Psychosocial, Medical and Demographic Variables Associated with Parent Mealtime Behavior in Young Children Recently Diagnosed with Type 1 Diabetes

Abstract

Objective:

Managing young children’s mealtime concerns can be challenging following type 1 diabetes (T1D) diagnosis, due to developmental factors and diabetes management demands. To identify potential intervention targets, we evaluated medical, psychosocial, and demographic factors in relation to parents’ engagement in problem mealtime behaviors (e.g., pressure to eat, restriction).

Method:

Parents (N=157) of young children (age 1–6 years) reported on psychosocial variables (parent fear of hypoglycemia, family functioning, parent problem-solving, and parents’ problem mealtime behavior frequency and perceptions of being problematic) within two months post T1D diagnosis. Hierarchical regression analyses examined associations among psychosocial variables, demographics (child sex, parent race/ethnicity), child continuous glucose monitor (CGM) use, and parents’ problem mealtime behaviors

Results:

Parents of children using CGM reported parents’ mealtime behaviors as more problematic than non-users, but there were no differences for other medical or demographic variables. Models predicting parents’ problem mealtime behavior frequency and problem perceptions that included psychosocial variables, demographic variables, and CGM use led to significant R² of .14 and .16, respectively. CGM use and parent problem solving were significantly associated with parent mealtime behaviors being perceived as more problematic.

Conclusion:

Shortly following T1D diagnosis in young children, medical and parent psychosocial factors related to how frequently parents engaged in problem mealtime behaviors and the degree to which parents perceived them as problematic. Other factors may further explain the complexities of mealtime management. Considering parents’ problem solving skills and child treatment regimens may help guide interventions targeting mealtime challenges during the new diagnosis period.

INTRODUCTION

Mealtimes are a particularly challenging aspect of parenting young children (under 6 years). Young children engage in a variety of behaviors that can disrupt mealtimes, such as picky eating, extended mealtimes, and leaving the table before the meal has finished.^1,2 To manage children’s behaviors during meals, parents may engage in nonresponsive parenting behaviors, including pressuring children to eat, food restriction, preparing alternate meals, lack of reciprocity with the child, and inattention to children’s satiety cues^3–5. In contrast, responsive mealtime behaviors include having predictable mealtime routines, setting clear expectations for behavior at the table, and responding consistently to child hunger cues.

Chronic conditions such as type 1 diabetes (T1D) can introduce specific challenges to mealtimes with young children. One central component of T1D management is accurate insulin dosing, which requires calculation of carbohydrates the child is anticipated to consume at every meal and snack.⁶ Ideally, parents administer insulin before children eat to maximize insulin action in the body;⁶ however, developmental factors of early childhood, including variable eating patterns and food preferences, make these calculations difficult to predict.⁷ When children do not eat the types or amount of food expected, parents must manage these child behaviors and try to match their carbohydrate consumption with the insulin dose already administered to reach target blood glucose level.⁷ Parent adherence to mealtime schedules has been suggested to improve glycemic levels in newly diagnosed young children.⁷

Psychosocial factors could contribute to how parents respond to developmentally expected child behaviors at mealtimes, especially while managing a new T1D diagnosis. Parents of young children with T1D report less confidence at mealtimes, perceive mealtimes as more difficult, give more direct and indirect commands to eat, and experience more stress during mealtimes than parents of children without T1D.^8–10 Despite these unique challenges, there is little research on factors related to post diagnosis parent mealtime behavior. Research with families of youth with established T1D may not generalize to early experiences in the immediate post-diagnosis period when families are adjusting to new routines.

Given the strain on parents and families following T1D diagnosis, understanding factors that relate to parents’ mealtime behavior can help guide intervention and education efforts.⁹ Kazak’s Pediatric Psychosocial Preventative Health Model (PPPHM¹¹) is a social ecological model that recognizes factors at multiple levels that relate to management of a health condition. PPPHM posits that if a family microsystem is vulnerable or has pre-existing difficulties at diagnosis, these difficulties may affect long term health outcomes and may require higher levels of professional support.¹¹ Thus, understanding psychosocial, medical, and demographic factors in the family system soon after diagnosis may help identify guide future care needs.

At the family psychosocial level, in T1D, many parents experience fear of hypoglycemia (i.e., disruptive worry about the child having a low blood glucose level), which may lead them to use a variety of strategies to persuade the child to eat.¹² Given the challenges of planning meals in coordination with insulin dosing and child behavior management, parents of young children with T1D need strong problem solving skills to account for carbohydrates for unplanned snacks and meals, unfinished food, and blood glucose fluctuations.¹³ Additionally, family cohesion and positive family functioning are associated with higher engagement in T1D management tasks overall¹⁴ and with mealtimes specifically in young children.⁹

At the medical level, devices, such as continuous glucose monitors (CGM), may influence parents’ mealtime behaviors. CGMs are diabetes management devices worn on the body that provide measurements of glucose concentrations every 5 minutes.¹⁵ Early adoption of CGM soon after diagnosis can help parents understand how children’s blood glucose values respond to different eating and physical activity patterns.¹⁵ The data can be helpful in monitoring glucose trends to understand patterns, especially in the newly diagnosed period.¹⁶ However, the continuous data CGMs provide can be overwhelming, and frequent diabetes-related disruptions to parents’ attention can result in increased parent stress.¹⁶ It is possible that some parents may pressure children to eat in response to glucose data they observe on the CGM.^15,16

Family sociodemographic factors including sex, race, and ethnicity may also be associated with parent mealtime behaviors, though there are no data on these patterns in youth with T1D. For example, mothers of children without T1D praise girls twice as much as boys for eating during mealtime.¹⁷ In the general US population, limited research indicates differences in mealtime functioning and communication, with parents of racially and ethnically minoritized backgrounds (e.g., Hispanic and Asian American parents) exhibiting more behaviors that are perceived as directive from a Western lens (e.g., encouragement to eat all food presented) than White Non-Hispanic parents.¹⁸

Aligned with PPPHM, we aimed to examine family demographic, medical, and psychosocial factors related to parents’ mealtime behaviors to better understand early functioning across multiple domains. The results of this study will advance the field by providing information about how parents establish patterns of mealtime behaviors soon after T1D diagnosis, with a focus on a particularly challenging and relatively understudied population of very young children. These early patterns have potential to influence longer-term patterns in mealtime behaviors and diabetes management. We evaluated the degree to which parents’ psychosocial factors (i.e., fear of hypoglycemia, problem solving, family functioning), family demographic variables (i.e., child sex, parent race/ethnicity), and children’s medical characteristics (i.e., CGM use) related to engagement in problem mealtime behaviors and perceptions of these behaviors being problematic among parents of young children with newly diagnosed T1D. We expected that parents with lower fear of hypoglycemia, higher problem solving skills, and higher family functioning would report engaging less frequently in parent problem mealtime behaviors and would perceive their behaviors to be less problematic. We also explored whether parent race/ethnicity, parent and child sex, and child CGM use were related to parents’ problem mealtime behaviors.

METHODS

Participants and Procedures

This was a secondary data analysis of baseline data from a randomized clinical trial conducted at two academic medical centers in the Mid-Atlantic and Southwestern United States (NCT02527525). The trial tested a stepped care behavioral intervention for parents of children age 1–6 years within two months after T1D diagnosis. Parents were recruited while the child was hospitalized for a new diagnosis of T1D or shortly following discharge from the hospital. Parents’ eligibility criteria included: age ≥21 years, fluent in English, and self-identified as the primary caregiver. Parents were not eligible if their child had any other life-threatening illnesses or developmental disability in addition to T1D, per parent report. Baseline data collection took place within eight weeks of diagnosis (M=29±15 days). Following recruitment and consent, parents completed baseline surveys distributed through a link to REDCap, a secure, HIPAA-compliant research platform.¹⁹

Between 2016–2019, the study teams approached 364 primary caregivers for recruitment, of which 298 were able to be reached for screening and 217 were confirmed to be eligible and consented to participate. Of those, 158 primary caregivers were randomized and 157 completed baseline data. Detailed information about trial recruitment and enrollment have been published.^20,21 Of participants who were eligible but did not participate, the main reasons included time commitment, changed location of diabetes care, and not completing baseline surveys within the eligibility window^20,21. Participant data are presented in Table 1.

Table 1.

Sample Demographic and Medical Characteristics

Characteristics	Total Sample N	CGM users % (N) or M±(SD)
		Full sample (n=157)	CGM users (n=38)
Child Age, years		4.5 ±1.6	4.2± years (1.5)
Days since Diagnosis		29±15	31±.2 days (16)
HbA1c at baseline (%)		8.4±1.4	8.0±1.2
Child sex
Female		55% (86)	58% (22)
Male		45% (71)	42% (16)
Parent age, years		34.8±7.0	35.7± years (5.5)
Parent sex
Female		91% (143)	89% (34)
Male		9% (14)	11% (4)
Parent Employment
Full time		57% (89)	55% (21)
Part time		8% (12)	5% (2)
Not employed outside the home		29% (46)	32% (12)
Student		3% (5)	3% (1)
Declined to answer		3% (5)	0
Parent Race/Ethnicity
Non-Hispanic White		62% (97)	76% (29)
Non-Hispanic Black		14% (23)	5% (2)
Hispanic/Latino/a/x		12% (19)	11% (4)
Asian/Asian American		12% (12)	8% (3)
Multiracial		2.5% (4)	0-
American Indian/Alaskan Native		<1% (1)	0-
Parent Level of Education
High school diploma to 2 year college		46% (73)	29% (11)
4 year college and above		54% (84)	71% (27)
Child’s health insurance, private		72% (112)	95% (36)
DKA at diagnosis		37% (57)	100% (38)

Measures

Parents’ Problem Mealtime Behaviors.

Participants completed the Behavioral Pediatric Feeding Assessment Scale (BPFAS), which assesses problem mealtime behaviors in children with medical conditions and their parents.⁵ The BPFAS includes 35 total items, 10 of which specifically address parents’ engagement in problem mealtime behaviors, like pressuring to eat and using threats, and their feelings about managing mealtime behavior, (e.g., “I get frustrated and/or anxious when feeding my child”). First, respondents rate how often they engage in each behavior on a 5-point Likert scale (which results in the Frequency score), then they answer the question “Is this a problem for you?” with a yes/no option (which results in the Problem score).⁵ Only the 10 Parent items were used in this analysis, resulting in two subscales: Parent Frequency Score (how often parents felt they exhibited each of the 10 parent mealtime behaviors), and Parent Problem Score (how problematic parents perceived their engagement in the 10 parent mealtime behaviors to be).⁵ Higher scores represent more frequent and more problematic parent mealtime behavior. Clinical cut-offs are: Frequency=20, Problem=2. Internal consistency in this sample was: Frequency α=.82; Problem α=.80.

Fear of Hypoglycemia.

Parents completed the Hypoglycemia Fear Survey (HFS)¹², modified by this study team to be appropriate for parents of young children. The HFS is scored on a 5-point Likert scale with 1 for “never” and 5 for “always” with higher scores representing greater fear. We included only the Worry subscale, which reflects parent concerns about their child experiencing hypoglycemia. The internal consistency in this sample was α=.92.

Problem solving.

Participants completed a 25-item version of the Social Problem Solving Inventory Revised-Short Form (SPSI-R:S)²². The items address current problem solving skills and are scored on a 5-point Likert scale with 0 for “not at all true of me” to 4 for “extremely true of me”, with higher scores indicating better overall problem solving ability.²² The internal consistency was high in this sample (α=.90).

Family Functioning.

Participants completed the 5-item Family Functioning/Resiliency subscale of the Protective Factors Survey (PFS)²³, which measures the current well-being of the family, including the degree to which they openly share positive and negative experiences with one another and persevere in times of crisis. Respondents use a 7-point Likert scale with 1 for “never” and 7 for “always,” with higher scores meaning higher functioning.²³ The internal consistency was high in this sample (α=.89).

Demographic and medical variables.

Parents completed a questionnaire about family demographics and child medical variables. Child’s CGM use at baseline was corroborated by medical records, with rates and patterns of CGM use in this sample reported previously.²⁴

Analysis

To test for differences in BPFAS scores across demographic and medical variables, we conducted t-tests for dichotomous variables (e.g., CGM users and nonusers) and ANOVA for variables with more categories (e.g., parent race/ethnicity). Given some race/ethnicity groups with very small cell sizes, we combined the race/ethnicity categories reported in Table 2 to four groups (Non-Hispanic White, 62%; Non-Hispanic Black, 14%; Hispanic/Latino/a/x, 12%, and parents of another race/ethnicity (Asian/Asian-American, Multiracial and American Indian/Alaskan Native), 12%), which allowed us to include all cases for analysis. Correlations determined strength and direction of associations between the psychosocial variables (HFS, SPSI-R:S, PFS) and parent mealtime behavior (BPFAS Scales). Two separate hierarchical regressions were then used to examine factors related to BPFAS scores (dependent variables: Parent Frequency Scale, Parent Problem Scale). In each hierarchical regression, demographic variables (child sex, parent sex, and parent race/ethnicity) were entered first to control for any group differences in parent mealtime behavior, the medical variable (CGM use) was entered second to understand how diabetes treatment may relate to parent mealtime behavior beyond demographic factors, and the psychosocial variables were entered in the third step. Standardized β values were reported to account for differing scales used in analysis. A p-level of 0.05 was used to determine statistical significance.

Table 2.

Correlations among psychosocial variables

Scale
	M	SD	Hypoglycemia Fear Survey-Worry Scale	Protective Factors Survey- Family Functioning Scale	Social Problem-Solving Inventory	BPFAS Frequency Score	BPFAS Problem Score
Hypoglycemia Fear Survey-Worry Scale	44.52	11.35	-	−.21**	−.26**	.20*	.18*
Protective Factors Survey- Family Functioning Scale	5.63	1.00	-	-	.46**	−.23**	−.21**
Social Problem-Solving Inventory	74.56	13.43	-	-	-	−.31**	−.33**
Behavioral Pediatrics Feeding Assessment Scale- Frequency	18.91	4.98					.75**
Behavioral Pediatrics Feeding Assessment Scale- Problem	1.29	2.16

^**Correlation is significant at the 0.01 level (2-tailed)

^*Correlation is significant at the 0.05 level (2-tailed)

RESULTS

Bivariate Analyses

On average, parents reported engaging in problem mealtime management behaviors somewhat frequently (M=18.84, SD=4.95) and that they perceived their mealtime behaviors as somewhat problematic (M=1.26, SD=2.16). In this sample, 65 (41%) parents had elevated Frequency scores and 41 (26%) had elevated Problem scores.

Parents of children using CGM did not have significantly different BPFAS-Parent Frequency scores than non-users (CGM-users: M=20.02±5.27; CGM non-users: M=18.55±4.86). However, parents of children using CGM had higher BPFAS-Parent Problem scores (M=1.89±2.42) than non-users (M= 1.10±2.04). BPFAS Frequency and Problem subscale scores did not differ significantly across parent or child sex or parent race/ethnicity. We also explored the degree that parent race/ethnicity and both parent sex and child sex were related to mealtime behavior, all of which were not significant.

All psychosocial variables were significantly correlated with one another (Table 2). Higher BPFAS-Parent Frequency scores were correlated with higher fear of hypoglycemia, lower family functioning, and lower social problem solving scores. Higher BPFAS-Parent Problem scores were also correlated with higher fear of hypoglycemia, lower family functioning, and lower social problem solving scores.

Hierarchical Regressions

Results of regression analyses for BPFAS-Frequency and BPFAS-Problem scores are reported in Table 3. For the BPFAS-Frequency regression, the model with only demographic variables did not significantly predict frequency of parents’ problem mealtime behavior. The addition of CGM use did not lead to a significant increase in frequency. However, the full model including psychosocial variables led to a significant increase in R². The full model including demographic, medical, and psychosocial variables was significant and explained 14% of the variance in BPFAS-Parent Frequency scores. Examination of the beta coefficients showed that problem solving was significantly associated with BPFAS-Parent Frequency scores (β=−.27, p<.01) indicating that lower problem solving skills were associated with higher reported frequency of parents’ problem mealtime behavior.

Table 3.

Hierarchical Regressions

Predictor Variables	Model 1: Demographic Variables	Model 2: Demographic + Medical Variables	Model 3: Demographic + Medical + Psychosocial Variables
	Standardized β	Standardized β	Standardized β
BPFAS Frequency Score
Step 1: Demographic
Parent Race/Ethnicity	.02	.04	.05
Parent Sex	−.13	−.14	−.11
Child Sex	.03	.03	−.01
Step 2: Medical
CGM Use		.11	.15
Step 3: Psychosocial
HFS-Worry			.11
PFS-Family Functioning			−.05
SPSI			−.27**
Model Statistics
R 2	0.02	0.03	.14
ΔR 2		0.01	.11
F	0.92	1.63	6.26
p	0.43	0.20	.00**
BPFAS Problem Score
Step 1: Demographic
Parent Race/Ethnicity	−.03	−.01	.01
Parent Sex	−.07	−.07	.02
Child Sex	.04	.04	−.01
Step 2: Medical
CGM Use		.15	.20**
Step 3: Psychosocial
HFS-Worry			.08
PFS-Family Functioning			−.04
SPSI-R:S			−.32**
Model Statistics
R 2	.01	.03	.16
ΔR 2	-	.02	.13
F	.35	3.44	7.32
p	.79	.07	.00**

HFS-Worry: Hypoglycemia Fear Survey Worry Scale

PFS-Family Functioning: Protective Factors Survey- Family Functioning Scale

SPSI-R:S: Social Problem-Solving Inventory

BPFAS Frequency Score: Behavioral Pediatrics Feeding Assessment Scale- Frequency

BPFAS Problem Score: Behavioral Pediatrics Feeding Assessment Scale- Problem

For the BPFAS-Problem regression (Table 3), the model with only demographic variables did not significantly predict parents’ perceptions of whether their mealtime behaviors were problematic. The addition of CGM use to the demographic variables did not demonstrate a significant increase in R². However, the full model including psychosocial variables resulted in a significant increase in R² of .13. Thus, the full model of demographic medical and psychosocial variables was significant and explained 16% of the variance in BPFAS-Problem scores. Examination of the beta coefficients showed that CGM use and problem solving scores were significantly correlated with BPFAS-Parent Problem scores, indicating that CGM use and lower problem solving skills were associated with higher perceived problematic mealtime behavior.

DISCUSSION

Soon after a child’s diagnosis of T1D, parent psychosocial factors were significant when controlling for medical and demographic factors in relation to the frequency of parents’ problem mealtime behaviors and their perceptions of those behaviors being problematic. Of the factors considered, psychosocial variables were related to both frequency and perception of parents’ problem mealtime behavior above and beyond CGM use, emphasizing the relative importance of parents’ social problem solving skills in particular, especially in the early period after their child’s T1D diagnosis. Other literature has described parents of children with established (>6 months duration) T1D reporting increased stress, lower confidence, and increased anxiety.^9,10 Our study added to literature that many parents of young children who have been diagnosed for <2 months find mealtimes at least somewhat challenging. This evidence about early mealtime family behaviors is important to consider as part of clinical efforts to support families of young children as they adjust to management of a new chronic illness.²⁵

On average, our results did not reach published cutoff scores for parent problem mealtime behaviors but were similar to data published from samples of children with T1D.⁸ Yet a substantial minority of the sample (41%) did endorse elevated problem mealtime behavior perceptions, suggesting that although the average score was not elevated, many parents may need support to establish positive mealtime routines. Results partially supported initial hypotheses, in that parents who experienced higher fear of hypoglycemia, lower problem solving ability, and lower levels of family functioning reported higher frequency and more problematic perceptions of their mealtime behaviors. However, these psychosocial variables together accounted for only a small portion of the variance in parent mealtime behavior. Though the combination of fear of hypoglycemia, problem solving skills, and family functioning was significantly related to parent mealtime behavior, this percentage of variance was relatively low.

Of these factors, only problem solving was significantly related to the parent behavior outcomes in the final models, suggesting this is a relatively important psychosocial factor during parents’ initial adjustment to a T1D diagnosis in young children. Being able to effectively solve challenging problems during stressful mealtime situations may contribute to how parents respond to their young children during mealtimes (such as restricting foods or pressuring their children to eat) and how they feel about these behaviors. This finding aligns with other research emphasizing the relative importance of problem solving skills. For example, Chisholm et al. reported that parents of young children recently diagnosed with T1D who encouraged higher problem solving and communication skills at mealtimes had higher dietary adherence and engaged in more diabetes management behaviors overall.²⁶ That fear of hypoglycemia and family functioning were not uniquely significantly linked with parent mealtime behaviors in the regression analyses suggests that the action orientation of problem solving may be more closely and directly related to mealtime behaviors, possibly superseding more distal links with fear of hypoglycemia and general family functioning.

Given the overall low variance explained by the models, other factors should be considered to more fully understand problem mealtime behaviors among parents of young children with newly diagnosed T1D. Other factors not included in this analysis may be more relevant to parent mealtime behaviors, such as parent stress or mood concerns. For example, in a study measuring pediatric parenting stress in parents of young children diagnosed with T1D for at least one year, depressive symptoms and fear of hypoglycemia accounted for 68% of variance in pediatric parenting stress.²⁷ In another study with parents of young children diagnosed with T1D for under a year, parents with elevated depressive and distress symptoms at baseline had prolonged symptoms of diabetes distress over time.²⁸

The finding that parents of children using CGM perceived their own mealtime behavior as more problematic advances the field by bringing awareness to potential challenges that some families may experience when adjusting to new technologies soon after diagnosis. While many parents of young children with T1D have reported that using CGM increased their confidence in diabetes management and reduced their fear of hypoglycemia, features of the CGM (e.g., the volume of data, frequent alerts) may increase parent worry.¹⁶ It is also possible that parents that seek out CGMs early may have higher baseline worry.²⁴ These patterns have not been previously published in the new diagnosis period. Parents that adopt CGMs soon after diagnosis may be more aware of blood glucose fluctuations, and it is plausible that this increased data may influence how parents respond to young child mealtime behaviors like food refusal or leaving the table. This awareness may contribute to parent behaviors to prevent food refusal (e.g., preparing alternative foods or pressuring to eat) to meet blood glucose goals. Because of the correlational nature of this study, we cannot determine causal direction between these factors. While psychosocial variables accounted for relatively more variance in parent mealtime behavior than CGM, CGM use may be affected by psychosocial factors and warrants further research. Given increasing rates of CGM use in young children, especially soon after diagnosis,¹⁶ these patterns highlight the need for research about CGM use and parent mealtime functioning to help support families in managing these challenges as they adjust to a new T1D diagnosis in young children.

No demographic factors were associated with parents’ problem mealtime behavior, in contrast to some previous evidence of sex and race/ethnicity differences in approaches to food and mealtimes in children without T1D.^17,29 It is possible that there are no differences in these constructs for families of young children with newly diagnosed T1D, and it is also possible that the demographics of our sample did not reflect a wide enough range to capture differences that may exist. Further, mealtime practices are nuanced across cultures, and meaning assigned to certain mealtime behaviors is culturally derived, so a more detailed assessment approach is necessary, given the measure we used primarily present items that lean towards Eurocentric eating norms. Further studies with a targeted focus on cultural norms, or other factors like SES status, food insecurity, or mealtime structure, may capture more relevant demographic considerations relevant to mealtime environment and family interactions in this population.

This study has several strengths and limitations. This study had a majority Non-Hispanic White participants, but generalizability was enhanced by this multisite sample being more demographically diverse compared to other studies with young children with T1D,^25,27 with 15% Non-Hispanic Black, 12% Hispanic/Latino/a/x, 7% Asian/Asian American, 2.5% Multiracial and American Indian/Alaskan Native participants, and over one-quarter on public health insurance. Moreover, recruitment occurred across 2 large hospital systems in major metropolitan areas of the United States, enhancing the geographic diversity of the sample. This analysis collapsed multiple racial/ethnic groups to maximize participant data which limited our ability to detect meaningful differences across racial, ethnic, or cultural groups. Additional considerations for generalizability include that the sample of self-identified primary caregivers was mostly mothers, who may have different behaviors or perceptions than fathers or other caregivers³⁰, and the scope of this study focused on typically developing children with T1D in respect to eating behavior. The findings in this study may not extend to older children with T1D or those with disordered eating behavior and T1D due to different developmental needs surrounding mealtimes for each of these groups. This study comprised families participating in a clinical trial and was a secondary analysis of cross-sectional data at baseline of the trial, so we could not comment on temporal or causal links between variables. A longitudinal or experimental design would provide data about family dynamics around mealtime behaviors over time, including later in a child’s T1D diagnosis and later in childhood. Except for the fear of hypoglycemia measure, the instruments used to assess family psychosocial factors (i.e., problem solving, family functioning) and parent problem mealtime behaviors were general rather than specific to diabetes because of the recency of T1D diagnosis; using diabetes-specific measures may have captured disease-relevant experiences more precisely and provided differing results.

Future research directions include investigating other psychosocial, medical, and demographic factors associated with parent mealtime management, such as parent stress, socioeconomic status, or HbA1c. Additionally, adding a detailed measure of children’s eating behaviors such as the Child Feeding Questionnaire³¹ could provide more detailed information on the categorized types of parent mealtime behaviors (i.e. restriction, pressure to eat) not included in the BPFAS. Together, these data could help inform future interventions to improve family mealtimes during the challenging new-onset period and in this age range (under six years).

Clinically, identifying family factors that predict parents’ mealtime difficulties early in a T1D diagnosis can help diabetes care teams recognize families who are struggling and help them receive targeted levels of care.¹¹ In line with Kazak’s PPPHM Model, this early intervention period is critical to learning how to prevent potential later issues. Education and support for parents of young children during the initial post-diagnosis adjustment period may benefit from an increased focus on teaching parents problem solving skills that can be used for common mealtime challenges. Screening for mealtime difficulties at diagnosis may help identify the subset of families with scores above clinical cutoffs, who may need focused support around managing mealtimes. While this study considered the BPFAS problem scale as an indicator of suboptimal parent mealtime behaviors, it is possible that parents’ recognition of their mealtime behaviors as problematic is a necessary part of making improvements. Thus, professional consultation may be helpful to understand parents’ experiences and perceptions about mealtimes and support them in establishing positive mealtime behaviors and routines. Finally, CGMs are becoming more widespread and more financially accessible³² to families with T1D. Therefore, it may be useful to integrate more guidance around behavior management at mealtimes into CGM education protocols. Family education and support focusing on using CGM data effectively at mealtimes could help with diabetes management and may address parental concerns about hypoglycemia or assist with problem solving.