OBJECTIVE

Estimated time in range (eTIR) obtained from DCCT glucose profiles (pre- and postprandial and bedtime) was recently reported to be associated with microvascular outcomes and was recommended as a clinical trial outcome, but without consideration of HbA1c.

RESEARCH DESIGN AND METHODS

The associations of eTIR with diabetic retinopathy and microalbuminuria were assessed without and with adjustment for HbA1c and baseline covariates.

RESULTS

Adjusted for HbA1c and covariates, eTIR was marginally significantly associated with retinopathy in the full cohort (hazard ratio [HR] 1.12 per 10% lower eTIR [95% CI 1.0, 1.26], P = 0.042). Conversely, HbA1c was significantly associated with both outcomes (HR ≥1.19 per 0.5% higher HbA1c, P ≤ 0.0002) in five of six adjusted analyses.

CONCLUSIONS

The association of eTIR with complications is largely explained by its correlation with HbA1c. HbA1c, not eTIR or continuous glucose monitoring TIR, remains the preferred outcome in clinical studies of type 1 diabetes complications.

The time-in-range (TIR) glucose from continuous glucose monitoring (CGM), generally defined as the percentage of time in the ∼70–180 mg/dL range, is commonly used to guide diabetes care (1,2). Beck et al. (3) recently used seven-point fingerstick blood glucose profiles (before and after meals and at bedtime) obtained in the Diabetes Control and Complications Trial (DCCT), as measured by the DCCT central biochemistry laboratory (4), to calculate an estimated TIR (eTIR). These eTIR values were associated with DCCT complications, and Beck et al. concluded that TIR measured from CGM could be “an acceptable end point for clinical trials.”

While eTIR is an imperfect measure of mean glucose, it is still correlated with HbA1c, which, in turn, is strongly associated with microvascular outcomes (5,6). Therefore, some, if not all, of the association of eTIR with outcomes may just reflect the role of mean HbA1c rather than that of eTIR per se. To determine whether the eTIR has an independent association with outcomes beyond that of HbA1c and could substitute for, or provide added value to, HbA1c in future clinical investigations, we have replicated the analyses of Beck et al. (3) by including both eTIR and HbA1c together.

The DCCT showed that HbA1c was a major determinant of the risk of progression of complications (7). The 1,441 participants were randomly assigned to intensive or conventional diabetes therapy and followed for an average of 6.5 years. Of these, the primary prevention cohort (n = 726) had no preexisting microvascular complications and 1–5 years duration of type 1 diabetes (T1D), and the secondary intervention cohort (n = 715) had mild preexisting complications and 1–15 years duration of T1D.

Average follow-up was 6.5 years with 6 monthly assessments of diabetic retinopathy on fundus photographs. The primary outcome was sustained progression of retinopathy at two successive visits. The albumin excretion rate was measured annually, with a secondary outcome being the onset of microalbuminuria (albumin excretion rate ≥30 mg/24 h) also sustained.

HbA1c and the blood glucose profiles were measured quarterly by the central laboratory. Complete glucose profiles were obtained in 67% of the ∼37,000 expected quarterly collections. As previously described (8), multiple imputation (9) generated 10 complete data sets using the multiple imputation by chained equations algorithm (10), and the Rubin-Schenker average was obtained (11). The same data and programs were used by Beck et al. (3) and herein. The quarterly eTIR was the percentage of the seven glucose values in the 70–180 mg/dL range. The mean eTIR and mean HbA1c up to each visit (every 3 months) were then used as time-dependent factors in discrete-time Cox proportional hazards models of the outcome event (visit) times with and without adjustment for other covariates. Simple P values without correction for multiple tests are presented.

Table 1 presents the hazard ratio (HR) for each outcome per 10% lower eTIR with no adjustment for other covariates as in the Beck et al. (3) analyses, then with adjustment for HbA1c alone and adjustment for both HbA1c and other covariates (3). Table 1 also presents the HR per 0.5% higher HbA1c when adjusted for eTIR and other covariates. Analyses are presented for the full DCCT cohort and separately for the primary and secondary cohorts.

In unadjusted analyses, as in Beck et al. (3), eTIR was significantly associated with the six outcomes. However, when adjusted for HbA1c alone, these associations were substantially diminished (HR closer to 1.00), and only the association of eTIR with nephropathy in the primary cohort reached significance (HR 1.40 per 10% lower eTIR, P = 0.027). When adjusted for HbA1c and other covariates, this eTIR association remained unchanged (HR 1.39, P = 0.032), and the association with retinopathy in the full cohort was barely significant (HR 1.12, P = 0.042).

Conversely, HbA1c had a statistically significant association with all outcomes when unadjusted, when adjusted for eTIR, and when adjusted for eTIR and other covariates (each P < 0.003), except for microalbuminuria in the primary cohort where P = 0.193. However, only 36 primary cohort subjects developed microalbuminuria, inadequate for a definitive conclusion.

In the analyses by Beck et al. (3), a higher eTIR was significantly associated with a decreased risk of progression of retinopathy and nephropathy in the DCCT cohort. However, in the analyses herein, those associations are largely eliminated with adjustment for HbA1c, whereas the association of HbA1c with outcomes remains largely significant when adjusted for eTIR. Therefore, these results do not provide direct evidence that higher TIR computed from CGM, or the DCCT eTIR, contributes to the risk of progression of complications beyond that conferred by, or independent from, HbA1c levels.

The analyses herein address these issues by fitting models with eTIR and HbA1c separately and together. Adjusted for HbA1c alone, eTIR was not significantly associated with the outcome in five of the six analyses. When also adjusted for other covariates, eTIR was no longer nominally significant in four of the six analyses. After correcting for six multiple tests (data not shown), all eTIR P values became nonsignificant. Likewise, HbA1c had a significant association in five of the six analyses when adjusted for eTIR (all P < 0.001). These associations also remained significant (P < 0.006) when corrected for multiple tests. Thus, eTIR fails to replace or negate the association of HbA1c with outcomes and does not augment the association of HbA1c with outcomes.

The analyses by Beck et al. (3) showed that eTIR by itself was associated with outcomes, whereas the current analyses show that those associations are largely negated when adjusted for HbA1c. Conversely, the HbA1c associations are not affected by adjustment for eTIR. Thus, if a treatment were to increase the eTIR, it is possible that it might show a beneficial association with outcomes. However, that benefit will depend to a greater extent on the treatment’s effects on HbA1c, not eTIR. The potential benefit of a treatment’s effect on CGM-measured TIR is not known.

Clearly, the DCCT eTIR is an imperfect representation of TIR as would be provided by CGM. Furthermore, a substantial fraction of the profile data was missing and replaced using multiple imputation. While previous analyses validated this approach (3,8), the missing data are a weakness. Nevertheless, the quarterly DCCT seven-point profiles provide the best-available opportunity to assess the association of eTIR with progression of retinopathy and nephropathy during the DCCT, since it is unlikely that a DCCT-like study using CGM will ever be performed. Thus, the eTIR serves as a surrogate for a bona fide measure of TIR obtained from CGM.

In conclusion, our analyses with eTIR do not support the conclusion by Beck et al. (3) that “a compelling case can be made that TIR is strongly associated with the risk of microvascular complications and should be an acceptable end point for clinical trials.”

Clinical trial reg. nos. NCT00360893 and NCT00360815, clinicaltrials.gov

*

A complete list of the members of the DCCT/EDIC Research Group can be found in N Engl J Med 2017;376:1507–1516.

This article is featured in a podcast available at diabetesjournals.org/journals/pages/diabetes-core-update-podcasts.

Funding. The DCCT/EDIC has been supported by cooperative agreement grants (1982–1993, 2012–2022) and contracts (1982–2012) with the Division of Diabetes Endocrinology and Metabolic Diseases, National Institute of Diabetes and Digestive and Kidney Diseases (current grants U01 DK094176 and U01 DK094157), and through support by the National Eye Institute, National Institute of Neurologic Disorders and Stroke, General Clinical Research Centers Program (1993–2007), and Clinical Translational Science Center Program (2006–present).

Duality of Interest. Free or discounted supplies or equipment to support participants’ adherence to the study have been provided by Abbott Diabetes Care, Animas, Bayer Diabetes Care, Becton Dickinson, Eli Lilly, Extend Nutrition, Insulet Corporation, Lifescan, Medtronic Diabetes, Nipro Home Diagnostics, Nova Diabetes Care, Omron, Perrigo Diabetes Care, Roche Diabetes Care, and Sanofi. No other potential conflicts of interest relevant to this article were reported.

Industry contributors have had no role in the DCCT/EDIC study.

Authors Contributions. J.M.L. and D.M.N. wrote the initial draft of the manuscript. I.B. and X.G. conducted the statistical analyses. I.B. and B.Z. contributed to revising the manuscript. All authors approved the final content. J.M.L. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

1.
Agiostratidou
G
,
Anhalt
H
,
Ball
D
, et al
.
Standardizing clinically meaningful outcome measures beyond HbA1c for type 1 diabetes: a consensus report of the American Association of Clinical Endocrinologists, the American Association of Diabetes Educators, the American Diabetes Association, the Endocrine Society, JDRF International, The Leona M. and Harry B. Helmsley Charitable Trust, the Pediatric Endocrine Society, and the T1D Exchange
.
Diabetes Care
2017
;
40
:
1622
1630
2.
Battelino
T
,
Danne
T
,
Bergenstal
RM
, et al
.
Clinical targets for continuous glucose monitoring data interpretation: recommendations from the International Consensus on Time in Range
.
Diabetes Care
2019
;
42
:
1593
1603
3.
Beck
RW
,
Bergenstal
RM
,
Riddlesworth
TD
, et al
.
Validation of time-in-range as an outcome measure for diabetes clinical trials
.
Diabetes Care
2019
;
42
:
400
405
4.
The DCCT Research Group
.
The Diabetes Control and Complications Trial (DCCT). Design and methodologic considerations for the feasibility phase
.
Diabetes
1986
;
35
:
530
545
5.
The DCCT Research Group
.
The relationship of glycemic exposure (HbA1c) to the risk of development and progression of retinopathy in the Diabetes Control and Complications Trial
.
Diabetes
1995
;
44
:
968
983
6.
Lachin
JM
,
Genuth
S
,
Nathan
DM
,
Zinman
B
;
DCCT/EDIC Research Group
.
Effect of glycemic exposure on the risk of microvascular complications in the Diabetes Control and Complications Trial--revisited
.
Diabetes
2008
;
57
:
995
1001
7.
Nathan
DM
,
Genuth
S
,
Lachin
J
, et al.;
Diabetes Control and Complications Trial Research Group
.
The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus
.
N Engl J Med
1993
;
329
:
977
986
8.
Lachin
JM
,
Bebu
I
,
Bergenstal
RM
, et al.;
DCCT/EDIC Research Group
.
Association of glycemic variability in type 1 diabetes with progression of microvascular outcomes in the Diabetes Control and Complications Trial
.
Diabetes Care
2017
;
40
:
777
783
9.
Rubin
DB
.
Multiple Imputation for Nonresponse in Surveys
.
New York
,
John Wiley & Sons
,
1987
10.
White
IR
,
Royston
P
,
Wood
AM
.
Multiple imputation using chained equations: issues and guidance for practice
.
Stat Med
2011
;
30
:
377
399
11.
Rubin
DB
,
Schenker
N
.
Multiple imputation for interval estimation from simple random samples with ignorable nonresponse
.
J Am Stat Assoc
1986
;
81
:
366
374
Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at https://www.diabetesjournals.org/journals/pages/license.