Continuous Glucose Monitoring Associated With Less Diabetes-Specific Emotional Distress and Lower A1c Among Adolescents With Type 1 Diabetes

Abstract

Background:

Psychosocial impact research of continuous glucose monitoring (CGM) and continuous subcutaneous insulin infusion (CSII) among adolescents with type 1 diabetes (T1D) is limited. The present study assesses associations between diabetes technology use on adolescent- and parent-perceived diabetes-specific distress and A1c.

Method:

Adolescents with T1D and parents (N = 1040; primarily mothers) completed measures of diabetes distress. Adolescents were categorized by technology use: CGM Alone, CSII Alone, CGM+CSII, or No Technology. ANOVA, regression, and Cohen’s d were used for group comparisons on measures of diabetes distress and A1c. Analyses also compared groups on clinical elevations of distress.

Results:

CGM use was associated with less adolescent distress compared to No Technology (d = 0.59), CGM+CSII (d = 0.26), and CSII Alone (d = 0.29). Results were similar but with smaller effect size for parent-reported distress, although CGM+CSII showed equivocal association with parent distress compared to No Technology (d = 0.18). CGM Alone was associated with lower A1c compared to No Technology (d = 0.48), to CSII Alone (d = 0.37), and was comparable to CGM+CSII (d = 0.03). CGM+CSII conferred advantage over CSII Alone (d = 0.34). Clinical elevation of distress was associated with not using any technology particularly for adolescents.

Conclusions:

Technology use is associated with lower adolescent distress than lower parent distress. CGM Alone is associated with lower adolescent and parent distress than CSII or CGM+CSII. This appears to be clinically meaningful based on cut scores for measures. CGM is associated with lower A1c independent of being used alone or with CSII.

Keywords

adolescent distress glucose sensor insulin pump type 1 diabetes

The American Diabetes Association recommends a target A1c of 7.5% for youth with type 1 diabetes (T1D).¹ According to the T1D Exchange, the average A1c for adolescents in the United States is 9.0%. Only 16% obtain an A1c less than 7.5%.² Technologies, such as continuous subcutaneous insulin infusion (CSII) and continuous glucose monitoring (CGM), may improve glycemic control.

Recent studies have identified a small, but clinically significant reduction in A1c with CSII use. Pooled analyses from three national registries found lower mean A1c associated with CSII versus multiple daily injections (MDI) among youth ages 8-15 years.³ Similar findings were identified in one large study⁴ and in two longitudinal, case-control studies, showing a decrease in A1c among youth using CSII compared with MDI over a 6-month⁵ and a 7-year period.⁶ In addition, results of a meta-analysis of randomized controlled trials comparing mode of insulin delivery on change in A1c found an overall reduction of 0.32% in youth using CSII versus MDI.⁷ Regardless of the positive associations found with CSII use, only 58% of adolescents in the United States use CSII.² Demographic factors associated with CSII use (compared to MDI) include being female, being Caucasian, having caregivers with college degrees, having private insurance, and living with two caregivers.^3,8-10

Compared to CSII, CGM uptake among adolescents is much lower with only 5% of those participating in the T1D Exchange using CGM,² although other registries show 16-17% use.¹¹ Higher socioeconomic status, having private health insurance, longer duration of T1D, and CSII use are all associated with CGM uptake.¹² Near-daily CGM use is associated with higher quality of life, more frequent glucose checks, and decreased risk for omitting insulin.^13-15

CGM use has a clinically significant impact on glycemic control when used ≥6 days/week.^14-18 Youth who consistently used CGM showed significantly lower A1c initially, and again 12 months later, compared to those who used CGM less often.^16,19 This finding holds across developmental stages as the same reduction in A1c associated with near-daily use was found for both children and adolescents.¹⁴ In large registries of youth with T1D (N > 20 000), CGM use, regardless of insulin delivery method, is associated with lower A1c.¹¹ While several studies suggest that consistent CGM use is associated with improved glycemic control, other studies do not show such improvement. Two studies examining large national samples of adolescents did not find a significant difference in mean A1c between consistent CGM users and inconsistent users or nonusers.^12,20 Although the association between CGM use and A1c is mixed, there is an association between adolescent CGM use and both fewer episodes of hypoglycemia and increased time in range.²¹

Technology use in youth may be associated with lower A1c^21,22 although the magnitude of this association needs further evaluation given recent technological advancements. Technology use may be associated with psychological benefits, or it may be associated with increased distress. A greater understanding of how diabetes technology is associated with diabetes-related psychological distress is needed.

Diabetes-specific emotional distress refers to the typical negative emotions that individuals experience in relation to the psychological burden and time involved with ongoing daily diabetes care diabetes.²³ A majority of adolescents and parents report experiencing elevated levels of distress, with approximately one-third reporting moderate to high levels.^24-26 Diabetes distress has been linked with poorer glycemic control,^24,27-29 increased depressive symptoms, poorer quality of life, and poorer self-care in youth with T1D.^30-33

Limited studies have reported on associations between diabetes distress and technology use. Three studies examining CSII use showed associations that were not significant between CSII use and diabetes distress.^32,34,35 Studies assessing the association between CGM use and diabetes distress showed no difference in distress among parents,³⁶ youth,¹⁵ or adults.³⁷ Furthermore, Naranjo and colleagues compared both CSII and CGM use (alone and in combination) in adults with T1D and found no differences in using CSII or CGM, alone or together, with respect to diabetes distress.³⁸ To our knowledge, no studies directly comparing the associations of CSII and CGM technologies with diabetes distress among youth with T1D exist.

The purpose of the present study was to compare type of technology used (CGM Alone, CSII Alone, CGM+CSII, or no technology use) on diabetes distress and A1c in adolescents. Previous studies have shown no clear associations of CGM or CSII use with diabetes distress. Therefore, it is unclear how using both technologies compare to using either one alone or not using technology at all. Current analyses examined the association between type of technology use and adolescent- and parent-reported diabetes distress. We hypothesized that using both technologies or using one technology would be associated with lower A1c than using no technology. It was unclear how using one technology would compare to each other or to both technologies with regard to distress; therefore, investigating these comparisons was exploratory.

Methods

Participants

Data stem from a study investigating psychosocial outcomes of youth attending diabetes camps throughout the United States. All study procedures were approved by the Institutional Review Board at Ann & Robert H. Lurie Children’s Hospital of Chicago. Adolescents aged 12-18 years (N = 1216) and their parents were recruited to complete participant-specific questionnaires about diabetes distress. Links to the online survey were sent via email prior to the start of camp.

Measures

Diabetes-Specific Emotional Distress

Adolescents (PAID-T) and parents (P-PAID-T) completed a 14-item and 15-item scale, respectively, measuring their own diabetes-specific emotional distress.^30,39 Adolescents and their parents rate how much of a problem (1 = not a problem to 6 = serious problem) diabetes management and diabetes-specific emotional stress are in their everyday lives. Cut scores of ≥ 44 for the PAID-T and ≥ 54 for the P-PAID-T have recently been empirically derived.³⁹ Both measures demonstrate good psychometric properties. Internal reliability for the current study was α = .93 for both measures.

Hemoglobin A1c

For this study, hemoglobin A1c was obtained by parent-report and based on youth’s last outpatient endocrinology visit. Because of the web-based nature of this national study, the investigators were unable to obtain laboratory-based A1c results for each youth.

Demographics

Parents reported on their child’s age, race/ethnicity, family income, and diabetes variables including use of CGM and/or CSII. Two questions were included on the parent survey which specifically asked (1) if their adolescent uses CGM or a standard blood glucose meter most of the time and (2) if their adolescent receives insulin via a pump or insulin pen/syringe most of the time.

Data Analysis

Adolescents were categorized into one of four technology use groups: CGM Alone, CSII Alone, CGM+CSII, and no technology use (ie, using a glucose meter and administering insulin through a pen/syringe). Analysis of variance (ANOVA) with Games-Howell post hoc tests for individual group differences was used to assess the mean differences among technology groups on adolescent- and parent-reported diabetes distress and hemoglobin A1c. Games-Howell tests were chosen as they do not assume equal variance or sample sizes between groups. Cohen’s d effect sizes were calculated between technology groups to measure magnitude of mean difference on outcomes. Effect sizes are less dependent on sample size than null hypothesis tests and may offer less-biased estimates of population differences. Interpretation of effect sizes per recommended guidelines is as follows: d < 0.2 is negligible, 0.2 ≤ d < 0.5 is small, 0.5 ≤ d < 0.8 is medium, and d ≥ 0.8 is large.⁴⁰ To further examine differences between technology groups on distress, chi square analyses were completed to determine differences between groups on percentage of participants scoring above clinical cutoff scores on the PAID-T and P-PAID-T respectively. Finally, demographics that were significantly associated with each primary outcome were entered into regression models along with dummy-coded technology groups. Regressions were used to determine the robustness of the comparisons between technology groups when accounting for other significant covariates. Cohen’s d values were calculated for regression comparisons.⁴¹

Results

Sample Characteristics

Participants consented were 1216 adolescents and their parents; however, 176 adolescents were not categorized into technology groups due to missing parent-reported data either on how adolescents administer insulin or check blood sugars. Consequently, the study sample total size was 1040. The uncategorized sample (n = 176) had a higher proportion of minority youth (25.3% versus 9.2%, P < .001) and higher mean A1c (8.75 versus 7.89, P = .001). Of those who were categorized (N = 1040), 19 (1.8%) used CGM Alone, 664 (63.8%) used CSII Alone, 112 (10.8%) used CGM+CSII, and 245 (23.6%) did not use technology.

Table 1 displays the demographics of the total sample and each technology use group. Differences in gender distribution were noted as more girls used CSII alone than no technology. CSII alone was more common in adolescents with higher family income than those not using technology. Tables 2 and 3 present the descriptive statistics for adolescent- and parent-reported diabetes distress and for A1c.

Table 1.

Demographic Characteristics of Sample and Differences Among Technology Groups.

	Total (N = 1040)		CGM Alone (n = 19)		CSII Alone (n = 664)		CGM+CSII (n = 112)		No Technology (n = 245)
	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD
Adolescent age	14.38	1.51	14.41_a	1.82	14.36_a	1.47	14.27_a	1.48	14.50_a	1.58
Adolescent age	n	%	n	%	n	%	n	%	n	%
Gender: female	587	56.7	9	47.4_ab	400	60.5_b	58	51.8_ab	120	49.2_a
Youth race
White	901	90.8	15	83.3_a	588	92.0_a	102	94.4_a	196	86.3_a
Black/African American	37	3.7	2	11.1_b	25	3.9_ab	1	0.9_a	9	4.0_a
Latino	44	4.2	1	5.6_ab	18	2.8_b	5	4.6_ab	20	8.8_a
Native American	6	0.6	0	0_a	4	0.6_a	0	0_a	2	0.9_a
Asian	4	0.4	0	0_a	4	0.6_a	0	0_a	0	0_a
Caregiver relationship
Mother	922	89.4	17	89.5_a	599	90.5_a	99	89.2_a	207	86.6_a
Father	95	9.2	0	0_a	56	8.4_a	12	10.8_a	27	11.3_a
Grandmother	11	1.1	0	0_a	7	1.1_a	0	0_a	4	1.7_a
Grandfather	1	0.1	1	5.3_a	0	0_a	0	0_a	0	0_a
Other guardian	2	0.2	1	5.3_b	0	0_a	0	0_ab	1	0.4_ab
Family income
<$25 000	80	9.1	1	7.7_ab	39	6.9_b	6	6.7_ab	34	16.2_a
$26-50 000	154	17.5	3	23.1_ab	92	16.2_b	6	6.7_b	53	25.2_a
$51-75 000	127	14.4	0	0_a	82	14.5_a	13	14.6_a	32	15.2_a
$76-100 000	140	15.9	1	7.7_a	98	17.3_a	11	12.4_a	30	14.3_a
$101-125 000	135	15.4	1	7.7_a	94	16.6_a	13	14.6_a	27	12.9_a
$126-150 000	75	8.5	2	15.4_ab	48	8.5_ab	14	15.7_b	11	5.2_a
$151-175 000	55	6.3	0	0_a	40	7.1_a	7	7.9_a	8	3.8_a
>$175 000	113	12.9	5	38.5_b	74	13.1_ac	19	21.3_bc	15	7.1_a

Different subscripts across row indicate statistically significant differences between groups (P < .05).

Table 2.

Correlations, Means, and Standard Deviations of Continuous Measures.

	1	2	3	4
1. Age	14.38 (1.51)
2. A1c	.089	7.89 (1.50)
3. P-PAID-T	−.025	.389	48.35 (15.88)
4. PAID-T	.084	.298	.468	40.51 (16.16)

Significant correlations (P < .05) are in bold. Overall sample mean (SD) is presented on the diagonal for each variable.

Table 3.

Means and Standard Deviations of Diabetes Distress (PAID-T and P-PAID-T scales) and A1c by Technology Use Group.

Technology group	PAID-T		P-PAID-T		A1c
Technology group	Mean	SD	Mean	SD	Mean	SD
CGM Alone	34.05	17.11	43.37	15.92	7.36	1.29
CSII Alone	39.63	15.72	47.47	15.24	7.86	1.38
CGM+CSII	38.56	16.24	48.20	15.37	7.40	1.31
No Technology	44.42	16.67	48.35	15.88	8.23	1.83

Demographic Differences in Diabetes Distress and A1c

Adolescent distress was positively associated with age (P = .016) and being female (P < .001) and negatively associated with family income (P < .001). Parent distress was positively associated with racial minority status (P < .001) and negatively associated with family income (P < .001). A1c was positively associated with age (P = .006) and racial minority status (P < .001) and negatively associated with family income (P < .001). Higher A1c was significantly associated with higher levels of both adolescent-and parent-reported distress (r = .30, P < .001 and r = .39, P < .001 respectively).

Differences in Diabetes Distress and A1c Among Technology Groups

Table 4 shows the comparisons between technology use groups including statistical significance of Games-Howell post hoc tests and Cohen’s d effect size on distress and A1c. Comparisons using the CGM Alone group were likely underpowered to detect statistical difference due to low sample size; effect sizes are likely more representative of population differences. CGM Alone had the lowest mean adolescent-reported distress and the No Technology group had the highest. The remaining groups were in between and similar to each other (d = 0.07, P = .934) (see Tables 3 and 4). CGM Alone was not significantly different than the other three technology groups on adolescent-reported diabetes distress, likely due to being statistically underpowered given the small sample size in this group. Despite not being statistically significant, CGM Alone demonstrated small to moderate effect size differences with the other groups. CSII and CGM+CSII groups demonstrated significant, small effect size differences from the No Technology group.

Table 4.

Cohen’s d Effect Sizes and Associated Statistics and ANOVA Games-Howell Post Hoc Test Group Comparison P-values.

Outcome		d	SE of d	95% CI	P-value
PAID-T	Reference: No Technology
	CGM Alone	−0.62	0.27	(–1.15, –0.09)	.148
	CSII Alone	−0.30	0.08	(–0.46, –0.14)	.003
	CGM+CSII	−0.35	0.13	(–0.60, –0.11 )	.024
	Reference: CGM+CSII
	CGM Alone	−0.27	0.28	(–0.82, 0.27)	.776
	CSII Alone	0.07	0.11	(–0.15, 0.29)	.934
	Reference: CSII Alone
	CGM Alone	−0.35	0.26	(–0.87, 0.16)	.608
P-PAID-T	Reference: No Technology
	CGM Alone	−0.45	0.24	(–0.92, 0.02)	.201
	CSII Alone	−0.24	0.08	(–0.38, –0.09)	.018
	CGM+CSII	−0.18	0.11	(–0.40, 0.05)	.363
	Reference: CGM+CSII
	CGM Alone	−0.31	0.25	(–0.80, 0.18)	.615
	CSII Alone	−0.05	0.10	(–0.25, 0.15)	.966
	Reference: CSII Alone
	CGM Alone	−0.27	0.23	(–0.72, 0.19)	.689
A1c	Reference: No Technology
	CGM Alone	−0.48	0.26	(–0.99, 0.03)	.085
	CSII Alone	−0.24	0.08	(–0.39, –0.09)	.029
	CGM+CSII	−0.49	0.12	(–0.72, –0.26)	<.001
	Reference: CGM+CSII
	CGM Alone	−0.03	0.27	(–0.56, 0.49)	.999
	CSII Alone	0.34	0.10	(0.13, 0.54)	.005
	Reference: CSII Alone
	CGM Alone	−0.37	0.25	(–0.86, 0.13)	.436

Reference groups refer to the group being compared to the others below it in the calculation of d. For each comparison, negative effect sizes signify a larger value for the reference group, whereas positive effect sizes signify a smaller value for the reference group.

Technology group means on parent-reported distress showed the same pattern as adolescent-reported distress with CGM Alone being the lowest, No Technology being the highest, and the remaining groups in between and similar to each other (d = 0.05, P = .966). Effect sizes for parent-reported distress were similar to adolescent-reported distress although most were smaller. CGM Alone was not statistically different from the other three technology groups for parent-reported distress; however, CGM Alone demonstrated small-moderate effect size differences with the other groups. CGM+CSII and No Technology had essentially no statistical difference in means (d = 0.18, P = .363), while CSII Alone conferred statistically significant and small effect (d = 0.24, P = .018) relative to No Technology.

CGM Alone and CGM+CSII demonstrated the lowest mean A1c (effect size between these groups was negligible, d = 0.03, P = .999), the No Technology group had the highest mean A1c, and the CSII Alone group mean was between the CGM groups and the No Technology group. CGM Alone was not statistically different from CSII Alone or No Technology; however, CGM Alone demonstrated small to moderate effect size differences from these groups. Use of CGM was not influenced by mode of insulin administration as effect sizes involving CGM+CSII and CGM Alone compared to CSII Alone and No Technology were similar.

Differences in Clinical Elevation of Distress among Technology Groups

Overall, 41.0% of adolescents in this sample scored above the empirically derived clinical cutoff score of the PAID-T (≥44) with 33.3% scoring above in the CGM Alone group, 37.9% scoring above in the CGM+CSII group, 37.5% scoring above in the CSII Alone group, and 52.8% scoring above in the No Technology group. There was a statistically significantly difference among technology groups, χ²(3) = 14.88, P = .002, with CGM+CSII and CSII Alone showing a lower proportion of adolescents being clinically elevated compared to No Technology. CGM alone had the lowest proportion of clinical elevation among the groups but statistical significance from No Technology was likely not achieved due to low sample size.

Among parents, 37.0% of parents scored above the clinical cutoff score of the P-PAID-T (≥54) with 31.6% scoring above in the CGM Alone group, 36.9% scoring above in the CGM+CSII group, 34.8% scoring above in the CSII Alone group, and 43.6% scoring above in the No Technology group. There were no statistically significant differences between technology groups with respect to clinical elevation, χ²(3) = 6.20, P = .102.

Technology Group Differences on A1c and Distress Covarying for Demographic Factors

Technology group, adolescent age, adolescent gender, and family income were entered into a multiple regression to predict adolescent-reported diabetes distress (Table 5). The overall model was statistically significant, F(6, 673) = 11.65, P < .001, adjusted R² = .086. Using any technology was associated with significantly lower adolescent distress than using no technology with small-moderate effect size. There were no statistical differences among CGM Alone, CSII Alone, or CGM+CSII groups; however, CGM alone had a small effect difference, with lower adolescent distress, compared to CGM+CSII and CSII Alone.

Table 5.

Regression Analyses Presenting Technology Use Group Comparisons, While Controlling for Demographic Variables, for Diabetes Distress (PAID-T and P-PAID-T) and A1c.

	A1c				PAID-T				P-PAID-T
	b	P-value	d	95% CI	b	P-value	d	95% CI	b	P-value	d	95% CI
Age	0.13	<.001			1.22	.002			—	—
Gender (female)	—	—			4.36	<.001			—	—
Racial minority	0.43	.015			—	—			4.22	.023
Annual income	−0.13	<.001			−1.55	<.001			−1.13	<.001
Reference: No Technology
Intercept	8.06	<.001			36.91	<.001			50.11	<.001
CGM Alone	−0.90	.039	−0.62	(–1.13, –0.11)	−9.98	.048	−0.65	(–1.18, –0.12)	−5.67	.207	−0.36	(–0.83, 0.11)
CSII Alone	−0.16	.200	−0.11	(–0.26, 0.04)	−4.12	.005	−0.27	(–0.43, –0.10)	−2.55	.054	−0.16	(–0.31, –0.02)
CGM+CSII	−0.47	.016	−0.33	(–0.55, –0.10)	−4.42	.047	−0.29	(–0.53, –0.04)	−0.18	.929	−0.01	(–0.24, 0.21)
Reference: CGM+CSII
Intercept	7.59	<.001			32.48	<.001			49.93	<.001
CGM Alone	−0.43	.341	−0.29	(–0.82, 0.23)	−5.56	.286	−0.36	(–0.91, 0.19)	−5.48	.237	−0.35	(–0.84, 0.14)
CSII Alone	0.31	.073	0.21	(0.01, 0.42)	−0.30	.880	0.02	(–0.20, 0.24)	−2.37	.190	−0.15	(–0.35, 0.05)
Reference: CSII Alone
Intercept	7.90	<.001			32.78	<.001			47.56	<.001
CGM Alone	−0.74	.084	−0.51	(–1.00, –0.01)	−5.86	.236	−0.38	(–0.89, 0.13)	−3.11	.478	−0.20	(–0.66, 0.26)

Racial minority variable was dummy coded with being white as the reference group and a being a racial minority coded as one. Age and annual income were mean-centered. — indicates that variable was not included in the model as it was not significantly associated with outcome initially. Three regressions were run varying the reference group for the dummy-coded technology use category variable to make all group comparisons. All results are presented above including Cohen’s d effect sizes.

Technology group, adolescent minority status, and family income were entered to predict parent-reported diabetes distress (Table 5). The overall model was statistically significant, F(5, 831) = 7.67, P < .001, adjusted R² = .035. There were no statistically significant comparisons between groups. CGM alone showed a small effect (but not significant) compared to No Technology, CSII Alone, and CGM+CSII with lower reported parent distress.

Technology group, adolescent age, racial minority status, and family annual income were entered to predict A1c (Table 5). The overall model was statistically significant, F(6, 772) = 11.28, P < .001, adjusted R² = .073. Using CGM Alone or CGM+CSII compared to No Technology was statistically different, with CGM-use groups showing small-moderately lower A1c. CGM alone demonstrated moderately lower, albeit not significant, A1c than CSII alone.

Discussion

The present study compared T1D technology use groups (CGM Alone, CSII Alone, CGM+CSII, and No Technology) on adolescent- and parent-perceived diabetes-specific emotional distress and A1c. CGM use was associated with lower distress for both adolescents and their parents compared to those who did not use technology, to those who used with CSII, and to those who used CSII alone. Notably, smaller effect sizes were found for parent distress. Lower A1c was associated with CGM or CGM+CSII use. In terms of clinical elevations, a relatively large proportion of parents and adolescents noted elevated diabetes distress based on empirically derived cutoff scores. Not using technology was associated with greatest elevation in distress, primarily for adolescent distress. Conclusions from these findings support the association between use of CGM and lower A1c as well as decreased diabetes distress particularly for adolescents.

The small-medium effect size differences for CGM Alone compared to No Technology corresponds to about a half standard deviation between groups for both adolescent and parent-reported diabetes distress. Given that distress is associated with clinical outcomes such as lower quality of life, increased depressive symptoms, and lower adherence, such a difference in distress is clinically meaningful. It appears that use of CGM by itself is associated with lower distress and use of CSII with or without CGM increases distress; however, these differences are small for adolescents and negligible for parents after controlling for demographic correlates of distress. Furthermore, analyses examining differences between technology groups on clinical diabetes distress cutoff scores demonstrated lower rates of participants scoring at or above these benchmarks when adolescents are using CGM, CSII, or CGM+CSII providing further evidence of clinically meaningful association.

About a 0.5% lower A1c is associated with CGM use over use of CSII alone, and this difference is even larger when comparing CGM use to use of no technology; findings of the current study do not support a significant additive difference in A1c associated with using CGM+CSII. Given that an absolute increase of 1% in A1c standard deviation is linked with increased risks for retinopathy and neuropathy,⁴² current results are clinically significant. These findings must be considered in light of the cross-sectional design of this study. Interpretation of these findings are bidirectional meaning that CGM use may improve diabetes distress and A1c or that lower diabetes distress and lower A1c predicts CGM use. Causality cannot be inferred from the current data.

Despite associations of CGM technology use with lower distress and A1c shown in the current study, past studies suggest either no psychosocial improvements or increases in psychological burden.^15,43 With every medical technology, consideration of both advantages and disadvantages of device use and of individual preferences is necessary. Previous burdens related to using CGM include wearability, data overload/inaccuracy, and alarm fatigue.^44,45 Despite these findings, many adolescents perceive general benefits of CGM including making diabetes management easier, having better detection of hypoglycemia, and improved glycemic control.^44,46,47 Studies comparing CGM to CGM+CSII have demonstrated improved A1c similar to the associations in the current study although statistical significance was not achieved.^48,49 Experiences of technology use, both positive and negative, that may mediate the relationship between CGM use, distress, and glycemic control should be more fully investigated in future studies to elucidate links between CGM use alone and CGM+CSII on distress and A1c. Development of a model incorporating technology type, diabetes distress, and glycemic control along with these mediators would be beneficial to understanding the psychosocial and metabolic impact of diabetes technology.

Limitations of the current study should be noted. First, the sample was composed of mostly Caucasian participants from higher income families which is not representative of all youth with T1D. Second, A1c was based on parent report and not lab results. Because this was a national, web-based survey, investigators did not have the resources to collect lab data or to procure medical chart data. Research has shown a high correlation between adult self-reported A1c and lab-based A1c (r = .84).²⁹ In addition, the authors examined the relationship between parent-reported A1c values and laboratory-based values from another dataset of adolescents with T1D⁵⁰ and found a correlation of .84. These data provide support that parent-reported A1c values demonstrate validity. Third, this study did not include other measures of glycemic variability beyond A1c which limits the generalizability of our findings. Future work examining associations of technology use with other metrics of glycemic variability including other biomarkers and various metrics calculated from blood glucose meters would be beneficial.⁵¹ Fourth, this study was cross-sectional and therefore unable to make causal attributions of technology use in relation to diabetes distress or A1c. Fifth, there was a small number of participants in the CGM Alone technology use group, likely impacting the statistical power of analyses involving this group as standard errors are typically larger in small samples. Future studies investigating CGM without CSII on psychosocial and medical outcomes in a larger sample would contribute to the literature on diabetes technology. Sixth, based on available data, we were unable to ascertain provider perceptions of use of technology and factors that may influence prescribing CGM and CSII to adolescents with T1D. There may be underlying latent factors that influence prescription rates not accounted for in this study. Seventh, the extent of CGM use (eg, how many days per week) was not assessed in this study, precluding the assessment of dose effects of CGM Alone and CSII on diabetes distress and A1c. Finally, due to missing data, we were unable to categorize 176 (14.5%) adolescents into a technology group. There was a higher percentage of minority youth and higher mean A1c in the uncategorized group as compared to the categorized group. Research on technology use and associations with barriers to use and with psychosocial outcomes in youth with T1D from underrepresented populations would benefit both research and clinical practice. Strengths of this study include a large sample from 44 diabetes camps across the United States. Moreover, this is the first study to examine differences among diabetes technology use groups in relation to diabetes distress and A1c in a sample of adolescents.

Conclusion

In this study, use of CGM, alone or in combination with CSII, was found to be associated with lower adolescent-perceived diabetes distress and lower A1c. Ongoing research examining the causal impact of CGM on diabetes distress and A1c is essential.

Footnotes

Abbreviations

ANOVA, analysis of variance; CGM, continuous glucose monitoring; CI, confidence interval; CSII, continuous subcutaneous insulin infusion; MDI, multiple daily injections; PAID, Problem Areas in Diabetes; SE, standard error; T1D, type 1 diabetes.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported in part by a grant from the Helmsley Charitable Trust foundation.

References

American Diabetes Association. Standards of medical care in diabetes—2016 abridged for primary care providers. Clin Diabetes. 2016;34:3-21.

Miller

Foster

Beck

et al . Current state of type 1 diabetes treatment in the US: updated data from the T1D Exchange clinic registry. Diabetes Care. 2015;38:971-978.

Sherr

Hermann

Campbell

et al . Use of insulin pump therapy in children and adolescents with type 1 diabetes and its impact on metabolic control: comparison of results from three large, transatlantic paediatric registries. Diabetologia. 2016;59:87-91.

Szypowska

Schwandt

Svensson

et al . Insulin pump therapy in children with type 1 diabetes: analysis of data from the SWEET registry. Pediatr Diabetes. 2016;17(suppl 23):38-45.

Brorsson

Viklund

Örtqvist

Lindholm Olinder

Does treatment with an insulin pump improve glycaemic control in children and adolescents with type 1 diabetes? A retrospective case-control study. Pediatr Diabetes. 2015;16:546-553.

Johnson

Cooper

Jones

Davis

EA.

Long-term outcome of insulin pump therapy in children with type 1 diabetes assessed in a large population-based case-control study. Diabetologia. 2013;56:2392-2400.

Benkhadra

Alahdab

Tamhane

McCoy

Prokop

Murad

MH.

Continuous subcutaneous insulin infusion versus multiple daily injections in individuals with type 1 diabetes: a systematic review and meta-analysis. Endocrine. 2017;55:77-84.

Wong

Dolan

Yang

Hood

KK.

Insulin pump use and glycemic control in adolescents with type 1 diabetes: predictors of change in method of insulin delivery across two years. Pediatr Diabetes. 2015;16:592-599.

Willi

Miller

DiMeglio

et al . Racial-ethnic disparities in management and outcomes among children with type 1 diabetes. Pediatrics. 2015;135:424-434.

10.

McKnight

Wild

Lamb

MJE

et al . Glycaemic control of type 1 diabetes in clinical practice early in the 21st century: an international comparison. Diabet Med. 2015;32:1036-1050.

11.

DeSalvo

Miller

Hermann

et al . Continuous glucose monitoring (CGM) and glycemic control among youth with type 1 diabetes (T1D): international comparison from the T1D exchange (T1DX) and the DPV initiative 2017. Pediatr Diabetes. 2017;18:36.

12.

Wong

Foster

Maahs

et al . Real-time continuous glucose monitoring among participants in the T1D exchange clinic registry. Diabetes Care. 2014;37:2702-2709.

13.

Diabetes Research in Children Network (DirecNet) Study Group. Psychological aspects of continuous glucose monitoring in pediatric type 1 diabetes. Pediatr Diabetes. 2006;7:32-38.

14.

Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group. Factors predictive of use and of benefit from continuous glucose monitoring in type 1 diabetes. Diabetes Care. 2009;32:1947-1953.

15.

Giani

Snelgrove

Volkening

Laffel

LM.

Continuous glucose monitoring (CGM) adherence in youth with type 1 diabetes: associations with biomedical and psychosocial variables. J Diabetes Sci Technol. 2017;11:476-483.

16.

Chase

Beck

Xing

et al . Continuous glucose monitoring in youth with type 1 diabetes: 12-month follow-up of the Juvenile Diabetes Research Foundation continuous glucose monitoring randomized trial. Diabetes Technol Ther. 2010;12:507-515.

17.

Deiss

Bolinder

Riveline

J-P

et al . Improved glycemic control in poorly controlled patients with type 1 diabetes using real-time continuous glucose monitoring. Diabetes Care. 2006;29:2730-2732.

18.

Diabetes Research in Children Network (DirecNet) Study Group. Continuous glucose monitoring in children with type 1 diabetes. J Pediatr. 2007;151:388-393.e1-2.

19.

Rachmiel

Landau

Boaz

et al . The use of continuous glucose monitoring systems in a pediatric population with type 1 diabetes mellitus in real-life settings: the AWeSoMe Study Group experience. Acta Diabetol. 2015;52:323-329.

20.

Ludwig-Seibold

Holder

Rami

Raile

Heidtmann

Holl

RW.

Continuous glucose monitoring in children, adolescents, and adults with type 1 diabetes mellitus: analysis from the prospective DPV diabetes documentation and quality management system from Germany and Austria. Pediatr Diabetes. 2012;13:12-14.

21.

Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group. The effect of continuous glucose monitoring in well-controlled type 1 diabetes. Diabetes Care. 2009;32:1378-1383.

22.

Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group. Effectiveness of continuous glucose monitoring in a clinical care environment. Diabetes Care. 2010;33:17-22.

23.

Fisher

Gonzalez

Polonsky

WH.

The confusing tale of depression and distress in patients with diabetes: a call for greater clarity and precision. Diabet Med. 2014;31:764-772.

24.

Berlin

Rabideau

Hains

AA.

Empirically derived patterns of perceived stress among youth with type 1 diabetes and relationships to metabolic control. J Pediatr Psychol. 2012;37:990-998.

25.

Hagger

Hendrieckx

Sturt

Skinner

Speight

Diabetes distress among adolescents with type 1 diabetes: a systematic review. Curr Diab Rep. 2016;16:9.

26.

Whittemore

Jaser

Chao

Jang

Grey

Psychological experience of parents of children with type 1 diabetes: a systematic mixed-studies review. Diabetes Educ. 2012;38:562-579.

27.

Delamater

Patiño-Fernández

Smith

Bubb

Measurement of diabetes stress in older children and adolescents with type 1 diabetes mellitus. Pediatr Diabetes. 2013;14:50-56.

28.

Markowitz

Volkening

Butler

Antisdel-Lomaglio

Anderson

Laffel

LMB

. Re-examining a measure of diabetes-related burden in parents of young people with type 1 diabetes: The Problem Areas in Diabetes Survey–Parent Revised version (PAID-PR). Diabet Med. 2012;29:526-530.

29.

Hessler

Fisher

Polonsky

Johnson

Understanding the areas and correlates of diabetes-related distress in parents of teens with type 1 diabetes. J Pediatr Psychol. 2016;41:750-758.

30.

Weissberg-Benchell

Antisdel-Lomaglio

Diabetes-specific emotional distress among adolescents: feasibility, reliability, and validity of the Problem Areas in Diabetes–Teen Version. Pediatr Diabetes. 2011;12:341-344.

31.

Franklin

MD.

The Relationship Between Psychosocial Factors, Self-Care Behaviors, and Metabolic Control in Adolescents With Type 1 Diabetes [doctoral dissertation]. Nashville, TN: Graduate School of Vanderbilt University; 2008.

32.

Boland

Grey

Mezger

Tamborlane

WV.

A summer vacation from diabetes: evidence from a clinical trial. Diabetes Educ. 1999;25:31-40.

33.

Grey

Boland

Sullivan-Bolyai

Tamborlane

WV.

Personal and family factors associated with quality of life in adolescents with diabetes. Diabetes Care. 1998;21:909-914.

34.

Forsander

Bøgelund

Haas

Samuelsson

Adolescent life with diabetes-gender matters for level of distress. Experiences from the national TODS study. Pediatr Diabetes. 2016;18:651-659.

35.

Lašaitė

Dobrovolskienė

Danytė

et al . Diabetes distress in males and females with type 1 diabetes in adolescence and emerging adulthood. J Diabetes Complications. 2016;30:1500-1505.

36.

Markowitz

Pratt

Aggarwal

Volkening

Laffel

LM.

Psychosocial correlates of continuous glucose monitoring use in youth and adults with type 1 diabetes and parents of youth. Diabetes Technol Ther. 2012;14:523-526.

37.

Hermanns

Kulzer

Gulde

et al . Short-term effects on patient satisfaction of continuous glucose monitoring with the GlucoDay with real-time and retrospective access to glucose values: a crossover study. Diabetes Technol Ther. 2009;11:275-281.

38.

Naranjo

Tanenbaum

Iturralde

Hood

KK.

Diabetes technology: uptake, outcomes, barriers, and the intersection with distress. J Diabetes Sci Technol. 2016;10:852-858.

39.

Shapiro

Vesco

Weil

LEG

Evans

Hood

Weissberg-Benchell

. Psychometric properties of the Problem Areas in Diabetes: Teen and Parent of Teen Versions. J Pediatr Psychol. 2018;43:561-571.

40.

Cohen

Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Hillsdale, NJ: Erlbaum; 1988.

41.

Feingold

A regression framework for effect size assessments in longitudinal modeling of group differences. Rev Gen Psychol. 2013;17:111-121.

42.

Kilpatrick

Rigby

Atkin

SL.

A1c variability and the risk of microvascular complications in type 1 diabetes: data from the Diabetes Control and Complications Trial. Diabetes Care. 2008;31:2198-2202.

43.

Patton

Clements

MA.

Psychological reactions associated with continuous glucose monitoring in youth. J Diabetes Sci Technol. 2016;10:656-661.

44.

Pickup

Holloway

Samsi

Real-time continuous glucose monitoring in type 1 diabetes: a qualitative framework analysis of patient narratives. Diabetes Care. 2015;38:544-550.

45.

Polonsky

Hessler

Perceived accuracy in continuous glucose monitoring: understanding the impact on patients. J Diabetes Sci Technol. 2015;9:339-341.

46.

Tansey

Laffel

Cheng

et al . Satisfaction with continuous glucose monitoring in adults and youths with type 1 diabetes. Diabet Med. 2011;28:1118-1122.

47.

Hussain

Akle

Nagelkerke

Deeb

Comparative study on treatment satisfaction and health perception in children and adolescents with type 1 diabetes mellitus on multiple daily injection of insulin, insulin pump and sensor-augmented pump therapy. SAGE Open Med. 2017;5: 1-6.

48.

Kordonouri

Pankowska

Rami

et al . Sensor-augmented pump therapy from the diagnosis of childhood type 1 diabetes: results of the paediatric onset study (ONSET) after 12 months of treatment. Diabetologia. 2010;53:2487-2495.

49.

Battelino

Phillip

Bratina

Nimri

Oskarsson

Bolinder

Effect of continuous glucose monitoring on hypoglycemia in type 1 diabetes. Diabetes Care. 2011;34:795-800.

50.

Weissberg-Benchell

Rausch

Iturralde

Jedraszko

Hood

A randomized clinical trial aimed at preventing poor psychosocial and glycemic outcomes in teens with type 1 diabetes (T1D). Contemp Clin Trials. 2016;49:78-84.

51.

Wright

Hirsch

IB.

Metrics beyond hemoglobin A1c in diabetes management: time in range, hypoglycemia, and other parameters. Diabetes Technol Ther. 2017;19(suppl 2):s16-s26.