OBJECTIVE

Type 1 diabetes is associated with an increased risk of psychiatric morbidities. We investigated predictors and diabetes outcomes in a pediatric population with and without psychiatric comorbidities.

RESEARCH DESIGN AND METHODS

Data from the Danish Registry of Childhood and Adolescent Diabetes (DanDiabKids) and National Patient Register were collected (1996–2015) for this population-based study. We used Kaplan-Meier plots to investigate whether age at type 1 diabetes onset and average glycated hemoglobin (HbA1c) levels during the first 2 years after onset of type 1 diabetes (excluding HbA1c at debut) were associated with the risk of being diagnosed with a psychiatric disorder. Mixed-effects linear and logistic regression models were used to analyze HbA1c, BMI, severe hypoglycemia (SH), or ketoacidosis as outcomes, with psychiatric comorbidities as explanatory factor.

RESULTS

Among 4,725 children and adolescents with type 1 diabetes identified in both registers, 1,035 were diagnosed with at least one psychiatric disorder. High average HbA1c levels during the first 2 years predicted higher risk of psychiatric diagnoses. Patients with psychiatric comorbidity had higher HbA1c levels (0.22% [95% CI 0.15; 0.29]; 2.40 mmol/mol [1.62; 3.18]; P < 0.001) and an increased risk of hospitalization with diabetic ketoacidosis (1.80 [1.18; 2.76]; P = 0.006). We found no associations with BMI or SH.

CONCLUSIONS

High average HbA1c levels during the first 2 years after onset of type 1 diabetes might indicate later psychiatric comorbidities. Psychiatric comorbidity in children and adolescents with type 1 diabetes increases the risk of poor metabolic outcomes. Early focus on the disease burden might improve outcomes.

Type 1 diabetes in childhood has been found to be associated with an increased risk of psychiatric comorbidities (13), which might intensify the burden of disease and accelerate metabolic deterioration (46), subsequently increasing the risk of mortality and long-term complications such as retinopathy, nephropathy, and neuropathy (79).

Metabolic dysregulation is closely linked to age and diabetes duration, showing a peak in adolescence and early adulthood (10,11). Early adolescence is also characterized as a time of psychological vulnerability (12), in which the incidence of major psychiatric disorders increases (13). A diagnosis of type 1 diabetes in early adolescence seems to increase psychological distress (1,2), and three large population-based studies have shown higher rates of psychiatric disorders in children and adolescents with type 1 diabetes compared with the general population (13). In particular, increased risk was seen for depression, anxiety, and eating disorders, where the pathogenesis is considered to involve reactive mechanisms and imbalances in the diathesis-stress system (13,14). One of these population-based studies, based on the Danish National Patient Register (NPR), found the highest risk of psychiatric comorbidity among children diagnosed with type 1 diabetes at age 10–14 years and among children with diabetes duration >5 years (2).

Another population-based study found increased glycated hemoglobin (HbA1c) levels before diagnosis of a psychiatric disorder (3). However, a small longitudinal study failed to confirm this (15). It is unclear how soon before the diagnosis of psychiatric comorbidity that elevated HbA1c levels can be detected.

Several clinical studies (1620) and a few population-based studies (46) on children and adolescents with type 1 diabetes have detected high HbA1c levels after the diagnosis of a psychiatric comorbidity. The German/Austrian diabetes register study group investigated HbA1c levels and other diabetes-related outcomes in patients with psychiatric comorbidities, including depression, eating disorders, and psychosis. They obtained mixed results regarding BMI but observed that overall diabetes outcomes were poorer and characterized by higher HbA1c levels, more episodes of severe hypoglycemia (SH) and diabetic ketoacidosis (DKA), and longer, more frequent hospitalizations (46).

Despite clinical and research evidence that a child with type 1 diabetes often receives more than one psychiatric diagnosis (1,3), most studies evaluate one disorder at a time (46,1620).

Motivated by findings that Danish children and adolescents with type 1 diabetes have a higher risk of developing a psychiatric disorder compared with the background population (2), we performed two studies based on the NPR and the Danish Registry of Childhood and Adolescent Diabetes (DanDiabKids). First, we investigated age at diabetes onset and metabolic control within the first 2 years of type 1 diabetes as predictors of psychiatric comorbidity, including all psychiatric disorders (study A). Second, we investigated the association of psychiatric comorbidities with a broader range of diabetes-related outcomes in children and adolescents (study B).

Study Design and Population

This population-based study used personal identification numbers to merge data from national registries within Denmark. All Danish citizens have a unique civil personal registration (CPR) number (21). The CPR number can be used to accurately merge data from different national registers to generate comprehensive patient records (21). Danish citizens have free access to health care services, and almost all contact that children and adolescents have with hospitals, including psychiatric inpatient and outpatient settings, is with public hospitals. All children and adolescents with type 1 diabetes are monitored in public hospitals.

The NPR contains psychiatric and somatic diagnoses from all inpatient admissions to Danish public hospitals since 1977. From 1995 and onwards, the register includes inpatient, outpatient, and emergency details. The register has used the ICD-10 since 1994 (22,23). Data on registration of psychiatric and type 1 diabetes diagnoses were collected from the NPR, covering 1996 to April 2015.

DanDiabKids collects information on all children and adolescents diagnosed with type 1 diabetes before the age of 15 years and monitors them until they are transferred to adult clinics at ∼18 years of age. All public hospital pediatric units must supply annual data on all patients with diabetes to DanDiabKids. The classification of type 1 diabetes is based on the guidelines drafted by the International Society for Pediatric and Adolescent Diabetes (ISPAD). DanDiabKids contains annual data on all registered patients since 1996, including information on quality indicators, demographic variables, associated conditions, diabetes classification, diabetes family history, growth, self-management, and treatment variables. DanDiabKids now covers 99% of all Danish children and adolescents diagnosed with type 1 diabetes before the age of 15 years. However, the coverage of clinical data included in the register differs from year to year and ranges between 80% and 95% (24). Clinical data on all patients included in this study was collected from DanDiabKids for the years 1996–2015.

Our study population was generated by merging data from DanDiabKids and the NPR. The inclusion criteria were registration with type 1 diabetes in DanDiabKids, age at onset <15 years, year of onset 1995–2014, and year of birth after 1980. Clinical data were included until 30 April 2015, with visits until the age of 18 years. Data on registration of a psychiatric diagnosis were included for all patients during the entire observational period (1996–April 2015) (e.g., before onset of type 1 diabetes, while in pediatric care, and after referral to adult care).

Measures of Psychiatric Morbidity

In Denmark, the majority of children and adolescents in need of psychiatric assessment and treatment are referred to public hospitals. In these settings, patients are examined and diagnosed according to the clinical and diagnostic guidelines described by ICD-10 (13,23). Thereafter, records of the diagnoses are kept in the NPR.

The ICD-10 diagnoses included in this study as “any psychiatric disorder” were all diagnoses on mental and behavioral disorders from the F section, F00–99, and diagnoses from the X section, X60–84. The diagnoses are listed in Tables 1 and 2 (2,13).

Table 1

Frequencies of psychiatric diagnoses

ICD-10 diagnosesNumber
F0–9: Organic, including symptomatic, mental disorders 14 
F10–19: Mental and behavioral disorders due to psychoactive substance use 190 
F20–29: Schizophrenia, schizotypical and delusional disorders 55 
F30–39: Mood (affective) disorders 205 
F40–48: Neurotic, stress-related, and somatoform disorders, including anxiety disorder 492 
F50: Eating disorders 75 
F51–59: Behavioral syndromes associated with physiological disturbances and physical factors 12 
F60–69: Disorders of adult personality and behavior 104 
F70–79: Mental retardation 56 
F80–83: Specific developmental disorders 72 
F84: Pervasive developmental disorders, including autism spectrum disorders 114 
F88–89: Other disorders of psychological development 14 
F90, F98.8: Hyperkinetic disorder (ADHD) and ADD 172 
F91–95, F98.0–98.6: Behavioral and emotional disorders with onset usually occurring in childhood and adolescence (excluding F90 and F98.8) 310 
X60–X84: Intentional self-harm, including suicide attempts 27 
Total number of diagnoses 1,912 
ICD-10 diagnosesNumber
F0–9: Organic, including symptomatic, mental disorders 14 
F10–19: Mental and behavioral disorders due to psychoactive substance use 190 
F20–29: Schizophrenia, schizotypical and delusional disorders 55 
F30–39: Mood (affective) disorders 205 
F40–48: Neurotic, stress-related, and somatoform disorders, including anxiety disorder 492 
F50: Eating disorders 75 
F51–59: Behavioral syndromes associated with physiological disturbances and physical factors 12 
F60–69: Disorders of adult personality and behavior 104 
F70–79: Mental retardation 56 
F80–83: Specific developmental disorders 72 
F84: Pervasive developmental disorders, including autism spectrum disorders 114 
F88–89: Other disorders of psychological development 14 
F90, F98.8: Hyperkinetic disorder (ADHD) and ADD 172 
F91–95, F98.0–98.6: Behavioral and emotional disorders with onset usually occurring in childhood and adolescence (excluding F90 and F98.8) 310 
X60–X84: Intentional self-harm, including suicide attempts 27 
Total number of diagnoses 1,912 

Total number of diagnoses registered in the NPR for the 1,035 patients with both type 1 diabetes and a psychiatric comorbidity.

Table 2

Categorization of psychiatric disorders

Disorders of neurodevelopment and constitutional origin 
 F20–29 Schizophrenia, schizotypical and delusional disorders 
 F60–69 Disorders of adult personality and behavior 
 F70–79 Mental retardation 
 F80–83 Specific developmental disorders 
 F84 Pervasive developmental disorders, including autism spectrum disorders 
 F88–89 Other disorders of psychological development 
 F90, F98.8 Hyperkinetic disorder (ADHD) and ADD 
 F94 Disorders of early social functioning 
 F95 Tic disorders 
Disorders of potentially reactive origin 
 F10–19 Mental and behavioral disorders due to psychoactive substance use 
 F30–39 Mood (affective) disorders 
 F40–48 Neurotic, stress-related, and somatoform disorders, including anxiety disorders 
 F50, F98.2 Eating disorders 
 F91–93, F98.0–98.6 Behavioral and emotional disorders with onset in childhood and adolescence, excluding F98.2 
 X60–X84 Intentional self-harm, including suicide attempts 
Uncategorized 
 F0–9 Organic, including symptomatic, mental disorders 
 F51–59 Behavioral syndromes associated with physiological disturbances and physical factors, excluding eating disorders 
Disorders of neurodevelopment and constitutional origin 
 F20–29 Schizophrenia, schizotypical and delusional disorders 
 F60–69 Disorders of adult personality and behavior 
 F70–79 Mental retardation 
 F80–83 Specific developmental disorders 
 F84 Pervasive developmental disorders, including autism spectrum disorders 
 F88–89 Other disorders of psychological development 
 F90, F98.8 Hyperkinetic disorder (ADHD) and ADD 
 F94 Disorders of early social functioning 
 F95 Tic disorders 
Disorders of potentially reactive origin 
 F10–19 Mental and behavioral disorders due to psychoactive substance use 
 F30–39 Mood (affective) disorders 
 F40–48 Neurotic, stress-related, and somatoform disorders, including anxiety disorders 
 F50, F98.2 Eating disorders 
 F91–93, F98.0–98.6 Behavioral and emotional disorders with onset in childhood and adolescence, excluding F98.2 
 X60–X84 Intentional self-harm, including suicide attempts 
Uncategorized 
 F0–9 Organic, including symptomatic, mental disorders 
 F51–59 Behavioral syndromes associated with physiological disturbances and physical factors, excluding eating disorders 

Further, we grouped the included ICD-10 diagnoses in the following two groups: 1) neurodevelopmental/constitutional disorders and 2) potentially reactive disorders, as shown in Table 2. These subdivisions are based on the literature (14), previous research findings (13), and current classification schemes (23). We made the a priori assumption that disorders of early developmental and, hereby, constitutional origin (mental retardation, specific learning disorders, autism spectrum disorders, attention-deficit/hyperactivity disorder (ADHD), and other neurobehavioral and personality disorders) would show a different pattern of risk associations, compared with disorders that are considered more stress dependent and reactive in nature (emotional and affective disorders, eating disorders, disorders of substance abuse, self-harm, and suicide attempts).

Diabetes Outcome Variables and Covariates

Clinical data on all patients registered in DanDiabKids and the NPR were collected from the DanDiabKids register for the period 1996 to April 2015. The variables and covariates included were as follows.

HbA1c

HbA1c levels were measured centrally using a high-pressure liquid chromatography method (Tosoh Bioscience, South San Francisco, CA) with a normal range of 4.3–5.8% (23.5–40 mmol/mol). HbA1c levels were categorized as optimal control (HbA1c <7.5% [<58 mmol/mol]), intermediate control (7.5–8.5% [58–70 mmol/mol]), and high HbA1c (HbA1c >8.5% [>70 mmol/mol]).

SH

SH was defined according to the ISPAD guidelines as an event associated with severe neuroglycopenia resulting in coma or seizure and requiring glucagon or intravenous glucose (25).

DKA

DKA was defined by ISPAD guidelines as blood glucose >11 mmol/L, venous pH <7.3 or bicarbonate <15 mmol/L, and ketonemia and ketonuria (26).

BMI

BMI was calculated as height (cm)/weight (kg)2 and converted to a BMI SD score (BMI-SDS) using national reference material (27).

Background Variables

The background variables were sex, age, diabetes duration (divided into three classes: 0–2, >2–4, and >4 years), and ethnicity (defined as Danish when at least one parent was of Danish origin, or immigrants and offspring of immigrants).

Treatment

Individuals self-monitoring blood glucose (SMBG) or using a continuous glucose monitoring sensor (CGMS) were categorized as SMBG (<3, 3–5, 6–9, or >10 times daily) or using a CGMS for a minimum of 6 days/month. Insulin treatment was categorized according to the basal type of insulin used and the injection method: Mixtard insulin, NPH insulin, or analog or pump treatment. The number of injections/boluses over a given period was included as a continuous variable.

Statistics

Descriptive Data

Data were described by suitable summary statistics: mean (SD) or number (n [%]). Time-dependent variables were summarized across all visits. Descriptive data for all children included mean age at type 1 diabetes diagnosis and mean diabetes duration at last visit. Frequencies were calculated for sex and ethnicity and minimum and maximum for birth year. HbA1c levels, treatment (insulin type and dose), and self-management (insulin bolus injection and SMBG or use of CGMS) were summarized by the mean of all visits or sum over all visits. Observations were presented for all children together and divided into two categories: with a psychiatric disorder or with type 1 diabetes only. For each child, visits prior to the first registered diagnosis of any psychiatric disorder were included in the category with diabetes only, whereas visits after this date were included in the category with any psychiatric disorder.

Study A: Predictors of Psychiatric Comorbidity

To investigate associations between psychiatric comorbidities and age at type 1 diabetes onset or mean HbA1c within the first 2 years of type 1 diabetes onset, we estimated the cumulative incidence of psychiatric diagnoses using Proc Lifetest in SAS statistical software (version 9.4; SAS Institute, Inc., Cary, NC). Age at type 1 diabetes onset was grouped (0–5, 5–10, and 10–15 years old). Children with a psychiatric disorder before onset of type 1 diabetes were excluded. Mean HbA1c levels were based on recordings made within 2 years after type 1 diabetes onset, excluding values measured at onset. The 2-year limit was chosen because DanDiabKids data are only collected once a year around the birthday of the child; therefore many children could still be in remission if only the first HbA1c value registered in DanDiabKids was used to categorize HbA1c groups. Mean duration of type 1 diabetes for the included HbA1c levels was 1.26 years (range 0.5–2 years). In the analyses of the association between HbA1c and the risk of psychiatric comorbidity, children diagnosed with a psychiatric disorder before type 1 diabetes onset or during the first 2 years were excluded. Unadjusted group comparisons were made using a log-rank test comparing the cumulative incidence between groups.

Study B: Outcome Measurements and Psychiatric Comorbidity

To compare HbA1c levels between groups with or without psychiatric comorbidity, we used mixed-effects linear models for repeated measurements, with a random effect at the person-level, and an autoregressive residual correlation structure (SAS Proc Mixed). The first model with HbA1c as outcome and psychiatric disorder as explanatory variable was adjusted for age (cubic), sex, diabetes duration (<2 years, 2–4 years, or >4 years), ethnicity, and visit year (5-year intervals). In model 2, we added SMBG measurements and treatment as explanatory variables. All visits prior to a psychiatric diagnosis were categorized as time with no psychiatric disorder, whereas all subsequent visits were categorized as time with a psychiatric disorder. The linear models were repeated with psychiatric diagnoses divided into two main groups of psychiatric disorders (neurodevelopmental/constitutional disorders and potentially reactive disorders, as shown in Table 2) and a third group of children and adolescents diagnosed with a combination of neurodevelopmental/constitutional and potentially reactive disorders. All visits were coded according to the diagnoses the patient had at the time of the visit.

To test for differences in BMI-SDS between psychiatric groups, mixed-effects linear models were used. A mixed-effects logistic regression model (SAS Proc Glimmix) was used to compare the odds of SH events or DKA between the psychiatric groups. The models for BMI, SH events, and DKA were all adjusted for HbA1c levels in addition to the covariates included in model 2.

Ethical Approval

The study was approved by the Danish Data Protection Agency. DanDiabKids was approved by ethics committee number KA 95139M. In Denmark, further ethical approval is unnecessary for studies that use only registry data.

Basic Characteristics

After merging DanDiabKids with the NPR, 4,725 children and adolescents with type 1 diabetes were identified; 259 with a diagnosis of type 1 diabetes were missing from the NPR and therefore excluded. The missing case subjects were evenly distributed through the years 1996–2014. Characteristics for the included subjects were as follows: mean age at onset of diabetes was 8.98 years (SD 3.81), birth year ranged from 1980 to 2013, mean age at last visit was 14.6 years (3.7), 2,462 (52.1%) were boys, mean duration of diabetes at last visit was 5.65 years (3.7), 4,434 (93.8%) were of Danish origin, 254 (5.4%) were immigrants or offspring of immigrants, and 36 (0.8%) had unknown ethnicity. Treatment and self-management characteristics can be found in Table 3.

Table 3

Treatment and self-management characteristics

VariablesAllNo psychiatric disorderAny psychiatric disorder
HbA1c*, % (mmol/mol) 8.3 (66.7) 8.2 (66.4) 8.5 (69.5) 
SMBG**, n (%)    
 <3 daily 1,588 (5.2) 1,393 (5.0) 195 (6.4) 
 3–5 daily 11,127 (36.2) 9,895 (35.7) 1,232 (40.7) 
 6–9 daily 8,726 (28.4) 7,833 (28.2) 893 (29.5) 
 >10 daily 1,731 (5.6) 1,533 (5.5) 198 (6.5) 
 CGMS (minimum 6 days/month) 276 (0.9) 244 (0.9) 32 (1.1) 
 Missing 7,324 (23.8) 6,847 (24.7) 477 (15.8) 
Treatment**, n (%)    
 Mixtard 8,254 (26.8) 7,830 (28.2) 424 (14.0) 
 NPH 5,302 (17.2) 4,881 (17.6) 421 (13.9) 
 Analog 6,201 (20.1) 5,303 (19.1) 898 (29.7) 
 Pump treatment 8,660 (28.1) 7,598 (27.4) 1,062 (35.1) 
 Missing 2,355 (7.7) 2,133 (7.7) 222 (7.3) 
Number of injections/day* 2.93 2.86 3.65 
Insulin per kg* 0.90 0.89 0.96 
VariablesAllNo psychiatric disorderAny psychiatric disorder
HbA1c*, % (mmol/mol) 8.3 (66.7) 8.2 (66.4) 8.5 (69.5) 
SMBG**, n (%)    
 <3 daily 1,588 (5.2) 1,393 (5.0) 195 (6.4) 
 3–5 daily 11,127 (36.2) 9,895 (35.7) 1,232 (40.7) 
 6–9 daily 8,726 (28.4) 7,833 (28.2) 893 (29.5) 
 >10 daily 1,731 (5.6) 1,533 (5.5) 198 (6.5) 
 CGMS (minimum 6 days/month) 276 (0.9) 244 (0.9) 32 (1.1) 
 Missing 7,324 (23.8) 6,847 (24.7) 477 (15.8) 
Treatment**, n (%)    
 Mixtard 8,254 (26.8) 7,830 (28.2) 424 (14.0) 
 NPH 5,302 (17.2) 4,881 (17.6) 421 (13.9) 
 Analog 6,201 (20.1) 5,303 (19.1) 898 (29.7) 
 Pump treatment 8,660 (28.1) 7,598 (27.4) 1,062 (35.1) 
 Missing 2,355 (7.7) 2,133 (7.7) 222 (7.3) 
Number of injections/day* 2.93 2.86 3.65 
Insulin per kg* 0.90 0.89 0.96 

*Presented as the mean of all visits for all the groups with no psychiatric and any psychiatric disorder.

**Presented as the sum over all visits for all the groups with no psychiatric and any psychiatric disorder.

The observed number of SH and DKA events per 100 person-years was respectively 10.7 (SH) and 3.2 (DKA) in patients with neurodevelopmental/constitutional psychiatric disorder, 12.1 (SH) and 3.7 (DKA) in patients with potentially reactive psychiatric disorder, 12.3 (SH) and 6.4 (DKA) in patients with both types of psychiatric disorders, and 8.1 (SH) and 1.8 (DKA) in patients without psychiatric disorder.

Incident Psychiatric Comorbidity

Among the 4,725 children and adolescents included in the study, 1,035 were diagnosed with at least one psychiatric disorder at some point. Of these, a total of 175 received their first psychiatric diagnosis before the onset of type 1 diabetes, 575 during pediatric care, and 285 were diagnosed after referral to adult care.

Psychiatric diagnoses among the 1,035 patients with psychiatric comorbidity are shown in Table 1. Anxiety disorders were the most common (n = 492), followed by “behavioral and emotional disorders” (n = 310), mood disorders (n = 205), psychoactive substance misuse disorders (n = 190), and disorders of inattention and hyperactivity (ADHD/attention-deficit disorder [ADD]) (n = 172). Of the 1,035 patients, 46% were diagnosed with two or more psychiatric disorders and 22.8% were diagnosed with three or more psychiatric disorders. The co-occurrence of diagnoses is presented in Supplementary Table 1.

Study A: Predictors of Psychiatric Comorbidity

The cumulative incidence of psychiatric disorders differed among age-groups (P = 0.001) (Fig. 1). Shortly after type 1 diabetes diagnosis, a higher estimated risk of psychiatric disorders was evident among patients who were 10–15 years old at onset of type 1 diabetes. However, after 15–20 years with diabetes, the differences among the groups leveled out at a risk of ∼30% (Fig. 1). The cumulative incidence of psychiatric disorders also differed significantly among HbA1c groups (P = 0.02) (Fig. 2). Children with high mean HbA1c levels (>8.5% [>70 mmol/mol]) during the first 2 years showed the highest estimated risk of developing a psychiatric disorder, although these differences also appear to level out after 15–20 years with type 1 diabetes.

Figure 1

Risk of a psychiatric diagnosis as a function of diabetes duration (from onset of type 1 diabetes until end of the observational period in the NPR [30 April 2015]) in children and adolescents divided into three groups according to age at type 1 diabetes onset (0–5 years, blue line; 5–10 years, red line; 10–15 years, green line).

Figure 1

Risk of a psychiatric diagnosis as a function of diabetes duration (from onset of type 1 diabetes until end of the observational period in the NPR [30 April 2015]) in children and adolescents divided into three groups according to age at type 1 diabetes onset (0–5 years, blue line; 5–10 years, red line; 10–15 years, green line).

Figure 2

Risk of psychiatric diagnosis as a function of diabetes duration (from onset of type 1 diabetes until end of the observational period in the NPR [30 April 2015]) in children and adolescents divided into three groups according to average HbA1c values during the first 2 years after type 1 diabetes onset (excluding values measured at onset). Good metabolic control (<7.5% [<58 mmol/mol]), green line; intermediate metabolic control (7.5–8.5% [58–70 mmol/mol]), blue line; poor metabolic control (>8.5% [>70 mmol/mol]), red line.

Figure 2

Risk of psychiatric diagnosis as a function of diabetes duration (from onset of type 1 diabetes until end of the observational period in the NPR [30 April 2015]) in children and adolescents divided into three groups according to average HbA1c values during the first 2 years after type 1 diabetes onset (excluding values measured at onset). Good metabolic control (<7.5% [<58 mmol/mol]), green line; intermediate metabolic control (7.5–8.5% [58–70 mmol/mol]), blue line; poor metabolic control (>8.5% [>70 mmol/mol]), red line.

Study B: Outcome Measurements and Psychiatric Comorbidity

The mean HbA1c level was higher in children with a psychiatric disorder (0.22% [95% CI 0.15; 0.29]; 2.45 mmol/mol [1.67; 3.22]) compared with children with no psychiatric disorder (P < 0.001) (model 1); also after adjusting for treatment and self-management, the estimated difference in HbA1c was 0.22% (0.15; 0.29) (2.40 mmol/mol [1.62; 3.18]; P < 0.001) (model 2).

There was no significant elevation in HbA1c levels in patients with neurodevelopmental/constitutional disorders compared with patients with no psychiatric disorder (0.02% [−0.13; 0.16]; 0.21 mmol/mol [−1.38; 1.80]; P = 0.80), whereas HbA1c levels were elevated by 0.28% (0.19; 0.37) (3.04 mmol/mol [2.05; 4.03]; P < 0.001) in patients with potentially reactive disorders and by 0.29% (0.13; 0.44) (3.19 mmol/mol [1.46; 4.92]; P < 0.001) in patients with both types of psychiatric disorders. After additional adjustment for treatment and self-management (model 2), these differences changed to 0.01% (−0.13; 0.15) (0.12 mmol/mol [−1.40; 1.65]; P = 0.88) for patients with neurodevelopmental/constitutional disorders, 0.28% (0.19; 0.37) (3.08 mmol/mol [2.12; 4.04]; P < 0.001) for patients with potentially reactive disorders, and 0.28% (0.12; 0.43) (3.03 mmol/mol [1.36; 4.70]; P = 0.001) for patients with both types of disorders.

No association was found between any psychiatric disorder and BMI outcome (estimated difference in BMI-SDS 0.029 [−0.026; 0.084]; P = 0.29), regardless of psychiatric subgroups (P = 0.68). The odds of hospitalization with DKA was increased (odds ratio 1.80 [1.18; 2.76]; P = 0.006), whereas no increased risk of SH events was found (odds ratio 0.94 [0.84; 1.34]; P = 0.63). No differences in odds of DKA (P = 0.29) or SH (P = 0.91) were found regarding psychiatric subgroups.

This is the first nationwide register-based investigation showing associations between diabetes outcomes in young patients with type 1 diabetes and any comorbid psychiatric disorder. High HbA1c levels in the early period after type 1 diabetes onset seem to be a possible indicator for subsequent psychiatric disorders, and having a psychiatric disorder was associated with higher HbA1c levels, especially in patients with disorders of putative reactive pathogenesis. Given that the Kaplan-Meier plots showed that the estimated risk of being diagnosed with a psychiatric disorder within a period of 15–20 years of type 1 diabetes onset was close to 30% in most groups, our finding highlights an important clinical problem.

Findings in Relation to Other Studies

Previous studies have demonstrated that type 1 diabetes in early adolescence is associated with increased risk of psychiatric comorbidities (1,2), suggesting a particularly sensitive developmental period regarding patients learning to cope with a chronic disease. We found that type 1 diabetes onset at 10–15 years of age was associated with a more rapid onset of psychiatric disorders; although children with earlier type 1 diabetes onset showed similar risk of psychiatric disorders when they reached adolescence. This finding underscores a particularly increased psychiatric vulnerability in early adolescence for young people facing type 1 diabetes (1214).

Few studies have examined HbA1c levels prior to the diagnosis of a psychiatric disorder. The higher rate of psychiatric disorders in children with high average HbA1c levels early after onset of type 1 diabetes is consistent with results obtained by Cooper et al. (3), who found a hazard ratio for any psychiatric disorder of 1.31, associated with a 1% (10 mmol/mol) increase in mean HbA1c level, across all clinic visits between the diagnosis of type 1 diabetes and the psychiatric disorder. Many factors may affect metabolic control and the disease burden. These include the severity of the disease, socioeconomic status of the family, and the patient’s ability to cope with a severe chronic condition. Problems associated with mental health (diagnosed or undiagnosed disorders) and cognitive capabilities (known or unknown disabilities) may also be exacerbated in the child and/or parents due to the many pressures and tasks required to control diabetes. Even though this register-based study does not explain the cause-effect relationship between HbA1c levels and psychiatric comorbidities, poor metabolic control seen during the early years after onset of type 1 diabetes may be a marker of psychological burden.

Our finding that having a psychiatric disorder is associated with compromised treatment outcomes is consistent with previous clinical (1620) and population-based studies (46) in pediatric populations; although previous studies have investigated individual disorders and none included the entire range of psychiatric disorders.

Our categorization of psychiatric disorders into overarching groups seems justified because nearly half of the patients had co-occurring psychiatric diagnoses. We found that having a psychiatric diagnosis from the potentially reactive disorder group or a combination of both types of disorder was associated with higher HbA1c levels. In comparison, the German/Austrian diabetes register study generally found greater differences in metabolic outcome (0.3–0.8% [3.5–9 mmol/mol]) when they investigated individual diagnoses (46).

In the current study, no increased risk of higher HbA1c levels was found in children with neurodevelopmental/constitutional disorders. Our finding may reflect the influences of health care on children with neurodevelopmental disorders in Denmark, possibly including the support of parents to follow the diabetes guidelines. Still, for the young people themselves, neurodevelopmental problems and associated impairments in executive functioning may also result in poor adherence to diabetes treatments and a deterioration in metabolic control (19,20).

There were few events of SH or DKA in our population, indicating effective prevention of severe acute complications. However, we confirmed a link established by earlier studies (46) between psychiatric comorbidity and hospitalization due to DKA. We observed no effect of psychiatric comorbidity on BMI. This might be due to bidirectional effects on BMI caused by the particular groupings.

The greater differences in HbA1c values and higher rates of acute complications in the German/Austrian studies compared with our results might be due to methodological differences. In the German/Austrian studies, registration depended on the diabetes team being aware of the psychiatric comorbidity and/or medication, whereas we used registrations from the NPR that were independent of diabetes treatment. When registration depends on the diabetes team, patients with high HbA1c levels or more obvious psychiatric symptoms may be more likely to be registered. Additionally, the German/Austrian studies only included data from visits close to the observed diagnosis, whereas we included all subsequent visits.

Implications

Individual variance of HbA1c trajectories in patients with childhood type 1 diabetes over time suggests that deterioration of metabolic control is not an inexorable condition of adolescents but a preventable complication (28). Our results point to an increased risk from disorders with a suggested diathesis-stress imbalance (14). The psychological burden raises the levels of counterregulatory hormones, increasing glucose production (29). This can amplify the demands/difficulties in adjusting insulin dosing in children and adolescents with type 1 diabetes and might influence cognitive function, increase family conflicts, affect school performance, and deteriorate health-related quality of life (3032).

Reactive psychiatric disorders might be prevented if preliminary symptoms are targeted early and if patients are supported in their everyday life to normalize life with diabetes (33,34). Psychological and problem-solving interventions have provided some success in improving metabolic control and psychosocial function (3538).

Strengths and Limitations

The primary strength of this study is its nationwide scope, achieved by using a database that includes the vast majority of children and adolescents with type 1 diabetes. All psychiatric diagnoses were confirmed by a medical doctor and registered in the NPR with 100% coverage (13).

One limitation is that patients with psychiatric disorders diagnosed prior to 1995 in an outpatient public clinic and patients diagnosed by a private psychiatrist were not registered in the NPR. We do not have data on the number of patients treated outside of public hospitals but suspect this is uncommon for children (2). Additionally, the number of patients diagnosed prior to 1995 would be limited because only patients with diabetes onset after 1996 were included. As we do not have dates of remission or recovery for potentially reversible psychiatric disorders, all patients were registered with a psychiatric disorder from the time of diagnosis, and this disorder was used in all subsequent visits in the analysis of diabetes treatment outcomes. This is a potential limitation regarding reactive disorders, but taking into account the risk of recurrence or persistency of these disorders in youngsters facing chronic stressful experiences (12), we believe that our approach is acceptable. We did not have information on the treatment of psychiatric disorders, diabetes-related family stress, quality of life, or socioeconomic factors, all of which might influence diabetes outcomes and the burden of the disease (4,5,32,39,40).

Summary

The estimated risk of developing a psychiatric disorder during the 15–20 years after type 1 diabetes diagnosis is high. The most vulnerable period appeared to be adolescence. Patients with poorly regulated diabetes shortly after onset had a higher estimated risk of developing psychiatric comorbidities. Young patients diagnosed with a psychiatric disorder had more episodes of DKA, and those diagnosed within the reactive spectrum had higher HbA1c levels. Children and adolescents with type 1 diabetes, and in particular those who fail to reach treatment goals, should be systematically evaluated regarding psychological vulnerabilities.

Acknowledgments. The authors thank DanDiabKids and all the patients and health care professionals for their time and effort in providing and managing data for the register and for making this study possible.

Funding. This project received funding from the Ronald McDonald Children’s Foundation, Helsefonden (15-B-0112), and the Mrs. C. Hermansens Fund.

The funding sources were not involved in any parts of the project planning, execution, or description.

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

Author Contributions. S.M.S. conceived the idea for the study, designed the study, interpreted the results, participated in the discussion, wrote the first draft of the manuscript, and revised the manuscript and approved the final version. N.B. performed the statistical analyses, interpreted the results, participated in the discussion, and revised the manuscript and approved the final version. E.B.L., J.S.T., and K.A.B. interpreted the results, participated in the discussion, and revised the manuscript and approved the final version. G.K.T. conceived the idea for the study, interpreted the results, participated in the discussion, and revised the manuscript and approved the final version. A.M.S. designed the study, interpreted the results, participated in the discussion, and revised the manuscript and approved the final version. J.S. conceived the idea for the study, designed the study, performed statistical analyses, interpreted the results, participated in the discussion, wrote the first draft of the manuscript, and revised the manuscript and approved the final version. J.S. 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. An abstract of this study was a poster presentation at the 57th Annual Meeting of the European Society of Paediatric Endocrinology, 27–29 September 2018, Athens, Greece.

1.
Butwicka
A
,
Frisén
L
,
Almqvist
C
,
Zethelius
B
,
Lichtenstein
P
.
Risks of psychiatric disorders and suicide attempts in children and adolescents with type 1 diabetes: a population-based cohort study
.
Diabetes Care
2015
;
38
:
453
459
[PubMed]
2.
Dybdal
D
,
Tolstrup
JS
,
Sildorf
SM
, et al
.
Increasing risk of psychiatric morbidity after childhood onset type 1 diabetes: a population-based cohort study
.
Diabetologia
2017
;
14
:
1
8
[PubMed]
3.
Cooper
MN
,
Lin
A
,
Alvares
GA
,
de Klerk
NH
,
Jones
TW
,
Davis
EA
.
Psychiatric disorders during early adulthood in those with childhood onset type 1 diabetes: rates and clinical risk factors from population-based follow-up
.
Pediatr Diabetes
2017
;
18
:
599
606
[PubMed]
4.
Plener
PL
,
Molz
E
,
Berger
G
, et al
.
Depression, metabolic control, and antidepressant medication in young patients with type 1 diabetes
.
Pediatr Diabetes
2015
;
16
:
58
66
[PubMed]
5.
Galler
A
,
Bollow
E
,
Meusers
M
, et al.;
German Federal Ministry of Education and Research (BMBF) Competence Network Diabetes Mellitus
.
Comparison of glycemic and metabolic control in youth with type 1 diabetes with and without antipsychotic medication: analysis from the nationwide German/Austrian Diabetes Survey (DPV)
.
Diabetes Care
2015
;
38
:
1051
1057
[PubMed]
6.
Scheuing
N
,
Bartus
B
,
Berger
G
, et al.;
DPV Initiative
;
German BMBF Competence Network Diabetes Mellitus
.
Clinical characteristics and outcome of 467 patients with a clinically recognized eating disorder identified among 52,215 patients with type 1 diabetes: a multicenter German/Austrian study
.
Diabetes Care
2014
;
37
:
1581
1589
[PubMed]
7.
White
NH
,
Cleary
PA
,
Dahms
W
,
Goldstein
D
,
Malone
J
,
Tamborlane
WV
;
Diabetes Control and Complications Trial (DCCT)/Epidemiology of Diabetes Interventions and Complications (EDIC) Research Group
.
Beneficial effects of intensive therapy of diabetes during adolescence: outcomes after the conclusion of the Diabetes Control and Complications Trial (DCCT)
.
J Pediatr
2001
;
139
:
804
812
[PubMed]
8.
Olsen
BS
,
Sjølie
AK
,
Hougaard
P
, et al.;
Danish Study Group of Diabetes in Childhood
.
The significance of the prepubertal diabetes duration for the development of retinopathy and nephropathy in patients with type 1 diabetes
.
J Diabetes Complications
2004
;
18
:
160
164
[PubMed]
9.
Sandahl
K
,
Nielsen
LB
,
Svensson
J
, et al
.
Increased mortality in a Danish cohort of young people with type 1 diabetes mellitus followed for 24 years
.
Diabet Med
2017
;
34
:
380
386
[PubMed]
10.
Miller
KM
,
Foster
NC
,
Beck
RW
, et al.;
T1D Exchange Clinic Network
.
Current state of type 1 diabetes treatment in the U.S.: updated data from the T1D Exchange clinic registry
.
Diabetes Care
2015
;
38
:
971
978
[PubMed]
11.
Hofer
SE
,
Raile
K
,
Fröhlich-Reiterer
E
, et al.;
Austrian/German Diabetes Patienten Verlaufsdokumentation DPV Initiative
;
German Competence Network for Diabetes Mellitus
.
Tracking of metabolic control from childhood to young adulthood in type 1 diabetes
.
J Pediatr
2014
;
165
:
956
961.e1-2
[PubMed]
12.
Patton
GC
,
Coffey
C
,
Romaniuk
H
, et al
.
The prognosis of common mental disorders in adolescents: a 14-year prospective cohort study
.
Lancet
2014
;
383
:
1404
1411
[PubMed]
13.
Pedersen
CB
,
Mors
O
,
Bertelsen
A
, et al
.
A comprehensive nationwide study of the incidence rate and lifetime risk for treated mental disorders
.
JAMA Psychiatry
2014
;
71
:
573
581
[PubMed]
14.
Rutter
M
,
Caspi
A
,
Moffitt
TE
.
Using sex differences in psychopathology to study causal mechanisms: unifying issues and research strategies
.
J Child Psychol Psychiatry
2003
;
44
:
1092
1115
[PubMed]
15.
Northam
EA
,
Matthews
LK
,
Anderson
PJ
,
Cameron
FJ
,
Werther
GA
.
Psychiatric morbidity and health outcome in type 1 diabetes--perspectives from a prospective longitudinal study
.
Diabet Med
2005
;
22
:
152
157
[PubMed]
16.
Young
V
,
Eiser
C
,
Johnson
B
, et al
.
Eating problems in adolescents with type 1 diabetes: a systematic review with meta-analysis
.
Diabet Med
2013
;
30
:
189
198
[PubMed]
17.
Johnson
B
,
Eiser
C
,
Young
V
,
Brierley
S
,
Heller
S
.
Prevalence of depression among young people with type 1 diabetes: a systematic review
.
Diabet Med
2013
;
30
:
199
208
[PubMed]
18.
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
[PubMed]
19.
Nylander
C
,
Toivonen
H
,
Nasic
S
,
Söderström
U
,
Tindberg
Y
,
Fernell
E
.
Children and adolescents with type 1 diabetes and high HbA1c -- a neurodevelopmental perspective
.
Acta Paediatr
2013
;
102
:
410
415
[PubMed]
20.
McNally
K
,
Rohan
J
,
Pendley
JS
,
Delamater
A
,
Drotar
D
.
Executive functioning, treatment adherence, and glycemic control in children with type 1 diabetes
.
Diabetes Care
2010
;
33
:
1159
1162
[PubMed]
21.
Schmidt
M
,
Pedersen
L
,
Sørensen
HT
.
The Danish Civil Registration System as a tool in epidemiology
.
Eur J Epidemiol
2014
;
29
:
541
549
[PubMed]
22.
Schmidt
M
,
Schmidt
SAJ
,
Sandegaard
JL
,
Ehrenstein
V
,
Pedersen
L
,
Sørensen
HT
.
The Danish National Patient Registry: a review of content, data quality, and research potential
.
Clin Epidemiol
2015
;
7
:
449
490
[PubMed]
23.
World Health Organization
.
The ICD-10 Classification of Mental and Behavioural Disorders: Diagnostic Criteria of Research
.
Geneva, Switzerland
,
World Health Organization
,
1993
24.
Svensson
J
,
Cerqueira
C
,
Kjærsgaard
P
, et al
.
Danish registry of childhood and adolescent diabetes
.
Clin Epidemiol
2016
;
8
:
679
683
[PubMed]
25.
Ly
TT
,
Maahs
DM
,
Rewers
A
,
Dunger
D
,
Oduwole
A
,
Jones
TW
;
International Society for Pediatric and Adolescent Diabetes
.
ISPAD Clinical Practice Consensus Guidelines 2014. Assessment and management of hypoglycemia in children and adolescents with diabetes
.
Pediatr Diabetes
2014
;
15
(
Suppl. 20
):
180
192
[PubMed]
26.
Wolfsdorf
JI
,
Allgrove
J
,
Craig
ME
, et al.;
International Society for Pediatric and Adolescent Diabetes
.
ISPAD Clinical Practice Consensus Guidelines 2014. Diabetic ketoacidosis and hyperglycemic hyperosmolar state
.
Pediatr Diabetes
2014
;
15
(
Suppl. 20
):
154
179
[PubMed]
27.
Nysom
K
,
Mølgaard
C
,
Hutchings
B
,
Michaelsen
KF
.
Body mass index of 0 to 45-y-old Danes: reference values and comparison with published European reference values
.
Int J Obes Relat Metab Disord
2001
;
25
:
177
184
[PubMed]
28.
Schwandt
A
,
Hermann
JM
,
Rosenbauer
J
, et al.;
DPV Initiative
.
Longitudinal trajectories of metabolic control from childhood to young adulthood in type 1 diabetes from a large German/Austrian registry: a group-based modeling approach
.
Diabetes Care
2017
;
40
:
309
316
[PubMed]
29.
Musselman
DL
,
Betan
E
,
Larsen
H
,
Phillips
LS
.
Relationship of depression to diabetes types 1 and 2: epidemiology, biology, and treatment
.
Biol Psychiatry
2003
;
54
:
317
329
[PubMed]
30.
Gonder-Frederick
LA
,
Zrebiec
JF
,
Bauchowitz
AU
, et al
.
Cognitive function is disrupted by both hypo- and hyperglycemia in school-aged children with type 1 diabetes: a field study
.
Diabetes Care
2009
;
32
:
1001
1006
[PubMed]
31.
Whittemore
R
,
Jaser
S
,
Guo
J
,
Grey
M
.
A conceptual model of childhood adaptation to type 1 diabetes
.
Nurs Outlook
2010
;
58
:
242
251
[PubMed]
32.
Hoey
H
,
Aanstoot
H-J
,
Chiarelli
F
, et al
.
Good metabolic control is associated with better quality of life in 2,101 adolescents with type 1 diabetes
.
Diabetes Care
2001
;
24
:
1923
1928
[PubMed]
33.
Weissberg-Benchell
J
,
Rausch
J
,
Iturralde
E
,
Jedraszko
A
,
Hood
K
.
A randomized clinical trial aimed at preventing poor psychosocial and glycemic outcomes in teens with type 1 diabetes (T1D)
.
Contemp Clin Trials
2016
;
49
:
78
84
[PubMed]
34.
Babler
E
,
Strickland
CJ
.
Helping adolescents with type 1 diabetes “figure it out”
.
J Pediatr Nurs
2016
;
31
:
123
131
[PubMed]
35.
Winkley
K
,
Ismail
K
,
Landau
S
,
Eisler
I
.
Psychological interventions to improve glycaemic control in patients with type 1 diabetes: systematic review and meta-analysis of randomised controlled trials
.
BMJ
2006
;
333
:
65
[PubMed]
36.
McBroom
LA
,
Enriquez
M
.
Review of family-centered interventions to enhance the health outcomes of children with type 1 diabetes
.
Diabetes Educ
2009
;
35
:
428
438
[PubMed]
37.
Northam
EA
,
Todd
S
,
Cameron
FJ
.
Interventions to promote optimal health outcomes in children with type 1 diabetes--are they effective
?
Diabet Med
2006
;
23
:
113
121
[PubMed]
38.
Fitzpatrick
SL
,
Schumann
KP
,
Hill-Briggs
F
.
Problem solving interventions for diabetes self-management and control: a systematic review of the literature
.
Diabetes Res Clin Pract
2013
;
100
:
145
161
[PubMed]
39.
Tsiouli
E
,
Alexopoulos
EC
,
Stefanaki
C
,
Darviri
C
,
Chrousos
GP
.
Effects of diabetes-related family stress on glycemic control in young patients with type 1 diabetes: systematic review
.
Can Fam Physician
2013
;
59
:
143
149
[PubMed]
40.
Gesuita
R
,
Skrami
E
,
Bonfanti
R
, et al
.
The role of socio-economic and clinical factors on HbA1c in children and adolescents with type 1 diabetes: an Italian multicentre survey
.
Pediatr Diabetes
2017
;
18
:
241
248
[PubMed]
Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at http://www.diabetesjournals.org/content/license.

Supplementary data