Background

Since 1980, the adult population living with diabetes has increased four-fold to approximately 422 million according to the most recent World Health Organization’s Global Report on Diabetes. This sharp rise can be attributed to overweight and obesity, which have resulted in an increase in type 2 diabetes [1]. The prevalence of diabetes in Turkey has recently been reported as 13.2% [2].

The most common reason for vision loss in diabetic patients is diabetic macular edema (DME). Unfortunately, the absolute prevalence of DME may be increasing due to the overall increase in the prevalence of diabetes in industrialized nations [3]. Population-based studies have reported the prevalence of DME in type 1 diabetic patients as 4.2–7.9%, while the rate for type 2 diabetes patients ranges from 1.4–12.8% [4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27]. In a Cochrane review of the DME prevalence evaluated using optical coherence tomography (OCT), the prevalence rates covered a wide range (19%–65%) [28].

In recent years, the use of OCT has become more widespread for the objective measurement of retinal thickness and the other elements of macular edema [29,30,31]. The Diabetic Retinopathy Clinical Research network (DRCR.net) has adopted standard OCT DME assessments in multicenter studies of diabetic retinopathy (DR). Since this assessment is quantitative with the use of OCT, rather than qualitative when applying photography or biomicroscopy, this is considered to be a significant advantage.

The epidemiology and disease burden have not yet been fully elucidated, and there is limited information on the current state of DME in Turkey. Therefore, the aim of this study was to evaluate the prevalence, demographic characteristics of the patients, and systemic associations of DME utilizing OCT in Izmir, Turkey.

Methods

This cross-sectional study was conducted in the departments of ophthalmology and endocrinology at the Dokuz Eylul University School of Medicine in Izmir. A total of 413 eyes of 413 diabetic patients who were followed up in the clinics between January 2011 and July 2012 were enrolled. The demographic data, diabetes type, diabetic age, treatment modality, smoking and alcohol consumption habits, as well as the systemic blood pressure, renal functional test results, hemoglobulin A1c (HbA1c) level, serum lipid profile, 24-h urine albumin level, and the existence of neuropathy were noted and statistically analyzed. The ophthalmological evaluation of each participant included the best corrected visual acuity (BCVA), slit-lamp biomicroscopy, intraocular pressure (IOP) measurement, and dilated fundoscopy. Fluorescein angiography and a central macular thickness (CMT) analysis with OCT were also performed. The relationships between the systemic findings and the prevalence of DME were studied.

Those patients ≥18 years old with type 1 or 2 diabetes diagnosed by an endocrinologist at the Dokuz Eylul University Hospital Endocrinology Clinic between January 2011 and July 2012, who were then referred to the Ophthalmology Department Retina Unit for DME and DR screenings, were included in this study.

The exclusion criteria were as follows: eyes with an ocular abnormality other than DME (vitreomacular traction, epiretinal membrane, etc.) and media opacities interfering with the reliability of OCT imaging (dense cataract, uveitis, etc.), and those patients with insufficient data for the study protocol.

Ophthalmological examination

The BCVA was evaluated using the Bailey-Lovie chart after correcting for refractive errors. An anterior segment examination was conducted using slit-lamp biomicroscopy and dilated fundoscopy. The IOP was obtained with a Goldmann applanation tonometer, and Heidelberg retinal angiography (HRA) and OCT were performed using the Spectralis HRA-OCT II (Heidelberg, Germany). After obtaining a fixation point for the patient, 6 OCT shots were lined up with the radial line scan and each other at an angle of 30°. The eyes were evaluated for clinically significant macular edema (CSME) as defined by the Early Treatment Diabetic Retinopathy Study (ETDRS) and with a central macular thickness (CMT) (mean thickness at the point of the intersection of 6 radial scans) via OCT ≥ 250 μm attributable to DME [32].

Statistical analysis

The data from all of the subjects who fulfilled the inclusion/exclusion criteria were analyzed using SPSS 16.0 software. For the descriptive analysis, the mean, standard deviation, and percentage were used. The chi-squared test, Fisher’s exact test, and t-test were applied for the univariate analysis. A p value < 0.05 was considered to be statistically significant.

Results

Of the 425 patients who met the study criteria, 413 were included for evaluation. DME was detected in 15.3% (63) of the patients and DR was determined in 32% (132) of the patients. Moreover, DME was found in 14.8% (4) of the patients with type 1 diabetes and in 15.3% (59) of the patients with type 2 diabetes (p = 0.604). Of the 63 DME patients, 15 received previous focal/grid laser treatments, 8 received previous intravitreal anti-vascular endothelial growth factor (VEGF) or steroid treatments, and 5 received previous combined focal/grid laser and anti-VEGF/steroid treatments. In addition, 9 patients without DME received previous focal/grid treatments and one patient underwent a vitrectomy.

The demographic and laboratory characteristics of the patients are summarized in Table 1. DME was significantly more prevalent in the males than the females (p = 0.031), and the male subjects had higher HbA1c levels than the female subjects (8.30 ± 2.25% and 7.89 ± 2.13%, respectively) (p = 0.054). Although there was no direct statistical correlation between the HbA1c levels and DME, a significant increase in the frequency of DME was observed particularly in those subjects with HbA1c values of 7.0% or more (p = 0.037). While the type of diabetes did not have an effect on DME, the duration of diabetes was significantly longer in the DME patients, particularly in those diagnosed between 10 and 20 years previously (p < 0.001). Those patients without DME were determined to have a significantly higher rate of antihyperlipidemic drug usage and a higher level of high density lipoprotein cholesterol (HDL-C) (p = 0.040 and p = 0.046, respectively). The mean serum creatinine levels in those patients with and without DME were 1.13 ± 0.81 mg/dL and 0.87 ± 0.63 mg/dL, respectively, and this difference was statistically significant (p = 0.021).

Table 1 Comparison of the demographic and laboratory characteristics of the patients with and without DME
Full size table

In the comparison of the normoalbuminuric, microalbuminuric, and macroalbuminuric patients in terms of the DME frequency, a statistically significant difference was seen between the 3 groups (p < 0.001). While 11.0% of the patients without nephropathy had DME, 29.0% of patients with microalbuminuria and 26.7% of the patients with macroalbuminuria had DME (p < 0.001). Peripheral neuropathy was also significantly frequent in those patients with DME (p = 0.006). The mean BCVAs of the eyes with and without DME were 0.55 ± 0.59 logMAR and 0.04 ± 0.10 logMAR, respectively (p < 0.001). The mean IOPs of the eyes with and without DME were 14.91 ± 2.45 mmHg and 15.12 ± 2.64 mmHg, respectively, and no statistical difference was seen (p = 0.562).

The prevalence of DME was 28.6% in those patients with mild to moderate non-proliferative diabetic retinopathy (NPDR) and 72.6% in those patients with severe NPDR to proliferative diabetic retinopathy (PDR) (p < 0.001). The DME prevalences in the phakic and pseudophakic eyes were 12.9% (49) and 43.7% (14), respectively (p < 0.001). Assuming the possible effects of cataract surgery on DME and evaluating only the phakic patients showed that the duration of diabetes, nephropathy, neuropathy, and antihyperlipidemic drug use significantly affected the DME in similar ways (p < 0.001, p = 0.020, p = 0.012, and p = 0.038, respectively). However, in the phakic patients, the gender, creatinine level, and HDL-C level did not have statistically significant effects on the DME (p = 0.610, p = 0.227, and p = 0.233, respectively).

Discussion

There is a known increasing worldwide prevalence of DME. Correspondingly, an increase in diabetes-related complications is expected with the increase in diabetes mellitus cases in Turkey. In one study from Turkey, the prevalence of DME was found to be 14.2% in the pre-OCT era [33]. Most studies have used non-stereoscopic fundus photography; therefore, the accuracy of the DME assessment is in doubt. The use of stereoscopic slit-lamp biomicroscopy alone may also lead to both the underdiagnosis and overdiagnosis of DME. Macular edema was defined using the CSME criteria in approximately one-half of the previous studies, and thus, only covered the more severe DME spectrum. The clinical use of OCT has enabled the detection of DME that was previously overlooked in a stereoscopic fundus examination. When compared to a clinical examination, the OCT detection and assessment of DME is more objective and reproducible, ensuring greater uniformity in the interventions applied and the treatment outcomes when compared to the pre-OCT era [34, 35]. According to DRCR.net, for DME trial inclusion and retreatment eligibility, the central subfield mean thickness on a Stratus OCT must be ≥250 μm. The current study used the Spectralis HRA-OCT II, which produces high resolution histological macular images, and the prevalence of DME was found to be 15.3%. This ratio was higher than the prevalence in a previous study conducted in 2006, and thus supports the sensitivity of the OCT.

The DME prevalence is related to the disease duration. In the present study, the prevalence of DME was 2.8% within 5 years of the diabetes diagnosis and 22.0% 5 years after the diagnosis (p < 0.001). After 10 years, the prevalence rose prominently. In a study by Aiello et al. [36], the prevalence was 5% within the first 5 years after the diagnosis and 15% at 15 years.

The males in this study exhibited DME more frequently than the females, and the odds ratio (OR) for the males was 1.811 (95% CI: 1.051 < OR < 3.121) (p = 0.031). In addition, the HbA1c levels were significantly higher in the males than the females; therefore, and it can be suggested that not only gender, but also worse diabetic control in male patients can indicate a higher prevalence of DME. The HbA1c level in the patients with DME (8.39 ± 1.97%) was slightly higher than that in the patients without DME (8.02 ± 2.23%), but this difference was not statistically significant (p = 0.226). The prevalences of DME in those patients with HbA1c levels < 7.0% and ≥ 7.0% were 10.62% and 18.18%, respectively (p = 0.037). In the Diabetes Control and Complications Trial (DCCT), it was shown that the strict control of blood glucose in type 1 diabetes patients led to a 29% decrease in the cumulative incidence of macular edema at the 9-year follow-up, and halved the application of focal laser treatment for DME [37, 38]. Even if there is a deterioration in control later in life, the effects of improved glycemic control sustained over many years have been shown to persist. In the Epidemiology of Diabetes Interventions and Complications (EDIC) study, which was an extension of the DCCT in which the level of glycemic control of the former intensive and conventional control groups converged, it was reported that the former intensive control group continued to fare better than the former conventional control group. Four years after the end of the DCCT, the CSME incidence was 2% in the former intensive control group, compared to the 8% rate in the former conventional control group (p < 0.001) [39].

In the UK Prospective Diabetes Study, an analogous, randomized clinical trial of type 2 diabetes patients, it was reported that strict blood glucose control resulted in a 29% reduction in laser treatment in a follow-up period of 10 years; of the laser treatments required, 78% were for DME [40]. In the current study, the prevalence of DME was conspicuously higher in the insulin-taking patients (p < 0.000). In previous studies, taking insulin has been reported to trigger the development of DME in the acute period. In this period, the hypoxia-inducible factor connects to the VEGF promotor region, and the VEGF transcription increases. Subsequently, the blood-retina barrier breaks down and permeability increases with the activation of protein kinase C. In the chronic period, insulin shows anti-inflammatory and anti-apoptotic effects and reduces oxidative stress [41]. The high prevalence of DME in the insulin-taking patients in the present study may be the result of the poor glycemic control in these patients.

The UK Prospective Diabetes Study also reported that the mean systolic blood pressure was reduced by 10 mmHg and the diastolic blood pressure was reduced by 5 mmHg in a median follow-up period of 8.4 years, which resulted in a 35% decrease in the retinal laser treatments, 78% of which were for DME [42]. In addition, the Wisconsin Epidemiologic Study of Diabetic Retinopathy determined that systemic hypertension increased the prevalence of DME 3-fold. It has been suggested that not only is hypertension a risk factor for macular edema development, but the treatment may have important benefits in patients with uncontrolled hypertension [43]. In the current study, the prevalences of DME in those patients with and without systemic hypertension were 17.4% and 12.3%, respectively (p = 0.158). In addition, there was no statistically significant difference with respect to the systolic and diastolic blood pressure levels between those patients with and without DME. However, anti-hypertensive medications may affect these results. The beneficial effects of anti-hypertensive medications that target the renin-angiotensin-aldosterone system (RAAS) in DR and DME have been evaluated in several clinical trials, such as the Diabetic Retinopathy Candesartan Trials (DIRECT) and Renin-Angiotensin System Study (RASS).

A recent meta-analysis revealed that patients with DME or PDR were more likely to have incident cardiovascular disease (CVD) and fatal CVD when compared to those without DME or PDR in type 2 diabetes mellitus [63]. For instance, the Singapore Malay Eye Study found an association between ocular hypertension and diabetes, but not glaucoma [64].

Diabetes is associated with the early and rapid development of cataracts, and cataract surgery, other types of intraocular surgery, and ocular inflammatory disease may produce inflammatory and angiogenic mediators that can produce macular edema in eyes with or without DR [65,66,67,68,69]. In accordance with this, in the present study, the DME prevalences in the phakic and pseudophakic eyes were 12.9% (49) and 43.7% (14), respectively (p < 0.001).

Conclusions

In 2010, the prevalence of diabetes in Turkey was 13.7% as reported in the Turkish Diabetes Epidemiology II (TURDEP-II) study. In the USA, DR is the leading cause of blindness in individuals aged < 60 years old, and DME is the most common cause of visual loss in those with DR [56, 66]. Fortunately, permanent vision loss can be prevented by the early diagnosis and treatment of DME. The DME prevalence has been reported at a wide range of rates in numerous studies in the literature, but there have been no previous studies in Turkey on this topic. The development of DME may be avoided or limited and the response to treatment may be improved by the regulation of the DME risk factors. In this study, the prevalence of DME was associated with male gender, diabetes duration, HbA1c ≥ 7.0%, insulin usage, alcohol consumption, low HDL-C levels, nephropathy, neuropathy, severity of DR, and previous cataract surgery. However, antihyperlipidemic drugs may be protective against DME. The cross-sectional design could be considered a limitation of this study; therefore, longitudinal studies with more subjects are needed.