Sixty patients with type 2 DM and 60 age-and sex-matched healthy controls were recruited for this case-control study. All procedures adhered to the tenets of the Declaration of Helsinki, and the local ethics committee (CEICA, ethics committee for scientific research in Aragon) approved the experimental protocol. All subjects provided informed written consent to participate in the study. The study was performed at Miguel Servet University Hospital, Zaragoza, Spain.
The diagnosis of type 2 DM was established by an experienced endocrinologist based on the American Diabetes Association criteria  and only subjects without diabetic retinopathy (at least 12 months prior recruitment) were eligible for the study.
The endocrine evaluation was carried out by a trained specialist from the Endocrine and Nutrition department of Miguel Servet University Hospital and provided demographic and disease related parameters. Age at the time of diagnosis, disease duration, prescribed treatments (for DM2 and other comorbidities), presence of micro- and macrovascular complications (such as cardiovascular disease, cerebrovascular disease, diabetic neuropathy or diabetic nephropathy) and presence of conditions associated with chronic complications in DM (such as arterial hypertension, hypercholesterolemia and obesity) were recorded. Endocrine variables such as glycosylated hemoglobin (HbA1C; most recent measured levels), microalbumin/creatinine ratio and total, HDL (high density lipoprotein) and LDL (low density lipoprotein) cholesterol levels were obtained through blood analysis during routine check-ups and recorded for the study. Arterial blood pressure and body mass index (BMI) were also registered. Subjects with arterial blood pressure levels of >140/90 mmHg or under current hypotensive treatment were classified as “hypertensive”; subjects with a calculated BMI of ≥ 25 were classified as “overweight/obese”. Smoking habit in patients and controls was also registered.
Inclusion criteria were: confirmed type 2 DM diagnosis of at least 6 months, best-corrected visual acuity (BCVA) of +0.4 Log MAR in each eye to allow performance of the protocol, and intraocular pressure less than 21 mmHg. Exclusion criteria were: the presence or past history of diabetic retinopathy, confirmed by indirect funduscopy or retinography images, presence of significant refractive errors (≥5 diopters of spherical equivalent refraction or 3 diopters of astigmatism), axial length >26 mm or < 22 mm, intraocular pressure ≥21 mmHg; media opacifications, concomitant ocular diseases, including history of glaucoma or retinal pathology, and systemic conditions that could affect the visual system, including neurodegenerative disorders such as Parkinson’s disease, multiple sclerosis or dementia. Healthy controls had no history nor evidence of ocular or neurologic disease of any nature; their BCVA was ≤+0.2 Log MAR. Only one eye per subject was randomly selected and included.
All subjects underwent a complete neuro-ophthalmic evaluation that included an in-depth funduscopic examination. Visual function was evaluated following a systematic protocol described previously by our team , using the ETDRS chart (for BCVA analysis), the CVS-1000E and Pelli-Robson tests (for contrast sensitivity vision –CSV- assessment), the Farnsworth desaturated D15 and L´Anthony desaturated D15 tests (for color vision analysis) and the automated perimeter Easyfield (Oculus Optikgeräte GmbH, Wetzlar, Germany).
Visual acuity (VA) was assessed at three different contrast levels: 100% (high contrast VA, using ETDRS chart), 2.50%, and 1.25% (low contrast VA, using Low-Contrast Sloan Letter Charts -Precision Vision, LaSalle, IL-). All measurements were obtained under controlled lighting conditions and monocular vision with best correction, BCVA at all contrast levels was recorded in Log Mar.
Contrast sensitivity was evaluated to obtain more complete information about visual function than results provided by visual acuity tests. CSV was assessed using the Pelli-Robson chart and the CVS-1000E test. The Pelli-Robson chart evaluates CSV at a frequency of 1 cycle per degree (cpd). The test comprises horizontal lines of capital letters organized into triplets, with two triplets per line. Contrast decreases from one triplet to the next. All patients were evaluated under controlled photopic conditions (85 cd/m2) at a distance of 1 meter from the chart. The score corresponding to the last triplet of letters seen by the patient was recorded.
The CSV-1000E instrument is used worldwide for standardized CSV and glare testing. The chart comprises four rows with 17 circular patches each. The patches present a grating that decreases in contrast moving from left to right across the row. All patients were evaluated at a distance of 2.5 meters from the chart under monocular vision at 4 different spatial frequencies (3, 6, 12, and 18 cpd). Each contrast value for each spatial frequency was transformed into a logarithmic scale according to standardized values.
Color vision was assessed using the Color Vision Recorder (CVR) program. CVR software analyzes chromatic discrimination by classification of hues. The program includes the classic test of Farnsworth 100-hue (FM-100), Farnsworth - Munsell D15, and L´Anthony D15. Only the last two were used in our study and both tests were performed under monocular vision. The different output parameters such as the Color Confusion Index (CCI, representing the ratio between the radius or distance between caps), the Confusion angle (Conf Ang, which is the axis of color deficiency), and the Scatter Index (S-index, the parallelism of confusion vectors to the personal confusion angle) were recorded.[12,13]
Automated perimetry was evaluated with the Easyfield perimeter (Oculus Optikgeräte GmbH, Wetzlar, Germany). This visual field analyzer is equipped with LED stimuli of a maximum of 10,000 apostles (asb), presented on a background of 31.5 asb. The QuickSpark protocol (with a brief short and fast luminous flash), which performs a rapid (less than three minutes) but very accurate measurement of the thresholds in the central field of vision was used in all subjects. Refraction for adequate proximal vision was calculated for each subject; the test was performed in monocular vision. The parameters recorded were mean sensitivity (arithmetic mean of the sensitivities of all the points studied), mean defect (arithmetic mean of the differences in sensitivities of each point with respect to the normal value for a healthy person of the same age), deviation on the pattern, and reliability factor (percentage of failed answers). As a criterion of reliability, in case of results with more than 50% of losses in detected stimuli, the test was repeated a second time. In 3 of the subjects, a reliable test was not obtained despite a second attempt, so the result of this test was discarded and was not included in the statistical analysis.
All data analyses were performed using SPSS software version 20.0 (SPSS Inc., Chicago, IL). Due to the parametric distribution of the data, differences between evaluations of diabetic patients and healthy subjects were compared using Student´s t-test. The linear correlation between endocrine (age at the time of diagnosis, duration of the disease, HbA1C, total, HDL and LDL cholesterol levels, systolic and diastolic arterial blood pressure and BMI) and functional parameters was determined using Pearson’s correlation coefficient. A binary logistic regression using Forward Wald method was performed to evaluate the predictor ability of all visual function test parameters to distinguish between DM eyes and control subjects.
Additionally, a sub-classification of subjects was performed, dividing DM2 patients into 2 different groups depending on the presence/absence of some particular characteristics: disease duration ≥ or < to 10 years, presence/absence of micro-macrovascular complications and presence/absence of HbA1C levels >7 mmol/ml. Values of p less than 0.05 were considered to indicate statistical significance. To avoid a high false positive rate, the Bonferroni correction for multiple tests was calculated and the corrected p-values were added to the previously calculated data (see all tables).