Advanced Glycation Endproducts in Children with Diabetes

Shreepal Shah, MD1, Eileen A. Baez, MD1, Dania L. Felipe, MD1, John D. Maynard, MS3,

James M. Hempe, PhD2, and Stuart A. Chalew, MD1,2

Objectives To estimate skin content of advanced glycation endproducts (AGEs) by measurements of skinintrinsic fluorescence (SIF) from youth with diabetes in comparison with a population of youth and adults withoutdiabetes.Study design Using a specialized instrument, skin AGEs were estimated from skin auto-fluorescence induced at420 nm and corrected for skin pigmentation (SIF420[kx0.5, km0.5]) in children with types 1 and 2 diabetes, as well aschildren and adults without diabetes. The effect of age, sex, ethnicity, and diabetes status on SIF420[kx0.5, km0.5] wasanalyzed.Results SIF420[kx0.5, km0.5] increased with chronologic age and was higher in children with diabetes comparedwith children without diabetes (P = .0001). SIF420[kx0.5, km0.5] from 43% of children with type 1 diabetes and 55%with type 2 diabetes overlapped the range of adults without diabetes. SIF420[kx0.5, km0.5] was higher in girls thanboys in patients with diabetes patients. However, there was no effect of sex or race on SIF420[kx0.5, km0.5] in subjectswithout diabetes.Conclusions After 4-6 years’ exposure to diabetes, many children will have precociously high estimates of skinAGEs, comparable with levels that would naturally accumulate only after �25 years of chronologic aging. Poten-tially, this technology identifies children who are at increased risk for complications. (J Pediatr 2013;163:1427-31).

Diabetes mellitus is associated with the development of serious microvascular and macrovascular complications.1

Early identification and treatment of diabetes complications is an important goal of diabetes management. One ofthe potential mechanisms underlying development of diabetes complications is believed to be from damage

caused by advanced glycation endproducts (AGEs).2-4 AGEs can be formed from proteins, lipids, and nucleic acids inseveral ways, including non-enzymatic glycation of amino groups through the Maillard reaction, by oxidation of sugarsand/or lipids to form dicarbonyls that bind to amino groups, from oxidation of amino groups, and through polyolpathway metabolism of glucose.4 AGEs accumulate in skin and other tissues over time and their tissue concentrationis also a marker of chronologic aging.2,5 AGE accumulation is associated with gradual decline in tissue and organ func-tion with time.6,7 Adults with diabetes have increased levels of tissue AGEs compared with subjects matched for chrono-logical age without diabetes.2,5 Increased tissue AGE levels from biopsied skin in patients with type 1 diabetes arepredictive of progression of microvascular complications.8,9 Thus, clinical assessment of AGEs may be a useful biomarkerfor complications.

Previously, studies of tissue AGEs have relied on direct chemical measurement from blood and body fluids10-15 or tissue bi-opsy samples.2,5,8,9,16 Based on the fluorescent properties of some AGEs, noninvasive methods have been developed to estimateskin AGEs in vivo from measurement of induced skin fluorescence.17-19 As measurement of skin fluorescence is noninvasiveand takes only a few moments to ascertain, it could be useful in children. We previously reported that selecting an appropriate“intrinsic correction” to adjust the raw skin auto-fluorescence data to yield measurements of skin intrinsic fluorescence (SIF) isimportant for the interpretation of estimated AGE in children with type 1 diabetes.20 Intrinsic correction of skin auto-fluorescence mitigates distortion caused by melanin, hemoglobin, and light scattering, and can enhance the ability to detectinfluences of age, duration of diabetes, and sex on auto-fluorescence. In a brief report, Barat et al found that skinauto-fluorescence without intrinsic correction was higher in children with type 1 diabetes compared with their siblings withoutdiabetes.21 However, SIF in children with type 1 and type 2 diabetes has not previously been evaluated in comparison with areference population of children and adults without diabetes. We hypothesized that children with diabetes would have higher

From the 1Division of Pediatric Endocrinology/Diabetes,Louisiana State University-Health Sciences Center;2

SIF in comparison with children without diabetes even after adjustment for chro-nologic age, similar to findings from blood and tissue biopsy studies. In thisstudy, we compared SIF measures previously shown to be associated with chro-nological age, duration of diabetes and hemoglobin A1c (HbA1c), but not skin

Research Institute for Children, Children’s Hospital,New Orleans, LA; and 3VeraLight, Inc, Albuquerque, NM

The authors declare no conflicts of interest.

Portions of this study were presented at the AmericanDiabetes Association’s Scientific Sessions, June 8-12,2012, Philadelphia, PA.

0022-3476/$ - see front matter. Copyright ª 2013 Mosby Inc.

All rights reserved. http://dx.doi.org/10.1016/j.jpeds.2013.06.044

AGE Advanced glycation endproduct

HbA1c Hemoglobin A1c

LED Light-emitting diode

SIF Skin intrinsic fluorescence

1427

Table. Patient characteristics by group at time of skin AGE evaluation

Established type 1diabetes

Established type 2diabetes

Children withoutdiabetes

Adults withoutdiabetes

n 133 20 91 129Mean age (y) 13� 3a 15� 2a 10� 5b 40� 8c

(range) (5-21) (10-19) (2-19) (21-67)Sex (M/F) 65/68 6/14 41/50 29/100Race(White/African American/Hispanic/others)

87/42/1/3 15/4/0/1 41/42/8/0 72/52/1/4

BMI (kg/m2) 22� 5a 31� 3b - -HbA1c (%) 9.3� 2.0 9.6� 3.0 - -Duration (y) 6.0� 3.6 4.0� 2.0 - -Mean SIF420[kx0.5, km0.5] (AU)(unadjusted )

1.96� 0.59a 2.20� 0.55a 1.52� 0.35b 3.20� 0.75a

Least squares mean SIF420[kx0.5, km0.5] (AU)(adjusted for age, sex, and race)

2.31a 2.50a 1.99b 2.44a

AU, auto-fluorescence unit; BMI, body mass index.a b cDifferent designator letters indicate statistical differences of mean between groups within row.

THE JOURNAL OF PEDIATRICS � www.jpeds.com Vol. 163, No. 5

pigmentation,20 in youth with established type 1 and type2 diabetes compared with children and adults withoutdiabetes.

Methods

Children with diabetes of at least 1-year duration attendingdiabetes clinics at Children’s Hospital of New Orleans, Loui-siana, were invited to participate. Data regarding duration ofdiabetes, height, weight, date of birth, race, sex, HbA1c, andmean blood glucose were collected at the time of visit.Normal healthy children (without history of chronic diseaseor smoking) were enrolled from outpatient pediatric conti-nuity clinics and orthopedic clinics. Parents of pediatricdiabetes participants were invited to participate to obtainan adult-age reference group. Parents with history of diabetesor smoking were excluded from the study. Height, weight,glucose, and HbA1c measurements were not obtained fromsubjects without diabetes. The study was approved by theInstitutional Review Boards at Louisiana State UniversityHealth Sciences Center, New Orleans and the Children’sHospital of New Orleans. Informed consent was obtainedfrom parents and consent/assent was obtained as appropriatefrom the children.

SIF was assessed as an estimate of skin AGEs as previouslydescribed.20 During a clinic visit, induced skin fluorescencewas measured noninvasively from the volar surface of theleft forearm from each subject using a SCOUTDS instrument(VeraLight Inc, Albuquerque, New Mexico). The devicesequentially excites the skin surface using different light-emitting diodes (LEDs) that have peak excitation wave-lengths of 375, 405, and 420 nm. Fluorescence generatedfrom each LED was detected over a 435-655 nm emissionwindow. Skin reflectance is measured for each excitationLED, and a white light LED is used to measure skin reflec-tance over the emission region.20

The detected skin fluorescence has intrinsic correctionsapplied to yield measures of SIF as previously described.20

We previously demonstrated that SIF results using the420 nm excitation LED, with kx and km set at 0.5 and

1428

0.5, respectively (SIF420[kx0.5, km0.5]) were associated withage, duration of diabetes, and level of HbA1c but not skinpigmentation among children with diabetes.20 Therefore,we have used SIF420[kx0.5, km0.5] as our primary outcomemeasure in this report. SIF420[kx0.5, km0.5] has previouslybeen referred to as SIF 420c.20 Similar results were obtainedwith SIF excited by 405-nm LEDs, with kx and km set at 0.5and 0.5 (previously referred to as SIF 405c20). SIF data arereported in arbitrary auto-fluorescence units.

Statistical AnalysesUnless otherwise noted, data are reported as mean � 1 SD.Between-group SIF comparisonsweremade from least squaresmeans adjusted for the influence of age, sex, and race using SASsoftware, PROC GLM (SAS Institute, Cary, North Carolina).Statistical significance was considered to be at P < .05.

Results

Subjects were divided into 4 groups: children withoutdiabetes, children with type 1 diabetes, children with type2 diabetes, and adults without diabetes. Characteristics ofthe groups at the time of SIF measurement are presentedin the Table. Mean age was statistically different betweengroups. SIF420[kx0.5, km0.5] increased with age in thereference group without diabetes, as well as patients withdiabetes (Figure 1).When adjusted for age, sex, and race, SIF420[kx0.5, km0.5]

was higher among children with diabetes than those withoutdiabetes (P < .05) but not different from adults without dia-betes (Table). The statistical model accounted for 64% ofthe variance in SIF420[kx0.5, km0.5]. Ten percent of childrenwithout diabetes overlapped the range of the adultswithout diabetes; 43% of children with type 1 diabetes and55% of children with type 2 diabetes had SIF420[kx0.5,km0.5]

within the range of adults without diabetes (Figure 2). Inpatients with diabetes, SIF420[kx0.5, km0.5] was higher in girlscompared with boys, and increased with higher HbA1c andduration of diabetes.20 There was no influence of race orsex in the reference population without diabetes.

Shah et al

Figure 1. Estimated skin AGEs (SIF420[kx0.5, km0.5]) increasewith age in subjects with and without diabetes. A, Subjectswithout diabetes, SIF420[kx0.5, km0.5] = 0.051*age + 1.063. B,Children with diabetes, SIF420[kx0.5, km0.5] = 0.072*age + 0.98.SIF420[kx0.5, km0.5] expressed in terms of arbitrary AU.

Figure 2. Comparison of estimated skin AGEs (SIF420[kx0.5,km0.5]) for individuals with and without diabetes. Unadjustedmean SIF420[kx0.5, km0.5] of each group represented by hori-zontal line; 43% of type 1 patients and 55% of type 2 patientsoverlapped the range of adults without diabetes. DM,diabetes mellitus.

November 2013 ORIGINAL ARTICLES

Discussion

Tissue accumulation of AGEs is believed to mediate compli-cations of diabetes.2-4 This study assesses skin AGEs, esti-mated quickly and noninvasively from SIF, in youth withtype 1 and type 2 diabetes compared with a reference popula-tion of children and adults without diabetes. Intrinsic correc-tion of auto-fluorescence data to generate measures of SIF isessential in mitigating the impact of skin pigmentation, he-moglobin content, and light scattering. We previously foundthat excited fluorescence at 420 nm using intrinsic correctionof kx = 0.5 and km = 0.5 (SIF420[kx0.5, km0.5]) increased withage, duration of diabetes, andHbA1c20, similar to findings re-ported from tissue biopsy studies.5-7 Remarkably, in childrenwith type 1 diabetes SIF420[kx0.5, km0.5] was associated withboth current and updated HbA1c andHemoglobin GlycationIndex but not the current or updated mean blood glucose.20

This suggests that SIF may be influenced by factors associatedwith glycation that are not blood glucose concentration-

Advanced Glycation Endproducts in Children with Diabetes

dependent. Furthermore SIF420[kx0.5, km0.5] is not influencedby the degree of skin pigmentation, which readily facilitatescomparison between racial/ethnic groups and individuals ofdiffering skin pigmentation.20 This may not be the casewith other reported auto-fluorescence measures, which areuncorrected and can be distorted by differences in skinpigmentation.We found that skin AGEs estimated by SIF increased with

age in subjects with and without diabetes. Children withdiabetes, both type 1 and type 2, had highermean levels of esti-mated skin AGE than children without diabetes when adjustedfor chronological age and sex. This is similar to findings ofhigher tissue AGE levels from biopsy samples reported fromadults with diabetes compared with age-matched subjectswithout diabetes,2,5 and a prior small study of skin auto-fluorescence in children with type 1 diabetes vs their siblingswho did not have diabetes.21 A novel observation of the presentstudy is that a large percentage of children with diabetes (43%type 1 and 55% type 2) had levels of skin AGEs that overlappedthe range of adults without diabetes. Thus, approximately 4-6years of diabetes exposure in some childrenmaybe sufficient toincrease skin AGEs to levels that would naturally accumulateonly after �25 years of chronological aging. A small numberof children without diabetes were also found to have increasedSIF levelswithin the range of the adults, and the implications ofthis finding will require further investigation.A number of prior studies of AGEs in children and adults

sampled only serum or urine levels.11,12,14,15,22 Serum andurine AGEs levels are generally higher in patients with dia-betes compared with individuals without diabetes. To date,

1429

THE JOURNAL OF PEDIATRICS � www.jpeds.com Vol. 163, No. 5

there have not been comparisons performed between serumand/or urine AGEs with SIF data for effectiveness in predict-ing long-term diabetes complications. Serum and urine AGElevels may have limited predictive ability as levels may betransient in nature and may not reflect long-term trends intissue AGE accumulation, which may be a better predictorof long-term diabetes complications.9,16 Potentially, SIF,which provides an estimate of tissue AGE accumulation,may be an indicator of long-term “metabolic memory” anda more robust predictor of diabetes complications.23,24

Several limitations of the current study should be presented.Although SIF offers a novel, noninvasive technology for estima-tion of skin AGEs, not all AGEs fluoresce, so SIF mayunderestimate total tissue burden of AGEs. The present studywas cross-sectional in design and would not capture individualvariation over time that might be present in a longitudinalstudy. Diet may contain AGEs and it is not clear how dietaryfactors may influence SIF measurements.25 The diabetes andsmoking status of the reference youth and adult populationwas based on self-report; there was no biochemical confirma-tion (eg, bymeasurementofHbA1c, glucose levels, or cotinine).As type 1 and type 2 diabetes in children have a characteristicclinical symptomatology, it is unlikely that the pediatric popu-lation without diabetes inadvertently contained patients withundetected diabetes. Individuals with unrecognized diabetesin the adult cohort would have raised the upper limit of SIFfor this group, leaving the lower limit unaffected and, thus,would not change our conclusions regarding the precociouslyhigh levels of AGEs in children with diabetes. Potentially, indi-viduals, particularly youth, might have been reluctant to admitsmoking and, thus, smokers may have inadvertently beenincluded in the study population, which may possibly explainsome higher SIF levels in some youths without diabetes.

In biopsy studies, tissue AGE levels have been associatedwith complications of diabetes and age-related disease.3,4,6,9,26

Skin AGEs estimated noninvasively by assessment of SIF havebeen correlated with complications of diabetes in adults.24,27-29 Thus, precociously high levels of estimated skin AGEs inchildren with diabetes may be a biomarker of early diabetescomplications. Measurement of SIF in children and adoles-cents with diabetesmay facilitate early identification of youngpatients at highest risk for complications and potentially leadto better management planning. A further longitudinal studywill be needed to define the importance of SIF as an early pre-dictor of complications in children. n

The authors thank Suzanne Lefevre, MD, for facilitating access topatients without diabetes at the Louisiana State University-HealthSciences Center Pediatric Ambulatory Care Clinics.

Submitted for publication Dec 12, 2012; last revision received Mar 4, 2013;

accepted Jun 20, 2013.

Reprint requests: Stuart A. Chalew, MD, Children’s Hospital of New Orleans,

200 Henry Clay Ave, New Orleans, LA 70118. E-mail: schale@lsuhsc.edu

References

1. The Diabetes Control and Complications Trial Research Group. The

effect of intensive treatment of diabetes on the development and

1430

progression of long-term complications in IDDM. N Engl J Med 1993;

329:977-83.

2. Cerami A. The unexpected pathway to the creation of theHbA1c test and

the discovery of AGE’s. J Intern Med 2012;271:219-26.

3. Ahmed N, Thornalley PJ. Advanced glycation endproducts: what is

their relevance to diabetic complications? Diabetes Obes Metab 2007;

9:233-45.

4. Goh SY, Cooper ME. Clinical review: the role of advanced glycation end

products in progression and complications of diabetes. J Clin Endocrinol

Metab 2008;93:1143-52.

5. Dyer DG, Dunn JA, Thorpe SR, Bailie KE, Lyons TJ, McCance DR, et al.

Accumulation of Maillard reaction products in skin collagen in diabetes

and aging. J Clin Invest 1993;91:2463-9.

6. Monnier VM. Nonenzymatic glycosylation, the Maillard reaction and

the aging process. J Gerontol 1990;45:B105-11.

7. Ramasamy R, Vannucci SJ, Yan SS, Herold K, Yan SF, Schmidt AM.

Advanced glycation end products and RAGE: a common thread in aging,

diabetes, neurodegeneration, and inflammation. Glycobiology 2005;15:

16R-28R.

8. Araszkiewicz A, Naskret D, Niedzwiecki P, Samborski P, Wierusz-

Wysocka B, Zozulinska-Ziolkiewicz D. Increased accumulation of skin

advanced glycation end products is associated with microvascular com-

plications in type 1 diabetes. Diabetes Technol Ther 2011;13:837-42.

9. Genuth S, SunW, Cleary P, Sell DR, DahmsW,Malone J, et al. Glycation

and carboxymethyllysine levels in skin collagen predict the risk of future

10-year progression of diabetic retinopathy and nephropathy in the dia-

betes control and complications trial and epidemiology of diabetes inter-

ventions and complications participants with type 1 diabetes. Diabetes

2005;54:3103-11.

10. Wu Y, Monnier V, Friedlander M. Reliable determination of furosine in

human serum and dialysate proteins by high-performance liquid chro-

matography. J Chromatogr Biomed Appl 1995;19:328-32.

11. Chiarelli F, de Martino M, Mezzetti A, Catino M, Morgese G,

Cuccurullo F, et al. Advanced glycation end products in children and ad-

olescents with diabetes: relation to glycemic control and early microvas-

cular complications. J Pediatr 1999;134:486-91.

12. Kostolanska J, Jakus V, Barak L. HbA1c and serum levels of advanced

glycation and oxidation protein products in poorly and well controlled

children and adolescents with type 1 diabetes mellitus. J Pediatr Endocri-

nol Metab 2009;22:433-42.

13. Jakus V, Bauerova K, Michalkova D, Carsky J. Values of markers of early

and advanced glycation and lipoxidation in serum proteins of children

with diabetes mellitus. Bratisl Lek Listy 2000;101:484-9.

14. Berg TJ, Dahl-Jorgensen K, Torjesen PA, Hanssen KF. Increased serum

levels of advanced glycation end products (AGEs) in children and ado-

lescents with IDDM. Diabetes Care 1997;20:1006-8.

15. Galler A, Muller G, Schinzel R, Kratzsch J, Kiess W, Munch G. Impact of

metabolic control and serum lipids on the concentration of advanced

glycation end products in the serum of children and adolescents with

type 1 diabetes, as determined by fluorescence spectroscopy and N

epsilon-(carboxymethyl)lysine ELISA. Diabetes Care 2003;26:2609-15.

16. Monnier VM, Bautista O, Kenny D, Sell DR, Fogarty J, Dahms W, et al.

Skin collagen glycation, glycoxidation, and crosslinking are lower in sub-

jects with long-term intensive vs conventional therapy of type 1 diabetes:

relevance of glycated collagen products vs HbA1c as markers of diabetic

complications. DCCT Skin Collagen Ancillary Study Group. Diabetes

Control and Complications Trial. Diabetes 1999;48:870-80.

17. Meerwaldt R, Graaff R, Oomen PH, Links TP, Jager JJ, Alderson NL,

et al. Simple noninvasive assessment of advanced glycation endproduct

accumulation. Diabetologia 2004;47:1324-30.

18. Hull EL, Ediger MN, Brown CD, Maynard JD, Johnson RD, inventors;

Determination of a measure of a glycation end-product or disease state

using tissue fluorescence. US Patent 7139598, 2006.

19. Mulder DJ, Water TV, Lutgers HL, Graaff R, Gans RO, Zijlstra F, et al.

Skin autofluorescence, a novel marker for glycemic and oxidative

stress-derived advanced glycation endproducts: an overview of current

clinical studies, evidence, and limitations. Diabetes Technol Ther 2006;

8:523-35.

Shah et al

November 2013 ORIGINAL ARTICLES

20. Felipe DL, Hempe JM, Liu S, Matter N, Maynard J, Linares C, et al. Skin

intrinsic fluorescence is associated with hemoglobin a1c and hemoglobin

glycation index but not mean blood glucose in children with type 1 dia-

betes. Diabetes Care 2011;34:1816-20.

21. Barat P, Cammas B, Lacoste A, Harambat J, Vautier V, Nacka F, et al.

Advanced glycation end products in children with type 1 diabetes: family

matters? Diabetes Care 2012;35:e1.

22. Jakus V, Bauerova K, Michalkova D, Carsky J. Serum levels of

advanced glycation end products in poorly metabolically controlled chil-

dren with diabetes mellitus: relation to HbA1c. Diabetes Nutr Metab

2001;14:207-11.

23. Meerwaldt R, Links T, Zeebregts C, Tio R, Hillebrands JL, Smit A. The

clinical relevance of assessing advanced glycation endproducts accumu-

lation in diabetes. Cardiovasc Diabetol 2008;7:29.

24. Orchard TJ, Lyons TJ, Cleary PA, Braffett BH, Maynard J, Cowie C, et al.

The DCCT/EDIC Research Group.The association of skin-intrinsic fluo-

rescence with type 1 diabetes complications in the DCCT/EDIC Study.

Diabetes Care. 2013 Jun 28. Eup ahead of print.

Advanced Glycation Endproducts in Children with Diabetes

25. Uribarri J, Woodruff S, Goodman S, Cai W, Chen X, Pyzik R, et al.

Advanced glycation end products in foods and a practical guide to their

reduction in the diet. J Am Diet Assoc 2010;110:911-6.e12.

26. Monnier VM, Vishwanath V, Frank KE, Elmets CA, Dauchot P,

Kohn RR. Relation between complications of type I diabetes mellitus

and collagen-linked fluorescence. N Engl J Med 1986;314:403-8.

27. Conway B, Edmundowicz D, Matter N, Maynard J, Orchard T.

Skin fluorescence correlates strongly with coronary artery calcifica-

tion severity in type 1 diabetes. Diabetes Technol Ther 2010;12:

339-45.

28. Conway BN, Aroda VR, Maynard JD, Matter N, Fernandez S, Ratner RE,

et al. Skin intrinsic fluorescence correlates with autonomic and distal

symmetrical polyneuropathy in individuals with type 1 diabetes. Dia-

betes Care 2011;34:1000-5.

29. Conway BN, Aroda VR, Maynard JD, Matter N, Fernandez S, Ratner RE,

et al. Skin intrinsic fluorescence is associated with coronary artery disease

in individuals with long duration of type 1 diabetes. Diabetes Care 2012;

35:2331-6.

1431