Impact of Fat, Protein, andGlycemic Index on PostprandialGlucoseControl inType 1Diabetes:Implications for Intensive DiabetesManagement in the ContinuousGlucose Monitoring EraDiabetes Care 2015;38:1008–1015 | DOI: 10.2337/dc15-0100

BACKGROUND

Continuous glucose monitoring highlights the complexity of postprandial glucosepatterns present in type 1 diabetes and points to the limitations of currentapproaches tomealtime insulin dosing based primarily on carbohydrate counting.

METHODS

A systematic review of all relevant biomedical databases, including MEDLINE,Embase, CINAHL, and the Cochrane Central Register of Controlled Trials, wasconducted to identify research on the effects of dietary fat, protein, and glycemicindex (GI) on acute postprandial glucose control in type 1 diabetes and prandialinsulin dosing strategies for these dietary factors.

RESULTS

All studies examining the effect of fat (n = 7), protein (n = 7), and GI (n = 7) indicatedthat these dietary factors modify postprandial glycemia. Late postprandial hyper-glycemia was the predominant effect of dietary fat; however, in some studies,glucose concentrations were reduced in the first 2–3 h, possibly due to delayedgastric emptying. Ten studies examining insulin bolus dose and delivery patternsrequired for high-fat and/or high-protein meals were identified. Because of meth-odological differences and limitations in experimental design, study findings wereinconsistent regarding optimal bolus delivery pattern; however, the studies in-dicated that high-fat/protein meals require more insulin than lower-fat/proteinmeals with identical carbohydrate content.

CONCLUSIONS

These studies have important implications for clinical practice and patient edu-cation and point to the need for research focused on the development of newinsulin dosing algorithms based onmeal composition rather than on carbohydratecontent alone.

1CharlesPerkinsCentreand theSchool ofMolecularBioscience, The University of Sydney, Sydney,Australia2Joslin Diabetes Center, Boston, MA3Department of Paediatric Endocrinology andDiabetes, John Hunter Children’s Hospital,Newcastle, Australia4Hunter Medical Research Institute, School ofMedicine and Public Health, University ofNewcastle, Rankin Park, Australia5Children’s Hospital, Boston, MA6Harvard Medical School, Boston, MA

Corresponding author: Howard A. Wolpert,howard.wolpert@joslin.harvard.edu.

Received 15 January 2015 and accepted 26February 2015.

This article contains Supplementary Data onlineat http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc15-0100/-/DC1.

© 2015 by the American Diabetes Association.Readers may use this article as long as the workis properly cited, the use is educational and notfor profit, and the work is not altered.

See accompanying articles, pp. 968,971, 979, 989, 997, 1016, 1030, and1036.

Kirstine J. Bell,1,2 Carmel E. Smart,3,4

Garry M. Steil,5,6 Jennie C. Brand-Miller,1

Bruce King,3,4 and Howard A. Wolpert2,6

1008 Diabetes Care Volume 38, June 2015

TYPE1DIABETES

ATACROSSROADS

Currently, carbohydrates are consid-ered the predominant macronutrientaffecting postprandial glucose controland the primary determinant for calcu-lating mealtime insulin doses in type 1diabetes (1). Emerging evidence fromrecent research and the use of continu-ous glucose monitoring have shownthat other nutritional properties offood, including fat, protein, and glyce-mic index (GI), can significantly affectpostprandial glucose excursions. Thesefindings highlight the need for alterna-tive mealtime insulin dosing algorithmsand have important implications for nu-trition education and counseling in pa-tients with diabetes.The American Diabetes Association

(ADA) recommends that people with di-abetes who have mastered carbohy-drate counting receive education onthe glycemic impacts of protein andfat (1). To our knowledge, neither acomprehensive review of the literatureexamining the relative effects of pro-tein, fat, and GI on postprandial glyce-mia nor guidelines for clinicians on howinsulin doses should be adjusted in type1 diabetes for various meal composi-tions exist.This article reviews the evidence ad-

dressing the following questions:

1. What effects do GI, protein, and fathave on acute postprandial glucoseconcentrations in type 1 diabetes?

2. What prandial insulin dosing strate-gies work best for GI, protein, and fatin type 1 diabetes?

In light of the evidence reviewed inthese questions and our clinical insights,this article also discusses the followingquestions:

3. What are the implications for clinicalpractice?

4. What are the knowledge gaps, andhow can technology be leveraged toimprove postprandial glucose control?

Current clinical approaches to inten-sive diabetes management tend to beinsulin-centric; however, a focus on di-etary quality and mealtime routine,with referral to a registered dietitianfor medical nutrition therapy, may bejust as important for optimizing glyce-mic control (1,2). In the coming years, ascontinuous glucose monitoring becomes

the standard of care in the managementof type 1 diabetes (3), the challenges ofkeeping postprandial glucose concentra-tions in rangewill become an inescapableand increasing focus in the daily lives ofpeople with diabetes.

SYSTEMATIC REVIEW PROCEDURE

For questions 1 and 2, a search of allrelevant biomedical databases was con-ducted, including MEDLINE, Embase,CINAHL (Cumulative Index to Nursing andAllied Health Literature), Thomson Reu-ters Web of Science (formerly ISI Web ofKnowledge), and the Cochrane CentralRegister of Controlled Trials. The follow-ing key words were used to identifypotentially relevant articles: “type 1diabetes” AND “blood glucose” OR “in-sulin” AND “dietary carbohydrate” OR“dietary protein” OR “dietary fat” OR“glycemic index.” The search strategywas deliberately broad to identify all rel-evant articles to answer the two reviewquestions.

Original controlled studies were in-cluded if they were published in Englishbetween March 1995 and November2014, conducted in adults and/or chil-dren with type 1 diabetes, used rapid-acting insulin analogs (lispro, aspart,or glulisine), and 1) compared post-prandial glycemia following differentfoods/meals with the same insulin dos-ing strategy or 2) different prandial dos-ing strategies for the same foods/meals.A start date of March 1995 was inten-tionally chosen because this was whenthe first rapid-acting insulin analog (lis-pro) was available in the U.S. Studies ofparticipants of any age, sex, or racewere included. Studies where both themeal and the insulin dose were concur-rently altered or the meal compositionwas not controlled and studies in preg-nancy or involving exercise wereexcluded.

The literature search identified 143studies, and 13 more studies were iden-tified through hand-searching refer-ence lists (Supplementary Fig. 1). Ofthese studies, 127 were excluded pri-marily due to duplication (n = 23) andnot meeting the inclusion criteria forthis review, including 21 in type 2 dia-betes, 6 comparing various types of in-sulin or regimens, and 4 varying insulindosing and meals concurrently. The re-maining 29 studies were included in thisreview.

QUESTION 1: WHAT EFFECTS DOGI, PROTEIN, AND FAT HAVE ONACUTE POSTPRANDIAL GLUCOSECONCENTRATIONS IN TYPE 1DIABETES?

As summarized in Table 1, 16 studiesexamining the effects of nutritional fac-tors on postprandial glycemia wereidentified (4–19) (see SupplementaryTable 1 for more details).

FatSeven studies examining the effect ofdietary fat on postprandial glycemia in103 subjects with type 1 diabetes wereidentified (5,7,8,11,15,18,19). All stud-ies added fat to the test meal, whichconcurrently increased the energy con-tent of the test meal. The amount of di-etary fat added to the control mealsranged from 6.6 to 52 g.

All studies reported that dietary fatmodified postprandial glycemia. Onestudy did not find an increase in glucoseconcentration (5) possibly because thepostprandial monitoring period wasonly 3 h. Two studies reported that theaddition of dietary fat reduced the areaunder the curve in the first 2–3 h (7,8).This may be due to fat delaying gastricemptying, as indicated by one of thestudies demonstrating that the gastricemptying rate was significantly reducedin the first 2 h after consuming a high-fatmeal (7). This in turn delayed the rise inpostprandial glycemia, with four studiesreporting a lag in time to peak bloodglucose concentrations between 6 and90 min (mean 29 min) (5,7,15,18).

Evidence suggests thatmeals contain-ing carbohydrates and that are high indietary fat cause sustained late post-prandial hyperglycemia. One studyshowed the addition of 35 g dietary fatsignificantly increased postprandial glu-cose concentrations by 2.3 mmol/L at 5h (15). Wolpert et al. (19) demonstratedthat the addition of 50 g fat caused sig-nificant hyperglycemia over 5 h, evenwhen additional insulin was adminis-tered using a closed-loop glucose con-trol system. Free fatty acids (FFAs)directly induce insulin resistance, andone study postulated that the mecha-nism for the delayed hyperglycemic ef-fect of dietary fat is FFA-induced insulinresistance with increased hepatic glu-cose output (20). Consistent with theobserved time course of hyperglycemiafollowing higher-fat meals in type 1

care.diabetesjournals.org Bell and Associates 1009

diabetes, Gormsen et al. (21) demon-strated in healthy subjects that insulinresistance develops at least 120 min af-ter circulating FFAs increase.

ProteinSeven studies examining the effect ofdietary protein on postprandial glyce-mia in 125 subjects with type 1 diabetes(range 8–33 subjects) were identified(5,8,11,13,15–17). All studies kept car-bohydrate content consistent, with sixadding protein to the test meal, therebyconcurrently increasing the energy con-tent. Winiger et al. (17) was the onlystudy to keep energy levels constant;however, to keep energy and carbohy-drates constant, both protein and fatwere simultaneously varied.All included studies reported that

postprandial glycemia was modified bythe addition of protein, with all but one(5) reporting significant differences.Smart et al. (15) reported that the addi-tion of 35 g protein to 30 g carbohy-drates significantly increased bloodglucose levels by 2.6 mmol/L at 5 h.

The effect of fat and protein was addi-tive, with blood glucose concentrationsincreasing by 5.4 mmol/L at 5 h, the sumof the individual incremental increasesfor protein and fat. Paterson et al. (13)was the only study examining the effectof protein only (in the absence of carbo-hydrates and fat) and found that theaddition of 12.5–50 g protein did notsignificantly affect glycemia, althoughthe addition of 75 and 100 g significantlyincreased glucose concentrations,reaching a peak at the conclusion ofthe 5-h study and causing an increasein glucose concentrations similar tothat of 20 g carbohydrates givenwithoutinsulin. These results suggest that pro-tein has differential effects when con-sumed with and without carbohydrates.

All studies agreed that protein affectsblood glucose concentrations in the latepostprandial period. When protein wasthe only macronutrient consumed, glu-cose concentrations began to rise after100 min for protein loads of $75 g(13). For meals where carbohydrateswere also consumed, increased glucose

concentrations were noted after 3–4 h(11,15,17).

Glycemic IndexSeven studies examining the effects ofGI on postprandial glycemia in 98 sub-jects with type 1 diabetes (range 8–20subjects) were identified (4,6,8–10,12,14). All studies compared higher-versus lower-GI foods or meals. Twostudies concurrently varied the energyand macronutrient compositions of thetest meals, confounding the interpreta-tion (4,8).

All seven studies reported significantdifferences in blood glucose concentra-tions, with low-GI foods and meals pro-ducing lower glycemic responses. Threestudies suggested that the risk of mildhypoglycemia is greater with low-GIthan with high-GI foods (6,9,10); how-ever, the timing of the episodes wasnot reported. One study suggestedthat low-GI foods are more likely tocause early hypoglycemia, reporting acorrelation between GI and time to hy-poglycemia, with each unit increase in GI

Table 1—Summary of systematic review: effects of fat, protein, and GI on acute postprandial glycemia in type 1 diabetes

Nutritional factor Summary of findings Clinical implications

Fat c Seven studies (total 103 subjects) (5,7,8,11,15,18,19).c All studies reported significant differences inglycemia with addition of fat.

c Fat reduces early glucose response (first 2–3 h) (7,8) anddelays peak blood glucose (5,7,15,18) dueto delayed gastric emptying.

c Fat leads to late postprandial (.3 h) hyperglycemia(18,19).

c Addition of 35 g fat can increase blood glucose by2.3 mmol/L (15), and in some individuals, 50 g of fatcan increase insulin requirements by twofold (19).

c Marked interindividual differences in the glycemiceffect of fat.

c Increase in dose required for coverage of higher-fatmeals needs to be individualized.

c Delicate balance in calculation and timing of insulinaction: needs more insulin to prevent late postprandialhyperglycemia; however, if too much insulin upfront,there is a risk for early postprandial hypoglycemia.

Protein c Seven studies (total 125 subjects) (5,8,11,13,15–17).c All studies reported significant differences inglycemia with addition of protein.

c Effect of protein is delayed (effects seen ;100 minpostmeal) (11,13,15,17).

c Protein has different effects when consumedwith and without carbohydrates [e.g., 30 g proteinwith carbohydrates will affect blood glucose (15,16),whereas at least 75 g protein is needed to see aneffect when consumed in isolation (13)].

c Protein-only meals (e.g., $230 g lean steak with salad)may require a different insulin dosing strategy than forprotein and carbohydrate meals.

GI c Seven studies (total 98 subjects) (4,6,8–10,12,14).c All studies reported significant differences inglycemia with differing GI (same carbohydrate).

c High-GI foods have rapid glucose spike (9,14).c Low-GI foods lower overall glucose response(8–10,12,14), reduce glucose peak (4,9,14), andincrease risk of hypoglycemia (when usual CIR isused) (6,9,10).

c Mismatch between insulin action and carbohydrateabsorption following high-GI foods can be problematic,leading to a rapid glucose spike.

c Total carbohydrate content still important: a largecarbohydrate serving of low-GI food will still causelarge glycemic response.

c Low-GI foods with high fructose and/or sucrose content(e.g., fruit juice) will still produce a rapid glucose spike.

See Supplementary Table 1 for details.

1010 Postprandial Glycemia and Insulin Dosing in T1D Diabetes Care Volume 38, June 2015

delaying hypoglycemia by 1 min (9). Thisis consistent with the reduced overallglycemic response with low-GI foods/meals.

QUESTION 2: WHAT PRANDIALINSULIN DOSING STRATEGIESWORK BEST FOR GI, PROTEIN, ANDFAT IN TYPE 1 DIABETES?

Fifteen studies examining the effects ofvarying prandial dosing strategies, in-cluding various bolus types, timing ofthe meal bolus, and methods to calcu-late the bolus dose, on postprandial gly-cemia were identified (14,19,22–34)(Supplementary Table 2).

Fat and ProteinTen studies examining insulin bolusdose and delivery patterns and timingrequired to cover meals high in fatand/or protein were identified (19,22–24,26–30,33).

Insulin Dose

As summarized in Table 2, six studiesinvestigated the additional insulin re-quirements for food or meals high infat or protein (19,22,23,28,29,33). Theamount of additional insulin variedacross studies depending on themethod used to estimate the bolus in-sulin requirements. Lee et al. (29) used apredetermined value of 50% of thecarbohydrate-to-insulin ratio (CIR) to

calculate fat-to-insulin and protein-to-insulin ratios, with the additional insulindelivered as an extended bolus. Simi-larly, another two studies (28,33) calcu-lated additional insulin for fat-proteinunits (FPUs) (defined as 100 kcal or420 kJ fat and/or protein) (35), withthe duration of the extended bolus ad-justed from 3 to 6 h depending on thenumber of FPUs consumed. A major lim-itation of this equation was the high rateof clinically significant hypoglycemiacompared with carbohydrate counting(28,33). Another study using closed-loop glucose control measured theadditional insulin required to maintainglucose control after a high-fat meal(19). This study revealed that the addi-tion of 50 g of fat markedly increasedinsulin requirements over the 5-h post-prandial period, with one subjectneeding a more than twofold increase.This study identified marked interindi-vidual differences in fat sensitivity, high-lighting the need for individualizedincremental dosing for fat and pointingto the limitations of equations that cal-culate individual insulin dosing for fatand protein as a proportion of the indi-vidual’s CIR (28,29,33).

Two studies investigated anothernovel insulin dosing algorithm, theFood Insulin Index (FII), to calculatethe insulin demand for individual foods

(23) and mixed meals (22). The FII isbased on the physiological insulin de-mand evoked by 1,000-kJ food portionsas measured in healthy subjects. Be-cause food energy is the constant, themethod accounts for all nutritional andmetabolic factors that influence insulindemand, not just the macronutrientcontent. With this method, bolus insulinis calculated for foods not traditionallycovered by insulin, including eggs andsteak. Both studies reported signifi-cantly improved glycemia in the 3-hpostprandial period. However, the rela-tively short postprandial monitoring pe-riod may not have detected the delayedglycemic impact of fat and protein. Therates of hypoglycemia were high in boththe FII and the carbohydrate-countinggroups in one study (23), suggestingthat the initial CIR and basal rates werenot adequately optimized at study com-mencement. A pilot study using the FII inpractice did not find an increased risk ofhypoglycemic events (36).

Bolus Timing and Type

Whereas all the studies in the literatureexamining the optimal timing of theprandial bolus (25,26,34) uniformlydemonstrated that delivering a bolus15–20 min before eating rather thanimmediately before or after the mealsignificantly improves postprandialglycemia, studies examining the optimalbolus type required to cover higher-fatmeals showed discrepant results. Fourstudies reported that the combo wavebolus was the most effective method(24,27,29,33), whereas one suggestedthat the standard bolus with insulin de-livered 15 min before the meal was su-perior (26) and another reported nodifferences in glucose excursions amongnormal, combo wave, and extended bo-luses after pasta meals (30).

Methodological differences and is-sues confound the evaluation of thepublished literature. The duration andsplit of the bolus types varied acrossstudies. Furthermore, failure to opti-mize basal rates as part of the study pro-tocol could have affected the findings ofsome studies. In addition, differences inthe GI and macronutrient content of thetest meals and duration and split of thebolus types in the studies may have con-tributed to the differences in the results.Of note, some studies had a relativelyshort postprandial monitoring period

Table 2—Summary of systematic review: studies evaluating prandial insulindosing equations and requirements for high-fat meals with or without protein

Methoda Limitation

Two studies (28,33) calculated additionalinsulin for fat and protein using a complexdosing equation with FPUs (35) and thesubject’s CIR.

High rate of clinically significantpostprandial hypoglycemiacompared with carbohydratecounting.

One study (29) used a predetermined value of50% of the CIR to calculate fat-to-insulinand protein-to-insulin ratios and added theadditional insulin to extended bolus.

Suboptimal postprandial glucosecontrol.

Two studies (22,23) used the FII to calculateadditional insulin for high-fat and/or-protein meals. Both studies reportedsignificantly improved glycemia in the3-h postprandial period.

Because of the short postprandialmonitoring period, may havefailed to detect delayed impactof fat and protein on glucoseconcentrations.

One study (19) used closed-loop control todetermine additional insulin required tomaintain postprandial glycemic controlafter a high-fat meal; definitivelyestablished that high-fat meals need moreinsulin coverage than lower-fat mealswith identical carbohydrate content.

Findings need to be translated intoa dosing equation for use inclinical practice.

See Supplementary Table 2 for details. aMethod used to calculate the additional insulin and theamount given varied across studies.

care.diabetesjournals.org Bell and Associates 1011

and, thus, may not have detected thedelayed hyperglycemia from dietary fat.

Glycemic IndexTwo studies examining insulin dosingstrategies to improve postprandial gly-cemia for low-GI meals were identified(14,32). One study in patients using apump demonstrated that a 50:50%combo wave bolus over 2 h resultedin a 47% decrease in the postprandialglucose area under the curve comparedwith a standard bolus for a low-GI meal(32). Another study examined bolusstrategies for low-GI meals in childrenusing multiple daily injections and foundthat insulin administration 15min beforethe meal resulted in better postprandialglycemic control than insulin administra-tion 15 min after the meal (14).No studies investigated changes in in-

sulin dose for meals of a higher GI. Themismatch between the action of analoginsulins and the rapid glucose spikecaused by high-GI meals remains a clini-cal challenge, and practical insulin dosingstrategies are needed to address this.

QUESTION 3: WHAT ARE THEIMPLICATIONS FOR CLINICALPRACTICE?

From the findings of this review, GI, pro-tein, and fat substantially modulatepostprandial glucose concentrations inindividuals with type 1 diabetes. Theimpact on 3-h postprandial glucoseconcentrations of the addition of 35 gfat and 40 g protein to a meal (37) isequivalent to that resulting from theconsumption of 20 g carbohydrateswithout insulin (13). The addition of 50 gfat to a meal can increase insulin re-quirements by more than twofold (19).This demonstrates that to optimize post-prandial glucose control, somemealtimeinsulin doses need to be adjusted basedon the complete meal compositionrather than solely on the carbohydratecontent.Whereas meals high in fat and protein

require more insulin to control latepostprandial hyperglycemia thanlower-fat and -protein meals with thesame carbohydrate content, as outlinedin the previous section, the actualamount of additional insulin and deliv-ery pattern required for high-fat or high-protein meals is not clear. Preliminarystudies have indicated that in someindividuals, a combowave bolus extend-ing over 2–2.5 h with as much as a 125%

increase in the total insulin dose is re-quired for meals containing 40 g satu-rated fat (38). As outlined in Fig. 1, anempirical approach is necessary to iden-tify an individual’s nutrient sensitivitiesand optimize insulin dosing for complexmeals. As a starting point, initial insulindoses are calculated based on the car-bohydrate content of the meal and theindividual’s CIR, with subsequent incre-mental dose increases guided by retro-spective review of the patient’s previouspostprandial glucose responses. This re-view of records can also identify alter-native favorite foods with less glycemiceffect, and these insights can be help-ful in patient education and nutritioncounseling.

Of note, in covering higher-fat meals,the amount of insulin given upfrontneeds to be adjusted to minimize the

risk for early postprandial hypoglycemiafrom delayed gastric emptying inducedby the dietary fat. Based on physiologi-cal considerations, it seems likely thatthe percentage of the total insulindose delivered during the initial phaseof a combo wave bolus for higher-fatmeals will also need to vary dependingon the GI of the meal (more insulin up-front for higher-GI carbohydrate loads).However, as outlined in the previoussection, limited scientific data are avail-able regarding the optimal split and du-ration of the advanced pump boluses forvarious meal types.

In individuals on multiple dose insulintherapy, strategies frequently sug-gested in clinical practice for high-fat/low-GI meals (e.g., pizza) are to injectan additional insulin bolus 1 h after themeal to match the delayed absorption

Figure 1—Clinical application of insights about the effects of fat, protein, and GI on postprandialglucose control.

1012 Postprandial Glycemia and Insulin Dosing in T1D Diabetes Care Volume 38, June 2015

or, alternatively, to cover the mealwith a preprandial injection of regularwith or without analog insulin. How-ever, these approaches have not beeninvestigated in controlled studies.Interindividual differences in the im-

pact of macronutrients on postprandialglucose control are another factor com-pounding the development of stan-dardized insulin dosing algorithms forcomplex meals (15,19). To date, studieshave not been able to identify pheno-type markers that correlate with interin-dividual differences in the glycemiceffect of dietary fat. Furthermore, thesedifferences do not appear to correlatewith the CIR, and there is no scientificfoundation for the use of the CIR as thebasis for calculating incremental insulindoses to cover fat [as has been sug-gested by others (33)].The mismatch between the rapid glu-

cose absorption following higher-GImeals and the relatively delayed actionof subcutaneously administered insulinleads to an initial postprandial glycemicspike that in practice can be difficult toovercome. As indicated in the ADAguidelines, substituting low–glycemicload foods for higher–glycemic loadfoods may modestly improve glycemiccontrol (2). Increasing the insulin doseto reduce the spike can lead to overin-sulinization in the late postprandial pe-riod with an associated increased riskfor hypoglycemia. Use of a superbolus(an increased insulin bolus upfront fol-lowed by basal rate reduction) has beenproposed to provide a better match be-tween insulin action and glucose ab-sorption following higher-GI meals.Technosphere insulin (Afrezza), the in-haled prandial insulin recently releasedin the U.S., has an action profile thatmay allow for better coverage ofhigher-GI carbohydrates than insulinanalogs delivered subcutaneously (39).The timing of the insulin dose in re-

lation to the meal is an important factorin optimizing postprandial control. Earlydelivery of a preprandial insulin bolusfor higher-GI carbohydrate loads is an-other strategy that can be at least par-tially effective in blunting the initialpostprandial spike. In this context, itshould be mentioned that unless the in-dividual has gastroparesis, preprandial in-sulin is preferable to insulin administeredduringor after themeal for allmeal types,including low-GI and high-fat meals.

Finally, clinical education and recom-mendations should reinforce healthyeating principles and not simply focuson insulin adjustments. In clinical prac-tice, it may bemore pertinent to addressissues in dietary quality and meal-time routines to improve postprandialcontrol rather than to just focus onfine-tuning mealtime insulin therapy. Aregistered dietitian can provide a thor-ough assessment and medical nutritiontherapy (1,2).

QUESTION 4: WHAT ARE THEKNOWLEDGE GAPS, AND HOWCAN TECHNOLOGY BE LEVERAGEDTO IMPROVE POSTPRANDIALGLUCOSE CONTROL?

Further research in this area is crucial toprovide a stronger evidence base forclinical decision-making. Table 3 out-lines some of the issues of practical im-portance to address in future research.This increased knowledge about the im-pact of macronutrients on postprandialglucose control and improved insulindosing algorithms will be of practical rel-evance only if a pathway for translatingthe research findings into clinical care isestablished.

Digital health tools and cloud com-puting will open opportunities to de-velop sophisticated analytic systems toautomatically evaluate postprandial glu-cose data and provide dosing recom-mendations. Carbohydrate countingis a challenging aspect to diabetes self-management, and requiring that fat andprotein intake also be quantitated andincorporated in insulin dosing decisionswill create an additional burden that fewpatients will be able to accomplish. Theneed for both practical simplicity andwidespread use of advanced dosing al-gorithms will ultimately be resolvedwith the development of data analysisand decision-support tools that evaluate

meal patterns to identify whether mac-ronutrients are contributing to glycemicfluctuations and to provide individual-ized dosing recommendations to pa-tients for their common meals, therebyeliminating the need for patients to rou-tinely count carbohydrates and othermacronutrients.

An opportunity exists to capitalize onthe control algorithms used in artificialpancreas (AP) technology to optimizecurrent open-loopmealtime insulin dos-ing. Although AP technology ultimatelyhas the potential to remove any need tocalculate insulin dosing, considerableregulatory and technical hurdles mustbe overcome before an AP system willbecome a routine tool for the manage-ment of type 1 diabetes. Going backalmost a decade, methods based on pro-portional integral derivative controlstrategies have been used to effectchanges in basal insulin delivery withimproved fasting glucose control(19,40,41). Metabolic models for pre-dicting future glucose excursions havebeen developed for use with AP systems(36), and the same approach can be usedto analyze postprandial glucose patternsand provide adaptive open-loop bolusdose recommendations (38).

The introduction of continuous glu-cose monitoring into diabetes care hasrevealed the complex picture of post-prandial glucose patterns in type 1 dia-betes, including rapid glucose spikesfrom higher-GI carbohydrates and latepostprandial hyperglycemia from dietaryfat and protein, and highlights the limi-tations of the traditional carbohydrate-based approaches for mealtime insulindosing. New insights about the effectof dietary macronutrients on post-prandial glucose control has been in-cluded in the recent ADA standardsof medical care, which mention that“for selected individuals. . .the impact

Table 3—Unanswered questions about the effect of dietary fat and protein onpostprandial glucose control in type 1 diabetesHow much fat does there need to be in a meal before a clinically significant glycemic effect

becomes apparent?

Is there a threshold and/or dose response (i.e., more fat requires more insulin)?

Do all types of fat and protein have similar effects?

Are there phenotypic characteristics that can be used as markers to identify individualswith diabetes who are more nutrient sensitive and will require more insulin to coverhigher-fat/protein meals?

What are the optimal insulin dose adjustments needed for common meals with varying fatand protein content?

care.diabetesjournals.org Bell and Associates 1013

of protein and fat on glycemic excursionscan be incorporated into diabetes man-agement” (1). At a practical level, in re-viewing patient glucose records, thepossible role of dietary fat and proteinas contributors to glycemic fluctuationsneeds to be considered. Although thisreview shows that high-fat meals requiremore insulin coverage than lower-fatmeals with identical carbohydrate con-tent and provides some indication ofthe additional insulin dosages de-manded for higher-fat foods (19), wedo not yet have simple and easy-to-useinsulin dosing algorithms for dietary fat.Ongoing research in conjunction with thedevelopment of digital tools to as-sist patients in dosage decision-makingpromise solutions to address this needfor a better approach to optimizing post-prandial insulin coverage that could bereadily and widely applied in the man-agement of type 1 diabetes.

Duality of Interest. No potential conflicts ofinterest relevant to this article were reported.Author Contributions. K.J.B. contributed tothe conception and design of the study, col-lected and interpreted the data, contributedto the writing of the first draft of the manuscriptand subsequent revisions, critically reviewedthe drafts of the manuscript, and contributedto intellectual content. C.E.S. contributed tothe design of the study, collected and inter-preted the data, contributed to the writing ofthe first draft of the manuscript, criticallyreviewed the drafts of the manuscript, andcontributed to intellectual content. G.M.S.contributed to the conception and design ofthe study, contributed to the writing of themanuscript, critically reviewed the drafts ofthe manuscript, and contributed to intellectualcontent. J.C.B.-M. and B.K. critically reviewedthe drafts of the manuscript and contributedto intellectual content. H.A.W. conceived anddesigned the study, contributed to the data in-terpretation, contributed to the writing of thefirst draft of the manuscript and subsequentrevisions, critically reviewed the drafts of themanuscript, and contributed to intellectualcontent.

References1. American Diabetes Association. Approachesto glycemic treatment. Sec. 7. In Standards ofMedical Care in Diabetesd2015. Diabetes Care2015;38(Suppl. 1):S41–S482. American Diabetes Association. Foundationsof care: education, nutrition, physical activity,smoking cessation, psychosocial care, and immu-nization. Sec. 4. In Standards of Medical Care inDiabetesd2015. Diabetes Care 2015;38(Suppl.1):S20–S303. Klonoff DC, Buckingham B, Christiansen JS,et al.; Endocrine Society. Continuous glucosemonitoring: an Endocrine Society Clinical

Practice Guideline. J Clin Endocrinol Metab2011;96:2968–29794. Elleri D, Allen JM, Harris J, et al. Absorptionpatterns of meals containing complex carbohy-drates in type 1 diabetes. Diabetologia 2013;56:1108–11175. Garc ıa-Lopez JM, Gonzalez-Rodriguez M,Pazos-Couselo M, Gude F, Prieto-Tenreiro A,Casanueva F. Should the amounts of fat andprotein be taken into consideration to calculatethe lunch prandial insulin bolus? Results from arandomized crossover trial. Diabetes TechnolTher 2013;15:166–1716. Lafrance L, Rabasa-Lhoret R, Poisson D,Ducros F, Chiasson JL. Effects of different gly-caemic index foods and dietary fibre intake onglycaemic control in type 1 diabetic patients onintensive insulin therapy. Diabet Med 1998;15:972–9787. Lodefalk M, Aman J, Bang P. Effects of fatsupplementation on glycaemic response andgastric emptying in adolescents with type 1 di-abetes. Diabet Med 2008;25:1030–10358. MacDonald K, Lowe JM, Barker D,MenschM,Attia J. Effect of popular takeaway foods onblood glucose levels in type 1 diabetes mellituspatients on intensive insulin therapy. Int J ClinPract 2009;63:189–1949. Mohammed NH, Wolever TMS. Effect of car-bohydrate source on post-prandial blood glu-cose in subjects with type 1 diabetes treatedwith insulin lispro. Diabetes Res Clin Pract2004;65:29–3510. Nansel TR, Gellar L, McGill A. Effect of vary-ing glycemic index meals on blood glucosecontrol assessed with continuous glucose mon-itoring in youth with type 1 diabetes on basal-bolus insulin regimens. Diabetes Care 2008;31:695–69711. Neu A, Behret F, Braun R, et al. Higher glu-cose concentrations following protein- and fat-richmeals - the Tuebingen Grill Study: a pilot studyin adolescents with type 1 diabetes. Pediatr Di-abetes 20 October 2014 [Epub ahead of print]12. Parillo M, Annuzzi G, Rivellese AA, et al. Ef-fects of meals with different glycaemic index onpostprandial blood glucose response in patientswith type 1 diabetes treated with continuoussubcutaneous insulin infusion. Diabet Med2011;28:227–22913. Paterson MA, Smart CE, McElduff P, et al.Influence of pure protein on postprandialblood glucose levels in individuals with type1 diabetes mellitus (Abstract). Diabetes 2014;63:A1514. Ryan RL, King BR, Anderson DG, Attia JR,Collins CE, Smart CE. Influence of and optimalinsulin therapy for a low-glycemic index meal inchildrenwith type 1 diabetes receiving intensiveinsulin therapy. Diabetes Care 2008;31:1485–149015. Smart CEM, Evans M, O’Connell SM, et al.Both dietary protein and fat increase postpran-dial glucose excursions in children with type 1diabetes, and the effect is additive. DiabetesCare 2013;36:3897–390216. Uthoff H, Lehmann R, Sprenger M, WiesliP. Skipping meals or carbohydrate-free mealsin order to determine basal insulin require-ments in subjects with type 1 diabetes melli-tus? Exp Clin Endocrinol Diabetes 2010;118:325–327

17. Winiger G, Keller U, Laager R, Girard J,Berger W. Protein content of the eveningmeal and nocturnal plasma glucose regulationin type-I diabetic subjects. Horm Res 1995;44:101–10418. Wolever TMS, Mullan YM. Sugars and fathave different effects on postprandial glucoseresponses in normal and type 1 diabetic sub-jects. Nutr Metab Cardiovasc Dis 2011;21:719–72519. Wolpert HA, Atakov-Castillo A, Smith SA,Steil GM. Dietary fat acutely increases glucoseconcentrations and insulin requirements inpatients with type 1 diabetes: implications forcarbohydrate-based bolus dose calculation andintensive diabetes management. Diabetes Care2013;36:810–81620. Roden M, Price TB, Perseghin G, et al.Mechanism of free fatty acid-induced insulinresistance in humans. J Clin Invest 1996;97:2859–286521. Gormsen LC, Nielsen C, Jessen N, JørgensenJO, Møller N. Time-course effects of physio-logical free fatty acid surges on insulin sensi-tivity in humans. Acta Physiol (Oxf) 2011;201:349–35622. Bao J, Gilbertson HR, Gray R, et al. Improv-ing the estimation of mealtime insulin dose inadults with type 1 diabetes: the Normal InsulinDemand for Dose Adjustment (NIDDA) study.Diabetes Care 2011;34:2146–215123. Bell KJ, Gray R, Munns D, et al. Estimatinginsulin demand for protein-containing foods us-ing the food insulin index. Eur J Clin Nutr 2014;68:1055–105924. Chase HP, Saib SZ, MacKenzie T, HansenMM, Garg SK. Post-prandial glucose excursionsfollowing four methods of bolus insulin admin-istration in subjects with type 1 diabetes. DiabetMed 2002;19:317–32125. Cobry E, McFann K, Messer L, et al. Timingof meal insulin boluses to achieve optimal post-prandial glycemic control in patients with type 1diabetes. Diabetes Technol Ther 2010;12:173–17726. De Palma A, Giani E, Iafusco D, et al. Low-ering postprandial glycemia in children withtype 1 diabetes after Italian pizza “margherita”(TyBoDi2 Study). Diabetes Technol Ther 2011;13:483–48727. Jones SM, Quarry JL, Caldwell-McMillan M,Mauger DT, Gabbay RA. Optimal insulin pumpdosing and postprandial glycemia following apizza meal using the continuous glucose moni-toring system. Diabetes Technol Ther 2005;7:233–24028. Kordonouri O, Hartmann R, Remus K, BlasigS, Sadeghian E, Danne T. Benefit of supplemen-tary fat plus protein counting as compared withconventional carbohydrate counting for insulinbolus calculation in childrenwith pump therapy.Pediatr Diabetes 2012;13:540–54429. Lee SW, CaoM, Sajid S, et al. The dual-wavebolus feature in continuous subcutaneous insu-lin infusion pumps controls prolonged post-prandial hyperglycaemia better than standardbolus in type 1 diabetes. Diabetes Nutr Metab2004;17:211–21630. Lindholm-Olinder A, Runefors J, Smide B,Kernell A. Post-prandial glucose levels followingthree methods of insulin bolusing: a study inadolescent girls and in compliance with girls

1014 Postprandial Glycemia and Insulin Dosing in T1D Diabetes Care Volume 38, June 2015

without diabetes. Pract Diabetes Int 2009;26:110–11531. Lopez P, Smart C, Morbey C, McElduff P,Paterson M, King BR. Extended insulin bolusescannot control postprandial glycemia as wellas a standard bolus in children and adults usinginsulin pump therapy. BMJ Open Diab Res Care2014;2:e000005032. O’Connell MA, Gilbertson HR, Donath SM,Cameron FJ. Optimizing postprandial glycemia inpediatric patientswith type1diabetes using insulinpump therapy: impact of glycemic index and pran-dial bolus type. Diabetes Care 2008;31:1491–149533. Pankowska E, Błazik M, Groele L. Does thefat-protein meal increase postprandial glucoselevel in type 1 diabetes patients on insulinpump: the conclusion of a randomized study.Diabetes Technol Ther 2012;14:16–22

34. Scaramuzza AE, Iafusco D, Santoro L, et al.Timing of bolus in children with type 1 diabetesusing continuous subcutaneous insulin infusion(TiBoDi Study). Diabetes Technol Ther 2010;12:149–15235. Pankowska E, Szypowska A, Lipka M,Szpotanska M, Błazik M, Groele L. Applicationof novel dual wave meal bolus and its impact onglycated hemoglobin A1c level in children withtype 1 diabetes. Pediatr Diabetes 2009;10:298–30336. Bell K, Gray R, Munns D, et al. Food InsulinIndex (FII) vs. traditional carbohydrate countingfor glycemic control in adults with type 1 diabe-tes: a 3-month pilot study (Abstract). Diabetes2014;63(Suppl.):A18937. Smart CE, Ross K, Edge JA, King BR,McElduff P,Collins CE. Can children with Type 1 diabetes and

their caregivers estimate the carbohydrate contentofmeals and snacks?DiabetMed2010;27:348–35338. Bell K, Toschi E, Wolpert H, Steil GM. Use ofmodel predicted bolus estimation to optimizeinsulin dosing strategies covering high fat meals(Abstract). Diabetes Technol Ther 2015;17:A-2239. NufferW, Trujillo JM, Ellis SL. Technosphereinsulin (Afrezza): a new, inhaled prandial insulin.Ann Pharmacother 2015;49:99–10640. Dauber A, Corcia L, Safer J, Agus MS, Einis S,Steil GM. Closed-loop insulin therapy improvesglycemic control in children aged ,7 years:a randomized controlled trial. Diabetes Care2013;36:222–22741. Steil GM, Agus MS. Critical illness hyper-glycemia: is failure of the beta-cell to meetextreme insulin demand indicative of dysfunc-tion? Crit Care 2009;13:129

care.diabetesjournals.org Bell and Associates 1015