香京julia种子在线播放

Food is key to personal health [e.g., (1)], as well as to the health of the planet given that current patterns of food production and consumption have considerable environmental impacts (2). Conversely, disasters such as the COVID-19 pandemic can disrupt our food system (3) and change our relationship with food. For instance, in an effort to reduce the spread of infection, border and other logistic restrictions limiting the flow of goods and people increased the risk of food shortages due to impaired supply chains, including those related to labor shortages [as can be seen in the US and Europe, (4, 5)]. Furthermore, the partial or complete lockdown measures introduced at regional and national levels, such as the closure of schools, universities, workplaces, non-essential shops and restaurants, banned events, and travel and mobility restrictions, likely changed the way people accessed their food, where they ate, and how their food was prepared. Some of these measures have served as a further obstacle to the distribution of food to vulnerable populations. For example, some programmes that provide main meals for school children were not operational during confinement. Additionally, quarantine due to illness or coming into contact with infected people may have further restricted people's access to food.

A variety of COVID-19 related psychological changes might have also affected food-related behaviors. Even in areas with relatively low disease risks, people were exposed to extensive communication about the risks of COVID-19, which was likely to have caused some of them stress. Such people may try to cope through stress-related eating, in which they attempt to make themselves feel better by eating or drinking when under stress [e.g., (6, 7)]. For example, during lockdown in Italy, people increased their consumption of processed “comfort foods,” such as chocolate, chips, and snacks (8, 9), and in some cases this was due to anxiety about their eating habits during COVID-19 (10). A study from Denmark also observed a higher degree of emotional eating during the lockdown, e.g., increased consumption of pastries and alcohol (11). In Norway, it was found that consumption of high sugar food and beverages was greater for those with increased COVID-19 related worries and general psychological distress than the overall population (12).

Risk perception associated with COVID-19 may influence people's food purchase and consumption behaviors. For example, people may try to minimize the risk of being infected by increasing their use of delivery services, purchasing more packaged food, which is seen as being more hygienic (8), buying food with a longer shelf-life (and thus purchasing less fresh food), in order to limit their shopping trips, or eating more healthy food in an attempt to boost their immune system [e.g., (13)]. Additionally, people's concern about possible food shortages may have influenced purchasing behavior, e.g., stocking up on certain foods [e.g., (8)].

It has been shown recently that COVID-19 might present additional health risks due to the metabolic impact of overeating under conditions of home confinement (14). Ammar et al. (44) reported an increase in unhealthy eating patterns based on their international survey on physical activity and eating behavior (N = 1,047, April 2020), something that was also observed during lockdown in a Polish national cross-sectional study (N = 1,097) by Sidor and Rzymski (15). About half of the participants reported more eating and snacking, while these tendencies were more pronounced in overweight individuals (15).

In Italy, which was affected much earlier and more seriously by COVID-19 than most other European countries in the first wave of the virus, a total lockdown was introduced at the national level in March 2020. A study by Scarmozzino and Visioli (9) was conducted on 1,939 participants (using a snowballing sampling approach) in April 2020 and showed that about 20% of them gained weight. This study also found and highlighted the increased consumption of processed “comfort foods,” such as chocolate, desserts, and snacks. These observations were partially confirmed by a food consumption study which investigated changes in the sale of food in over 10,000 Italian stores (8), showing an increase in the consumption of pasta, flour, eggs, long-life milk and frozen foods, alongside a reduction of fresh food purchases. This study also reported a drop in the sale of snacks, particularly sweet ones, in relation to homemade desserts, although there was an increase in savory snacks. Interestingly, the results of a COVIDiet Study, conducted on a very large sample (N = 7,514; snowball sampling approach) in Spain (a country also severely affected by COVID-19) showed that confinement in general led to the adoption of healthier dietary behaviors, measured as adherence to the Mediterranean diet (13).

While the above-mentioned studies focused on the general population, some studies specifically targeted younger people. A study of 820 adolescents (aged 10 to 19 years) from Italy, Spain, Chile, Colombia, and Brazil showed that COVID-19 confinement notably influenced dietary habits and modified consumption of both processed foods and fruits and vegetables (16). Gallo et al. (45) investigated the impact of COVID-19 isolation measures on Australian university students and observed increased snacking frequency and the energy density of consumed snacks. Increased energy intake was observed in females (but not males), while physical activity was impacted for both sexes – the proportion of students with “sufficient” physical activity levels was about 30% lower, in comparison with data collected in the years 2018 and 2019.

Studies by consulting companies – addressing changes in shopping behavior during COVID-19 across different product categories (food and others) – reported a marked shift across all categories to “mindful” shopping, “trading-down” to less expensive items (17), and in particular a strong focus on “essentials” (17–19). Groceries was the only product category in which consumers across all countries consistently anticipated spending more (17, 19).

The above literature regarding changes in food purchase/consumption patterns during COVID-19 documents general trends, but does not relate them to specific changes in people's circumstances due to the pandemic and resulting lockdown. Making such linkages is important in order to be able to identify the mechanisms underlying such changes, so that more accurate projections of the effects of COVID-19 can be forecast, and measures can be effectively targeted toward minimizing their negative effects on food consumption. Therefore, the main aim of our research was to understand the changes in food consumption behavior and identify the factors influencing individual changes in the food consumption frequencies of different food categories, such as fresh food, preserved food, sweet snacks, and alcoholic drinks.

To do this, we examined three countries that were similarly affected by COVID-19 infection rates in the first wave, but which varied in the extent of their lockdown measures: namely, Denmark, Germany, and Slovenia. The specific examples of government measures in the three study countries (Supplementary Table 1) illustrate how different households were affected by restrictions in different ways, e.g., not everybody was required to work from home.

To avoid some confounding factors, the study was conducted simultaneously using online panel surveys in late April and early May 2020 in three European Union countries – Denmark, Germany, and Slovenia. The three countries are comparable in terms of all having prompt and extensive government restrictions imposed at the beginning of the pandemic. On the one hand, these lockdown measures considerably limited the spread of the disease at a very early stage of the first wave of the pandemic, but on the other hand, they seriously affected people's lives. Although this paper is focused on changes in food consumption, given the scale of the pandemic and its effects on the food supply system, changes in people's food-related behavior are also likely to have implications for the resilience of food systems.

We developed a conceptual framework of factors that potentially caused changes in food consumption at the level of the individual consumer during the pandemic (Figure 1), building on two strands of literature: food choice process, and behavior change.

The interplay between food-related behaviors forms the core of our framework (Figure 1), i.e., the processes of consuming (what, where, with whom, how often), obtaining (where, how, how often), and preparing food (what, how). Food-related behaviors are influenced by the personal food system, i.e., food-related values and strategies, which in turn are influenced by personal factors, resources, and ideals (20, 21). We introduced a dynamic perspective by recognizing that food consumption during the pandemic is related to food consumption before the pandemic.

The framework further recognizes that individual-level (changes in) food consumption patterns are embedded in a complex system of multilevel factors (22), including the household level and the broader micro- and macro-context (23). We further drew upon dynamic behavior change models (24) based on Bandura's (25) social cognitive theory and concept of reciprocal determinism, postulating that personal, contextual, and behavioral factors create a feedback loop and influence each other. We thus suggest that personal experiences with changes in food-related behaviors during the pandemic potentially influence future behavior after the pandemic and might also lead to changes in personal food-related values and strategies.

Figure 1 maps the most important changes the pandemic (potentially) brought about in the study countries. This illustrates that government restrictions and lockdown measures (along with restrictions imposed by the private sector) had profound impacts on the micro- and macro-contexts of food choice. For instance, the closure of physical workplaces and the closure of schools and day care institutions interrupted people's daily life and potentially changed how, where and with whom individuals ate meals and snacks. The same applies to the closure of work canteens, cafés and restaurants, and the restrictions on private gatherings. Government recommendations to stay at home are likely to have affected how often (and where) people went food shopping.

At the personal level, we expected that the individual risk perception of COVID-19 might have caused changes in food consumption. One proposition is that people concerned about the disease would eat more healthily in order to strengthen their immune system [e.g., (13)]. An alternative proposition is that people anxious about COVID-19 might drink more alcohol and eat more comfort foods, such as snacks and cake, in order to better cope with the situation [e.g., (6, 7, 11). The pandemic also had potential impacts on households' food-related resources, i.e., money and time. Some people faced income losses, e.g., due to reduced working hours. In terms of time, households were affected by the pandemic in very different ways; some people faced severe time constraints while others had more time available for food preparation and consumption than before. In our empirical analysis, we tested the effects that pandemic-related changes at a personal level and contextual changes had on food consumption.

The results chapter starts with a description of the socio-demographic composition of the sample (section Socio-demographic characteristics of the sample) and the main COVID-19 impacts (section Main COVID-19 impacts), before presenting the observed changes in food-related behaviors (section Changes in food-related behaviors), and the analysis of factors significantly related to increases and decreases of food consumption frequencies (section Factors related to changes in food consumption frequencies).

The main objective of the present research was to map changes in the consumption and purchase of food before and during lockdown of the COVID-19 pandemic in Denmark, Germany, and Slovenia. Broadly speaking, people across the three countries shopped less frequently during lockdown and there was an overall reduction in the consumption of fresh foods, but an increase in the consumption of food with a longer shelf life in Denmark and Germany. However, it is worth noting that within these overall trends, in all food categories there were subgroups of people that exhibited an opposite pattern of behavior, e.g., an increase in the consumption of fresh food. It is interesting that we observed changes in food consumption frequency in all food categories during the pandemic. Depending on the food category, 15 to 42% of respondents reported a change. The findings are in line with another study from Denmark (11). Only a small proportion of respondents (15% in DK; 14% in DE; 8% in SI) reported no change across the eleven food categories investigated. About half of the participants in Denmark and Germany and two-thirds in Slovenia reported changes in consumption frequency of three or more food categories, which represents a shift in food choice patterns. The higher rates of change in consumption frequencies in Slovenia were perhaps due to the fact that Slovenia had stricter restrictions in place compared to Germany and Denmark, and is likely to be related to the fact that greater reductions in grocery shopping frequency were reported in Slovenia than in the other two countries.

What is interesting is that differences in individual consumption levels before the pandemic tended to even out during the pandemic, perhaps because many people were facing similar conditions during the lockdown period, leading to a convergence of eating patterns, and so individual differences in macro- and micro- contexts that exerted a greater influence pre-pandemic had less of an effect during lockdown. While it is widely recognized in the literature that food choices are dynamic and evolve over the life course (28), they are also considered to be fairly stable and largely driven by habit when looking at shorter time spans. Significant shifts or turning points in food choice patterns are usually initiated by major life events such as leaving school, changing employment, or entering/leaving personal relationships (29). Our findings suggest that the COVID-19 pandemic and the related restrictions impacting on people's daily lives caused – at least temporary – shifts in food choice patterns for a large part of the population. This is an interesting finding suggesting that the COVID-19 pandemic potentially had similarly large effects on food transitions as major life events.

The question then becomes: are the observed changes a temporary phenomenon limited to the lockdown period in spring 2020, or will the changes last and become new habits? Theories on behavior change suggest that positive experiences with the behavior in question are strong drivers of enduring behavior change; negative experiences, however, are likely to have an opposite effect (24). For example, people who try a new vegetarian dish, and happen to like the taste of it, are likely to eat this food again and change their eating patterns toward more vegetarian dishes, in contrast to people who try the same dish but do not like its taste. In our study, it remains open as to what extent the respondents perceived the changes in food consumption they underwent as positive or negative experiences. It is also worth noting that repeated exposure to food has been proven to be a powerful driver of food choice in the literature on food choice trajectories (30). We therefore expect that food shifts experienced in the spring of 2020 will partly lead to new food habits, in particular since the COVID-19 pandemic and related restrictions will continue beyond the year 2020 in many countries around the world.

It is thus worthwhile looking at the observed changes in the different food categories in more detail. Interestingly, we were able to identify factors driving changes in individual food consumption frequency at food category level by analyzing pandemic-related contextual and personal factors as well as socio-demographic variables. The conceptual framework we developed proved very useful. The factors that were found to be associated with changes in food consumption in a more consistent pattern were decreased shopping frequency, loss of income due to the pandemic, COVID-19 risk perception, closure of physical workplaces, cafés and restaurants, having a household with children, and gender. Consumption levels before the pandemic constitute an important control variable. Overall, our findings illustrate how food choice is largely influenced by contextual factors, i.e., where and with whom people eat, and by food shopping frequency.

People tended to reduce their intake of fresh food during lockdown, particularly fruits, vegetables and meat, which is consistent with the findings of Bracale and Vaccaro (8) based on the sales data of Italian food stores. This was related to a reduction in shopping frequency in all three countries, with the greatest decrease occurring in Slovenia compared with Denmark and Germany. This could be because a government recommendation to reduce shopping frequency was in place in Slovenia but not in the other countries, together with travel restrictions (people were only allowed to travel within their municipality of residence). Moreover, it should be noted that Slovenia shares a border with Northern Italy, which was very strongly affected by COVID-19 in the spring of 2020, and this affected policy measures and, presumably, also people's behaviors. With fewer government restrictions in Germany and Denmark, we see that the reduction in shopping frequency appeared to be at least partly attributable to people's anxiety about COVID-19. That is, the greater they perceived the risk of COVID-19 to be, the less frequently they shopped (perhaps to minimize risk of infection), and thus they tended to consume less fresh produce given its perishable nature.

In Slovenia, respondents from households that had lost income due to the pandemic were more likely to decrease their fruit and vegetable intake, perhaps because income loss was much more pronounced in this country (53%) than the others (DK: 9%, DE: 23%). This household income loss in Slovenia was also associated with an increased intake of bread, and a decrease in ready-made meals and alcohol, suggesting that those who had reduced income may have shifted toward buying more affordable food.

The observed reduction in the consumption of fresh food is in contrast to the results of a study showing an increase in healthier dietary patterns (including increased intake of fruit and vegetables) in Spain during the COVID-19 confinement (13). However, it is important to note that the sample of the Spanish study comprised over 70% women (compared to the current study that was more or less representative in terms of gender) and our results show that women in all three countries were more likely to increase their intake of fruit and vegetables during lockdown. Although women generally consume more of these types of foods than do men [e.g., (31, 32)], it is interesting that this difference was further emphasized during the lockdown. This may be because women feel more strongly that eating fruit and vegetables is good for health, and they feel that they have more control over this behavior, which in turn causes them to consume more fruit and vegetables (33). Given that the pandemic likely highlighted people's health concerns, this could be why the gender difference in fruit and vegetable intake increased even more during lockdown. It remains to be seen to what extent such dietary differences may contribute to the greater severity of clinical outcomes of COVID-19 in men compared to women [e.g., (34, 35)].

Against the trend of an overall reduction in fresh foods, households with children tended to increase their intake of fresh produce in Germany during lockdown, and to a lesser extent in the other two countries. This may be due to households having to replace school lunches during lockdown.

Although people consumed more ready-made meals, sweet snacks and alcohol in Denmark and Germany during lockdown, there was no such increase in Slovenia, with even a decrease in ready-made meals. Other data from this study suggests two possible reasons for this. Firstly, because there appears to be greater anxiety in Slovenia about acquiring food than in the other two countries, and so consumers in Slovenia were less likely to increase spending on less essential food products. Consistent with this, loss of income was related to a decrease in non-essential food in Slovenia but not in Denmark nor Germany. Secondly, because households in Slovenia suffered greater income loss due to the pandemic (and so presumably had less work), participants may have had more time to prepare meals than those in Germany and Denmark.

Unsurprisingly, the closure of cafes and restaurants was associated with increased consumption of ready-made meals in Germany and Slovenia among those who had eaten out at least once a week before lockdown. Inversely, the consumption of such meals actually decreased in Germany, Slovenia, and to some extent Denmark, for those whose physical workplaces had closed. This might suggest these people had more time for meal preparation since they spent more time at home compared to their daily routine before the pandemic. However, given that there was no consistent increase in the consumption of fresh food amongst this segment, this suggests that they were not using more fresh ingredients when eating from home.

There was an overall increase in the consumption of sweet snacks in Denmark and Germany, although not in Slovenia. The rise in snack consumption is consistent with a study of Canadian middle to high income families with young children who also commonly reported consuming more snacks during the COVID-19 pandemic (36). Further, and as noted above, it is likely that the large income loss experienced in Slovenia mitigated against purchasing and eating snacks given that they tend to be a relatively expensive food item.

However, it is also worth noting that the relationship between lockdown and snacking is not always consistent – with for example, one Italian study showing an increase in the consumption of sweet snacks (9), whereas another Italian study showed a decrease in the sales of sweet snacks (8). This may partly be because people's snacking behavior may change over time the longer lockdown continues (37). Furthermore, what is considered to be a snack may not always be clear to consumers, as the same food may be a snack or part of a meal depending on when and how it is consumed.

A major strength of this study is its controlled sampling conducted over a short duration and at the same time in the three case countries. In contrast to most other studies where convenience or snowball sampling was used, the use of online panels enabled the collection of national samples which are more comparable with census data in terms of gender, age, and geography than other samples. However, our samples also exhibit deviations from census data in terms of age and education. It should be noted that strict lockdown measures were in place during data collection, so that an online study was, in practice, the only possible method. Quota sampling secured valid responses geographically across all three countries. Such an approach is, however, also subject to some limitations. For example, there was ready access to populations with internet connections but this probably mitigated against the inclusion of participants with lower socio-economic status. On the other hand, the compatibility of the survey with handheld devices should have reduced this problem to a good extent, and all three countries have relatively good internet infrastructures and digitally literate populations compared with the European average. Even so, it is likely this explains the difficulties in achieving representativeness of the study samples in terms of educational level. While in Denmark's sample the educational level is very similar to the country average, we observed slightly over-representation of higher education levels in the German and Slovenian datasets.

The current findings suggest that it is more difficult for people to eat healthily during a COVID-19 confinement in terms of fresh fruit and vegetables. However, the fact that their reduced consumption is linked to reduced shopping frequency, and to some extent, greater risk perception of COVID-19, suggests that increasing the accessibility of such produce, such as through deliveries or pick-up points, may be a way of increasing their consumption. Although online delivery systems may be overburdened by individual orders in many cases, measures could be implemented to prioritize the more frequent delivery of more essential foods. This might include having a standard care package of food that could be produced/prepared in bulk, which could also increase the affordability of such produce. Furthermore, the delivery of such items could be prioritized by banning the delivery of non-essentials. For example, by temporarily banning the sales of bottled water, the online British supermarket chain Ocado was able to deliver to 6,000 additional homes during the pandemic (38).

The reduction in the intake of fresh fruit and vegetables during lockdown was also related to gender and loss of income. One factor that has been shown to increase fruit and vegetable intake in previous studies is the use of food banks, which provide such food at a nominal price (39, 40). Relatedly, providing fruit and vegetables for free increases their uptake among socially disadvantaged men (39) and families (41). Although our data show overall only small increases in the use of food banks and anxiety about acquiring food especially amongst participants with a lower education, such changes are significant and might be expected to be higher over a longer timeframe than the initial 5–6 weeks of the pandemic covered by this study. Differences between the three countries also show that as more households have experienced income loss, this has gone hand-in-hand with an increase in food anxiety and a greater reduction in the purchase and consumption of fresh fruit and vegetables. This suggests that those suffering from the greatest income loss shift to more affordable food and away from fresh foods, whilst those with much lower income loss instead tend to seek compensation for the huge inconveniences of lockdowns and changing routines by shifting to greater consumption of ready-made meals, sweets and alcohol, echoing the findings of a US American study among low-income adults (42). This has important policy implications, for example through the provision of greater income support for those who cannot work because of the pandemic and/or for vulnerable households that have the largest squeezes on their incomes. In terms of research, it is clear these issues also require further detailed examination.

Because women were more likely than men to increase their consumption of fruit and vegetables during lockdown, further research could examine whether such increases are due to the fact that women more strongly perceive the health benefits of these foods, and thus may increase their intake in order to support their immunity. If this is indeed the case, then perhaps increasing the awareness of the benefits of fruit and vegetables in helping to protect against COVID-19, particularly among men, may also help to increase the consumption of fruit and vegetables.

Other issues with important policy implications and that require further research include the composition of households which show important differences, for example in terms of whether or not younger children are present. Our study shows that many such households simultaneously tend to both increase their consumption of fruit and vegetables and of sweets and chocolate. These households are likely to require even more focused support than other groups, both in terms of income and of food advice, especially when they fall into the more vulnerable population categories.

Many aspects of this study have shown the importance of local and national food cultures in determining both before- and during-pandemic patterns of food purchasing, preparation and consumption and thus how people change their food behavior. Policies need to be more nuanced in taking account of these differences, working with them where they help shift people to more healthy and sustainable food, as well as attempting to mitigate their influence when they work the other way. Much more research is needed on this issue.

Finally, the results reported in this paper have revealed, as expected, the critical importance of how lockdown measures, as well as the response of the numerous actors along the food value chain, alongside food culture, have impacted moves toward or away from more healthy and sustainable diets. The paper has thrown up many unanswered questions, as well as relevant insights, so it is clear that much more research is required on these issues if policy makers are to be more successful in improving food wellbeing and reducing the negative environmental effects of the food system.

Across Denmark, Germany and Slovenia people tended to reduce their consumption of fresh food, except for households with children. This change is related to reduced shopping frequency during the pandemic in all countries, and increased risk perceptions of COVID-19 in Denmark and Germany. In Denmark and Germany, people also increased their intake of non-perishable foods, but not in Slovenia, which may have been due to more people in Slovenia being affected by loss of income. Those who ate out once a week before the pandemic increased their intake of convenience foods in Germany and Slovenia during the first lockdown. Women were more likely than men to increase their intake of fresh fruit and vegetables.

The results presented here suggest that changes in eating behavior during the first wave of the pandemic were driven by contextual factors such as lockdown conditions, and personal factors such as anxiety related to COVID-19, loss of income, household composition, and gender. These results help to identify populations that are particularly vulnerable to nutritional changes during the pandemic, and potential avenues that could be explored to minimize the negative effects of the pandemic on food intake in consumers.

Requests to access the datasets should be directed to Jeremy Millard, jeremy.millard@3mg.org.

The studies involving human participants were reviewed and approved by Research ethics officer, Danish Technological Institute, Aarhus, Denmark; Research ethics officer, Copenhagen Business School, Department of Management, Society and Communication, Frederiksberg, Denmark; Bioethical committee at the Higher School of Applied Sciences, Ljubljana, Slovenia. The participants provided their written informed consent to participate in this study.

MJ and HH analyzed the data with assistance from JM. All authors contributed to early drafts which were then collated, rewritten and substantially updated by MJ. All authors read and approved the final manuscript, interpreted the data, and made final revisions.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.