Eating and sociodemographic data of adults with type 1 Diabetes mellitus during COVID-19’s social distancing

Objective: Evaluate the association between eating habits and sociodemographic factors in adults with Type 1 Diabetes Mellitus (T1DM) during COVID-19’s social distancing. Design: Transversal, descriptive and analytical study carried out in July 2020, using a form on the platform Google Forms®. Sociodemographic data, information on social distancing and food choices were collected. For statistical analysis it was performed the Chi-Square Test, considering p<0.05. Results: There was an association between social distancing and being female (p=0.014), living with individuals with diabetes (p=0.037) and having income maintained during the pandemic (p=0.032). Being female was associated with increased consumption of snacks (p=0.003) and sweets (p=0.016). Being between 18 and 24 years old was associated with changing food consumption (p=0.001), increasing consumption (p=0.001), especially snacks (p=0.005). Being over 45 years old was associated with maintaining the eating pattern (p=0.001). Having income maintained or increased was associated with maintaining the eating pattern (p=0.042). Conclusion: Age, gender and income can influence the eating habits of adults with T1DM during social distancing, which alerts to the need to maintain nutritional assistance in order to ensure a healthy eating pattern, as well as to have minimum income guarantee programs to ensure access to good quality food.


Introduction
In 2020, after the World Health Organization declared the Coronavirus Disease 2019 (COVID-19) pandemic, social distancing was widely recommended, in order to control the spread of the Coronavirus and also protect the people most vulnerable to the disease's complications (Bedford & Enria, 2020).
Individuals with diabetes are in the risk group for COVID-19, probably associated with chronic complications and of the lack of glycemic control (Gentile, et al., 2020). In the study by Holmann et al. (2020) it was observed that people with glycated hemoglobin (HbA1c) values above 10% had a higher risk of a worse prognosis and death by COVID-19, both for patients with Type 1 diabetes mellitus (T1DM) and Type 2 (T2DM).
There is a significant impact of food on improving glycemic control and to preventing the early appearance of complications (Brazilian Diabetes Society [BDS], 2019). The study by Verma et al. (2020) demonstrated that the pandemic context can negatively affect the glycemic control of patients with T1DM, one of the causes being poor adherence to the diet.
In the period prior to the pandemic, studies discuss the importance of understanding the sociodemographic conjuncture to guide clinical conduct. According to Andrade et al. (2017), sociodemographic factors play an important role in the adherence to treatment of people with T1DM, with low glycemic control of these patients being common in Brazil, associated with low education and the perception of low adherence to diet and insulin therapy. Furthermore, Mendoza et al. (2018) concluded in their study that low and medium levels of food insecurity in people with T1DM are associated with higher levels of HbA1c. Marsola et al. (2020), when evaluating the reason for food choices and how socioeconomic characteristics and risk perception of chronic diseases affect them, observed that socioeconomic factors, such as income and education, strongly affect food choices.
Thus, despite a wide discussion about the importance of glycemic control in people with T1DM in COVID-19 pandemic, there is still a scarcity of studies that evaluated food choices and sociodemographic aspects by this public during the pandemic of COVID-19. Therefore, this study aims to assess the association between eating habits and sociodemographic factors in the context of social distancing in people with T1DM.

Study type and participants
Transversal, descriptive and analytical study carried out in July 2020. Transversal studies look forward to investigate variables in only one moment and they also allow the investigator to observe the variables without follow-ups and interventions in the sample. In addition, descriptive studies are characterized by describing in details the variables analyzed and analytical studies aim to analyze relations between the variables (Zangirolami-Raimundo et al., 2018). The call for participants took place through the internet in social networks of a research and extension project linked to a public university and in several other social networks that approach the T1DM theme (Facebook®, Instagram® and WhatsApp®).
Individuals aged 18 years old or over, diagnosed with T1DM and who agreed to participate in the online survey were included, selecting the option "I read the Informed Consent Form (ICF) and accept to participate in the survey" available at the beginning of the form. Then, the participants had to select which condition they fit, among the options: being an adult and having a diagnosis of T1DM; not having diabetes; have other types of diabetes; being a child or adolescent with T1DM; being a caregiver of a person with T1DM. If the selection did not correspond to the audience of adults with T1DM, the survey was automatically closed. Therefore, people who marked any alternative other than the inclusion criteria, people under 18 years of age, those who did not complete the form filling out or who did not agree with the ICF were excluded from the study.

Data collection
An online form was created on the Google Forms® platform, in the opinion survey format, in accordance with Resolution 510 of 2016, of the National Health Council (Brazil, 2016). The following information was collected: -Sociodemographic data: sex, age, housing region (state, city), level of education, number of people with diabetes at home.
-Financial status: family income and receiving emergency aid.
-Eating during quarantine in relation to pre-pandemic pattern over a 30-day period: change in eating (has changed a lot; changed a little; or did not change); increase in food consumption (increased a lot; increased a little; remains the same; or decreased); number of daily meals (less than five meals; or greater than or equal to five meals); fruit consumption (adequate, for the consumption of two to three servings daily; and inadequate, for the consumption of less than two servings or greater than three servings daily); consumption of vegetables (adequate, when consumption was greater than or equal to two daily servings; and inadequate, when consumption was less than two daily servings); consumption of snacks, considering: packaged snacks, microwave popcorn, stuffed biscuits, biscuits or salty biscuits (more than before; equal to or less than before); consumption of ultra-processed foods, considering: ready-to-eat frozen foods such as nuggets, cheese bread, pizza or pasta (more than before; equal to or less than before); consumption of industrialized beverages (more than before; equal to or less than before); consumption of sweets (much more than before; a little more than before; the same as before; or decreased); and delivery orders (do not order or order less; order the same; or more than before).
-Social distancing: practiced total social distancing (was not going out for any activity); partial (going out only to buy food and/or medicine); did not practice social distancing, as they needed to go out to work; did not practice social distancing because did not agree with isolation; practiced social distancing for family reasons, despite not agreeing. Research, Society and Development, v. 11, n. 2, e54911225847, 2022 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v11i2.25847 4 Informed consent was obtained from all individual participants included in the study.

Statistical analysis
For statistical analysis, the Statistical Package for Social Science software, version 21.0 was used (SPSS Inc., Chicago, IL, 2012). The Kolmogorov-Smirnov normality test was applied to verify the sample distribution. For descriptive analysis, measures of central tendency and dispersion were calculated. For analytical statistics, the social distancing variable had its categories grouped in order to classify into "yes", I was doing it, and "no", regardless of the reason. Pearson's Chi-Square test was performed with adjusted residual analysis (statistical significance p<0.05).

Results and Discussion
472 people were evaluated with a mean age of 30.24±9.74 years old, of which 86% were women (p<0.000). Most lived in the Southeast region of Brazil (47%; p<0.000). About 85% had adhered to social distancing (p<0.000). As for education, 26.9% had completed higher education (p<0.000). Regarding income, it was observed that 32.4% had an income between 3 to 5 minimum wages (p<0.000) and most continued in the same income range during the pandemic (50.6%; p<0.000). Regarding the aid provided by the government, during the pandemic period, 39.2% reported having received it (p<0.000), in which 59.7% (p<0.000) stated that the benefit was used to pay basic bills (water, energy, gas, internet or telephone) -and 57.5% of these used the money to purchase food (p<0.000).
Regarding food during social distancing, 51.5% reported that food consumption changed a little and 40.5% reported that food consumption increased a little (p<0.000). Regarding the purchase of food, 76.9% reported buying food in the same amount as before and 47.7% had 3 to 4 meals a day (p<0.000). As for the consumption of fruits and vegetables, most consumed 2 to 3 servings of each food group, with 42.6% referring to the consumption of fruits and 53.4% to vegetables (p<0.000). It was observed that most participants continued to consume ultra-processed foods (38.1%; p<0.000), industrialized beverages (45.3%; p<0.000), snacks (38.6%; p<0.000), sweets (37.1%; p<0.000) and delivery orders (26.9%; p<0.204) equally as in the period before the pandemic.
It was observed that being female (p=0.014), living with another person with diabetes in the same household (p=0.037) and having a family income maintained or increased during the pandemic (p=0.032) was associated with social distancing. In addition, having reduced income in this period was associated with not making social distancing (p=0.032) ( Table 1). No association was found between adherence to social distancing and eating habits adopted during the pandemic (Table   2). It was observed that being female was associated with having higher consumption of sweets consumption than before and being male, it was associated with maintaining the usual pattern (p=0.016). It was observed that having higher consumption of snacks than before was associated with the female sex and consumption equal to or less than before associated with the male sex (p=0.003) ( Table 3).
It was found that being between 18 and 24 years old was associated with changing one's food consumption. As well as being aged between 25 and 44 years old was associated with making few changes and not having changes in the eating pattern, and being aged 45 years or over was associated with not making alterations in eating (p=0.001). Furthermore, it was observed that being between 18 and 24 years old was associated with having increased food consumption. As well as being aged 25 to 44 years old was associated with having a little changes or decreased consumption, and being 45 years old or more was associated with having the same or decreased consumption (p=0.001). There was an association between being aged 18 to 24 years and Research, Society and Development, v. 11, n. 2, e54911225847, 2022 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v11i2.25847 consumption of snacks higher than before; as well as being aged from 25 to 44 years was associated with the consumption equal to or less than before (p=0.005) (Table 3). It was observed that having the income maintained or increased was associated with maintaining the eating pattern during social distancing (p=0.042) and having had a decrease in income was associated with making changes in eating (Table   4).

Discussion
In the present study, it was possible to evaluate the association between eating and sociodemographic factors of people with T1DM in the context of social distancing due to COVID-19 in Brazil. It was observed that being female was associated with performing social distancing, which corroborates the findings of Coriou et al. (2020), who in seeking to identify barriers and facilitators of adherence to social distancing, observed that women were more likely to avoid socialize in person and keep a safe distance when in public. Data from the Ministry of Health indicate that women, in general, express greater concerns and care for their health, and attend the health system more, seeking not only care for themselves, but also for family members, neighbors or friends (Brazil, 2011). Therefore, the hypothesis is suggested that due to this profile, they were taking the measures of social distancing more seriously when compared to men.
Living with another person with diabetes was associated with the practice of distancing. According to Gentile et al. (2020), COVID-19 mainly affects men in advanced age and with the presence of comorbidities such as DM. Besides, in China the mortality rate in people with DM was 7.3%, being considered high (Wu & McGoogan, 2020). Therefore, the hypothesis is suggested that this high mortality rate, in line with the characteristics of the most affected groups, are factors that generate fear and concern to individuals and that may explain the association with the greater practice of social distancing.
The equal or increased income was associated to practicing social distancing. As Afifi et al. (2020) discusses, the social conditions of individuals directly affect their ability to avoid the disease or diminish its consequences. These conditions are related to the level of financial resources, knowledge and power of individuals, and interfere or make impossible the recommended prevention strategies, such as hand washing, social distancing, orders to stay at home and isolation in case of infection. As well as Burnquist et al. (2020) stated that those most affected by COVID-19 are those workers whose families depend on income from daily activities, since family support depends on the service performed, and thus, it is difficult to practice social distancing without a source of income to ensure basic needs.
Associations were found between age and changes in food consumption. Ruiz-Roso et al. (2020) observed that older individuals with T2DM did not change consumption compared to younger ones, and it is argued that this may be due to the higher level of awareness regarding the consequences of inadequate eating habits in the metabolic control of DM. Moreover, when evaluating knowledge, attitudes and practices regarding COVID-19 in young adults with T1DM, Pal et al. (2020) mention that young people are in the transition between pediatric and adult care, which makes their engagement in health-related behaviors more unstable, besides being more likely to engage in behaviors that pose health risks.
It was observed that decreased income was associated with alterations in eating. In the study by Grabia et al. (2020), in which the majority of the sample was composed of individuals with T1DM, 21% were unemployed and the rest were students, of whom 60% affirmed to have improved their eating habits, but, the authors did not evaluate the association between these variables. However, in deriving trajectories of eating alterations during isolation and associated risk factors in UK adults, Herle et al. (2021) found that people with lower incomes were less likely to eat more during social distancing, probably due to financial problems.
No studies were found with individuals with T1DM that made an association between age and increased food consumption, but when evaluating the eating habits and lifestyle changes of the Italian population during quarantine by COVID-19, Di Renzo et al. (2020) observed that 34.4% reported having more appetite during this period and people who had more appetite were younger. In the study by Grabia et al. (2020), the authors found that 65% of the participants started to eat more regularly during the pandemic, especially the main meals. When evaluating the effects of quarantine on people with T1DM and their caregivers, Tejera-Perez et al. (2021) identified that 40.4% of participants reported weight gain associated with increased food consumption. In addition, participating caregivers reported higher than normal glycemic control and insulin requirements.
As identified by Grabia et al. (2020), 40% of the participants had problems controlling their weight during social distancing, probably due to the increase in the number of daily meals.
Associations were found between being female and consuming more sweets. No studies were found to evaluate these associations in people with T1DM, however, in general, Mattioli et al. (2020) mention that in situations of chronic stress, such as the pandemic period, people seek and consume foods with greater energy density. When evaluating eating changes among Brazilians during the pandemic, Malta et al. (2020) observed that women increased the frequency of consumption of sweets.
Thus, when evaluating Saudi women during the COVID-19 pandemic, Al-Musharaf et al. (2020) found that the consumption of sweet foods was one of the indicators of emotional eating and an adaptive mechanism for dealing with the negative emotions of the social distancing period.
No studies were found that evaluated the association of snack consumption with age and gender in people with T1DM, however, in the study by Ruiz-Roso et al. (2020), when evaluating people with T2DM, they observed that during the period of social distancing, women increased their consumption of snacks. Si Hassen et al. (2018), when evaluating socioeconomic factors associated with snacking consumption in adults, observed that women snacked more frequently during the week.
In the study by Grabia et al. (2020), more than 30% of individuals affirmed that they had increased the frequency of eating snacks between meals during social distancing. Nogueira Bezerra et al. (2018), when seeking to describe the eating habits of Brazilian people, observed that young adults have a lower frequency of consumption of healthy foods and a higher frequency of unhealthy foods when compared to middle-aged adults. Corroborating the results, the study by Malta et al. (2020) with Brazilian adults in the COVID-19 pandemic, young adults increased the frequency of consumption of unhealthy foods during the pandemic.
The present study had the limitations of not having assessed glycemic control, so there is no way to say whether eating changes caused positive or negative changes in this aspect. In addition, conducting an online survey may not reach individuals with T1DM who do not have access to the internet and/or electronic media. However, it was observed a scarcity of studies that explored eating changes and associated factors in people with T1DM during the COVID-19 pandemic, so this study allows us to contribute to the literature and highlight the need for further research addressing this topic, in order to generate a greater understanding of how the context of the pandemic influenced the eating habits of these individuals and other factors that may be associated.

Conclusion
In this study, it was observed the association between social isolation and being female, living with individuals with diabetes and having family income maintained during the COVID-19 pandemic. Furthermore, it was noted that being female was associated with increased consumption of snacks and sweets. Being between 18 and 24 years old was associated with increasing food consumption, especially snacks. Having income maintained or increased during the pandemic was associated with maintaining the eating pattern. It is important to highlight the need for maintenance of nutritional assistance by professionals during this period, in order to ensure a healthy eating pattern, as well as to implement government actions to continue the minimum income guarantee programs to ensure the population's access to good quality food.
Therefore, it is suggested new studies about the effects of the COVID-19 pandemic in the eating habits of people with T1DM, in order to observe the impacts of the changes and their relation sociodemographic factors, and if the adopted behaviors were sustained or modified in a long-term period. In addition, further studies are suggested to assess the association between eating habits and glycemic control during the COVID-19 pandemic.