Immediate Impact of the COVID-19 Conﬁnement on Physical Activity Levels in Spanish Adults

: Enacted measures to control the spread of COVID-19 disease such as compulsory conﬁnement may inﬂuence health behaviors. The present study investigated changes in physical activity (PA) levels during the ﬁrst days of conﬁnement. Using an online survey, the Spanish population (n = 2042, 54% women, age 35.9 (SD 13.6) years) replied to questions concerning sociodemographic characteristics as well as PA behavior before and during the ﬁrst week of enacted isolation. Physical activity vital sign (PAVS) short form was used to estimate weekly minutes of PA before and during the isolation period. Statistical analysis used the following tests: Mc Nemar Chi-squared tests, independent and paired samples t -test, and e ﬀ ect size (Cohen’s d). During the ﬁrst week of conﬁnement, participants reduced their weekly PA levels by 20% (~45.2 weekly minutes (95% CI: 37.4 − 53.0)). This led to a decrease from 60.6% to 48.9% (di ﬀ erence: 11.7%) ( p < 0.0001) in the number of participants meeting the recommended World Health Organization (WHO) PA levels. Subgroups including men, participants aged 43 or over, and those not holding a university degree had the greatest reductions in both weekly minutes of PA and adherence to guidelines. The PA levels of the Spanish population generally declined during the ﬁrst days of COVID-19 conﬁnement.


Introduction
Coronavirus disease 2019 (COVID- 19) was declared a global pandemic by the World Health Organization (WHO) on 11th March of 2020 [1]. At the time of writing the present document (29 March 2020), 634,813 cases had been diagnosed globally, resulting in 29,891 fatalities thus far [2], and Spain ranked as one of the most inflicted countries, with a total of 85,195 cases diagnosed and 7340 fatalities thus far [3]. On the 15th March of 2020, the Spanish Government approved a period of confinement in order to fight the spread of COVID-19 [4,5]. During the confinement period, the Spanish population had to stay at home and, in consequence, interrupt their usual activities; owing to these restrictions, physical activity (PA) levels can be dramatically affected since Spanish citizens are recommended to telework and do PA at home instead of training in either sports facilities or outdoors [4][5][6]. Time spent outdoors has been associated with higher levels of PA, higher physical fitness, and lower sedentary behavior, whereas outdoor sports have been linked to a wide-range of health benefits; thus, Spanish citizens without access to open spaces or green areas might experience a decline in their PA levels [7,8].
Although exhaustive exercise, overtraining, or high-intensity exercise may lead to immunosuppression [9], emerging evidence suggests that regular PA enhances the immune function and increases resistance to infections compared with a sedentary lifestyle [10][11][12][13][14][15][16]. Even when performing PA below the recommended guidelines, the long-term risk of death from infectious disease lowered 40% [17]. The effects of PA on the immune function are also acute, as PA also produces an acute immune response [18]. Furthermore, one prospective study found PA associated with a decreased risk of community-acquired pneumonia among women [19], and pneumonia is one of the complications caused by COVID-19 in some infected people [20]. Moreover, this is not only a question of physical health but also a mental health issue, since a social isolation context and perception of loneliness predicted both anxiety and depression symptoms in older people, and could lead to both poor mental health as well as a higher likelihood of suicidal ideation [21][22][23], which might be attenuated through PA [24][25][26].
Furthermore, moderate PA improves mood [27] and helps to prevent and reduce anxiety and depression [28]. Also, previous research has found that PA produces not only long-term benefits in mental health, but also immediate psychological benefits for mood, anxiety, and stress due to the acute effects of PA [29][30][31]. Therefore, given the benefits of PA to physical and mental health, it seems very important that people meet the global recommendations for PA set out by WHO during the confinement period [32], although during this confinement period the PA should be carried out at home to ensure government social distancing requirements are being maintained. Throughout a week, WHO recommends that adults perform at least 150 min of moderate-intensity aerobic PA, or 75 min of vigorous-intensity PA, or an equivalent combination of moderate-and vigorous-intensity aerobic PA; WHO also recommends that muscle-strengthening activities should be performed involving major muscle groups on two or more days a week [32].
Some recent studies with a large representative sample of Spanish adults, carried out before the confinement period, have indicated that 70% of Spanish adults meet recommended PA guidance set out by WHO, although these results were based on self-report and may be overestimated [33]. Moreover, since performing PA outdoors has been restricted during the COVID-19 confinement, data from Spain might be lower than in other European countries, likely having a greater impact on the population's health. Given the importance of PA to health, this study aimed to quantify the change in PA of the Spanish adult population during the COVID-19 confinement compared with before confinement. Quantification of this change is important to provide practical recommendations, e.g., on physical activities to perform at home.
We hypothesized that PA will decrease in Spanish adults during the first days of confinement. The results of this study can be very useful for countries where PA outdoors is still forbidden and for future coronavirus resurgence that can make necessary new confinements, helping to design and implement measures that guarantee that the population can do adequate levels of PA during confinements.

The Survey
A cross-sectional online survey design concerning health habits was conducted during the COVID-19 outbreak in Spain. A web-form link available from the 22nd to the 29th of March 2020 (i.e., from the seventh official day of Government-enacted national confinement) was launched on social media encouraging users to answer the survey. Convenience sampling was used to select the participants of the study; according to server analytics, 2850 media users covering all the Spanish regions were asked to participate. Participants were previously informed about the purposes of the study, gave informed consent to participate, and indicated whether they were in a confinement situation. Provided data were anonymous and treated according to Spanish law regarding general data protection. Once the survey was completed, the participants were provided with advice on how to be active at home and its benefits. The present study retrieved data from the survey in relation to the following variables: PA (i.e., before and during the confinement), age, gender, civil status, occupation, and education.
The study followed the principles of the World Medical Declaration of Helsinki, and got the approval of the Ethics Committee of Research in Humans of the University of Valencia (register code 1278789). The reporting of this study follows the guidelines of the Strengthening the Reporting of Observational Studies in Epidemiology statement (STROBE) [34].

Physical Activity
Assessment of PA was conducted through physical activity vital sign (PAVS) short form, in which participants answered two questions regarding the number of days they were engaged in PA during a usual week, and the average minutes they usually performed PA in those days; participants could select between 0 to 7 days of PA a week and 10, 20, 30, 40, 50, 60, 90, and 150 or more daily minutes [35]. The PAVS was asked twice to the participants; first referred to the PA they usually performed before the confinement, and second referred to the isolation period. In order to apply PAVS to that particular context of confined population, the word "usual" was replaced for "isolation" when referring to the week they performed PA. Following PAVS original procedure, weekly minutes of PA were calculated by multiplying days with minutes. Participants were later classified in accordance with the WHO recommendations for moderate-intensity aerobic PA.

Demographic Characteristics
Given previous research [34,35], the study also estimated age, gender, civil status, occupation, and education as factors. All variables were self-reported by the participants and categorized as follows: Civil status ("married" or "not married"), occupation ("employed" or "not employed"), and education ("having a university degree" or "not having a university degree").

Statistical Analyses
Statistical analyses were conducted through the Statistical Package for Social Sciences (SPSS) version 23.0 (SPSS Inc., Chicago, IL, USA). The Kołmogorov-Smirnov test was applied to check normality. Paired t-tests were performed for each subgroup to examine PA differences within each subgroup for each characteristic. Independent t-tests were performed on the subgroup differences to determine whether the subgroup differences varied within each characteristic. Differences in relation to prevalence of WHO guidelines were evaluated by Mc Nemar Chi-squared tests, providing the frequencies and percentages of each group, the percentage difference, χ2 value, degres of freedom, and p-values. To check differences regarding demographic characteristics, the analyses were stratified for each one of the factors. A cut-off point for age (i.e., 43 years) was estimated through a data-driven approach using receiver operating characteristic (ROC) curve analysis. There were no missing data. Levels of significance were set at p < 0.05. In addition, the effect size was calculated using Cohen's d, classified as small (0.20), medium (0.50), and large (0.80).

Results
A total of 2042 participants on average aged 35.9 (SD 13.6) completed the survey. Table 1 presents the descriptive statistics for the sample, in which 1104 participants (54.1%) were women, 1001 (49.0%) were married, 1293 (63.3%) held a university degree, and 1287 (63.0%) were employed. Regarding WHO guidelines concerning PA, participants from all categories presented lower values during the isolation in Table 2. The overall percentage of reduction was 11.7% (p < 0.0001) with values remaining significant for all the subgroups except for women. The subgroup of men and participants aged 43 or over showed higher reductions of WHO guidelines achievement with respectively 17.1% (p < 0.0001) and 17.2% (p < 0.0001). Concerning differences in the number of weekly PA minutes reduced during the COVID-19 isolation period, the subgroups of women (p < 0.0001), participants holding a university degree (p < 0.005), and employed (p < 0.05) showed significantly lower reductions of PA weekly minutes as regards their counterparts.
The subgroup aged 43 and over was the only one not reaching WHO guidelines regarding PA before and in the course of the confinement period. Minutes of PA declined during the isolation period for all subgroups; Table 1 also showed a significant difference of 45.2 min (p < 0.0001) between the PA performed before and during the isolation for the overall sample. When examining the decline for each demographic characteristic, men, adults with no university education, and adults not employed reported significantly larger declines in minutes of PA.
As regards the way PA was performed during the COVID-19 isolation period, Figure 1 lists a range of activities that the participants performed. Household ranked first in this list (62.2%), in which social networks and applications devices respectively accounted for 27.4% and 19.4% of the participants.

Discussion
Findings from the present study in a large sample of the Spanish adult population showed that PA weekly minutes significantly decreased by 20% with the COVID-19 confinement (Figure 2).

Discussion
Findings from the present study in a large sample of the Spanish adult population showed that PA weekly minutes significantly decreased by 20% with the COVID-19 confinement (Figure 2).

Discussion
Findings from the present study in a large sample of the Spanish adult population showed that PA weekly minutes significantly decreased by 20% with the COVID-19 confinement (Figure 2). This reduction was particularly pronounced in men, older participants, and those with lower education levels. The results of the present study support the general notion that an isolation context associates with less PA, as observed in prior research with older adults [36]. Further, this study strengthens the findings found in recent research, in which adult populations showed lower PA levels during the COVID-19 confinement in all types of PA intensity levels, as well as in both physically This reduction was particularly pronounced in men, older participants, and those with lower education levels. The results of the present study support the general notion that an isolation context associates with less PA, as observed in prior research with older adults [36]. Further, this study strengthens the findings found in recent research, in which adult populations showed lower PA levels during the COVID-19 confinement in all types of PA intensity levels, as well as in both physically active and inactive participants [37,38]. Nevertheless, our study provides novel data from an unknown setting, in which participants have not voluntarily decided to isolate themselves, and citizens of all ages had been officially ordered to stay at home for at least a month. Importantly, our study showed that only a week of isolation might have a great impact on PA reduction.
A report from the European Union (EU) found Spanish PA levels to be on average regarding other European countries, with 56% never or seldom doing PA [39]. However, a study by Fernández-Navarro et al. [40] suggests a worse situation, in which around 73% of the Spanish population does not meet the WHO guidelines concerning PA, with a subsequent higher risk of chronic diseases; as observed in that research, our study also showed men to be more physically active than women, as well as the important role of education. On the other hand, our study points at a higher percentage of Spanish residents achieving PA recommended levels, particularly before the COVID-19 isolation, which could be due to substantial differences regarding age (i.e., our study sample of adults is younger compared to other studies). In fact, our study found lower levels of PA in older participants, which might be partly explained by the ageing process itself [41]. Another highlight was that university degree holders performed higher PA, and presented a lower decline of PA levels as regards their counterparts; one plausible explanation for this finding might be a higher awareness of having a healthy lifestyle, particularly in such an exceptional setting, since educational differences have been underscored to be a reliable predictor of leisure-time PA [42]. In accordance with previous research [43], unemployed participants of this study showed lower PA levels than those employed, although the fact that specific job roles (e.g., health workers or security forces and corps) are exempt from complying with the COVID-19 isolation governmental normative might partially explain these differences (i.e., working duties of specific job roles such as health workers and security corps might contribute to increase their PA levels).
Our study observed significant reductions in performed weekly PA as well as in adherence to WHO guidelines in all subgroups; however, such reductions mainly showed a small to medium effect size. Thus, according to this data, PA reductions are wide-spread within the sample without particularly focusing too much on any specific subgroup. Nevertheless, it is important to notice that the effect seemed to be more pronounced in groups with lower education as well as in the elderly, i.e., groups that are already at increased risk of lifestyle diseases. Thus, confinement may amplify existing inequalities in health across groups in society.
As expected, household chores were observed to be a prevalent PA content among the participants of this study. Due to the amount of time the participants are supposed to spend at home, the household accounts for an important part of the performed weekly PA. It is noticeable that social media and application devices are moderately used for PA purposes during this period, although our percentages are similar to others found in previous research [44]. Performing home-based PA through wearables and smartphone applications might be a way forward to promote sufficient PA during this confinement period, although the effects of their implementation to increase PA levels have been generally estimated between small to moderate, and it is still a challenge to reach the most physically inactive people with these type of tools [45].
Overall as well as subgroup reductions of weekly PA might be explained for the lack of social support and the feelings of loneliness, which have been associated with lower levels of PA and poor health conditions in both younger and older adults [36,46,47]. In fact, a prolonged time of social isolation has been observed to associate with cardiovascular diseases, poor mental health, and to increase all-cause mortality [22,48]. Social isolation and feelings of loneliness are current issues that, in this particular situation of confinement due to COVID-19, might be exacerbated by the fact that physical human interaction has been constrained for a significant period. In addition, natural light and natural environment exposure deprivation might lead to a worse mood and poor mental health, which consequently, might predispose for less PA [49]. Thus, anxiety may increase as the COVID-19 related mortality increases if the quarantine time continues [50]. Besides, prior research has observed lower socioeconomic status (i.e., lower educational level and unemployed) to increase the odds for joint health-risk behaviors involving physical activity, screen exposure, sleep time, fruits and vegetable consumption, alcohol consumption, and smoking habit [51] among Spanish adults confined due to COVID-19; hence future research focused on deprived populations, such as those from poorer backgrounds or those financially affected by the COVID-19 pandemic, would be of special interest.
Strengths of the current study include using a large and widespread sample of Spanish adults (i.e., participants from all the Spanish regions) and the use of a validated tool to assess PA. Moreover, the fact that significant reductions of PA levels and adherence to WHO guidelines have significantly decreased in all the study subgroups within a short-term, provides consistency to our findings. However, several limitations should be highlighted for this study. First, due to the fact that PA and other variables are self-reported, it might lead to information bias. Second, since young adults are overrepresented in the study sample, estimated PA levels might be higher than in the general Spanish population of adults; owing to the type of sampling, there is a possibility of selection bias. Third, due to the study design, the present work does not allow drawing causal inferences. Fourth, adjusted analyses for health status might lead to different results. Also, important aspects regarding PA (i.e., intensity levels) or critical health-related habits such as diet have not been reflected in this study. Last, because the collected data mainly referred to the first week of the COVID-19 outbreak in Spain, it might not properly reflect the big picture of the influence of social isolation and confinement over PA (i.e., more days of isolation might associate with even lower levels of PA). Thus, future research should investigate how PA levels are varying during this complete period.

Conclusions
The PA levels of the Spanish population generally declined during the first days of COVID-19 confinement. The promotion of home-based physical activities is particularly recommendable in this specific context to reach the recommended PA levels set out by WHO.