Abstract
Objective: Understanding how parent–child relationships influence diabetes management in youth with type 1 diabetes is critical for minimizing the risk of short- and long-term complications. We examined how classes of diabetes-specific parenting behaviors are associated with disease management and well-being for youth with type 1 diabetes. Research Design and Methods: The Family Management of Diabetes clinical trial tested the efficacy of a 2-year behavioral intervention for families of youth with type 1 diabetes. Three hundred and ninety youth diagnosed with type 1 diabetes and their primary caregiver were recruited from four pediatric endocrinology centers in the US Classifications of parental involvement utilized baseline parent and youth reports of task involvement, collaborative involvement, and parent–youth conflict. Class differences in baseline glycemic control (HbA1c), regimen adherence, general and diabetes quality of life, and depressive symptoms, and 2-year change in HbA1c were examined. Results: Latent profile analysis identified three classes: (1) high in task and collaborative involvement, low in conflict (Harmonious), (2) low in task involvement, collaborative involvement, and conflict (Indifferent), (3) high in task involvement and conflict, low in collaborative involvement (Inharmonious). The Harmonious group demonstrated the best adherence, glycemic control, and psychosocial well-being. The Inharmonious and Indifferent groups had similar diabetes management, but youth from Inharmonious families showed poorer psychosocial well-being. The intervention effect on glycemic control did not differ across the classes. Conclusions: The interplay of parental involvement and conflict resulted in distinct parenting classes that differed in disease management and well-being. However, the classes benefitted similarly from the behavioral intervention.
Original language | English |
---|---|
Pages (from-to) | 1133-1142 |
Number of pages | 10 |
Journal | Pediatric Diabetes |
Volume | 23 |
Issue number | 7 |
DOIs | |
State | Published - Nov 2022 |
Keywords
- juvenile-onset diabetes
- parent–child relationship
- self-management
- type 1 diabetes
Fingerprint
Dive into the research topics of 'Latent classifications of parental involvement in diabetes management for youth with type 1 diabetes: A randomized clinical trial'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver
}
In: Pediatric Diabetes, Vol. 23, No. 7, 11.2022, p. 1133-1142.
Research output: Contribution to journal › Article › peer-review
TY - JOUR
T1 - Latent classifications of parental involvement in diabetes management for youth with type 1 diabetes
T2 - A randomized clinical trial
AU - Temmen, Chelsie D.
AU - Lu, Ruijin
AU - Gee, Benjamin T.
AU - Chen, Zhen
AU - Nansel, Tonja R.
N1 - Funding Information: This research was supported by the intramural research program of the National Institute of Child Health and Human Development (NICHD), contract #'s N01‐HD‐3‐3360, N01‐HD‐4‐3361, N01‐HD‐4‐3362, N01‐HD‐4‐3363, N01‐HD‐4‐3364. Eunice Kennedy Shriver Funding Information: All study protocols were approved by the Eunice Kennedy Shriver National Institute of Child Health and Human Development Institutional Review Board and the institutional review boards of all participating institutions. This research was supported by the intramural research program of the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD), contract #'s N01-HD-3-3360, N01-HD-4-3361, N01-HD-4-3362, N01-HD-4-3363, N01-HD-4-3364. Families were recruited during routine clinic visits and followed for 2 years. Parents provided informed written consent and youth provided written assent. Baseline assessments were conducted in participant homes (or other convenient location if preferred by participants) by a data coordinating center, who were blinded to study assignment and not affiliated with the clinical sites. Clinic study assessments were conducted by members of the research team. The study protocol was approved by the institutional review board of each clinical site and the Eunice Kennedy Shriver National Institute of Child Health and Human Development. The data for this secondary analysis come from the Family Management of Diabetes clinical trial conducted from 2006 to 2009. This multi-site, parallel-group trial tested the efficacy of a clinic-integrated, family-based behavioral intervention for families of youth with type 1 diabetes. Participants were enrolled from four large, geographically diverse pediatric endocrinology centers in the United States (Boston, MA; Chicago, IL; Jacksonville, FL; Houston, TX). Eligible families were those with a child between 9 and 14.9 years of age who was diagnosed with type 1 diabetes for at least 3 months; glycated hemoglobin (HbA1c) between 6% (42 mmol/mol) and 12% (108 mmol/mol) for those diagnosed <1 year or greater than 6% (42 mmol/mol) for those diagnosed >1 year; daily insulin dosage of 0.5 μg/kg/day for those diagnosed for ≥1 year or 0.2 μg/kg/day for those diagnosed for <1 year, with ≥2 injections or units of insulin pump; and not affected by any other major chronic disease (with the exception of well-controlled asthma, celiac, or thyroid disease) or serious cognitive/psychiatric disorder. Parent and family inclusion criteria included fluency in English, telephone access, attendance to at least two clinic visits in the past year, and no psychiatric diagnoses in participating parents. The sample size was determined through the estimation of detecting significant differences in HbA1c levels between treatment and usual care at a given time point, as described previously.21 Families were recruited during routine clinic visits and followed for 2 years. Parents provided informed written consent and youth provided written assent. Baseline assessments were conducted in participant homes (or other convenient location if preferred by participants) by a data coordinating center, who were blinded to study assignment and not affiliated with the clinical sites. Clinic study assessments were conducted by members of the research team. The study protocol was approved by the institutional review board of each clinical site and the Eunice Kennedy Shriver National Institute of Child Health and Human Development. Participating families were randomized to usual care or intervention following the baseline assessment; randomization procedures have been described previously.21 Randomization was stratified by age (9–<12 and 12–<15 years) and HbA1c [≤8.3% (67 mmol/mol) and >8.3% (67 mmol/mol)]. Research staff provided usual care families with clinical liaison support, including assistance with appointment-scheduling. The intervention group participated in family sessions with a trained research assistant at each clinic visit. The “WE-CAN” manage diabetes intervention was designed to improve adherence and family management practices across the developmental period of pre- and early adolescence. Family sessions were structured around an applied problem-solving approach—a semi-structured process incorporating assessment and specification of target behaviors, identification of barriers and motivators, collaborative setting of goals, facilitation of problem-solving and coping skills, and provision of follow-up and support. Supplementary materials addressed common family issues including communication, conflict, and responsibility-sharing. A detailed description of the intervention has been published previously.21 This study utilized self-report data from youths and one parent (89.6% mothers) and biospecimens from youth. All measures used are from the baseline assessment with the exception of HbA1c and parenting constructs, for which values are examined at both baseline and 2-year follow-up. Youth completed the 16-item Collaborative Parent Involvement Scale. This measure assesses aspects of parent involvement that reflect a collaborative role in diabetes management, such as consulting, supporting, planning, problem-solving, and troubleshooting. Items include assisting with planning diabetes care to fit the youth's schedule, helping the youth learn how to manage difficulties with diabetes, knowing when to give the youth more autonomy, and knowing when the youth requires assistance. Response options range from almost never [1] to always [5]. Higher average scores indicate greater collaborative involvement. The measure has previously shown good internal consistency, expected age-related changes, and differential relationships with adherence.22 Cronbach alpha (α) was 0.93. The revised Diabetes Family Conflict Scale, completed by youth and parents, queries family conflict around 19 aspects of diabetes management such as arguing about remembering to check blood sugars.23 Response options are on a 3-point scale, from never argue [1] to always argue [3]. Higher average scores indicate greater conflict (αchild = 0.92; αparent = 0.89). Both youth (Mean(SD) = 27.32(7.95), Skew = 1.48) and parent (Mean(SD) = 28.24(6.08), Skew = 2.11) scores were logarithmically transformed due to skewness. The Diabetes Family Responsibility Questionnaire, completed by youth and parents, assesses the degree of parent involvement in 17 diabetes management tasks, such as taking insulin, adjusting insulin, deciding what to eat, and remembering to do blood sugar checks.24 For each item, participants indicate whether responsibility for the task belongs to the child [1], is shared equally between child and parent [2], or belongs to the parent [3]. The sum of the items indicates overall parent involvement in diabetes management tasks (αchild = 0.69; αparent = 0.73). The PedsQL Generic Core Scales25 and PedsQL Diabetes Module26 were completed by youth. The measures have well-established validity in healthy and patient populations including youth with diabetes.25,26 Scores on the Generic Core Scales reliably differentiate between healthy children and those with acute or chronic conditions, are related to indicators of morbidity and illness burden, and display a factor-derived solution consistent with a priori conceptually driven scales. Response options are on a five-point scale, from never a problem [0] to a lot of a problem [4]. Higher scores indicate better general quality of life (PedsQL Generic Core Scales; α = 0.90) and diabetes quality of life (PedsQL Diabetes Module; α = 0.87). Youth completed the Children's Depression Inventory (CDI27), a 27-item measure validated in children aged 7–17 years.28 Responses for the 27 depressive symptoms are rated from no symptom [0] to distinct symptom [2], with response options specific to each symptom. CDI summed scores range from 0 to 54; higher scores indicate greater depressive symptoms (α = 0.89). Blood samples were collected for HbA1c assay (Tosoh A1c 2.2 Plus Glycohemoglobin Analyzer, Tosoh Medics, South San Francisco, CA) at a centralized laboratory (Joslin Diabetes Center, Boston, MA). Samples were also processed with the DCA-2000 (Siemens Healthcare Diagnostics, Deerfield, IL) onsite and used to impute replacement values for lost or damaged samples (1.2% of values). Higher scores indicate poorer glycemic control. Diabetes management adherence was assessed with the Diabetes Self-Management Profile, a well-validated structured interview conducted with parents.29,30 The interview assesses overall adherence to a diabetes management regimen in the past 3 months across five domains: diet, exercise, blood glucose testing, insulin administration, and management of low blood sugar. Parallel versions are used for youth on flexible and conventional regimens. Higher scores indicate greater adherence. Participant characteristics including age, gender, family composition, income, education, race/ethnicity were collected from the electronic medical record or reported by parents. All analyses were conducted using R statistical software.31 Latent profile analysis using the FlexMix package32 was conducted with baseline measures of diabetes-specific parenting behaviors to identify potential subgroups. Latent profile analysis is an exploratory approach that identifies subgroups based on participants' observed responses to a set of variables.33 This analytic strategy clusters individuals with similar response patterns across a set of variables into distinct classes. The variables included were baseline youth-reported parent task involvement, parent–child conflict, and collaborative involvement, and parent-reported parent task involvement and parent–child conflict. Five models in total were estimated. The first model estimated a single class, and each model following estimated an additional latent class. The best fitting model was determined by examining the AIC,34 BIC,35 and ICL36 values of each model, with lower values indicating better model fit than the previously estimated model. Additionally, we examined each model's entropy values,37 which indicate how distinct each estimated class is from the other estimated classes in a model, and the significance of the bootstrap likelihood ratio tests (BLRT38), which indicate whether an estimated model is better at explaining the data than the previously estimated model. Criteria for the best fitting model were that the AIC, BIC, and ICL values no longer decreased meaningfully as the number of classes estimated increased, the entropy value was greater than 0.70, and the BLRT p-value was no longer significant at the 0.05 level. The model that best meets these cutoff criteria has adequately distinct classes that present an acceptable representation of the data utilized to create the classes. Additionally, we considered how well each model produced distinct and meaningful classes.39 Complete baseline data were available on youth-reported constructs, while data on parent-reported constructs (parent task involvement and diabetes-related conflict) were missing in six families. We generated multiple imputations for these incomplete constructs by Gibbs sampling, using multivariate imputation by chained equations as implemented in the R “mice” package.40 Next, we examined whether the classes differed in baseline glycemic control (HbA1c), regimen adherence, diabetes quality of life, general quality of life, and depressive symptoms using ANCOVA. Each model included the parenting classes estimated from the latent profile analyses as the independent variable, one outcome, and was adjusted by youth sex and race/ethnicity. To determine whether the intervention effect on glycemic control (HbA1c) differed between the parenting classes, a factorial ANCOVA examined the interaction between intervention group and latent classes on the change in HbA1c from baseline to 2-year follow-up, controlling for youth sex and race/ethnicity. Significant main effects for the ANCOVA and factorial ANCOVA analyses were probed using Tukey's HSD multiple comparisons. Tukey's HSD multiple comparisons is a post-hoc test used to determine if individual pairwise comparisons within a factorial ANCOVA analysis are significantly different from one another.41 A p-value of less than 0.05 indicates that mean for one group is significantly different from the mean of the other group in the pairwise comparison. Additionally, we tested the intervention effect at 2-year follow-upon each of the diabetes-specific parenting constructs (task involvement, collaborative involvement, conflict) using the same approach. Across the ANCOVA models, 22–28 families had missing data on outcomes or covariates. This missingness was handled using a maximum likelihood approach. Publisher Copyright: Published 2022. This article is a U.S. Government work and is in the public domain in the USA.
PY - 2022/11
Y1 - 2022/11
N2 - Objective: Understanding how parent–child relationships influence diabetes management in youth with type 1 diabetes is critical for minimizing the risk of short- and long-term complications. We examined how classes of diabetes-specific parenting behaviors are associated with disease management and well-being for youth with type 1 diabetes. Research Design and Methods: The Family Management of Diabetes clinical trial tested the efficacy of a 2-year behavioral intervention for families of youth with type 1 diabetes. Three hundred and ninety youth diagnosed with type 1 diabetes and their primary caregiver were recruited from four pediatric endocrinology centers in the US Classifications of parental involvement utilized baseline parent and youth reports of task involvement, collaborative involvement, and parent–youth conflict. Class differences in baseline glycemic control (HbA1c), regimen adherence, general and diabetes quality of life, and depressive symptoms, and 2-year change in HbA1c were examined. Results: Latent profile analysis identified three classes: (1) high in task and collaborative involvement, low in conflict (Harmonious), (2) low in task involvement, collaborative involvement, and conflict (Indifferent), (3) high in task involvement and conflict, low in collaborative involvement (Inharmonious). The Harmonious group demonstrated the best adherence, glycemic control, and psychosocial well-being. The Inharmonious and Indifferent groups had similar diabetes management, but youth from Inharmonious families showed poorer psychosocial well-being. The intervention effect on glycemic control did not differ across the classes. Conclusions: The interplay of parental involvement and conflict resulted in distinct parenting classes that differed in disease management and well-being. However, the classes benefitted similarly from the behavioral intervention.
AB - Objective: Understanding how parent–child relationships influence diabetes management in youth with type 1 diabetes is critical for minimizing the risk of short- and long-term complications. We examined how classes of diabetes-specific parenting behaviors are associated with disease management and well-being for youth with type 1 diabetes. Research Design and Methods: The Family Management of Diabetes clinical trial tested the efficacy of a 2-year behavioral intervention for families of youth with type 1 diabetes. Three hundred and ninety youth diagnosed with type 1 diabetes and their primary caregiver were recruited from four pediatric endocrinology centers in the US Classifications of parental involvement utilized baseline parent and youth reports of task involvement, collaborative involvement, and parent–youth conflict. Class differences in baseline glycemic control (HbA1c), regimen adherence, general and diabetes quality of life, and depressive symptoms, and 2-year change in HbA1c were examined. Results: Latent profile analysis identified three classes: (1) high in task and collaborative involvement, low in conflict (Harmonious), (2) low in task involvement, collaborative involvement, and conflict (Indifferent), (3) high in task involvement and conflict, low in collaborative involvement (Inharmonious). The Harmonious group demonstrated the best adherence, glycemic control, and psychosocial well-being. The Inharmonious and Indifferent groups had similar diabetes management, but youth from Inharmonious families showed poorer psychosocial well-being. The intervention effect on glycemic control did not differ across the classes. Conclusions: The interplay of parental involvement and conflict resulted in distinct parenting classes that differed in disease management and well-being. However, the classes benefitted similarly from the behavioral intervention.
KW - juvenile-onset diabetes
KW - parent–child relationship
KW - self-management
KW - type 1 diabetes
UR - http://www.scopus.com/inward/record.url?scp=85136466186&partnerID=8YFLogxK
U2 - 10.1111/pedi.13397
DO - 10.1111/pedi.13397
M3 - Article
C2 - 36250647
AN - SCOPUS:85136466186
SN - 1399-543X
VL - 23
SP - 1133
EP - 1142
JO - Pediatric Diabetes
JF - Pediatric Diabetes
IS - 7
ER -