Introduction
Diabetic macular edema (DME) is a serious complication of diabetes and is currently
one of the leading causes of visual impairment and blindness. Cases of diabetes
complications such as DME and diabetic retinopathy (DR) are rising [1], and so far, restoring cell loss of the
retina and reverse visual impairment has not been possible [2]. Therefore, new strategies to detect
early retinal changes are urgently needed to prevent the development of DME and
DR.
Physiologically, retinal substrate supply and oxidative stress protection are
dependent on regular choroidal function [3]. The choroid represents the only source of metabolic exchange for the
avascular fovea centralis due to a high blood flow. Reduced choroidal blood
circulation can result in retinal dysfunction resulting in vision loss [4]. Alterations in choroidal vascularity
caused by tight junction disassembly and endothelial cell-mediated leukostasis are
considered major mechanisms for retinal edema and ischemia [5]. Furthermore, DME is caused by leakage
of fluid from the choroidal vessels. The fluid accumulates within the neurosensory
retina resulting in an increased thickness of the central retina [6]. The choroid also maintains the highly
metabolically active photoreceptor cells, and hypoperfusion leads to outer retina
dysfunction [5]. Precise and
three-dimensional retinal and choroidal morphology can be assessed by optical
coherence tomography (OCT) with a high resolution of 7 µm optical
and 3.5 µm digital axis [7]. To diagnose DME, measuring choroidal thickness (ChT) by OCT
represents a well-established method in clinical practice. Measurements by OCT have
been evaluated to be more precise than conventional diagnostic methods such as
stereoscopic slit lamps fundus examination [8]. However, OCT is only used in clinical practice to assess DME when DR
is already diagnosed or to follow up the treatment of DR. It is known that ChT
increases in progressive DR and decreases after treatment of DR. In addition, ChT
is
thicker in patients with DME [9]. While
patients with type 1 diabetes are at a higher risk of developing DR, patients with
type 2 diabetes (T2D), especially when treated with insulin, tend to develop DME.
However, the development of DME is due to changes in the choroid and its blood flow
[10]. Thus, we hypothesize ChT as a
parameter to detect early metabolic complications in posterior structures of the
eye. Therefore, this pilot study aimed to assess fasting state ChT in healthy
normal-weight controls and compare it to measures in healthy obese people as well
as
in patients with T2D.
Methods
Subjects
Three respective cohorts of participants (healthy normal-weight, healthy obese
with BMI≥30 kg/m2, and obese with T2D) were
enrolled in this mono-center study at the Metabolic Core Unit of the Centre of
Brain, Behavior and Metabolism, University of Luebeck, Germany. T2D was
diagnosed according to the American Diabetes Association criteria [11]. Healthy normal-weight subjects had
no history of chronic illness and long-term medication. All patients with T2D
were treated with insulin and oral antidiabetics. Exclusion criteria for all
groups were substance abuse, any ophthalmological diagnosis or previous
treatment, myopia or hyperopia with a spherical equivalent of<-7.0 dpt.
or>+7.0 dpt., wearing contact lenses in the past 12 h, a
glomerular filtration rate<60 mL/min./1.73, any
type of insulin-deficient diabetes, and a blood
pressure>140/90 mmHg. Further exclusion criteria for
healthy normal-weight and healthy obese participants were medication and acute
or chronic illness of any kind. The study protocol was approved by the ethics
committee of the University of Luebeck according to the declaration of Helsinki.
All participants gave written informed consent prior to study enrolment.
Study Design
Measurements were taken at 8 am in a fasting state. Blood pressure was measured
right before ChT measurements. ChT was defined as the subfoveal choroid measured
by OCT (Heidelberg Engineering GmbH, Heidelberg, Germany) at the center of the
foveola according to standard criteria: ChT was measured as the distance between
the hyperreflective outer border on the retinal pigment epithelium and the
interface line behind the large vessel layers of the choroid. The outer border
of the retinal pigment epithelial layer was automatically detected by the
instrument, and the interface line behind the vessels was defined manually [12] ([Fig. 1]). Simultaneously, blood
samples were drawn for duplicate measurement of fasting blood glucose
(EKF-Diagnostics Biosen C-line, Barleben, Germany), insulin and insulin-like
growth factor-1 (IGF-1) were assessed by immunoassays (Immulite 2000, Siemens
Healthcare Diagnostics, UK). The model assessment-estimated insulin resistance
(HOMA-IR) was calculated according to standard equation (insulin
(µU/mL) x fasting blood glucose (mg/dL)/405)
[13].
Fig. 1
a) Image of the retina. The horizontal line shows the area where
the measurement was taken by optical coherence tomography (OCT).
b) Example image of the choroidal thickness measurements by
OCT (Heidelberg Engineering GmbH, Heidelberg, Germany) at the center of
the foveola. The green line shows the subfoveal area. Red lines
highlight the measured choroidal thickness of
288 µm.
Statistical Analyses
One-way ANOVA calculation was performed to compare ChT between cohorts.
Pearson’s correlation coefficients were calculated to determine
bivariate relationships between ChT and BMI, HOMA-IR, and IGF-1, respectively.
Stepwise multiple linear regression was performed to test independent
confounding variables (BMI, HOMA-IR, age, sex, and IGF-1) of variance of ChT.
All data are presented as mean±SEM. Analyses were performed with SPSS
22.0 for Mac (SPSS Inc, Chicago, II). P-values of<0.05 were accepted as
being statistically significant. [Fig.
2] was created using GraphPad Prism 7 for Mac (San Diego, CA).
Fig. 2 Choroidal thickness (ChT) is significantly higher in people
with obesity compared to healthy participants, but not compared to
people with type 2 diabetes (T2D). ChT is also elevated in people with
T2D compared to healthy participants
(***p<0.0001).
Results
Study Group
In total, data of 53 participants were included for the final analysis: 17
healthy (aged 27.6±1.7 years, BMI
24.4±0.6 kg/m2), 20 healthy obese (aged
37.8±2.5 years, BMI 34.9±1.0 kg/m2), and 16
obese subjects with T2D (aged 56.4±3.0 years, BMI
34.7±1.1 kg/m2), ([Table 1]).
Table 1 Clinical characteristics of healthy participants,
people with obesity, and those with T2D.
|
Participants
|
Metabolic Healthy
|
Obesity
|
T2D
|
|
Age (years)
|
27.6 (±1.7)
|
37.8 (±2.5)
|
56.4 (±3.0)
|
|
BMI (kg/m2)
|
24.4 (±0.6)
|
34.9 (±1.0)
|
34.7 (±1.1)
|
|
Fasting blood glucose (mg/dL)
|
77.3 (±1.6)
|
87.7 (±2.0)
|
154.2 (±15.9)
|
|
Insulin (µU/mL)
|
4.8 (±0.8)
|
13.5 (±2.2)
|
37.5 (±13.3)
|
|
HOMA-Index
|
1.1 (±0.2)
|
3.0 (±0.5)
|
12.8 (±3.8)
|
|
IGF-1 (ng/mL)
|
158.9 (±12.2)
|
122.7 (±10.0)
|
122.7 (±10.0)
|
|
ChT (µm)
|
220.6 (±12.6)
|
309.4 (±9.7)
|
309.4 (±9.7)
|
ChT in the obese healthy (309.40±9.66 µm) as well as
obese with T2D cohort (328.82±15.65 µm) was
significantly thicker as compared to physiological ChT in the healthy
normal-weight cohort (220.60±12.60 µm),
(P<0.0001; [Fig. 2]).
However, there was no difference in ChT between healthy obese and obese with T2D
subjects (P=0.826). Overall, ChT correlated positively with BMI
(r=0.625, P<0.0001), HOMA-IR (r=0.274, P<0.045),
and negatively with IGF-1 (r=-0.268, P=0.050). The cohort with
T2D showed no correlation between ChT and diabetes duration (r=-0.058,
P=0.851). Stepwise multiple linear regression revealed that BMI
(R2=0.209; P=0.002; beta=0.388) and
HOMA-IR (R2=0.074; P=0.015; beta=0.305) are
independent variables of ChT explaining 20.9 and 7.4% of its variance
(both p<0.016), whereas age, sex, and IGF-1 were not significant
confounders of ChT (p>0.975).
Discussion
These data reveal the first evidence that increased metabolic burden, i. e.,
obesity and insulin resistance, is associated with thicker ChT as measured by OCT.
Choroid morphology has been previously proposed as possibly an important determinant
for diabetes-associated changes in posterior eye structures, and ChT has been found
to be thicker in patients with diabetes [3]
[14]
[15]. Furthermore, data suggest that in
patients with DR, both ChT and the sclera are thicker than in healthy eyes [15]. Another study reported increased ChT
after a hyperglycemic episode in hospitalized patients with diabetes [16]. In line, Ferreira et al. described
that ChT increases with higher glycemic levels (>160 mg/dL)
in patients with diabetes [17]. This is of
special clinical interest because the choroid is the central structure for supplying
the retinal pigment epithelium and the photoreceptors of the retina with nutrient
substrates and oxygen [3]. However, T2D is
a chronic disease derived from increasing insulin resistance, showing a huge overlap
with obesity as another main criterion of the metabolic syndrome. Until now, there
was no data on ChT in people with obesity but without manifested T2D. Here we show
that higher BMI, as a measure of obesity, is independently associated with thicker
ChT. To further account for the continuity in the development of metabolic
conditions, such as diabetes and the broad overlap with obesity, we used HOMA-IR as
a clinically established measure of insulin resistance [18]
[19]
[20] for our analysis. Apart
from obesity, insulin resistance was positively associated with ChT, implying that
distinct changes in ChT morphology arise even during early impairment of glucose
metabolism and before the manifestation of diabetes.
Since this study reports associations between metabolic characteristics and OCT
measures, we can only speculate on underlying mechanistic pathways. Therefore, we
included IGF-1 as a proxy of somatotropic axis activity in an exploratory and
hypothesis-generating analysis. IGF-1 represents a well-known modulator of energy
metabolism and is key for the regulation of body composition [21]. In people with long-standing insulin
resistance or manifested T2D, IGF-1 levels are reported to be lower as compared to
healthy controls due to elevated IGF-1 binding proteins (IGF-1BP). Hence, low IGF-1
levels have been debated as a predictive factor for the development and progression
of diabetes [22]
[23]
[24]. In addition, elevated levels of IGF-1 binding protein within the
vitreous body were found in patients with DR, contributing to the progress of
proliferative DR and DME [25]. Fitting
into this current state of knowledge, our exploratory analysis of IGF-1 revealed a
significant negative correlation between IGF-1 and ChT. However, further research
is
needed to assess the IGF-1 pathway as a potential link between metabolic traits and
changes in eye physiology.
This study had some limitations that need to be addressed. Cohorts differed
significantly in age, which might have an impact on ophthalmological measures. In
a
previous study, age was found not to be associated with ChT in young people [3] and decreased with age in adults [26]. In line, regression analysis revealed
that age was not an independent variable of ChT in our study. Furthermore, this
pilot study reports only correlations between metabolic state and ChT. Future
– and in the best case scenario, interventional – studies are needed
to prove a causal link between metabolically modulated dynamics in ChT and the
development of DME and DR.
In summary, our findings reveal that increased metabolic burden is associated with
thicker ChT as measured by OCT. Currently, there are no diagnostic methods for
detecting the early development of metabolic eye complications, e. g., DR.
Our preliminary findings might add clinical implications for early eye screening in
patients with metabolic disease. However, further studies are needed to confirm our
findings and translate them into clinical routine.
Author Contributions
S.M.M., H.J.G., B.W., and S.M. designed the study. A.J.P., M.S., N.K., and S. M.
collected the data, and R.C., B.W., A.K., S.M.M., and S.M. analyzed and interpreted
the data. S.M. researched literature and wrote the manuscript. B.W., R.C., A.K.,
A.J.P., M.S., N.K., A.S.S., E.P., H.J.G., and S.M.M. reviewed the manuscript. S.M.M.
takes responsibility for the integrity of the data and the accuracy of the data
analysis.
