The current pandemic situation caused by SARS-CoV-2 has caused a deep change in citizen’s lifestyle. Measures focused on the protection of public health have been implemented worldwide. In particular, the Spanish government imposed a national lockdown on March 14^th, 2020, due to the severity of the situation. For nearly 3 months, Spanish population had to adapt their lifestyles to the new situation of staying at home, in a context where outdoor activities are an essential part of their cultural habits. In addition to this, the stressful environment created by the rising of COVID-19 cases and deaths (Lima et al. 2020), the fear of infection, the uncertainty about the near future and the possible socio-economic consequences of the pandemic (Balluerka et al. 2020; Garre-Olmo et al. 2020), are just some factors that have a psychological impact on the population (Brooks et al. 2020).

Indeed, in the previous outbreaks of Severe Acute Respiratory Syndrome (SARS) and the Middle East Respiratory Syndrome Coronavirus (MERS-CoV), psychological symptoms like anxiety (Jeong et al. 2016), depression, and post-traumatic stress disorder (PTSD) (Ko et al. 2006; Reynolds et al. 2008; Lam et al. 2009; Mak et al. 2009; Yoon et al. 2016) were detected in individuals who were isolated, and those symptoms remained for some time (Wu et al. 2008; Jeong et al. 2016). Regarding COVID-19 pandemic, a study comparing the Spanish and Greek populations during the first outbreak showed that living in Spain was associated with increased anxiety symptoms. Authors attributed these differences to the severe situation and stricter lockdown measures in Spain (Papandreou et al. 2020).

The effect of psychological symptoms on lifestyle has been described in traumatic situations, like terrorist attacks (Boscarino, Adams, and Galea, 2006) or natural disasters (Lebeaut, Tran, and Vujanovic, 2020), which were associated with an increased risk of addictive behaviors such as alcohol abuse (García-Álvarez et al., 2020). Therefore, it should be considered the potential risk that COVID-19 lockdown may exert on the development of unhealthy habits. Studies conducted during COVID-19 lockdown showed an increase in the consumption of alcohol or comfort foods (Balluerka et al., 2020; Sidor and Rzymski, 2020). It has been pointed out those negative emotions such as anxiety or depression, may prompt Emotional Eating (EE) behaviors (Matcht, 2008; Konttinen et al. 2010; Clum et al., 2014; Ferrel, Watford, and Braden, 2020). EE is defined as the tendency to eat, usually hyper-palatable foods (Macht, 2008; Camilleri et al. 2014; Muscogiuri et al., 2020), in the response to positive or negative emotions (Cardi, Leppanen, and Treasure, 2015). Some authors indicate that this eating behavior could mediate between depressive symptoms and body mass index (BMI) (Clum et al., 2014), waist circumference, and percent body fat (Konttinen et al., 2010).

Moreover, physical activity (PA) has become a paramount concern due to the confinement situation. For instance, Spanish citizens were not allowed to exercise outdoor for at least seven weeks, which implied that PA had to be adapted to be practiced at home. This fact may have impaired exercising, as several investigations have recently shown (Balluerka et al., 2020; Garre-Olmo et al., 2020; López-Bueno et al., 2020; Rodríguez-Pérez et al., 2020). Additionally, PA during the lockdown should have compensated for the decrease of the activities of daily living (Jiménez-Pavón, Carbonell-Baeza, and Lavie, 2020). Short- term physical inactivity and a positive energy balance can increase risk factors for developing metabolic syndrome, owing to reduced insulin sensitivity, increased total body, central fat and a proinflammatory status (Martínez-Ferrán, Sanchis-Gomar, and Pareja Galeano, 2020). Overall, habit changes have been detected in Spanish population during the lockdown, observing differences in sociodemographic and living conditions (Navas-Martín et al., 2021). In a society where Non-Communicable Diseases (NCDs) are responsible for almost 70% of all deaths in the world (https://www.who.int/health-topics/noncommunicable-diseases#tab=tab_1), is necessary to pay attention to these circumstances which could lead to the establishment of unhealthy habits in a long term. In particular, it is interesting to evaluate how citizens adapt their lifestyle to this exceptional situation from a working level perspective, due to the significant changes that the confinement causes in this specific area.

Therefore, the aim of this study was to determine the effect of the COVID-19 lockdown on workers’ lifestyle from Spanish National Research Council (CSIC), the largest public research institution in Spain. In most cases, CSIC workers had to adapt to work from home, which led to a new situation since the telecommuting was not established. Consequently, assessing lifestyle changes under such special situations, especially on dietary patterns, PA practice and emotional changes becomes necessary.

To investigate the healthy habits of CSIC workers, a survey was sent during the lockdown applied in Spain from March 14^th, 2020. The questionnaire had an online Google docs format, and it was approved by CSIC Ethics Committee. Participation was shared through the CSIC institutional mailing list, reaching only the internal CSIC staff and not external services, and responses were accepted from April 20^th to June 14^th, 2020. Participation in the online questionnaire was voluntary and anonymized, being necessary for a previous acceptance of the participant.

The survey consisted of 39 questions structured into 4 sections: socio-demographic characteristics, dietary patterns, physical activity, and emotional state during confinement (Figure 1).

Socio-demographic characteristics: gender, age, height, weight, training in nutrition and dietetics, current work modality, cohabitants, living conditions, and the presence of COVID-19 symptoms were asked. Body Mass Index of the respondents was also calculated with the self-reported height and weight.

Changes in dietary patterns: perception of changes in eating and consumption habits during the confinement (shopping activities, availability of food and meals per day) and 14 questions about food group consumption were asked. Participants had to indicate an increase, decrease, or no change in the intake of water, energetic drinks, alcoholic beverages, fruits and vegetables, wholegrain cereals, legumes, fish, dairy products, red and cured meat, pre-cooked products, fast food, savory snacks, sweet snacks and bakery products.

Two score systems were created to evaluate the consumption of food considered a more or less healthy food. For the healthy food score, points addressed to intake of fruits and vegetables, pulses, fish and whole-grain cereals were assigned, obtaining 2 points if the consumption increased, 1 point if it did not change and 0 points if it decreased. Then the total number of points was divided between the highest rating that it could be obtained, resulting in a score from 0 to 1, where 0 meant the participant decreased the consumption of all the considered healthy food, while 1 meant an improvement of its intake. For the unhealthy food score, the following food groups were included: energetic drinks, alcoholic beverages, savory and sweet snacks, bakery products, red and cured meat, pre-cooked products and fast food. The score was calculated using the same procedure, but 2 points were assigned when the intake of these products decreased, and 0 points were assigned if its consumption increased. Thus, the higher the score was the lower intake of these non-recommended products, and vice versa. Therefore, higher scores in both indicators meant better dietary changes.

Changes in physical activity (PA): practice during the confinement, changes in usual routine, motivation and factors that increase motivation were inquired.

Changes in emotional states (ES): anxiety, boredom, sadness, irritability, and sleep disorders were asked through Likert scale questions, rating from 1-5 (1 no change, 5 complete change). Checkbox type questions were used to evaluate the attitudes and strategies adopted to cope with emotional changes.

An Emotional Eating Score (EE score) was created by the inclusion of those items associated with overeating [“It is difficult for me to control the amount of hyper-palatable food (sweets, fried snacks, etc.) that I eat”, “I eat more because of anxiety, boredom, etc.”, “Snacking” ]. One point was assigned when each item was not checked, thus the higher punctuation, the lesser emotional eating behaviors were followed.

Statistical analyses were conducted using IBM SPSS Statistics for Windows, Version 26.0. (IBM Corp. Released 2019, Armonk, NY, USA). To explore personal and sociodemographic characteristics, dietary patterns, physical activity and emotional state, frequencies, percentages (%) and means with standard deviations were employed. For the analysis of the variables, Chi-square was applied when comparing categorical variables and ANOVA to test differences between the means of groups. Correlations between continuous variables were calculated using Pearson correlation.

During the lockdown, 33% of the participants increase the preparation of traditional meals, 18% increased meal planning and 15% used shopping lists. 70% of the participants did not change the number of meals per day, where the most common behavior was 3 meals per day.

Almost half of the participants (45.3%) perceived changes in their diet during the confinement (Table 2). From those, 35% expressed that their eating habits had worsened, exhibiting a higher BMI average in this group in relation to participants with a different perception (25.1 ±4.61; 23.6±3.73; 24.3±3.59 kg/m² perceived worsening, no change and perceived improvement, respectively; p<0.001). Moreover, women expressed proportionally worse eating habits during the lockdown than men (18.7 and 11.5%, respectively; p<0.001). Concerning age, 21% of the participants under 35 years old expressed that they changed their eating behavior negatively.

Overall, changes in diet were detected (Table 2), observing a rise in the number of participants who increased their consumption of pulses, fruits and vegetables, fish, dairy products, just as no recommended products like savory snacks and bakery goods. Conversely, several participants reduced the intake of red and cured meat, precooked products and fast food.

No significant differences in BMI and nutritional knowledge were detected for healthy and unhealthy scores, result that it should be highlighted. Nevertheless, the group that perceived worse eating habits during the lockdown obtained significant lower marks in the healthy score (0.49 ±0.17; 0.534 ±0.123; 0.651 ±0.157 (perceived worsening, no change and perceived improvement, respectively); p<0.001) and unhealthy score (0.386 ±0168; 0.544 ±0.149; 0.651 ±0.175 (perceived worsening, no change and perceived improvement, respectively; p<0.001). Additionally, a positive correlation between both scores (r=0.205; p<0.001) was found. Moreover, participants who increased their PA were significantly younger than those who reduced it (42.7±12.7 and 45.4±11.8 y, respectively; p<0.001).

Regarding the relationship between dietary patterns and PA, it was observed that, those who had increased their training during the lockdown obtained higher eating scores than those who reduced their PA, for both healthy (0.584 ±0.154 vs. 0.547 ±0.16 (increased vs. decreased PA); p<0.001) and unhealthy scores (0.576 ±0.184 vs. 0.529 ±0.182 (increased vs. decreased PA); p<0.001).

All emotional markers were positively correlated with each other (p<0.001), and a strong correlation between anxiety (anx.), irritability (irr.) and sadness (sad.) was detected (r=0.578 (anx.-irr.), 0.626 (anx.-sad.) and 0.569 (irr.-sad.). Additionally, a negative correlation between age and all emotional indicators was also found (p<0.001).

The COVID-19 pandemic has brought about a global health crisis which led to the application of severe restrictive measures on the general population like confinement. Previous outbreaks like SARS o MERS showed that people under isolation can develop psychological disorders (Ko et al., 2006; Reynolds et al., 2008; Lam et al., 2009; Mak et al., 2009; Jeong et al., 2016; Yoon et al., 2016) which may lead to other health problems, as alcohol dependence (Wu et al., 2008), besides the possible effect on physical activity and diet. For this reason, health organizations (https://www.euro.who.int/en/health-topics/health-emergencies/coronavirus-covid-19/publications-and-technical-guidance/food-and-nutrition-tips-during-self-quarantine; Academia Española de Nutrición y Dietética y el Consejo General de Colegios Oficiales de Dietistas-Nutricionistas, 2020) specifically established health and nutrition recommendations for a self-quarantine situation during the COVID-19 outbreak.

According to our study conducted with workers telecommuting from CSIC, it seems that these recommendations, in general, were followed, since a tendency for a greater intake of fruits, vegetables, pulses, and a decreased consumption of pre-cooked meals, red and cured meat, and fast food was identified. Our findings correspond with similar studies during the COVID-19 confinement in Spanish adults (Laguna et al., 2020; López-Moreno et al., 2020; Pérez-Rodrigo et al., 2020; Rodríguez-Pérez et al., 2020) and other Mediterranean countries (Scarmozzino and Visioli 2020). Some authors reported an increase in the consumption of fruits, vegetables and pulses during confinement that was related to motivations of health or weight control, lower price and long shelf-life (in the case of pulses and canned goods), whereas the decrease in non-perishable ready meals was associated its unhealthy profile (Laguna et al., 2020) and more available time for cooking (Laguna et al., 2020; Scarmozzino and Visioli, 2020).

Regarding PA, several studies conducted in Spanish population (Balluerka et al., 2020; Garre-Olmo et al., 2020; López-Bueno et al., 2020; Rodríguez-Pérez et al., 2020) found out that ~50% of the people exercised less during COVID-19 lockdown, results matching with those presented here. It is worth noting that the participants with worse eating scores, women and young people, were the ones who declared an increase in their PA routine during the confinement. This trend was also observed by García-Tascón et al. (2020), who hypothesized that men are more likely to exercise for social, ludic and competitive reasons, while women prefer non-competitive and relaxed sports and, hence, they were able to adapt exercise at home easily. These statements are in line with the study conducted by Sánchez-Sánchez et al. (2020), who indicated that men invested more time in exercising before the lockdown than women, whereas during confinement, men have decreased more their training routine than women. Therefore, it seems that men adapted worse their PA practice at home.

Another relevant possible effect of the COVID-19 pandemic over the population is the emotional changes that this situation may trigger. Psychological symptoms such as anxiety, depression or poor sleep quality, have been described during the COVID-19 outbreak in adults (Balluerka et al., 2020; Chacón-Fuertes, Fernández Hermida, and García Vera, 2020; DiRenzo et al., 2020; Garre-Olmo et al., 2020; Ingram, Maciejewski, and Hand, 2020; Jiménez-Pavón, Carbonell-Baeza, and Lavie, 2020; Moccia et al., 2020; Ozamiz-Etxebarria et al., 2020; Rajkumar, 2020; Beck et al., 2021), as well as in previous outbreaks like SARS in 2003 (Hawryluck et al., 2004; Ko et al., 2006; Wu et al., 2008; Lam et al., 2009; Mak et al., 2009) or MERS-CoV in 2012 (Jeong et al., 2016; Yoon et al., 2016). Several factors related to the emotional impact caused by COVID-19 pandemic on the population have been described. For instance, the concern for oneself infection or their relatives (Balluerka et al., 2020; Wang et al., 2020) or an adverse work situation, especially for those whose economy may be compromised (Garre-Olmo et al., 2020), could be the origin of emotional alterations. In this sense, CSIC participants did not jeopardize their job during the COVID-19 crisis, but they experienced emotional changes, especially sleep disorders and anxiety. According to Garre-Olmo et al. (2020), changing work conditions, including telecommuting, is associated with depressive symptoms, which was the situation of the majority of our participants. Moreover, young participants experienced more changes in ES than older workers did. Similar results have been reported in other studies (Balluerka et al., 2020; ; DiRenzo et al., 2020; Ozamiz-Etxebarria et al., 2020; Qiu et al., 2020; Beck et al., 2021). The effect of overload information or infodemic could be a possible explanation to this behavior, which has been pointed out as a cause of anxiety and depression (DiRenzo et al., 2020; Gao et al., 2020), especially in young people who are usually more exposed to social media (Qiu et al., 2020). Lastly, women have shown more anxiety, sadness, sleep disorders and irritability than men have. Even though, Ozamiz-Etxebarria et al. (2020) described more depressive symptoms in men and similar levels of stress and anxiety in relation to gender. However, other studies reported more emotional changes in women (Balluerka et al., 2020; DiRenzo et al., 2020; López-Moreno et al., 2020; Qiu et al., 2020; Wang et al., 2020; Beck et al., 2021). For instance, Navas-Martín et al. (2021) highlighted the greater impact on habit changes, especially those related to home care, in women and people living with children. According to Moscoso-Sánchez (2020), this could be related to the women’s triple role (reproductive, housekeeping and work outside the home) that still prevails.

Furthermore, some authors linked psychological behaviors such as emotional eating (Nonogaki, Nozue and Oka, 2007; Rodríguez-Martín and Meule 2015; Ferrell et al., 2020; Yilmaz and Gökmen, 2020); or alcohol use (Boscarino, Adams, and Galea, 2006; Wu et al., 2008; Lebeaut, Tran, and Vujanovic, 2020; Sallie et al., 2020) as strategies to cope with stressful situations. In fact, after the SARS outbreak Wu et al. (2008) reported an association between individuals who were in quarantine, with later alcohol abuse or dependence symptoms. Our results showed that 23% of the participants decreased their alcohol consumption while 20% increased it during the confinement, in accordance with other recent studies (Balluerka et al., 2020; López-Moreno et al., 2020; Sallie et al., 2020). Additionally, the mean age of those who increased their alcohol consumption (45.4±11.3y) was higher than those who decreased it (40.2±13.6y), findings concurring with those presented by Sallie et al. (2020), who postulated that older individuals, parents with children, essential workrs, people with depressive and anxiety symptoms or an impulsive personality might have a greater risk for alcohol misuse.

Concerning EE, our results showed a greater intake of snacks during the lockdown, which is in line with consumer reports from grocery stores (Los españoles confinados picotean hasta un 50% más; El ticket medio recupera la estabilidad. ¿Cómo evolucionan las compras de alimentación y bebidas en el confinamiento?) and other investigations (Laguna et al., 2020; Naja and Hamadeh, 2020; Ruiz-Roso et al., 2020; Sánchez-Sánchez et al., 2020; Scarmozzino and Visioli, 2020; Sidor and Rzymski, 2020) during the same situation. This could be related to the fact that negative emotions can trigger the intake of energy-dense food, such as snacks or sweets (Macht, 2008; Camilleri et al., 2014; Clum et al., 2014; Charbonnier et al., 2015). Additionally, women showed an overall greater consumption of these products in our study, as it has been recently reported (Balluerka et al., 2020; Pérez-Rodrigo et al., 2020). This behavior could be associated with higher anxiety levels experienced in women, who increased their savory and sweet snacks intake during the lockdown. In fact, it seems that emotional changes have a greater influence on eating behaviors in women, as 33% of them in our study declared that emotions exert a negative effect on their diet vs. 25% of men. These findings are supported by the results of López-Moreno et al. (2020).

Diverse hypotheses correlate specific eating behaviors with higher levels of stress and negative emotions. Physiologically, food intake is suppressed under stress-induced psychological changes due to the activation of the sympathetic nervous system, which may interfere with digestion processes such as gastrointestinal transit or by delaying glucose absorption (Tsigos et al., 2000). However, overeating behavior under negative emotions or stress conditions has well been reported. The reason for these mixed behaviors is a complex problem still unresolved and may involve physiological, behavioral, emotional and cognitive mechanisms (Macht, 2008; DiRenzo et al., 2020; Muscogiuri et al., 2020). Thus, food-induced emotion may affect the control of food choice, where energy-dense foods usually rich in fat and sugar, would evoke positive affective responses causing a greater intake, whereas bitter compounds would provoke the opposite effect (Charbonnier et al., 2015). Another class of emotion-induced changes that could trigger overeating is the impairment of cognitive eating controls. Lastly, a learning perspective could be the motive in which eating would be used as a strategy to regulate emotions. In this case, the learning process would involve the negative emotions as the stimuli and eating palatable food as the operant behavior, which would result in a reduction of the negative emotions due to the mood-elevating effect of this kind of food. The repetition of this behavior may reinforce it and might be used to cope with negative emotions (Rodríguez-Martín and Meule, 2015). For instance, in the study of DiRenzo et al. (2020) about the relationship between eating habits, mental and emotional mood among the Italian population, almost half of the participants claimed that they used food when they felt anxious as a tool to feel better.

Regarding the effect of boredom on EE, Koball et al. (2012) pointed out that boredom should be considered in a separate dimension of EE, since the eating response, which triggers is different from the response activated by other emotions. This may explain why our results did not show gender differences when participants felt bored. Laguna et al. (2020) suggested that the lockdown forced the population to transfer the usual leisure activities to home, which would raise the intake of snacks and homemade bakery. In fact, higher levels of boredom were associated with the intake of bakery goods and all kinds of snacks (Muscogiuri et al. 2020). Moreover, according to consumer reports (Cotos 2020; https://www.revistainforetail.com/noticiadet/el-ticket-medio-recupera-la-estabilidad/738fe8425e3be42b8b685a0751b23265) the sales of products related to homemade bakery and appetizers remarkably increased, reaching even to be sold out during the confinement.

This study presents some limitations as the possible bias owing to the self-reported data. In addition, this kind of online questionnaires cannot be excessively longer because of the risk of a low engagement level.

Our study was designed to explore the effect of the lockdown during the first and more severe period. For this reason, the data cannot be considered representative, not being possible, for instance, to determine in a deeper level the gender differences regarding the home care. A longer study targeting the same population in different stages of the pandemic, as the inclusion of other variables, would have contribute to establish stronger and additional correlations.

The strength of this study is the large representation of all CSIC centers and the notable number of participants considering the total CSIC workers.

The COVID-19 health crisis and the ad doc lockdown measures implemented in Spain seem to cause a negative effect on CSIC workers’ lifestyles, especially from a psychological status point of view. This work shows the effects of the lockdown on a specific population, which did not have risk of losing their jobs, but had to change their working conditions to a telecommuting mode.

Despite emotional changes were detected, a general improvement in the dietary patterns was identified, probably owing to an increased health concern and more available time for cooking. However, the intake of unhealthy products such as snacks and bakery goods seem to have increased during the confinement, especially in women. This consumption could be associated with greater emotional changes in women, particularly for anxiety, and the strategies adopted by them to cope with these emotional alterations. In this way, emotional eating behaviors, like snacking, lack of control over the consumption of hyperpalatable food and intake related to emotions, have been associated with greater levels of anxiety. In addition, it should be highlighted that in a context where food is undoubtedly involved in Spanish leisure activities, it is not surprising the link of boredom and a greater intake of snacks and bakery products.

Moreover, the mobility restrictions established during confinement had a great effect on the practice of physical activity that was reduced in almost half of the population. In addition, it seems that men adapted their PA practice worse to the confinement situation, possibly due to the type of sport practiced before the pandemic.

Further investigations are needed to determine the long-term effects of COVID-19 lockdown on the population’s lifestyles. Therefore, our study provides initial information about how the lockdown influenced this specific population and then develop future interventions to mitigate it.