The impact of hypertension on retinal nerve fiber layer thickness and its association with carotid intima media thickness

Abstract

Objective. Our aim was to investigate retinal nerve fiber layer (RNFL) thickness in hypertensive patients using spectral-domain optical coherence tomography (SD-OCT) and to evaluate the relationship between RNFL thickness and carotid intima media thickness (CIMT). Methods. This study included 59 patients with hypertension (HT) (53.6 ± 10.7 years) and 54 age-matched healthy controls (51.0 ± 8.1 years). We evaluated anthropometric and metabolic parameters as well as RNFL and CIMT measurements in patients with hypertension and controls. Results. The average RNFL thickness was 86.60 ± 10.86 μm in hypertensive patients and 93.63 ± 7.30 μm in healthy controls (p < 0.001). Selective thinning of the RNFL was found in the superior and inferior quadrants. Mean CIMT values were higher in patients with HT (0.80 ± 0.15 mm) than the healthy subjects (0.71 ± 0.1 mm) (p < 0.001). The average, inferior and nasal RNFL thickness were negatively associated with diastolic blood pressure respectively (r = − 0.112, r = − 0.210, r = − 0.225). There was an inverse correlation between RNFL thickness in the average and superior retinal quadrant and CIMT (r = − 0.201, r = − 0.185). There were no correlations between RNFL thickness and age, body mass index, fasting plasma glucose, lipid parameters, high-sensitive C-reactive protein and microalbuminuria. Conclusion. RNFL thickness is reduced in hypertensive patients and may be associated with atherosclerosis.

Key Words::

• Carotid intima media thickness
• hypertension
• retinal fiber layer thickness

Introduction

Hypertension is an established risk factor of atherosclerosis, and associated with a higher risk of death and ischemic events (1). It is important to identify patients with hypertension at risk of developing cardiovascular events. Carotid intima media thickness (CIMT) is a validated parameter for detecting subclinical atherosclerosis (2,3). In numerous studies, CIMT was shown to predict the occurrence of both stroke and myocardial infarction, independently of traditional cardiovascular risk factors (4–7).

Hypertension also has a high association with an increased risk of retinal microvascular abnormalities, such as retinal arteriolar narrowing and retinopathy (8–10). Retinal arteriolar caliber was shown previously to be related inversely to blood pressure (BP) levels (11). In some studies, retinal microvascular disorders were found to be associated with cardiovascular disease and mortality (12–15). Recently, it was demonstrated that pulse-wave velocity of optic nerve head circulation was correlated with brachial–ankle pulse-wave velocity and CIMT (16).

Evaluation of the retinal nerve fiber layer (RNFL) has been a well-established clinical and investigational tool (17). The retinal nerve fibers occur in the innermost layer of the retina and are formed by ganglion cell axons, which react to impulses from the photoreceptors to the central vision by way of the optic nerve. Axon damage is characterized by thinning of the RNFL. RNFL thickness can be affected in numerous neurodegenerative diseases (18,19). Optical coherence tomography (OCT) can be used for this purpose via obtaining histological sections of the retina and RNFL thickness measurements (20).

Hypertension plays an important role in the development of ischemic disorders of retina and the other systemic risk factors are diabetes mellitus, ischemic cardiovascular disease, hyperlipidemia and atherosclerosis (21). As diabetes mellitus induces the ischemia of retina, it results in the thinning of the RNFL (22). Numerous studies have also demonstrated the association between thinning of the RNFL and diabetes mellitus (23–26). However, RNFL thicknesses have not yet been evaluated in systemic hypertension.

The aim of the current study was to evaluate the impact of systemic hypertension on RNFL thickness using spectral-domain OCT (SD-OCT). Our second aim was to investigate the relationship between RNFL thickness and CIMT.

Materials and Methods

Study population

Fifty-nine patients with essential hypertension (mean age: 53.62 ± 10.7 years) and age-matched 54 healthy subjects (mean age: 51.0 ± 8.18 years) were enrolled in the current study. All the participants gave a written consent. The study protocol was approved by the local ethical committee.

Arterial BP measurements were performed by the same calibrated sphygmomanometer after 5 min of rest in a sitting position; measurements were performed on both arms at 5-min intervals. The average of the two measurements obtained in each arm was taken as the subject's BP. Patients with BP values greater than 140/90 mmHg were considered hypertensive (3). Newly diagnosed hypertensive patients were excluded from the study. We included patients with essential hypertension who were receiving anti-hypertensive drugs at least for 1 year. The mean duration of hypertension was 6.7 years (1–30 years) − 40.6% of the patients were treated with angiotensin-converting enzyme inhibitors or angiotensin receptor blockers, 25.4% were receiving calcium channel blockers, 28.8% of the patients were using beta-blockers and 50.8% of them were treated with diuretics; 56.1% of the patients were treated with monotherapy, 23.7% of them were treated with two drugs and 20.2% of them were treated with three or more drugs. Almost half (45.6%) of the patients were poorly treated.

Subjects with glaucoma (diagnosis was based on cup-to-disc ratio > 50%, cup-to-disc ratio asymmetry between two eyes > 20%, corrected intraocular pressure > 21 mmHg and glaucomatous visual field defects), pseudoexfoliation syndrome, high myopia and hypermetropia (greater than 6.0 diopters), anomalous optic disc, age-related macular degeneration, macular edema (macular thickness measured by optic coherence tomography greater than 300 μm), peripheral vasospasm, family history of glaucoma, history of ocular trauma, optic neuropathy, use of corticosteroid and glaucoma medications, diabetes mellitus, chronic kidney disease and a previous ischemic event were excluded from the study.

Clinical assessment and biochemical measurements

Weight, height and waist circumference were measured. The body mass index (BMI) was calculated by dividing the body weight in kilograms by the square of the height in meters (kg/m²). Waist circumference (WC) was measured at the narrowest level between the costal margin and iliac crest.

Venous blood samples were obtained from all subjects following an 8–12-h overnight fast. The levels of fasting plasma glucose (FPG), total cholesterol, triglyceride (TG), high-density lipoprotein-cholesterol (HDL-cholesterol), low-density lipoprotein cholesterol (LDL-cholesterol), high-sensitive C-reactive protein (hs-CRP), urea, creatinine and microalbuminuria were measured.

The hexokinase method was used to measure glucose levels, and photometric method (Abbott Architect c16000 autoanalyzer) was used to measure the total cholesterol, TG, HDL- and LDL-cholesterol levels. The concentration of hsCRP was measured using immunotubidimetric method with Abbott Architect C16000 autoanalyzer (Abbott Diagnostic, USA). The cut-off for hsCRP was taken as < 0.5 mg/dl.

Urinary albumin excretion (as an indicator of nephropathy) was measured in an early morning urine sample as albumin-to-creatinine ratio using a Cobas Integra 800 (Roche Diagnostics, Mannheim, Germany). Micro- or macro-albuminuria was present if urinary albumin excretion in at least two of three consecutive urine samples, 2 months apart, was 30–299 mg/g creatinine or ≥ 300 mg/g creatinine, respectively (27).

Measurement of retinal fiber layer thickness

All patients and controls underwent ophthalmologic testing, including assessment of visual acuity and color vision, slit-lamp examination with intraocular pressure measurement, visual fields with automated perimetry (Octopus 900; Haag Streit, Koeniz, Switzerland) and funduscopy. OCT imaging was performed with the Cirrus HD SD-OCT (Carl Zeiss Meditec, Dublin, CA). The OCT imaging was performed with the participants’ pupils dilated to obtain higher signal quality. Scans that had signal strengths lower than 8 were not included in the study to obtain images with minimal measurement errors. Average temporal, nasal, inferior, superior quadrant peripapillary RNFL thicknesses were obtained from the OCT with optic disc 200 × 200 cube scan protocol along a circle with a diameter of 3.46 mm around the center of the optic disc. OCT imaging was performed by the same ophthalmologist blinded to the patients.

Measurement of carotid intima media thickness

Ultrasonography was performed on all participants using a high-resolution ultrasonography scanner (Aplio 500, Toshiba Medical Systems, Tokyo, Japan) with a PLT-805AT linear array transducer. Measurements were performed on the right and left carotid arteries. The patient was lying supine with the head directed away from the side of interest and the neck slightly extended. The transducer was manipulated so that the near and far walls of the common carotid artery (CCA) were parallel, and the lumen diameter was maximized in the longitudinal plane. The region 1 cm proximal to the carotid bifurcation was identified, and the CIMT of the far wall was evaluated as the distance between the lumen–intima interface and the media–adventitia interface. The CIMT was measured on a frozen frame of a suitable longitudinal image, with the image magnified to achieve a higher resolution of detail. The CIMT measurement was obtained from four contiguous sites at 1-mm intervals on each carotid artery, and the average of all eight measurements was used for analysis. All measurements were performed by the same cardiologist, who was blinded to patient data (28).

Statistical analysis

Data were analyzed using SPSS Software (Version 17, SPSS, Inc., Chicago, IL, USA). Results were expressed as mean± standard deviation. The Mann–Whitney U test was used to compare the continuous variables and the chi-square test was used to compare categorical variables. Spearman's rank correlation test was used for calculation of associations between variables. Multiple regression analyses were performed to identify the independent determinants for RNFL thickness. A p-value of less than 0.05 was considered statistically significant.

Results

Clinical characteristics of the subjects

The demographic and clinical characteristics of the study population are shown in Table I. There were no significant differences in age, sex, BMI and waist circumference between patients with hypertension and healthy controls. Serum levels of FPG were higher in patients with hypertension (100.13 ± 9.44 mg/dl) compared with the healthy subjects (95.71 ± 7.09 mg/dl) (p = 0.006). Among the lipid parameters, total-cholesterol, LDL-cholesterol and TG levels were significantly higher in the hypertensive group in comparison with the controls, respectively (p = 0.002, p = 0.018, p = 0.028). Although serum urea and creatinine levels were in normal range in all patients with hypertension, values of urea were higher compared with those in healthy subjects (p = 0.039). Serum levels of uric acid and hsCRP were also higher in the patient group in comparison with the control group (Table I). Mean CIMT values were higher in patients with hypertension (0.80 ± 0.15 mm) than the healthy subjects (0.71 ± 0.10 mm) (p = 0.001).

Table I. Comparison of the clinical and biochemical parameters in patients with hypertension and control group.

	Patients with hypertension (n = 59)	Healthy subjects (n = 54)	p
Age (years)	53.6 ± 10.8	51.0 ± 8.2	0.146
Gender (% female)	52.1	47.9	0.514
BMI (kg/m^2)	32.8 ± 5.8	32.1 ± 6.6	0.562
WC (cm)	101.5 ± 10.5	103.9 ± 15.1	0.318
SBP (mmHg)	146.7 ± 19.9	127.5 ± 13.1	< 0.0001
DBP (mmHg)	87.5 ± 10.9	76.4 ± 8.1	< 0.0001
FPG (mg/dl)	100.1 ± 9.4	95.7 ± 7.1	0.006
Total cholesterol (mg/dl)	247.1 ± 33.5	225.5 ± 40.9	0.002
LDL-cholesterol (mg/dl)	162.8 ± 33.9	145.6 ± 38.1	0.018
HDL-cholesterol (mg/dl)	48.2 ± 10.5	51.5 ± 19.7	0.264
Triglyceride (mg/dl)	167.5 ± 79.7	136.8 ± 65.0	0.028
Urea (mg/dl)	30.7 ± 10.33	27.1 ± 7.712	0.039
Creatinine (mg/dl)	0.8 ± 0.2	0.8 ± 0.1	0.247
Uric acid (mg/dl)	6.9 ± 8.1	4.6 ± 1.3	0.044
Microalbuminuria (mg/g)	94.5 ± 266.6	13.7 ± 96.7	0.029
hsCRP (mg/dl)	0.6 ± 0.9	0.5 ± 0.7	0.470
CIMT (mm)	0.8 ± 0.1	0.71 ± 0.1	0.001

Values are expressed as means± SD. BMI, body mass index; WC, waist circumference; SBP, systolic blood pressure; DBP, diastolic blood pressure; FPG, fasting plasma glucose; LDL, low-density lipoprotein; HDL, high-density lipoprotein; hs-CRP, high-sensitive C-reactive protein; CIMT, carotid intima media thickness.

Comparison of RNFL thickness between the groups

According to quadrants, the mean RNFL thickness of the two groups is shown in Table II. The average RNFL thickness was 86.64 ± 10.86 μm in hypertensive patients and 93.63 ± 7.30 μm in healthy controls (p < 0.001). Selective thinning of the RNFL was found in the superior and inferior quadrants (Table II). Mean superior RNFL thickness was 106.98 ± 17.51 μm in patients with hypertension and 118.91 ± 14.74 μm in controls (p < 0.001). Mean inferior RNFL thickness values were lower in hypertensive patients (109.2 ± 18.69 μm) compared with the control group (118.59 ± 13.81 μm) (p = 0.003). The RNFL thickness did not differ significantly between patients with hypertension and controls in temporal and nasal quadrants respectively (p = 0.401, p = 0.150).

Table II. Retinal nerve fiber layer thickness in patients with hypertension and healthy controls.

Location	RNFL thickness		p-value
	HT	Controls
Average	86.6 ± 10.7	93.6 ± 7.3	< 0.001
Superior	106.9 ± 17.5	118.9 ± 14.7	< 0.001
Temporal	61.7 ± 14.3	63.7 ± 11.1	0.401
Inferior	109.0 ± 18.7	118.6 ± 13.8	0.003
Nasal	68.7 ± 10.9	71.9 ± 12.5	0.150

RNFL, retinal nerve fiber layer; HT, hypertension.

The correlations between RNFL thickness, CIMT and cardiometabolic parameters

The average (r = − 0.112, p = 0.038), inferior (r = − 0.210, p = 0.026) and nasal (r = − 0.225, p = 0.016) RNFL thickness were negatively associated with diastolic BP. There was also an inverse correlation between RNFL thickness in the average and superior retinal quadrant and CIMT (r = − 0.201, r = − 0.185) (Table III). The negative correlation between average RNFL thickness and CIMT is shown in Figure 1. There were no correlations between RNFL thickness and age, BMI, FPG, lipid parameters, hsCRP and microalbuminuria.

Figure 1. The correlation between average retina nerve fiber layer (RNFL) thickness and carotid intima media thickness (CIMT). The patient who had a CIMT of 1.5 mm was excluded from the calculation.

Table III. The correlation between retinal nerve fiber layer thickness, carotid intima media thickness (CIMT) and cardiometabolic parameters in the study population.

Variable	Average		Superior		Temporal		Inferior		Nasal
	r	p	r	p	r	p	r	p	r	p
Age	− 0.173	0.067	− 0.113	0.235	− 0.172	0.069	− 0.176	0.062	− 0.030	0.755
BMI	0.116	0.223	0.123	0.195	0.029	0.757	0.070	0.464	0.119	0.241
SBP	− 0.127	0.181	− 0.180	0.056	0.008	0.935	− 0.003	0.975	− 0.177	0.061
DBP	− 0.112	0.038	− 0.165	0.082	0.093	0.327	− 0.210	0.026	− 0.225	0.016
FPG	− 0.033	0.733	− 0.054	0.571	0.070	0.460	0.029	0.758	0.120	0.206
Total cholesterol	− 0.066	0.487	0.072	0451	0.073	0.441	− 0.177	0.061	− 0.145	0.124
Triglyceride	0.059	0.536	0.123	0.191	0.038	0.696	0.015	0.873	− 0.049	0.610
LDL-cholesterol	− 0.039	0.680	0.066	0.489	0.144	0.127	− 0.136	0.151	− 0.176	0.062
HDL-cholesterol	0.011	0.680	− 0.023	0.806	− 0.083	0.380	0.009	0.925	0.129	0.173
hsCRP	− 0.010	0.984	0.059	0.536	− 0.094	0.327	− 0.007	0.942	− 0.014	0.880
Uric acid	0.038	0.704	− 0.079	0.427	0.220	0.026	− 0.003	0.973	0.029	0.771
MA	0.002	0.982	− 0.015	0.874	0.169	0.075	− 0.006	0.948	− 0.129	0.175
CIMT	− 0.201	0.034	− 0.185	0.049	− 0.104	0.277	− 0.063	0.508	− 0.156	0.100

r indicates Spearman's rho correlation coefficient; correlation is significant at the 0.05 level. BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; FPG, fasting plasma glucose; LDL, low-density lipoprotein; HDL, high-density lipoprotein; hs-CRP, high-sensitive C-reactive protein; MA, microalbuminuria.

Multiple linear regression analyses

We performed a multiple linear regression analyses including RNFL thickness as dependent variable and age, uric acid, systolic BP, diastolic BP, LDL-cholesterol, TG, hsCRP and CIMT as explanatory variables. In this analysis, we did not find any significant and independent predictor for RNFL thickness.

Discussion

Our study showed that RNFL of hypertensive patients was thinner than healthy controls and this difference was most prominent in the superior and inferior quadrants. Furthermore, average and superior RNFL thickness was inversely associated with CIMT.

Blood supply of retina originates from choroidal arteries, posterior ciliary arteries and in part from pial arteries networks. Thinning of the RNFL was demonstrated in patients with migraine (29). Associated with neurovascular pathogenesis of migraine, vasoconstriction in particular during the migraine attacks affects retrobulbar artery. Consequently, due to this effect, retinal hypoperfusion and retinal anomalies related to ganglionic retinal cell death occur.

Arterial hypertension can influence the retina blood flow in a number of ways. Increased vascular resistance in the terminal arterioles can derange retina blood supply. Secondary hypertensive vascular changes also occur in retina vessels, and finally blood flow autoregulation is disturbed (21). As a result, these changes make the optic nerve prone to perfusion deficiency in hypertension and retinal ischemia can develop. In the current study, we hypothesized whether RNFL thickness may be reduced in patients with systemic hypertension.

RNFL thinning has been described in patients with type 2 diabetes mellitus in a number of studies (23–26). However, it has not yet been evaluated in hypertension. Systemic hypertension is a systemic condition and affects the optic nerve vascular circulation. In the present study, we showed that RNFL of hypertensive patients was thinner than healthy controls. The average, inferior and nasal RNFL thickness were negatively associated with diastolic BP. Thinning of the RNFL was found to be more prominent in the superior quadrant of retina in diabetic patients in most of the studies (24,30). In our study, the superior and inferior quadrants of RNFL were thinner in patients with hypertension. We suggest that particularly superior and inferior quadrants of retinal nerve are affected in systemic hypertension.

Atherosclerotic changes in the large arteries may play a role in the development of ischemic disorders of the retina (31). Inflammation and endothelial dysfunction play a role in the pathogenesis of both large atherosclerosis and retinal microvascular diseases (32). Retinal vascular abnormalities were found to be related to various markers of atherosclerosis in a number of studies (33–36). Carotid intima-media thickness is a surrogate marker of atherosclerosis. Lim et al. demonstrated the upper limit of carotid IMT in healthy population (37). In our study, the control group had similar values of CIMT to the results from the study of Lim et al. We also found higher CIMT values in patients with hypertension than the control group.

In the present study, we hypothesized that carotid atherosclerosis may also be associated with thinning of RNFL. Recently, we have shown a negative association between RNFL thinning and carotid atherosclerosis in patients with type 2 diabetes mellitus (38). We also found that there was a negative correlation between RNFL thickness and CIMT in the current study. Hayreh & Jonas showed reduced RNFL of rhesus monkeys in chronic experimental arterial hypertension and atherosclerosis (39). Our data provide support for the hypothesis that systemic atherosclerosis may be associated with thinning of RNFL in patients with systemic hypertension. Additionally, further research is needed to understand the link between atherosclerosis and RNFL thinning.

Retinal examination is recommended as part of the routine assessment of people with hypertension and its value has been questioned, partly because clinical ophthalmoscopy is highly unreliable for detecting retinal microvascular lesions (40). SD-OCT can provide B-scan images of the retina with a resolution of 8–10 μm. Thus, the thickness of RNFL can be measured reliably and accurately by the installed software.

In conclusion, our data suggest that retinal fiber layer thickness is reduced in hypertensive patients and may be associated with atherosclerosis.

Acknowledgments

The authors declare no competing interests. No financial support was received for this paper.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.