Patients and methods: Eighty apparently healthy subjects screened for cardio- ... Conclusion: In apparently healthy subjects with normal renal function, retinal ...... Iyengar SK (2006): Genome-wide link- ... Email: [email protected].
Acta Ophthalmologica 2013
Retinal vascular caliber is associated with renal function in apparently healthy subjects Vincent Daien,1,2 Ryo Kawasaki,3 Max Villain,1 Jean Ribstein,4 Guilhem Du Cailar,4 Albert Mimran4 and Pierre Fesler4 1
Department of Opthalmology, Hoˆpital Gui de Chauliac, Montpellier, France National Institute for Health and Medical Research (INSERM), Montpellier, France 3 Centre for Eye Research Australia, Royal Victorian Eye and Ear Hospital, University of Melbourne, Melbourne, Victoria, Australia 4 Department of Internal Medicine, Hoˆpital Lapeyronie, Montpellier, France 2
ABSTRACT. Purpose: To assess the relation between retinal vascular caliber and renal function. Patients and methods: Eighty apparently healthy subjects screened for cardiovascular risk factors (mean age 47 years, 51% female, 36% hypertensive, without diabetes or renal dysfunction) were recruited. Retinal vascular calibers were measured from fundus photographs and expressed as central retinal artery and venular equivalent. Renal function was assessed by measurement of glomerular filtration rate (urinary clearance of 99mTc-DTPA) and urinary albumin ⁄ creatinine ratio. Results: Mean glomerular filtration rate was 117 ml ⁄ min ⁄ 1.73m2. Overall, central retinal artery and venular equivalent were positively correlated with glomerular filtration rate (r = +0.31, p = 0.005 and r = +0.30, p = 0.006, respectively). In addition, central retinal artery equivalent was negatively correlated with urinary albumin ⁄ creatinine ratio (r = )0.34, p = 0.002). No significant relationship was found between central retinal venular equivalent and urinary albumin ⁄ creatinine ratio (r = +0.12, p = 0.32). The observed relations between retinal vascular calibers and renal function parameters remained significant after adjusting for potential confounding factors. Conclusion: In apparently healthy subjects with normal renal function, retinal arteriolar and venular calibers were negatively correlated with kidney function, suggesting common determinants of these preclinical target organ damages. Key words: glomerular filtration rate – hypertension – microcirculation – retinal vessels – target organ damages
Acta Ophthalmol. 2013: 91: e283–e288 ª 2013 The Authors Acta Ophthalmologica ª 2013 Acta Ophthalmologica Scandinavica Foundation. Published by Blackwell Publishing Ltd.
doi: 10.1111/aos.12094
Introduction In essential hypertension, several studies have suggested that microvascular
dysfunction of subcutaneous capillaries and arterioles may be predictive of cardiovascular events (Levy et al. 2008).
Fundus photography provides a noninvasive approach of the retinal microcirculation. Changes in the caliber of these retinal vessels reflect cumulative effects of the ageing process (Leung et al. 2003a), cardiovascular risk factors (Wong et al. 2003; Kawasaki et al. 2006), inflammation (Wong et al. 2006a), endothelial dysfunction (Delles et al. 2004), and genetic factors (Xing et al. 2006). In addition, a relationship between retinal vascular calibers and cardiovascular complications of hypertension such as stroke (Ikram et al. 2006) and coronary heart disease has been reported (Wong et al. 2002a,b; Wong et al. 2006b). Following a long period characterized by isolated renal hemodynamic abnormalities such as glomerular hyperfiltration (Mimran et al. 1996), long-standing hypertension is associated with structural alteration of intrarenal microvessels. Within a follow-up period of 8.6 years, a positive correlation between media-to-lumen ratio of subcutaneous resistance arteries and the evolution of estimated glomerular filtration rate (GFR) was observed in a sample of sixty normotensive and hypertensive subjects (Boari et al. 2010). During the last decade, association of the decrease in GFR with cardiovascular risk was observed in large population studies (Astor et al. 2008; Efstratiadis et al. 2008; Matsushita et al. 2010).
e283
Acta Ophthalmologica 2013
In patients with renal failure, retinal arteriolar and venular calibers were found to progressively decrease with increasing stage of chronic kidney disease (CKD) (Ooi et al. 2011). Whether a relationship between renal and retinal microcirculation exists in subjects without renal failure remains to be documented.
Patients and Methods Study population
Subjects were recruited from the outpatient clinic of the department of medicine, where individuals came directly or were referred by general practitioners for the detection or investigation of cardiovascular risk factors, through a local programme of screening for cardiovascular risk (Montpellier area, France). In this study, we only considered apparently healthy subjects, without known cardiovascular, renal or metabolic disease and who had never received antihypertensive, antidiabetic or lipid-lowering therapy. Patients with clinical evidence of atherosclerosis (either stroke or coronary and peripheral artery disease), heart failure, renal failure (serum creatinine >1.2 mg ⁄ dl), diabetes mellitus (fasting blood glucose >126 mg ⁄ dl), marked obesity (body mass index ‡35 kg ⁄ m2), a history of alcohol abuse (>5 drinks ⁄ day) and secondary hypertension were excluded. Written informed consent was obtained from each participant conforming to the provisions of the Declaration of Helsinki in 1995 (as revised in Edinburgh 2000). This study was approved by the University Hospital of Montpellier. Blood pressure and heart rate
Blood pressure (BP) and heart rate were measured every 3 min with an automatic device (Model Dinamap V100; GE Healthcare, Little Chalfont, UK) (Reinders et al. 2006) and reported values are the average of at least 10 measurements following a 10-min period of rest in the supine position. Mean blood pressure was calculated as systolic Blood Pressure + [2* diastolic Blood Pressure] ⁄ 3. Hypertension was defined as systolic BP ‡140 mmHg and ⁄ or diastolic BP ‡90 mmHg (International
e284
Society of Hypertension Guidelines 1999). Determination of renal function
On the day of investigation, patients came to the ward with 24-hr urine collection for the determination of creatinine, electrolytes and albumin (nephelometry). Prior to investigations, subjects were weighed (Model 703; CKSeca, Hambourg, Deutschland) and sized (Microtoise 04-116; Stanley Black & Decker, New Britain, CT, USA). As previously described, glomerular filtration rate was measured by urinary clearance of technetiumlabelled diethylene triaminopentaacetic acid (99mTc-DTPA) using the continuous infusion technique. Three to four 30-min urine collections were obtained by spontaneous voiding (Fesler et al. 2004). Fasting blood samples were obtained prior to clearance determinations for the measurement of serum creatinine (enzymatic method based on the conversion of creatinine to ammonia and N-methylhydantoine by creatinine deiminase), glucose, electrolytes, total cholesterol, triglycerides and lipid levels. In a previous study conducted in 20 subjects, the reproducibility of day-to-day measurements of GFR, expressed as coefficient of variation, was 6.4% (Mimran & Deschodt 1983). Retinal vascular calibers measurements
Retinal photographs of both eyes were obtained, using a nonstereoscopic, 45 nonmydriatic fundus camera (CR5-NM45; Canon Inc., Tokyo, Japan; and TRC, Topcon Inc., Tokyo, Japan), centred on the optical disc. Diameters of all vessels coursing through a specified area (0.5–1 disc diameter surrounding the optic disc) were measured using image analysis software (IVAN, Department of Ophthalmology Visual Science, University of Wisconsin, Madison, WI). The calibers of the central retinal artery and vein were estimated using the ‘Big-6 formula’ and summarized as the central retinal artery and vein equivalents (CRAE and CRVE) (Sherry et al. 2002; Knudtson et al. 2003). The CRAE was divided by the CRVE to obtain the arteriole-to-venule ratio (AVR). The reproducibility of retinal vascular measurements was high, with
intragrader correlation coefficients of 0.97 [95% confidence interval (CI): 0.96–0.98] for CRAE and 0.95 (95% CI: 0.94–0.96) for CRVE. Due to the high inter-eye correlation reported in previous studies (Leung et al. 2003a,b), we analysed only one eye per subject (the right eye, or, if unavailable or ungradable, the left eye). Statistical analysis
Statistical analysis software (SAS version 9.2; SAS Institute, Cary, NC, USA) was used for data analysis. Albuminuria was calculated as the urine albumin-to-creatinine ratio (UACR) to account for the possibility of inadequate urine collection. Due to a skewed distribution, UACR values were log-transformed prior to statistical analysis. Adjusted means comparing retinal vascular caliber of normotensive and hypertensive subjects were obtained using multiple covariance analysis. The relationship between retinal parameters with renal parameters, age, gender, mean BP, smoking status, fasting blood glucose, body mass index, total cholesterol, triglycerides, was first tested by linear univariate analysis. Multivariate linear regression was used to detect the independent relationship between retinal and renal parameters. Significance levels were set at p < 0.05.
Results Population characteristics
The study population consisted of 80 subjects (41 women and 39 men, aged 22–77 years), of whom 35% (n = 29) were hypertensive. The characteristics of the study group are summarized in Table 1. The mean value of measured GFR was 117 ± 20 ml ⁄ min ⁄ 1.73m2, with no significant difference between normotensive and hypertensive subjects. Comparison of retinal vascular caliber between normotensive and hypertensive subjects
As summarized in Table 2, mean CRAE and AVR (adjusted for age, gender, body mass index, smoking status, total cholesterol, triglycerides, fasting blood glucose and GFR) were
Acta Ophthalmologica 2013
significantly lower in hypertensive compared with normotensive subjects (135 ± 24 versus 142 ± 18 lm for CRAE and 0.65 ± 0.12 versus 0.70 ± 0.09 for AVR, respectively). No significant difference in adjusted mean CRVE was detected. Correlations between retinal vascular calibers, age and blood pressure
As shown in Fig. 1, CRAE and CRVE were negatively correlated with age (r = )0.47; p < 0.001 and r = )0.36; p < 0.001, respectively). CRAE, but not CRVE, was negatively correlated with mean BP (r = )0.31, p = 0.005 for CRAE and r = )0.14, p = 0.22 for CRVE). The univariate analysis did not find correlations at a significant level between CRAE and body mass index (p = 0.94), triglyceride levels (p = 0.16), fasting blood glucose (p = 0.93), gender (p = 0.08), smoking (p = 0.26), whereas a negative one was found with cholesterol level (r = )0.25, p = 0.02). In univariate analysis, no correlation was found between CRVE and body mass index (p = 0.15), triglyceride levels (p = 0.09), fasting blood glucose (p = 0.26), cholesterol level (p = 0.27), gender (p = 0.09) and smoking (p = 0.13).
Table 1. Population characteristics. Number of subjects
Overall 80
Normotensive (51)
Hypertensive (29)
Age (year) Male gender (%) Current smokers (%) Body mass index (kg ⁄ m2) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Mean blood pressure (mmHg) Heart rate (bpm) Fasting blood glucose (mg ⁄ dl) Total cholesterol (mg ⁄ dl) HDL-cholesterol (mg ⁄ dl) Triglycerides (mg ⁄ dl) Serum creatinine (mg ⁄ dl) Glomerular filtration rate (ml ⁄ min ⁄ 1.73m2) Urinary albumin ⁄ creatinine ratio (mg ⁄ mmol)
47 ± 14 49 30 25.7 ± 4 135 ± 19 81 ± 12 99 ± 14 66 ± 10 88 ± 9 211 ± 45 60 ± 17 98 ± 46 0.8 ± 0.2 117 ± 20
46 ± 14 28 22 24.7 ± 4 124 ± 11 75 ± 10 92 ± 10 66 ± 11 88 ± 9 206 ± 46 61 ± 18 90 ± 38 0.8 ± 0.1 115 ± 20
50 ± 13 21 8 26.7 ± 4 156 ± 16 91 ± 11 113 ± 11 67 ± 9 90 ± 11 220 ± 41 59 ± 15 114 ± 57 0.8 ± 0.2 120 ± 19
0.59 [0.39–1.14]
0.52 [0.36–0.89]
0.83 [0.49–1.67]
Values are expressed as percentage, mean ± standard deviation or mean [interquartile range].
Table 2. Mean (±SD) central retinal artery equivalent, central retinal vein equivalent and arteriole-to-venule ratio in normotensive and hypertensive subjects.
Adjusted central retinal artery equivalent (lm) Adjusted central retinal vein equivalent (lm) Adjusted arteriole-to-venule ratio
Normotensive
Hypertensive
p value
142 ± 18 208 ± 30 0.70 ± 0.09
135 ± 24 203 ± 23 0.65 ± 0.12
0.03 0.28 0.007
* Adjusted for age, gender, body mass index, smoking status, total cholesterol, triglyceride levels, fasting blood glucose and glomerular filtration rate.
Association of renal function and retinal microvasculature
Central retinal artery and vein equivalents were positively correlated with measured GFR (r = +0.31, p = 0.005 and r = +0.30, p = 0.006, respectively). As a consequence, AVR was not correlated with GFR (r = +0.02; p = 0.83). Central retinal artery equivalents was negatively correlated with UACR (r = )0.34, p = 0.002); whereas, no relationship with CRVE was found (r = +0.12, p = 0.32). AVR was negatively correlated with UACR (r = )0.41; p = 0.0002). The independent contribution of measured GFR and UACR to retinal vascular caliber is depicted in Table 3, considering adjustments for age and mean BP in model 1, but also other potential confounding factors such as gender, body mass index, current smoking status, total cholesterol, triglyceride levels and fasting blood glucose in model 2.
Fig. 1. Correlations between retinal vascular calibers – that is central retinal artery equivalent (CRAE) and central retinal venular equivalent (CRVE) – and age, mean blood pressure, glomerular filtration rate and urinary albumin-to-creatinine ratio.
e285
Acta Ophthalmologica 2013
Table 3. Multivariate regression analysis of retinal vascular calibers as the dependent variables and renal outcomes as the explanatory variables. CRAE (lm) b ± SD Glomerular filtration rate (ml ⁄ min ⁄ 1.73m2) Crude 0.22 Adjusted model 1* 0.21 Adjusted model 2 0.15 Urinary albumin ⁄ creatinine ratio (mg ⁄ mmol, log) Crude )12.34 Adjusted model 1* )10.59 Adjusted model 2 )9.75
CRVE (lm) p-value
b ± SD
AVR p-value
b ± SD
p-value
± 0.07 ± 0.08 ± 0.07
0.005 0.01 0.04
0.29 ± 0.10 0.26 ± 0.10 0.23 ± 0.11
0.007 0.01 0.04
)8.10)4 ± 4.10)4 )2.10)4 ± 4.10)4 )5.10)4 ± 4.10)4
0.83 0.64 0.90
± 4.25 ± 4.88 ± 4.32
0.005 0.03 0.02
7.03 ± 6.27 14.01 ± 6.16 9.72 ± 6.77
0.26 0.15 0.15
)0.08 ± 0.02 )0.09 ± 0.03 )0.08 ± 0.03
0.004 0.003 0.003
CRAE = central retinal artery equivalent; CRVE = central retinal vein equivalent; AVR = arteriole-to-venule ratio. * Models 1 adjusted for age (years) and mean blood pressure (mmHg). Models 2 adjusted for age (years), gender (male, female), mean blood pressure (mmHg), body mass index (kg ⁄ m2), current smoking status (absent, present), total cholesterol (mg ⁄ dl), triglyceride levels (mg ⁄ dl), fasting blood glucose (mg ⁄ dl).
Of interest, when measured GFR was replaced by an estimated GFR (MDRD [Levey et al. 1999] or CKDEPI [Levey et al. 2009]), the renal function parameter remained significantly associated with CRAE and CRVE.
Discussion In the present study conducted in apparently healthy subjects with normal renal function, it was observed that a smaller retinal arterial caliber is associated with lower measured GFR and higher albuminuria and that a smaller retinal venular caliber is related to lower GFR but not to the degree of albuminuria, independently of other potential confounding factors. In a population-based study conducted in Asia, reduced CRAE was associated with the presence of CKD (estimated GFR
