Diabetes Stigma and Clinical Outcomes in Adolescents and Young Adults: The SEARCH for Diabetes in Youth Study

Health-related stigma is defined as a personal experience characterized by exclusion, rejection, or blame resulting in an adverse social judgment about a person with a specific health condition (1). Stigma has been well documented with HIV/AIDS, epilepsy, and mental illness (2–4). Community stigma leads to internalized stigma, resulting in adverse clinical and psychosocial outcomes (4). To our knowledge, the largest quantitative study of diabetes stigma used unvalidated questions to survey 5,422 people, of whom 96% were adults, and found that 76% of people with type 1 diabetes and 52% of people with type 2 diabetes perceived diabetes-related stigma (5).

There is a limited body of research examining the association between diabetes stigma, sociodemographic variables, and clinical outcomes. In a survey of 1,594 adults with type 1 diabetes, higher diabetes-related stigma scores were associated with female sex, younger age, and shorter diabetes duration (6). Diabetes stigmatization is associated with elevated HbA_1c (5,7,8) and having at least one episode of severe hypoglycemia in the past year (7). In Australia, use of validated Diabetes Stigma Assessment Scales 1 and 2 among adults with type 1 (n = 959) and type 2 (n = 1,129) diabetes demonstrated that diabetes stigma is associated with depressive and anxiety symptoms and diabetes distress, whereas associations with diabetes self-care and HbA_1c were interpreted as being statistically significant but not clinically meaningful (9). A qualitative study of adults with type 1 diabetes in Australia found that participants perceived or experienced diabetes-related stigma by association with type 2 diabetes and as blame, negative social judgments, exclusion, rejection, or discrimination (10). These perceptions or experiences led to several consequences of diabetes-related stigma, including emotional distress, impact on identity, further perpetuation of type 2 diabetes-related stigma, and reluctance in disclosing their diagnosis, especially in the workplace and social settings (10).

Adults with type 2 diabetes who report experiencing diabetes-related stigma tend to have a higher BMI, use insulin, and have higher HbA_1c (5). A qualitative study among adults with type 2 diabetes showed internalized diabetes-related stigma led to a change in attitude toward social participation and management of their diabetes as recommended (11), which may include delaying or skipping insulin injections. Another study of adults with type 2 diabetes found that those who perceived or experienced diabetes-related stigma were unwilling to disclose their diagnosis and believed that people with type 1 diabetes did not experience diabetes-related stigma (12).

The adolescent and young adult (AYA) population may be especially vulnerable to stigma because of the emphasis on personal identity, peer relationships, and establishing autonomy from parents at this developmental stage (13). Additionally, AYAs with diabetes tend to have increased HbA_1c levels during the adolescent period (14) and time of transition from pediatric to adult care for their diabetes (15–17), making this an important time to provide psychosocial support. One study has examined diabetes-related stigma in AYAs with type 1 diabetes (n = 380) and found a diabetes-related stigma prevalence of 65.5% (7). Additionally, AYAs who endorsed experiencing diabetes-related stigma had double the odds of having HbA_1c >9% (75 mmol/mol) or having at least one episode of severe hypoglycemia in the past year (7).

The aim of this study was to determine the frequency of diabetes stigma in AYA participants in the SEARCH for Diabetes in Youth (SEARCH) study with type 1 or type 2 diabetes and assess the association of diabetes stigma with clinical characteristics and outcomes, such as diabetes type, diabetes treatment plan, continuous glucometer (CGM) use, HbA_1c levels, acute complications, and long-term complications. To our knowledge, this research will be the first large, quantitative study of diabetes stigma in AYAs to include and stratify by both type 1 and type 2 diabetes. It will also address the current gaps in the literature, including the associations of diabetic ketoacidosis (DKA), long-term complications, insulin pump use, and CGM use with diabetes-related stigma in AYAs.

SEARCH is a multicenter study following the clinical course of type 1 and type 2 diabetes among youth in the U.S. SEARCH recruited participants from a membership-based health plan site in southern California, several other geographic sites (Washington, Colorado, Ohio, South Carolina, and selected American Indian reservations in Arizona and New Mexico under the direction of Colorado). At study visits, participants completed informed consent and assent when applicable, provided blood and urine samples for laboratory assessments, had a physical examination, and completed questionnaires. Individuals in selected incident years (2002–2006, 2008, and 2012) were invited for follow-up visits. These visits included additional questionnaires, blood samples, and a physical examination. A local institutional review board for each of the study sites approved this study protocol.

This study included SEARCH participants who were diagnosed in youth with type 1 diabetes or type 2 diabetes, completed an in-person SEARCH 4 cohort follow-up visit, were 10 to 24.9 years of age at the cohort visit, and completed a survey regarding diabetes-related stigma. The SEARCH 4 cohort study visits were conducted from 2016 to 2019 and were the fourth funding phase of the study.

Demographic characteristics collected at the cohort visit included age, sex as recorded in the medical record, age at diabetes diagnosis, current health insurance, and self-reported race and ethnicity, in order to consider racial and ethnic social constructs and the impacts on diabetes care and outcomes. Participants self-identified their race and ethnicity as one or more of the following categories: American Indian, Alaska Native, Asian, Hispanic, Multiracial, Non-Hispanic Black, Non-Hispanic White, or Pacific Islander. The SEARCH surveys did not collect self-reported gender. For participants aged ≥18 years, information on the highest education level attained and employment status was collected. For participants aged <18 years, information on the highest parental education level of the two parents and employment status was collected. Clinical characteristics included diabetes type determined by physician report, diabetes treatment plan, CGM use, BMI, HbA_1c levels, self-reported DKA episode in the past 12 months, self-reported hospitalization in the past 12 months, self-reported severe hypoglycemia episode in the past 12 months, diabetic retinopathy, nephropathy, and dyslipidemia. The in-person visit included a physical examination and the collection of blood and urine samples for assessments of diabetic retinopathy, nephropathy, hypertension, and dyslipidemia. The presence of diabetic retinopathy was determined by grading digital fundus images collected during the in-person visit, as previously described (18). Nephropathy was defined as an albumin to creatinine ratio ≥30 μg/mg or estimated glomerular filtration rate <60 mL/min/1.73 m². Dyslipidemia was defined as LDL cholesterol level >100 mg/dL or triglyceride level >150 mg/dL.

Treatment plans for participants with type 1 diabetes were categorized as insulin pump, multiple daily injections defined as long-acting insulin plus short- or rapid-acting insulin at least twice daily, and other injections defined as only short-acting insulin or only long-acting insulin. Treatment plans for participants with type 2 diabetes were categorized as oral medications, oral medications plus any insulin, insulin only, and no medications. BMI was categorized as <25 kg/m² (or <85th percentile), 25–30 kg/m² (or 85th to <95th percentile), and ≥30 kg/m² (or ≥95th percentile), using percentile for participants aged <18 years and absolute BMI for those aged ≥18 years. The SEARCH 4 cohort completed five questions assessing diabetes-related stigma, as further detailed in the next section.

The SEARCH 4 diabetes-related stigma survey was developed in 2014 on the basis of expert opinion and existing literature at the time, including the second Diabetes Attitudes, Wishes, and Needs (DAWN2) study that assessed psychosocial outcomes in people with diabetes across 17 countries (19) and the Browne et al. (10) qualitative study previously described. At that time, there were no widely used validated surveys for assessing diabetes-related stigma in AYAs. The Barriers to Diabetes Adherence Measure for adolescents was validated in 2011 and had a stigma component to the survey, which focused on social engagement and blame (20).

The first three questions of the SEARCH 4 diabetes-related stigma survey assess perception or experience of diabetes stigma and the last two questions ask about consequences of diabetes stigma (Table 1). Participant answers were scored on a 6-point Likert scale: never (1 point), less than once a year (2 points), a few times a year (3 points), a few times a month (4 points), at least once a week (5 points), and almost daily (6 points). The total diabetes-related stigma score was determined for each participant by adding up their responses for a total score ranging from 5 to 30 points, with the higher number indicating greater perception or experience of diabetes-related stigma.

A principal components analysis showed an Eigenvalue score of 2.77 for the total diabetes stigma score. The Eigen vectors for each stigma question ranged from 0.40 to 0.47, indicating they all contributed to the total diabetes stigma score relatively equally. In addition, the Cronbach α values for each stigma question ranged between 0.72 and 0.77, which indicates that the variables do work well with the score and there is good internal consistency.

Descriptive statistics were calculated for all variables of interest. For categorical measures, we examined counts and percentages; for continuous measures, we examined means and SDs. Each measure was examined in the full participants’ data set and then stratified by diabetes type (type 1 vs. type 2). Next, we examined a series of general linear models to determine which variables were associated with diabetes-related stigma score. We performed these analyses stratified by diabetes type. In each model, we examined the β coefficients from the fitted models and corresponding P values to determine the strength of the association between the measures and diabetes-related stigma scores.

Seven different models were fit that examined different sets of variables and their association with diabetes-related stigma score. The base model included HbA_1c level (measured continuously), treatment plan, race/ethnicity, age, sex, clinic site, duration of diabetes, education level of participant or parent, and health insurance. In model 2, we took the base model and added employment status of participant or parent. In model 3, we took the base model and added household income. In model 4, we refit model 1, replacing HbA_1c measured on a continuous scale with HbA_1c defined as a three-level ordinal variable: <7% (53 mmol/mol), 7–9% (53–75 mmol/mol), and >9% (75 mmol/mol). In model 5, we added CGM use to the base model. In model 6, we took the base model and then examined a set of acute outcomes (namely, DKA episode in the past year, severe hypoglycemia episode in the past year, and hospitalization in the past year). In model 7, we removed the acute outcomes and examined a set of diabetes complications and comorbidities (namely, retinopathy, nephropathy, and dyslipidemia). All analyses were performed using SAS (version 9).

The data that support the findings of this study are available from the SEARCH study, but restrictions apply to the availability of these data, which were used under license for the present study and, therefore, are not publicly available. Data are available, however, from the authors upon reasonable request and with permission of the SEARCH study.

Of the 1,608 participants included, 78% had type 1 diabetes, 56% were female, and 48.5% were non-Hispanic White (Table 2). For all participants, the mean (SD) age at diagnosis was 10.7 (4.49) years and the mean (SD) age at the SEARCH 4 cohort visit was 21.7 (5.11) years. Participants with type 2 diabetes tended to be older at diagnosis (14.3 [SD 2.66] years) and at time of cohort visit (24.6 [SD 4.31] years) compared with those with type 1 diabetes (9.7 [SD 4.37] years at diagnosis and 20.8 [SD 5.01] years at cohort visit). Of all participants, 14.9% had an HbA_1c value at the American Diabetes Association–recommended level of <7% (53 mmol/mol). The average (SD) HbA_1c was 9.2% (2.3%; 77 mmol/mol [2 mmol/mol]). The average (SD) diabetes-related stigma score was 10.9 (5.4) for participants with type 1 diabetes and 9.8 (5.6) for participants with type 2 diabetes.

Table 3 shows the multivariable linear modeling outcomes for participants with type 1 diabetes. For those participants, higher diabetes-related stigma scores were associated with higher HbA_1c (P < 0.001). Female sex was associated with a 1.96-point higher diabetes-related stigma score compared with male sex (P < 0.001). The only significant difference for race/ethnicity was that those who identified as multiracial had a 1.53-point lower diabetes-related stigma score than those who identified as non-Hispanic White (P = 0.0274). No statistically significant association was found between diabetes-related stigma scores and education, employment status, health insurance, or use of an insulin pump or CGM. Household income less than $25,000 was associated with a 1.16-point higher diabetes-related stigma score compared with a household income of $75,000 or greater (P = 0.03).

When adjusted for HbA_1c, the presence of a DKA episode (P = 0.0003) and a severe hypoglycemia episode in the past year (P = 0.002) were both associated with higher diabetes-related stigma scores. Independent of HbA_1c, the presence of retinopathy (P = 0.0002) and nephropathy (P = 0.04), but not dyslipidemia, were associated with higher diabetes-related stigma scores (β coefficient, 1.94 and 1.16, respectively).

Participants with type 2 diabetes had a similar pattern of higher diabetes-related stigma scores being associated with female sex (P = 0.002) and higher HbA_1c (P = 0.009), as shown in Table 4. There was no significant association between diabetes-related stigma scores and race/ethnicity, education, employment status, household income, health insurance, or CGM use. Use of insulin only was associated with a 1.79-point increase in diabetes-related stigma score (P = 0.04), and using insulin plus oral diabetes medication was associated with a 2.03-point increase (P = 0.02) compared with no medications. There was no statistically significant association between diabetes-related stigma scores and DKA episode, severe hypoglycemia episode, or hospitalization in the past year. Retinopathy was associated with a 1.98-point increase in diabetes-related stigma score (P = 0.02), but nephropathy and dyslipidemia were not associated with a difference in diabetes-related stigma score.

We have shown that diabetes-related stigma in AYAs with type 1 or type 2 diabetes is associated with female sex and elevated HbA_1c, but not with education, employment status, health insurance, or use of an insulin pump or CGM. There was no consistent pattern seen between race and ethnicity and diabetes-related stigma. We found that in participants with type 1 diabetes, there was an association between diabetes-related stigma and DKA episode in the past year, severe hypoglycemia episode in the past year, retinopathy, and nephropathy, all independent of HbA_1c level. In participants with type 2 diabetes, diabetes-related stigma was associated with insulin use and independent of HbA_1c, with retinopathy.

The adolescent and emerging adulthood period is marked by an emphasis on peer relationships, personal identity, and establishing autonomy (13), which may make AYAs with diabetes particularly vulnerable to stigma. Elevated HbA_1c levels have been well documented during the adolescent period (14,15) and during transition from pediatric to adult care (16–18). We have shown that diabetes stigma in AYAs is associated with elevated HbA_1c levels for type 1 and type 2 diabetes, which is consistent with reports in earlier literature (5,7,8). Qualitative research has demonstrated that community health–related stigma can lead to internalized self-stigma (2–4). This diabetes-related self-stigma is associated with decreased self-care behaviors (11,21), which contribute to elevated HbA_1c levels, in turn increasing risk for DKA (22), retinopathy, nephropathy, neuropathy, and cardiovascular disease (23,24). Similar to previously reported studies, we have demonstrated that diabetes stigma is associated with female sex (5–7). Female sex is also associated with a higher burden of other psychosocial comorbidities, including depression (25), diabetes distress (26), and disordered eating behaviors (27).

Independent of HbA_1c, we have shown that higher diabetes-related stigma scores are associated with acute complications such as DKA episodes and severe hypoglycemia in AYAs with type 1 diabetes. Similarly, Brazeau et al. (7) found that diabetes-related stigma in AYAs was associated with increased rate of severe hypoglycemia over 1 year (odds ratio 1.86; 95% CI 1.05–3.31), although DKA episodes were not examined. To our knowledge, our study is the first to examine the association of diabetes-related stigma and DKA in AYAs. We did not find an association with diabetes-related stigma and DKA or severe hypoglycemia in participants with type 2 diabetes, likely in part due to having fewer participants and because DKA and hypoglycemia occur less frequently with type 2 diabetes than type 1 diabetes. Additionally, only 56% of participants with type 2 diabetes were using insulin, which increases the risk for hypoglycemia (28).

Long-term complications associated with higher diabetes-related stigma scores included retinopathy and nephropathy for type 1 diabetes and retinopathy for type 2 diabetes when controlling for HbA_1c. Hansen et al. (6) found diabetes-related stigma experienced by adults with type 1 diabetes was associated with having at least one diabetes complication. Our study, to our knowledge, is the first to examine long-term complications and their association with diabetes-related stigma in AYAs. The SEARCH study previously has shown that diabetes-related complications and comorbidities (namely, retinopathy, nephropathy, neuropathy, arterial stiffness, and cardiovascular autonomic neuropathy) are prevalent in the AYA population with type 1 diabetes, with 23.2% having one complication and 4.7% having two complications (29). Medical and behavioral interventions, including addressing diabetes-related stigma in the pediatric and adolescent period, may prevent or delay diabetes-related complications. Additionally, a public health initiative on decreasing community diabetes-related stigma through public education may be considered to increase awareness of the cause of diabetes and what it looks like to live with diabetes. Interventions for reducing mental health stigma include presentation of facts, which can be effective at changing attitudes, and social contact or first-person narratives, which have been shown to reduce stigma (30).

In AYAs with type 2 diabetes, we have shown that insulin use is associated with increased experience of diabetes-related stigma, which is consistent with findings in previously reported literature (5). Interestingly, in our study, technology use such as an insulin pump and CGM was not associated with diabetes-related stigma score. To our knowledge, there have been no quantitative studies examining diabetes-related stigma experience related to insulin pumps and CGMs. There have been qualitative studies assessing patients’ perceptions of insulin pumps, with demonstration that women were more concerned about body image related to and social acceptance of the insulin pump (31).

Most prior studies examining diabetes stigma focused on adults (5,6,8–12). To our knowledge, there is only one previous study of diabetes stigma specifically in the AYA population (n = 380), and people with type 2 diabetes were excluded (7). Our study is unique in its large sample size of AYAs and distribution of participants across the U.S. Additionally, the diabetes-related stigma score was stratified by type 1 diabetes and type 2 diabetes. This is important because it has been well documented in the literature that the ways in which diabetes-related stigma is experienced may vary based on type of diabetes (5,10,12).

Our study is limited in that it is a cross-sectional analysis and had fewer AYAs with type 2 diabetes than with type 1 diabetes. Our five-question diabetes-related stigma survey is not validated. Next steps could include conducting a larger prospective study using a validated diabetes-related stigma questionnaire and collecting and analyzing psychosocial variables, such as socioeconomic status, depression, diabetes distress, disordered eating behaviors, and quality of life. Future studies can also include longitudinal assessment of diabetes-related stigma to further elucidate the causal and temporal relationship between diabetes stigma and clinical outcomes.

Diabetes-related stigma in AYAs is associated with female sex, elevated HbA_1c, and retinopathy, in addition to DKA and severe hypoglycemia in those with type 1 diabetes and insulin use in those with type 2 diabetes. It is critical to address diabetes stigma in comprehensive diabetes care, especially in the AYA period, as this is a key time for developing personal identity and autonomy in addition to transitioning to adult diabetes care.

Funding. The SEARCH 4 for Diabetes in Youth Cohort Study is funded by the National Institutes of Health, the National Institute of Diabetes and Digestive and Kidney Diseases (1R01DK127208-01, 1UC4DK108173) and supported by the Centers for Disease Control and Prevention. The Population-Based Registry of Diabetes in Youth Study is funded by the Centers for Disease Control and Prevention (grants 1U18DP006131, U18DP006133, U18DP006134, U18DP006136, U18DP006138, U18DP006139 and DP-15-002) and supported by the National Institutes of Health, the National Institute of Diabetes and Digestive and Kidney Diseases. SEARCH 1, 2, and 3 for Diabetes in Youth is funded by the Centers for Disease Control and Prevention (PA no. 00097, DP-05-069, and DP-10-001) and supported by the National Institute of Diabetes and Digestive and Kidney Diseases. Participating sites and their respective SEARCH grant numbers are as follows: Kaiser Permanente Southern California (U48/CCU919219, U01 DP000246, and U18DP002714), University of Colorado Denver (U48/CCU819241-3, U01 DP000247, and U18DP000247-06A1), Cincinnati’s Children’s Hospital Medical Center (U48/CCU519239, U01 DP000248, and 1U18DP002709), the University of North Carolina at Chapel Hill (U48/CCU419249, U01 DP000254, and U18DP002708), the Seattle Children’s Hospital (U58/CCU019235-4, U01 DP000244, and U18DP002710-01) and Wake Forest University School of Medicine (U48/CCU919219, U01 DP000250, and 200-2010-35171).

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. K.B.E., A.J.R., C.P., R.D’A., C.E.B., R.A.B., A.B., A.C., D.D., L.M.D., E.T.J., A.D.L., E.J.M.-D., K.R., and S.M.M. were involved in the conception, design, and conduct of the study and the analysis and interpretation of the results. K.B.E. wrote the first draft of the manuscript, and all authors edited, reviewed, and approved the final version of the manuscript. C.P. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Prior Presentation. Parts of this study were presented at the 82nd Scientific Sessions of the 2022 American Diabetes Association, New Orleans, LA, 4 June 2022; University of Washington Diabetes Day, Seattle, WA, 25 May 2022; and the 11th International Meeting of Pediatric Endocrinology, Buenos Aires, Argentina, 4–7 March 2023.