies have examined the associations between changes in PA, sedentary behavior, and online behavior with mental health and quality of life among the Chinese population during the COVID-19 pandemic. A study of 4,898 adolescents in China showed that higher PA levels were significantly related to lower negative mood scores and higher positive mood scores during the COVID-19 pandemic [23]. More engagement in physical activities was also associated with reduced anxiety and depression among Chinese college students after adjusting for confounding demographic factors [24,25]. While providing important insights, these studies often focused on subpopulations (e.g., adolescents, college students), and thus their findings cannot be generalized to the general population.
A study of the Chinese adult population found that compared with that before the lockdown, their time engaged in daily PA decreased and sedentary time increased, and PA is positively associated with life satisfaction [26]. However, the authors did not examine the effect of COVID-19-induced behavioral changes on adverse mental health outcomes. Although some studies have found that PA participation was positively associated with mental health among the Chinese adults during the COVID-19 pandemic [27,28], they did not specifically examine how the changes in PA and sitting time before and during the COVID-19 pandemic may affect mental health. Moreover, virtually no research has investigated how different types of online behavior during the pandemic may affect the Chinese population's psychological wellbeing during the COVID-19 pandemic.

The present study
Given the lack of research on the effects of COVID-19-induced behavioral changes on mental wellbeing among the Chinese population, this study aims to investigate the impacts of behavioral changes in PA, sedentary behavior, and online behavior on individual psychological wellbeing and life satisfaction among residents in Hubei province, China, the original epicenter of the pandemic. The data collection was carried out between March 23, 2020, and April 9, 2020, while Hubei's lockdowns were gradually eased. Such data creates a unique opportunity for us to investigate Hubei residents' behavioral changes and psychological wellbeing during the pandemic. The findings of this study may improve our understanding and inform public health policies regarding the pattern and consequences of behavioral changes during the COVID-19 pandemic. Based on previous literature, we hypothesized that (1)

Data
The data came from the "Public Attitude toward The Novel Coronavirus Epidemic in Hubei Province" survey, which was carried out by the China Academy of Science and Technology Development Strategy, the Social Policy Research Institute at Renmin University, and the Institute of Sociology of the Chinese Academy of Social Sciences. The survey was conducted between March 23, 2020, and April 9, 2020.
Authorities imposed a strict lockdown policy for cities in Hubei since January 23, 2020, and started to ease the lockdown since March 23, 2020, with the lockdown in Wuhan officially lifted on April 8, 2020.
The survey targeted all the residents aged between 18 and 80 in the urban and rural areas of Hubei. It was a combined online and phone survey. The online survey was conducted on a professional survey platform in China. The platform sent a notification to respondents in their sample bank with a link to allow access to the questionnaire. Only Hubei residents (with IP address locations in Hubei) can answer the questionnaire. No incentives were provided to the participants. A supplemented phone survey was carried out by trained research assistants to reach populations with limited access to the Internet (e.g., aged over 65, rural population). There are minimal differences between the two modes of data collection.
Our research staff will be asked not to provide any additional information but simply read the questions to the participants. Participants who did the online survey provided written consent. Participants who answered the questions via phone gave oral consent, and the interviewer signed a form pledging that he/she had gone through the proper procedures to obtain the verbal informed consent of the participant.

Measurements
Symptoms of mental disorders were measured by the 12-item Chinese Health Questionnaire scale (CHQ-12), which was widely used for screening non-psychotic mental health problems in the general population. Developed by , CHQ was adapted from the General Health Questionnaire [29] with the inclusion of culturally relevant items in Chinese societies. CHQ-12 has been validated in the general population in mainland China [30] [31], Taiwan [32] [33], and Chinese people in foreign countries [34] [35]. Ten items of CHQ-12 were negatively worded (e.g., "Been suffering from headache or pressure in your head") and two items were positively worded (e.g., "Have you been getting along well with your family or friends recently?"). The responses for all items ranged from 0 = "not at all" to 3 = "very often." The responses for positive ones were reversely coded so that higher values indicated higher levels of mental disorder. The internal validity of the scale was good (Chronbach's alpha = 0.78). In this study, the cut-off point against the diagnosis of non-psychotic disorders was 13/14. About 32.07% of the participants scored greater than or equal to 14 and could be regarded as having symptoms consistent with potential mental disorders. MVPA refers to exercises such as running, playing ball, swimming, dancing, brisk walking, etc., for ≥10 minutes, which cause breathing and heart rate to increase. The response categories ranged from "never" to "every day." Exercise training guidelines for a quarantine situation due to  suggested that healthy or asymptomatic persons should do aerobic exercises for at least three sessions per week (four or five seem optimal) [41]. Participants were categorized as "maintaining high PA" if they did PA > two times a week, as "decreased PA" if they did PA for > two times a week before the COVID-19 pandemic but did not do so afterward, as "increased PA" if they increased PA to > two times a week during the pandemic, and as "maintaining low PA" if they did PA for less than two times a week before and during the pandemic.
Sitting time. Participants reported average daily time spent sitting before (i.e., in early January 2020) and during the COVID-19 pandemic. Previous studies suggested that the association of daily sitting time and most long-term health outcomes is not linear, with a weekday sitting time below 8 hours/day associated with better perceived mental health and quality of life [40]. Participants were then classified as "maintaining low sitting time" if they reported ≤8h/day of sitting at both time points, as "increasing sitting time" if they reported ≤ 8h/day of sitting before the pandemic but reported > 8h/day of sitting after the pandemic, as "decreasing sitting time" if they reported > 8h/day of sitting before the pandemic but reported ≤ 8h/day of sitting after the pandemic, and as "maintaining high sitting time" if they reported > 8h/day of sitting at both time points.
Online behavior. Seven main and conceptually different types of Internet uses were accessed, including general information seeking; work; buying products (e.g., groceries); entertainment (e.g., playing online games, watching videos, listening to music); reading books/online class; communication with family and friends via online communication tools (e.g., WeChat); and investment and financial management.
The responses ranged from 0 = "never" to 4 = "everyday". Higher values indicated a higher frequency of using the Internet for the corresponding activity.
In addition, a series of sociodemographic variables were measured, including age (continuous), sex (male vs. female), education (middle school or below, high school, college or above), occupation (managerial/professional position, manual/service/part-time worker, or other), China Communist Party membership (Party member vs. non-Party member), monthly personal income (from no income to >8000 yuan), area (rural vs. urban), city (Wuhan vs. non-Wuhan).

Analysis
All the analyses were performed using Stata 16.0. Descriptive statistics were used to characterize the study population. Considering the variable of potential mental disorder was binary, we fit multiple logistic regression models to study the associations between changes in sitting time and PA during the COVID-19 pandemic and online behavior with symptoms of potential mental disorders. For the outcome variable of life satisfaction, we used OLS regression models to examine the effect of sedentary activity, PA, and online behavior after adjusting for sociodemographic variables. Variance inflation factors (VIFs) of the independent variables were estimated to check whether multicollinearity exists in the models. All the VIFs were below 2, suggesting that multicollinearity was not a significant concern.  Besides, our findings indicated that the overall time spent in sedentary behavior is positively associated with symptoms of mental disorders. Sedentary behavior could be associated with mental disorders via several possible pathways. For example, screen-based sedentary behaviors (e.g., TV viewing) are likely to induce addiction [47], internalizing symptoms [48], and low sleeping quality [49] , potentially leading to heightened levels of mental distress. Further, engaging in sedentary behaviors may displace time spent in other activities such as household or work-related responsibilities or PA, which may be effective coping strategies for symptoms of mental disorders [50]. It is worth noting that specific sedentary behaviors may contribute to mental health in different ways. For example, a study found that specific screen-based recreation was associated with higher self-esteem among Australian adolescents in low-income communities [51]. Thus, future studies may investigate associations between different types of sedentary behavior and mental health.

Results
Our findings showed that various Internet use types had different effects on mental health and life satisfaction during the COVID-19 pandemic among Hubei residents. To start with, online informationseeking behavior was associated with a lower likelihood of mental disorders and increased life satisfaction. Such findings seemed inconsistent with other studies showing that exposure to COVID-19-related information may increase psychological distress [52] [53]. The discrepancy between our study and others may be that we measured the general information-seeking behavior rather than COVID-19 specific information seeking. Future studies may further investigate whether informationseeking behavior in general and about COVID-19 has differential effects on psychological wellbeing.
Also, online communication activity played an important role in improving mental health and life satisfaction, which appeared contradictory to prior studies showing that online communication (especially the use of social media) may deteriorate mental health due to its negative impact on face-toface communications [54] [55]. Considering offline interactions were restricted by physical distancing and stay-at-home orders during the COVID-19 pandemic, online contact with friends and family may increase emotional support, decrease loneliness and social isolation, which leads to lower depression and improved life satisfaction. Besides, our results have shown that online entertainment can reduce symptoms consistent with potential mental disorders. Using the Internet to relax may be a positive coping strategy to alleviate feelings of stress and isolation [56]. Similarly, working and learning remotely during the pandemic may keep people connected with their co-workers and classmates, despite physical isolation, and thus were positively associated with life satisfaction. Using the Internet for work and study may also help regularize daily routines, which is essential for buffering the adverse impact of stress exposure during a crisis [57].
While respondents who used the Internet for information seeking, communication, study/work, reading books, and entertainment reported better mental health and life satisfaction, the commercial use of the Internet (e.g., shopping, investment) seemed to increase mental distress. We conjecture that the economic downturn and uncertainty caused by the pandemic may be why people who were involved in more online economic activities suffered more distress. Previous studies also found that online shopping may be a negative coping strategy for stress and was associated with increased symptoms of mental disorders [58] [59]. While online purchases were the dominant way for people to get groceries due to social distancing and stay-at-home orders, scholars have suggested setting a specific time and financial limits for online shopping to reduce the risk of disordered or addictive use [60].
Despite the significant findings, this study is not without limitations. First, due to the cross-sectional design employed in the present study, it cannot ascertain whether mental health problems lead to lower PA levels or whether lower levels of PA lead to mental health problems during the COVID-19 pandemic.
Additional longitudinal studies are needed to infer the direction of the association. Second, our samples are not representative since it was mainly conducted online. However, considering in-person surveys would not be appropriate or possible during the COVID-19 pandemic, online surveys are arguably the most feasible way to collect data in a timely matter. We have also supplemented the online survey with a telephone survey to reach groups with limited access to the Internet. Future studies may try various sampling methods to gain a population-based representative sample. Third, participants were asked to self-report their PA level, thus introducing self-reporting bias into the findings. Moreover, we asked participants to retrospectively report their physical activity and sitting time prior to the COVID-19 pandemic, which may be subject to recall bias due to inaccurate memory. Future studies may use objective or direct measures of physical activity to increase precision and validate the self-report measures. Lastly, due to the online survey's length limitation, we only measured the PA frequency and the total amount of sedentary behavior during the pandemic. Future studies may further specify the level and type of PA and sedentary behavior and their relationship with mental health and life satisfaction.

Conclusion
In conclusion, social distancing and quarantine measures to mitigate the COVID-19 pandemic may affect PA, mental health, and life satisfaction among the general population, with those experiencing a decrease in PA having lower life satisfaction, and those sitting more time reporting a higher likelihood of mental disorders. Besides, the increasing use of the Internet and digital technology throughout the pandemic also impacts individuals' mental health and life satisfaction. Our findings showed that participants who used the Internet for information seeking, communication, study/work, reading books, and entertainment more frequently reported better mental health and life satisfaction. In contrast, the commercial use of the Internet (e.g., shopping, investment) seemed to increase mental distress. This work was supported by the National Social Science Fund of China (grant no. 16ZDA079). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

behavior, and online activities
Numerous studies have shown that the COVID-19-related restriction measures disrupted daily routines and impacted lifestyle activities, including participation in sports and physical activity (PA) [9] [10]. The decline in PA was accompanied by increased sedentary (sitting) behavior [11]. Given the consistent benefits of regular PA for mental health [12] [13] [14], reductions in PA are likely to exacerbate COVID-19-fueled psychological distress. A study of the general population in the UK revealed a negative association between moderate-to-vigorous PA (MVPA) and depressive or anxious symptoms during the COVID-19 pandemic [15]. Similar findings were found among adults in the US [16], Austria [17], and Italy [18].  aged over 65, rural population). There are minimal differences between the two modes of data collection.
Our research staff will be asked not to provide any additional information but simply read the questions to the participants. Participants who did the online survey provided written consent. Participants who answered the questions via phone gave oral consent, and the interviewer signed a form pledging that he/she had gone through the proper procedures to obtain the verbal informed consent of the participant.

Measurements
Symptoms of mental disorders were measured by the 12-item Chinese Health Questionnaire scale (CHQ-12), which was widely used for screening non-psychotic mental health problems in the general population. Developed by , CHQ was adapted from the General Health Questionnaire [29] with the inclusion of culturally relevant items in Chinese societies. CHQ-12 has been validated in the general population in mainland China [30] [31], Taiwan [32] [33], and Chinese people in foreign countries [34] [35]. Ten items of CHQ-12 were negatively worded (e.g., "Been suffering from headache or pressure in your head") and two items were positively worded (e.g., "Have you been getting along well with your family or friends recently?"). The responses for all items ranged from 0 = "not at all" to 3 = "very often." The responses for positive ones were reversely coded so that higher values indicated higher levels of mental disorder. The internal validity of the scale was good (Chronbach's alpha = 0.78). In this study, the cut-off point against the diagnosis of non-psychotic disorders was 13/14.
About 32.07% of the participants scored greater than or equal to 14 and could be regarded as having symptoms consistent with potential mental disorders. MVPA refers to exercises such as running, playing ball, swimming, dancing, brisk walking, etc., for ≥10 minutes, which cause breathing and heart rate to increase. The response categories ranged from "never" to "every day." Exercise training guidelines for a quarantine situation due to COVID-19 suggested that healthy or asymptomatic persons should do aerobic exercises for at least three sessions per week (four or five seem optimal) [41]. Participants were categorized as "maintaining high PA" if they did PA > two times a week, as "decreased PA" if they did PA for > two times a week before the COVID-19 pandemic but did not do so afterward, as "increased PA" if they increased PA to > two times a week during the pandemic, and as "maintaining low PA" if they did PA for less than two times a week before and during the pandemic.
Sitting time. Participants reported average daily time spent sitting before (i.e., in early January 2020) and during the COVID-19 pandemic. Previous studies suggested that the association of daily sitting time and most long-term health outcomes is not linear, with a weekday sitting time below 8 hours/day associated with better perceived mental health and quality of life [40]. Participants were then classified as "maintaining low sitting time" if they reported ≤8h/day of sitting at both time points, as "increasing sitting time" if they reported ≤ 8h/day of sitting before the pandemic but reported > 8h/day of sitting after the pandemic, as "decreasing sitting time" if they reported > 8h/day of sitting before the pandemic but reported ≤ 8h/day of sitting after the pandemic, and as "maintaining high sitting time" if they reported > 8h/day of sitting at both time points.
Online behavior. Seven main and conceptually different types of Internet uses were accessed, including general information seeking; work; buying products (e.g., groceries); entertainment (e.g., playing online games, watching videos, listening to music); reading books/online class; communication with family and friends via online communication tools (e.g., WeChat); and investment and financial management.
The responses ranged from 0 = "never" to 4 = "everyday". Higher values indicated a higher frequency of using the Internet for the corresponding activity.
In addition, a series of sociodemographic variables were measured, including age (continuous), sex (male vs. female), education (middle school or below, high school, college or above), occupation (managerial/professional position, manual/service/part-time worker, or other), China Communist Party membership (Party member vs. non-Party member), monthly personal income (from no income to >8000 yuan), area (rural vs. urban), city (Wuhan vs. non-Wuhan).

Analysis
All the analyses were performed using Stata 16.0. Descriptive statistics were used to characterize the study population. Considering the variable of potential mental disorder was binary, we fit multiple logistic regression models to study the associations between changes in sitting time and PA during the COVID-19 pandemic and online behavior with symptoms of potential mental disorders. For the outcome variable of life satisfaction, we used OLS regression models to examine the effect of sedentary activity, PA, and online behavior after adjusting for sociodemographic variables. Variance inflation factors (VIFs) of the independent variables were estimated to check whether multicollinearity exists in the models. All the VIFs were below 2, suggesting that multicollinearity was not a significant concern.    Besides, our findings indicated that the overall time spent in sedentary behavior is positively associated with symptoms of mental disorders. Sedentary behavior could be associated with mental disorders via several possible pathways. For example, screen-based sedentary behaviors (e.g., TV viewing) are likely to induce addiction [47], internalizing symptoms [48], and low sleeping quality [49] , potentially leading to heightened levels of mental distress. Further, engaging in sedentary behaviors may displace time spent in other activities such as household or work-related responsibilities or PA, which may be effective coping strategies for symptoms of mental disorders [50]. It is worth noting that specific sedentary behaviors may contribute to mental health in different ways. For example, a study found that specific screen-based recreation was associated with higher self-esteem among Australian adolescents in low-income communities [51].  [56]. Similarly, working and learning remotely during the pandemic may keep people connected with their co-workers and classmates, despite physical isolation, and thus were positively associated with life satisfaction. Using the Internet for work and study may also help regularize daily routines, which is essential for buffering the adverse impact of stress exposure during a crisis [57].

Descriptive statistics
While respondents who used the Internet for information seeking, communication, study/work, reading books, and entertainment reported better mental health and life satisfaction, the commercial use of the Internet (e.g., shopping, investment) seemed to increase mental distress. We conjecture that the economic downturn and uncertainty caused by the pandemic may be why people who were involved in more online economic activities suffered more distress. Previous studies also found that online shopping may be a negative coping strategy for stress and was associated with increased symptoms of mental disorders [58] [59]. While online purchases were the dominant way for people to get groceries due to social distancing and stay-at-home orders, scholars have suggested setting a specific time and financial limits for online shopping to reduce the risk of disordered or addictive use [60].
Despite the significant findings, this study is not without limitations.

Response Letter
Editor's Comment 1.Please clarify the reasons for the restrictions on data availability. Stating "data available upon reasonable request from the corresponding author" is not sufficient. Please explain your exceptional situation.

Response:
We have clarified the reasons for the restrictions on data availability, as shown below.
"Data cannot be shared publicly because of data protection regulation. Data are available from the Chinese Social Quality Data Archive for researchers who meet the criteria for access to confidential data. The data are available for research upon reasonable request and with permission from the Chinese Social Quality Data Archive: http://csqr.cass.cn/index.jsp" 2. This is a cross-sectional study using survey methodology. Your limitations section (pages [13][14] does a very good job of conveying this material about the cross-sectional nature of the design. As you note, this type of design does not allow statements about one variable causing a change in another variable. For this reason, statements such as "the commercial use of the internet seemed to increase mental disorders" (in the abstract) should be phrased as "there was an association between commercial use of the internet and symptoms of mental disorders." (Note comment 4 below that mental disorders per se were not assessed; rather, a self-report measure of symptoms of mental disorders was completed). Likewise, the statement "we hypothesized that (1) the lockdown may cause a decrease in PA and an increase in sedentary behavior" (page 6, as numbered) is not appropriate for this design. Please rewrite as "….the lockdown may be associated with changes in PA and sedentary behavior." More generally, throughout the writing, please edit to make sure the language used reflects the cross-sectional design and does not suggest causality or a longitudinal design.
Response: Thanks for your comment. We have revised the manuscript accordingly, using language that reflects the cross-sectional design. We detailed the changes below.
(1) We have rephrased the sentence "the commercial use of the internet seemed to increase mental disorders" (in the Abstract) as "there was a positive association between commercial use of the internet and symptoms of mental disorders." (2) As the editor suggested, we have revised the first hypothesis as "the COVID-19 pandemic may be associated with changes in PA and sedentary behavior." (p.6) (3) We have revised the second hypothesis, "a decrease in PA may deteriorate mental health and life satisfaction," as "a decrease in PA may be negatively associated with mental health and life satisfaction." (p.6) 3. Please clarify the consent process for the participants who went through a phone survey. Did participants who answered the questions via phone give oral consent, and those who did the online survey provide written consent?

Response:
We have added information about the consent process in the revised manuscript (p.7).

Response to Reviewers
"Participants who did the online survey provided written consent. Participants who answered the questions via phone gave oral consent, and the interviewer signed on a form pledging that he/she had gone through the proper procedures to obtain the verbal informed consent of the participant." 4. The self-report measure used does not permit a true diagnosis of mental disorders. I recommend using language that indicates that the measure reports on symptoms consistent with a potential mental disorder.
Response: Thanks for the comment. We have revised the manuscript using language that indicates that the measure reports on symptoms consistent with a potential mental disorder. For example, (1) We rewrote the sentence "Mental health was measured by the 12-item Chinese Health Questionnaire scale (CHQ-12)" as "Symptoms of mental disorders were measured by the 12-item Chinese Health Questionnaire scale (CHQ-12)". (p.7) (2) We have rephrased the sentence, "About 32.07% of the participants scored greater than or equal to 14 and could be regarded as having potential mental disorders." as "About 32.07% of the participants scored greater than or equal to 14 and could be regarded as having symptoms consistent with a potential mental disorder." (p.8) (3) We have rephrased the sentence, "we fit multiple logistic regression models to study the associations between changes in sitting time and PA during the lockdown period of COVID-19 and online behavior with potential mental disorders." as "we fit multiple logistic regression models to study the associations between changes in sitting time and PA during the lockdown period of COVID-19 and online behavior with symptoms of potential mental disorders." 5. What is meant by depression, page 10 as numbered? I don't see a measure of depression. Is depression used as a synonym for "mental disorders"? Please harmonize the language. If the selfreport measure is a measure of mental disorders, then the term mental disorders should be used to refer to findings from this measure.
Response: Thanks for the comment. We used depression as a synonym for mental disorders in the original manuscript. We have harmonized the language in the revised manuscript, particularly in the results and discussion parts. For example, we rewrote the sentence, "kept sitting for more than 8 hours a day at both time points were more likely to develop depressive symptoms" as "kept sitting for more than 8 hours a day at both time points were more likely to experience symptoms of mental disorders" (p.1q).
We rewrote the sentence, "However, changes in the frequency of PA were not associated with depression" as "However, changes in the frequency of PA were not associated with symptoms of mental disorders" (p.11). Besides, we revised the sentence, "Previous studies also found that online shopping may be a negative coping strategy for stress and was associated with increased depression" as "Previous studies also found that online shopping may be a negative coping strategy for stress and was associated with increased symptoms of mental disorders." (p.13) 6. Please make sure all the dates and details given in the text about the pandemic are correct. It is beyond the scope of the review process to verify this information.

Response:
We have double-checked the dates and details about the pandemic and confirmed that the details are correct.