Skip to main content
NCBI home page
Journal of Pediatric Psychology logoLink to Journal of Pediatric Psychology
. 2021 Aug 7;46(9):1110–1118. doi: 10.1093/jpepsy/jsab042

Insulin Restriction, Emotion Dysregulation, and Depressive Symptoms in Late Adolescents with Diabetes

Aislinn B Beam 1,✉, Deborah J Wiebe 1, Cynthia A Berg 2
PMCID: PMC8634538  PMID: 34363674

Abstract

Objective

Insulin restriction occurs when an individual takes less insulin than recommended and is a serious concern for those with diabetes. General insulin restriction (IR) and insulin restriction for weight control (IRWC) have not been clearly distinguished in the literature, creating inconsistencies and limited understanding of factors that underlie this behavior. We examined whether these are distinct, and how emotion dysregulation and depressive symptoms relate to both forms of insulin restriction during late adolescence.

Methods

As part of a larger study, late adolescents (ages 17–18) with type 1 diabetes (N = 236) completed measures of depressive symptoms (Center for Epidemiologic Studies—Depression Scale [CES-D]), facets of Difficulties In Emotion Regulation Scale (DERS), diabetes self-management behaviors, insulin restriction, and hemoglobin A1c (HbA1c).

Results

IR and IRWC were not significantly associated with each other. IR was associated with self-management behaviors but not HbA1c, whereas the opposite was true for IRWC. All DERS subscales (M = 10.60–16.73) and CES-D (M = 16.56) were correlated with greater IRWC; CES-D and all but one DERS subscale were correlated with IR. Covariation with CES-D explained associations between DERS and IRWC. CES-D moderated associations with IR, indicating most subscales of the DERS were associated with IR only when CES-D was higher.

Conclusion

Emotion dysregulation and depressive symptoms are important correlates of the dangerous behavior of insulin restriction, but function differently when insulin is restricted specifically for weight control versus nonspecified reasons. Future research to understand these underlying processes will be necessary to develop emotion-based theory and evidence-based interventions for this dangerous behavior.

Keywords: adolescents, depression, diabetes, insulin restriction, insulin omission, risk behavior

Introduction

Insulin restriction is a dangerous practice in which individuals with type 1 diabetes reduce their insulin intake against treatment recommendations, often leading to hyperglycemia and serious health complications (see De Paoli & Rogers, 2018 for review). Insulin restriction may occur for a variety of reasons, such as mismanagement (e.g., forgetting to dose insulin before a snack), fear of hypoglycemia, or as a specific form of disordered eating. Of particular concern is the practice of restricting insulin for the purpose of perceived or actual weight loss, which occurs when an individual takes little or no insulin which can induce ketosis (causing weight loss by passing glucose through urine). Both general forms of insulin restriction (IR) and insulin restriction for the specific purpose of weight control (IRWC) have been found to have dangerous consequences for individuals with type 1 diabetes. General IR has been associated longitudinally with lower self-management and higher hemoglobin A1c (HbA1c) (Tracy et al., 2019), and predicted three times the risk of mortality over an 11-year period (Goebel-Fabbri et al., 2008). IRWC similarly has been associated with elevated HbA1c (Battaglia et al., 2006; Luyckx et al., 2019) and increased risk of ketoacidosis, retinopathy, and nephropathy (Takii et al., 2008). Unfortunately, disordered eating behaviors including taking less insulin are fairly common in adolescence (Araia et al., 2017; Nip et al., 2019; Wisting et al., 2013), with prevalence rates for IR specifically ranging from 7.4% to 33% (Neumark-Sztainer et al., 2002; Nip et al., 2019; Peveler et al., 2005). Given such findings, it is imperative to understand factors that underlie this dangerous behavior during adolescence.

Two interrelated factors—emotion dysregulation and depression—have often been studied as risk factors for disordered eating and are theorized to be associated with IR. Emotion dysregulation is a multidimensional construct that captures difficulties not only in modulating the intensity and duration of negative emotions, but also in one’s awareness, clarity, and acceptance of negative emotion, and one’s ability to control impulses and maintain goal-directed behavior when distressed (Gratz & Roemer, 2004). An expansive literature identifies emotion dysregulation as a cross-diagnostic risk factor underlying both depression (Weinberg & Klonsky, 2009) and disordered eating (Lavender et al., 2015), and emotion dysregulation has been specifically linked to bulimia symptoms in adolescents with type 1 diabetes (Young-Hyman et al., 2016). Theories about the development of disordered eating and IR among those with type 1 diabetes suggest a central role for negative emotion and emotion dysregulation (De Paoli & Rogers, 2018). There is evidence that negative emotion forms a proximal context that increases the risk of IR, potentially reflecting a maladaptive strategy for modulating negative affect. Using ecological momentary assessments, Merwin et al. (2015) found that when participants experienced momentary increases in negative affect, they were more likely to restrict insulin at their next meal, suggesting negative emotion sets an immediate context that triggers risk for IR. Indeed, depressive symptoms are associated with IR generally (Goebel-Fabbri et al., 2011) and with disordered eating behaviors including IR (Luyckx et al., 2019; Olmsted et al., 2008), with some evidence that depression may mediate links between emotion dysregulation and disordered eating in nondiabetes samples (Gilboa-Schechtmann et al., 2006). Such findings suggest that links between emotion dysregulation and IR primarily reflect shared variance with depression.

Some facets of emotion dysregulation may be associated with IR primarily in the presence of elevated negative emotion. For example, facets of emotion dysregulation that reflect difficulty controlling impulses or maintaining goal-directed behavior when distressed may increase risk for IR primarily when individuals are experiencing elevated negative emotion, but not in the absence of negative emotion. From this perspective, some dimensions of emotional dysregulation would be expected to interact with depressive symptoms to predict IR. Young-Hyman et al. (2016) found that depressive symptoms were more strongly linked to bulimic symptoms among individuals with type 1 diabetes who had higher scores on emotion dysregulation. Similarly, a recent study found that negative affect was linked to disordered eating behaviors only among adolescents with type 1 diabetes who reported higher “negative urgency,” a trait characterized by impulsivity when experiencing negative emotions (Rose et al., 2020). Although such findings support the possibility that facets of emotion dysregulation will interact with depression to predict IR, these studies did not examine IR specifically.

This study examined whether and how emotion dysregulation and depressive symptoms were associated with both IR and IRWC among late adolescents with type 1 diabetes. Measures of IR have not consistently identified the motivation underlying insulin manipulation, with some measures referring to general IR and others referring to IRWC. These varied measures have been argued to explain inconsistencies in the literature, with calls for researchers to more carefully identify the motivation underlying IR (Peyrot et al., 2010). We are not aware of studies that have simultaneously examined general IR and IRWC in a single sample. As an initial step to determine whether they are distinct or interrelated constructs, Aim 1 examined how each IR variable was associated with each other, as well as with self-management and HbA1c. Aim 2 examined whether aspects of emotion dysregulation were associated with IR, and whether associations remained when depressive symptoms were statistically controlled. We hypothesized that higher emotion dysregulation would be associated with higher levels of both IR and IRWC, but that these associations would at least partially reflect shared variance with depression. We explored whether aspects of emotion dysregulation showed different patterns of association among variables, but did not make specific hypotheses due to the inconsistent and limited literature. Aim 3 examined whether aspects of emotion dysregulation interacted with depressive symptoms to predict IR. We hypothesized that some subscales—particularly those that tap the ability to control impulses and maintain goal-directed behavior when distressed—would have stronger associations with IR among individuals experiencing higher depressive symptoms. Addressing these aims will provide novel information on differences between types of IR, and begin to identify the role of emotion dysregulation in IR during late adolescence, a time of high risk for diabetes management when disordered eating behaviors peak (Nip et al., 2019).

Materials and Methods

Participants

This study was a secondary analysis of data from a larger multisite longitudinal project examining diabetes management from late adolescence into early emerging adulthood (see Berg et al. 2019). All study procedures were approved by the relevant Institutional Review Boards (IRB). This study analyzed data collected at baseline (when participants were in their last year of high school), as this time captures a high-risk age for both diabetes management and disordered eating (Hanlan et al., 2013). High school seniors (61% female) aged 17–18 years (M = 17.7, SD = 0.40) were recruited from pediatric endocrinology clinics in two southwestern U.S. cities. Eligibility requirements included diagnosis of type 1 diabetes for more than 1 year (M = 7.36, SD = 3.92), English as their primary language, living with a parent, and no condition that would prohibit study completion (e.g., severe intellectual disability). Around half used an insulin pump (42.1%), and most were covered by their parents’ insurance (72.5%). Similar to national samples of youth with type 1 diabetes (Miller et al., 2015), most participants were non-Hispanic White (75%), with 14% identifying as Hispanic/Latino. Most participants’ parents had some college education or above (84.8% for mothers, 77.6% for fathers), and average household income for this sample was $87,609.

Of eligible individuals recruited, 48.7% agreed to participate in a 3-year longitudinal study (N = 247). Common reasons for declining were being too busy (34%) and lack of interest (33%); 20% declined to give a reason. At one site, IRB allowed review of medical records to compare those who did or did not participate (N = 118 vs. 98). Participants and nonparticipants did not differ on HbA1c, illness duration, pump status, gender, or race/ethnicity. Of those who agreed to participate, eleven were excluded from analyses due to invalid data (final N = 236).

Procedure

Participants were invited to an initial assessment in a laboratory setting, where they provided consent (18 or older) or assent with parental consent (<18). Participants were trained to use an online survey and provided a blood sample to assay blood glucose; they then completed confidential surveys online at home. Participants were paid $50 for the survey and HbA1c assay.

Measures

Hemoglobin A1c

HbA1c was used to indicate average blood glucose over the prior few months as it is the medical standard for measuring glycemic control. Higher HbA1c is associated with long-term microvascular complications (Škrha et al., 2016). HbA1c was analyzed via mail-in dried blood spot assay kits provided and processed by CoreMedica laboratories (accredited by the College of American Pathologists; https://www.coremedicalabs.com/). The average level for this sample was 8.27 (SD = 1.64). This measure was highly correlated with point-of-care HbA1c assays in medical records (r = .74, p < .001).

Diabetes Self-Management

The Diabetes Behavior Rating Scale (DBRS) measured diabetes self-care across 37 items that tap behaviors necessary for diabetes health. The DBRS has good concurrent validity with interview measures of self-care, and predictive validity of HbA1c (Iannotti Nansel et al., 2006). Scores were computed as a proportion ranging from 0 to 1 (present sample α = .840).

Emotional Dysregulation

Emotional dysregulation was assessed with the Difficulties in Emotion Regulation Scale (DERS; Gratz & Roemer, 2004). The scale is comprised of six subscales, each of which measures difficulties with important regulation skills including: awareness and acceptance of negative emotions, clarity of what emotions are being experienced, accessing strategies to regulate negative emotions, and impulse control and engaging in goal-directed behavior when experiencing negative emotions. Each subscale consists of five to eight items, which are answered using Likert-scale ranging from 1 (almost never) to 5 (almost always).

Scores are calculated by summing all relevant items. In the present sample, internal consistency was good for the composite score (α = .965), and all subscales (acceptance: α = .924; goals: α = .894; impulse control: α = .872; awareness: α = .816; strategies: α = .906; clarity: α = .842).

Depression Symptoms

Symptoms of depression were assessed via the Center for Epidemiologic Studies—Depression Scale (CES-D; Radloff, 1977). Participants rated their experience of depressive symptoms in the past week using a 4-point Likert scale ranging from 0 (rarely or none of the time) to 3 (most or all of the time). Internal validity was excellent (α = .931). Although the CES-D was analyzed as a continuous measure, 42 participants (17.9%) scored in the 16–23 range indicating mild depressive symptoms and 57 (24.4%) scored 24 or higher indicating moderate/severe depressive symptoms.

Insulin Restriction

IR has been assessed in a variety of ways. Some researchers have used a single item asking if individuals take less insulin than they should (Goebel-Fabbri et al., 2011; Peyrot et al., 2010), whereas others have used single items to assess IR for the purpose of weight management (d’Emden et al., 2013). We used both approaches, asking participants “How often do you take less insulin than you should?” and “How often do you take less insulin or skip a dose of insulin to lose weight or keep from gaining weight?” with Likert-scale answers of 0 (never), 1 (rarely), 2 (sometimes), 3 (often), 4 (usually), and 5 (always). These items were asked consecutively as part of a set of questions assessing risk behaviors. Such single item measures have predicted higher mortality rates (Goebel-Fabbri et al., 2008) and higher HbA1c (Battaglia et al., 2006).

Statistical Analyses

Preliminary analyses were conducted in SPSS version 26, while regression models were analyzed in Mplus version 7.4 (Muthén & Muthén, 1998–2012); full information maximum likelihood accounted for missing data (missing data < 4% for all variables). Preliminary analyses indicated that most participants never restrict insulin to lose weight (N = 197, 83.5%), with remaining participants reporting a range of frequencies. We thus dichotomized IRWC to indicate “never” versus “ever” engaging in this behavior; scores ≥ 1 were categorized as ever. Goebel-Fabbri et al. (2008) found higher mortality for a similarly dichotomized item, suggesting that it captures a significant risk regardless of frequency. In contrast, a range of frequencies were reported for the general IR item: never (11.5%), rarely (31.9%), sometimes (40.0%), often (12.3%), usually (3.8%), and always (0.4%). The full range of IR scores was thus analyzed.

We first conducted bivariate correlations to address Aims 1 and 2. Aims 3 and 4 were then tested in Mplus, using logistic regressions when the dichotomous IRWC measure was the outcome variable. In these models, we initially entered emotion dysregulation and depressive symptoms (centered on their means) simultaneously in Step 1 to determine whether each variable had unique associations with IR. We then entered their cross product on Step 2 to determine whether the emotion dysregulation × depressive symptoms interaction predicted additional variance in IR. In separate analyses, we examined models with the DERS composite, followed by each DERS subscale in order to explore whether specific facets of emotion dysregulation carried the effect when the composite score was found to be associated with IR, or whether the composite variable may have masked specific facets of emotion dysregulation when composite effects were not found. The variables of gender, ethnicity (non-Latino White vs. other), SES, illness duration, and pump status were considered as covariates as they have been linked to IR in prior work; these variables were excluded for parsimony as they were not correlated with either IR variable (r values ranged from −.070 to .084, p values > .201). Notably, when both males and females have been included in prior studies of IR, gender differences are not consistently found (Araia et al., 2017; Luyckx et al., 2019).

Results

Bivariate correlations among study variables are shown in Table I. General IR and IRWC were not correlated with each other and were associated with different diabetes outcome measures. Specifically, IRWC was associated with higher HbA1c but not with adherence, while the opposite pattern was true for general IR. It is notable that HbA1c was higher among individuals who indicated they do versus do not skip insulin to lose weight, M (SD) = 8.94% (1.93) versus 8.14% (1.54), t (232) = −2.852, p = .005. As expected, all facets of emotion dysregulation were correlated with higher depressive symptoms, and both emotion dysregulation and depressive symptoms were generally correlated with IR. The one exception was that the subscale capturing difficulties in Awareness were not correlated with general IR.

Table I.

Correlations Among Study Variables

Mean (SD) 1 2 3 4 5 6 7 8 9 10 11
1. IR 1.66 (0.992)
2. IRWC 0.165 (0.372) 0.06
3. HbA1c 8.27 (1.63) 0.12 0.18**
4. Self-management 0.607 (0.122) −0.34** −0.10 −0.24**
5. Acceptance 13.04 (6.42) 0.26** 0.18** 0.02 −0.24**
6. Goal-directed behavior 13.23 (5.04) 0.21** 0.15* 0.09 −0.28** 0.49**
7. Impulse control 11.37 (4.97) 0.20* 0.25** 0.21** −0.29** 0.54** 0.59**
8. Awareness 15.21 (5.20) 0.05 0.22** 0.05 −0.32** 0.26** 0.08 0.24**
9. Strategies 16.73 (7.28) 0.20** 0.26** 0.18** −0.32** 0.66** 0.70** 0.75** 0.23**
10. Clarity 10.60 (4.32) 0.18** 0.19** 0.08 −0.36** 0.51** 0.41** 0.53** 0.53** 0.60**
11. DERS composite 80.17 (25.14) 0.25** 0.28** 0.14* −0.39** 0.78** 0.73** 0.82** 0.49** 0.90** 0.77**
12. CES-D 16.56 (12.00) 0.16* 0.29** 0.24** −0.33** 0.54** 0.58** 0.62** 0.31** 0.77** 0.60** 0.77**
Open in a new tab

Note. IR = general insulin restriction; IRWC = insulin restriction specifically for weight control; HbA1c = hemoglobin A1c; DERS = Difficulty in Emotion Regulation Scale; CES-D = Center for Epidemiologic Studies Depression Scale.

*

p < .05;

**

p < .01

Results of regression analyses are shown in Table II, with Aim 3 addressed in Step 1 and Aim 4 addressed in Step 2. Focusing first on the IRWC (left columns), neither the composite nor depressive symptoms were uniquely associated with IRWC when entered simultaneously, potentially reflecting their high correlation (r = .77). However, in all subscale analyses, emotion dysregulation was no longer related to IRWC, while the association for depressive symptoms remained. This pattern is consistent with the hypothesis that associations of emotion dysregulation with IRWC reflect shared variance with depressive symptoms. There was no evidence that emotion dysregulation and depressive symptoms interacted to predict IRWC.

Table II.

Regression Analyses Examining Unique and Interacting Associations of Emotion Dysregulation and Depressive Symptoms Predicting Insulin Restriction

Insulin restriction for weight control
 General insulin restriction
Coefficient (SE) p CI Coefficient (SE) p CI
DERS composite Step 1 DERS 0.015 (0.011) .198 −0.008 to 0.037 0.013 (0.004) .001 0.005–0.020
CES-D 0.036 (0.023) .139 −0.012 to 0.085 −0.015 (0.009) .080 −0.032 to 0.002
Step 2 DERS*CES-D 0.000 (0.001) .984 −0.001 to 0.001 0.001 (0.000) .001 0.000–0.001
Acceptance of emotions (ACC) Step 1 ACC 0.019 (0.036) .591 −0.051 to 0.089 0.029 (0.012) .011 0.007–0.052
CES-D 0.055 (0.018) .002 0.020–0.090 −0.001 (0.006) .865 −0.013 to 0.011
Step 2 ACC*CES-D −0.001 (0.002) .811 −0.005 to 0.004 0.002 (0.001) .001 0.001–0.004
Goal-directed behavior (GOA) Step 1 GOA −0.015 (0.048) .756 −0.109 to 0.079 0.036 (0.015) .021 0.005–0.066
CES-D 0.064 (0.020) .001 0.026 − 0.102 0.001 (0.007) .866 −0.012 to 0.015
Step 2 GOA*CES-D −0.002 (0.003) .813 −0.007 to 0.005 0.002 (0.001) .038 0.000–0.004
Impulse control (IM) Step 1 IM 0.057 (0.046) .218 −0.034 to 0.147 0.027 (0.016) .095 −0.005 to 0.059
CES-D 0.044 (0.020) .023 0.006–0.083 0.002 (0.007) .823 −0.012 to 0.015
Step 2 IM*CES-D 0.000 (0.003) .967 −0.006 to 0.006 0.002 (0.001) .021 0.000–0.005
Awareness of emotions (AW) Step 1 AW 0.077 (0.040) .054 −0.001 to 0.155 0.002 (0.013) .894 −0.024 to 0.028
CES-D 0.051(0.016) .001 0.020–0.082 0.014 (0.006) .012 0.003–0.025
Step 2 AW*CES-D 0.000 (0.003) .919 −0.005 to 0.006 0.000 (0.001) .891 −0.002 to 0.002
Strategies for negative emotions (ST) Step 1 ST 0.028 (0.039) .478 −0.049 to 0.104 0.022 (0.013) .101 −0.004 to 0.048
CES-D 0.046 (0.024) .049 0.000–0.093 −0.006 (0.008) .494 −0.022 to 0.011
Step 2 ST*CES-D 0.000 (0.002) .947 −0.004 to 0.004 0.002 (0.001) .001 0.001 to 0.004
Clarity of emotions (CL) Step 1 CL 0.020 (0.056) .725 −0.089 to 0.128 0.026 (0.019) .161 −0.010 to 0.063
CES-D 0.057 (0.018) .002 0.021–0.093 0.007 (0.007) .323 −0.007 to 0.020
Step 2 CL*CES-D –0.001 (0.03) .738 −0.008 to 0.005 0.001 (0.001) .337 −0.001 to 0.004
Open in a new tab

Note. DERS = Difficulty in Emotion Regulation Scale; CES-D = Center for Epidemiologic Studies Depression Scale.

A different pattern was identified for analyses predicting the general IR variable (right columns, Table II). There was mixed evidence that emotion dysregulation had an association with general IR independent of depressive symptoms. The composite, as well as difficulties in acceptance of emotion and in engaging in goal-directed behavior when distressed remained associated with higher IR independent of depressive symptoms; the remaining emotion dysregulation scores did not. However, there was a significant interaction between emotion dysregulation and depressive symptoms for the composite and all subscales except awareness and clarity of emotions. Predicted means for the composite score are displayed in Figure 1; emotion dysregulation was associated with higher IR among participants with higher depressive symptoms, but not among those with lower depressive symptoms. Predicted means for interactions with the subscales showed identical patterns (see Supplementary Appendix A).

Figure 1.

Figure 1.

Open in a new tab

Predicted means for interaction of emotion dysregulation composite and depressive symptoms predicting general insulin restriction.

Discussion

This study provides important insights into the roles of emotion dysregulation and depression in the dangerous behavior of IR among late adolescents, a high-risk group for managing type 1 diabetes. Multiple facets of emotion dysregulation were associated with risk of IR generally, as well as specifically for the purpose of weight control, suggesting that a range of emotion regulation skills may be important for limiting this dangerous behavior. Findings also demonstrated that general IR and IRWC are different constructs. These variables were unrelated to each other, had different frequencies, and were associated with different aspects of diabetes management. Specifically, general IR was associated with lower self-management behaviors, while IRWC was associated with higher HbA1c. It is notable that the 0.8% difference between those who did versus did not restrict insulin for weight control is clinically meaningful given that a one percentage increase in HbA1c (e.g., 8–9%) is associated with a 40% increased risk of developing retinopathy (Hood et al., 2009). Finally, emotion dysregulation and depressive symptoms showed different patterns of association with general IR and IRWC. Understanding these underlying patterns may provide insights into different processes of risk for IR that can guide different points of intervention.

Study findings extend the literature on disordered eating in nondiabetes samples to enhance our understanding of IRWC among youth with type 1 diabetes. In nondiabetes samples, emotion dysregulation is consistently associated with disordered eating (Lavender et al., 2015), with some evidence that depression mediates that association (Gilboa‐Schechtman et al. 2006). Consistent with this pattern, we found every facet of emotion dysregulation was no longer associated with IRWC once shared variance with depressive symptoms was controlled, while depressive symptoms remained uniquely associated with IRWC. Thus, emotion dysregulation risks appear to extend to the dangerous behavior of IRWC, with depressive symptoms playing a central role in that relationship. The absence of an interaction of depression with emotion dysregulation further demonstrates its prominent role, as even those with good emotion regulation skills appeared at risk for IRWC when elevated depressive symptoms were present.

Our cross-sectional design does not allow us to discern the direction of associations among variables. Although findings are consistent with a model where emotion dysregulation enhances risk for depression which then forms a proximal context promoting IRWC, other explanations are possible. For example, depressive symptoms may limit access to emotion regulation resources, thus enhancing risk for IRWC. Theoretical reviews on the central role of affect regulation in disordered eating in both diabetes and nondiabetes samples have argued that emotion dysregulation may function as a risk factor for developing disordered eating and/or as a maintenance factor, where the repeated use of disordered eating behaviors to manage negative emotions functions to maintain symptoms across time (Lavender et al., 2015; De Paoli & Rogers, 2018). Research using alternative designs (e.g., ecological momentary assessments or longitudinal studies) will be a high priority to disentangle potential links between emotion dysregulation, depression, and IRWC.

Although emotion dysregulation and depressive symptoms did not interact to predict IRWC, this interaction was found for general IR. The absence of an interaction for IRWC was surprising given recent studies of such an effect when predicting disordered eating (Rose et al., 2020; Young-Hyman et al., 2016). Inconsistent findings may reflect numerous methodological differences across studies (i.e., different measures of emotion dysregulation, negative emotion, and disordered eating; different ages). In contrast to findings for IRWC, however, multiple facets of emotion dysregulation interacted with depressive symptoms to predict general IR. We had expected this interaction for difficulties in Impulse Control and Goal-Directed Behavior as, by definition, these facets capture difficulty in engaging in adaptive behavior specifically when experiencing negative emotion. This interaction was also found for difficulties in Accepting and in accessing Strategies to modulate negative emotion, suggesting that depressive symptoms form a risky context for late adolescents who experience multiple aspects of emotion dysregulation.

An important finding of the current study is that general IR and IRWC appear to be quite different constructs. This is the first study we are aware of to examine both forms of insulin manipulation in the same sample. Although these measures have been used somewhat interchangeably, Peyrot et al. (2010) argued for the need to identify motivations for IR. General restriction of insulin (e.g., skipping a dose, taking less insulin than one should, or not taking as prescribed) could have many causes beyond disordered eating such as fear of hypoglycemia, simply forgetting when busy, or trying to reduce health costs. Each of these may have different predictors and suggest different types of interventions. We suspect that IRWC is capturing an especially dangerous aspect of disordered eating, while general IR may reflect a domain of mismanagement or self-care. We did not have a broader measure of disordered eating that would have allowed us to explore this possibility but note that IRWC can occur in the absence of other disordered eating and that those with eating disorders may not restrict insulin (De Paoli & Rogers, 2018). Additional research to develop more precise measures and to understand the host of factors that may contribute to IR is necessary.

There are limitations to this study that should be considered when interpreting the findings. First, this was a cross-sectional correlational design and causal claims regarding the relationships cannot be made. Future research examining associations across time will be needed to discern directions of association. Second, this study was a secondary analysis of an existing dataset, limiting our approach to the measures available. Validated measures of disordered eating behaviors in diabetes were not available, and depressive symptomatology was the only measure of negative emotion; other negative emotions may show different associations. Third, some of our findings may have reflected the high correlation between depressive symptoms and emotion dysregulation (especially the composite score), making it difficult to detect unique associations. However, our correlations are similar to those found in other studies, and empirical and theoretical evidence indicates these are different constructs (Neumann et al., 2010). Fourth, although the race and ethnic composition of this sample was similar to youth who have type 1 diabetes in the United States (Miller et al., 2015), the sample was mostly non-Latino White, and all participants were seniors in high school. Findings may not generalize to different samples. Fifth, measures were primarily self-reported, with both types of IR indexed by single items. Similar items have been linked to important clinical outcomes in prior studies (Battaglia et al., 2006; d’Emden et al., 2013; Goebel-Fabbri et al., 2011), but our data suggest that, as a field, we need to better define and measure these constructs. Finally, the present study only examined two out of many variables that may be associated with IR. Future research should consider the role of other variables that are commonly associated with disordered eating (e.g., BMI, body image, drive for thinness).

In conclusion, this study demonstrated that emotion dysregulation and depressive symptoms are important correlates of the dangerous behavior of IR, but function differently when insulin is restricted specifically for weight control versus nonspecified reasons. Depressive symptoms played a central role for both forms of IR, accounting for the associations of emotion dysregulation with IRWC, and moderating the associations of multiple facets of emotion dysregulation with general IR. Future research using longitudinal designs will be necessary to clarify the extent to which emotion dysregulation functions as a risk factor that contributes to depressive symptoms and IR, or as a maintenance factor where emotion becomes dysregulated in response to IR which then maintains the behavior over time (De Paoli & Rogers, 2018). Understanding these underlying processes will be necessary to develop emotion-based theory and interventions for this dangerous behavior. If supported by future research, findings suggest it may be useful to screen for disordered eating and IR in tandem with routine screening for depression, and to consider emotion regulation skills in interventions. For example, emotion regulation training combined with cognitive behavioral therapy (CBT) for depression has been shown to be more effective than CBT alone (Berking et al., 2013), and may serve as a template for late adolescents at risk for engaging in the dangerous behavior of IR.

Supplementary Data

Supplementary data can be found at: https://academic.oup.com/jpepsy.

Funding

The data used in this study were supported by a grant from the National Institutes of Diabetes and Digestive and Kidney Diseases DK092939 awarded to Deborah Wiebe and Cynthia Berg.

Conflicts of interest: None declared.

Supplementary Material

jsab042_Supplementary_Data
Click here for additional data file. (33.6KB, docx)

References

  1. Araia E., Hendrieckx C., Skinner T., Pouwer F., Speight J., King R. M. (2017). Gender differences in disordered eating behaviors and body dissatisfaction among adolescents with type 1 diabetes: Results from diabetes MILES youth—Australia. The International Journal of Eating Disorders, 50(10), 1183–1193. doi:10.1002/eat.22746 [DOI] [PubMed] [Google Scholar]
  2. Battaglia M. R., Alemzadeh R., Katte H., Hall P. L., Perlmuter L. C. (2006). Brief report: Disordered eating and psychosocial factors in adolescent females with type 1 diabetes mellitus. Journal of Pediatric Psychology, 31(6), 552–556. doi:10.1093/jpepsy/jsj047 [DOI] [PubMed] [Google Scholar]
  3. Berg C. A., Wiebe D. J., Lee Tracy E., Kelly C. S., Mello D., Turner S. L., Butner J. E., Munion A. K., Mansfield J. H., White P. C., Murray M., Such Y. (2019). Parental involvement and executive function in emerging adults with type 1 diabetes. Journal of Pediatric Psychology, 44(8), 970–979. doi:10.1093/jpepsy/jsz025 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Berking M., Ebert D., Cuijpers P., Hofmann S. G. (2013). Emotion regulation skills training enhances the efficacy of inpatient cognitive behavioral therapy for major depressive disorder: A randomized controlled trial. Psychotherapy and Psychosomatics, 82(4), 234–245. doi:10.1159/000348448 [DOI] [PubMed] [Google Scholar]
  5. d’Emden H., Holden L., McDermott B., Harris M., Gibbons K., Gledhill A., Cotterill A. (2013). Disturbed eating behaviours and thoughts in Australian adolescents with type 1 diabetes. Journal of Paediatrics and Child Health, 49(4), E317–E323. doi:10.1111/jpc.12014 [DOI] [PubMed] [Google Scholar]
  6. De Paoli T., Rogers P. J. (2018). Disordered eating and insulin restriction in type 1 diabetes: A systematic review and testable model. Eating Disorders, 26(4), 343–360. doi:10.1080/10640266.2017.1405651 [DOI] [PubMed] [Google Scholar]
  7. Gilboa‐Schechtman E., Avnon L., Zubery E., Jeczmien P. (2006). Emotional processing in eating disorders: Specific impairment or general distress related deficiency? Depression and Anxiety, 23(6), 331–339. doi:10.1002/da.20163 [DOI] [PubMed] [Google Scholar]
  8. Gratz K. L., Roemer L. (2004). Multidimensional assessment of emotion regulation and dysregulation: Development, factor structure, and initial validation of the difficulties in emotion regulation scale. Journal of Psychopathology and Behavioral Assessment, 26(1), 41–54. doi:10.1023/B:JOBA.0000007455.08539.94 [Google Scholar]
  9. Goebel-Fabbri A. E., Anderson B. J., Fikkan J., Franko D. L., Pearson K., Weinger K. (2011). Improvement and emergence of insulin restriction in women with type 1 diabetes. Diabetes Care, 34(3), 545–550. doi:10.2337/dc10-1547 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Goebel-Fabbri A. E., Fikkan J., Franko D. L., Pearson K., Anderson B. J., Weinger K. (2008). Insulin restriction and associated morbidity and mortality in women with type 1 diabetes. Diabetes Care, 31(3), 415–419. doi:10.2337/dc07-2026 [DOI] [PubMed] [Google Scholar]
  11. Hanlan M. E., Griffith J., Patel N., Jaser S. S. (2013). Eating disorders and disordered eating in type 1 diabetes: Prevalence, screening, and treatment options. Current Diabetes Reports, 13(6), 909–916. doi:10.1007/s11892-013-0418-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Hood K. K., Peterson C. M., Rohan J. M., Drotar D. (2009). Association between adherence and glycemic control in pediatric type 1 diabetes: A meta-analysis. Pediatrics, 124(6), e1171–e1179. doi:10.1542/peds.2009-0207 [DOI] [PubMed] [Google Scholar]
  13. Iannotti R. J., Nansel T. R., Schneider S., Haynie D. L., Simons-Morton B., Sobel D. O., Zeitzoff L., Plotnick L. P., Clark L. (2006). Assessing regimen adherence of adolescents with type 1 diabetes. Diabetes Care, 29(10), 2263–2267. doi:10.2337/dc06-0685 [DOI] [PubMed] [Google Scholar]
  14. Lavender J. M., Wonderlich S. A., Engel S. G., Gordon K. H., Kaye W. H., Mitchell J. E. (2015). Dimensions of emotion dysregulation in anorexia nervosa and bulimia nervosa: A conceptual review of the empirical literature. Clinical Psychology Review, 40, 111–122. doi:10.1016/j.cpr.2015.05.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Luyckx K., Verschueren M., Palmeroni N., Goethals E. R., Weets I., Claes L. (2019). Disturbed eating behaviors in adolescents and emerging adults with type 1 diabetes: A one-year prospective study. Diabetes Care, 42(9), 1637–1644. doi:10.2337/dc19-0445 [DOI] [PubMed] [Google Scholar]
  16. Merwin R. M., Dmitrieva N. O., Honeycutt L. K., Moskovich A. A., Lane J. D., Zucker N. L., Surwit R. S., Feinglos M., Kuo J. (2015). Momentary predictors of insulin restriction among adults with type 1 diabetes and eating disorder symptomatology. Diabetes Care, 38(11), 2025–2032. doi:10.2337/dc15-0753 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Miller K. M., Foster N. C., Beck R. W., Bergenstal R. M., DuBose S. N., DiMeglio L. A., Maahs D. M., Tamborlane W. V. (2015). Current state of type 1 diabetes treatment in the US: Updated data from the T1D Exchange clinic registry. Diabetes Care, 38(6), 971–978. doi:10.2337/dc15-0078 [DOI] [PubMed] [Google Scholar]
  18. Muthén L. K., Muthén B. O. (1998–2012). Mplus User's Guide. Sixth Edition. Muthén & Muthén.
  19. Neumann A., van Lier P. A., Gratz K. L., Koot H. M. (2010). Multidimensional assessment of emotion regulation difficulties in adolescents using the difficulties in emotion regulation scale. Assessment, 17(1), 138–149. doi:10.1177%2F1073191109349579 [DOI] [PubMed] [Google Scholar]
  20. Neumark-Sztainer D., Patterson J., Mellin A., Ackard D. M., Utter J., Story M., Sockalosky J. (2002). Weight control practices and disordered eating behaviors among adolescent females and males with type 1 diabetes: Associations with sociodemographics, weight concerns, familial factors, and metabolic outcomes. Diabetes Care, 25(8), 1289–1296. doi:10.2337/diacare.25.8.1289 [DOI] [PubMed] [Google Scholar]
  21. Nip A. S., Reboussin B. A., Dabelea D., Bellatorre A., Mayer-Davis E. J., Kahkoska A. R., Lawrence J. M., Peterson C. M., Dolan L., Pihoker C.; SEARCH for Diabetes in Youth Study Group. (2019). Disordered eating behaviors in youth and young adults with type 1 or type 2 diabetes receiving insulin therapy: The SEARCH for Diabetes in Youth Study. Diabetes Care, 42(5), 859–866. doi:10.2337/dc18-2420 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Olmsted M. P., Colton P. A., Daneman D., Rydall A. C., Rodin G. M. (2008). Prediction of the onset of disturbed eating behavior in adolescent girls with type 1 diabetes. Diabetes Care, 31(10), 1978–1982. doi:10.2337/dc08-0333 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Peveler R. C., Bryden K. S., Neil H. A. W., Fairburn C. G., Mayou R. A., Dunger D. B., Turner H. M. (2005). The relationship of disordered eating habits and attitudes to clinical outcomes in young adult females with type 1 diabetes. Diabetes Care, 28(1), 84–88. doi:10.2337/diacare.28.1.84 [DOI] [PubMed] [Google Scholar]
  24. Peyrot M., Rubin R. R., Kruger D. F., Travis L. B. (2010). Correlates of insulin injection omission. Diabetes Care, 33(2), 240–245. doi:10.2337/dc09-1348 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Radloff L. S. (1977). The CES-D scale: A self report depression scale for research in the general population. Applied Psychological Measurement, 1(3), 385–401. doi:10.1177%2F014662167700100306 [Google Scholar]
  26. Rose M., Streisand R., Tully C., Clary L., Monaghan M., Wang J., Mackey E. (2020). Risk of disordered eating behaviors in adolescents with type 1 diabetes. Journal of Pediatric Psychology, 45(5), 583–591. doi:10.1093/jpepsy/jsaa027 [DOI] [PubMed] [Google Scholar]
  27. Škrha J., Šoupal J., Škrha J., Prázný M. (2016). Glucose variability, HbA1c and microvascular complications. Reviews in Endocrine & Metabolic Disorders, 17(1), 103–110. doi:10.1007/s11154-016-9347-2 [DOI] [PubMed] [Google Scholar]
  28. Takii M., Uchigata Y., Tokunaga S., Amemiya N., Kinukawa N., Nozaki T., Iwamoto Y., Kubo C. (2008). The duration of severe insulin restriction is the factor most closely associated with the microvascular complications of type 1 diabetic females with clinical eating disorders. The International Journal of Eating Disorders, 41(3), 259–264. doi:10.1002/eat.20498 [DOI] [PubMed] [Google Scholar]
  29. Tracy E., Berg C. A., Baker A. C., Mello D., Litchman M. L., Wiebe D. J. (2019). Health-risk behaviors and type 1 diabetes outcomes in the transition from late adolescence to early emerging adulthood. Children’s Health Care: Journal of the Association for the Care of Children’s Health, 48(3), 285–300. doi:10.1080/02739615.2018.1531758 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Weinberg A., Klonsky E. D. (2009). Measurement of emotion dysregulation in adolescents. Psychological Assessment, 21(4), 616–621. doi:10.1037/a0016669 [DOI] [PubMed] [Google Scholar]
  31. Wisting L., Frøisland D. H., Skrivarhaug T., Dahl-Jørgensen K., Rø Ø. (2013). Disturbed eating behavior and omission of insulin in adolescents receiving intensified insulin treatment: A nationwide population-based study. Diabetes Care, 36(11), 3382–3387. doi:10.2337/dc13-0431 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Young-Hyman D. L., Peterson C. M., Fischer S., Markowitz J. T., Muir A. B., Laffel L. M. (2016). Depressive symptoms, emotion dysregulation, and bulimic symptoms in youth with type 1 diabetes: Varying interactions at diagnosis and during transition to insulin pump therapy. Journal of Diabetes Science and Technology, 10(4), 845–851. doi:10.1177%2F1932296816645118 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

jsab042_Supplementary_Data
Click here for additional data file. (33.6KB, docx)

Articles from Journal of Pediatric Psychology are provided here courtesy of Oxford University Press

RESOURCES