Weight change-related factors during the COVID-19 pandemic: a population-based cross-sectional study using social cognitive theory

Introduction

Obesity is an emerging global pandemic that impacts both physical and mental health. Its prevalence increased two-fold in the last 40 years, regardless of age, gender, ethnicity, or social status. If current trends continue, half of the world’s population will be obese by 2030 (Mehrzad, 2020). Obesity is linked to several co-morbidities, including type-2 diabetes, osteoarthritis, cardiovascular disease, and psychological disorders, which have significant economic and health consequences (Dixon, 2010). On March 11, 2020, the World Health Organization (WHO) declared the Novel Coronavirus-19 (COVID-19) outbreak a global pandemic (Cucinotta & Vanelli, 2020). This pandemic necessitated lockdown, quarantine, and social isolation measures, including remote work and outdoor activity restrictions (Rehman & Ahmad, 2020), affecting people’s behaviors, weight, and overall health. There was a spectrum of weight change ranging from weight loss (35% of the population), no weight change (42%), to weight gain (23%) (Deschasaux-Tanguy et al., 2020; Di Renzo et al., 2020; He et al., 2020; Matsungo & Chopera, 2020; Scarmozzino & Visioli, 2020; Sidor & Rzymski, 2020; Zachary et al., 2020). Whether weight gain or loss, the average weight change was 2 kg during the weeks of lockdown (Deschasaux-Tanguy et al., 2020).

Current literature has shed light on some predictors of weight gain during the COVID-19 pandemic. The individual’s behavior change could explain weight gain during the COVID-19 pandemic. Individuals who gained weight were more likely to snack (ALMughamis, AlAsfour & Mehmood, 2020; Deschasaux-Tanguy et al., 2020; Sidor & Rzymski, 2020; Zachary et al., 2020), eat ready-made food, fast food, meat and dairy products (Al-Musharaf, 2020; Rehman & Ahmad, 2020; Scarmozzino & Visioli, 2020; Sidor & Rzymski, 2020), eat fewer fruits and vegetables (Di Renzo et al., 2020; Rehman & Ahmad, 2020), drink more alcohol (Cucinotta & Vanelli, 2020; He et al., 2020), smoke more (7, 11, 12) and exercise less (Cucinotta & Vanelli, 2020; He et al., 2020; Zhang et al., 2020). Furthermore, during the COVID-19 lockdown, stress has been linked to weight gain (Scarmozzino & Visioli, 2020), with higher levels of depression predisposing to weight gain (Deschasaux-Tanguy et al., 2020). Surprisingly, higher anxiety levels were protective against weight gain and linked to weight loss (Matsungo & Chopera, 2020). Most importantly, there was a complex interaction between obesity before the pandemic and weight gain during the lockdown. Being of average weight before the pandemic had a higher impact on weight gain during lockdown (ALMughamis, AlAsfour & Mehmood, 2020; Deschasaux-Tanguy et al., 2020; He et al., 2020). One explanation is that people who are overweight or obese have a mentality that is geared toward higher levels of concern with shape and weight, making them more resistant to weight gain during the pandemic (Haddad et al., 2020). Finally, female gender and young age (under 35 years old) were associated with weight gain during COVID-19 lockdowns (Deschasaux-Tanguy et al., 2020; He et al., 2020; Matsungo & Chopera, 2020; Sidor & Rzymski, 2020).

Identifying weight change-associated factors during the COVID-19 pandemic aids in developing targeted interventions to reverse behavior and prevent health consequences. The current literature has failed to depict a comprehensive understanding of the predictors, instead focusing on one or two predictors at a time, lacking the context of a framework. A theoretical framework is essential to guide the selection of the associated factors in such a complex multifactorial behavior change. For example, a systematic review of internet-based behavioral interventions has found that interventions that used theory to select participants tended to be more effective (Webb et al., 2010). The Social Cognitive Theory (SCT), one of the most extensive dynamic models used for behavior change, describes how people adjust their behavior through control and reinforcement to achieve sustainable goal-directed behavior (Bandura & Walters, 1977). The theory highlights the interplay between internal and external social factors, including cognitive, personal, environmental, and behavioral factors. Therefore, this research explores a broader range of factors associated with weight change during the COVID-19 pandemic using the SCT framework based on behavioral, environmental, and cognitive themes. Based on our findings, multiple interventions could be instituted to reduce quarantine’s acute and chronic impact on health and weight during this and future pandemics.

Materials and Methods

This study is a quantitative, cross-sectional, survey-based study conducted via an anonymous online questionnaire between July 15, 2021, and October 10, 2021. Participants were recruited from the list of American University of Beirut-Medical Center (AUBMC) patients who are active on the electronic medical record patient portal and have shown interest in being involved in research projects. The targeted population received a message notification at their portal. Three reminders were sent weekly. Furthermore, recruitment occurred through a post on AUBMC’s and researchers’ social media accounts. The following were the inclusion criteria: Lebanese adults at or above 18 years of age residing in Lebanon. Exclusion criteria include individuals under 18 years, women who have been pregnant during the previous year because their weight change may not be related to the COVID-19 pandemic, individuals with a nationality other than Lebanese, and Lebanese individuals residing outside Lebanon.

The survey included four domains: demographics, cognitive, behavioral, psychological, and environmental factors (Supplement 1). Appendix 1 shows the mapping of the various variables to the theoretical framework. Multiple validated tools were included in the survey: Generalized Anxiety Disorder-2 item (GAD-2), Patient Health Questionnaire-2 (PHQ-2), General Self Efficacy Scale (GSES), Alcohol Use Disorders Identification Test-Concise (AUDIT-C), and the dietary pattern evaluation tool.

Both the GAD-2 and PHQ-2 are first-step screening tools for generalized anxiety disorder (GAD) and major depressive disorder (MDD) for their ease of comprehension and administration (Hughes et al., 2018; Kroenke, Spitzer & Williams, 2003). PHQ-2 questionnaire is one of the most widely validated screening tools for major depressive disorder (MDD), with high sensitivity of 85% and specificity of 95% at a cutoff of three or more. The sensitivity and specificity of GAD are 76% and 81%, respectively.

The GSES evaluates an individual’s perceived self-efficacy to foresee mechanisms of handling and adapting to new stressful situations. It is a validated, 10-item questionnaire, scored according to a 4-point Likert scale for each item. A total score is the sum of the 10 items, ranging between 10 and 40, with higher scores indicating higher perceived general efficacy (Kusurkar, 2013). It has good scale reliability, with Cronbach’s alpha of 0.84 (Kusurkar, 2013). It was validated among a similar population of Qatari females (Crandall, Rahim & Yount, 2016).

The AUDIT-C is a tool adapted from the original 10-question AUDIT scale. It comprises three questions, scored on a scale of 0–12 (with zero indicating no alcohol use), identifying individuals at risk of hazardous drinking or alcohol use disorder. The cutoff in women is 3 and 4 in men, and the higher the score, the more likely the individual’s drinking affects their health (Reinert & Allen, 2007). It has good test-retest reliability with inter class correlation coefficient of 0.91.

The dietary pattern evaluation tool is a series of simple, easily administered 34 questions that allow assessment of dietary behavior patterns. It highlights four dietary patterns relevant to the COVID_19 pandemic: high fat/high calorie, overeating/binge, dietary impulse, and unbalanced food intake. The questions were scored on a 1–5 scale (1: not at all; 2: no; 3: average; 4: yes; 5: very much so) with no assigned cutoff (Do Lee et al., 2016).

Results

Discussion

This quantitative, cross-sectional online survey-based research aimed to determine a broader range of predictors of weight change during the COVID-19 lockdown in the Lebanese population in the setting of the Social Cognitive Theory framework. The factors included general demographic, behavioral, psychological, environmental, and cognitive factors. Weight change was along a spectrum from weight loss (28.7%), fluctuation (20.4%), stable weight (22.2%), and weight gain (28.7%), similar to what was reported in other studies (Deschasaux-Tanguy et al., 2020; Di Renzo et al., 2020; He et al., 2020; Matsungo & Chopera, 2020; Scarmozzino & Visioli, 2020; Sidor & Rzymski, 2020; Zachary et al., 2020). The mean weight change was −7.0 ± 6.0 kg and 6.4 ± 5.0 kg in the decreased and increased weight groups. Multinomial logistic regression had shown that it is more likely to have weight increase, as compared to staying the same, with higher scores on overeating/binge eating and unbalanced food pattern. It is more likely to have weight decrease, as compared to staying the same, with higher baseline BMI, higher anxiety scores, and being a smoker.

Our model has shown that different factors affected weight increase or decrease during the COVID pandemic. Similar to previous studies that have highlighted that weight gain was related to increased eating, especially snacking and fast food, our model has shown that overeating/binge eating and unbalanced food patterns were the most important factors. Nevertheless, most participants had access to local groceries (92.2%) and homemade food (87.4%), and very few ordered fast food (21.6%). Accordingly, eating behaviors were not related to environmental factors, such as access to fast food or groceries while staying home during the multiple COVID-19 lockdowns, but instead may be related to one’s own approach to food which is the product of the levels of anxiety and personal attributes of the individual (smoking and baseline BMI). Along the same line, previous studies have highlighted the protective effect of anxiety and baseline BMI against weight gain (ALMughamis, AlAsfour & Mehmood, 2020; Deschasaux-Tanguy et al., 2020; Haddad et al., 2020; He et al., 2020; Matsungo & Chopera, 2020). As probably expected during a pandemic, physical activity decreased in half of the sampled population, although most participants had a space to walk. The physical activity decrease was not distinct from the literature (Alqurashi, 2021; Sooriyaarachchi et al., 2021; Yang et al., 2021). In a similar sample in Sri Lanka, a significant decrease in physical activity and increased sitting and screen time were noted, along with one-third of the sample gaining weight (Sooriyaarachchi et al., 2021). An important consideration here is that even though environmental factors did not directly correlate to weight change, however, the COVID pandemic had its impact on the individual’s anxiety regarding their fear related to the COVID-19 virus and concerns related to the economy, including fear of being laid off or experiencing poorer personal economy.

In our study, baseline BMI, overeating or binge eating, unbalanced food pattern, anxiety, and smoking were the main factors of weight change. We hypothesize that weight gain during the COVID-19 pandemic was triggered intrinsically by one’s relationship with food and emotional eating at times of crisis and stress, including natural disasters and eminent fear of infection, such as during the COVID-19 lockdown. For example, different stressful situations have been found to manifest in increased eating, such as the observance of overeating accompanying high levels of distress post-earthquake (Kuijer & Boyce, 2012). Emotional eating is highly associated with the development of obesity, notably during stressful periods (Al-Musharaf, 2020; Konttinen, 2020). In our study, emotional eating was addressed in the form of eight questions entitled “Dietary Impulse Pattern” in the setting of the eating scale. GSES was associated with weight loss in the bivariate analysis (p-value = 0.043) and could be interpreted as the more control one has over their impulsivity to eat and the more refined ability to cope with various stimuli, the better control over weight change.

An intuitive question is whether emotional eating is inherited or acquired (Herle, Fildes & Llewellyn, 2018). Whether it is one or the other does not truly matter because we, as humans, are the product of both our nature and nurture. Still, the latter is the only entity we can control to alleviate the consequences of obesity in this current pandemic and in the future (Freitas et al., 2018). It was evident that the environment and demographic factors did not help one gain or lose weight during the COVID-19 lockdown; however, behavioral factors played an important role along with psychological factors such as anxiety. Weight seems closely related to one’s mental state, general self-efficacy, and eating patterns. Luckily behavior and mental health are entities that could be addressed and, as such, alleviate the effect of the COVID-19 lockdown on weight. Following our results, behavioral support was a major key in weight control during the COVID-19 pandemic (Zaman et al., 2022).

The strength of this research is that it was guided by a theoretical model that includes the interplay between demographic, behavioral, environmental, and cognitive factors. Previously published studies focused mainly on weight gain and its associated factors during COVID-19 lockdown rather than the whole spectrum ranging from weight loss to weight gain (Hammouri et al., 2022; Khubchandani et al., 2022; Seal et al., 2022; Tan, Tan & Tan, 2022). When compiled in the setting of a framework, the factors that drive weight gain are overeating, binge eating, and having an unbalanced food pattern which are all behavioral factors well known to cause weight gain and can, fortunately, be controlled. On the other hand, our study has uncovered factors associated with weight loss, including baseline BMI before the lockdowns, anxiety, and smoking.

Nevertheless, there are limitations to this study. First, the response rate could not be calculated as the participants recruited through the AUBMC health portal, MyChart, and social media is unknown, which may affect the external validity and generalization of the results. Second, the sample encompasses mostly females of high educational and socioeconomic status, mainly residing in two areas of Lebanon (Beirut and Keserwan), a pattern that occurred in similar studies affecting the generalizability of the results. Furthermore, only Lebanese individuals were recruited; hence specific cultural issues could be different in people of different nationalities (Alqurashi, 2021; Sooriyaarachchi et al., 2021; Yang et al., 2021). Third, the survey included self-reported data with inherent recall bias as individuals were asked to recall information after the lockdown had been withdrawn. They may have referred to the current moment to answer the questions regarding their emotions and behaviors.

Moreover, the survey relied on the self-report of the participants. Finally, the findings of this research only show an association between weight change and the associated factors. It cannot provide evidence of causality. Nevertheless, this study helps establish preliminary evidence that can be used to plan future studies when a cohort of participants is followed throughout the lockdown. Approaching participants through a systematic random sampling of neighborhoods may eliminate the selection bias.

Conclusions

COVID-19-related weight change is complex and multifactorial. The COVID pandemic has led to a spectrum of weight changes, with different factors affecting weight gain or loss. Environmental factors related to access to food and walking did not play a major role but instead eating patterns, anxiety, baseline BMI and smoking. Further research is warranted, preferably as qualitative studies, to uncover other factors contributing to the weight change. In future periods of lockdown, it is strongly advisable to spread more education on emotional eating and anxiety reduction through awareness campaigns to help address and possibly control two of the most important driving forces of weight change. This may prepare individuals not only for future COVID-19 lockdowns but maybe other future pandemics and stressors, including other natural occurrences.

Supplemental Information