Regarding the clinical and self-care aspects of participants living with T1DM, 43% experienced hypoglycemic episodes in the last year, and 13% required hospitalization due to diabetic ketoacidosis during the same period. These data were presented in Table 2. Nearly a quarter (23.4%) reported having another chronic disease in addition to T1DM, and 66.4% indicated having other health problems. More than half (52.3%) had a perception of control that was considered poor/fair/average in the last year, although the vast majority had undergone medical consultations for T1DM control (96.3%) and underwent HbA1c testing (95.3%) during the same period. A significant number of participants had performed four or more blood glucose tests a day in the past week and reported using only the prescribed amount of insulin in the last thirty days, with 35.5% never experiencing any delays in doses during the same period.

Half of the participants wore or carried something with them that identified them as having T1DM, and 86% carried some food or beverage in case of low blood glucose levels. There were very few smokers (Table 2). Around 50.5% had never participated in groups or associations related to the disease, and 72.9% reported having received support from someone for T1DM care during childhood and adolescence. Table 2 also indicates comparisons between response proportions of clinical and self-care indicators among people living with T1DM with Adequate (≤ 7%) and Inadequate (≥ 8%) control of HbA1c. The analyses indicated that participants with adequate glycemic control (HbA1c ≤ 7%) were more likely to never delay insulin doses and to perform blood glucose testing at a frequency of more than 4 times per day compared to participants with inadequate glycemic control (HbA1c ≥ 8%). In contrast, those with inadequate control perceived their T1DM control as inferior to those who had achieved adequate control (HbA1c ≤ 7%).

A factor analysis was conducted with the clinical and self-care variables to aggregate variability and avoid collinearity among the variables (Table 3). An initial assessment of the factor analysis revealed that two items (“Carries food for diabetes control” and “Did you smoke a cigarette?”) had factor loadings lower than 0.40. Additionally, one item (“Did you need to be hospitalized or go to the emergency room due to diabetic ketoacidosis in the last 12 months?”) had factor loadings higher than 0.40 in two factors. Therefore, a new analysis was implemented excluding these items. Table 3 presents the factor loadings of each evaluated variable and the polarization of the factor loadings. Initially, the model was observed with seven factors, but the scree plot indicated an inflection point in the descending curve at the fifth factor, used as a parameter (Ledesma et al., 2015).

Therefore, a new analysis was generated with five factors explaining 61.23% of the variance. The first factor was named “Daily Testing” and included variables related to insulin and glucose testing, and the second factor was named “Medical Care” because it included variables related to consultations and examinations carried out by people living with diabetes. In each of these factors, only two variables had factor loadings equal to or greater than 0.40, but as they were highly intercorrelated (above 0.80), the factors can be considered reliable (Knekta et al., 2019). The third factor was named “T1DM Control” as it included variables related to a positive perception of T1DM control, not delaying insulin doses, and not using more than what is prescribed; the fourth factor was named “Other Health Problems”, with variables related to the presence of other diseases or health problems; and finally, the fifth factor was named “Support from Others” because it included variables related to family, medical, and social support.

The factor loadings of the model were saved and included as predictors in logistic regression models of clinical and self-care indicators to predict glycemic control in T1DM (Table 4). The independent variables were the standardized scores on the five factors obtained through principal component analysis. The outcome considered was the glycemic control of people living with T1DM, evaluated as 0 = Adequate (HbA1c ≤ 7%) and 1 = Inadequate (HbA1c ≥ 8%). The results indicated that the factors “Daily Testing” and “T1DM Control” were responsible for predicting the odds ratio of having adequate glycemic control, while “Other Health Problems” was responsible for predicting the odds of having inadequate glycemic control. These variables were identified as protective and risk factors for T1DM management and self-care.

To assess the effects of demographic and economic characteristics on glycemic control, a logistic regression analysis was performed, with sex, education level, employment, personal income, BMI, social class, and age at diagnosis as explanatory variables. The results described in Table 5 indicate that individuals without income were 5.12 times more likely to have inadequate glycemic control compared to those with an income of three or more minimum wages. Those with an income between one and two minimum wages had 4.61 times higher odds of having the outcome compared to those with higher income. Furthermore, it can be observed that a one-unit increase in BMI represented 1.22 times higher odds of having inadequate glycemic control. Finally, it was observed that belonging to social classes A and B were protective factors, as they represented 0.04- and 0.25-times lower odds of having HbA1c > 8% compared to participants in social classes C/D/E (Table 5).

Glycemic control is an important health parameter for people living with T1DM, who require continuous monitoring of the disease throughout their lives (SBD, 2021). The factors associated with adequate glycemic control in this population are still poorly studied in Brazil. The results of the present study showed that a significant percentage of our sample did not have adequate glycemic control, corroborating international and national data (Coutinho & Silva Júnior, 2016; Fortins et al., 2019; Moreira et al., 2010; Sheleme et al., 2020).

It was also noted that inadequate glycemic control was higher among individuals with lower levels of education, income, and social class, and among those with higher BMI, in agreement with previous findings (Alramadan et al., 2018; Davison et al., 2014; McLarty et al., 2021; Pettus et al., 2019). The precarious health of a significant portion of the global population is related to poverty, as individuals with lower income and fewer individual resources have worse access to health services or preventive measures (WHO, 2018), such as adequate quantity and quality of food or regularly monitored physical activity.

In Brazil, as in other countries, the association between higher diabetes prevalence and low education has been identified as an important socioeconomic indicator and suggests risks for the health-disease process, as access to health services is less frequent, and the adoption of dietary care, physical activity, and disease prevention is not consistent with the opportunities and life context of the poor (Azharuddin et al., 2021; Malta et al., 2014). People with low income face more challenging demands to achieve glycemic control in T1DM, as access to medications and modern supplies is more difficult, as is the continuous support to help them manage the disease and educate them on diabetes (Huo et al., 2020; Nguyen et al., 2015). Thus, poorer glycemic control and treatment difficulties are associated with the financial condition and low socioeconomic status of people living with T1DM (Gomes et al., 2012; Rodrigues et al., 2010).

Nutritional status has also been associated with glycemic uncontrol, which corroborates the findings of the present study. In this regard, a study showed that participants who were overweight or obese had lower adherence to the prescribed diet, and smoking was more frequent (Davison et al., 2014). Other studies with Brazilian samples of adults (Moraes et al., 2003; Nunes et al., 2009) and adolescents (Marques et al., 2011) living with T1DM also found higher proportions of inadequate glycemic control among those with higher BMI, obesity, overweight, and/or at risk of overweight.

The present study explored a series of daily self-care indicators related to T1DM and overall health, finding sets of aspects that were most influential in the risk of high HbA1c. It was possible to observe that adequate glycemic control was predicted by good T1DM management (not delaying insulin use or exceeding doses, adequate perception of disease control) and daily testing (frequency of glucose and insulin testing). Conversely, individuals who reported having other health problems had a higher probability of having poorer glycemic control.

In this regard, routine care such as glucose monitoring has the primary purpose of determining the level of glycemic control for people living with diabetes, being the best option over time in preventing chronic and acute complications (Okido et al., 2017). A study conducted in Germany and Austria between 1995 and 2006 with over 27,000 children and adolescents living with T1DM observed that the concentration of HbA1c was proportional to the number of daily glucose monitoring (Ziegler et al., 2011). Another study also found that the higher the number of daily glucose checks through monitoring, the better glycemic control of HbA1c in people living with T1DM (Šoupal et al., 2020). People undergoing multiple daily insulin injections should monitor their blood glucose levels before and occasionally after meals, before bed and physical exercise, in cases of suspected hypoglycemia or hyperglycemia, and after their normalization, as a strict monitoring behavior (approximately six to eight times a day) is an indicator of reduced concentration of HbA1c in adults (ADA, 2022).

Furthermore, in another study, individuals living with T1DM who regularly performed glucose monitoring had fewer chances of omitting insulin treatment compared to those who monitored glucose levels occasionally (Peyrot et al., 2012). In addition to individual and social factors, a recent meta-analysis of 43 studies conducted in low- and middle-income countries like Brazil, indicated factors associated with low adherence to insulin treatment, including variability in disease expression and treatment, complexity of therapeutic regimens, and difficulties in the transportation of insulin (Azharuddin et al., 2021). In this regard, the presence of other health problems, whether diabetes-related or not, adds complexity to disease management and should be considered by healthcare teams and services.

The use of insulin is indispensable for the survival of individuals living with T1DM, and appropriate treatment is crucial for achieving prolonged glycemic control (ADA, 2022). However, it is important to emphasize the need to strengthen behavioral strategies and expand the use of resources and devices that facilitate insulin monitoring in daily life. Thus, it is believed that new technologies that enable the monitoring of these indicators and alert individuals living with diabetes and healthcare services can assist in care and prevent complications, sequelae, and premature deaths related to glycemic no control (Johnston et al., 2021). Nevertheless, it is important to keep in mind that such technological advancements should be available and accessible to all individuals in an equitable manner, considering the different profiles of individual and social vulnerability among those living with T1DM.

Our study has some limitations to be considered. Being a cross-sectional study, we cannot determine the direction of causality between the variables included in the analyses and glycemic control over time, as participants self-selected the group to participate in based on their monitoring abilities. Additionally, due to restrictions imposed by the Ministry of Health for non-essential activities during the COVID-19 pandemic period, the glucose monitoring results were obtained through self-reporting. Therefore, there is a possibility of overestimation of levels of glycemic control due to memory bias and social desirability. Future studies still need to assess the extent to which self-report measures underestimate the evaluation of glycated hemoglobin. However, it is worth noting that even though the study was based on a convenience sample that was more educated than the general Brazilian population, it was able to detect differences related to social inequalities; therefore, it emphasizes the importance of investigating the topic in more diverse and representative samples. In any case, our study investigated the influence of various self-care indicators on glycemic control in individuals living with T1DM using a robust analytical approach.

The process of glycemic control in T1DM is complex, and its mastery depends on several elements, including commitment to daily care, family support, a bond with the medical team, financial situation, cultural influences, and other factors directly related to the health of individuals living with diabetes. Particularly, diabetes education is recommended as a fundamental part of effective care of people living with diabetes to promote better disease management outcomes. Diabetes education should always be supported by an understanding of the psychological, social, and cultural needs of individuals living with T1DM, especially regarding difficulties in accessing and adhering to healthy eating and guided physical activity.

It is worth noting that the widespread availability of new technologies and treatment options that facilitate T1DM management is still limited in Brazil, which hinders aspects of disease management for a significant portion of the population. Therefore, our study highlights the importance of implementing health strategies that are sensitive to social inequalities and individual characteristics and difficulties related to T1DM control in everyday life. Additionally, these strategies can promote treatment adherence and maintenance.