Advancements in continuous glucose monitoring technology allow nuanced assessment of glycemic control. Glycemic variability (GV) and time spent with glucose in target range time-in-range (TIR) are now recognized risk predictors for diabetes-related complications and are associated with cardiovascular risk, even for patients with well-controlled HbA1c (1,2). Prior research by our group and others demonstrates that besides carbohydrate amount and type, the timing of carbohydrate consumption during a meal significantly impacts regulation of postprandial glucose levels (3–5). In acute crossover studies, consuming fibrous vegetables and protein 10 min before carbohydrate (i.e., carbohydrates-last [CL] food order) reduced incremental postprandial glucose peaks and glycemic variability for up to 3 h in comparison with the reverse sequence (carbohydrates-first [CF] food order) in patients with type 2 diabetes (T2D) (4). Here, in this study (clinical trial reg. no. NCT04738799, clinicaltrials.gov) we examine the glycemic effects of nutrient sequencing without a rest interval between consumption of meal components—under controlled and free-living conditions. The primary end point for sample size calculation was the incremental glucose peak (IGP) measured in controlled conditions, on the basis of our prior research (4).

A total of 20 adult participants (8 male, 12 female) with metformin-treated T2D were studied via within-subject crossover design. Mean ± SD age, BMI, HbA1c, and duration of T2D were 59.2 ± 8.5 years, 34.5 ± 4.9 kg/m2, 7.2% ± 0.46%, and 4.5 ± 2.5 years, respectively. Participants were randomized to consume eucaloric standardized meals reflective of diverse cultural eating patterns in either a carbohydrates-last (CL) or CF meal sequence for 6 days, followed by the reverse sequence for the subsequent 6 days. Meals were prepared by an academic clinical research metabolic kitchen with energy content of meals calculated using the Food Processor (ESHA Research) nutrient analysis and menu planning software. The menu for the study had two calorie levels: approximately 2,100 or 2,600 kcal/day. Participants were provided the calorie level closest to their eucaloric energy needs, estimated on study enrollment with the Mifflin St. Jeor formula. Meal components were packaged and color coded to facilitate adherence. The first 5 days of each intervention period took place in free-living conditions; the 6th occurred in controlled conditions. Average macronutrient distribution of the meals was 20%–25% protein, 30%–35% fat, 45%–50% carbohydrate, and 25–30 g fiber. Participants underwent continuous glucose monitoring via Dexcom G6 Pro (Dexcom) throughout the observation period. Participants maintained daily food logs recording the amounts and sequence of meal components consumed. Participants were instructed to maintain the same level of activity throughout the study period; this was captured via pedometers. In the controlled condition, venous blood glucose was measured at time points 0, 30, 60, 90, 120, 150, and 180 min following consumption of a sequenced breakfast meal (with no rest interval between meal components) after an overnight fast.

In controlled conditions, IGPs were reduced by 44% in the CL versus CF sequence (48.3 ± 25.8 vs. 85.7 ± 25.5 mg/dL, respectively; P < 0.001) (Fig. 1). Total area under the curve (26,040.8 ± 4,157.7 vs. 30,117.0 ± 4,014.8 mg/dL × 180 min; P < 0.0001) and incremental area under the curve (4,053.8 ± 2,970.1 vs. 8,112.0 ± 3,619.1 mg/dL × 180 min; P < 0.0001) for glucose were significantly lower for the CL versus CF sequence. In free-living conditions, average blood glucose was similar between CL and CF (146.0 ± 22.2 vs. 152.3 ± 25.1 mg/dL; P = 0.202); however, with the CL sequence there was significantly improved glycemic variability, assessed with coefficient of variation (CV) (19.2% ± 4.4% vs. 23.0% ± 4.4%; P = 0.001) and percent TIR (84.8% ± 14.0% vs. 78.6% ± 15.1%; P = 0.041). A subanalysis of CV and TIR, performed with data from days 3–5 to wash out a potential carryover effect, revealed similar findings and statistical significance.

Figure 1

A: Blood glucose concentration over 180 min in the controlled conditions. N = 19. B: IGP in the controlled conditions. N = 19. C: TIR in the free-living conditions. N = 20. D: Glycemic variability assessed with CV in the free-living conditions. N = 20. Data are means ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001.

Figure 1

A: Blood glucose concentration over 180 min in the controlled conditions. N = 19. B: IGP in the controlled conditions. N = 19. C: TIR in the free-living conditions. N = 20. D: Glycemic variability assessed with CV in the free-living conditions. N = 20. Data are means ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001.

Close modal

In this study, we demonstrated that food order substantially impacts short-term glycemic control in T2D. We confirmed that the effect of food order remains significant even without a rest interval between consumption of different meal components, enhancing its practical application. Strengths of the study include robust study design, statistically significant findings in both controlled and free-living conditions over a 2-week period, and translational nature of the intervention to real-world conditions. Limitations of the study include sample size, unclear generalizability to patients treated with insulin, and the inability to ensure absolute adherence to the study protocol in free-living conditions. Although participants were provided with eucaloric diets, there was an average weight loss of 1.97 kg over the 12-day period. However, there was no significant difference in body weights measured at the end of the CL versus CF intervention periods. Our findings indicate that food order may be a practical behavioral strategy to improve clinical outcomes in patients with T2D. This is meaningful, because in our experience patients are more amenable to changing the order of food consumption than they are to changing the overall quantity or macronutrient composition of meals, which may make this intervention more sustainable. Further investigations assessing the impact of food order on long-term T2D outcomes are indicated.

Clinical trial reg. no. NCT04738799, clinicaltrials.gov

Acknowledgments. The authors thank Dr. Zhengming Chen for statistical support, Lynelle Weis for assistance with data compilation, Clara Cho for assistance with subject recruitment, and Anthony Casper for regulatory support, all from Weill Cornell Medicine.

Funding. This work was supported by the Weill Cornell Friend Center Weight Fund. The research reported in this publication was supported by the National Center for Advancing Translational Sciences of the National Institutes of Health under award UL1TR002384. S.T. was supported by the Roberts Lifestyle and Diversity Scholars Award at Weill Cornell Medicine.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. S.T., K.C.H., and A.P.S. contributed to the development of the study concept and design. S.T., A.K., and A.P.S. contributed to drafting the manuscript. S.T., K.P., A.K., K.C.H., J.R., S.S., A.G., and A.S.Z. contributed to the conduct of study procedures and data acquisition. D.D. performed data analysis. S.T., K.P., A.K., K.C.H., J.R., A.S.Z., L.C.A., L.J.A., and A.P.S. performed data interpretation. All authors contributed to the review and editing of the manuscript. A.P.S. served as the principal investigator. A.P.S. 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.

Prior Presentation. Parts of this study were presented in abstract form at the 41st Annual Meeting of the Obesity Society, ObesityWeek, Dallas, TX, 13–17 October 2023.

Handling Editors. The journal editors responsible for overseeing the review of the manuscript were John B. Buse and Jennifer B. Green.

1.
Yapanis
M
,
James
S
,
Craig
ME
,
O’Neal
D
,
Ekinci
EI
.
Complications of diabetes and metrics of glycemic management derived from continuous glucose monitoring
.
J Clin Endocrinol Metab
2022
;
107
:
e2221
e2236
2.
Tang
X
,
Li
S
,
Wang
Y
, et al
.
Glycemic variability evaluated by continuous glucose monitoring system is associated with the 10-y cardiovascular risk of diabetic patients with well-controlled HbA1c
.
Clin Chim Acta
2016
;
461
:
146
150
3.
Shukla
AP
,
Iliescu
RG
,
Thomas
CE
,
Aronne
LJ
.
Food order has a significant impact on postprandial glucose and insulin levels
.
Diabetes Care
2015
;
38
:
e98
e99
4.
Shukla
AP
,
Andono
J
,
Touhamy
SH
, et al
.
Carbohydrate-last meal pattern lowers postprandial glucose and insulin excursions in type 2 diabetes
.
BMJ Open Diabetes Res Care
2017
;
5
:
e000440
5.
Ferguson
BK
,
Wilson
PB
.
Ordered eating and its effects on various postprandial health markers: a systematic review
.
J Am Nutr Assoc
2023
;
42
:
746
757
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.