Changes in Eating Habits and Physical Activity after COVID-19 Pandemic Lockdowns in Italy

The introduction of lockdowns and other containment measures during the COVID-19 pandemic substantially altered people’s lifestyle and dietary behavior. Several studies evaluated the short-term effects of these measures; yet reports on long-term consequences are scarce. We sought to address this gap in the literature by analyzing dietary and lifestyle data collected at an obesity center in Rome, Italy. The Italian region of Lazio was hit hard by the pandemic. To evaluate the potential health impacts, we compared the pre- and post-lockdown data of 118 individuals. Contrary to the common belief that lockdown had adverse effects solely on people’s dietary habits, we observed a significantly increased consumption of raw vegetables, whole grains, and water in our study sample. Favorable effects, however, were also accompanied by adverse trends, such as a higher prevalence of sleeping difficulties. Our data emphasize that the lockdowns associated with the pandemic also influenced participants’ social behavior, with less individuals reporting eating out or in company. Our study highlights the substantial impact of lockdowns on many dimensions of life. As such, it is of utmost importance in the critical evaluation of such stringent containment measures.


Introduction
The advent of the COVID-19 pandemic has caused substantial changes in people's eating and lifestyle behaviors. Due to several consecutive months of physical and social isolation, many individuals modified their dietary habits. Several studies evaluated the short-term effects of lockdown on nutritional habits, emphasizing both positive and negative changes [1,2].
The Italian region of Lazio experienced a tight lockdown phase (between 9 March and 3 May 2020), with a subsequent tightening of containment measures (between 4 May and 14 June 2020). Thereafter, from 15 March 2021 until June 2021, the Italian government again imposed several other strict lockdown periods. The most severe lockdown period was characterized by the suspension of common commercial activities, catering services, religious celebrations, and the banning of people gathering in public places. Regular office work was discontinued for most professions and was replaced by working from home. For students, in-person teaching was suspended [3]. From 6 August 2021, normal activities gradually returned, especially for vaccinated persons with a green pass [4]. The reactivation of activities showed the necessity to deal with new routines and habits [5]. Some studies have shown favorable changes in individuals' eating habits, with an increase in the number of meals prepared and consumed at home [3][4][5]. Favorable effects were accompanied by other negative ones, such as an increase in alcohol consumption [4], increased "comfort food" consumption due to emotional distress [5], and reduced sleep quality [6]. These changes were accompanied by a more sedentary lifestyle due to increased screen time [7], smart working and the resultant abolition of walking or public transport to work [8], and fewer opportunities for practicing physical activity [9]. Most studies have evaluated the short-term effects of social distancing and home isolation [5][6][7][8][9]. However, there is a lack of studies assessing the long-term effects of quarantine periods. The aim of this study was to evaluate the potential impact of lockdown periods on eating habits and health risk factors through the use of a questionnaire that was administered before and after the lockdown periods.

Materials and Methods
We collected data related to eating behaviors, food habits, and food taste before and after the aforementioned lockdown periods at an obesity center in Rome, Lazio, Italy. For this study, we defined the pre-lockdown period ("before") as the time before March 2020. All procedures performed were in accordance with the ethical standards of the 1964 Helsinki Declaration and its later amendments. The study was approved by the Ethics Committee of the University Hospital "Tor Vergata" in Rome (ID number 44.22) and date of approval is 06 April 2020.
The survey was conducted among patients in a nutrition practice who were asked to complete an online questionnaire in the Italian language prior to their first visit. The survey was disseminated through an online form created specifically for this purpose. Altogether, 1256 individuals answered the survey between June 2018 and January 2020. In this sample of subjects, we assessed gender differences in taste and lifestyle habits [10]. The inclusion and exclusion criteria were established before the first survey to reduce the number of external confounders and to ensure a homogenous sample. In summary, patients with age < 12 years or >75 years, pregnant or breastfeeding, or a BMI < 18.5 kg/m 2 were excluded.
All 1256 participants were contacted via a dedicated newsletter between May and July 2021, with the request to repeat the same online questionnaire. The online survey, which required approximately 30 min to complete, allowed participants to answer the test via any electronic device with internet access. When accessing the link, participants expressed their consent to the use of the data and had the option to consent or not to participate. We recorded all responses anonymously. The online questionnaire consisted of four parts, as described elsewhere in detail [10]. The first part of the survey investigated daily meals, hunger during the day, eating disorders, and sleep quality. The second part of the test asked subjects about the consumption frequency of several food groups individually, with the option to respond ("yes", "no", and "rarely"). The third part of the survey inquired about favorite dishes, a typical day's eating, and the daily consumption of water as well as alcoholic and sugary drinks. The last part was dedicated to physical activity and included number of hours per week devoted to sport, time of day when sport was practiced, and type of sport. Average weekly METs (the ratio of the work metabolic rate to the resting metabolic rate) were calculated using the time spent per week on each sport activity.
An expert registered dietician performed the dietary history, and body composition (BC) assessments were evaluated in the days following the first meeting. Weight and height were determined with the subjects fasting overnight and wearing only underwear. Weight and body composition were assessed from a standing position after overnight fasting using a Tanita BC-420 MA (TANITA Corporation, Sportlife Tokyo, Japan, range of 0-200 kg, accuracy: 100 g), an instrument validated using dual-X-ray absorptiometry (DXA) [11]. Patients were required to observe the following guidelines before BC evaluation: 3 h or more after meals and not eating or drinking too much the day before the measurement; 3 h after awakening and beginning normal daily activities; 12 h or more after a hard workout; urination before the measurement; avoiding alcohol 12 h before the visit. Women were asked to avoid the visit in the days before and during the menstrual cycle because this could affect the accuracy of the measurement. The measurement of body composition in female subjects was performed only during the period from day 6 to day 24 of the menstrual cycle. We considered for this study only two parameters that were measured during the visit: fat mass (FM) and fat-free mass (FFM).

Statistical Analysis
Statistical analysis was performed using StatTech v. 2.1.0 (Developer: StatTech LLC, Kazan, Russia). Quantitative variables were assessed for normality using the Shapiro-Wilk test (when the number of subjects was less than 50) or the Kolmogorov-Smirnov test (when the number of subjects was more than 50). Quantitative variables following non-normal distribution were described using median (Me) and lower and upper quartiles (Q1-Q3). The Wilcoxon test was used for comparison of quantitative variables following non-normal distribution between two matched samples. In light of the pre-planned study hypothesis, we did not make use of the Bonferroni correction [12].

Results
A total of 118 participants completed both questionnaires and were included in the final analysis. Table 1 shows the baseline characteristics of all participants. Data are expressed as means ± SD. Abbreviations-SD: standard deviation; n: number of participants; FM: fat mass; FFM: fat-free mass; MET: metabolic equivalent. MET is defined as the amount of oxygen consumed while sitting at rest and is equal to 3.5 mL O 2 per kg body weight x min.
In addition to the frequency of food types, we observed some significant changes in drinking habits among participants before and after the lockdowns (Table 3). We observed no differences with regard to alcohol intake. Data are expressed as means ± SD. Abbreviations-SD: standard deviation; n: number of subjects. Participants drank more water (2 L per day 35.6% "before" vs. 50.8% "after"; p < 0.001) and fewer sugar-sweetened beverages (0 drinks per day: 39.8% "before" vs. 60.2% "after"; p < 0.001). Table 4 shows other relevant changes in dietary and sleeping habits and sports frequency before and after lockdowns.  The context in which meals were eaten changed as well. Participants ate out of home less often (24.6% "before" vs. 48.3% "after"; p <0.001) and ate lunch less frequently at a company canteen (16% "before" vs. 6.8% "after"; p = 0.012). Patients reported an increased frequency of sleeping disorders. We found a 6.7% increase in the number of people who said they had difficulties falling asleep at night (p = 0.33). Furthermore, we observed no reduction in time spent for physical activities, even though 22% of the subjects stated that they practiced more sport in the morning, as compared to only 12.7% prior to lockdowns. (p = 0.016) Tables 2-4 highlight other, non-significant trends prior to and after the lockdown periods.

Discussion
The long periods of isolation had an impact on patients' preferences and eating habits. It is commonly believed that lockdown is a period characterized by home-confinement and consumption of unhealthy foods, creating an obesogenic environment [13,14]. On the contrary, for many people, the increased availability of time was an opportunity to improve their eating habits, partly due to greater involvement in the kitchen [15]. Previous studies demonstrated that a lack of willpower, time constraints, and taste preferences are the main difficulties in people eating healthy [16] It is likely that the lockdowns' substantial impact on people's life induced significant improvements in several nutritional behaviors. For example, a French study [17] revealed that 83% of patients reported increased time spent cooking during self-quarantine.
In our study, we sought to assess the possible changes in lifestyle habits subsequent to periods of isolation and the particular general situation induced by the pandemic. Confirming the aforementioned studies, our data also show lockdowns were associated with some good eating habits, possibly due to the increased time available for meal preparation.
One example is the increased preference for raw vegetables and whole grains in our sample. Other studies [18,19] have shown that the quality of diets slightly improved and the prevalence of food insecurity was reduced in healthy populations. One potential explanation is that governmental guidance for a healthy diet to prevent COVID-19 was widely followed and implemented by the Italian population. Undoubtedly, the increased time spent at home in our sample of smart workers also contributed to the preference for fresh foods, which often require extensive preparation. It is conceivable that regular office workers would otherwise not have had the time to cook such meals.
Several changes may have been the effect of governmental campaigns to prevent COVID. The World Health Organization recommended legumes, fruits, and vegetables as the best food items during self-quarantine or longer home stays [20]. Our data showed that participants also reported drinking more water as compared to the pre-lockdown period. This could be explained by the increased time spent at home and water consumption being a more health-conscious behavior. Many employees, especially those working in contact with the public, are forced to limit their daily water consumption to avoid having to leave the workplace to use toilets [21]. Smart working probably improved this, as it provides more flexibility in work schedules and better opportunities to take breaks.
The consumption of sugary drinks was reduced as well. At workplaces, snack bars or vending machines are frequently used to sell snacks, fruit juices, and soft drinks. These places also serve as important gathering points for breaks and for interpersonal communication. Working from home may have reduced these habits [22]. As expected, our patients reduced the number of meals outside the home, particularly at lunchtime. At the time the second survey was administered, a large number of employees still worked from home, and access to places to eat outside the home was limited, particularly at lunch time.
It has been previously demonstrated that eating out of the home may be an important risk factor for lower micronutrient and higher energy intake (mainly from fat) [23]. In contrast, a study of adults from 11 European countries showed that eating at work is similar to eating at home. Only the higher consumption of alcoholic beverages was responsible for the higher caloric intake of those who ate mainly in restaurants [24]. However, our data did not detect any substantial increase in alcohol consumption, supporting the first of both aforementioned studies [23].
We observed some changes in the physical activity habits of our participants. Although the type of sport and the amount of time spent on physical activity remained the same, the time of day that sports were practiced changed, with an increase in morning hours. While previous studies suggested a reduced physical activity during lockdowns [25,26], we observed a "return to normal" with regard to sports practice. One explanation is that the more liberal free-time management involved in smart working allowed people to exercise more often in the morning.
Our study sample shows an increased frequency of sleep disturbances, especially in the first phase of sleep (falling asleep). This suggests that the pandemic has had longlasting and substantial effects, since the last lockdown ended months ago. These data confirm other studies conducted during the COVID-19 pandemic, highlighting increased psychological distress and significant sleep disturbances in many individuals [27,28]. It is likely that the proximity of the most severe period of the pandemic left a legacy of uncertainty and psychological distress in our study sample. It will be interesting for future studies to investigate whether these alterations in sleep quality, which in turn have a great influence on dietary quality [29], remain when the pandemic is completely over.
Our study suffers from several limitations that warrant discussion. Many participants did not agree to repeat the survey a second time; as such, the number of participants is limited to n = 118. The questionnaire we used did not consider energy intake, and we did not stratify the sample by BMI. Undoubtedly, the physical state of our participants could have influenced the relationship with food, taste, and eating habits. We also have to consider that some patients had their last dietary examination months before the second survey, which could introduce a certain bias. Finally, all items were self-reported, potentially introducing reporting bias [30]. Despite these limitations, we believe that our results are important and highlight important changes in a very special population, from Lazio, Italy. Our data may support other researchers to understand the sequelae of lockdowns and help in future pandemic planning.

Conclusions
The aim of this study was to evaluate the potential impact of lockdown periods on eating habits and health risk factors through the use of a multi-topic questionnaire. Our results suggest that, overall, participants' food preferences improved slightly (e.g., participants consumed more raw vegetables, more cereals, and fewer sugary drinks). We propose that the increased time spent at home allowed participants more freedom and flexibility in preparing fresh foods. It is conceivable that regular office workers may lack the time to cook fresh meals. Additionally, we highlighted that participants in our sample ate out less frequently, especially at lunchtime. The reduced exposure to snack bars or vending machines (which are frequently used to sell snacks, fruit juices, and soft drinks) may also have contributed to improved eating behaviors. Moreover, physical activity in the morning hours increased substantially. Again, greater flexibility and a more liberal day schedule may have contributed to the increased morning physical activity. However, in light of the crosssectional nature of our study, this assumption requires further investigation. Ultimately, we also highlighted several negative aspects, e.g., an increase in reported difficulties in falling asleep, probably as a consequence of pandemic-induced (social) distress and discomfort. Additional trials are warranted to examine whether our results apply to other populations.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/nu13124522/s1, Table S1: Jobs performed by study subjects. Institutional Review Board Statement: The study was approved by the Ethics Committee of the University Hospital "Tor Vergata" in Rome (ID number 44.22) and date of approval is 06 April 2020.
Informed Consent Statement: Written informed consent was obtained from the patients to publish this paper.
Data Availability Statement: The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Conflicts of Interest:
The authors declare no conflict of interest.