Abstract
Purpose:
The purpose of this study is to evaluate whether higher emotional distress (depressive symptoms or diabetes distress) was associated with a lower likelihood of basal or rapid-acting insulin initiation among participants enrolled in the GRADE Emotional Distress Substudy (EDS).
Methods:
Individuals with type 2 diabetes <10 years duration on metformin alone were randomized to add 1 of 4 glucose-lowering drugs. Per protocol, participants were expected to start basal or rapid-acting insulin rescue therapy after reaching secondary or tertiary glycemic outcomes (A1C >7.5%). Depressive symptoms and diabetes distress were assessed ≤12 months prior to outcome confirmation. Multinomial and binomial logistic regression models examined associations of depressive symptoms and diabetes distress with basal insulin initiation and rapid-acting insulin initiation, respectively.
Results:
Of the 525 participants expected to start basal insulin, 30.9% initiated ≤6 weeks, 35.2% initiated >6 weeks, and 33.9% never initiated. Of the 325 participants expected to start rapid-acting insulin, 67.4% never initiated. Neither depressive symptoms nor diabetes distress were associated with starting basal or rapid-acting insulin.
Conclusions:
In the GRADE EDS, approximately one-third of participants did not start basal insulin, and two-thirds of participants did not start rapid-acting insulin. Emotional distress did not appear to play a role in insulin initiation among trial participants.
Therapeutic inertia is the lack of timely adjustment to the treatment regimen when a patient’s therapeutic targets are not met.1 Among patients with diabetes, the median time to intensifying a patient’s treatment after A1C was above goal is more than 1 year (range 0.3 to >7.2 years).2 Diabetes complications, higher A1C levels, younger age, and use of fewer noninsulin diabetes medications are associated with higher likelihood of starting insulin.3 Emotional distress, including depressive symptoms and diabetes-related distress, is higher in those already taking insulin compared to noninsulin users.4–7
Reluctance to use insulin therapy, or psychological insulin resistance, is driven by fear of insulin injections and self-monitoring of blood glucose, hypoglycemia, and weight gain; a perceived loss of control; poor self-efficacy; and concern about poor outcomes related to insulin therapy.8–11 Reluctance to use insulin is associated with lower likelihood of initiating or intensifying insulin therapy, contributing to therapeutic inertia.9,10,12 After starting insulin, patients experience reduced negative appraisal of insulin, negative attitudes toward injections, and feelings of guilt related to diabetes self-management, whereas those who did not start insulin had increased or unchanged negative appraisal of insulin.8,13 Depressive symptoms and diabetes distress may play a role in a participant’s reluctance to use insulin because insulin-naïve patients with emotional distress are more likely to have negative beliefs about insulin.14,15 However, 2 previous observational, longitudinal studies16,17 did not find differences in insulin initiation or time to insulin initiation by baseline depressive symptoms; these studies were limited by the length of time between the baseline assessment of depressive symptoms and the decision to initiate insulin.16,17 No studies have examined the relationship of emotional distress to insulin initiation at the time when insulin is indicated or when the decision is made to initiate insulin. Additionally, there are few, if any, studies evaluating initiation of basal and rapid-acting insulin separately, which is important given the more intensive regimen required with rapid-acting insulin. Research that advances the understanding of how emotional distress influences decisions regarding insulin initiation or intensification can provide important information to clinicians and patients to guide treatment and avoid therapeutic inertia.
The objective of this study was to evaluate whether depressive symptoms or diabetes distress were associated with initiating basal or rapid-acting insulin therapy among trial participants enrolled in the GRADE (Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study) Emotional Distress Substudy (EDS).18 Specifically, the aims of this study were to (1) examine the association of emotional distress (depressive symptoms and diabetes distress) and initiation of basal insulin and (2) examine the association of emotional distress and initiation of rapid-acting insulin.
Methods
Study Design and Participants
The GRADE study was a multicenter, parallel group, comparative effectiveness clinical trial comparing 4 commonly used glucose-lowering medications (basal insulin glargine U-100, sulfonylurea glimepiride, glucagon-like peptide 1 [GLP-1] receptor agonist liraglutide, or dipeptidyl peptidase 4 [DPP-4] inhibitor sitagliptin) in participants on metformin-monotherapy. The GRADE trial protocol and primary outcome articles have been published previously and describe randomization procedures.19,20 A total of 5047 participants were enrolled in GRADE from 36 clinical centers and 9 subsites in the United States. Eligibility criteria were age ≥30 years (age ≥20 years for American Indian and Alaska Natives), diagnosis of type 2 diabetes mellitus within 10 years, A1C of 6.8% to 8.5% (51–69 mmol/mol), treatment with metformin (minimum dose of 1000 mg/d), and willingness to take a second diabetes medication, including injectable medications.
Participants completed quarterly visits including measurement of A1C. The primary glycemic outcome was defined as A1C ≥7.0% (53 mmol/mol) confirmed at the next quarterly visit. Participants reached the secondary glycemic outcome (referred to hereafter as “secondary outcome”) if they had an A1C >7.5% (58 mmol/mol) confirmed at the next quarterly visit after a minimum of 6 months on maximally tolerated doses of the randomized study medication and metformin. The protocol specified that after secondary outcome confirmation, basal insulin was to be added in the non-glargine treatment groups and that rapid-acting insulin should be initiated in the glargine group. The time frame to initiate insulin was operationalized as within 6 weeks of outcome confirmation (the midpoint between visits) to allow participants time to initiate insulin between assessments. The tertiary glycemic outcome (referred to hereafter as “tertiary outcome”) was defined as reaching an A1C >7.5% (58 mmol/mol) confirmed at the next quarterly visit among participants who previously reached the secondary outcome. Per protocol, rapid-acting insulin was expected to be initiated after tertiary outcome confirmation for the non-glargine treatment groups, and randomized study medication was stopped so that participants were continued on metformin, basal insulin, and rapid-acting insulin.
The EDS was an embedded substudy of GRADE that aimed to assess emotional distress in type 2 diabetes; the EDS protocol was published previously.18 All GRADE sites were invited but not required to participate in EDS; 26 of 36 GRADE centers and 8 subsites participated. At participating sites, all participants who enrolled in GRADE were simultaneously consented and enrolled in EDS and received compensation for completing the additional substudy assessments. From October 2015 to August 2017, 1739 participants were enrolled. Site institutional review boards approved study procedures prior to initiation; written informed consent was obtained from all participants. CONSORT reporting guidelines were followed.21
The analyses were restricted to EDS participants, including the subset of EDS participants who were eligible to initiate basal insulin or rapid-acting insulin as outlined in the following. The basal insulin analyses were restricted to those participants from the 3 non-glargine treatment groups who met and confirmed the secondary outcome within 12 months after completion of an EDS questionnaire, allowing for 2 opportunities to complete the questionnaire, which was administered every 6 months, unless the secondary outcome confirmation occurred after completion of the EDS follow-up period. Rapid-acting insulin analyses were restricted to EDS participants from the glargine treatment group who met the secondary outcome and from the 3 non-glargine treatment groups who met the tertiary outcome. Of note, participants from the non-glargine treatment groups were only eligible to start rapid-acting insulin after the tertiary outcome if they previously started basal insulin after reaching the secondary outcome. The rapid-acting insulin analyses excluded participants from the non-glargine treatment groups who never initiated basal insulin. Additionally, the sample was restricted to those participants who had glycemic outcome confirmation within 12 months of a completed EDS questionnaire. Participants who had a questionnaire within 12 months were included in order to include as many participants in the analysis as possible and to reduce selection bias due to excluding participants who were missing an EDS questionnaire and/or who confirmed the outcome more than 6 months following completion of the EDS follow-up period.
Measures
Participants completed the EDS questionnaire (including Patient Health Questionnaire [PHQ-8] and Diabetes Distress Scale [DDS]) every 6 months for up to 3 years. Emotional distress was assessed using validated measures of depressive symptoms (PHQ-8) and diabetes distress (DDS). All analyses were based on depressive symptoms and diabetes distress assessed from the most recent EDS questionnaire prior to the glycemic outcome confirmation.
The validated PHQ-8 is an 8-item self-report questionnaire with items based on criteria for depression from the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders.22 Participants were asked to respond with how often they experienced a particular depressive symptom during the preceding 2 weeks with response options from 0 (not at all) to 3 (nearly every day). Responses are summed to yield a total score ranging from 0 to 24; higher scores indicate more severe depressive symptoms, and a score of ≥10 indicates the presence of moderate depressive symptoms. The PHQ-8 includes items related to somatic and cognitive-affective depressive symptoms, which were averaged and included as symptom domains. Internal reliability for the PHQ-8 in the current sample was good (standardized Cronbach α = 0.88 for the basal insulin analysis sample, α = 0.82 for the rapid-acting insulin analysis sample).
Diabetes-related distress was measured by the validated 17-item DDS, which asks participants the degree to which each item (e.g., “feeling overwhelmed by the demands of living with diabetes”) was currently problematic for them on a 6-point Likert scale from 1 (no problem) to 6 (serious problem).23 Responses are averaged to yield total DDS scores ranging from 1 to 6; higher scores indicate greater diabetes distress severity, with an average score of ≥2 indicating clinically significant diabetes distress.24 The DDS has 4 subscales (scores ranging from 1 to 6): (1) emotional burden of having diabetes, (2) regimen distress, (3) interpersonal distress, and (4) physician-related distress. Internal reliability for the total scale was excellent (standardized Cronbach α = 0.94 for the basal insulin analysis sample, α = 0.92 for the rapid-acting insulin analysis sample).
Participants were categorized based on initiation of basal insulin following the secondary outcome confirmation into 3 groups: initiation of insulin ≤6 weeks after outcome confirmation, initiation of insulin >6 weeks after outcome confirmation, and never initiated insulin before the end of study follow-up. Due to the small number of EDS participants who met the tertiary outcome and initiated rapid-acting insulin, participants were categorized as initiating rapid-acting insulin at any time after outcome confirmation before the end of study follow-up versus never initiating.
Self-reported participant baseline characteristics included age, sex, race/ethnicity, and diabetes duration. Diabetes duration was defined as the number of years between the reported diabetes diagnosis date and the time of outcome confirmation. The randomized treatment group assignment was used for all analyses: glargine (insulin), glimepiride (sulfonylurea), liraglutide (GLP-1 receptor agonist), or sitagliptin (DPP 4-inhibitor). Diabetes complications were defined as presence of at least 1 of the following at the time of outcome confirmation: prior amputation, retinopathy, nephropathy (presence of either albumin-to-creatinine ratio ≥30 mg/g or estimated glomerular filtration rate <60 ml/min/m2), neuropathy (presence of diabetic peripheral neuropathy determined by Michigan Neuropathy Screening Instrument score and clinical examination), or cardiovascular autonomic neuropathy, defined as presence of abnormal values for 2 heart rate variability indices.25
Statistical Analyses
Descriptive characteristics were summarized separately for the basal insulin analysis sample and rapid-acting insulin analysis sample as median and interquartile range (IQR) for continuous variables and counts and percentages for categorical variables, overall and stratified by insulin initiation category (ie, 3-category variable for the basal insulin analysis sample; binary indicator for the rapid-acting insulin analysis sample).
To assess the association of emotional distress with insulin initiation, separate multinomial logistic regression models were fit for the 3-category basal insulin initiation variable, and separate logistic regression models were fit for the 2-category rapid-acting insulin initiation variable, with PHQ-8 total score or total DDS score as a predictor variable (ie, 2 separate models for each insulin initiation outcome, 1 for each predictor variable). Models for basal insulin initiation were adjusted for these covariates: duration of diabetes and diabetes complications at the time of the secondary outcome confirmation and randomized treatment group. Due to the smaller sample size for the rapid-acting insulin analyses, models for this outcome were not adjusted for covariates. A likelihood ratio test of the model effect for the emotional distress predictor variable (ie, PHQ-8 total score or total DDS score) was used to test for an association between the emotional distress variable and the insulin initiation variable. Odds ratios and 95% confidence intervals were calculated based on the fitted model to report the estimated magnitude of the association. In addition, exploratory analyses were conducted by repeating the regression analyses with each of the PHQ-8 symptom types and DDS subscales as the predictor variable. Sensitivity analyses were conducted: (1) evaluating each of the aforementioned associations with basal insulin initiation categorized as a binary variable (initiated vs never initiated) with logistic regression models (for the basal insulin initiation outcome only) and (2) repeating the aforementioned analyses restricting the sample to participants who completed the EDS questionnaire within 6 months of outcome confirmation.
Results
In the basal insulin analyses, 525 participants were included from the 3 non-glargine treatment groups who reached the secondary outcome (A1C >7.5% [58 mmol/mol] and were expected to start basal insulin) and had completed measures of depressive symptoms and diabetes distress within 12 months prior to outcome confirmation (Figure 1A, Table 1). Participants were excluded from the basal insulin analyses for these reasons: early initiation of insulin or a missing outcome confirmation form (n = 9) or missing 2 consecutive EDS questionnaires or reaching the secondary outcome >12 months after completing the 3-year EDS follow-up (n = 43). Median age was 56.8 years (IQR 49.5, 65.1); participants were 36.2% female; 66.3% White, 17.9% Black, and 21.6% Hispanic. Of these, 162 participants (30.9%) initiated basal insulin ≤6 weeks, 185 (35.2%) initiated >6 weeks, and 178 (33.9%) never initiated. Racial differences were noted between insulin initiation categories: 40.4% of Black participants initiated insulin ≤6 weeks compared to 29.3% of White participants, and 40.4% of Black participants never initiated insulin compared to 30.7% of White participants, P = .003.
Figure 1.
Flowchart for the GRADE (Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study) Emotional Distress Substudy (EDS) sample. (A) Basal insulin analyses.a (B) Rapid-acting insulin analyses.b
aParticipants were excluded from the basal insulin analyses for these reasons: early initiation of insulin or a missing outcome confirmation form (n = 9) or missing 2 consecutive EDS questionnaires or reaching the secondary outcome >12 months after completion of the 3-year EDS follow-up (n = 43).
bParticipants were excluded from the rapid-acting insulin analyses for these reasons: not adding glargine after reaching the secondary outcome (n = 96), missing outcome confirmation form (n = 1), or missing 2 consecutive EDS questionnaires or reaching the outcome >12 months after completion of the 3-year EDS follow-up (n = 41). Non-glargine groups: Glimepiride (N = 439), Liraglutide (N = 450), Sitagliptin (N = 432).
Table 1.
GRADE (Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study) Emotional Distress Substudy (EDS) Participant Characteristics Stratified by Basal Insulin Initiationa
| Characteristics | Insulin initiation | ||||
|---|---|---|---|---|---|
| Allb | ≤6 wkc | >6 wkc | Neverc | P valued | |
| N total | 525 | 162 | 185 | 178 | |
| Age, ye | 56.8 [49.5, 65.1] | 55.3 [49.1, 62.0] | 56.1 [48.9, 65.8] | 58.8 [53.1, 66.7] | .008 |
| Sex, n (%) | .080 | ||||
| Female | 190 (36.2) | 70 (36.8) | 60 (31.6) | 60 (31.6) | |
| Male | 335 (63.8) | 92 (27.5) | 125 (37.3) | 118 (35.2) | |
| Race, n (%) | .003 | ||||
| White | 348 (66.3) | 102 (29.3) | 139 (39.9) | 107 (30.7) | |
| Black/African American | 94 (17.9) | 38 (40.4) | 18 (19.1) | 38 (40.4) | |
| All others | 83 (15.8) | 22 (26.5) | 28 (33.7) | 33 (39.8) | |
| Hispanic/Latino, n (%) | .386 | ||||
| Yes | 113 (21.6) | 34 (30.1) | 35 (31.0) | 44 (38.9) | |
| No | 409 (78.4) | 126 (30.8) | 150 (36.7) | 133 (32.5) | |
| Diabetes duration, ye | 5.4 [3.6, 7.6] | 5.3 [3.4, 7.7] | 5.1 [3.5, 7.0] | 5.6 [3.9, 7.9] | .293 |
| A1C (%)e | 8.2 [7.8, 8.8] | 8.2 [7.8, 8.6] | 8.4 [7.9, 9.2] | 8.1 [7.8, 8.6] | <.001 |
| Diabetes complications, N (%)e,f | .359 | ||||
| Yes | 230 (47.1) | 68 (29.6) | 91 (39.6) | 71 (30.9) | |
| No | 258 (52.9) | 85 (32.9) | 86 (33.3) | 87 (33.7) | |
| Treatment group, n (%) | |||||
| Glargine | N/A | N/A | N/A | N/A | |
| Glimepiride | 177 (33.7) | 49 (27.7) | 58 (32.8) | 70 (39.5) | |
| Liraglutide | 156 (29.7) | 53 (34.0) | 51 (32.7) | 52 (33.3) | |
| Sitagliptin | 192 (36.6) | 60 (31.2) | 76 (39.6) | 56 (29.2) | |
| PHQ-8 total scoreg | 2.0 [0.0, 5.0] | 2.0 [0.0, 5.0] | 2.0 [0.0, 6.0] | 2.0 [0.0, 4.0] | |
| ≥10 | 42 (8.0) | 16 (38.1) | 16 (38.1) | 10 (23.8) | |
| <10 | 483 (92.0) | 146 (30.2) | 169 (35.0) | 168 (34.8) | |
| Total DDS scoreg | 1.4 [1.1, 1.9] | 1.4 [1.1, 1.9] | 1.4 [1.2, 2.0] | 1.4 [1.1,1.9] | |
| ≥2 | 119 (22.7) | 39 (32.8) | 47 (39.5) | 33 (27.7) | |
| < 2 | 405 (77.3) | 123 (30.4) | 137 (33.8) | 145 (35.8) | |
Abbreviations: DDS, Diabetes Distress Scale; PHQ-8, Patient Health Questionnaire.
Continuous variables are summarized as median [interquartile range]; categorical variables are summarized as count.
Column percent.
Row percent.
The P values are presented for tests of differences by insulin initiation group, based on an analysis of variance global F test for continuous variables or Pearson’s chi-square test for categorical variables.
At the time of outcome confirmation.
Diabetes complications = at least 1 of the following: prior amputation, retinopathy, nephropathy, neuropathy, or cardiovascular autonomic neuropathy.
Emotional distress measures from the most recent EDS questionnaire prior to outcome confirmation.
In regression analyses adjusted for diabetes duration, treatment group, and diabetes complications, overall depressive symptoms (PHQ-8 total score) were not significantly associated with timing of basal insulin initiation (OR 1.00, 95% CI, 0.95–1.06 for insulin initiation ≤6 weeks and OR 1.04, 95% CI, 0.99–1.10 for insulin initiation >6 weeks compared to never initiating basal insulin, P = .267; Table 2). Similarly, total diabetes distress score was not significantly associated with timing of basal insulin initiation (OR 1.05, 95% CI, 0.78–1.42 for insulin initiation ≤6 weeks and OR 1.15, 95% CI, 0.86–1.53 for insulin initiation >6 weeks compared to never initiating basal insulin, P = .630). Additionally, no significant associations were found for depressive symptom types and diabetes distress subscales in separate models. For each of the aforementioned models, unadjusted models yielded similar results with no significant associations. The results of sensitivity analyses examining the association of emotional distress with basal insulin initiation as a binary variable are presented in Table 2; no significant associations were found.
Table 2.
Association of Measures of Emotional Distress With Basal Insulin Initiation, as a Three-Category Variable and as a Binary Variable, Adjusted Analysesa
| 3-category basal insulin initiationb | Binary basal insulin initiationc | ||||
|---|---|---|---|---|---|
| Emotional distress measure | Odds ratiod (≤6 wk vs never) | Odds ratiod (>6 wk vs never) | P value | Odds ratiod (anytime vs never) | P value |
| Depressive symptoms | |||||
| PHQ-8 total score | 1.00 [0.95, 1.06] | 1.04 [0.99, 1.10] | .267 | 1.02 [0.98, 1.08] | .262 |
| PHQ-8 somatic depressive symptoms | 1.04 [0.68, 1.59] | 1.35 [0.91, 1.99] | .247 | 1.20 [0.85, 1.73] | .213 |
| PHQ-8 cognitive-affective depressive symptoms | 1.01 [0.64, 1.60] | 1.27 [0.83, 1.93] | .450 | 1.15 [0.79, 1.71] | .485 |
| Diabetes distress | |||||
| Total DDS score | 1.05 [0.78, 1.42] | 1.15 [0.86, 1.53] | .630 | 1.10 [0.86, 1.44] | .576 |
| DDS subscales | |||||
| Emotional burden | 1.12 [0.85, 1.47] | 1.20 [0.92, 1.56] | .396 | 1.16 [0.92, 1.48] | .263 |
| Regimen distress | 1.06 [0.85, 1.32] | 1.13 [0.92, 1.40] | .494 | 1.10 [0.92, 1.33] | .415 |
| Physician distress | 0.95 [0.72, 1.26] | 0.84 [0.62, 1.14] | .512 | 0.90 [0.70, 1.15] | .303 |
| Interpersonal distress | 0.97 [0.76, 1.23] | 1.12 [0.90, 1.40] | .376 | 1.05 [0.86, 1.30] | .681 |
Abbreviations: DDS, Diabetes Distress Scale; PHQ-8: Patient Health Questionnaire.
Models adjusted for diabetes duration and diabetes complications at time of outcome confirmation in addition to randomized treatment group.
Multinomial logistic regression model for the 3-category basal insulin initiation outcome variable.
Logistic regression model for the binary basal insulin initiation outcome variable.
95% confidence intervals included in brackets.
In the rapid-acting insulin analyses, 325 participants were included from the 4 treatment groups who reached secondary outcome for the glargine treatment group or tertiary outcome for the non-glargine treatment groups (A1C >7.5% [58 mmol/mol]) and therefore were expected to start rapid-acting insulin and had completed measures of depressive symptoms and diabetes distress within 12 months prior to outcome confirmation (Figure 1B, Table 3). Participants were excluded from the rapid-acting insulin analyses for 3 reasons: not adding glargine before reaching the tertiary outcome (n = 96), missing outcome confirmation form (n = 1), and missing 2 consecutive EDS questionnaires or reaching the outcome >12 months after completion of the 3-year EDS follow-up (n = 41). Median participant age was 55.7 years (IQR 49.1, 62.3); the sample was 34.5% female and 62.5% White, 22.5% Black, and 24.4% Hispanic. Of 325 participants, 106 (32.6%) initiated rapid-acting insulin after outcome confirmation before the end of study follow-up. A higher proportion of Black participants did not initiate rapid-acting insulin (78.1%) compared to White participants (67.0%), P = .015. A higher proportion of Hispanic participants initiated rapid-acting insulin (43.0%) compared to non-Hispanic participants (29.4%), P = .035. Supplementary Table 1 presents participant characteristics stratified by glargine versus non-glargine groups.
Table 3.
GRADE (Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study) Emotional Distress Substudy (EDS) Participant Characteristics Stratified by Rapid-Acting Insulin Initiationa
| Insulin initiation | ||||
|---|---|---|---|---|
| Characteristics | Allb | Anytimec | Neverc | P valued |
| N total | 325 | 106 | 219 | |
| Age, ye | 55.7 [49.1, 62.3] | 52.3 [45.2, 60.6] | 56.3 [51.0, 62.8] | <.001 |
| Sex, n (%) | .451 | |||
| Female | 112 (34.5) | 33 (29.5) | 79 (70.5) | |
| Male | 213 (65.5) | 73 (34.3) | 140 (65.7) | |
| Race, n (%) | .015 | |||
| White | 203 (62.5) | 67 (33.0) | 136 (67.0) | |
| Black/African American | 73 (22.5) | 16 (21.9) | 57 (78.1) | |
| All others | 49 (15.1) | 23 (46.9) | 26 (53.1) | |
| Hispanic/Latino, n (%) | .035 | |||
| Yes | 79 (24.4) | 34 (43.0) | 45 (57.0) | |
| No | 245 (75.6) | 72 (29.4) | 173 (70.6) | |
| Diabetes duration, ye | 5.5 [3.8, 7.6] | 4.6 [2.6, 7.5] | 5.8 [4.1, 7.6] | .004 |
| Diabetes complications, n (%)e,f | .582 | |||
| Yes | 126 (42.7) | 40 (31.7) | 86 (68.3) | |
| No | 169 (57.3) | 60 (35.5) | 109 (64.5) | |
| A1C (%) | 8.1 [7.8, 8.8] | 8.1 [7.9, 9.2] | 8.1 [7.8, 8.6] | .034 |
| Treatment group, n (%) | ||||
| Glargine | 145 (44.6) | 68 (46.9) | 77 (53.1) | |
| Glimepiride | 67 (20.6) | 9 (13.4) | 58 (86.6) | |
| Liraglutide | 48 (14.8) | 10 (20.8) | 38 (79.2) | |
| Sitagliptin | 65 (20.0) | 19 (29.2) | 46 (70.8) | |
| PHQ-8 total scoreg | 2.0 [0.0, 5.0] | 2.0 [1.0, 5.0] | 2.0 [0.0, 5.0] | |
| ≥10 | 25 (7.7) | 6 (24) | 19 (76.0) | |
| <10 | 299 (92.3) | 100 (33.4) | 199 (66.6) | |
| Total DDS scoref | 1.5 [1.1, 1.9] | 1.5 [1.2, 1.9] | 1.4 [1.1, 1.9] | |
| ≥2 | 71 (21.9) | 25 (35.2) | 46 (64.8) | |
| <2 | 253 (78.1) | 81 (32.0) | 172 (68.0) | |
Abbreviations: DDS, Diabetes Distress Scale; PHQ-8: Patient Health Questionnaire.
Continuous variables are summarized as median [interquartile range]; categorical variables are summarized as count.
Column percent.
Row percent.
The P values are presented for tests of differences by insulin initiation group, based on an analysis of variance global F test for continuous variables or Pearson’s chi-square test for categorical variables.
At the time of outcome confirmation.
Diabetes complications = presence of at least 1: prior amputation, retinopathy, nephropathy, neuropathy, or cardiovascular autonomic neuropathy.
Emotional distress measures from the most recent EDS questionnaire prior to outcome confirmation.
In unadjusted regression analyses, overall depressive symptoms (PHQ-8 total score) were not significantly associated with initiation of rapid-acting insulin at any time prior to end of study follow-up (OR 0.99, 95% CI, 0.93–1.06, compared to never initiating, P = .822; Table 4). Similarly, total diabetes distress score was not significantly associated with initiation of rapid-acting insulin (OR 1.27, 95% CI, 0.94–1.72, compared to never initiating, P = .125). No significant associations were found for depressive symptom types and diabetes distress subscales.
Table 4.
Results of Regression Models Evaluating Association of Emotional Distress With Rapid Acting Insulin Initiation, Unadjusted Analyses
| Emotional distress measure | Odds ratioa (anytime vs never) | P value |
|---|---|---|
| Depressive symptoms | ||
| PHQ-8 total score | 0.99 [0.93, 1.06] | .822 |
| PHQ-8 somatic depressive symptoms | 0.98 [0.61, 1.56] | .935 |
| PHQ-8 cognitive-affective depressive symptoms | 0.91 [0.55, 1.45] | .695 |
| Diabetes distress | ||
| Total DDS score | 1.27 [0.94, 1.72] | .125 |
| DDS subscales | ||
| Emotional burden | 1.28 [0.97, 1.68] | .081 |
| Regimen distress | 1.08 [0.86, 1.35] | .498 |
| Physician distress | 1.13 [0.86, 1.48] | .371 |
| Interpersonal distress | 1.27 [1.01, 1.60] | .044 |
Abbreviations: DDS, Diabetes Distress Scale; PHQ-8: Patient Health Questionnaire.
95% confidence intervals included in brackets.
Finally, analyses restricting the sample to those participants who had an EDS questionnaire completed within 6 months of the outcome did not show significant associations between emotional distress and basal insulin initiation (N = 473, 90.1% of sample) or rapid-acting insulin initiation (N = 274, 84.3% of sample; Supplementary Tables 2 and 3).
Discussion
In GRADE EDS, after reaching secondary or tertiary glycemic outcomes, approximately one-third of participants did not start basal insulin, and two-thirds of participants did not start rapid-acting insulin. Neither depressive symptoms nor diabetes distress appear to have played a role in initiation of basal or rapid-acting insulin in this trial setting with routine assessments of emotional distress 6 to 12 months prior to the decision of whether to initiate insulin.
Although depressive symptoms and diabetes distress have been linked to greater negative beliefs about insulin, there was not a significant relationship between either measure of emotional distress and subsequent initiation of basal or rapid-acting insulin in this trial environment.14 Findings from the current analyses and the larger GRADE EDS study point to a potential shift in understanding how emotional distress relates to insulin therapy and informs clinical care. Previous studies4–7 have supported an association between emotional distress and insulin therapy, but in this longitudinal study, when adequately accounting for important factors such as disease progression, these previously established relationships between emotional distress and insulin initiation were not demonstrated. In fact, in Gonzalez et al,26 the EDS study team reports that depressive symptoms and diabetes distress decreased over the course of the 3-year GRADE EDS follow-up period. This change was particularly pronounced during the first year of the study among those assigned to the basal insulin treatment group, who experienced significantly lower diabetes distress compared to participants in the other 3 treatment groups. Thus, clinicians should not view elevated depressive symptoms or diabetes distress in a patient as an absolute barrier to initiating insulin. Clinicians may consider that insulin therapy could alleviate emotional distress initially and continue to focus on addressing specific potential concerns related to insulin use with their patients.27,28
It may be that reluctance to use insulin is driven by other factors not reflected in depressive symptoms or diabetes distress scales (ie, concerns or negative beliefs specific to insulin, fear of injections, or burden of increased self-monitoring of blood glucose).15 In another study reported for the larger GRADE cohort (N = 2387), adherence to randomized study medications was high, but similar to the currently presented analyses, Hollander et al found that 36% of participants did not initiate basal insulin and that 70% did not initiate rapid-acting insulin.29 In this prior article,29 other factors were reported as reasons for delaying or declining to initiate insulin, including desire to focus on other strategies to improve glucose (ie, lifestyle changes or increased adherence to current treatment regimen) and specific concerns related to insulin. The protocol of the GRADE trial attempted to mitigate some of these factors in that participants had regular follow-up and assessments with specified treatment protocols outlining when basal or rapid-acting insulin should be initiated. Additionally, participants were provided study medications at no cost, removing potential financial barriers to initiating insulin.30 These findings drive home the challenge of intensifying the treatment regimen with insulin, particularly rapid-acting insulin, when A1C is above target levels. Clinicians may consider these potential strategies to support patients as they initiate insulin, including demonstrating the injection process, explaining the benefits of insulin, sharing examples of success in using insulin therapy, managing expectations, and collaborative decision-making style, which have been described in previous studies27,28 as effective strategies by patients or clinicians.
One strength of this study is that participants within 10 years of diabetes diagnosis comprise the study sample; this extends the existing literature that has studied the relationship of emotional distress and insulin use in populations who are further along in their disease course or have more advanced disease. These results indicate the importance of studying the association of emotional distress and insulin initiation across a range of populations who have different durations of diabetes and levels of emotional distress to more completely characterize these relationships. Additionally, the racial/ethnic and socioeconomic diversity of the study sample is a strength of this analysis. Furthermore, this study separately examined the association of emotional distress with basal insulin and rapid-acting insulin initiation. Emotional distress was assessed within 12 months prior to outcome confirmation, whereas the gap between baseline assessment of distress and the time of the decision to initiate insulin was greater in previous studies.16,17
This study should be interpreted considering the following limitations. This analysis was conducted among clinical trial participants who completed a run-in period, agreed to start insulin as part of study eligibility assessment at the start of the study, and regularly completed assessments and were in contact with study personnel, so their experience differed from standard clinical practice. It is possible that the factors related to nonadherence to treatment regimens in this clinical trial setting may differ from those in standard clinical practice. Additionally, for participants in the non-glargine treatment groups who reached tertiary outcome, their randomized study medication was stopped when rapid-acting insulin was added, which may have played a role in their decision-making. Furthermore, the EDS cohort had lower levels of depressive symptoms (8% with elevated levels) and diabetes distress (22% to 23% with elevated levels) at baseline compared to other studies that included people with diabetes with longer duration or more varied treatment regimens. These factors may limit generalizability of these findings to other populations. Due to sample size, there were limits to the number of covariates adjusted for in the models; thus, other covariates, such as medication adherence, were not able to be included. Finally, participants were enrolled in GRADE EDS between 2015 and 2017. Recent studies showing the effectiveness of insulin in once-weekly injection along with the availability of other effective noninsulin injectable medications provide other options for treatment intensification among patients with type 2 diabetes when indicated; additionally, increased availability of continuous glucose monitoring may reduce the burden of self-monitoring.31–33
Conclusions
There was not an association between depressive symptoms or diabetes distress with insulin initiation in trial participants. More research is needed to identify the most effective strategies to support individualization and intensification of treatment for individuals when indicated to reduce therapeutic inertia and achieve target glycemic control.
Supplementary Material
Supplemental material for this article is available online.
Acknowledgments
The GRADE (Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study) Study Research Group is deeply grateful to our participants, whose loyal dedication made GRADE possible. Clinical trial registration: “A Comparative Effectiveness Study of Major Glycemia-Lowering Medications for Treatment of Type 2 Diabetes (GRADE),” No. NCT01794143, https://clinicaltrials.gov/study/NCT01794143.
Declaration of Conflicting Interests
CAP reports a grant from the National Center for Complementary and Integrative Health outside the submitted work. NMB reports participation on the Data and Safety Monitoring Board (DSMB) for the Fast Antibiotic Susceptibility Testing for Gram Negative Bacteriemia Trial (FAST) outside the submitted work. CVD reports a grant from Novo Nordisk; consulting fees from Novo Nordisk, Asahi, Bayer, and Merck; travel expenses from Novo Nordisk for a meeting; participation on the Nation VA DSMB for clinical trials (unpaid); and being the NEBRA board director (unpaid). CJH reports grants from the National Institute on Aging, National Institute of Diabetes and Digestive and Kidney Diseases, and Juvenile Diabetes Research Foundation outside the submitted work. NR reports grants or contracts from Novo Nordisk and Eli Lilly and consulting fees from Novo Nordisk and Eli Lilly outside the submitted work. HKS, PAH, VSL, EAL, CLM, JSG, and ALC have nothing to disclose.
Funding
The GRADE Study was supported by a grant from the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) of the National Institutes of Health under Award No. U01DK098246. The planning of GRADE was supported by a U34 planning grant from the NIDDK (U34-DK-088043). The American Diabetes Association supported the initial planning meeting for the U34 proposal. The National Heart, Lung, and Blood Institute and the Centers for Disease Control and Prevention also provided funding support. The Department of Veterans Affairs provided resources and facilities. Additional support was provided by Grant Nos. P30 DK017047, P30 DK020541, P30 DK020572, P30 DK072476, P30 DK079626, P30 DK092926, U54 GM104940, UL1 TR000170, UL1 TR000439, UL1 TR000445, UL1 TR001102, UL1 TR001108, UL1 TR001409, 2UL1TR001425, UL1 TR001449, UL1 TR002243, UL1 TR002345, UL1 TR002378, UL1 TR002489, UL1 TR002529, UL1 TR002535, UL1 TR002537, UL1 TR002541, and UL1 TR002548. Educational materials were provided by the National Diabetes Education Program. Material support in the form of donated medications and supplies were provided by Becton, Dickinson and Company, Bristol-Myers Squibb, Merck & Co., Inc., Novo Nordisk, Roche Diagnostics, and Sanofi. The GRADE Emotional Distress Substudy was supported by a grant from the NIDDK of the National Institutes of Health under Award No. R01 DK104845. Additional support was provided by grant number P30 DK111022.
The content of this article is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Footnotes
Ethical Considerations
At participating sites, all participants who enrolled in GRADE (Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study) were simultaneously consented and enrolled in the Emotional Distress Substudy (EDS) and received compensation for completing the additional substudy assessments. From October 2015 to August 2017, 1739 participants were enrolled. Site institutional review boards approved study procedures prior to initiation; written informed consent was obtained from all participants.
Guarantor Statement
Caroline Presley and Nicole Butera are the guarantors of this work and as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Data Availability
The GRADE (Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study) study is funded by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK). This article is based on follow-up data and outcome assessments from the 1739 participants enrolled in the Emotional Distress Substudy (EDS). The GRADE database, which includes the Emotional Distress Substudy (EDS), is available in the NIDDK Central Repository at https://repository.niddk.nih.gov/study/151.
References
- 1.Gabbay RA, Kendall D, Beebe C, et al. Addressing therapeutic inertia in 2020 and beyond: a 3-year initiative of the American Diabetes Association. Clin Diabetes. 2020;38(4):371–381. doi: 10.2337/cd20-0053 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Khunti K, Gomes MB, Pocock S, et al. Therapeutic inertia in the treatment of hyperglycaemia in patients with type 2 diabetes: a systematic review. Diabetes Obes Metab. 2018;20(2):427–437. doi: 10.1111/dom.13088 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Turchin A, Hosomura N, Zhang H, Malmasi S, Shubina M. Predictors and consequences of declining insulin therapy by individuals with type 2 diabetes. Diabet Med. 2020;37(5):814–821. doi: 10.1111/dme.14260 [DOI] [PubMed] [Google Scholar]
- 4.Wardian JL, Tate J, Folaron I, Graybill S, True M, Sauerwein T. Who’s distressed? A comparison of diabetes-related distress by type of diabetes and medication. Patient Educ Couns. 2018;101(8):1490–1495. doi: 10.1016/j.pec.2018.03.001 [DOI] [PubMed] [Google Scholar]
- 5.Nanayakkara N, Pease A, Ranasinha S, et al. Depression and diabetes distress in adults with type 2 diabetes: results from the Australian National Diabetes Audit (ANDA) 2016. Sci Rep. 2018;8(1):7846. doi: 10.1038/s41598-018-26138-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Delahanty LM, Grant RW, Wittenberg E, et al. Association of diabetes-related emotional distress with diabetes treatment in primary care patients with type 2 diabetes. Diabet Med. 2007;24(1):48–54. doi: 10.1111/j.1464-5491.2007.02028.x [DOI] [PubMed] [Google Scholar]
- 7.Bai X, Liu Z, Li Z, Yan D. The association between insulin therapy and depression in patients with type 2 diabetes mellitus: a meta-analysis. BMJ Open. 2018;8(11):e020062. doi: 10.1136/bmjopen-2017-020062 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Odawara M, Ishii H, Tajima N, Iwamoto Y. Impact of patient attitudes and beliefs to insulin therapy upon initiation, and their attitudinal changes after initiation: the DAWN Japan study. Curr Med Res Opin. 2016;32(4):681–686. doi: 10.1185/03007995.2015.1136605 [DOI] [PubMed] [Google Scholar]
- 9.Petrak F, Stridde E, Leverkus F, Crispin AA, Forst T, Pfutzner A. Development and validation of a new measure to evaluate psychological resistance to insulin treatment. Diabetes Care. 2007;30(9):2199–2204. doi: 10.2337/dc06-2042 [DOI] [PubMed] [Google Scholar]
- 10.Gherman A, Veresiu I, Sassu R, Schnur J, Scheckner B, Montgomery G. Psychological insulin resistance: a critical review of the literature. Pract Diabetes Int. 2011;28(3):125–128d. doi: 10.1002/pdi.1574 [DOI] [Google Scholar]
- 11.Leslie CS-RW, Matheson D, Stone R, Enfield G. Psychological insulin resistance: a missed diagnosis. Diabetes Spect. 1994;7(1):52–57. [Google Scholar]
- 12.Polonsky WH, Fisher L, Guzman S, Villa-Caballero L, Edelman SV. Psychological insulin resistance in patients with type 2 diabetes: the scope of the problem. Diabetes Care. 2005;28(10):2543–2545. doi: 10.2337/diacare.28.10.2543 [DOI] [PubMed] [Google Scholar]
- 13.Hermanns N, Mahr M, Kulzer B, Skovlund SE, Haak T. Barriers towards insulin therapy in type 2 diabetic patients: results of an observational longitudinal study. Health Qual Life Outcomes. 2010;8:113. doi: 10.1186/1477-7525-8-113 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Makine C, Karsidag C, Kadioglu P, et al. Symptoms of depression and diabetes-specific emotional distress are associated with a negative appraisal of insulin therapy in insulin-naive patients with type 2 diabetes mellitus. A study from the European Depression in Diabetes [EDID] Research Consortium. Diabet Med. 2009;26(1):28–33. doi: 10.1111/j.1464-5491.2008.02606.x [DOI] [PubMed] [Google Scholar]
- 15.Holmes-Truscott E, Skinner TC, Pouwer F, Speight J. Explaining psychological insulin resistance in adults with noninsulin-treated type 2 diabetes: the roles of diabetes distress and current medication concerns. Results from Diabetes MILES–Australia. Prim Care Diabetes. 2016;10(1):75–82. doi: 10.1016/j.pcd.2015.06.006 [DOI] [PubMed] [Google Scholar]
- 16.Upsher R, Yaquetto DA, Stahl D, Ismail K, Winkley K. Prospective study of the association between depressive symptoms at type 2 diabetes diagnosis and time to insulin initiation in the South London diabetes (SOUL-D) cohort. Prim Care Diabetes. 2022;16(4):502–508. doi: 10.1016/j.pcd.2022.05.007 [DOI] [PubMed] [Google Scholar]
- 17.Nefs G, Pop VJ, Denollet J, Pouwer F. The longitudinal association between depressive symptoms and initiation of insulin therapy in people with type 2 diabetes in primary care. PLoS One. 2013;8(11):e78865. doi: 10.1371/journal.pone.0078865 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Cherrington AL, Krause-Steinrauf H, Bebu I, et al. Study of emotional distress in a comparative effectiveness trial of diabetes treatments: Rationale and design. Contemp Clin Trials. 2021;107:106366. doi: 10.1016/j.cct.2021.106366 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Nathan DM, Buse JB, Kahn SE, et al. Rationale and design of the glycemia reduction approaches in diabetes: a comparative effectiveness study (GRADE). Diabetes Care. 2013;36(8):2254–2261. doi: 10.2337/dc13-0356 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Nathan DM, Lachin JM, Balasubramanyam A, et al. Glycemia reduction in type 2 diabetes - glycemic outcomes. N Engl J Med. 2022;387(12):1063–1074. doi: 10.1056/NEJMoa2200433 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Schulz KF, Altman DG, Moher D, CONSORT Group. CONSORT 2010 statement: updated guidelines for reporting parallel group randomised trials. Trials. 2010;11:32. doi: 10.1186/1745-6215-11-32 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Kroenke K, Strine TW, Spitzer RL, Williams JB, Berry JT, Mokdad AH. The PHQ-8 as a measure of current depression in the general population. J Affect Disord. 2009;114(1–3):163–173. doi: 10.1016/j.jad.2008.06.026 [DOI] [PubMed] [Google Scholar]
- 23.Polonsky WH, Fisher L, Earles J, et al. Assessing psychosocial distress in diabetes: development of the diabetes distress scale. Diabetes Care. 2005;28(3):626–631. doi: 10.2337/diacare.28.3.626 [DOI] [PubMed] [Google Scholar]
- 24.Fisher L, Hessler DM, Polonsky WH, Mullan J. When is diabetes distress clinically meaningful?: establishing cut points for the Diabetes Distress Scale. Diabetes Care. 2012;35(2):259–264. doi: 10.2337/dc11-1572 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Pop-Busui R, Backlund JC, Bebu I, et al. Utility of using electrocardiogram measures of heart rate variability as a measure of cardiovascular autonomic neuropathy in type 1 diabetes patients. J Diabetes Investig. 2022;13(1):125–133. doi: 10.1111/jdi.13635 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Gonzalez JS, Bebu I, Krause-Steinrauf H, et al. Differential effects of type 2 diabetes treatment regimens on diabetes distress and depressive symptoms in the Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study (GRADE). Diabetes Care. 2024;47(4):610–619. doi: 10.2337/dc23-2459 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Allen NA, Zagarins SE, Feinberg RG, Welch G. Treating psychological insulin resistance in type 2 diabetes. J Clin Transl Endocrinol. 2017;7:1–6. doi: 10.1016/j.jcte.2016.11.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Polonsky WH, Fisher L, Hessler D, et al. Identifying solutions to psychological insulin resistance: an international study. J Diabetes Complications. 2019;33(4):307–314. doi: 10.1016/j.jdiacomp.2019.01.001 [DOI] [PubMed] [Google Scholar]
- 29.Hollander PA, Krause-Steinrauf H, Butera NM, et al. The use of rescue insulin in the Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study (GRADE). Diabetes Care. 2024;47(4):638–645. doi: 10.2337/dc23-0516 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.McAdam-Marx C, Ruiz-Negron N, Sullivan JM, Tucker JM. The effects of patient out-of-pocket costs for insulin on medication adherence and health care utilization in patients with commercial insurance; 2007–2018. J Manag Care Spec Pharm. 2022;28(5):494–506. doi: 10.18553/jmcp.2022.21481 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.American Diabetes Association Professional Practice Committee. 9. Pharmacologic approaches to glycemic treatment: standards of care in diabetes-2024. Diabetes Care. 2024;47(suppl 1):S158–S178. doi: 10.2337/dc24-S009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Shetty S, Suvarna R. Efficacy and safety of once-weekly insulin icodec in type 2 diabetes: a meta-analysis of ONWARDS phase 3 randomized controlled trials. Diabetes Obes Metab. 2024;26(3):1069–1081. doi: 10.1111/dom.15408 [DOI] [PubMed] [Google Scholar]
- 33.Martens TW. Roadmap to the effective use of continuous glucose monitoring in primary care. Diabetes Spectr. 2023;36(4):306–314. doi: 10.2337/dsi23-0001 [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
Data Availability Statement
The GRADE (Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study) study is funded by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK). This article is based on follow-up data and outcome assessments from the 1739 participants enrolled in the Emotional Distress Substudy (EDS). The GRADE database, which includes the Emotional Distress Substudy (EDS), is available in the NIDDK Central Repository at https://repository.niddk.nih.gov/study/151.