Free
Retina  |   January 2014
Quantitative Analysis of Diabetic Macular Ischemia Using Optical Coherence Tomography
Author Affiliations & Notes
  • Dawn A. Sim
    National Institute for Health Research Biomedical Research Centre for Ophthalmology, Moorfields Eye Hospital NHS Foundation Trust, London, United Kingdom
    University College London, Institute of Ophthalmology, London, United Kingdom
  • Pearse A. Keane
    National Institute for Health Research Biomedical Research Centre for Ophthalmology, Moorfields Eye Hospital NHS Foundation Trust, London, United Kingdom
    University College London, Institute of Ophthalmology, London, United Kingdom
  • Simon Fung
    National Institute for Health Research Biomedical Research Centre for Ophthalmology, Moorfields Eye Hospital NHS Foundation Trust, London, United Kingdom
  • Michael Karampelas
    University College London, Institute of Ophthalmology, London, United Kingdom
  • Srinivas R. Sadda
    Doheny Eye Institute, University of Southern California, Los Angeles, California
  • Marcus Fruttiger
    University College London, Institute of Ophthalmology, London, United Kingdom
  • Praveen J. Patel
    National Institute for Health Research Biomedical Research Centre for Ophthalmology, Moorfields Eye Hospital NHS Foundation Trust, London, United Kingdom
    University College London, Institute of Ophthalmology, London, United Kingdom
    Medical Retina Service, Moorfields Eye Hospital NHS Foundation Trust, London, United Kingdom
  • Adnan Tufail
    National Institute for Health Research Biomedical Research Centre for Ophthalmology, Moorfields Eye Hospital NHS Foundation Trust, London, United Kingdom
    University College London, Institute of Ophthalmology, London, United Kingdom
    Medical Retina Service, Moorfields Eye Hospital NHS Foundation Trust, London, United Kingdom
  • Catherine A. Egan
    National Institute for Health Research Biomedical Research Centre for Ophthalmology, Moorfields Eye Hospital NHS Foundation Trust, London, United Kingdom
    Medical Retina Service, Moorfields Eye Hospital NHS Foundation Trust, London, United Kingdom
  • Correspondence: Dawn A. Sim, Moorfields Eye Hospital NHS Foundation Trust, 162 City Road, London EC1V 2PD, UK; [email protected]
Investigative Ophthalmology & Visual Science January 2014, Vol.55, 417-423. doi:https://doi.org/10.1167/iovs.13-12677
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Dawn A. Sim, Pearse A. Keane, Simon Fung, Michael Karampelas, Srinivas R. Sadda, Marcus Fruttiger, Praveen J. Patel, Adnan Tufail, Catherine A. Egan; Quantitative Analysis of Diabetic Macular Ischemia Using Optical Coherence Tomography. Invest. Ophthalmol. Vis. Sci. 2014;55(1):417-423. https://doi.org/10.1167/iovs.13-12677.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose.: We described the optical coherence tomography (OCT) features of diabetic macular ischemia (DMI), and correlate these findings with visual acuity (VA).

Methods.: Clinical and imaging data were collected from 100 patients with type 2 diabetes. Qualitative grading of DMI severity was determined according to criteria defined by the Early Treatment Diabetic Retinopathy Study. Quantitative analysis of foveal avascular zone (FAZ), and OCT images were performed using custom software.

Results.: In all eyes, the outer retina was thicker in eyes with DMI (167.4 ± 18.5 vs. 150.4 ± 31.4 μm, P = 0.04). However, subanalysis of eyes “without macular edema” revealed the converse; outer retinal thinning in eyes with DMI (223.1 ± 31.2 vs. 244.9 ± 37.2 μm, P = 0.007). A thinner retinal nerve fiber layer also was observed to correlate with increasing FAZ size (r = −0.231, P = 0.03), which strengthened in eyes “without macular edema” (r = −0.62, P = 0.001). In the choroid, quantification of its sublayers revealed a thicker Haller's large vessel layer in the presence of DMI (144.3 ± 51.0 vs. 103.5 ± 39.4 μm, P = 0.01). In eyes with DMI, a thicker retina was correlated with worsening VA (r = 0.52, P = 0.001). However, when eyes with macular edema were excluded, a thinner retina was associated with poor VA (r = −0.37, P = 0.004).

Conclusions.: In eyes with DMI, we observed thinning of the retinal nerve fiber layer, outer retina, and thickening of Haller's large vessel layer of the choroid. These parameters showed good correlation with VA and may serve as a useful tool for monitoring DMI in clinical practice or future clinical trials.

Introduction
Diabetic macular ischemia (DMI), an important category of diabetic retinopathy, is characterized by occlusion and atrophy of retinal capillaries in the macula, with narrowing or obliteration of precapillary arterioles. 1 The resulting tissue hypoxia leads to upregulation of growth factors, such as vascular endothelial growth factor, and, thus, contributes to the evolution of diabetic macular edema (DME), the most common cause of visual loss in diabetes. 2,3 In some cases, however, more extensive ischemia develops, and becomes the predominate feature of the maculopathy—in such eyes, profound and irreversible visual loss, out of proportion to that expected from any concomitant edema, may occur. 4,5  
Although the visual importance of macular ischemia was established by a number of seminal fluorescein angiographic studies in the 1970s, many clinical questions remain unclear. 68 In particular, the appearance or consequences of macular ischemia on optical coherence tomography (OCT) imaging remains ill defined—a significant shortcoming given the widespread adoption of OCT for monitoring of treatment in clinical trials of diabetic retinopathy. A number of preliminary studies have suggested severe macular ischemia is associated with thinning and disorganization of retinal layers on OCT. 911 In the bulk of OCT studies in diabetic retinopathy, however, the effects of macular ischemia are masked by coexisting DME. It perhaps is not surprising then, that only modest correlations have been observed to date, between OCT-derived retinal thickness and visual acuity (VA), in these studies. 10,12  
Fortunately, however, recent advances in OCT imaging offer the opportunity to characterize DMI morphology more precisely using OCT. In particular, analysis of OCT image sets has become more sophisticated in recent years, with a move from simple measurements of retinal thickness to more detailed quantitative subanalysis of individual retinal layers and specific disease features. 13 In addition, the formulation of “enhanced depth imaging” scanning protocols now allows high-resolution visualization of choroidal anatomy using the current generation of commercial OCT systems. 14 In a recent study, we combined these advances to describe novel segmentation protocols for the assessment of the choroid in diabetes, and observed that the choroidal vascular sublayers can be quantified with good reproducibility. 15  
In the current study, we performed detailed quantitative OCT analysis of retinal and choroidal morphology in a cohort of patients with DMI. Moreover, we attempted to account for the presence of coexisting DME, and to focus on specific retinal and choroidal layers hypothesized to be of functional significance in the context of DMI. 
Materials and Methods
Inclusion Criteria and Data Collection
Clinical and imaging data were collected retrospectively from patients attending medical retinal clinics at Moorfields Eye Hospital, London, United Kingdom. All data were collected over a six-month time period. Approval for data collection and analysis was obtained from the local ethics committee, and adhered to the tenets set forth in the Declaration of Helsinki. 
Patients with a diagnosis of type 2 diabetes mellitus who had undergone fluorescein angiography (FA) and spectral domain OCT scanning (Spectralis; Heidelberg Engineering, Heidelberg, Germany) were identified. In most cases (72/100 eyes), FA and OCT images were obtained on the date of clinic attendance in the clinic; however, patients also were included in the study if angiography and OCT images were acquired within 6 weeks of the study attendance date. For inclusion in the study, all angiographic and OCT image sets had to be of sufficient quality to allow grading of DMI severity and segmentation of retinal or choroidal boundaries. No image manipulation was performed before analysis. Boundaries were segmented by trained graders (SF, MK) who were masked to associated VA information at the time of grading. Patients with ocular comorbidities, including retinal arterial or venous occlusion, epiretinal membrane, neovascular age-related macular degeneration, inherited macular disease, or macular scarring of any etiology, were excluded. 
Patient demographic data, VA, and retinopathy/maculopathy grades, were obtained from standardized electronic reports in the United Kingdom National Screening Committee –Diabetic Eye Screening Programme, a grading system described in more detail previously. 16  
Acquisition and Analysis of Fluorescein Angiograms
Grading Methods.
All angiographic images were acquired with a digital retinal camera system (Topcon TRC 50IX; Topcon Medical Systems, Inc., Paramus, NJ). A macular centered FA image was dual-graded by two masked assessors using protocols and standard photographs from Early Treatment Diabetic Retinopathy Study (ETDRS) Report No. 11. 17 According to these criteria, DMI was classified as DMI grades of none, questionable, mild, moderate, or severe. 
The presence of additional areas of capillary nonperfusion, noncontiguous with the foveal avascular zone (FAZ), also was noted at the time of grading. “Papillomacular ischemia” was defined as the area of retina bordered by the temporal edge of the optic disc and the nasal edge of the FAZ, and corresponding to the nasal quadrants of the ETDRS grid. “Temporal ischemia” was defined as the area located one disc diameter temporal to the central fovea, and within the superior and inferior temporal vascular arcades. These novel methods of grading DMI have been described in detail previously. 15  
Quantification of the FAZ and Other Areas of Capillary Nonperfusion.
Quantitative analysis of all images was performed using a validated image viewer and grading software package (GRADOR; Doheny Image Reading Center, Los Angeles, CA). Using this software, the areas of the FAZ, and other areas of capillary nonperfusion, were assessed in square millimeters (mm2). 
Acquisition and Analysis of OCT Image Sets
Grading Methods.
All Spectralis OCT image sets contained a minimum of 13 B-scans distributed in a horizontal raster pattern overlying the area covered by the nine subfields of the ETDRS grid. Custom image analysis software (OCTOR; Doheny Image Reading Center), which has been described and validated in previous reports, was used for quantitative OCT analysis. 18,19 The DME was defined as a central foveal thickness (ETDRS central subfield) of greater than 275 μm. 
Segmentation Protocol for Retinal Spaces.
Boundaries were segmented manually in accordance with standardized OCT grading protocols (Fig. 1). 20 The retinal spaces were defined as retinal nerve fiber layer (RNFL), inner, outer, and total retina (Figs. 1ad). The “inner retina” was defined as the space lying between the inner aspect of the internal limiting membrane and the inner border of the outer plexiform layer, while the “outer retina” was defined as the space lying between the inner border of the outer plexiform layer and the inner aspect of the retinal pigment epithelium. 21,22 The RNFL, a hyperreflective layer that lies within the inner retina, is bordered by the internal limiting membrane and inner border of the ganglion cell layer. 
Figure 1
 
(a, b) Examples of a fluorescein angiograph in an eye with DMI. Green horizontal lines represent OCT B-scans (cf) acquired using the enhanced depth OCT protocol. (c) The RNFL area (green) was segmented in 13 B-scans acquired over the macula. (d) Similarly, measurements of the inner retina (red), outer retina (green), (e) total choroid (blue), (f) Sattler's medium vessel layer (yellow), and Haller's large vessel layer (blue) were taken.
Figure 1
 
(a, b) Examples of a fluorescein angiograph in an eye with DMI. Green horizontal lines represent OCT B-scans (cf) acquired using the enhanced depth OCT protocol. (c) The RNFL area (green) was segmented in 13 B-scans acquired over the macula. (d) Similarly, measurements of the inner retina (red), outer retina (green), (e) total choroid (blue), (f) Sattler's medium vessel layer (yellow), and Haller's large vessel layer (blue) were taken.
Segmentation Protocol for Choroidal Spaces.
The choroid was defined as the space between the outer border of the retinal pigment epithelium and the inner border of the sclera (Figs. 1a, 1b, 1e, 1f). The choroid was subdivided further into Haller's large vessel and Sattler's medium vessel layers (Fig. 1f). Haller's large vessel layer was defined as the outer choroid consisting of large hypointense spaces representing large vascular luminal spaces. Sattler's medium vessel layer consisted of small-to-medium size hypointense spaces, giving a mottled appearance on scans. For the purposes of grading, this layer also included the choriocapillaris, which is not easily indistinguishable on OCT images. 
The mean retinal choroidal thickness (μm) and volume (mm3) were calculated for the total macular circle (TMC, ETDRS areas 1–9), foveal central subfield (FCS, ETDRS central subfield), and areas over the temporal and papillomacular quadrants. 
Assessment of Segmentation Reproducibility.
A set of 20 image sets were selected randomly and segmented twice to assess intragrader reproducibility. To assess intergrader reproducibility of measurements, 51 image sets were segmented in a masked fashion by two graders. 
Statistical Analysis
Clinical and imaging data were analyzed with frequency and descriptive statistics. Snellen visual acuities were converted to logMAR (logarithm of the minimum angle of resolution) VA for the purposes of statistical analysis. The κ coefficient and Bland-Altman plots were used to assess intra- and intergrader reproducibility. Normality of the variables was examined using histograms. As anticipated, the distribution of all parameters was skewed and no simple transformation of data redressed the skewness. 
The Kruskal Wallis test and Spearman's correlation were used to assess the relationship between the EDTRS grade for macular ischemia, FA-derived measurements for areas of capillary nonperfusion, and OCT-derived parameters for the retinal and choroid, with VA. 
Statistical analysis was performed using SPSS software version 16.0 for Windows (SPSS, Inc., Chicago, IL). P values < 0.05 were considered significant. 
Results
Baseline Characteristics
A total of 100 consecutive patients with type 2 diabetes mellitus was included in the study. Of these, 60 patients underwent “enhanced depth imaging” (EDI). 14 In these cases, analysis of the choroid also was performed. 
A total of 12 patients did not have FA images of sufficient quality to permit grading of macular ischemia in either eye. These patients were excluded from analysis. From each patient, we selected the eye with the greatest macular ischemia severity, using permutated block randomization where macular ischemia severity was symmetrical (n = 32). This was to ensure inclusion of patients with a greater DMI severity within our patient cohort. 
For the 100 eyes that were analyzed, the male-to-female ratio was 1:3, and mean patient age was 62 years (SD 12). The proportions of eyes with ETDRS defined grades of retinopathy and maculopathy are presented in Table 1. In the 88 eyes graded for macular ischemia, 34 eyes (38.6%) had none, 12 (13.6%) questionable, 23 (26.1%) mild, 6 (6.8%) moderate, and 13 (14.8%) severe ETDRS-defined DMI grades (Table 1). 
Table 1
 
Demographics and Clinical Characteristics of Patients With Type 2 Diabetes Mellitus
Table 1
 
Demographics and Clinical Characteristics of Patients With Type 2 Diabetes Mellitus
Clinical Features
Age n (SD) 62 (12)
Sex M:F 1.3
Diabetic retinopathy grade, n = 100
 No DR n 5
 Nonproliferative DR n 54
 Proliferative DR n 41
Diabetic maculopathy grade, n = 100
 None n 32
 DME n 30
 CSME n 38
Diabetic macular ischemia grade, n = 88
 None n (%) 34 (38.6)
 Questionable n (%) 12 (13.6)
 Mild n (%) 23 (26.1)
 Moderate n (%) 6 (6.8)
 Severe n (%) 13 (14.8)
Intra- and Intergrader Reproducibility
For this assessment of ETDRS-defined DMI grades, we observed substantial intergrader agreement, with a weighted κ of 0.704 (SE, 0.087; 95% confidence interval [CI], 0.535–0.874). 15 Quantification of retinal, choroidal, Haller's, and Sattler's layer thicknesses also showed good intra- and intergrader agreement, with an intragrader 95% limits of agreement of 19.2, 26.9, 35.2, and 29.2 μm, respectively, and intergrader 95% limits of agreement of 27.9, 41.5, 38.6, 31.1 μm, respectively. 5,15  
The Effects of Ischemia on Retinal Morphology
Coexistence of Diabetic Macular Edema With Macular Ischemia (Table 2).
Table 2
 
Prevalence of DME in Different Grades of ETDRS-Defined Grades of DMI
Table 2
 
Prevalence of DME in Different Grades of ETDRS-Defined Grades of DMI
Diabetic Macular Ischemia Grades Macular Edema, n = 49 No Macular Edema, n = 39
None, n = 34, 38.6%
n % 19, 55.9% 15, 44.1%
Questionable, n = 12, 13.6%
n % 4, 33.3% 8, 66.7%
Mild, n = 23, 26.1%
n % 13, 56.5% 10, 43.5%
Moderate, n = 6, 6.8%
n % 4, 66.7% 2, 33.3%
Severe, n = 13, 14.8%
n % 9, 69.2% 4, 30.8%
All grades, n = 54, 61.4%
n % 30, 55.6% 24, 44.4%
The condition of DME, defined as a foveal retinal thickness of greater than 275 μm, was present in 49/100 eyes (49%). The proportions of DME, within different DMI grades, are presented in Table 1. The proportion of eyes with any grade of macular ischemia was comparable in both DME subgroups (55.6% “with DME” versus 44.4% “without DME”). There were no significant differences in age or sex between both macular ischemia or DME subgroups. 
Outer Retinal Thickness Measurements in DMI (Fig. 2, Table 3).
Table 3
 
Mean Thickness Values of the Total, Inner and Outer Retina, Nerve Fiber Layer, Choroid, and Haller's Large Vessel Layer in Eyes With and Without DMI
Table 3
 
Mean Thickness Values of the Total, Inner and Outer Retina, Nerve Fiber Layer, Choroid, and Haller's Large Vessel Layer in Eyes With and Without DMI
TMC P Value FCS P Value PM, Inner P Value PM, Outer P Value Temporal, Inner P Value Temporal, Outer P Value
Mean total retinal thickness ± SD, μm
No DMI 286.3 ± 34.6 0.37 268.0 ± 80.7 0.45 275.1 ± 31.4 0.03* 307.1 ± 40.1 0.01* 301.4 ± 37.1 0.12 270.1 ± 46.4 0.15
DMI 309.6 ± 58.3 326.6 ± 133.2 297.3 ± 50.3 341.5 ± 80.6 339.7 ± 93.6 285.2 ± 64.3
Mean outer retinal thickness ± SD, μm
No DMI 134.1 ± 13.0 0.87 150.4 ± 31.4 0.04* 123.9 ± 11.7 0.23 146.9 ± 18.8 0.91 149.3 ± 18.2 0.36 129.4 ± 18.2 0.78
DMI 133.0 ± 25.0 167.4 ± 18.5 132.77 ± 29.1 147.1 ± 35.2 144.3 ± 42.2 122.5 ± 26.3
Mean inner retinal thickness ± SD, μm
No DMI 142.4 ± 26.1 0.63 71.5 ± 34.0 0.34 138.2 ± 32.6 0.33 147.9 ± 39.5 0.31 146.9 ± 28.1 0.74 133.6 ± 18.3 0.87
DMI 148.9 ± 16.3 107.4 ± 76.8 150.3 ± 23.2 155.4 ± 28.0 152.8 ± 27.2 133.3 ± 25.5
Mean retinal nerve fiber layer thickness ± SD, μm
No DMI 38.2 ± 4.3 0.84 10.8 ± 5.7 0.87 48.6 ± 13.9 0.75 25.4 ± 4.8 0.77 22.0 ± 4.0 0.88 33.7 ± 15.2 0.88
DMI 30.9 ± 7.1 8.9 ± 2.5 39.0 ± 16.6 21.5 ± 5.8 19.0 ± 3.0 26.9 ± 6.6
Mean total choroidal thickness ± SD, μm
No DMI 223.5 ± 55.0 0.07 239.3 ± 65.8 0.11 192.3 ± 60.5 0.06 231.5 ± 65.9 0.13 232.7 ± 58.4 0.09 213.5 ± 49.7 0.10
DMI 264.5 ± 66.6 282.6 ± 70.7 240.1 ± 77.3 275.6 ± 75.9 270.2 ± 73.4 245.1 ± 66.3
Mean Haller's large vessel layer thickness ± SD, μm
No DMI 104.2 ± 30.9 <0.01† 103.5 ± 39.4 <0.01† 89.9 ± 35.7 0.03* 101.2 ± 42.1 0.03* 106.0 ± 31.4 0.23 103.8 ± 26.2 0.13
DMI 138.2 ± 42.6 144.3 ± 51.0 124.8 ± 50.5 141.3 ± 51.1 143.2 ± 48.2 134.0 ± 41.5
In the analysis of all eyes, a trend towards a negative correlation was observed between FAZ area and retinal thickness measurements (r = −0.19, P = 0.07). However, we observed that outer retina measurements were significantly thicker at the FCS in eyes “with macular ischemia” (167.4 ± 18.5 μm) compared to eyes with “no macular ischemia” (150.4 ± 31.4 μm, P = 0.04). When eyes with DME were excluded, the outer retina was significantly thinner in eyes “with macular ischemia” (223.1 ± 31.2 μm) compared to eyes with “no macular ischemia” (244.9 ± 37.2 μm, P = 0.007). Inner retinal measurements were not significantly different between groups. 
Figure 2
 
Scatterplots showing the a positive relationship between VA (logMAR) and outer retinal thickness measurements (μm [a]) in eyes with DME, and a negative correlation (b) in eyes without DME (regression coefficients, r 2 = 0.11 [with DME], r 2 = −0.20 [without DME]).
Figure 2
 
Scatterplots showing the a positive relationship between VA (logMAR) and outer retinal thickness measurements (μm [a]) in eyes with DME, and a negative correlation (b) in eyes without DME (regression coefficients, r 2 = 0.11 [with DME], r 2 = −0.20 [without DME]).
Correlation of RNFL Thickness With Area of the FAZ (Table 3).
We analyzed the RNFL thickness measurements in two groups: “all eyes” with FAZ measurements (n = 88) and in eyes “without DME” (n = 39). 
In “all eyes,” the FAZ area showed weak correlation with RNFL thickness at the TMC (r = −0.231, P = 0.03), but not at the FCS (r = −0.02, P = 0.88), temporal (outer, r = −0.10, P = 0.482; inner, r = −0.11, P = 0.43), or papillomacular quadrants (outer, r = −0.21, P = 0.482; inner, r = −0.25, P = 0.14). 
In eyes “without DME,” the FAZ area showed significant negative correlation with RNFL thickness over the papillomacular quadrant, particularly in the outer quadrant (outer, r = −0.62, P < 0.001; inner, r = −0.38, P = 0.04). This was not observed at the TMC (r = −0.12, P = 0.44), FCS (r = −0.02, P = 0.884), or temporal quadrants (outer, r = −0.12, P = 0.54; inner, r = −0.27, P = 0.16). 
The Effects of Ischemia on Choroidal Morphology
Haller's Large Vessel Layer Thickness in DMI (Table 3).
Haller's large vessel layer was significantly thicker in eyes “with macular ischemia,” compared to eyes with “no macular ischemia,” both at the FCS (144.3 ± 51.0 vs. 103.5 ± 39.4 μm, P = 0.01) respectively and TMC (138.2 ± 42.6 vs. 104.2 ± 30.9 μm, P = 0.01). There was no significant difference in total choroidal thickness between at the FCS (P = 0.11) or TMC (P = 0.07). There also was no significant difference in Haller's layer or choroidal thickness between eyes with and without DME. 
Correlation of Haller's Large Vessel Layer Thickness to FAZ Area.
In eyes “without DME,” Haller's layer over the TMC was correlated with the size of FAZ area (r = 0.43, P = 0.04). This was not observed in “all eyes” (r = 0.22, P = 0.16), or eyes “with DME” (r = 0.11, P = 0.64). Interestingly, there was no correlation between Haller's layer thickness and retinal thickness measurements (r = 0.16, P = 0.31). 
Relationships Between VA and OCT-Derived Parameters
Significant correlations were observed between VA and the FAZ area (r = 0.36, P = 0.02), papillomacular ischemia area (r = 0.613, P = 0.01), but not with the temporal ischemia area (r = 0.22, P = 0.29). These relationships are consistent with findings from our previous study. 5  
Retinal Thickness and VA.
In eyes with macular ischemia and edema, we observed a positive correlation between VA and total retinal thickness (r = 0.52, P = 0.001) and outer retinal thickness measurements (r = 0.33, P = 0.04) at the FCS. Interestingly, in eyes with macular ischemia, but without edema, we observed the converse, where VA was correlated negatively with total retinal thickness (r = −0.37, P = 0.004) and outer retinal thickness measurements (r = −0.44, P = 0.001) at the FCS. There were no significant associations in the analysis of “all eyes,” or between VA and inner retinal thicknesses measurements. 
RNFL and VA.
In eyes with macular ischemia, but without edema, we observed a negative correlation between VA and RNFL thickness over the papillomacular area (r = −0.37, P = 0.004), which was lost in the analysis of “all eyes” (r = 0.20, P = 0.18), and eyes with macular ischemia and edema (r = 0.335, P = 0.08). There were no associations between VA and RNFL thickness measurements in the analysis of “all eyes.” 
Subfoveal Choroidal Thickness and VA.
The VA was correlated moderately with choroidal thickness measurements at the FCS in macular ischemia grades “mild” to “severe” (n = 42, r = 0.47, P = 0.03). This correlation was strengthened further in eyes with macular ischemia grades “moderate” to “severe” (n = 19, r = 0.74, P = 0.009). 
Discussion
In this retrospective cross-sectional study, we perform detailed quantitative analyses of OCT images obtained from patients with Type 2 diabetes mellitus. Using this approach, we examined the relationships between FA-derived areas of capillary dropout and OCT-derived parameters of the retina and choroid, and evaluated its visual significance. We also highlighted novel parameters in choroidal analysis, including quantification of Haller's large vessel layer. 
Evaluating retinal thickness measurements in the context of DMI is challenging due to the coexistence of DME. In our study, we observed that a large proportion of eyes with macular ischemia had concurrent DME—30/54 eyes (55.6%, Table 2). In the analysis of OCT image sets across the entire cohort, retinal thickening, though not significant, was observed in the total and outer retinal layers. When we excluded eyes with DME, the inverse was seen—retinal thinning in the presence of macular ischemia, particularly in the outer retina. Previous studies have shown that cystoid spaces in DME occur primarily in the outer retinal layers, potentially masking any thinning that may occur. 2325 These findings provide one possible explanation for the degree of variation seen in many OCT studies of retinal thickness in diabetic eyes. Compared to nondiabetic controls, these studies have reported thickening, 2628 thining, 17,2932 or no difference, between eyes with minimal diabetic retinopathy and nondiabetic eyes. 33,34 In our study, we examined a diabetic population with a greater severity of disease, that is, those referred from a population screening program with potentially treatable maculopathy or retinopathy. In this cohort, we observed significant outer retinal thinning at the FCS in eyes with macular ischemia, but without edema. There was moderate negative correlation with VA, which suggested that thinning may, in fact, have a functional significance. 
Inner retinal layers, including the ganglion cell layer (GCL) and RNFL, have been shown to be thinned in patients with diabetes when compared to normal controls. 9,31,32,3537 Unlike these studies, which examined eyes with minimal retinopathy, we did not observe any significant differences with inner retinal thickness measurements (inner limiting membrane [ILM], RNFL, GCL, inner plexiform, and inner nuclear layers), or the RNFL in our cohort (Table 3). Instead, we found that the RNFL in the papillomacular nerve fiber bundle and adjacent to the optic disc was correlated negatively with the size of the FAZ, and correlated strongly with VA when eyes with DME were excluded. The RNFL at the fovea is very thin (with a mean thickness of 8–10 μm), making it difficult to make comparisons between groups. On the other hand, the RNFL at the area of the papillomacular nerve fiber bundle contains axons from the entire retina, which converge to form the optic cup, and consequently is much thicker (up to 49 μm). Therefore, it is not surprising that RNFL thinning in this area is associated with a reduction in visual function, and further supports the paradigm of diabetic neurodegeneration. We also demonstrated in this study that neurodegeneration, as implied by outer retinal and RFNL thinning, is present in the context of macular ischemia. 
Photoreceptors in the fovea, usually deficient of retinal capillaries, derive oxygen from the choroidal circulation. 38 In this study, increased Haller's large vessel layer thickness measurements were observed in eyes with macular ischemia, and also correlated with FAZ size (r = 0.43, P = 0.04). Only a few studies have examined choroidal thickness in patients with diabetes. In these studies, thinning was observed in the choroid, both compared to normal controls, and in the presence of DME. 3942 In this study, we observed significant thickening in eyes with macular ischemia, but without edema. One possible explanation for this discrepancy may be the differences in severity of disease between studies. In this study, we included a large number of patients with macular ischemia (of any grade, 61.4%). Comparatively, Esmaeelpour et al. 40 examined 18 patients with “diabetic retinopathy” of varying severity, including proliferative diabetic retinopathy, nonproliferative diabetic retinopathy, clinically significant macular edema, and 15 patients with no diabetic retinopathy, while Querques et al. 41 examined eyes with nonproliferative diabetic retinopathy with and without clinically significant macular edema. 
It has been suggested that decreased choroidal blood flow, as observed in Doppler flow studies, and consequent “diabetic choroidopathy” observed in histologic studies, are possible explanations for thinning of the choroid. 4346 However, flow dynamics of the choroidal circulation, though beyond the scope of this paper, do not have a straightforward relationship with choroidal thickness. 47,48 In the current report, we were unable to distinguish consistently the choriocapillaris from Sattler's medium vessel layer on OCT, and, therefore, were unable to assess “diabetic choroidopathy” in this layer. We hypothesized that the choroidal vasculature in the FAZ, an area of high metabolic demand, may be upregulated in the presence of macular ischemia. A similar mechanism, mediated by intrinsic choroidal neurons, has been described in changing choroidal thickness in response to retinal defocus. 49 These neurons reside in the muscular walls of arterioles, nonvascular smooth muscle cells, and span the stroma and lymphatic lacunae of the choroid. Intrinsic choroidal neurons are thought to be vasodilatative, regulating choroidal thickness via the release of nitric oxide. We speculated that local ischemic changes may trigger this vasodilatative response. 
The strengths of this study lie principally in the detailed quantitative analyses performed for both FA and OCT image sets. Our use of FA grading software equipped with standard planimetric tools allowed quantification of FAZ area, enabling detection of small changes over time. The quantitative OCT subanalyses used in this study have been validated previously, and allowed quantification over the entire length of scan, as compared to single caliper measurements used by most studies. 20 The limitations of this study are inherent in its retrospective nature, and the inclusion of eyes with more severe diabetic retinopathy than typically would be observed in a clinical setting. However, this may better reflect a clinical practice or trial settings, where patients who are receiving treatment would have a similar level of pathology. In addition, 28/100 eyes did not have FA and OCT at the same sitting, but within 6 weeks of each other. It is possible that the anatomic states of macular edema and ischemia may have changed in the ensuing 6 weeks; however, the limited evidence available on the natural history of both entities suggests that progression occurs over a longer period of time. Regarding macular edema, most of the natural history data originate from the time of the ETDRS, when OCT had yet to come into clinical practice. The more recent Diabetic Retinopathy Clinical Research Network (DRCRnet) studies, have a follow-up interval of 3 to 4 months between treatments, suggesting that changes in retinal thickness measurements occur within that time frame. In macular ischemia, we have observed, in a separate study, that the rate of change is slow even in established ischemia, where the area of the FAZ increases at a rate of 5% to 10% of baseline per year. 50  
In summary, in patients with type 2 diabetes mellitus, DMI was associated with thinning of the outer retina at the fovea, and the RNFL in the papillomacular bundle. The presence of these factors was associated with impairment of visual function. We also observed thickening of Haller's large vessel layer of the choroid in the presence of ischemia. These parameters, if confirmed in prospective studies, may be a useful noninvasive tool for monitoring of ischemia in diabetic maculopathy. 
Acknowledgments
Supported by Fight For Sight UK, Grant 1987 (DAS, MF), partly by the Department of Health's NIHR Biomedical Research Centre for Ophthalmology at Moorfields Eye Hospital and UCL Institute of Ophthalmology (PAK, PJP, CAE, DAS, AT), by travel grants from the Allergan European Retina Panel (PAK, DAS, PJP), and by Carl Zeiss Meditec, Optos, and Optovue, Inc. (SRS). The authors alone are responsible for the content and writing of the paper. 
Disclosure: D.A. Sim, Allergan European Retina Panel (R); P.A. Keane, Allergan European Retina Panel (R); S. Fung, None; M. Karampelas, None; S.R. Sadda, Carl Zeiss Meditec (F, C), Optos (F, C), Optovue (F), Genentech, Inc. (C), Regeneron (C), Allergan, Inc. (C); M. Fruttiger, None; P.J. Patel, Allergan European Retina Panel (R), Novartis UK (S); A. Tufail, Novartis (S), Pfizer (S), GSK (S), Thrombogenics (S), Bayer (S), Allergan, Inc. (S); C.A. Egan, None 
References
Arend O Wolf S Jung F Retinal microcirculation in patients with diabetes mellitus: dynamic and morphological analysis of perifoveal capillary network. Br J Ophthalmol . 1991; 75: 514–518. [CrossRef] [PubMed]
Aiello LP Wong JS. Role of vascular endothelial growth factor in diabetic vascular complications. Kidney Int . 2000; 77 (suppl): S113–S119. [CrossRef]
Grant MB Afzal A Spoerri P The role of growth factors in the pathogenesis of diabetic retinopathy. Expert Opin Investig Drugs . 2004; 13: 1275–1293. [CrossRef] [PubMed]
Arend O Wolf S Harris A Reim M. The relationship of macular microcirculation to visual acuity in diabetic patients. Arch Ophthalmol . 1995; 113: 610–614. [CrossRef] [PubMed]
Sim DA Keane PA Zarranz-Ventura J The effects of macular ischemia on visual acuity in diabetic retinopathy. Invest Ophthalmol Vis Sci . 2013; 54: 2353–2360. [CrossRef] [PubMed]
Kohner EM Henkind P. Correlation of fluorescein angiogram and retinal digest in diabetic retinopathy. Am J Ophthalmol . 1970; 69: 403–414. [CrossRef] [PubMed]
Bresnick GH Engerman R Davis MD Patterns of ischemia in diabetic retinopathy. Trans Sect Ophthalmol Am Acad Ophthalmol Otolaryngol . 1976; 814(pt 1): OP694–OP709.
Bresnick GH De Venecia G Myers FL Retinal ischemia in diabetic retinopathy. Arch Ophthalmol . 1975; 93: 1300–1310. [CrossRef] [PubMed]
Byeon SH Chu YK Lee H Foveal ganglion cell layer damage in ischemic diabetic maculopathy: correlation of optical coherence tomographic and anatomic changes. Ophthalmology . 2009; 116: 1949–1959. [CrossRef] [PubMed]
Lee DH Kim JT Jung DW The relationship between foveal ischemia and spectral-domain optical coherence tomography findings in ischemic diabetic macular edema. Invest Ophthalmol Vis Sci . 2013; 54: 1080–1085. [CrossRef] [PubMed]
Reznicek L Kernt M Haritoglou C In vivo characterization of ischemic retina in diabetic retinopathy. Clin Ophthalmol . 2010; 5: 31–35. [CrossRef] [PubMed]
Alasil T Keane PA Updike JF Relationship between optical coherence tomography retinal parameters and visual acuity in diabetic macular edema. Ophthalmology . 2010; 117: 2379–2386. [CrossRef] [PubMed]
Keane PA Heussen FM Ouyang Y Assessment of differential pharmacodynamic effects using optical coherence tomography in neovascular age-related macular degeneration. Invest Ophthalmol Vis Sci . 2012; 53: 1152–1161. [CrossRef] [PubMed]
Spaide RF Koizumi H Pozzoni MC. Enhanced depth imaging spectral-domain optical coherence tomography. Am J Ophthalmol . 2008; 146: 496–500. [CrossRef] [PubMed]
Sim DA Keane PA Mehta H Repeatability and reproducibility of choroidal vessel layer measurements in diabetic retinopathy using enhanced depth optical coherence tomography. Invest Ophthalmol Vis Sci . 2013; 54: 2893–2901. [CrossRef] [PubMed]
Mackenzie S Schmermer C Charnley A SDOCT imaging to identify macular pathology in patients diagnosed with diabetic maculopathy by a digital photographic retinal screening programme. PLoS One . 2011; 6: e14811. [CrossRef] [PubMed]
Classification of diabetic retinopathy from fluorescein angiograms. ETDRS report number 11. Early Treatment Diabetic Retinopathy Study Research Group. Ophthalmology . 1991; 98 (5 suppl): 807–822. [PubMed]
Sadda SR Wu Z Walsh AC Errors in retinal thickness measurements obtained by optical coherence tomography. Ophthalmology . 2006; 113: 285–293. [CrossRef] [PubMed]
Sadda SR Joeres S Wu Z Error correction and quantitative subanalysis of optical coherence tomography data using computer-assisted grading. Invest Ophthalmol Vis Sci . 2007; 48: 839–848. [CrossRef] [PubMed]
Keane PA Liakopoulos S Ongchin SC Quantitative subanalysis of optical coherence tomography after treatment with ranibizumab for neovascular age-related macular degeneration. Invest Ophthalmol Vis Sci . 2008; 49: 3115–3120. [CrossRef] [PubMed]
Pons ME Garcia-Valenzuela E. Redefining the limit of the outer retina in optical coherence tomography scans. Ophthalmology . 2005; 112: 1079–1085. [CrossRef] [PubMed]
Costa RA Calucci D Skaf M Optical coherence tomography 3: automatic delineation of the outer neural retinal boundary and its influence on retinal thickness measurements. Invest Ophthalmol Vis Sci . 2004; 45: 2399–2406. [CrossRef] [PubMed]
Otani T Kishi S Maruyama Y. Patterns of diabetic macular edema with optical coherence tomography. Am J Ophthalmol . 1999; 127: 688–693. [CrossRef] [PubMed]
Otani T Kishi S. Correlation between optical coherence tomography and fluorescein angiography findings in diabetic macular edema. Ophthalmology . 2007; 114: 104–107. [CrossRef] [PubMed]
Yeung L Lima VC Garcia P Correlation between spectral domain optical coherence tomography findings and fluorescein angiography patterns in diabetic macular edema. Ophthalmology . 2009; 116: 1158–1167. [CrossRef] [PubMed]
Lattanzio R Brancato R Pierro L Macular thickness measured by optical coherence tomography (OCT) in diabetic patients. Eur J Ophthalmol . 2002; 12: 482–487. [PubMed]
Sugimoto M Sasoh M Ido M Detection of early diabetic change with optical coherence tomography in type 2 diabetes mellitus patients without retinopathy. Ophthalmologica . 2005; 219: 379–385. [CrossRef] [PubMed]
Fritsche P van der Heijde R Suttorp-Schulten MS Polak BC. Retinal thickness analysis(RTA): an objective method to assess and quantify the retinal thickness in healthy controls and in diabetics without diabetic retinopathy. Retina . 2002; 22: 768–771. [CrossRef] [PubMed]
Nilsson M von Wendt G Wanger P Martin L. Early detection of macular changes in patients with diabetes using Rarebit Fovea Test and optical coherence tomography. Br J Ophthalmol . 2007; 91: 1596–1598. [CrossRef] [PubMed]
Verma A Rani PK Raman R Is neuronal dysfunction an early sign of diabetic retinopathy? Microperimetry and spectral domain optical coherence tomography (SD-OCT) study in individuals with diabetes, but no diabetic retinopathy. Eye (Lond) . 2009; 23: 1824–1830. [CrossRef] [PubMed]
van Dijk HW Verbraak FD Stehouwer M Association of visual function and ganglion cell layer thickness in patients with diabetes mellitus type 1 and no or minimal diabetic retinopathy. Vision Res . 2011; 51: 224–228. [CrossRef] [PubMed]
van Dijk HW Verbraak FD Kok PH Early neurodegeneration in the retina of type 2 diabetic patients. Invest Ophthalmol Vis Sci . 2012; 53: 2715–2719. [CrossRef] [PubMed]
Bressler NM Edwards AR Antoszyk AN Retinal thickness on Stratus optical coherence tomography in people with diabetes and minimal or no diabetic retinopathy. Am J Ophthalmol . 2008; 145: 894–901. [CrossRef] [PubMed]
Kashani AH Zimmer-Galler IE Shah SM Retinal thickness analysis by race, gender, and age using Stratus OCT. Am J Ophthalmol . 2010; 149: 496–502. [CrossRef] [PubMed]
Sugimoto M Sasoh M Ido M Retinal nerve fiber layer decrease during glycemic control in type 2 diabetes [published online ahead of print August 8, 2010]. J Ophthalmol . doi:10.1155/2010/569215 .
Verma A Raman R Vaitheeswaran K Does neuronal damage precede vascular damage in subjects with type 2 diabetes mellitus and having no clinical diabetic retinopathy? Ophthalmic Res . 2012; 47: 202–207. [CrossRef] [PubMed]
Shahidi AM Sampson GP Pritchard N Retinal nerve fibre layer thinning associated with diabetic peripheral neuropathy. Diabet Med . 2012; 29: e106–e111. [CrossRef] [PubMed]
Linsenmeier RA Braun RD. Oxygen distribution and consumption in the cat retina during normoxia and hypoxemia. J Gen Physiol . 1992; 99: 177–197. [CrossRef] [PubMed]
Esmaeelpour M Povazay B Hermann B Mapping choroidal and retinal thickness variation in type 2 diabetes using three-dimensional 1060-nm optical coherence tomography. Invest Ophthalmol Vis Sci . 2011; 52: 5311–5316. [CrossRef] [PubMed]
Esmaeelpour M Brunner S Ansari-Shahrezaei S Choroidal thinning in diabetes type 1 detected by 3-dimensional 1060 nm optical coherence tomography. Invest Ophthalmol Vis Sci . 2012; 53: 6803–6809. [CrossRef] [PubMed]
Querques G Lattanzio R Querques L Enhanced depth imaging optical coherence tomography in type 2 diabetes. Invest Ophthalmol Vis Sci . 2012; 53: 6017–6024. [CrossRef] [PubMed]
Regatieri CV Branchini L Carmody J Choroidal thickness in patients with diabetic retinopathy analyzed by spectral-domain optical coherence tomography. Retina . 2012; 32: 563–568. [CrossRef] [PubMed]
Hidayat AA Fine BS. Diabetic choroidopathy. Light and electron microscopic observations of seven cases. Ophthalmology . 1985; 92: 512–522. [CrossRef] [PubMed]
McLeod DS Lutty GA. High-resolution histologic analysis of the human choroidal vasculature. Invest Ophthalmol Vis Sci . 1994; 35: 3799–3811. [PubMed]
Nagaoka T Kitaya N Sugawara R Alteration of choroidal circulation in the foveal region in patients with type 2 diabetes. Br J Ophthalmol . 2004; 88: 1060–1063. [CrossRef] [PubMed]
Schocket LS Brucker AJ Niknam RM Foveolar choroidal hemodynamics in proliferative diabetic retinopathy. Int Ophthalmol . 2004; 25: 89–94. [CrossRef] [PubMed]
Sogawa K Nagaoka T Takahashi A Relationship between choroidal thickness and choroidal circulation in healthy young subjects. Am J Ophthalmol . 2012; 153: 1129–1132. e1. [CrossRef] [PubMed]
Nickla DL Wallman J. The multifunctional choroid. Prog Retin Eye Res . 2010; 29: 144–168. [CrossRef] [PubMed]
Poukens V Glasgow BJ Demer JL. Nonvascular contractile cells in sclera and choroid of humans and monkeys. Invest Ophthalmol Vis Sci . 1998; 39: 1765–1774. [PubMed]
Sim DA Keane PA Zarranz-Ventura J Predictive factors for the progression of diabetic macular ischemia. Am J Ophthalmol . 2013; 156: 684–692. [CrossRef] [PubMed]
Figure 1
 
(a, b) Examples of a fluorescein angiograph in an eye with DMI. Green horizontal lines represent OCT B-scans (cf) acquired using the enhanced depth OCT protocol. (c) The RNFL area (green) was segmented in 13 B-scans acquired over the macula. (d) Similarly, measurements of the inner retina (red), outer retina (green), (e) total choroid (blue), (f) Sattler's medium vessel layer (yellow), and Haller's large vessel layer (blue) were taken.
Figure 1
 
(a, b) Examples of a fluorescein angiograph in an eye with DMI. Green horizontal lines represent OCT B-scans (cf) acquired using the enhanced depth OCT protocol. (c) The RNFL area (green) was segmented in 13 B-scans acquired over the macula. (d) Similarly, measurements of the inner retina (red), outer retina (green), (e) total choroid (blue), (f) Sattler's medium vessel layer (yellow), and Haller's large vessel layer (blue) were taken.
Figure 2
 
Scatterplots showing the a positive relationship between VA (logMAR) and outer retinal thickness measurements (μm [a]) in eyes with DME, and a negative correlation (b) in eyes without DME (regression coefficients, r 2 = 0.11 [with DME], r 2 = −0.20 [without DME]).
Figure 2
 
Scatterplots showing the a positive relationship between VA (logMAR) and outer retinal thickness measurements (μm [a]) in eyes with DME, and a negative correlation (b) in eyes without DME (regression coefficients, r 2 = 0.11 [with DME], r 2 = −0.20 [without DME]).
Table 1
 
Demographics and Clinical Characteristics of Patients With Type 2 Diabetes Mellitus
Table 1
 
Demographics and Clinical Characteristics of Patients With Type 2 Diabetes Mellitus
Clinical Features
Age n (SD) 62 (12)
Sex M:F 1.3
Diabetic retinopathy grade, n = 100
 No DR n 5
 Nonproliferative DR n 54
 Proliferative DR n 41
Diabetic maculopathy grade, n = 100
 None n 32
 DME n 30
 CSME n 38
Diabetic macular ischemia grade, n = 88
 None n (%) 34 (38.6)
 Questionable n (%) 12 (13.6)
 Mild n (%) 23 (26.1)
 Moderate n (%) 6 (6.8)
 Severe n (%) 13 (14.8)
Table 2
 
Prevalence of DME in Different Grades of ETDRS-Defined Grades of DMI
Table 2
 
Prevalence of DME in Different Grades of ETDRS-Defined Grades of DMI
Diabetic Macular Ischemia Grades Macular Edema, n = 49 No Macular Edema, n = 39
None, n = 34, 38.6%
n % 19, 55.9% 15, 44.1%
Questionable, n = 12, 13.6%
n % 4, 33.3% 8, 66.7%
Mild, n = 23, 26.1%
n % 13, 56.5% 10, 43.5%
Moderate, n = 6, 6.8%
n % 4, 66.7% 2, 33.3%
Severe, n = 13, 14.8%
n % 9, 69.2% 4, 30.8%
All grades, n = 54, 61.4%
n % 30, 55.6% 24, 44.4%
Table 3
 
Mean Thickness Values of the Total, Inner and Outer Retina, Nerve Fiber Layer, Choroid, and Haller's Large Vessel Layer in Eyes With and Without DMI
Table 3
 
Mean Thickness Values of the Total, Inner and Outer Retina, Nerve Fiber Layer, Choroid, and Haller's Large Vessel Layer in Eyes With and Without DMI
TMC P Value FCS P Value PM, Inner P Value PM, Outer P Value Temporal, Inner P Value Temporal, Outer P Value
Mean total retinal thickness ± SD, μm
No DMI 286.3 ± 34.6 0.37 268.0 ± 80.7 0.45 275.1 ± 31.4 0.03* 307.1 ± 40.1 0.01* 301.4 ± 37.1 0.12 270.1 ± 46.4 0.15
DMI 309.6 ± 58.3 326.6 ± 133.2 297.3 ± 50.3 341.5 ± 80.6 339.7 ± 93.6 285.2 ± 64.3
Mean outer retinal thickness ± SD, μm
No DMI 134.1 ± 13.0 0.87 150.4 ± 31.4 0.04* 123.9 ± 11.7 0.23 146.9 ± 18.8 0.91 149.3 ± 18.2 0.36 129.4 ± 18.2 0.78
DMI 133.0 ± 25.0 167.4 ± 18.5 132.77 ± 29.1 147.1 ± 35.2 144.3 ± 42.2 122.5 ± 26.3
Mean inner retinal thickness ± SD, μm
No DMI 142.4 ± 26.1 0.63 71.5 ± 34.0 0.34 138.2 ± 32.6 0.33 147.9 ± 39.5 0.31 146.9 ± 28.1 0.74 133.6 ± 18.3 0.87
DMI 148.9 ± 16.3 107.4 ± 76.8 150.3 ± 23.2 155.4 ± 28.0 152.8 ± 27.2 133.3 ± 25.5
Mean retinal nerve fiber layer thickness ± SD, μm
No DMI 38.2 ± 4.3 0.84 10.8 ± 5.7 0.87 48.6 ± 13.9 0.75 25.4 ± 4.8 0.77 22.0 ± 4.0 0.88 33.7 ± 15.2 0.88
DMI 30.9 ± 7.1 8.9 ± 2.5 39.0 ± 16.6 21.5 ± 5.8 19.0 ± 3.0 26.9 ± 6.6
Mean total choroidal thickness ± SD, μm
No DMI 223.5 ± 55.0 0.07 239.3 ± 65.8 0.11 192.3 ± 60.5 0.06 231.5 ± 65.9 0.13 232.7 ± 58.4 0.09 213.5 ± 49.7 0.10
DMI 264.5 ± 66.6 282.6 ± 70.7 240.1 ± 77.3 275.6 ± 75.9 270.2 ± 73.4 245.1 ± 66.3
Mean Haller's large vessel layer thickness ± SD, μm
No DMI 104.2 ± 30.9 <0.01† 103.5 ± 39.4 <0.01† 89.9 ± 35.7 0.03* 101.2 ± 42.1 0.03* 106.0 ± 31.4 0.23 103.8 ± 26.2 0.13
DMI 138.2 ± 42.6 144.3 ± 51.0 124.8 ± 50.5 141.3 ± 51.1 143.2 ± 48.2 134.0 ± 41.5
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×