Two distinct psychological concepts: Diabetes depression and diabetes burnout profiles.

Introduction

Diabetes mellitus is on the rise worldwide,[1] and in Iran, the prevalence of diabetes was 15.0% (95% CI 12.6–17.3).[2] Specifically, the prevalence of type 2 diabetes in Chaharmahal Bakhtiari in Iran was 9.8%.[3]

Individuals with diabetes face various psychological challenges that are often intricate, including depression, distress, and burnout. Diabetes burnout refers to the exhaustion and frustration that arises from the daily demands of diabetes self-care, or not being able to achieve desired blood sugar levels, which can ultimately result in neglecting or inconsistent self-care behaviors.[45] Although the exact nature of the association between diabetes burnout and diabetes complications remains unclear, substantial evidence suggests that diabetes burnout may lead to adverse health outcomes.[678] In addition, diabetes outcomes are related to other psychological challenges such as depression.[9]

Nursing theorists have recently focused on the concept of burnout in the field of diabetes. Abdoli et al. have extensively studied this concept in the past few years[6810111213] and have identified three interdependent dimensions for diabetes burnout (exhaustion, detachment, and loss of control). These dimensions interact with one another and give rise to various diabetes burnout profiles that individuals may experience over time.[681011] These profiles are defined based on different combinations of burnout dimensions. Therefore, it is crucial to consider both burnout profiles and the primary dimensions of burnout (exhaustion, detachment, and loss of control) that may contribute to the onset of diabetes burnout.

While diabetes burnout represents a unique experience of emotional and behavioral disengagement with diabetes, it shares some similarities with other psychological concepts. Researchers have found it challenging to differentiate diabetes burnout from depression symptoms.[1415] However, scholars like Abdoli argue that diabetes burnout is a complex and multifaceted concept that is distinct from depression and reflects a crisis in the relationship between individuals and diabetes. Following previous studies of Abdoli et al. to complete the dimensions of the concept of diabetes burnout, further research is needed to differentiate diabetes burnout from depressive symptoms; this study aims to examine diabetes burnout profiles and their association with depressive symptoms in individuals with type 2 diabetes, to gain a better understanding of diabetes burnout.

Understanding diabetes burnout and its relation to other psychological concepts can aid healthcare providers and patients in identifying underlying mechanisms to prevent burnout or implement strategies to alleviate it, making diabetes care more comprehensive and focused beyond diabetes depression.[8]

Material and Methods

A cross-sectional descriptive–analytical study was conducted in all 10 counties of the Chaharmahal and Bakhtiari Provinces of Iran in 2023.

The study included individuals who had been diagnosed with type 2 diabetes and had been living with diabetes for at least 1 year. The participants were required to be at least 18 years old, able to comprehend Farsi, literate, and free from any self-reported diagnosed mental disorders. Additionally, participants had to be residents of the province. Convenience sampling with proportional allocation was conducted from each of the 10 counties. In total, 1300 individuals with type 2 diabetes participated in the study, with 1118 of them completing the questionnaire.

The study was approved by the Ethics Committee of Shahrekord University of Medical Sciences (IR.SKUMS.REC.1401.055). Participants were selected if they expressed interest in participating in the study. They were informed that the questionnaire was confidential and that there were no right or wrong answers, as it solely aimed to gather their personal opinions. Participants completed an informed consent form.

The three questionnaires using the self-report method included demographic information (age, gender, ethnicity, marital status, education level, occupation, and year of diabetes), Personal Health Questionnaire Depression Scale (PHQ), and Diabetes Burnout Scale (DBS).

PHQ-8 consisting of eight items was utilized to measure depressive symptoms. It presents questions with four response options (0: never; 1: several days; 2: more than half the day; and 3: nearly every day). The total score for all eight items was calculated, with scores ranging from zero to 24. A score higher than 10 indicates moderately severe depression, while a score greater than 20 suggests a severe depression.[16]

DBS is comprised of 12 questions, which are rated on a 5-point Likert scale (1: strongly disagree; 2: disagree; 3: neutral; 4: agree; and 5: strongly agree). The questionnaire is organized into three subscales, representing the main dimensions of diabetes burnout: exhaustion (four questions—1, 5, 8, and 12), detachment (five questions—2, 4, 7, 10, and 11), and loss of control (three questions—3, 6, and 9). To calculate the total burnout score, the scores of all 12 items are added and divided by 12. To calculate the score for each subscale, the scores of the relevant questions are added and divided by the number of questions in that subscale.[17]

The validity and reliability of both the DBS and the PHQ-8 questionnaire have been confirmed in a previous study conducted in the United States, Brazil, and Iran.[18] The reliability of the questionnaires was assessed using Cronbach’s alpha method.

Unsupervised clustering methods were utilized to determine the optimal number of profiles and assign individuals to their respective profiles. These methods involve analyzing and clustering unlabeled data, and uncovering hidden or distinct groups without any human intervention. Once different burnout profiles were identified, the Kruskal–Wallis test was conducted to examine the difference in diabetes depression score across these profiles. The statistical significance was determined by a P value of less than 0.05.

Results

Data analysis included 1118 participants, of whom 62.4% were women and 37.6% were men. In terms of marital status, 15.9% of participants were widowed or divorced, 79.7% were married, and 4.2% were single. Regarding education level, 38% of participants were illiterate, 36.8% had a high school education, 17.8% had a diploma, and 7.4% had a university education. On average, participants had been living with diabetes for 11.052 years with a standard error (SE) of 0.214. An average age was 64.47 (SE = 0.396). Mean scores for diabetes depression and burnout are presented in Table 1.

Mean score of diabetes depression and burnout

To uncover the latent profiles of diabetes burnout, three burnout dimensions—detachment, loss of control, and exhaustion—were treated as independent variables. To facilitate analysis, the independent variables were first merged into a single variable using the principal component analysis (PCA) method. This approach, known as dimension reduction, simplifies the problem by reducing the number of inputs and enhances clustering performance.

The K-means clustering method was employed to cluster the data. The number of clusters was tested from two to 27, and the optimal number of clusters was found to be five based on the silhouette coefficient, which is an indicator of cluster appropriateness. The silhouette coefficient can range from negative one to one (negative values indicating poor clustering and values close to zero indicating overlapping clusters. A value closer to one indicates better clustering). In this study, the average total silhouette coefficient was 0.545, indicating relatively good clustering. Notably, changing the number of clusters did not improve the silhouette coefficient. To compare the clusters, the data distribution was first checked within each group. The normality assumption was rejected at 0.05; so, the Kruskal–Wallis non-parametric test was used to compare the profiles [Table 2].

Comparing the mean score of burnout dimensions in different profiles

Table 2 shows that the P value for all three variables is less than 0.01, implying that the mean scores of each burnout dimension have a significant difference in at least two profiles. To identify homogeneous subgroups, Dunn’s post hoc test was used to compare the profiles [Tables 3].

Homogeneous groups in terms of detachment, exhaustion, and loss of control

Table 3 indicates that all five obtained profiles exhibit significant differences in the score of exhaustion, detachment, and loss of control. This implies that the five profiles are distinct from one another, and thus, five burnout profiles can be identified in the data [Table 4].

Burnout mean score in five obtained profiles

Table 4 presents the mean of each burnout dimension, which was used to name the profiles. The mean and standard error for each profile were computed for the three burnout dimensions (detachment, exhaustion, and loss of control). Profile number five had an average of almost 3.50 or higher in all three burnout dimensions and was therefore named “fully burned-out” (high on all burnout dimensions). Profile number three had an average of almost three or higher in all three dimensions but did not reach the critical value of profile number five; so, it was named “overextended” (high on exhaustion and loss of control). Profile number one had a detachment and loss of control score of less than three, while the exhaustion score was higher than three, and thus, it was named “exhausted.” (High on exhaustion only). Profile number four had a score of almost 2.30, which is lower than three, in all three burnout dimensions. Consequently, it was named the “threshold of burnout” profile. Profile number four suggests that if no intervention is taken, the individual may progress to the exhausted and then to the overextended and fully burned-out phases. The table reveals that the mean score in the exhaustion dimension is higher than the other two dimensions, indicating that individuals in profile number four (on the threshold of burnout) may first enter the exhausted profile. Profile number two had low scores (below two) in all three burnout dimensions, indicating that they have not yet reached the on the threshold of burnout or exhausted phase. Therefore, it was named the “not-burned-out” profile.

Initially, the normality of burnout scores for each profile was assessed to determine whether there was a significant difference between them. However, upon inspection, it was found that none of the profiles exhibited normal distribution of burnout scores. Consequently, the Kruskal–Wallis test was employed to compare the mean burnout scores across various profiles.

Test statistics in the Kruskal–Wallis test was 1051.846. The P value was 0.000 (degrees of freedom = 4), implying that there is a significant difference in burnout scores between at least two profiles. To delve deeper into the matter, Dunn’s post hoc test was performed [Table 5].

Homogeneous groups in terms of burnout score

Table 5 reveals that there is a significant discrepancy in the overall burnout score among all profiles. As anticipated, the average burnout score escalates as one transition from a not-burned-out profile to a fully burned-out one. These findings suggest that the profiles obtained are plausible, even when considering the total burnout score.

Table 1 displays the mean depression score. However, Table 6 reveals that 77.3% of individuals experience mild to moderate depression, 22.4% experience moderately severe depression, and 0.4% experience severe depression, indicating an overall prevalence of moderately severe and severe depression caused by diabetes at 22.8%. Conversely, Table 4 reports that 65.8% of people encounter at least one level of diabetes-related burnout (fully burned-out, overextended, and exhausted). These findings suggest that there may be a disconnect between these two concepts; that is, individuals may experience burnout even if they are not depressed.

Distribution of depressed participants in different profiles

Table 4 illustrates that when individuals in the not-burned-out and threshold of burnout profiles are considered as not-burned-out, they make up 34.2% of the study’s population. Additionally, as previously mentioned, 77.3% of the population did not experience moderately severe and severe depression. These results suggest that even though most individuals experience mild to moderate depression, they may have entered one of the burnout profiles, indicating a possible distinction between burnout and depression. This observation is further supported by Table 6, which presents the frequency of individuals in each burnout and depression profile.

Table 6 indicates that several individuals experience high levels of burnout despite having low levels of depression. However, the majority of those who experience moderately severe or severe depression are at the exhausted, overextended, or fully burned-out levels. This finding suggests that as the level of depression increases, the level of burnout also increases. Nevertheless, this does not imply that individuals with low levels of depression do not experience burnout. While the results indicate that individuals who are depressed also experience burnout, those who are burned-out are not necessarily depressed. This is a crucial outcome in distinguishing between these two concepts.

The mean score of depression was 2.33 (SE = 0.200) for the not-burned-out profile, 5.21(SE = 2.22) for the threshold of burnout, 7.03 (SE = 0.239) for the exhausted, 8.53 (SE = 0.238) for the overextended profile, and 11.31 (SE = 0.39) for the fully burned-out profile. These results indicate that the mean depression score increases as we move toward fully burned-out profile.

To investigate the difference in depression score across various burnout profiles, we first examined the distribution of data for each profile. It has no normal distribution; then, the Kruskal–Wallis test was employed to compare the depression score across different profiles. P value was 0.000 (test statistics was 335.801; degrees of freedom was 4). Consequently, we can conclude that there is a significant difference in the mean depression score between at least two profiles. Dunn’s post hoc test was conducted to identify profiles that differ significantly from one another [Table 7].

Table 7 reveals that there are significant differences in depression score across all profiles, and as the level of burnout increases, the average depression score also increases.

Discussion

In this study, we identified five burnout profiles in individuals with type 2 diabetes: fully burned-out, overextended, exhausted, on the threshold of burnout, and not-burned-out. Meanwhile, Abdoli et al.[19] conducted a cross-sectional online study across the United States to identify latent profiles of diabetes burnout in adults with type 1 diabetes. Their results revealed seven latent profiles of burnout, with 25.02% of their participants classified as not exhibiting any burnout symptoms. Various qualitative studies confirm the existence of different dimensions of diabetes burnout.[681011] Abdoli et al.[11] noted that while exhaustion is the entry point to burnout, detachment from the disease identity, diabetes self-care, and support system is more crucial components. Their findings suggest that diabetes burnout may manifest in different profiles among individuals, with some experiencing extreme exhaustion while struggling to find motivation and energy to participate in their own care. Burnout is a transitional state leading to a more complex experience in which individuals lose interest and feel disconnected from their illness identity, self-care, and support system.[12]

The study findings indicate that there is a significant difference in the total burnout score among all the identified profiles. As expected, the mean burnout score increases as we move from the not-burned-out profile to the fully burned-out profile, suggesting that the profiles obtained are reasonable even when considering the total burnout score. It has been estimated that 25% of people with type 1 diabetes and 12% of people with type 2 diabetes in the United States experience burnout.[20] In Abdoli et al.’s[18] study, 26.4% of participants reported moderate to severe levels of burnout. In a cross-national study of adults with type 1 diabetes conducted during the coronavirus disease 2019 (COVID-19) pandemic in the United States, Brazil, and Iran, participants from Iran had higher levels of diabetes burnout.[18] The prevalence of burnout in people with diabetes is high, as evidenced by the results of this study and others. Diabetes burnout is a common experience among those with diabetes and is a significant factor in creating gaps in diabetes care. Finding a solution to address diabetes burnout may help control diabetes in affected individuals.

In this study, individuals with type 2 diabetes had an estimated average diabetes depression score of 7.04, and 22.8% of them experienced moderately severe to severe diabetes depression. Additionally, 65.8% of these individuals experienced one of the levels of diabetes burnout. This suggests that even if individuals are not depressed, they may still experience burnout. In fact, most individuals experience mild to moderate depression and have entered one of the burnout profiles. The findings indicate that as depression levels increase, burnout levels also tend to increase, although this does not necessarily mean that individuals with low depression levels do not experience burnout. It is noteworthy that while depressed individuals are also burned-out, not all individuals experiencing burnout are necessarily depressed. This is a crucial finding in distinguishing between these two concepts.

By analyzing 44 studies that included 10,349 Iranian patients with diabetes, Khalighi et al.[21] estimated the overall prevalence of depression to be 61.8%, with mild depression at 24.4%, moderate depression at 19.1%, severe depression at 11.4%, and highly severe depression at 4.6%. Dehesh et al.[22] conducted a study in Iran on 1500 individuals with type 2 diabetes and estimated the depression rate to be 59%. In Saudi patients with type 2 diabetes, the prevalence of diabetes depression was reported to be 20%.[23] Hasanovic et al.[24] found that 36% of patients with type 2 diabetes in Bosnia experienced depression, with severe depression observed in 8% of cases among those with depression. Meanwhile, Zhang et al.[25] estimated the prevalence of depression at 60% in middle-aged and elderly patients with type 2 diabetes.

According to Abdoli et al.,[6] there was a significant association (P < 0.01) between diabetes burnout and depression in adults with type 1 diabetes in the United States. They further suggested that while diabetes burnout, depressive symptoms, and distress were significant predictors of multiple diabetes outcomes, the overall diabetes burnout was no longer significantly associated with these outcomes after controlling for depressive symptoms and distress.

In Abdoli et al.’s[12] study, participants noted that burnout is a distinct concept but closely related to distress and depressive symptoms. Meanwhile, in another study, the relationship between depression and the exhaustion dimension was found to be high, with the loss of control dimension being moderate, and the detachment dimension being low.[18] The findings of these studies highlight the significant relationship between burnout and depression, which is consistent with this study. Abdoli et al.[8] also emphasized that diabetes burnout may have long-term adverse outcomes, such as poor compliance with treatment, diabetes complications, and diabetes depression, not only in individuals with diabetes but also in their caregivers and families.

As evident from all the aforementioned studies, the occurrence of depression in people with diabetes is highly likely and can have a direct impact on the consequences of diabetes and its associated psychological problems. It appears that individuals who experience greater levels of burnout are more likely to experience depression. However, severe and chronic forms of diabetes burnout may exhibit similar characteristics to depression, and individuals with depression may also display symptoms of burnout. If burnout is left unaddressed, it may not only lead to adverse diabetes outcomes but also to other psychosocial disorders.

This study has some limitations that indicate the need for further research. Despite efforts to select representative samples, the majority of participants were married women, which may introduce biases and limit the generalizability of the findings. The study’s results are based on self-reported data from a cross-sectional study, and other profiles may be identified in different studies. The small number of participants in the fully burned-out profile may suggest a selection bias, as individuals who are severely depressed or burned-out may be less likely to participate in diabetes-related activities and difficult to reach.

While these findings provide an initial perspective and insight into the profiles, investigating the stability or changes of these profiles over time and their relationship with other psychosocial concepts and diabetes outcomes requires more prospective longitudinal studies, which represent a new avenue for research.

Conclusion

Diabetes burnout is a prevalent experience, and addressing it could significantly aid in diabetes management. This study identified five diabetes burnout profiles and found a significant relationship between each profile and depression. The study demonstrates that diabetes burnout is not a uniform experience, challenging the notion that it is synonymous with depression. Addressing burnout based on profiles could be more effective for individuals with diabetes. For instance, those who are exhausted may require interventions that may not be suitable or effective for other profiles. Individuals who have entered the stages of burnout but are not fully burned-out may not have stopped their self-care behaviors but they may be at risk of progressing to the fully burned-out profile. Ignoring such individuals may lead to increased complications and negative outcomes of diabetes.