To assess anxiety and risk perception among parents whose children screened positive for islet autoantibodies, indicating elevated risk for type 1 diabetes (T1D).
The Autoimmunity Screening for Kids (ASK) study identified 319 children age 1 to 17 years at risk for T1D via screening for islet autoantibodies; 280 children with confirmed islet autoantibodies and their caregivers enrolled in a follow-up education and monitoring program to prevent diabetic ketoacidosis at diagnosis. Parents completed questionnaires at each monitoring visit, including a 6-item version of the State Anxiety Inventory (SAI), to assess anxiety about their child developing T1D, and a single question to assess risk perception.
At the first ASK follow-up monitoring visit, mean parental anxiety was elevated above the clinical cutoff of 40 (SAI 46.1 ± 11.2). At the second follow-up monitoring visit (i.e., visit 2), mean anxiety remained elevated but started to trend down. Approximately half (48.9%) of parents reported their child was at increased risk for T1D at the initial follow-up monitoring visit (visit 1). Parents of children with more than one islet autoantibody and a first-degree relative with T1D were more likely to report their child was at increased risk.
Most parents of autoantibody-positive children have high anxiety about their child developing T1D. Information about the risk of developing T1D is difficult to convey, as evidenced by the wide range of risk perception reported in this sample.
Introduction
Screening for presymptomatic type 1 diabetes (T1D) by measuring islet autoantibodies (IAs) has historically been used in clinical research and often in combination with genetic testing where only those identified at increased genetic risk are invited to test for IAs. However, screening for IAs is rapidly increasing in the general population. Children identified through screening and who participate in monitoring studies or programs may experience health benefits, including significantly reduced risk of diabetic ketoacidosis (DKA), a life-threatening complication, at diagnosis. Furthermore, immunomodulatory drugs are becoming available to delay insulin dependence (1) in youth at high risk for developing type 1 diabetes; therefore, it is reasonable to expect an increase in screening now that there is an approved intervention to delay the onset of clinical T1D (i.e., stage 3 T1D).
Data from prospective monitoring studies of at-risk children, such as The Environmental Determinants of Diabetes in the Young (TEDDY), Prospective Assessment in Newborns for Diabetes Autoimmunity (PANDA), and Diabetes Prevention Trial of Type 1 Diabetes (DPT-1), have assessed the psychological experience of parents participating in T1D screening and monitoring programs. In general, parental anxiety is highest when notified about a child’s IA+ test results and tends to decline with time. However, this is not the case for female caregivers of children with multiple positive IAs, who tend to remain highly anxious (2). These studies have also highlighted how challenging it is for participants and families to understand T1D risk. In TEDDY, 40% of mothers from the general population (i.e., those without a family history of T1D) underestimated risk (3). Furthermore, PANDA demonstrated that risk perception accuracy tends to decline over time without intervention or re-education from study staff (4).
Behavioral and psychological characteristics of participants are important predictors of engagement and retention in monitoring studies and programs (5). Given that screening alone will not prevent adverse outcomes, active participation and retention are critical for screening and monitoring programs to achieve their goals. As screening and monitoring expands in the general population, it is essential we understand the participant experience and use this knowledge to support retention and decrease distress associated with the presymptomatic T1D period, especially among individuals from the general population and racial/ethnic minority groups, who have historically been understudied in screening and monitoring programs.
The Autoimmunity Screening for Kids (ASK) study aims to identify IA+ children from the general population and offer an education and monitoring follow-up program to prevent DKA (6). The purpose of this article is to describe the demographic, clinical, and psychological characteristics of those found to be IA+ through the ASK study.
Research Design and Methods
Procedures
In January 2017, ASK began population-based screening of Colorado children age 1 to 17 years for IAs, indicating increased risk for T1D, and transglutaminase autoantibodies, indicating increased risk for celiac disease. Participants are required to speak English or Spanish. Screening is offered at no cost to participants and in a variety of settings, including family practices, hospital settings, and community organizations. Islet autoantibodies to insulin, GAD, islet antigen 2 (IA-2), and zinc transporters (ZnT8) are measured by two methods: radiobinding assay and electrochemiluminescence assay, which is a high-affinity assay using a capillary sample of blood (7–9). Children positive for two or more IAs by either radiobinding assay or electrochemiluminescence assay have the highest risk of progressing to T1D (i.e., 70% risk of progressing to clinical diabetes in 10 years) (10). Those positive for one autoantibody by both methods have a 50% risk of progressing to clinical diabetes in 10 years (11). Children positive for one autoantibody by only one method have an estimated 2% risk of progressing to clinical diabetes in 10 years (unpublished data). All study procedures are approved by the Colorado Multiple Institutional Review Board.
Participants who screen positive for one or more IAs receive a phone call by an ASK research assistant (referred to as a study clinician) to discuss results and recommend families schedule a visit to Barbara Davis Center for Diabetes to confirm IA results. No information about risk level is provided during this phone call. All participants who tested positive for one or more IA are encouraged to come in for confirmation. At this confirmation visit, a venipuncture blood sample is obtained to confirm IA results (using the same assays but with a new blood sample), obtain HbA1c, and test random blood glucose (BG) to ensure the child does not already have clinical T1D. The study clinician provides initial diabetes education, which includes explaining the stages of T1D as well as signs and symptoms of stage 3 T1D (i.e., clinical T1D). The ASK study team includes Spanish-speaking clinicians and educational materials so that education and all visits are provided in families’ preferred language.
At this confirmation visit, a glucometer and BG strips are provided to families of children who screened positive for multiple IAs or a single IA by two methods. The study clinician trains the family on how to use the glucometer and recommends checking BG 2 h after the largest meal two to four times per month and daily when ill. As of late 2020, this recommendation changed to one BG check per week 2 h after the largest meal and daily when ill, given the trend toward increased T1D diagnoses associated with the COVID-19 pandemic. Written information is given to aid home glucose testing, including a chart defining normal, elevated, and high BG levels and instructions on steps to take should the participant have an elevated or high BG. This information and training are provided at the confirmation visit, even though IA results are not yet confirmed, as a preventative measure in case the family elects not to return for follow-up monitoring.
Children who are confirmed positive for one or more IAs are recommended to enroll in the follow-up monitoring program, which includes visits to the Barbara Davis Center for Diabetes every 3 to 12 months, depending on risk of developing T1D. Children with multiple IAs or a single IA by two methods are recommended to follow up every 3 to 6 months and children positive for a single IA by one method are recommended to follow up every 6 to 12 months, depending on clinical and laboratory history.
Consistent and supportive communication is a critical component of the protocol. At the end of each visit, the family is given a visit summary report, which includes the child’s current HbA1c, random BG, signs and symptoms of T1D, study clinician contact information, suggested home glucose testing, and recommended frequency of follow-up visits. Between study visits, the family receives monthly reminders via text, e-mail, and phone to complete glucose testing and contact the study staff should they need additional supplies or support. To build rapport and increase retention, participants work with the same study clinician throughout the course of their participation whenever possible. The assigned clinician partners with the family to track home glucose testing results, identify any trends indicating progression to the next stage of T1D, be a point of contact for questions and concerns, and support family engagement. Families are encouraged to be proactive and contact their study clinician as soon as any signs or symptoms of diabetes arise. The clinician–family interactions also provide reassurance to the family that resources are available to support them to keep their child safe in the event of disease progression.
Participants
As of March 2020, the ASK study had screened 23,847 children, of whom 319 (1.3%) were confirmed positive for one or more IAs and therefore were eligible for the follow-up monitoring program. Most (n = 289 [91%]) enrolled in the follow-up monitoring program. Enrolled children did not differ from those whose parents refused participation by child age at screening, sex, race/ethnicity, family history of T1D, or IA status. Of the 289 child–parent dyads, 280 provided complete data for the outcomes of interest at the first visit of the follow-up and monitoring program (visit 1). As of March 2020, 126 participants had completed a second follow-up visit (visit 2), with a median of 6.1 months between visits (interquartile range 5.3–8.5 months). Of these, 109 (multiple IA+ n = 34; single IA+ by two methods n = 20; single IA+ by one method n = 55) visits were completed by the same parent at both time points and were used to examine changes in anxiety in the first 6 months of study participation (Fig. 1). There was no significant difference in parent psychological characteristics (i.e., risk perception, anxiety) or child clinical or demographic characteristics (i.e., IA status, family history of T1D, child race/ethnicity, parent respondent) between those who completed their second visit compared with those who had not yet completed this visit.
Measures
Demographic Information
Demographic variables collected by parental self-report include child age at confirmation, race/ethnicity of the child, age and education level of the responding primary caregiver, and family history of T1D. The number of participants in each race and ethnicity category is presented in Table 1. For the purpose of the analyses and because of the small numbers in several of the race/ethnicity categories, the following groupings were used: Hispanic of any race, non-Hispanic White, and all other. A positive family history of T1D is defined as the child having a first-degree relative (parent or sibling) with T1D.
. | Total sample . | Multiple IA+ (n = 69 [24.6%]) . | Single IA+ by two methods (n = 65 [23.2%]) . | Single IA+ by one method (n = 146 [52.1%]) . |
---|---|---|---|---|
Caregiver | ||||
Mother | 250 (89.3) | 59 (85.5) | 57 (87.7) | 134 (91.8) |
Father | 24 (8.6) | 6 (8.7) | 7 (10.8) | 11 (7.5) |
Other | 6 (2.1) | 4 (5.8) | 1 (1.5) | 1 (0.7) |
Caregiver age, years | 39.0 (34.0–43.0) | 40 (35–44) | 39 (36–45) | 38 (33–43) |
Caregiver education | ||||
High school diploma or less than high school | 102 (38.4) | 21 (32.3) | 18 (29.5) | 63 (45.0) |
Postsecondary | 69 (25.9) | 18 (27.7) | 15 (24.6) | 36 (25.7) |
College or graduate degree | 95 (35.7) | 26 (40.0) | 28 (46.0) | 41 (29.3) |
Child age, years | 10.3 (6.9–13.5) | 9 (5.6–12.2) | 12 (8.1–14.7) | 10 (7.2–13.3) |
Child race/ethnicity* | ||||
Non-Hispanic White | 113 (40.4) | 39 (56.5) | 34 (52.3) | 40 (27.4) |
Hispanic, any race | 140 (50.0) | 25 (36.2) | 27 (41.5) | 88 (60.3) |
Asian | 6 (2.1) | 1 (1.5) | 0 (0) | 5 (3.4) |
African American | 13 (4.6) | 2 (2.9) | 3 (4.6) | 8 (5.5) |
American Indian/Alaskan Native | 2 (0.7) | 1 (1.5) | 0 (0) | 1 (0.7) |
More than one race | 6 (2.1) | 1 (1.5) | 1 (1.5) | 4 (2.7) |
Child with first-degree relative with T1D† | ||||
No | 256 (91.4) | 59 (85.5) | 56 (86.2) | 141 (96.6) |
Yes | 24 (8.6) | 10 (14.5) | 9 (13.8) | 5 (3.4) |
Mother | 6 (2.1) | 2 (2.9) | 3 (4.6) | 1 (0.7) |
Father | 8 (2.9) | 3 (4.4) | 3 (4.6) | 2 (1.4) |
Sibling | 10 (3.6) | 5 (7.2) | 3 (4.6) | 2 (1.4) |
Mean (SD) anxiety score | 46.1 (11.2) | 47.0 (9.0) | 44.7 (12.5) | 46.3 (11.6) |
Parent perception of child’s T1D risk† | ||||
Not at increased risk | 143 (51.1) | 23 (33.3) | 29 (44.6) | 91 (62.3) |
At increased risk | 137 (48.9) | 46 (66.7) | 36 (55.4) | 55 (37.7) |
. | Total sample . | Multiple IA+ (n = 69 [24.6%]) . | Single IA+ by two methods (n = 65 [23.2%]) . | Single IA+ by one method (n = 146 [52.1%]) . |
---|---|---|---|---|
Caregiver | ||||
Mother | 250 (89.3) | 59 (85.5) | 57 (87.7) | 134 (91.8) |
Father | 24 (8.6) | 6 (8.7) | 7 (10.8) | 11 (7.5) |
Other | 6 (2.1) | 4 (5.8) | 1 (1.5) | 1 (0.7) |
Caregiver age, years | 39.0 (34.0–43.0) | 40 (35–44) | 39 (36–45) | 38 (33–43) |
Caregiver education | ||||
High school diploma or less than high school | 102 (38.4) | 21 (32.3) | 18 (29.5) | 63 (45.0) |
Postsecondary | 69 (25.9) | 18 (27.7) | 15 (24.6) | 36 (25.7) |
College or graduate degree | 95 (35.7) | 26 (40.0) | 28 (46.0) | 41 (29.3) |
Child age, years | 10.3 (6.9–13.5) | 9 (5.6–12.2) | 12 (8.1–14.7) | 10 (7.2–13.3) |
Child race/ethnicity* | ||||
Non-Hispanic White | 113 (40.4) | 39 (56.5) | 34 (52.3) | 40 (27.4) |
Hispanic, any race | 140 (50.0) | 25 (36.2) | 27 (41.5) | 88 (60.3) |
Asian | 6 (2.1) | 1 (1.5) | 0 (0) | 5 (3.4) |
African American | 13 (4.6) | 2 (2.9) | 3 (4.6) | 8 (5.5) |
American Indian/Alaskan Native | 2 (0.7) | 1 (1.5) | 0 (0) | 1 (0.7) |
More than one race | 6 (2.1) | 1 (1.5) | 1 (1.5) | 4 (2.7) |
Child with first-degree relative with T1D† | ||||
No | 256 (91.4) | 59 (85.5) | 56 (86.2) | 141 (96.6) |
Yes | 24 (8.6) | 10 (14.5) | 9 (13.8) | 5 (3.4) |
Mother | 6 (2.1) | 2 (2.9) | 3 (4.6) | 1 (0.7) |
Father | 8 (2.9) | 3 (4.4) | 3 (4.6) | 2 (1.4) |
Sibling | 10 (3.6) | 5 (7.2) | 3 (4.6) | 2 (1.4) |
Mean (SD) anxiety score | 46.1 (11.2) | 47.0 (9.0) | 44.7 (12.5) | 46.3 (11.6) |
Parent perception of child’s T1D risk† | ||||
Not at increased risk | 143 (51.1) | 23 (33.3) | 29 (44.6) | 91 (62.3) |
At increased risk | 137 (48.9) | 46 (66.7) | 36 (55.4) | 55 (37.7) |
Data are presented as n (%) or median (interquartile range) unless otherwise noted.
Indicates significant difference among groups at P < 0.01.
Indicates significant difference among groups at P < 0.001.
Metabolic Monitoring
At each visit, a blood sample is obtained to measure HbA1c, random BG, and IAs. Height and weight are also measured. Participants positive for multiple IAs or a single IA by two methods are offered a continuous glucose monitor every 3 to 6 months and oral glucose tolerance tests every 6 months to identify dysglycemia.
Parental Anxiety
Parental anxiety about the child’s risk of T1D was assessed with a 6-item measure adapted from the 20-item state component of the State Anxiety Inventory (SAI) (12–14). Respondents are asked to think about their child’s risk for T1D when responding to the items, which measure anxiety at a single time (i.e., state anxiety). The 6-item score is then converted into a total score comparable to the 20-item SAI score. Studies that use the SAI for screening purposes have identified a score of >40 as denoting high anxiety (2,15,16).
Parent Risk Perception
A parent’s risk perception is assessed based on a single question: “Compared with other children, do you think your child’s risk for developing diabetes is a) much lower, b) somewhat lower, c) about the same, d) somewhat higher, or e) much higher?” This question has been used to assess risk perception and risk perception accuracy in screening and monitoring studies, such as TEDDY (3,17). For the purposes of this manuscript, “much lower,” “somewhat lower,” and “about the same” were considered to mean the parent does not perceive the child to be at increased risk for T1D. Reponses of “somewhat higher” and “much higher” were considered to mean the parent perceives the child to be at increased risk for T1D.
Statistical Analyses
All statistical analyses were performed using SAS version 9.4 (SAS Institute, Inc., Cary, NC). Continuous variables were analyzed using the Student t test or one-way ANOVA for differences in means or the Wilcoxon rank sum test for differences in medians. Categorical variables were analyzed using χ2 tests or Fisher exact tests. Multiple linear regression was used to examine the association between parental anxiety at visit 1 and covariates, such as IA status, family history of T1D, parent education, parent age, child age, race/ethnicity, and parent risk perception. Multiple logistic regression was used to examine the factors associated with parent risk perception at visit 1 and covariates, such as IA status, family history of T1D, parent education, parent age, child age, and race/ethnicity. Multiple linear regression was used to examine the association between change in anxiety (from visit 1 to visit 2) and covariates, such as anxiety at visit 1, parent age, child age, parent education, parent risk perception, and days between visit 1 and visit 2. For all tests, P values < 0.05 were considered significant with a two-tailed test.
Results
Parental Anxiety and Risk Perception at Visit 1
The characteristics of the overall study sample and by IA risk level are shown in Table 1. Approximately half (52.1%) of the youth were positive for a single autoantibody by one method. The remaining were roughly equally split between single autoantibody positivity by two methods (23.2%) and multiple autoantibody positivity (24.6%). There were differences in demographic characteristics (i.e., child race/ethnicity and child with first-degree relative with T1D) by IA status. There was also an overall group difference for parent risk perception by IA status. All regression models included these covariates to control for group differences.
At the first visit of the monitoring program, mean parental anxiety was 46.1 ± 11.2, with 74.4% reporting an SAI score >40, indicating most parents experienced high anxiety about their child’s T1D risk. Multiple linear regression was used to examine factors associated with parental anxiety about their child developing T1D. There was a strong dose response for parent education such that anxiety increased for each unit of decrease in parent educational attainment; parents with a high school diploma or less were the most anxious, with a mean adjusted anxiety score of 52.1 (SE 1.8), compared with a mean anxiety score of 44.0 (SE 1.9) for parents with a college degree or higher (Table 2). Parents identifying as Hispanic were more anxious than parents identifying as non-Hispanic White (β = 4.2; P = 0.03); there was a similar trend for parents identifying as any other non-Hispanic White race (β = 4.9; P = 0.08), but this difference only approached significance. Finally, parent perception that their child is at increased risk for T1D was positively associated with anxiety (β = 3.7; P = 0.02)
. | β Estimate . | 95% CI . | P . |
---|---|---|---|
Intercept | 37.0 | 27.6–46.5 | <0.0001 |
Diabetes risk by autoantibodies (ref = single IA+) | |||
Single IA+ by two methods | 0.0 | −3.7 to 3.7 | 0.83 |
Multiple IAs+ | 1.4 | −2.3 to 5.1 | 0.45 |
Age of child (per year) | 0.0 | −0.4 to 0.4 | 0.92 |
Age of parent (per year) | 0.0 | −0.2 to 0.2 | 0.98 |
Parent education (ref = college or graduate degree) | |||
Postsecondary | 4.1 | 0.0–8.1 | 0.05 |
High school diploma or less | 8.1 | 3.9–12.3 | <0.001 |
Child race/ethnicity (ref = non-Hispanic White) | |||
Hispanic, any race | 4.2 | 0.4–8.0 | 0.03 |
All others | 4.9 | −0.6 to 10.3 | 0.08 |
Child has first-degree relative with T1D (ref = no) | |||
Yes | 3.3 | −1.7 to 8.4 | 0.20 |
Parent’s perception of child’s T1D risk (ref = not at increased risk) | |||
At increased risk for T1D | 3.7 | 0.6–6.9 | 0.02 |
. | β Estimate . | 95% CI . | P . |
---|---|---|---|
Intercept | 37.0 | 27.6–46.5 | <0.0001 |
Diabetes risk by autoantibodies (ref = single IA+) | |||
Single IA+ by two methods | 0.0 | −3.7 to 3.7 | 0.83 |
Multiple IAs+ | 1.4 | −2.3 to 5.1 | 0.45 |
Age of child (per year) | 0.0 | −0.4 to 0.4 | 0.92 |
Age of parent (per year) | 0.0 | −0.2 to 0.2 | 0.98 |
Parent education (ref = college or graduate degree) | |||
Postsecondary | 4.1 | 0.0–8.1 | 0.05 |
High school diploma or less | 8.1 | 3.9–12.3 | <0.001 |
Child race/ethnicity (ref = non-Hispanic White) | |||
Hispanic, any race | 4.2 | 0.4–8.0 | 0.03 |
All others | 4.9 | −0.6 to 10.3 | 0.08 |
Child has first-degree relative with T1D (ref = no) | |||
Yes | 3.3 | −1.7 to 8.4 | 0.20 |
Parent’s perception of child’s T1D risk (ref = not at increased risk) | |||
At increased risk for T1D | 3.7 | 0.6–6.9 | 0.02 |
Ref, reference.
Slightly fewer than half (48.9%) of participants perceived their child as at increased risk for diabetes at the first monitoring visit (visit 1). Multiple logistic regression was used to examine factors associated with parent perception that their child is at increased risk for T1D (Table 3). The child having a first degree relative with T1D was the strongest predictor of a parent reporting the child is at increased risk (odds ratio [OR] 4.8; P = 0.01). Parents of children with multiple IAs were more likely to report their child had an increased risk of developing T1D compared with parents of children positive for a single IA by one method (OR 2.85; P < 0.01). Educational level of the parent was also associated with the parent reporting their child as being at increased risk for T1D. Parents with a high school diploma or parents who did not complete high school were significantly less likely to report their child as being at increased risk for T1D compared with parents with a college or graduate degree (OR 0.3; P < 0.01). Finally, parents of children who identified as Hispanic were less likely to report their child as being at increased risk for T1D compared with parents of non-Hispanic White children (OR 0.4; P = 0.03).
Effect . | OR . | 95% Wald confidence limits . | Pr > χ2 . |
---|---|---|---|
Diabetes risk by autoantibodies (ref = single IA+) | |||
Single IA+ by two methods | 1.3 | 0.6–2.6 | 0.53 |
Multiple IA+ | 2.8 | 1.3–6.0 | <0.01 |
Child has first-degree relative with T1D (ref = no) | |||
Yes | 4.8 | 1.4–16.7 | 0.01 |
Age of child (per year) | 1.1 | 1.0–1.1 | 0.23 |
Age of parent (per year) | 1.0 | 0.9–1.0 | 0.15 |
Parent education (ref = college or graduate degree) | |||
Postsecondary | 0.8 | 0.3–1.7 | 0.49 |
High school diploma or less | 0.3 | 0.1–0.7 | <0.01 |
Child race/ethnicity (ref = non-Hispanic White) | |||
Hispanic | 0.4 | 0.2–0.9 | 0.03 |
All other | 1.1 | 0.4–3.0 | 0.90 |
Effect . | OR . | 95% Wald confidence limits . | Pr > χ2 . |
---|---|---|---|
Diabetes risk by autoantibodies (ref = single IA+) | |||
Single IA+ by two methods | 1.3 | 0.6–2.6 | 0.53 |
Multiple IA+ | 2.8 | 1.3–6.0 | <0.01 |
Child has first-degree relative with T1D (ref = no) | |||
Yes | 4.8 | 1.4–16.7 | 0.01 |
Age of child (per year) | 1.1 | 1.0–1.1 | 0.23 |
Age of parent (per year) | 1.0 | 0.9–1.0 | 0.15 |
Parent education (ref = college or graduate degree) | |||
Postsecondary | 0.8 | 0.3–1.7 | 0.49 |
High school diploma or less | 0.3 | 0.1–0.7 | <0.01 |
Child race/ethnicity (ref = non-Hispanic White) | |||
Hispanic | 0.4 | 0.2–0.9 | 0.03 |
All other | 1.1 | 0.4–3.0 | 0.90 |
OR, odds ratio; ref, reference.
Parental Anxiety Change Over Time
To explore short-term changes in parental anxiety, we examined change in anxiety in 109 parents who completed the first two monitoring visits (visit 1 and visit 2), which occurred an average of 6 months apart. Mean anxiety decreased slightly from 45.0 ± 11.8 to 43.3 ± 11.7 (P = 0.03). Although this change was statistically significant, mean anxiety remained >40, indicating the sample was still highly anxious about their child’s risk for developing T1D. A similar percentage of parents scored >40 at visit 1 (74%) compared with visit 2 (69%). In multiple regression, the only predictor of change in anxiety was the initial anxiety score such that those with high anxiety experienced less of a decline in anxiety between the first and second monitoring visits.
Conclusions
New therapeutic interventions that may delay stage 3 T1D have generated increased attention to and interest in screening for IAs to identify individuals at risk for developing clinical T1D. Much of what is currently known about the psychological experience of participating in screening and monitoring for presymptomatic T1D is from studies that included mostly individuals with first-degree relatives with T1D (e.g., BABYDIAB) or large cohort studies that followed children from birth and included genetic risk before screening for IAs (e.g., TEDDY). As screening for IAs in the general population expands, it is important to better understand the psychological experience of these programs in the general population and in more diverse samples, both of which are lacking in the existing literature. Furthermore, there may be important differences between individuals identified as high risk by genetic screening before IA testing compared with those identified through IA testing alone. The ASK study includes a sample with increased diversity, predominantly from families with no prior knowledge of T1D or IA, thereby providing unique data to help fill this important gap in the literature.
Differences in study protocols (i.e., when a family is informed about the child’s risk as well as when risk and anxiety are assessed) make it difficult to directly compare our data with the TEDDY data. However, based on the available data, the ASK sample seems to be more anxious than parents in TEDDY. In ASK, approximately three-quarters of parents (74%) had SAI scores >40 at visit 1, which is the first visit after the child is confirmed to be positive for IAs. A similar percentage had elevated scores (69%) at the second visit, which occurred a mean of 6 months later. In contrast, at the first TEDDY visit, 47% of mothers and 34% of fathers had elevated anxiety (2). These data were collected after notification of the child’s increased genetic risk, not after IA+ test results. However, all data are collected shortly after initial notification to a family that their child is at increased risk for T1D. TEDDY then examined anxiety immediately after learning one’s child became IA+. The mean anxiety score after IA+ test results was ∼37 in mothers and ∼35 in fathers (2). In comparison, the mean anxiety score in ASK was 46. Taken together, these results suggest that anxiety in the ASK cohort is higher than anxiety in the TEDDY cohort.
There are several plausible explanations for this difference in anxiety. First, approximately half of the ASK participants identify as Hispanic, in contrast to those in TEDDY, where a majority are non-Hispanic White. The ASK sample is also more diverse in terms of educational attainment. Both race/ethnicity and educational attainment were shown to be correlated with anxiety about a child’s T1D risk in the TEDDY study (17), and this finding was replicated in the current study such that lower educational attainment and identification as a racial/ethnic minority are associated with higher anxiety about child’s risk for T1D. Finally, TEDDY participants are determined to be eligible based on a genetic risk score and do not complete IA testing until increased genetic risk is confirmed. Therefore, news that a child tested IA+ may be less of a surprise to parents in TEDDY compared with those in ASK, where the initial screening is for IAs and there is no genetic risk screening. Simply the fact that parents are enrolled in TEDDY and familiar with the monitoring provided by TEDDY may provide reassurance. In contrast, parents in ASK may be less familiar with T1D and do not have the benefit of already having an established relationship with a study clinician when they learn about their child’s IA+ test results.
TEDDY also demonstrated that anxiety about a child’s T1D risk tends to decline with time, except in mothers of children with multiple IA positivity (2). Initial analyses presented in this article highlight minimal change in anxiety from the first assessment at visit 1 to the second assessment at visit 2 for ASK participants. These data suggest that anxiety needs to be monitored, because it may not decline with time in this population, and additional psychological support will be necessary to appropriately support individuals from the general population identified as IA+. These results also highlight the critical importance of closely monitoring anxiety in individuals with less education and from racial and ethnic minority backgrounds.
Slightly fewer than half of the parents reported their child was at somewhat or much higher risk for T1D compared with other children. Consistent with previous literature, we found that parents of children with a first-degree relative with T1D were more likely to perceive their child as being at increased risk for T1D and that lower education and Hispanic ethnicity were both associated with not perceiving the child to be at increased risk. We did not find parents who identified as any other non-White race or ethnicity to be less likely to perceive their child is at risk for T1D; however, given the small number in this category, it will be an important question to monitor when our sample size is larger.
Risk perception is an important predictor of study retention and engagement in TEDDY (5). Therefore, it is plausible that risk perception may be related to participant adherence to recommended self-monitoring outside of study visits; however, we did not assess this in the current study. Future general population studies should consider tailored interventions to increase engagement in populations that tend to underestimate their child's risk for T1D. This is of particular importance because ethnic minorities have higher rates of DKA at diagnosis (18), and DKA at onset has been shown to lead to long-term differences in HbA1c (19).
These data and analyses must be interpreted with an understanding of their limitations. The present analyses do not include families of children with IA+ who declined to enroll in the ASK follow-up monitoring program or those for whom the response to their invitation was pending. These groups did not significantly differ with regard to demographic characteristics compared with the participants included in these analyses. However, it is possible that these groups may differ based on characteristics such as anxiety about T1D risk, which is not collected before the participants agree to be in the follow-up monitoring program. Risk perception is also not collected before enrollment in the follow-up monitoring program. Higher anxiety and underestimation of risk were important predictors of early withdrawal in TEDDY (5). Therefore, it is possible that the current sample may underestimate the level of anxiety and/or overestimate the percentage of participants reporting their child is at increased risk for T1D. There are also limitations with our sample size. Specifically, there were only 27 parents in the “all others” race/ethnicity category, and this variable approached significance in the multiple linear regression model predicting parental anxiety at baseline. A larger sample of participants from other racial and ethnic minority groups would be beneficial in future studies to increase power and better understand this trend.
In conclusion, we found that anxiety about child T1D risk was overall higher in the ASK study compared with previous cohorts, such as TEDDY. Consistent with previous cohorts, education level and racial/ethnic minority status seem to be important demographic factors associated with anxiety about T1D risk and risk perception (20–24). In addition, having a first-degree relative with T1D is associated with increased likelihood of perceiving one’s child to be at increased risk (21,23,24), and this perception of one’s child as being at increased is associated with higher anxiety (2). As screening for IAs increases and moves from research to clinical care, it is essential that anxiety and risk perception are assessed, monitored, and intervened upon as needed to ease the burden on participants and increase engagement in follow-up. Particular attention and adaptations may be needed for groups that have historically not been well represented in IA screening and follow-up monitoring studies.
See accompanying article, p. 2123.
This article contains supplementary material online at https://doi.org/10.2337/figshare.23989248.
H.K.O. and C.G.R. contributed equally to this work.
A full list of ASK Study Group members can be found in the supplementary material online.
Article Information
Funding. The ASK study (SRA-2018-564-M-N and SRA-2021-1065-M-N) is funded by JDRF International, the Leona M. and Harry B. Helmsley Charitable Trust, and Janssen Research and Development, LLC.
Duality of Interest. K.M.S. has received funding from Provention Bio and has received consulting fees from Provention Bio and Sanofi; however, this work does not pose a conflict to the work presented in this article. B.I.F. has served as a consultant for Provention Bio; however, this consultancy does not pose a conflict to the work presented in this article. M.J.R. has served as a consultant for Sanofi, Janssen, and Provention Bio; however, these consultancies do not pose a conflict to the work presented in this article. M.J.R. has received payment from Medscape and France Foundation for an online educational event and lecture, respectively; however, these events do not pose a conflict to the work presented in this article. No other potential conflicts of interest relevant to this article were reported.
Author Contributions. H.K.O. interpreted data. H.K.O. and C.G.R. wrote the manuscript. F.D. analyzed data and reviewed and edited the manuscript. K.M.S., A.K.S., B.I.F., K.B., M.J.R., and J.B. reviewed and edited the manuscript. H.K.O. 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.