Mental health comorbidities in adolescents and young adults (AYAs) with type 1 diabetes have a significant impact on chronic disease management and quality of life (1,2). Diabetes distress (DD) is described as the negative emotional impact of living with diabetes (3). It is prevalent in youth with type 1 diabetes and is associated with lower self-esteem, satisfaction with life, and self-efficacy (4). In the Diabetes MILES Youth study of AYAs aged 13–19 years, 36% of participants reported high levels of DD, which was associated with higher self-reported A1C levels (5).
Depression is prevalent in youth with type 1 diabetes, with a pooled rate of 30% of youth with type 1 diabetes reporting depressive symptoms in one meta-analysis (6). Depressive symptoms have been found to correlate with elevated A1C in some studies (7,8). In the Diabetes MILES Youth study, DD was found to mediate the relationship between symptoms of depression and A1C (5). This finding points to the overlap in psychosocial comorbidities in youth with type 1 diabetes and the potential relationship between DD and depression. We are not aware of studies exploring the longitudinal relationship between depressive symptoms and A1C in AYAs.
Suicide risk has been found to be present in 9% of AYAs with type 1 diabetes (9). Individuals with insulin-dependent diabetes may be at increased risk for suicide given the potential lethality of insulin when used as a form of self-harm, underscoring the importance of screening for psychological comorbidities and suicide risk in this population (9). Of the aforementioned psychosocial comorbidities, suicide risk is the least studied comorbidity in the population of AYAs with type 1 diabetes. In one study of adolescents with type 1 diabetes, higher depressive symptoms and having public health insurance were associated with endorsement of suicidal ideation, although other clinical variables such as A1C were not (10). Longitudinal trends have not been evaluated.
Consensus guidelines now recommend routine screening for psychosocial comorbidities in clinical practice (11,12). Although a few studies have examined sociodemographic characteristics associated with these psychosocial comorbidities in adolescents with type 1 diabetes (5,13), gaps remain in identifying which patients may be at highest risk for developing mental health comorbidities, and particularly suicidality. There is also a lack of research exploring the association between DD, depression, and suicide risk with A1C over time.
The aim of this study was to determine the longitudinal association between DD, depressive symptoms, and suicide risk with A1C. We hypothesized that these mental health measures would be associated with higher A1C and that these associations would be stable over time. In addition, we sought to examine associations between DD, depressive symptoms, and suicide risk with sociodemographic and clinical characteristics to identify AYAs who may benefit from targeted interventions to address mental health comorbidities.
Research Design and Methods
The ambulatory diabetes clinic mental health screening protocol for AYAs with diabetes was initiated in March 2018 and has been previously described as a part of a quality improvement initiative (14). At the beginning of a participant’s diabetes clinic visit, English- and Spanish-speaking patients between the ages of 13 and 21 years completed questionnaires via an electronic tablet assessing DD, depression, and suicide risk. Psychosocial screening was conducted at all routine ambulatory diabetes clinic visits. Elevated depressive symptoms or suicidal ideation led to a social work referral for evaluation, safety planning, and support with community resources. Elevated DD alerted providers, but no standardized intervention was implemented.
Data for this study were collected from mental health screenings performed between March 2018 and March 2020. Data after March 2020 were not included in this analysis given the impact of the coronavirus disease 2019 (COVID-19) pandemic on psychosocial screening after the adoption of telemedicine for ambulatory diabetes care, as in many U.S. diabetes centers during the first year of the pandemic (15). Reliable psychosocial screening and A1C evaluation were not possible for the majority of telemedicine visits in 2020 and 2021.
The Seattle Children’s Research Institute Institutional Review Board issued a waiver of consent, as this study was designed to use information obtained from a clinical protocol consistent with standard of care (12).
Measures
The Problem Areas in Diabetes–Teen (PAID-T) questionnaire was used to assess DD (16). This tool aims to assess diabetes-specific emotional distress for teens. The PAID-T has demonstrated construct validity in adolescents and includes 14 items on a 6-point Likert scale, with a possible score range of 14–84. A severity cutoff score of ≥44 maximizes sensitivity and specificity for differentiating teens with greater DD and more negative emotional symptoms (16).
The Patient Health Questionnaire-9 (PHQ-9) was used to assess symptoms of depression and suicide risk (17). This nine-item assessment has been validated in adolescents >13 years of age and adults in detecting major depression and includes nine items on a 4-point Likert scale (0–3). The PHQ-9 asks about frequency of symptoms respondents have experienced in the past 2 weeks. Response options range from 0 (not at all) to 3 (nearly every day). Summary scores are created by summing item responses, with a possible range of 0–27. In adolescents, a value of ≥11 has been found to optimize sensitivity and specificity and was used to define moderate-severe depressive symptoms (17). Suicide risk was determined by a response of ≥1 to the suicide risk item on the PHQ-9, which asks participants, “How often have you been bothered by . . . thoughts that you would be better off dead or of hurting yourself in some way” in the past 2 weeks?
Clinical and demographic characteristics from the baseline visit were extracted from the medical record. Diabetes type was determined based on International Classification of Diseases, 10th revision, diagnostic codes. A1C levels were measured via a point-of-care laboratory test carried out on the same encounter date as the mental health screening at multiple time points. Only A1C levels associated with an ambulatory diabetes visit at which PHQ-9 screening occurred were used in this analysis.
Data Analysis
Participant characteristics were summarized using descriptive statistics. χ2 or Fisher exact tests were used to assess differences in dichotomized PHQ-9 scores, PHQ-9 suicide risk, and PAID-T scores by sociodemographic characteristics. Linear mixed-effects models were used to evaluate the effects of PAID-T, PHQ-9 severity, and suicide risk (as measured by the PHQ-9) on A1C over time, adjusting for race/ethnicity, BMI, age, sex, health insurance status, continuous glucose monitoring (CGM) use, and insulin pump use. Results were reported as estimates with 95% confidence intervals. A P value of 0.05 was considered statistically significant. SAS, v. 9.4, software (SAS Institute, Cary, NC) was used for all analyses.
Results
The sample consisted of 596 AYAs with type 1 diabetes. The mean age of participants was 16.2 ± 2.1 years, with a mean A1C of 9.2 ± 2.2% (Table 1). Seventeen percent of participants endorsed DD, 13.9% experienced moderate-severe depressive symptoms, and 8.6% endorsed suicide risk at baseline. Nearly 8% of participants reported both elevated symptoms of depression and DD at baseline.
Table 1.
Participant Characteristics at Baseline (N = 596)
| Characteristic | Value |
|---|---|
| Age, years | 16.2 ± 2.1 |
| Female sex | 288 (48.3) |
| Race/ethnicity Non-Hispanic White Hispanic Non-Hispanic Black or African American Other Patient did not disclose |
391 (65.7) 80 (13.5) 35 (5.9) 64 (10.8) 25 (4.2) |
| Preferred language for care English Spanish |
569 (95.5) 27 (4.5) |
| Insurance* Commercial Medicaid Other |
375 (65.6) 187 (32.7) 10 (1.8) |
| BMI z score* | 0.6 ± 0.9 |
| Moderate-severe depressive symptoms (PHQ-9 ≥11)* | 83 (13.9) |
| Positive suicide risk | 51 (8.6) |
| High DD (PAID-T ≥44)* | 103 (17.4) |
| Baseline A1C, %* | 9.2 ± 2.2 |
| Insulin pump use* | 269 (50.3) |
| CGM use* | 337 (57.4) |
Data are mean ± SD or n (%).
Missing data included insurance (24 participants), BMI (60 participants), question 9 on the PHQ-9 (1 participant), PAID-T score (4 participants), A1C (21 participants), insulin regimen (61 participants), and CGM use (9 participants).
Baseline mental health scores and associations with demographic and clinical characteristics are presented in Table 2. At baseline, older age (≥18 years) (P = 0.003), female sex (P = 0.001), and a multiple daily injection (MDI) insulin regimen (P = 0.017) were associated with moderate-severe depressive symptoms. Use of an MDI regimen (P = 0.017) and older age (P = 0.045) were associated with suicide risk. Older age (P = 0.004) and female sex (P = 0.004) were associated with higher DD. CGM use and health insurance status were not associated with mental health measures.
Table 2.
Mental Health Baseline Scores and Associations With Demographic and Clinical Characteristics
| Characteristic | Depressive Symptoms | PHQ-9 Suicide Risk | Diabetes Distress | ||||||
|---|---|---|---|---|---|---|---|---|---|
| PHQ-9 ≥11 | PHQ-9 <11 | P | Yes | No | P | PAID-T ≥44 | PAID-T <44 | P | |
| Age, years 13–17 ≥18 |
54 (65.06) 29 (34.94) |
410 (79.92) 103 (20.08) |
0.003 |
34 (66.67) 17 (33.33) |
419 (78.86) 115 (21.14) |
0.045 |
69 (66.99) 34 (33.01) |
391 (79.96) 98 (20.04) |
0.004 |
| Sex Male Female |
29 (34.94) 54 (65.06) |
279 (54.39) 234 (45.61) |
0.001 |
23 (45.10) 28 (54.90) |
285 (52.39) 259 (47.61) |
0.319 |
35 (33.98) 68 (66.02) |
271 (55.42) 218 (44.58) |
<0.001 |
| Race/ethnicity Non-Hispanic White Black or African American Hispanic Other Patient did not disclose |
55 (66.27) 5 (6.02) 6 (7.23) 14 (16.87) 3 (3.61) |
337 (65.69) 30 (5.85) 74 (14.42) 50 (9.75) 22 94.29) |
0.184 |
34 (66.67) 2 (3.92) 4 (7.84) 10 (19.61) 1 (1.96) |
357 (65.63) 33 (6.07) 76 (13.97) 54 (9.93) 24 (4.41) |
0.171 |
63 (61.17) 12 (11.65) 10 (9.71) 13 (12.62) 5 (4.85) |
326 (66.67) 23 (4.70) 69 (14.11) 51 (10.43) 20 (4.09) |
0.059 |
| Insurance Commercial Medicaid Other |
47 (58.02) 32 (39.51) 2 (2.47) |
328 (66.80) 155 (31.57) 8 (1.63) |
0.296 |
27 (54.00) 21 (42.00) 2 (4.00) |
347 (66.60) 166 (31.86) 8 (1.54) |
0.127 |
64 (62.14) 38 (36.89) 1 (0.97) |
308 (66.09) 149 (31.97) 9 (1.93) |
0.531 |
| BMI z score <−1 −1 to 1 >1 |
3 (4.23) 38 (53.52) 30 (42.25) |
26 (5.59) 263 (56.56) 176 (37.85) |
0.732 |
1 (2.33) 21 (48.84) 21 (48.84) |
28 (5.69) 279 (56.71) 185 (37.60) |
0.278 |
6 (6.52) 47 (51.09) 39 (42.39) |
23 (5.22) 251 (56.92) 167 (37.87) |
0.576 |
| Injections versus pump Injections Pump |
43 (63.24) 25 (36.76) |
223 (47.75) 244 (52.25) |
0.017 |
28 (66.67) 14 (33.33) |
238 (48.37) 254 (51.63) |
0.023 |
55 (56.12) 43 (43.88) |
210 (48.39) 224 (51.61) |
0.167 |
| CGM use Yes No |
44 (55.70) 35 (44.30) |
293 (57.68) 215 (42.32) |
0.741 |
25 (51.02) 24 (48.98) |
311 (57.91) 226 (42.09) |
0.350 |
55 (54.46) 46 (45.54) |
280 (57.97) 203 (42.03) |
0.516 |
Data are n (%).
In adjusted models, DD was associated with higher A1C (coefficient 0.032, P <0.001), and PHQ-9 score was associated with higher A1C (coefficient 0.054, P <0.001) but not suicide risk. When both PAID-T and PHQ-9 scores were included in the same model, only PAID-T scores remained significantly associated with higher A1C (coefficient 0.031, P <0.001). The association between mental health comorbidities (PAID-T, PHQ-9, and suicide risk) and A1C levels did not vary significantly over time.
Discussion
Our study shows evidence of an association between DD and depressive symptoms with A1C over time in AYAs with type 1 diabetes. Consistent with previous research, DD was significantly associated with higher A1C, and when depression and DD were in the same model, depression was no longer significantly associated with A1C (5). This finding suggests that DD is the primary driver of the impact depression has on A1C and that the two mental health comorbidities are interrelated; DD likely contributes to depressive symptoms, and depression likely contributes to increased DD. We found that a 10-point increase in PAID-T score was associated with a 0.31% increase in A1C.
This study was unique in that it demonstrated that this strong relationship between DD and A1C persists longitudinally. In a previous study of adolescents with type 1 diabetes, one-third of participants had stable, moderate, or high levels of DD (as opposed to low or decreasing level of distress) over time; those with chronic DD had higher A1C levels over time (18). A randomized controlled trial of the STePS (Supporting Teen Problem-Solving) curriculum, a behavioral intervention for DD and depression, demonstrated decreased DD and symptoms of depression over a 3-year period; yet, there was no significant decrease in A1C over time (19). This result suggests that, in the case of this intervention, there was an uncoupling of the association between DD and A1C, although this relationship was not specifically looked at over time. In contrast, our study was based on a clinical screening protocol that did not offer standardized intervention for DD. However, our study underscores the importance of implementing clinical interventions to target DD, as evidenced by its longitudinal impact on A1C and, thus, future risk of developing diabetes complications.
A significant proportion of participants screened positive for DD in the current study. A Canadian study of younger youth aged 9–13 years with type 1 diabetes found an identical rate of DD of 17% (20), although this rate is lower than previously reported in AYAs with type 1 diabetes from Australia (17 vs. 36%) (5). Our study population has a similar age as the Diabetes MILES Youth study of Australian adolescents (mean age 15.7 years). However, the MILES study required active enrollment in the research study, whereas we looked at data obtained as part of the standard clinical screening protocol and thus included all eligible patients seen in clinic. This difference in methods may explain some of the differences seen in DD rates, as participants in our study were not asked to opt in to screening, thus reducing selection bias. There may also be cultural differences that could affect rates of DD in these different AYA populations. Additionally, older age and female sex were associated with elevated DD in our study. Age has also been demonstrated to be positively correlated with DD in other studies (5).
A significant portion of our sample reported moderate to severe depressive symptoms, although lower than levels reported elsewhere (5,6), which may be associated with differences in the questionnaires used to assess symptoms. Additionally, we used a higher cutoff point of a score ≥11 on the PHQ-9, as this cutoff has been demonstrated to optimize sensitivity and specificity in adolescents compared with the commonly used cutoff point for adults of ≥10 (17). In our study, older age, female sex, and MDI insulin regimen were associated with greater depression severity. Prior studies have similarly shown that adolescents with type 1 diabetes with more depressive symptoms were more likely to be female, have parents without a college education, and be on an MDI insulin regimen (13).
The high rate of suicide risk (9%) in our AYA population matched that described in previous studies looking at suicidality in AYAs with type 1 diabetes (9). In our study, MDI insulin regimen and older age were associated with higher suicide risk. One explanation for higher prevalence of MDI use in those with suicide risk may be related to the common practice of restricting access to insulin pump use during increased periods of risk. In a previous study of youth with diabetes, 6% of AYAs with suicide risk cited insulin overdose as a means of prior suicide attempt or as a part of a current plan (9), which raised the need for further research around the effectiveness of changing an insulin regimen from pump to MDI to restrict insulin access in this population. Additionally, the rates presented in our study likely under-identify AYAs with diabetes based on recent data on suicide risk with depression screeners, and the true rates in this population are likely higher (21).
Strengths of this study include the use of a large clinical sample of AYAs with type 1 diabetes over time and data obtained from a clinical protocol, thus reducing selection bias. This also made the study population more representative of the general population of AYAs with type 1 diabetes, given that participants did not have to enroll in a research study to be included. Furthermore, the use of a clinical population in our study provides an opportunity to observe the relationships between DD and A1C over time, outside of a formal research intervention setting. It is worth noting that, although there was no formal intervention to address mental health comorbidities, it is possible that increased provider awareness of DD, depressive symptoms, or suicide risk from routine clinic screening may have influenced delivery of diabetes care, thus resulting in our study not being purely observational. This study also adds to the paucity of literature on suicide risk and risk factors in this unique population.
A potential weakness of this study is its reliance on item 9 of the PHQ-9 to assess suicide risk, which may under-identify suicidality compared with comprehensive psychological assessments such as the Columbia-Suicide Severity Rating Scale (21). Additionally, the availability of clinical screening tools in only English or Spanish resulted in an underrepresentation of patients whose language for care was not English or Spanish. This study did not include data beyond March 2020, given the impact of the COVID-19 pandemic on screening, but it is important to do additional studies to understand how the pandemic affected mental health outcomes and A1C levels.
Conclusion
Our findings confirm that AYAs with psychological comorbidities have higher A1C levels, and this relationship is stable over time. AYAs with older age, female sex, and an MDI insulin regimen appear to experience higher rates of DD and depressive symptoms. Suicide risk was associated with older age and MDI insulin use. Tailored interventions to address and prevent psychological comorbidities in this population are warranted.
Article Information
Acknowledgments
A.J.R. was supported by the Seattle Children’s Hospital and University of Washington Department of Pediatrics Quality Improvement Scholars Program. J.P.Y.-F. has received funds for unrelated work from the National Institutes of Health. F.S.M. was supported in part by a K23 Career Development Award from the National Institute of Diabetes and Digestive and Kidney Diseases of the National Institutes of Health (DK119465).
Duality of Interest
No potential conflicts of interest relevant to this article were reported.
Author Contributions
A.J.R. conceptualized and designed the study, coordinated and supervised data collection, and drafted the initial manuscript. K.C. contributed to data acquisition, the analysis plan and execution, and data interpretation. J.P.Y.-F., A.M., M.G., and F.S.M. conceptualized and designed the study and contributed to data interpretation. All authors reviewed, revised, and approved the final manuscript. A.J.R. 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.
References
- 1. Bernstein CM, Stockwell MS, Gallagher MP, Rosenthal SL, Soren K. Mental health issues in adolescents and young adults with type 1 diabetes: prevalence and impact on glycemic control. Clin Pediatr (Phila) 2013;52:10–15 [DOI] [PubMed] [Google Scholar]
- 2. Stahl-Pehe A, Lange K, Bächle C, Castillo K, Holl RW, Rosenbauer J. Mental health problems among adolescents with early-onset and long-duration type 1 diabetes and their association with quality of life: a population-based survey. PLoS One 2014;9:e92473. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Dennick K, Sturt J, Speight J. What is diabetes distress and how can we measure it? A narrative review and conceptual model. J Diabetes Complications 2017;31:898–911 [DOI] [PubMed] [Google Scholar]
- 4. Powers MA, Richter SA, Ackard DM, Craft C. Diabetes distress among persons with type 1 diabetes. Diabetes Educ 2017;43:105–113 [DOI] [PubMed] [Google Scholar]
- 5. Hagger V, Hendrieckx C, Cameron F, Pouwer F, Skinner TC, Speight J. Diabetes distress is more strongly associated with HbA1c than depressive symptoms in adolescents with type 1 diabetes: results from Diabetes MILES Youth-Australia. Pediatr Diabetes 2018;19:840–847 [DOI] [PubMed] [Google Scholar]
- 6. Buchberger B, Huppertz H, Krabbe L, Lux B, Mattivi JT, Siafarikas A. Symptoms of depression and anxiety in youth with type 1 diabetes: a systematic review and meta-analysis. Psychoneuroendocrinology 2016;70:70–84 [DOI] [PubMed] [Google Scholar]
- 7. McGrady ME, Hood KK. Depressive symptoms in adolescents with type 1 diabetes: associations with longitudinal outcomes. Diabetes Res Clin Pract 2010;88:e35–e37 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Kristensen LJ, Birkebaek NH, Mose AH, Hohwü L, Thastum M. Symptoms of emotional, behavioral, and social difficulties in the Danish population of children and adolescents with type 1 diabetes: rresults of a national survey. PLoS One 2014;9:e97543. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Majidi S, O’Donnell HK, Stanek K, Youngkin E, Gomer T, Driscoll KA. Suicide risk assessment in youth and young adults with type 1 diabetes. Diabetes Care 2020;43:343–348 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Matlock KA, Yayah Jones NH, Corathers SD, Kichler JC. Clinical and psychosocial factors associated with suicidal ideation in adolescents with type 1 diabetes. J Adolesc Health 2017;61:471–477 [DOI] [PubMed] [Google Scholar]
- 11. American Diabetes Association . 13. Children and adolescents: Standards of Medical Care in Diabetes—2021. Diabetes Care 2021;44(Suppl. 1):S180–S199 [DOI] [PubMed] [Google Scholar]
- 12. Young-Hyman D, de Groot M, Hill-Briggs F, Gonzalez JS, Hood K, Peyrot M. Psychosocial care for people with diabetes: a position statement of the American Diabetes Association. Diabetes Care 2016;39:2126–2140 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Harrington KR, Shapira A, Volkening LK, et al. Associations of diabetes self-management characteristics, HbA1c, and psychosocial outcomes with depressive symptoms in a contemporary sample of adolescents with type 1 diabetes. J Diabetes Complications 2021;35:107838. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Roberts AJ, Barry D, Yi-Frazier J, Rutman L, Pihoker C, Malik FS. Screening for mental health comorbidities in a pediatric diabetes clinic setting. Clin Diabetes 2021;39:97–101 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Lee JM, Carlson E, Albanese-O’Neill A, et al. Adoption of telemedicine for type 1 diabetes care during the COVID-19 pandemic. Diabetes Technol Ther 2021;23:642–651 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Shapiro JB, Vesco AT, Weil LEG, Evans MA, Hood KK, Weissberg-Benchell J. Psychometric properties of the Problem Areas in Diabetes: Teen and Parent of Teen versions. J Pediatr Psychol 2018;43:561–571 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Richardson LP, McCauley E, Grossman DC, et al. Evaluation of the Patient Health Questionnaire-9 Item for detecting major depression among adolescents. Pediatrics 2010;126:1117–1123 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Iturralde E, Rausch JR, Weissberg-Benchell J, Hood KK. Diabetes-related emotional distress over time. Pediatrics 2019;143:e20183011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Weissberg-Benchell J, Shapiro JB, Bryant FB, Hood KK. Supporting Teen Problem-Solving (STEPS) 3 year outcomes: preventing diabetes-specific emotional distress and depressive symptoms in adolescents with type 1 diabetes. J Consult Clin Psychol 2020;88:1019–1031 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Wasserman RM, Eshtehardi SS, Anderson BJ, Weissberg-Benchell JA, Hilliard ME. Profiles of depressive symptoms and diabetes distress in preadolescents with type 1 diabetes. Can J Diabetes 2021;45:436–443 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Moss AC, Roberts AJ, Yi-Frazier JP, et al. Identifying suicide risk in adolescents and young adults with type 1 diabetes: are depression screeners sufficient? Diabetes Care 2022;45:1288–1291 [DOI] [PMC free article] [PubMed] [Google Scholar]