Individuals with type 1 diabetes (T1D) spend over 8000 hours per year self-managing diabetes outside of the medical setting [1]. Optimal management includes numerous daily behaviors and decisions such as checking blood glucose levels, counting carbohydrates, calculating doses of and administering insulin, and changing dietary habits. Following this regimen is challenging for anyone, but for people with developmental disorders, such as autism spectrum disorder (ASD), this complex care can be even more demanding. Individuals with ASD can have impairments in social communication and interaction as well as repetitive or restrictive behaviors and preferences [2]. Behaviors and activities that individuals with ASD perceive as unimportant or do not understand, such as self-care, health and safety awareness, and activities of daily living, can be very difficult for them to manage independently, which can make the complexity of self-managing diabetes even more challenging. In addition, over 30% of people with ASD also have an intellectual disability (ID), which can add additional challenges in the self-management of T1D [3].

T1D prevalence is increasing, affecting 1.25 million people in the United States in 2014, and 5 million people are expected to be affected by 2050 [4-6]. Up to 1 in 59 children in the United States have been identified with ASD [7], and the reported prevalence of ASD is also increasing [3,7]. There is little research regarding the prevalence of individuals with both ASD and T1D in the United States [3,6-8], although a recent study found the prevalence of ASD in the pediatric T1D population in 1 clinic to be 1.16% (95% CI 0.96-1.26) [9]. An additional study reported that among 10,032 participants from the T1D Exchange Clinic Network, 159 (1.58%) had T1D and ASD [10]. These studies suggest that there is no increased prevalence of ASD in children with diabetes. However, research has documented that people with ID alone may be more likely to develop T1D as a result of having genetic conditions such as Down syndrome [11,12]. In addition, 31% of children with ASD also have an ID [3]. Although ASD alone may not lead to increased diabetic ketoacidosis, severe hypoglycemia, or worsening glycemic control [10], T1D patients with both ASD and comorbid ID are more likely to have very poor glycemic control, higher emergency room utilization, and lower rates of achieving national standards of diabetes management, including lower rates of lipid screening and retinopathy screening [11-13]. With this comes greater risk for negative outcomes, and indeed diabetes-related morbidity and mortality have been identified as a particular concern for people with ID [14,15]. As people with ASD and ID live longer, it has become even more important that they achieve better long-term disease management, including glycemic control and control of other modifiable risk factors, to decrease the occurrence of potentially preventable complications as they age. The interplay between ASD with or without ID and impacts on clinical, behavioral, and psychosocial outcomes related to T1D continue to be areas where little has been explored. Many factors may be involved in such interplay and outcomes, including access to social support and social networks [16], which may be valuable resources in raising a child with both T1D and autism.

Although there have been great advances over the last decades in health care delivered to children with chronic conditions, not all children have benefited equally [17]. Although there is a well-developed evidence base regarding diabetes care and education for populations without ASD or ID, very little research has been done to better understand optimal ways to deliver education and behavioral interventions and to align clinical care with the needs of families raising a child with both T1D and ASD, with or without an ID.

This study sought to better understand self-management challenges and successes experienced by families with experience in raising a child with both T1D and ASD to lay the groundwork for future studies. This will enable the exploration of interventions to improve diabetes outcomes and reduce health disparities in this special population and other unique populations of children living with diabetes and other chronic medical conditions. To the best of our knowledge, this is the first study that utilizes a qualitative component to allow further exploration of the lived experience of raising a child with both T1D and ASD. We also sought to gather some exploratory information on glycemic control related to the degree of adaptive function in light of the coexisting ASD, hypothesizing that adaptive impairment may be associated with hemoglobin A_1c (HbA_1c).

This study involved an innovative approach consisting of 2 phases and included the use of social media as well as traditional interviews to capitalize on the strengths of each. In the first phase, secondary data analysis of web-based content was conducted to understand the challenges and successes families raising a child with T1D and ASD experience, similar to our other studies using web-based content to evaluate barriers and facilitators of living with or raising a child with T1D [18,19]. The themes generated from this study were used to create an interview guide that was used in phase 2 to interview caregivers of children with varying degrees of glycemic control and ASD severity to provide richer exploration of the lived experience in raising a child with both T1D and ASD. This 2-step approach allowed further exploration of initial themes identified from the secondary data analysis of web-based content, with the advantage of being able to ask follow-up questions.

This study was approved by the Penn State College of Medicine institutional review board through expedited review and was considered to be no greater than minimal risk. No procedures or participant interactions took place before receiving this approval.

There were 398 posts between 2009 and 2016 from a publicly available forum for caregivers raising a child with both T1D and ASD (117 unique members) and 25 individual blog posts from 2 blogs authored by a caregiver to a child with T1D and ASD. In all, 4 themes emerged from phase 1 (Table 1).

In phase 2, 12 participants were selected for a telephone interview from among 29 respondents who completed the eligibility survey. Information related to participant demographics and study variables such as HbA_1c and ABAS-3 parent form scores were obtained (Table 2). On average, both interview participants and the broader group of survey respondents were similar in current age, age of child’s diagnosis with T1D, age of child’s diagnosis with ASD, and self-reported HbA_1c. For both interview participants and survey respondents, the majority reported that they were white, not Hispanic or Latino, lived in a suburban setting, and visited their child’s provider every 3-6 months. The majority of both groups reported using continuous glucose monitoring (CGM). The majority of survey respondents reported using an insulin pump, whereas interview participants were evenly split.

In the study sample, 8 individuals had the most recent HbA_1c >7.5%, with an average of 9.7% (SD 1.89), or 82.5 mmol/mol (SD 20.5); 4 individuals had the most recent HbA_1c <7.5%, averaging 6.9% (SD 0.29), or 52.8 mmol/mol (SD 2.9). There was no statistically significant difference in GAC scores between those with HbA_1c >7.5% (mean 68.6, SD 14.22) and those with HbA_1c <7.5% (mean 76.3, SD 11.73; t₁₀=0.92; P=.38).

Overall, 3 themes emerged from an inductive thematic analysis of content obtained from interviews during phase 2 (Table 3).

Caregivers raising a child with comorbid T1D and ASD face unique challenges that are not well understood by families with a child with only 1 of the individual conditions or by health care providers, as evident by the themes that emerged from phase 1. Themes were generally related to the combination of T1D and ASD and focused on coping techniques, advice on needles and technology, health care experience, and sensory issues related to technology (eg, insulin pumps). Themes from phase 2 are complementary. Although some could be interpreted as applying to nearly anyone performing T1D management (with or without coexisting ASD), many of the diabetes-specific challenges were influenced (and generally amplified) by the coexistence of ASD.

To the best of our knowledge, this is the first study with a large qualitative component to better understand the experience of raising a child with both T1D and ASD. The use of both social media and interviews as qualitative data sources in this special population provided a deeper understanding than using either type of data source alone, much like we found in another study using social media and interviews in a sample of adults with T1D [18].

Overall, our study results show that children who have greater impairment in function based on adaptive skills areas do not achieve target HbA_1c values. On the basis of the themes that emerged from our interviews (phase 2), this may be due to the additional barriers that come with the comorbidity, overwhelming amount of information to learn, and challenges in the school setting. More detailed comparisons around our exploratory aim were not conclusive due to the small sample size, not surprisingly.

Caregivers who shared the many challenges they face in raising a child with T1D and ASD might benefit from expanded social networks to provide social support, including emotional, instrumental, informational, and appraisal support. Both in-person support and referral to components of the online caregiver community specific to comorbid T1D and ASD could be valuable resources, as supportive ties may enhance well-being and health of the caregivers, which has been shown to lead to positive outcomes in children with T1D [28-34]. Social networks can provide much needed emotional support, exemplified by 1 participant who, upon finding such a web-based resource, “WEPT tears of joy that there are other parents out there who know what this is like.” Instrumental support might result if caregivers to a child with comorbid T1D and ASD find they are geographically able to provide assistance such as babysitting services, because finding caregivers for any child with T1D presents challenges, and these challenges may be significantly increased in a child with both T1D and ASD [19]. This study shows that caregivers identified numerous coping techniques, including focusing on T1D management first, establishing concrete rules and routines, advice on types of needles and technology, and suggestions for health care appointments. Prioritizing disease-specific tasks when diabetes coexists with other health conditions has been reported elsewhere [35]. Referral to social networks where this kind of informational support can be shared may be highly valuable. Appraisal support in the form of feedback may be especially useful to caregivers with a child newly diagnosed with either T1D or ASD in the setting of having previously been diagnosed with the other condition, as receiving feedback that the challenges they are facing are real and being handled to the best of the caregiver’s ability may help the caregiver in coping with the added new diagnosis.

A modified conceptual model for the relationship of social networks and social support to health can be applied to this unique population (Figure 1, adapted from Glanz et al [16]). Pathway 1 shows that through the expansion of social networks, the supportive ties developed may enhance caregiver health and well-being, which is linked in multiple studies to improved diabetes outcomes in their youth. Pathway 2 represents the hypothesized effect of social networks and social support on individuals’ coping resources. By being able to access new contacts and information to help solve the unique challenges faced by families raising a child with T1D and ASD, there may be a sense or perception of increased control that may also improve the physical and mental health of caregivers. The community of health care providers caring for these children should consider referral to web-based caregiver sites specific for families raising children with both T1D and ASD. By providing resources to these families to strengthen their social networks, the community of health care providers caring for these children may also feel more empowered in caring for this challenging population, as shown in pathway 4. Pathways 2a and 4a demonstrate the buffering effect, whereby the negative effects of the stressful experience of raising a child with T1D and ASD described in this study may be diminished through an expanded social network and having health care providers who understand the potential role of peer support in coping. Pathway 3 illustrates how receiving information through an expanded social network may reduce the amount of stress experienced, which, in turn, can lead to improved mental and physical health in caregivers of children with T1D and ASD. Finally, pathway 5 describes the potential effects of social networks and support on health behaviors. By exchanging information and support, caregivers to a child with T1D and ASD may be more supported in health behaviors that improve the care of their child and themselves as well.

Our study demonstrated that the majority of survey respondents (8/12, 67%) and interview participants (13/17, 77%) had a child who used CGM technology. However, 82% (14/17) of the survey respondents reported their child using an insulin pump, compared with only 50% (6/12) of interview participants. Among those using insulin pumps, the majority use tubeless pumps (12/20, 60% overall, 5/6, 83% of interviewees and 7/14, 50% of survey respondents).

These findings complement the work of Stanek et al [9] and Bethin et al [10], who found that CGM use was the same among children with T1D and ASD as it was among those with T1D but without ASD, whereas insulin pump use was lower for those with both T1D and ASD. This qualitative work adds some explanatory context, as caregivers identified that sensory issues can preclude the successful use of insulin pumps and that successful solutions for some have included the use of tubeless pumps and desensitization techniques, such as the routine use of band-aids before attempting to use diabetes technology.

HbA_1c among study participants was not statistically correlated with the degree of function related to ASD, but it was limited by sample size. However, the trends of 2 comparisons (mean GAC scores of those achieving target HbA_1c vs those not achieving target HbA_1c and correlation of GAC scores with HbA_1c) were both in the direction that one would expect: children who scored higher in skills of daily living had improved HbA_1c values (<7.5%). It is not surprising that this study did not achieve statistical significance, as it was not powered to do so and the sample size was chosen for the greater qualitative focus here. This also echoes comments by Bethin et al [10] and Stanek et al [9] that further work is needed to be able to better support these children and their families. From this study specifically, we propose that further research with larger sample sizes is needed to continue to explore the potential relationships between the degree of functioning related to ASD and glycemic control.

Limitations of this study include the lack of representation of caregivers from diverse ethnic backgrounds and rural areas. In addition, although attempts were made to recruit both from an endocrinology clinic and the diabetes online community (DOC), a minority of participants were recruited through the clinic. An additional limitation is that the search strategy included autism to capture ASDs but did not specifically include Asperger and may, therefore, have limited the scope of search returns. However, this would not fully exclude representation of Asperger’s-related content as it is an ASD, and it would also tend to skew the sample toward more severe levels of impairment, increasing the ability to identify challenges within the sample. This study aligns with previous work involving the DOC that demonstrates a larger web presence of white individuals living in suburban or urban areas, which is not likely representative of the greater population with comorbid T1D and ASD [36]. This may represent a health disparity whereby those in rural areas or from diverse ethnic backgrounds may not be aware of the DOC, may not have access, or may not feel welcome or included in that space.

Raising a child with both ASD and T1D presents significant self-management challenges. However, caregivers of this population identify various coping techniques and strategies and find support in social media sites specific to this unique population. Further research is needed to develop new ways for the health care community to partner with this population.