4. Discussion
This large population-based study presents data on the prevalence of emotional eating and its association with worries and mental distress during the COVID-19 lockdown in Norway and is the first to explore associations between emotional eating and worries related to health and personal economy during a pandemic—findings that are in line with those presented in a study on eating habits during the COVID-19 lockdown period in Italy [4]. Due to limited evidence on the baseline levels of emotional eating in the study population, we are unable to directly compare these numbers with those from pre-COVID-19 times. However, emotional eating and associated eating behaviors are closely related to stressful life events and to perceived life stress [21,22,23]. The COVID-19 pandemic and the NPIs that were implemented during the lockdown period could represent a significant source of stress to many, and it is therefore not implausible that the prevalence of emotional eating was in fact increased during the study period compared to habitual levels. Literature from another collectively stressful event, namely, an earthquake in New Zealand, showed increased over-eating in subjects that were already prone to emotional eating and who reported high levels of stress related to the event [24].
Female participants reported more emotional eating than men, consistently with findings in other studies [12,25]. An interesting hypothesis relates this gender difference to dietary restraint. Dietary restraint refers to the cognitive effort to control food and calorie intake, as is the case with dieting. In general, women exhibit higher dietary restraint than men [26]. Moreover, experimental studies have shown that dieters eat more when exposed to stress or negative emotion than non-dieters, suggesting that dieting could in fact be a possible risk factor for emotional eating [9,27]. Interestingly, di Renzo et al. found that those who had been dieting prior to the study period were more prone to emotional eating [4].
Worries concerning the consequences of COVID-19 were weakly associated with emotional eating. This association was stronger for worries related to personal economy and job security compared to those related to health and disease transmission. Possibly, participants with health-related worries could be more conscious of their eating behavior and therefore have some inherent resistance to unhealthy and emotional eating. It is also possible that the prospect of negative outcomes on personal economy could have a stronger impact on worriers than the prospect of disease transmission and contraction, i.e., it is possible that the economic worriers were, in fact, more worried. A study comparing the mediation effect of emotional eating on the association of depression with BMI in Denmark and Spain highlighted the role of unemployment in explaining this mediation in the latter country [28]. This was seen in the context of the 2008–2014 Spanish financial crisis, leading to an upsurge in unemployment. During our study period, the COVID-19 lockdown led to a more than four-fold rise in unemployment in Norway [29]. Importantly, there was a social gradient in the loss of work related to COVID-19, where employees with the lowest income were the most at risk of losing their job. In Norway, several measures to reduce economic vulnerability were implemented early in the lockdown period [30]. Thus, it can be assumed that the associations we present could have been even stronger in the absence of such measures.
Emotional eating was strongly associated with psychological distress. This is in line with findings reported in the literature which present positive associations between depressive symptoms and emotional eating [11,12,13,14]. Similarly, the intake of high-sugar foods and beverages was markedly higher in those scoring above the threshold of psychological distress compared to the overall population. Stress is known to induce a shift in individuals to favor energy-dense foods containing high amounts of fat and/or sugar, especially among emotional eaters [7,10]. An experimental study on the mood-enhancing effect of chocolate found that chocolate, in particular, highly palatable milk chocolate, had a short-lived effect on improving negative mood, and that this effect was stronger in high-degree emotional eaters than in the other participants [31]. It is unlikely that eating certain foods has an actual role in enhancing mood and lowering stress other than short-lived reward or relief [21,32]. Still, these short-term effects, or even the belief that eating certain foods will provide comfort, may potentially support a tendency towards stress-induced (over-) eating. This habit could, on the long term, represent an added health burden in the form of adverse weight gain and obesity. With regard to the current pandemic, it is worth noting that obesity and related co-morbidities are emerging as risk factors for poor outcomes in COVID-19 patients [33].
Our study clearly demonstrates that psychological distress is associated with emotional eating and a higher consumption of high-sugar foods and beverages. The same is seen, but to a lesser extent, in those reporting substantial COVID-19-related worries. Moreover, we found that worries related to personal economy and job security had a stronger association with emotional eating compared to worries related to health and disease transmission. To our knowledge, this is the first study demonstrating this. This finding raises the question of whether those vulnerable to financial stress, e.g., those with a lower socioeconomic status, are more at risk of emotional eating during periods of economic uncertainty. In fact, emotional eating could play a role in the association of low socioeconomic status with higher BMI—a role that is dependent on emotional and psychological distress [34,35]. Future research, preferably in the form of large-scale longitudinal studies, could further elucidate this and the possible preventive health measures aimed at populations at risk of making adverse food choices during moments of hardship and distress. Interestingly, recent randomized controlled trials point to dietary improvement as a promising treatment strategy for depression [36,37]. Interventions aimed at healthy eating could have a dualistic beneficial effect on improving eating behaviors, while at the same time reducing psychological distress for those vulnerable to emotional eating.

Strengths and limitations
Our study provides an overview of the prevalence of emotional eating and its association with worry and psychological distress on a large scale amid a worldwide state of emergency. One strength of this study is its large sample size, allowing analyses with high precision and statistical power. Moreover, the study period coincided with the most invasive implementation of NPIs in Norway yet, and this offers insight into a phase that exemplifies the impact of a large-scale pandemic.
One inherent limitation of this study is its cross-sectional design, limiting causal inferences. This also prevents us from drawing conclusions about the effect of COVID-19 and related exposures on outcomes, specifically. Another limitation of this study is that it relies on self-report and therefore is subject to recall bias and dependent on the participants’ own insights. It is also based on relatively few, but validated, questions related to eating habits and could therefore provide a limited range of detail and nuance compared to a larger questionnaire. There is also an inherent selection bias in our study due to the questionnaire being written in Norwegian and distributed solely through digital means, which excludes people without access to the internet and with limited proficiency in the Norwegian language—e.g., elderly inhabitants and first-generation immigrants. Moreover, the senior citizens that did participate in this study are likely more self-reliant and healthier than those in the same age group that were unable to participate, providing a potentially unbalanced view of the oldest age group.