Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2015 Sep 1.
Published in final edited form as: J Diabetes Complications. 2014 Jun 17;28(5):658–661. doi: 10.1016/j.jdiacomp.2014.06.007

Differential reduction in corneal nerve fiber length in patients with type 1 or type 2 diabetes mellitus

Maxwell S Stem 1, Munira Hussain 1, Stephen I Lentz 2, Nilesh Raval 3, Thomas W Gardner 1, Rodica Pop-Busui 2, Roni M Shtein 1
PMCID: PMC4146399  NIHMSID: NIHMS606502  PMID: 25044236

Abstract

Aim

To examine the relationship between corneal nerve fiber length (CNFL) and diabetic neuropathy (DN) status in patients with type 1 or type 2 diabetes mellitus (DM)

Methods

In this cross-sectional study, we examined 25 diabetic patients without DN, 10 patients with mild DN, 8 patients with severe DN, and 9 controls without diabetes. DN status was assigned based on a combination of clinical symptoms, signs, and electrophysiological testing. Patients underwent corneal confocal microscopy (CCM) of the sub-basal nerve plexus. Post-hoc analysis of the CCM images was performed to quantify the average CNFL, and ANOVA was used to assess for differences in CNFL.

Results

All 25 subjects without DN had type 1 DM, and subjects with DN had type 2 DM. Participants with severe DN had significantly lower CNFL (12.5 ± 6.1 mm/mm2) compared to controls (20.7 ± 2.2 mm/mm2) (p=0.009). However, lower CNFL was also found in participants with type 1 DM who did not have DN (15.1 ± 4.7 mm/mm2) relative to controls (p=0.033).

Conclusions

CCM of the sub-basal nerve plexus may be an indicator of early peripheral nerve degeneration in type 1 DM. Type of diabetes, in addition to degree of neuropathy, may influence the extent of corneal nerve damage.

Keywords: Corneal confocal microscopy, diabetes mellitus, neuropathy, corneal nerves

INTRODUCTION

Diabetic neuropathy (DN) is a leading form of neuropathy in the western world, and affects up to 50% of patients with diabetes.1 In spite of continuous efforts, multiple agents targeting pathogenic mechanisms to reverse DN have failed in human trials. One explanation for the failure of most DN clinical trials is related to the complex anatomy of the peripheral nervous system and heterogeneous spectrum of clinical manifestations of DN. Numerous diagnostic and outcome measures have been used to assess therapeutic efficacy or monitor for disease progression in clinical trials, contributing to poor reproducibility and inconclusive findings.2 In addition, nerve conduction studies (NCS) considered for many years as “gold standard” diagnostic criteria provide information solely on large fiber dysfunction. The more recent use of intraepidermal nerve fiber density (IENFD) to assess small fibers3 is limited by its invasiveness.4 Given these limitations, there is a need for a more optimal measure of neuropathy severity that combines the best attributes from these other tests into a single metric.

Several groups of investigators have reported that corneal confocal microscopy (CCM) of the sub-basal nerve plexus provides a non-invasive, repeatable5 method of quantifying small fiber nerve damage from diabetes,3,6–12 and that it can discriminate between controls, patients with diabetes and no neuropathy, and diabetic patients with neuropathy.13 Furthermore, it was suggested that CCM may detect increased corneal nerve density14 after simultaneous pancreas and kidney transplantation before improvements can be detected by traditional measures of peripheral neuropathy severity.15 Taken together, these findings suggest that CCM may be a useful indicator of neuropathy severity that can be used to assess therapeutic efficacy in clinical trials. Indeed, examination of the corneal sub-basal nerve plexus has already been used to assess therapeutic efficacy of a vasopeptidase inhibitor in treating DN in an animal model of diabetes.16

Despite the promising role that CCM can play in stratifying patients with DN according to the stage of their disease, it remains unclear whether type of diabetes, in addition to stage of neuropathy, influences corneal nerve density. In the present study, we used CCM to examine the relationship between CNFL and DN status in patients with type 1 or type 2 diabetes mellitus (DM).

MATERIALS AND METHODS

Patients

Four groups of participants were recruited for this study: healthy controls, diabetic patients without DN, patients with mild DN, and those with severe DN. The main inclusion criteria were age ≥ 18 years and presence of diabetes as defined by the American Diabetes Association. Exclusion criteria for all participants were: 1) any systemic neuropathy other than DN; 2) history of corneal disease or any eye surgery; 3) any neurodegenerative condition, such as Parkinson’s disease or multiple sclerosis; 4) history of stroke or cancer; 5) history of spinal stenosis or previous back surgery; 6) not willing/able to provide consent.

This study was approved by the University of Michigan Institutional Review Board and complied with the tenets of the Declaration of Helsinki. Written informed consent was obtained from each subject prior to testing.

Diabetic neuropathy severity

Mild DN was defined as: the presence of ≥ 1 abnormal attribute (of amplitude, latency, F-wave, or nerve conduction velocity) in ≥ 2 separate nerves among the median, peroneal, and sural nerves in NCS studies, and the presence of small fiber dysfunction as evaluated by neurological examination performed by a board certified neurologist and by abnormal quantitative sensory testing (QST).17,18

Severe DN was defined as: a combination of symptoms and signs of distal sensorimotor polyneuropathy with two or more of the following: neuropathic symptoms, decreased distal sensation (as assessed by neurological examination performed by a board certified neurologist and by abnormal QST), and unequivocally decreased or absent ankle reflexes plus abnormalities in NCS as described above and absence of a recordable sural nerve amplitude.

Corneal Confocal Microscopy

All subjects underwent CCM imaging of the central corneal sub-basal nerve plexus using the Heidelberg Retina Tomograph-2 (HRT-2) Rostock cornea module (Heidelberg Engineering, Heidelberg, Germany). After instilling one drop of topical proparacaine hydrochloride 0.5% into the right eye, the patient placed his/her head into the headrest of the confocal microscope. A small amount of Genteal eye gel (Novartis Pharmaceuticals, East Hanover, NJ) was applied to the lens of the microscope, and a sterile plastic cap (Tomocap, Heidelberg Engineering, Heidelberg, Germany) was placed over the lens. The patient was instructed to look straight ahead as the plastic-covered lens contacted the patient’s cornea. “Sequence” mode was used to record consecutive 400 × 400 μm images of the sub-basal nerve layer at 8 frames per second.

Analysis of the corneal nerve images began with selection of five representative, focused, non-overlapping images of the sub-basal nerve plexus from the central cornea from each participant. A masked grader (NR) used NeuronJ software (publicly available at http://www.imagescience.org/meijering/software/neuronj/) to calculate the sum length of all nerve fibers within each image. From this number, the CNFL was derived by dividing the sum length of the nerve fibers in each image by the area of the image (1600 μm2). The five CNFL measurements from each individual were averaged to determine the final CNFL per patient.

Statistical Analysis

Means and standard deviations were calculated for continuous variables, and frequencies and percentages were used for categorical variables. We used analysis of variance (ANOVA) to compare the average CNFL measurements obtained from each group.

RESULTS

Patient Characteristics

This study included 25 diabetic patients without DN, 10 patients with mild DN, 8 patients with severe DN, and 9 controls without diabetes. Participant characteristics are listed in Table 1. All of the patients without DN had type 1 DM, and patients with mild or severe DN had type 2 DM. On average, individuals without DN were younger than those with mild (p < 0.001) or severe (p = 0.019) DN. 72% of patients without DN were female, and 20% of those with mild DN and 12.5% of individuals with severe DN were female (p = 0.002). There were no statistically significant differences in duration of diabetes or hemoglobin A1c (HBA1c) values among the three groups, but those with type 1 DM tended to have a longer duration of DM compared to patients with type 2 DM and either mild or severe DN (p = 0.058).

Table 1.

Participant characteristics according to severity of diabetic neuropathy. Abbreviations: DN, Diabetic Neuropathy; DM, Diabetes Mellitus; HbA1c, Hemoglobin A1c; SD, Standard Deviation; T1DM, Type 1 Diabetes Mellitus; T2DM, Type 2 Diabetes Mellitus.

CONTROLS NO DN (T1DM) MILD DN (T2DM) SEVERE DN (T2DM) P-value
Total participants, n (%) 9 (17.3) 25 (48.1) 10 (19.2) 8 (15.4)
Age, mean (SD) 43.9 (10.2) 38.7 (14.2) 57.7 (8.1)† 53.4 (8.0)† p < 0.001‡
HbA1c, mean % (SD) [mmol/mol] 5.4 (0.2) [35.5] 8.1 (1.0)* [65.0] 7.9 (1.1)* [62.8] 8.3 (1.5)* [67.2] p < 0.001‡
Duration of DM in years, mean (SD) - 13.5 (6.7) 7.6 (3.8) 10.6 (3.6) p = 0.058‡
Sex, n (%) p = 0.002§
 Males 3 (33.3) 7 (28.0) 8 (80.0) 7 (87.5)
 Females 6 (66.7) 18 (72.0) 2 (20.0) 1 (12.5)
Type of Diabetes, n (%) p < 0.001§
 Type 1 - 25 (100) 0 (0) 0 (0)
 Type 2 - 0 (0) 10 (100) 8 (100)
Open in a new tab
*

p < 0.05 vs. Control

†

p < 0.05 vs. No DN

‡

ANOVA

§

Chi-square test

Corneal Nerve Fiber Length

Figure 1 depicts representative CCM images from a participant within each group. Patients with type 2 DM and severe DN had fewer corneal nerves (CNFL ± 95% confidence interval 19 = 12.5 ± 5.1 mm/mm2) compared to controls (20.7 ± 1.7 mm/mm2) (p=0.009), and those with type 1 DM without DN also had lower CNFL (15.1 ± 1.9 mm/mm2) relative to controls (p=0.033) (Figure 2). However, individuals with mild DN did not exhibit significantly lower CNFL compared to controls (p=1.000).

Figure 1.

Figure 1

Open in a new tab

Representative images of the corneal sub-basal nerve plexus from patients with varying degrees of diabetic neuropathy. The tortuous white lines in each picture are the corneal nerves, and the small white dots are Langerhans cells. Participants with diabetes and no neuropathy and those with diabetes and severe neuropathy appear to have fewer corneal nerves compared to controls. Abbreviations: T1DM, Type 1 diabetes mellitus; T2DM, Type 2 diabetes mellitus; DN, diabetic neuropathy.

Figure 2.

Figure 2

Open in a new tab

Average corneal nerve fiber length (CNFL) for each group ± 95% confidence interval. CNFL is defined as the sum length of all corneal nerves within the image divided by the area of the image (0.16 mm2). Patients with type 1 diabetes and no neuropathy (n=25) and those with type 2 diabetes and severe neuropathy (n=8) had, on average, fewer nerves per area of cornea relative to controls (n=9). Patients with type 2 diabetes and mild DN (n=10) had a similar CNFL compared to controls. The maximum/minimum CNFL values were as follows: Control (24.4/17.6), No DN (24.5/6.5), Mild DN (29.1/12.1), and Severe DN (19.4/1.2 mm). Abbreviations: DN, Diabetic Neuropathy; mm, millimeters. Symbols: *, p < 0.05 vs. Control; #, p < 0.05 vs. Mild DN.

DISCUSSION

The goal of this cross-sectional study was to determine whether type of diabetes influences the level of corneal nerve loss across the spectrum of diabetic neuropathy. Using CCM, we found that patients with type 1 DM and no DN had 27% lower CNFL relative to controls, but patients with type 2 DM and mild DN did not have lower CNFL compared to controls. Patients with type 2 DM and severe DN however, exhibited 40% lower CNFL relative to the control group.

Our finding that patients with severe DN have lower levels of corneal nerves compared to controls is consistent with several other reports of corneal nerve loss in DN.6,20 However, unlike recent studies which showed a progressive loss of CNFL across the disease spectrum,13,21 we found that patients with type 1 DM who did not meet any conventional criteria for DN exhibited lower CNFL relative to patients with type 2 DM who had mild DN, in spite of equivalent HBA1c levels. Importantly, the tools that we used to stratify the participants into different neuropathy subgroups are the same tools other investigators have used to categorize patients according to neuropathy severity,13,21 so it seems unlikely that neuropathy group assignment error could account for our different results. Furthermore, gender does not appear to influence corneal nerve density, so it is unlikely that differences in gender distribution between the groups would account for our observations.22 Studies evaluating the effects of age on corneal innervation have yielded mixed results, with one study showing no effect of age on corneal nerve density23 and another finding that increasing age is correlated with reduced corneal innervation.22 In either case, it is unlikely that differences in age among the neuropathy subgroups would account for our results because the patients with DM but no neuropathy were younger, on average, than patients with mild DN or severe DN. Diabetes duration, however, is inversely correlated with CNFL,13 and in our study, patients with type 1 DM tended to have a longer duration of DM compared to patients with type 2 DM and mild DN. It is possible that the effect on CNFL that was observed in the type 1DM patients is rather due to their longer exposure to hyperglycemia than to their specific type of DM. However, given that type 2 DM patients may be completely asymptomatic for many years prior to a formal diabetes diagnosis it is likely that the duration is grossly underestimated in patients with type 2 DM.24 Thus, it seems reasonable to attribute the differences in CNFL between the no DN and mild DN subgroups to differences in type of diabetes rather than differences in DM duration.

There are several potential explanations for our finding that type of DM appeared to influence the extent of corneal nerve damage among the study participants. The fundamental difference in pathophysiology between type 1 and type 2 DM could theoretically account for the discrepancy. That is, in type 1 DM there is a systemic lack of insulin while in type 2 DM insulin abounds but its target cells are resistant to its effects. As insulin has been shown to be important for neuronal growth and survival,25 the differences in insulin production/utilization between type 1 and type 2 DM could influence corneal nerve degeneration in disparate ways.

This study, while small in scope, had several strengths. First, subjects were extensively characterized with neurophysiologic testing, clinical exams, and patient history in a standardized fashion, which ensured that all participants were placed into appropriate neuropathy categories based on disease severity. Second, measurements of CNFL using NeuronJ were done in a masked fashion by the same person in all of the study participants, which eliminates potential bias in image assessment.

Despite the strengths of this study, several limitations need to be acknowledged. This study involved a small number of patients, so definitive conclusions about the effect of type of diabetes on corneal nerve destruction cannot be drawn. Also, we did not enroll any patients with type 2 DM without DN or any individuals who had type 1 DM with neuropathy. Lastly, we were unable to control for potential confounding factors in this study, such as duration of diabetes.

In conclusion, the results from this cross-sectional study suggest that type of diabetes may influence the extent of corneal nerve loss in patients with DN. Specifically, CCM measurements of CNFL may be a more sensitive indicator of early nerve damage in patients with type 1 DM compared to those with type 2 DM. The results of this study, if confirmed in subsequent investigations, have important implications for the use of CCM in the detection of DN. For example, the diagnostic threshold of CNFL used to detect DN might be different for type 1 DM compared to type 2 DM (lower for type 1, and higher for type 2). While CCM shows promise as a technique for assessing DN severity and progression, it will need to be more rigorously tested in larger cohorts of patients with both types of diabetes, and across multiple centers to more fully understand the implications of this outcome measure.

Acknowledgments

This work was supported by: Michigan Diabetes Research Center funded by NIH P60DK020572 from the National Institute of Diabetes and Digestive and Kidney Diseases (RMS and SIL); American Diabetes Association-Merck (MSS); Juvenile Diabetes Research Foundation, Taubman Institute, and Research to Prevent Blindness Physician Scientist Award and EY20582 (TWG); HL102334 (RP).

Footnotes

No conflicting relationship exists for any author regarding any material discussed in this manuscript.

A version of this manuscript was presented as a poster at the Association for Research in Vision and Ophthalmology in Seattle, WA in 2013.

The sponsor or funding organization had no role in the design or conduct of this research.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • 1.Tesfaye S, Boulton AJM, Dyck PJ, et al. Diabetic neuropathies update on definitions, diagnostic criteria, estimation of seventy, and treatments Diabetes Care (vol 33, pg 2285, 2010) Diabetes Care. 2010 Dec;33(12):2725–2725. doi: 10.2337/dc10-1303. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Dyck PJ, Norell JE, Tritschler H, et al. Challenges in design of multicenter trials: end points assessed longitudinally for change and monotonicity. Diabetes Care. 2007 Oct;30(10):2619–2625. doi: 10.2337/dc06-2479. [DOI] [PubMed] [Google Scholar]
  • 3.Quattrini C, Tavakoli M, Jeziorska M, et al. Surrogate markers of small fiber damage in human diabetic neuropathy. Diabetes. 2007 Aug;56(8):2148–2154. doi: 10.2337/db07-0285. [DOI] [PubMed] [Google Scholar]
  • 4.Shtein RM, Callaghan BC. Corneal confocal microscopy as a measure of diabetic neuropathy. Diabetes. 2013 Jan;62(1):25–26. doi: 10.2337/db12-1114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Efron N, Edwards K, Roper N, et al. Repeatability of measuring corneal subbasal nerve fiber length in individuals with type 2 diabetes. Eye Contact Lens. 2010 Sep;36(5):245–248. doi: 10.1097/ICL.0b013e3181eea915. [DOI] [PubMed] [Google Scholar]
  • 6.Rosenberg ME, Tervo TM, Immonen IJ, Muller LJ, Gronhagen-Riska C, Vesaluoma MH. Corneal structure and sensitivity in type 1 diabetes mellitus. Invest Ophthalmol Vis Sci. 2000 Sep;41(10):2915–2921. [PubMed] [Google Scholar]
  • 7.Ahmed A, Bril V, Orszag A, et al. Detection of diabetic sensorimotor polyneuropathy by corneal confocal microscopy in type 1 diabetes: a concurrent validity study. Diabetes Care. 2012 Apr;35(4):821–828. doi: 10.2337/dc11-1396. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Efron N. The Glenn A. Fry award lecture 2010: Ophthalmic markers of diabetic neuropathy. Optom Vis Sci. 2011 Jun;88(6):661–683. doi: 10.1097/OPX.0b013e3182171020. [DOI] [PubMed] [Google Scholar]
  • 9.Hertz P, Bril V, Orszag A, et al. Reproducibility of in vivo corneal confocal microscopy as a novel screening test for early diabetic sensorimotor polyneuropathy. Diabet Med. 2011 Oct;28(10):1253–1260. doi: 10.1111/j.1464-5491.2011.03299.x. [DOI] [PubMed] [Google Scholar]
  • 10.Kallinikos P, Berhanu M, O’Donnell C, Boulton AJ, Efron N, Malik RA. Corneal nerve tortuosity in diabetic patients with neuropathy. Invest Ophthalmol Vis Sci. 2004 Feb;45(2):418–422. doi: 10.1167/iovs.03-0637. [DOI] [PubMed] [Google Scholar]
  • 11.Midena E, Brugin E, Ghirlando A, Sommavilla M, Avogaro A. Corneal diabetic neuropathy: a confocal microscopy study. J Refract Surg. 2006 Nov;22(9 Suppl):S1047–1052. doi: 10.3928/1081-597X-20061102-08. [DOI] [PubMed] [Google Scholar]
  • 12.Messmer EM, Schmid-Tannwald C, Zapp D, Kampik A. In vivo confocal microscopy of corneal small fiber damage in diabetes mellitus. Graefes Arch Clin Exp Ophthalmol. 2010 Sep;248(9):1307–1312. doi: 10.1007/s00417-010-1396-8. [DOI] [PubMed] [Google Scholar]
  • 13.Edwards K, Pritchard N, Vagenas D, Russell A, Malik RA, Efron N. Utility of corneal confocal microscopy for assessing mild diabetic neuropathy: baseline findings of the LANDMark study. Clin Exp Optom. 2012 May;95(3):348–354. doi: 10.1111/j.1444-0938.2012.00740.x. [DOI] [PubMed] [Google Scholar]
  • 14.Mehra S, Tavakoli M, Kallinikos PA, et al. Corneal confocal microscopy detects early nerve regeneration after pancreas transplantation in patients with type 1 diabetes. Diabetes Care. 2007 Oct;30(10):2608–2612. doi: 10.2337/dc07-0870. [DOI] [PubMed] [Google Scholar]
  • 15.Tavakoli M, Mitu-Pretorian M, Petropoulos IN, et al. Corneal Confocal Microscopy Detects Early Nerve Regeneration in Diabetic Neuropathy After Simultaneous Pancreas and Kidney Transplantation. Diabetes. 2012 Sep 21; doi: 10.2337/db12-0574. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Davidson EP, Coppey LJ, Yorek MA. Early loss of innervation of cornea epithelium in streptozotocin-induced type 1 diabetic rats: improvement with ilepatril treatment. Invest Ophthalmol Vis Sci. 2012 Dec;53(13):8067–8074. doi: 10.1167/iovs.12-10826. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Albers JW, Herman WH, Pop-Busui R, et al. Effect of prior intensive insulin treatment during the Diabetes Control and Complications Trial (DCCT) on peripheral neuropathy in type 1 diabetes during the Epidemiology of Diabetes Interventions and Complications (EDIC) Study. Diabetes Care. 2010 May;33(5):1090–1096. doi: 10.2337/dc09-1941. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Dyck PJ, Albers JW, Andersen H, et al. Diabetic Polyneuropathies: Update on Research Definition, Diagnostic Criteria and Estimation of Severity. Diabetes/metabolism research and reviews. 2011 Jun 21; doi: 10.1002/dmrr.1226. [DOI] [PubMed] [Google Scholar]
  • 19.De Cilla S, Ranno S, Carini E, et al. Corneal subbasal nerves changes in patients with diabetic retinopathy: an in vivo confocal study. Invest Ophthalmol Vis Sci. 2009 Nov;50(11):5155–5158. doi: 10.1167/iovs.09-3384. [DOI] [PubMed] [Google Scholar]
  • 20.Malik RA, Kallinikos P, Abbott CA, et al. Corneal confocal microscopy: a non-invasive surrogate of nerve fibre damage and repair in diabetic patients. Diabetologia. 2003 May;46(5):683–688. doi: 10.1007/s00125-003-1086-8. [DOI] [PubMed] [Google Scholar]
  • 21.Tavakoli M, Quattrini C, Abbott C, et al. Corneal confocal microscopy: a novel noninvasive test to diagnose and stratify the severity of human diabetic neuropathy. Diabetes Care. 2010 Aug;33(8):1792–1797. doi: 10.2337/dc10-0253. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Niederer RL, Perumal D, Sherwin T, McGhee CN. Age-related differences in the normal human cornea: a laser scanning in vivo confocal microscopy study. Br J Ophthalmol. 2007 Sep;91(9):1165–1169. doi: 10.1136/bjo.2006.112656. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Erie JC, McLaren JW, Hodge DO, Bourne WM. The effect of age on the corneal subbasal nerve plexus. Cornea. 2005;24(6):705–709. doi: 10.1097/01.ico.0000154387.51355.39. [DOI] [PubMed] [Google Scholar]
  • 24.Harris MI, Klein R, Welborn TA, Knuiman MW. Onset of NIDDM occurs at least 4–7 yr before clinical diagnosis. Diabetes Care. 1992 Jul;15(7):815–819. doi: 10.2337/diacare.15.7.815. [DOI] [PubMed] [Google Scholar]
  • 25.Bassil F, Fernagut PO, Bezard E, Meissner WG. Insulin, IGF-1 and GLP-1 signaling in neurodegenerative disorders: Targets for disease modification? Progress in neurobiology. 2014 Feb 28; doi: 10.1016/j.pneurobio.2014.02.005. [DOI] [PubMed] [Google Scholar]

RESOURCES