The Influence of Real-Time Continuous Glucose Monitoring on Psychosocial Outcomes in Insulin-Using Type 2 Diabetes

Abstract

Background:

To examine the impact of real-time continuous glucose monitoring (rtCGM) on psychosocial outcomes in adults with insulin-using type 2 diabetes (T2D).

Methods:

A total of 174 insulin-using adults with T2D completed questionnaires assessing diabetes distress, hypoglycemic confidence, hypoglycemic fear, device-related emotional burden, and device-related trust before and after a six-month trial of rtCGM. Hemoglobin A1c (HbA1c) was assessed at the same time points; impaired hypoglycemic awareness (IAH) was assessed at baseline. Change in psychosocial outcomes was examined with t tests, then modeled as a function of baseline HbA1c and IAH and simultaneous change in HbA1c in multiple regression analyses.

Results:

Respondents were predominantly male (57.5%) and non-Hispanic white (67.8%). Significant improvement over the trial was observed in hypoglycemic fear (P = .031), hypoglycemic confidence (P < .001), diabetes distress (P < .001), and device-related emotional burden (P < .001). Impaired hypoglycemic awareness at baseline predicted greater improvement in hypoglycemic fear (P = .002), hypoglycemic confidence (P = .003), diabetes distress (P = .013), and device-related emotional burden (P < .001). Higher baseline HbA1c was linked with greater improvement in hypoglycemic fear (P = .030); HbA1c change over the trial was positively associated with change in diabetes distress (P = .010) and device-related emotional burden (P = .003).

Conclusions:

Introduction of rtCGM in adults with insulin-using T2D was associated with significant improvements in diabetes-related psychosocial outcomes over six months. Gains were significantly greater among participants reporting IAH and those with higher HbA1c at baseline, thus providing the first evidence regarding which users might more likely benefit.

Keywords

continuous glucose monitoring psychosocial outcomes type 2 diabetes impaired hypoglycemic awareness

Introduction

Although real-world observational evidence and randomized controlled trial (RCT) findings point to positive psychosocial benefits of initiating continuous glucose monitoring among adults with type 1 diabetes (T1D),^1,2 it is not at all certain that individuals with type 2 diabetes (T2D) experience similar outcomes. Early RCTs found that use of both real-time continuous glucose monitoring (rtCGM)^3,4 and intermittently scanned CGM (isCGM)^5,6 led to significant glycemic benefits for T2D participants, but significant improvement was not observed in any of the key psychosocial dimensions, including diabetes distress, diabetes quality of life (QoL), hypoglycemic concerns, and overall well-being. Significant gains in treatment satisfaction, however, have been reported among patients with T2D in several published isCGM trials.^5-7

Several prospective observational studies^8-10 have included both T1D and T2D participants and found that isCGM initiation was linked to improvements in global measures of QoL and well-being. However, when the two groups were examined separately, significant psychosocial benefits were observed for T1D, but not T2D, participants.^8,10 Finally, in the most recent prospective cohort trial, the Landmark study, Gilbert et al¹¹ observed that diabetes distress and hypoglycemic concerns improved significantly over an initial 12-week period of rtCGM use in a mixed group of insulin users. Subgroup analyses revealed these significant positive changes were apparent in both the T1D and the insulin-using T2D subsets. While the Landmark findings are promising, they stand mostly alone and are inconsistent with prior CGM findings. Furthermore, the Landmark observation period was limited to a relatively brief, 12-week window and the T2D sample was small (n = 66). Therefore, aside from Landmark and the limited findings regarding treatment satisfaction gains, there is a lack of consistent empirical support, from RCTs or real-world observational studies, for the notion that CGM use contributes to positive psychosocial outcomes specifically in adults with T2D.

A recently published prospective observational study, the Continuous Glucose Monitoring for Control of Hypoglycemia (COACH) study, evaluated an rtCGM intervention (using the Dexcom™ G5 CGM System^®; Dexcom, Inc, San Diego, California) in a cohort of adults with T1D and insulin-using T2D. Subjects first completed a six-month control phase during which they based treatment decisions solely on self-monitoring of blood glucose (SMBG), then participated in a second six-month phase where rtCGM was introduced. Participants included 345 T1D adults and 174 insulin-using T2D adults. Mean frequency of severe hypoglycemic events decreased significantly from the SMBG phase to the rtCGM phase; hemoglobin A1c (HbA1c) remained stable over the SMBG phase, then dropped significantly over the rtCGM phase.¹² In addition, diabetes-specific psychosocial data—including measures of diabetes distress, hypoglycemic concerns, and satisfaction with one’s glucose monitoring system—were collected before and after the rtCGM phase, but have not yet been examined.

Investigation of these data in the T2D subset of COACH is an ideal opportunity to further elucidate the potential impact of CGM on psychosocial outcomes. Technological advances have given way to new generations of CGM systems that have not yet been evaluated for psychosocial benefit among adults with T2D, apart from the Landmark study. Given the promising results from Landmark, we hypothesized that positive psychosocial changes would emerge over the six-month rtCGM period among this T2D group. In addition, we expected these gains to be greatest among those who evidenced greater glycemic problems (specifically, elevated HbA1c and/or impaired hypoglycemic awareness [IAH]) at baseline, who were therefore more likely to benefit from the intervention. We also tested whether improvements in glycemic problems over the course of the rtCGM period were significantly predictive of simultaneous psychosocial gains over the same period.

Participants and Methods

COACH was a 12-month, prospective observational study conducted at 19 sites across the United States. Based in part on recommendations from an FDA advisory panel, the study was designed to see whether nonadjunctive, routine use of rtCGM in insulin-using adults affected the safety of diabetes management. For the complete study protocol, see the original trial results publication.¹² Relevant aspects of the protocol are summarized below.

Participants

The major inclusion criteria were broad: age ≥21 years, T1D or insulin-using T2D (established treatment three or more months), and any baseline HbA1c value. As described earlier, subjects first completed a six-month SMBG-only control phase, followed by a six-month intervention phase where rtCGM (Dexcom™ G5 CGM System^®) became the key means for making all treatment decisions. Diabetes management and device training were conducted as per the usual approach at each study site.

Procedures

Study visits occurred at months 0, 6, and 12, during which HbA1c values were obtained. Between-visit calls were placed each month over the two study phases to determine how many moderate or severe hypoglycemic events or episodes of diabetes ketoacidosis may have occurred. At month 6, four psychosocial measures were completed, then rtCGM training was provided and rtCGM use began. At month 12, the psychosocial measures were again completed. Since the current study analyses focus primarily on the rtCGM period, we will henceforth refer to the month 6 visit as “baseline” and the month 12 visit as “study end.”

Measures

The Diabetes Distress Scale (DDS) is a 17-item instrument that assesses worries and concerns specifically related to diabetes and its management and has been shown to be a good marker of diabetes-related QoL.¹³ The DDS yields a total mean-item score with four moderately inter-correlated subscale scores: distress concerning emotional burden, regimen management, problems regarding emotional support from others, and concerns about obtaining satisfactory medical care. Each item is assessed on a six-point Likert scale, from 1 (“Not a Problem”) to 6 (“A Very Serious Problem”). An overall mean score <2.0 indicates little or no distress, 2.0 to 2.9 indicates moderate distress, and ≥3.0 indicates high distress. Any overall or subscale mean score ≥2.0 is considered clinically significant.

The Hypoglycemic Confidence Scale (HCS) is a nine-item questionnaire that yields a total mean-item score and evaluates the degree to which respondents feel able, secure, and comfortable regarding their ability to stay safe from hypoglycemic-related problems.¹⁴ Each HCS item is rated on a four-point scale from 1 (“Not Confident at All”) to 4 (“Very Confident”). Higher mean scores indicate greater confidence in managing hypoglycemia-related issues; scores ≥3 suggest relatively high confidence.

The worry subscale of the revised, 18-item Hypoglycemia Fear Survey (HFS-II) assesses the frequency of common worries regarding hypoglycemia, with items summed for a total scale score. Each item is rated on a five-point scale from 0 (“Never”) to 4 (“Almost Always”). Higher mean scores indicate greater hypoglycemic fear.^15,16

To assess satisfaction with one’s current glucose monitoring system, participants completed two of the four subscales of the 15-item Glucose Monitoring Satisfaction Survey (GMSS), T1D version: GMSS emotional burden (four items) highlights the potential emotional stresses and strains of monitoring systems, and GMSS trust (three items) focuses on the perceived reliability of the monitoring system and the degree of confidence that the results are accurate (ref). Each item is assessed on a five-point Likert scale, from 1 (“Strongly Disagree”) to 5 (“Strongly Agree”). Higher mean scores on GMSS emotional burden point to greater dissatisfaction in this particular dimension, while higher mean score on GMSS trust reflects greater satisfaction in this area.¹⁷

In addition to these four measures, the Gold score, a single question designed to identify IAH, was completed by all participants at baseline only.¹⁸ The item (“Do you know when your low blood sugar events are beginning?”) is rated on a seven-point Likert scale, from 1 (“Always Aware”) to 7 (“Never Aware”). A Gold score >4 reflects the presence of IAH.

Data Analysis

Data analysis was conducted in IBM SPSS Statistics 28. All analyses reported here were conducted only on COACH participants with T2D (N = 174). Descriptive statistics were computed for sample baseline characteristics, psychosocial outcomes, and glycemic outcomes at each time point. Paired samples t-tests were first conducted to examine changes in psychosocial and glycemic outcomes during the six-month rtCGM period from baseline to study end. Cohen’s d was computed to estimate effect sizes for paired mean differences. Difference (change) scores for all psychosocial outcomes and HbA1c were computed by subtracting the baseline (month 6 assessment) score from the study end (month 12 assessment) score, such that a positive value represents an increase from pre-intervention to post-intervention.

The focal analysis involved a series of two multiple regression models that were estimated for each psychosocial outcome change score. In the absence of randomization, change score analysis is the preferred approach over modeling residualized change (analysis of covariance).¹⁹ In all models, the following covariates were included: age, gender, and length of time diagnosed with T2D. The frequency of moderate/severe hypoglycemic events was not included as a covariate due to the very low number of events. The first set of regression analyses included two focal predictor variables in addition to these covariates: pre-rtCGM HbA1c and IAH (both assessed at baseline). This analysis tested the question of whether individuals with greater glycemic problems prior to initiating rtCGM demonstrated more improvement on the psychosocial outcomes over the rtCGM period. In the second set of regression analyses, the focal predictor variable was the HbA1c difference score, representing change in HbA1c from baseline to study end. Baseline HbA1c was included as an additional covariate to determine the unique effect of change in HbA1c above and beyond starting level.

Results

Sample Characteristics

A total of 174 insulin-using adults with T2D completed both phases of the trial. Sample baseline characteristics are shown in Table 1. The majority were non-Hispanic white (67.8%) and 42.5% were female. Mean age was 60.5 (±11.0) years, mean diabetes duration was 18.4 (±8.8) years, and mean A1C at baseline was 7.8% (±1.3%). Although severe hypoglycemic events over the six months prior to rtCGM initiation were rare (only 2.3% reported having one or more events), IAH was reported by 31.8% of the sample.

Table 1.

Sample Characteristics.

	N (%)
Age, M (SD)	60.5 (11.0)
Female	74 (42.5%)
Non-Hispanic/Latino white	118 (67.8%)
Diabetes duration, M (SD)	18.4 (8.8)
Years on insulin, M (SD)	10.5 (7.5)
HbA1c %, M (SD)	7.8 (1.3)
Gold score ≥4	55/173 (31.8%)
Recent hypoglycemic event	4 (2.3%)

N = 174. Gold score ≥4 indicates impaired hypoglycemic awareness. Recent hypoglycemic event represents the occurrence of at least one hypoglycemic event over the six months prior to starting the real-time continuous glucose monitoring phase.

Abbreviations: M, mean; SD, standard deviation; HbA1c, hemoglobin A1c.

Change in Psychosocial Outcomes Over the Study Period

Paired samples t tests revealed significant improvement from baseline to study end in hypoglycemic fear (HFS; P = .031), hypoglycemic confidence (HCS; P < .001), and diabetes distress (DDS; P < .001; Table 2). Among the four DDS subscales, significant reductions were apparent in emotional burden and regimen distress (in both cases, P < .001). Regarding glucose monitoring satisfaction, a significant drop was observed in device-related emotional burden (GMSS emotional burden subscale, P < .001), though there was no change in device-related trust (GMSS trust subscale; P = .808). Significant effects were generally small-to-moderate in magnitude, with Cohen’s d ranging between 0.165 (hypoglycemic fear) and 0.480 (DDS regimen distress). Of note, moderate or severe hypoglycemic episodes continued to be rare over the rtCGM period, with only 1.1% of participants reporting one or more events.

Table 2.

Descriptive Statistics for Psychosocial and Glycemic Outcomes at Each Time Point, M (SD).

	Baseline	Study end	Mean difference	P	D
Hypoglycemia Fear Scale	15.559 (12.473)	13.626 (10.621)	1.932 (11.719)	.031	0.165
Hypoglycemic Confidence Scale	3.351 (0.615)	3.536 (0.431)	−0.185 (0.640)	<.001	−0.289
Diabetes Distress Scale
Emotional burden	2.194 (1.099)	1.893 (0.899)	0.301 (0.918)	<.001	0.328
Interpersonal distress	1.657 (0.989)	1.544 (0.828)	0.113 (0.765)	.053	0.148
Provider distress	1.226 (0.706)	1.194 (0.556)	0.032 (0.696)	.550	0.045
Regimen distress	2.303 (1.099)	1.825 (0.805)	0.478 (0.997)	<.001	0.480
Total score	1.904 (0.842)	1.647 (0.661)	0.257 (0.719)	<.001	0.357
Glucose Monitoring Satisfaction Survey
Emotional burden	2.474 (0.996)	2.019 (0.825)	0.455 (1.125)	<.001	0.405
Trust	3.666 (0.971)	3.642 (1.005)	0.024 (1.301)	.808	0.018
HbA1c	7.836 (1.320)	7.467 (1.179)	0.370 (1.167)	<.001	0.317
Number of hypoglycemic events in last six months	0.040 (0.292)	0.017 (0.169)	0.023 (0.339)	.373	0.068

Two-tailed P value shown is from paired samples t test with df = 173. Baseline refers to the study’s month 6 assessment, while study end refers to the study’s month 12 assessment. Cohen’s d represents effect size of paired mean difference.

Abbreviations: M, mean; SD, standard deviation; HbA1c, hemoglobin A1c.

Predictors of Psychosocial Change

Turning to the focal regression analyses, observed benefits in the psychosocial outcomes were not significantly related to age, gender, or diabetes duration (Table 3). However, baseline Gold score emerged as a relatively consistent, significant predictor: greater IAH was related to significantly more improvement in hypoglycemic fear (P = .002), hypoglycemic confidence (P = .003), diabetes distress total score (P = .013), and device-related emotional burden (P = .033). Significantly greater reductions in three of the four DDS subscales, emotional burden (P = .005), interpersonal distress (P = .021), and regimen distress (P = .028), were also associated with IAH. Baseline HbA1c was also linked with improvements in several psychosocial outcomes; specifically, improvement in hypoglycemic confidence (P = .030) and reductions in two of the DDS subscales, emotional burden (P = .043) and regimen distress (P = .027), were significantly greater in subjects with higher baseline HbA1c.

Table 3.

Baseline Predictors of Change in Psychosocial Outcomes From Baseline to Study End.

	Age		Female		Diabetes duration		Baseline HbA1c		Baseline Gold
Outcome variable (change score)	b (SE)	P	b (SE)	P	b (SE)	P	b (SE)	P	b (SE)	P
Hypoglycemia Fear Scale	0.071 (0.091)	.434	0.650 (1.886)	.731	−0.005 (0.111)	.964	−1.062 (0.732)	.149	−1.610 (0.517)	.002
Hypoglycemic Confidence Scale	−0.006 (0.005)	.256	−0.112 (0.101)	.272	−0.004 (0.006)	.553	0.086 (0.039)	.030	0.084 (0.028)	.003
Diabetes Distress Scale
Emotional burden	0.015 (0.007)	.034	0.002 (0.143)	.990	0.005 (0.008)	.557	−0.113 (0.055)	.043	−0.112 (0.039)	.005
Interpersonal distress	0.013 (0.006)	.025	−0.051 (0.121)	.675	0.009 (0.007)	.192	0.006 (0.047)	.896	−0.077 (0.033)	.021
Provider distress	0.004 (0.006)	.429	−0.018 (0.115)	.877	0.007 (0.007)	.331	0.036 (0.045)	.423	−0.011 (0.032)	.733
Regimen distress	0.018 (0.007)	.017	−0.005 (0.155)	.974	0.003 (0.009)	.767	−0.135 (0.060)	.027	−0.095 (0.043)	.028
Total score	0.013 (0.005)	.017	−0.014 (0.112)	.900	0.005 (0.007)	.410	−0.063 (0.044)	.148	−0.077 (0.031)	.013
GMSS
Emotional burden	0.015 (0.009)	.089	−0.023 (0.179)	.896	0.009 (0.011)	.386	−0.039 (0.070)	.574	−0.106 (0.049)	.033
Trust	−0.013 (0.010)	.184	0.348 (0.208)	.097	−0.010 (0.012)	.439	−0.072 (0.081)	.374	0.053 (0.057)	.358

N = 173. Each row of table shows results from a multiple regression model, where the outcome variable (first column) is a change or difference score from pre-intervention to post-intervention and the predictor variables and corresponding parameters are listed in columns. Two-tailed P values and unstandardized coefficients shown. Diabetes duration measured in years. Baseline refers to the study’s month 6 assessment, while study end refers to the study’s month 12 assessment.

Abbreviations: GMSS, Glucose Monitoring Satisfaction Survey; SE, standard error; HbA1c, hemoglobin A1c.

Toward an explanation of these findings, post hoc mean comparisons of psychosocial outcomes of participants considered to have IAH versus unimpaired hypoglycemic awareness (using cutoff Gold score ≥4) were conducted at baseline and study end. Results are shown in Table 4. Of note, participants with IAH at baseline reported significantly lower hypoglycemic confidence (P = .003) and greater hypoglycemic fear (P = .001), diabetes distress (P = .002), and device-related emotional burden (P < .001) than participants without IAH, but the two groups were no longer different at study end (Table 4). Further inspection of the group means indicates that positive change over time in most of the psychosocial outcomes occurred primarily among participants with IAH, with little change observed in the unimpaired group.

Table 4.

Descriptive Statistics M (SD) in Psychosocial Outcomes at Baseline and Study End by Baseline Gold Score.

	Baseline			Study end
	Unimpaired	Impaired	P	Unimpaired	Impaired	P
Hypoglycemia Fear Scale	13.010 (9.570)	20.855 (15.987)	.001	13.093 (9.687)	14.491 (12.344)	.420
Hypoglycemic Confidence Scale	3.466 (0.479)	3.108 (0.789)	.003	3.550 (0.432)	3.505 (0.434)	.523
Diabetes Distress Scale
Emotional burden	1.988 (0.915)	2.607 (1.320)	.002	1.851 (0.830)	1.945 (1.003)	.515
Interpersonal distress	1.556 (0.824)	1.885 (1.257)	.081	1.545 (0.786)	1.552 (0.923)	.963
Provider distress	1.186 (0.613)	1.309 (0.879)	.290	1.174 (0.423)	1.241 (0.773)	.462
Regimen distress	2.161 (0.971)	2.589 (1.295)	.032	1.834 (0.827)	1.771 (0.718)	.628
Total score	1.774 (0.702)	2.169 (1.042)	.013	1.633 (0.614)	1.659 (0.749)	.807
Glucose Monitoring Satisfaction Survey
Emotional burden	2.292 (0.856)	2.891 (1.142)	<.001	2.008 (0.803)	2.027 (0.879)	.889
Trust	3.757 (0.870)	3.452 (1.140)	.082	3.630 (0.997)	3.709 (0.990)	.627

N = 173 (unimpaired awareness n = 118; impaired awareness n = 55; missing Gold score n = 1). Gold score ≥4 indicates impaired hypoglycemic awareness. P-values are two-tailed, obtained from independent samples t-tests of mean differences between unimpaired and impaired groups at each time point. Baseline refers to the study’s month 6 assessment, while study end refers to the study’s month 12 assessment.

Abbreviations: M, mean; SD, standard deviation.

Table 5 shows the results of models examining change in HbA1c as a predictor of concurrent change in the psychosocial outcomes. Change in HbA1c over the course of the six-month rtCGM period was significantly associated with concurrent change in device-related emotional burden (P = .003) and diabetes distress, including total DDS (P = .010) and the emotional burden (P = .015), regimen distress (P = .004), and provider distress (P = .032) subscales. Specifically, greater HbA1c reductions over the study period were linked to larger drops in each psychosocial outcome.

Table 5.

Does Concurrent Change in HbA1c Predict Change in Psychosocial Outcomes From Baseline to Study End?

Outcome variable (change score)	Age		Female		Diabetes duration		Baseline HbA1c		Change in HbA1c
Outcome variable (change score)	b (SE)	P	b (SE)	P	b (SE)	P	b (SE)	P	b (SE)	P
Hypoglycemia Fear Scale	0.081 (0.094)	.392	−0.236 (1.907)	.902	0.036 (0.114)	.754	−0.391 (0.874)	.655	0.649 (0.941)	.492
Hypoglycemic Confidence Scale	−0.005 (0.005)	.342	−0.039 (0.102)	.704	−0.007 (0.006)	.246	0.107 (0.047)	.023	0.071 (0.050)	.162
Diabetes Distress Scale
Emotional burden	0.017 (0.007)	.015	−0.048 (0.141)	.733	0.006 (0.008)	.513	−0.007 (0.065)	.916	0.172 (0.070)	.015
Interpersonal distress	0.013 (0.006)	.028	−0.101 (0.121)	.406	0.012 (0.007)	.110	0.021 (0.055)	.701	−0.001 (0.060)	.980
Provider distress	0.006 (0.006)	.283	−0.012 (0.112)	.918	0.004 (0.007)	.517	0.095 (0.051)	.066	0.120 (0.055)	.032
Regimen distress	0.021 (0.007)	.005	−0.031 (0.152)	.836	0.002 (0.009)	.801	−0.007 (0.070)	.916	0.218 (0.075)	.004
Total score	0.015 (0.005)	.007	−0.044 (0.110)	.691	0.005 (0.007)	.417	0.022 (0.051)	.665	0.142 (0.055)	.010
GMSS
Emotional burden	0.018 (0.009)	.033	−0.049 (0.174)	.778	0.009 (0.010)	.367	0.113 (0.080)	.158	0.262 (0.086)	.003
Trust	−0.015 (0.010)	.149	0.352 (0.208)	.092	−0.013 (0.012)	.311	−0.135 (0.095)	.159	−0.09 (0.103)	.384

Abbreviations: GMSS, Glucose Monitoring Satisfaction Survey; SE, standard error; HbA1c, hemoglobin A1c.

Discussion

In these data from the COACH study, we found that the introduction of rtCGM in adults with insulin-using T2D was associated with significant improvements in diabetes distress, hypoglycemic fear, hypoglycemic confidence, and device-related satisfaction over a six-month period. These findings are broadly consistent with the T2D results from the Landmark study, which—though a smaller (N = 66) and briefer (12 weeks) trial—noted similarly positive changes in diabetes distress and hypoglycemic concerns after beginning rtCGM.¹¹ We also discovered that the observed gains in these psychosocial outcomes were significantly greater among participants reporting IAH and, among those with higher HbA1c values, prior to initiating rtCGM. In addition, psychosocial gains were significantly related to concurrent glycemic benefit (ie, reduced HbA1c). Collectively, these findings support the hypothesis that rtCGM use in adults contributes to positive, diabetes-related psychosocial outcomes in insulin-using T2D (as has been previously observed in adults with T1D) and, furthermore, provide the first data as to which users (ie, those with IAH and elevated HbA1c) might more likely benefit.

Although reports of severe hypoglycemic episodes were rare at baseline (only 2.3% reported one or more events in the six months prior to rtCGM initiation), IAH prevalence was markedly higher (31.8%) than has been reported among adults with insulin-using T2D in previous reports.^20-23 We suspect this may be partly due to subject self-selection, with individuals reporting IAH more likely to be worried about hypoglycemia, even though they may not have been recently troubled by significant episodes, and thus perhaps more motivated to participate in a study that provided devices that could warn them of impending hypoglycemia. Indeed, we found that IAH at baseline was significantly associated with greater hypoglycemic fear and lower hypoglycemic confidence. Furthermore, given that by study end there were no longer significant differences in psychosocial outcomes between those with and without IAH, rtCGM may be especially beneficial in reducing diabetes-related psychosocial distress in this vulnerable group, helping participants with IAH to feel less out of control with T2D (the underlying theme of diabetes distress)^13,24 and less fearful about hypoglycemia, even though the actual rates of significant hypoglycemia began, and remained very low during the study period.

Following from the original COACH publication, which reported that HbA1c dropped significantly over the rtCGM period (for both adults T1D and T2D participants),¹² the current study found that these reductions in HbA1c in the T2D group were significantly associated with simultaneous drops in diabetes distress over the rtCGM period. It is not possible to determine whether there may be a causal relationship here, but we speculate that as participants experienced and observed the glycemic benefits of rtCGM, they began to feel more hopeful, confident, and in control of T2D, thus leading to the reduction in diabetes distress. Future studies utilizing more intensive longitudinal assessment during rtCGM intervention would be valuable in elucidating the directionality and causal nature of these associations.

The primary strengths of this study are the sample size, the broad inclusion criteria, and the real-world–like conditions which reflected usual practice. There were, however, significant methodological limitations, which highlight the need for caution when interpreting these findings. First, there was no control condition, so we cannot be certain that the observed changes in psychosocial outcomes were a direct result of rtCGM use. While all subjects did participate in a six-month SMBG-only period prior to rtCGM initiation, no psychosocial data were collected at the beginning of the SMBG period (month 0). As result, no comparison of psychosocial changes during the two study periods was possible. Furthermore, if change in one or more of the psychosocial outcomes over the SMBG period did occur (eg, an increase in diabetes distress due to the lack of any positive intervention), this may have influenced the appearance of psychosocial benefit observed during the rtCGM period. Second, though the single-item Gold score is recognized as an adequate method for assessing IAH, it may be that more comprehensive, multi-item scales (eg, Clarke Questionnaire,²⁵ Hypoglycemia Awareness Questionnaire)²⁶ might have provided a more detailed view of IAH in this T2D sample. Third, although there was relatively little attrition in the parent study (~16%),¹³ psychosocial outcomes among those who dropped out early from the study are not known. Fourth, the rtCGM used in this study, the Dexcom™ G5 CGM System^®, was not the most current of the newer CGM systems and still required twice-daily calibration. Use of any of the newer CGM systems that no longer require calibration may have led to different results. Finally, the majority of study participants were non-Hispanic whites, and so not representative of the broader T2D population. It is evident that further research on the potential value of rtCGM in more ethnically diverse populations is needed.

Conclusions

We conclude that rtCGM may lead to diabetes-specific psychosocial benefits for adults with insulin-using T2D, especially for individuals with IAH. Future RCTs focusing on the insulin-using T2D population with the newest generations of CGM systems in more diverse populations are needed to provide support for these findings.

Footnotes

Abbreviations

CGM, continuous glucose monitoring; rtCGM, real-time CGM; isCGM, intermittently scanned CGM; T2D, type 2 diabetes; T1D, type 1 diabetes; IAH, impaired hypoglycemic awareness; QoL, quality of life; DKA, diabetes ketoacidosis; DDS, Diabetes Distress Scale; HCS, Hypoglycemic Confidence Scale; HFS-II, Hypoglycemia Fear Survey; GMSS, Glucose Monitoring Satisfaction Survey.

Declaration of Conflicting Interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: W.H.P. has served as a consultant for Dexcom and Abbott Diabetes Care.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by Dexcom.

ORCID iDs

Emily C. Soriano

William H. Polonsky

References

Polonsky

. Psychosocial aspects of diabetes technology: adult perspective. Endocrinol Metab Clin North Am. 2020;49:143-155.

Klak

Manczak

Owoc

Olszewski

. Impact of continuous glucose monitoring on improving emotional well-being among adults with type 1 diabetes mellitus: a systematic review and meta-analysis. Pol Arch Intern Med. 2021;131:808-818.

Beck

Riddlesworth

Ruedy

, et al. Continuous glucose monitoring versus usual care in patients with type 2 diabetes receiving multiple daily insulin injections: a randomized trial. Ann Intern Med. 2017;167:365-374.

Vigersky

Fonda

Chellappa

, et al. Short- and long-term effects of real-time continuous glucose monitoring in patients with type 2 diabetes. Diabetes Care. 2012;35:32-38.

Haak

Hanaire

Ajjan

, et al. Use of flash glucose-sensing technology for 12 months as a replacement for blood glucose monitoring in insulin-treated type 2 diabetes. Diabetes Ther. 2017;8:573-586.

Yaron

Roitman

Aharon-Hananel

, et al. Effect of flash glucose monitoring technology on glycemic control and treatment satisfaction in patients with type 2 diabetes. Diabetes Care. 2019;42:1178-1184.

Wada

Onoue

Kobayashi

, et al. Flash glucose monitoring helps achieve better glycemic control than conventional self-monitoring of blood glucose in non-insulin-treated type 2 diabetes: a randomized controlled trial. BMJ Open Diabetes Res Care. 2020;8:e001115. doi:10.1136/bmjdrc-2019-001115.

Lameijer

Fokkert

Edens

, et al. Two-year use of flash glucose monitoring is associated with sustained improvement of glycemic control and quality of life (FLARE-NL-6). BMJ Open Diabetes Res Care. 2021;9:e002124. doi:10.1136/bmjdrc-2021-002124.

Fokkert

van Dijk

Edens

, et al. Improved well-being and decreased disease burden after 1-year use of flash glucose monitoring (FLARE-NL4). BMJ Open Diabetes Res Care. 2019;7:e000809. doi:10.1136/bmjdrc-2019-000809.

10.

Mitsuishi

Nishimura

Harashima

, et al. The effect of novel glucose monitoring system (flash glucose monitoring) on mental well-being and treatment satisfaction in Japanese people with diabetes. Adv Ther. 2018;35:72-80.

11.

Gilbert

Noar

Blalock

Polonsky

. Change in hemoglobin A1c and quality of life with real-time continuous glucose monitoring use by people with insulin-treated diabetes in the Landmark Study. Diabetes Technol Ther. 2021;23:S35-S39.

12.

Beck

Kelly

Price

, et al. Non-adjunctive continuous glucose monitoring for control of hypoglycaemia (COACH): results of a post-approval observational study. Diabet Med. 2022;39:e14739. doi:10.1111/dme.14739.

13.

Polonsky

Fisher

Earles

, et al. Assessing psychosocial distress in diabetes: development of the Diabetes Distress Scale. Diabetes Care. 2005;28:626-631.

14.

Polonsky

Fisher

Hessler

Edelman

. Investigating hypoglycemic confidence in type 1 and type 2 diabetes. Diabetes Technol Ther. 2017;19:131-136.

15.

Cox

Irvine

Gonder-Frederick

Nowacek

Butterfield

. Fear of hypoglycemia: quantification, validation, and utilization. Diabetes Care. 1987;10:617-621.

16.

Irvine

Cox

Gonder-Frederick

. The fear of hypoglycaemia scale. In: Bradley

, ed. Handbook of Psychology and Diabetes. New York, NY: Harwood Academic Publishers; 1994: 133-155.

17.

Polonsky

Fisher

Hessler

Edelman

. Development of a new measure for assessing glucose monitoring device-related treatment satisfaction and quality of life. Diabetes Technol Ther. 2015;17:657-663.

18.

Gold

MacLeod

Frier

. Frequency of severe hypoglycemia in patients with type I diabetes with impaired awareness of hypoglycemia. Diabetes Care. 1994;17:697-703.

19.

Fitzmaurice

. A conundrum in the analysis of change. Nutrition. 2001;17:360-361.

20.

Henderson

Allen

Deary

Frier

. Hypoglycaemia in insulin-treated type 2 diabetes: frequency, symptoms and impaired awareness. Diabet Med. 2003;20:1016-1021.

21.

Schopman

Geddes

Frier

. Prevalence of impaired awareness of hypoglycaemia and frequency of hypoglycaemia in insulin-treated type 2 diabetes. Diabetes Res Clin Pract. 2010;87:64-68.

22.

Meijel

de Vegt

Abbink

, et al. High prevalence of impaired awareness of hypoglycemia and severe hypoglycemia among people with insulin-treated type 2 diabetes: the Dutch Diabetes Pearl Cohort. BMJ Open Diabetes Res Care. 2020;8:e000935. doi:10.1136/bmjdrc-2019-000935.

23.

Alkhatatbeh

Abdalqader

Alqudah

MAY

. Impaired awareness of hypoglycaemia in insulin-treated type 2 diabetes mellitus. Current Diabetes Rev. 2019;15:407-413.

24.

Polonsky

Fisher

Hessler

Desai

King

Perez-Nieves

. Toward a more comprehensive understanding of the emotional side of type 2 diabetes: a re-envisioning of the assessment of diabetes distress. J Diabetes Complications. 2022;36:108103. doi:10.1016/j.jdiacomp.2021.108103.

25.

Clarke

Cox

Gonder-Frederick

Julian

Schlundt

Polonsky

. Reduced awareness of hypoglycemia in adults with IDDM. A prospective study of hypoglycemic frequency and associated symptoms. Diabetes Care. 1995;18:517-522.

26.

Speight

Barendse

Singh

, et al. Characterizing problematic hypoglycaemia: iterative design and preliminary psychometric validation of the Hypoglycaemia Awareness Questionnaire (HypoA-Q). Diabet Med. 2016;33:376-385.