Blood pressure development and hypertensive retinopathy - Nature

Journal of Human Hypertension (2010) 24, 505–513 & 2010 Macmillan Publishers Limited All rights reserved 0950-9240/10 www.nature.com/jhh

ORIGINAL ARTICLE

Blood pressure development and hypertensive retinopathy: 20-year follow-up of middle-aged normotensive and hypertensive men H Gudmundsdottir1,4, NCBB Taarnhøj2,6, AH Strand3,4, SE Kjeldsen3,4,5, A Høieggen1,4 and I Os1,4,5 1

Department of Nephrology, Ulleval University Hospital, Oslo, Norway; 2Department of Ophthalmology, Ulleval University Hospital, Oslo, Norway; 3Department of Cardiology, Ulleval University Hospital, Oslo, Norway; 4Department of Cardiovascular and Renal Research Center, Ulleval University Hospital, Oslo, Norway; 5Faculty of Medicine, University of Oslo, Oslo, Norway and 6Department of Ophthalmology, Glostrup Hospital, University of Copenhagen, Copenhagen, Denmark

Screening for hypertensive organ damage is important in assessing cardiovascular risk in hypertensive individuals. In a 20-year follow-up of normotensive and hypertensive men, signs of end-organ damage were examined, focusing on hypertensive retinopathy. In all, 56 of the original 79 men were reexamined for hypertensive organ damage, including by digital fundus photography. The diameters of the central retinal artery equivalent (CRAE) and vein were estimated and the artery-to-vein diameter ratio calculated. Components of metabolic syndrome were assessed. Fifty percent of the normotensive men developed hypertension during follow-up. Significant differences appeared in CRAE between the different blood pressure groups (P ¼ 0.025) while no differences were observed for other markers of

hypertensive organ damage. There were significant relationships between CRAE and blood pressure at baseline (r ¼ –0.466, P ¼ 0.001) and at follow-up (r ¼ –0.508, Po0.001). A linear decrease in CRAE was observed with increasing number of components of the metabolic syndrome (b ¼ –3.947, R2 ¼ 0.105, P ¼ 0.023). Retinal vascular diameters were closely linked to blood pressures and risk factors of the metabolic syndrome. The diversity in the development of hypertensive organ damage, with changes in retinal microvasculature preceding other signs of damage, should encourage more liberal use of fundus photography in assessing cardiovascular risk in hypertensive individuals. Journal of Human Hypertension (2010) 24, 505–513; doi:10.1038/jhh.2009.94; published online 10 December 2009

Keywords: hypertensive retinopathy; blood pressure; metabolic syndrome

Introduction Guidelines for the management of hypertension stress the importance of assessing subclinical target organ damage to enable stratification of cardiovascular risk and to guide treatment decisions. Hypertensive retinopathy refers to retinal microvascular signs that develop in response to raised blood pressure. Large epidemiological studies have shown associations between retinal vessel abnormalities and systemic disease such as hypertension, cardiovascular disease, diabetes mellitus and stroke.1–4

Correspondence: Dr H Gudmundsdottir, Department of Nephrology, Ulleva˚l University Hospital, 0407 Oslo, Norway. E-mail: [email protected] Received 6 September 2009; revised 9 November 2009; accepted 10 November 2009; published online 10 December 2009

These studies showed a relationship between hypertensive retinopathy and current as well as past blood pressure values. The eye is the only place in the body where it is possible to examine blood vessels of the central nervous system noninvasively and nondestructively. Physiologically, retinal arteries constrict in response to elevated blood pressure as a result of autoregulation. The primary pathologic change in hypertensive retinopathy is therefore protective vasoconstriction, generalized arterial narrowing and arterial straightening.5,6 International guidelines recommend examination of the optic fundus for initial evaluation of patients with hypertension, although with some variations.7–9 The additional value of the information provided has been questioned, mainly because of the limitations caused by large interobserver variation.10

Blood pressure and retinopathy: a 20-year follow-up H Gudmundsdottir et al 506

Computer-assisted methods allow reliable and precise measurements of retinal vessel calibre from digital retinal photographs.4,11–13 Fouguet et al.14 concluded that retinography allowed a more accurate stratification of cardiovascular risk in the first evaluation after diagnosis of hypertension, causing approximately 10% of the patients to be reclassified to a higher risk group. This underlies the importance of taking into account hypertensive retinopathy as target organ damage in evaluation and treatment of hypertensive individuals. The aim of this study was to analyze the relationship between blood pressure and the development of hypertensive target organ damage, with a focus on hypertensive retinopathy, in a group of normotensive and hypertensive, middle-aged Caucasian men followed for 20 years.

Methods Study population

At baseline in 1984, 79 Caucasian men (35 hypertensive and 44 normotensive) were recruited from The Oslo Study at an age of 42.1 ± 0.5 years. The hypertensive group (n ¼ 35) consisted of men with untreated mild essential hypertension with systolic blood pressure between 140 and 170 mm Hg and/or diastolic blood pressure (DBP) between 90 and 100 mm Hg when participating in the Oslo study. Participants were included in this study if at baseline they had DBP between 94 and 105 mm Hg on two separate occasions.15,16 The normotensive men (n ¼ 44) were recruited from the same population, that is, from the normotensive participants in The Oslo Study, and had blood pressure below 140/90 mm Hg. All 79 men had normal ocular fundi, electrocardiograms, urinalysis and kidney function estimated by creatinine clearance. At 20-year follow-up, all 79 individuals from 1984 were located and of these 56 (70.1%) participated in the follow-up examination: 34 of the original 44 normotensive men and 22 of the original 35 hypertensive men. In all, 7 men had died (5 normotensive and 2 hypertensive), 6 men had emigrated or moved to another county (3 normotensive and 3 hypertensive) and 10 men (2 normotensive and 8 hypertensive) did not respond or responded negatively to a letter of invitation. Those who remained normotensive through the 20-year follow-up were defined as sustained normotensive ( n ¼ 17). At the time of follow-up these men had office systolic blood pressure o140 and DBP o90 mm Hg and mean 24-h ambulatory blood pressure o125/80 mm Hg. Men who developed hypertension during the follow-up were defined as new hypertensives (n ¼ 17). They had either office systolic blood pressure X140 mm Hg or office DBP X90 mm Hg and 24-h ambulatory systolic blood pressure X125 mm Hg and 24-h ambulatory Journal of Human Hypertension

DBP X80 mm Hg or were taking antihypertensive medication (n ¼ 5) at the time of follow-up. The same criteria were used for the sustained hypertensive men (n ¼ 22) as for the new hypertensives, but they were hypertensive in 1984 as well as at 20year follow-up. Of these 22 sustained hypertensive men, 16 (72.7%) used antihypertensive drugs. Of the 17 men defined as new hypertensives, only 5 (29.4%) were taking antihypertensive drugs. Characteristics of the study population are listed in Table 1. Study protocol in 1984

The study protocol in 1984 has been described earlier.15,16 All participants were examined in an outpatient setting by the same physicians. They were studied at the same time of day in a quiet room and at a constant room temperature. All were familiar with the clinical examination and blood pressure recording. They fasted and abstained from smoking for 8 h and from alcohol for 24 h before examination. Blood pressure was recorded semiautomatically (oscillometric technique) with an Omega 1000 Adult/pediatric blood pressure (In Vivo Research Laboratories Inc., Tulsa, OK, USA). Short Teflon catheters (Venflon, former Viggo AB, Helsingborg, Sweden) were introduced under local anaesthesia (Xylocaine without adrenaline, (Astra, So¨derta¨lje, Sweden)) by one investigator in the left brachial artery. The institutional committee of ethics had approved the study and informed consent was obtained from each participant. Study protocol in 2004

The study protocol in 2004 has been described earlier.17 The participants were studied in the same facilities as 20 years ago and at the same time of day. All were familiar with the clinical examination and blood pressure measurements. They fasted and abstained from smoking for 8 h and from alcohol for 24 h before the examination. They were all examined by the same physicians who were not aware of their previous blood pressure status. Blood pressure was measured in duplicate with a calibrated mercury sphygmomanometer after 5 min rest in a sitting position. The arm was supported at heart level and an appropriately sized cuff was used. The mean of the two measurements was used for statistical analyses. Mean blood pressure (MBP) was calculated using the formula: MBP ¼ DBP þ (pulse pressure/3). Heart rate was registered after 5 min in a sitting position. Height, weight, waist circumference and body mass index were measured and calculated using standard methods. The National Committee of Medical Research Ethics in Norway approved the follow-up study, and a concession was granted from the National Data Inspectorate in Norway. All the participants gave written informed consent.


Table 1 Characteristics of the study population; NTNT: sustained normotensives, NTHT: new hypertensives, HTHT: sustained hypertensives NTNT (n ¼ 14)

NTHT (n ¼ 16)

HTHT (n ¼ 19)

P-value

1984 BMI (kg m–2) SBP (mm Hg) DBP (mm Hg) MBP (mm Hg) HR (b.p.m.) Haematocrit CrCl (ml min–1)

24.0±2.2 117.9±6.2 71.1±6.2 86.7±5.4 61.7±8.4 0.40±0.004 122.9±17.3

24.3±3.7 121.8±5.9 74.5±5.2 90.2±4.5 65.6±7.1 0.42±0.005 123.8±21.7

24.9±2.4 138.2±8.1 91.2±4.7 106.9±4.8 74.7±15.8 0.42±0.004 121.1±21.3

NS o0.001 o0.001 o0.001 o0.001 0.012 NS

2004 BMI (kg m–2) SBP (mm Hg) DBP (mm Hg) MBP (mm Hg) HR (b.p.m.) CRAE (mm) CRVE (mm) AVR Haematocrit CrCl (ml min–1) ACR (mg mmol–1) Carotid plaque score Mean IMT (mm) LV mass index (g m–2) LV hypertrophy (n) Diastolic dysfunction (n)

26.2±2.8 134.8±11.7 85.8±6.5 102.2±7.0 64.6±6.9 166.6±18.1 229.8±21.7 0.73±0.07 0.40±0.005 102.7±12.9 0.20±0.14 1.50±1.41 0.82 ±0.24 102±17 1 2

26.9±4.0 149.5±11.4 95.0±8.3 113.1±8.2 63.9±9.4 156.6±18.9 227.5±17.2 0.69±0.08 0.41±0.007 109.1±24.6 0.45±0.44 1.53±1.62 0.86±0.18 111±25 6 6

28.1±3.8 165.2±23.7 102.3±8.2 123.3±12.1 70.7±10.0 148.4±18.2 215.7±17.9 0.69±0.08 0.41±0.007 109.7±22.3 0.26±0.22 2.64±1.86 0.98±0.30 114±20 10 13

ns o0.001 o0.001 o0.001 0.037 0.025 NS NS NS NS NS NS NS NS 0.025 0.010

Variable

Abbreviations: AVR, artery-to-vein ratio; ACR, albumin/creatinine ratio; BMI, body mass index; b.p.m., beats per minute; CRAE, central retinal artery equivalent; CrCl, creatinine clearance; CRVE, central retinal vein equivalent; DBP, diastolic blood pressure; HR, heart rate; IMT, intima media thickness; LV, left ventricular; SBP; systolic blood pressure.

Methods

At 20-year follow-up, digital colour fundus photographs (451, 768 576 pixels centred on the macula or optic disc) of both eyes of all participants were recorded using a retinal camera (TOPCON TRC.NW5S, TOPCON Corporation, Tokyo, Japan). The photography was performed by an experienced investigator. In 49 individuals, the quality of the photographs was sufficient to perform retinal vessel analysis. Eyes were considered ungradable if one of the six largest arteries or veins could not be measured or if the image was of poor quality (low contrast), as judged by the grader with reference to a standard image of least acceptable quality and contrast. In 39 individuals, an image of the left eye was used in the analysis and in 10 individuals an image of the right eye was used, based on evaluation of the visual quality of the images. All fundus photographs were evaluated and graded by the same experienced ophthalmologist who was unaware of the individual’s blood pressure status. Retinal vessel diameters were assessed using a custom-developed semiautomatic computer programme. The vascular morphology measurement tool used the active contour model, the so-called ‘snake’, which is initiated by the grader and subsequently delineates the blood column at a subpixel level, and hence performs very precise measurements. This technique has been described elsewhere.18 The vessel

walls and the sleeve of plasma that surround the blood column are invisible. A standard grid containing three concentric circles was placed on the images, and only diameters of the vessels crossing the circular zone (zone B) from 0.5 disc diameter to 1.0 disc diameter from the outer rim of the optic disc were measured. The grader identified arteries and veins, using red and blue lines respectively to delineate the vessels. A segment of each vessel within zone B that was most suitable for measurement was chosen based on image quality, contrast, straightness of the vessel, and distance from branching points and crossings. When bifurcating or branching was found within the zone of interest, the trunk was preferred to the branches, unless the trunk was shorter than 80 mm. At the end of the measurement session, the computer programme identified the six widest arteries and the six widest veins and calculated the central retinal artery equivalent (CRAE) and the central retinal vein equivalent (CRVE), using the formulas developed by Knudtson et al.19 The artery-to-vein ratio (AVR) was defined as CRAE/CRVE. Absolute distances were determined assuming a uniform vertical optic nerve head diameter of 1800 mm (Figure 1). Intima media thickness and plaques were measured to detect structural changes in the carotid artery. The participants were examined in a supine position with an ultrasound scanner (Acuson 128, Journal of Human Hypertension


Figure 1 Fundus image showing a standard grid with measured arteries and veins delineated in red and blue, respectively. The inner circle demarcates an average optic disc, the middle circle demarcates 0.5 disc diameters from the outer rim of the optic disc and the outer circle demarcates 1.0 disc diameter from the outer rim of the optic disc. Zone B was defined as the region from 0.5 disc diameter to 1.0 disc diameter from the disc margin.

Mountain View, CA, USA) with a 7.0 MHz linear array transducer.20 The plaque score was visually determined online as grade 0: no plaque; grade 1: small plaques (o10 mm2) or intima media thickening o1.2 mm2; grade 2: two or more small plaques (o10 mm2); grade 3: one plaque X10 mm2; grade 4: one plaque X10 mm2; grade 5: circumferential plaque and/or two or more plaques X10 mm2 and/ or large plaques (50% stenosis). All scans were performed by the same sonographer who was unaware of the participants’ blood pressure status. Echocardiography was carried out as described earlier21 by one experienced investigator using a GE-Vingmed Vivid 7 echocardiograph (Horten, Norway) with a 1.7 MHz probe in second harmonic mode. The investigator doing the echocardiography had no knowledge of the participant’s blood pressure status. End-diastolic left ventricular (LV) dimensions were used to calculate LV mass by an anatomically validated formula (r ¼ 0.9 versus necropsy LV mass).22 LV hypertrophy was defined as LV mass index 4116 g m–2.23 All participants were in sinus rhythm and measurements of up to three cycles were averaged. Definition of the components of the metabolic syndrome was based on the International Diabetes Federation criteria.24 The variables were dichotomized. The variables used as risk factors were hypertension at follow-up, waist circumference X94 cm, fasting glucose X5.6 mmol l–1, high-density lipoprotein cholesterol o 1.0 mmol l–1 or specific treatment for this disorder and triglycerides X1.7 mmol l–1 or specific treatment for this disorder. Biochemical assays

Fasting blood samples were taken as described earlier. Haematocrit was measured using an ORTHO-ELT 800/WS (Ortho Diagnostic Systems, Westwood, MA, USA) (1984) and Sysmex XE-2100 (System Corp., Kobe, Japan) (2004). At follow-up, Journal of Human Hypertension

serum and urine creatinine, fasting glucose, total cholesterol, high-density lipoprotein and triglycerides were measured with an enzymatic method on a routine clinical chemistry analyzer (Cobas Integra, Roche, Basel, Switzerland). Creatinine clearance was calculated using the Cockcroft–Gault formula.25 Urinary albumin excretion was determined using albumin/creatinine ratio in the first voided morning urine sample on the day of the clinical examination. Urine albumin was measured turbimetrically on a Hitachi 912 autoanalyzer, using Tinaquant reagents from Roche, Basel, Switzerland. Blood and urine samples were analyzed in a mixed order by technicians who were unaware of the participants’ blood pressure status. Statistics

SPSS 15.0 (SPSS Inc., Chicago, IL, USA) was used for data management and statistical analysis. Results are presented as means ± s.d. unless otherwise stated. Differences are presented as means with 95% confidence interval. Parametric tests were used for normally distributed data, while log transformations of data or nonparametric tests were applied if data were skewed. Differences between two groups were assessed either by Student’s t-test or with the Mann– Whitney test and within-group differences with either the Student’s t-test or Wilcoxon test. Comparisons between multiple groups were performed by one-way analysis of variance or Kruskal–Wallis (analysis of variance). Univariate relationships between variables were assessed by the Spearman (r) correlation coefficient or by the Pearson’s correlation coefficient. A two-tailed P-value o0.05 was considered statistically significant. Multiple regression models were created with CRAE, CRVE and arteriovenous ratio as dependant variables, and MBP, haematocrit, body mass index, creatinine clearance and smoking status at baseline as independent variables. The selection of variables was


Results In all, 56 of the 79 men were accessible for examination at 20-year follow-up, as shown in the flowchart in Figure 2. All were Caucasian and of similar age. Characteristics of the men are given in Table 1. Of the original normotensive participants, 17 (50%) were categorized as new hypertensives because they had developed hypertension during follow-up. Among the hypertensive individuals, 16 were using antihypertensive agents at follow-up, either monotherapy or different combinations. Twelve of these were treated with agents inhibiting the renin–angiotensin system. Retinal photographs were obtained from all 56 individuals, with quality sufficient to perform retinal evaluations in 87.5%. In the 49 individuals with usable photographs, the mean CRAEs was 156.3±19.5 mm and the mean CRVEs 223.6± 19.5 mm. AVR ranged from 0.42 to 0.87 (median 0.70). There were significant differences in CRAE between the different blood pressure groups as shown in Figure 3, but not with regard to CRVE and AVR (Table 1). In the whole study population, significant relationships were observed between CRAE and blood pressure at baseline and at 20-year follow-up (r ¼ –0.466, P ¼ 0.001; r ¼ –0.508, Po0.001, respectively, Figures 4a and b). Although haematocrit differed between the blood pressure groups at baseline (P ¼ 0.012), the difference was lost at 20-year follow-up (Table 1). Haematocrit at baseline was significantly associated with CRAE (r ¼ –0.308, P ¼ 0.032,) and AVR at follow-up (r ¼ –0.466, P ¼ 0.001, Figures 5a and b). A significant difference in CRVE was observed between smokers and nonsmokers at baseline. Smokers had a larger CRVE than nonsmokers (233.0±20.7 mm and 219.8±18.1 mm, respectively;

P ¼ 0.031). There was no significant difference in CRAE and AVR between these two groups. Fifty-one (91%) of the participating individuals had one or more components of the metabolic syndrome at follow-up. A linear decrease in CRAE was observed with an increase in the number of components of the metabolic syndrome (b ¼ –3.947, R2 ¼ 0.105, P ¼ 0.023, Figure 6). Multiple regression analysis with CRAE as the dependant variable, and MBP, haematocrit, body mass index, creatinine clearance and smoking status at baseline as independent variables revealed that MBP was the only independent predictor of CRAE (b ¼ –0.792, R2 ¼ 0.181, P ¼ 0.002). Using the same model with CRVE as the dependant variable, MBP and haematocrit were independent predictors of CRVE (b ¼ –0.919, R2 ¼ 0.267, P ¼ 0.011). When AVR was entered as the dependant variable in the same model, haematocrit at baseline was the only independent predictor of AVR at follow-up (b ¼ –1.735, R2 ¼ 0.217, P ¼ 0.001).

***

200 Central retinal artery equivalent (µm)

based on clinical relevance. Furthermore, a regression analysis was performed with CRAE as the dependant variable and the number of variables defined as components of the metabolic syndrome as independent variables.

*

**

180

160

140 ANOVA p = 0.025 120 NTNT

NTHT

HTHT

Figure 3 Box plot of central retinal artery equivalents (CRAEs) at 20-year follow-up. NTNT, sustained normotensives; NTHT, new hypertensives; HTHT, sustained hypertensives, (n ¼ 49). Analysis of variance (ANOVA), P ¼ 0.025; *NTNT versus NTHT, P ¼ NS; **NTHT versus HTHT, P ¼ NS; ***NTNT versus HTHT, P ¼ 0.008.

Participants in 1984 n = 79

Normotensives in 1984 n = 44 Dropouts n = 10

Sustaiend Normotensives in 2004 n = 17

Hypertensives in 1984 n = 35 Dropouts n = 13

New Hypertensives in 2004 n = 17

Sustained Hypertensives in 2004 n = 22

Figure 2 Flowcharts of the study population at baseline and at 20-year follow-up. Journal of Human Hypertension


200 Central retinal artery equivalent (µm)

Central retinal artery equivalent (µm)

200

180

160

140

120

r = -0.466 p = 0.001

180

160

140

120

r = -0.308 p = 0.032

100 100 0.375 70

80

90

100

110

120

0.40

0.425

0.45

0.475

Hematocrit at baseline (%)

Mean blood pressure at baseline ( mmHg) 0.90

0.80 180

Artery-to-vein ratio


200

160

140

0.70

0.60

0.50 120 r = -0.508 p < 0.001

r = -0.466 p = 0.001

0.40 0.375

100 80

100

120

140

Mean blood pressure at 20-year follow-up (mmHg) Figure 4 Relationship between central retinal artery equivalent (CRAEs) at 20-year follow-up and mean blood pressure (MBP) at baseline (a) and at 20-year follow-up (b), (n ¼ 49).

At follow-up there were no significant differences in albumin/creatinine ratio, creatinine clearance and carotid plaque score between the different blood pressure groups. There were, however, significant differences between the groups with respect to number of individuals with LV hypertrophy and diastolic dysfunction, but not with regard to LV mass index (Table 1).

Discussion Our data suggest that hypertensive retinopathy measured by fundus photography could be considered an early marker of hypertensive organ damage. Individuals with hypertension at 20-year follow-up had significantly more hypertensive retinopathy than those who remained normotensive. Furthermore, we found that MBP predicted CRAE at Journal of Human Hypertension

0.40

0.425

0.45

0.475

Hematocrit at baseline (%) Figure 5 Relationship between haematocrit at baseline and central retinal artery equivalent (CRAE) (a) and artery-to-vein ratio (AVR) (b) at 20-year follow-up, (n ¼ 49).

follow-up in the whole group. The observation is in keeping with those of others reporting that generalized retinal arteriolar narrowing was inversely related to current blood pressure as well as to blood pressure measured three to 8 years earlier.2,4,12 This suggests that retinal arteriolar narrowing may be the result of long-term microvascular damage caused by elevated blood pressure.12 Structural changes in the retinal vasculature have long been acknowledged as a predictor of hypertensive organ damage.26 Our results extend these findings to a 20-year follow-up, albeit in a small group of individuals. The relationship between mean blood pressure at baseline and at follow-up and CRAE seemed stronger than the relationship between blood pressure and CRVE, implying that, as suggested earlier, the effect of blood pressure on arterioles may be more prominent than that on venules.2,27



200 p = 0.023 180

160

140

120 0

1 2 3 4 5 Components of the metabolic syndrome

Figure 6 Central retinal artery equivalent (CRAE) stratified according to components of the metabolic syndrome at 20-year follow-up (n ¼ 49). Zero indicates no component (n ¼ 5); 1 indicates one component (n ¼ 12); 2 indicates two components (n ¼ 11); 3 indicates three components (n ¼ 10); 4 indicates four components (n ¼ 2) and 5 indicates five components (n ¼ 9) of the metabolic syndrome. Bars represent 1 s.d.

We did not observe a relationship between the development of hypertensive retinopathy and LV hypertrophy. Conventional echocardiography did not discriminate differences in LV mass index in the different blood pressure groups as reported earlier.21 There were, however, significantly more hypertensive individuals than normotensives who had developed LV hypertrophy and diastolic dysfunction during the follow-up.21 There were no differences in other markers of end-organ damage, that is, creatinine clearance, urinary albumin excretion and carotid plaque score, between the different blood pressure groups. This diversity in development of end-organ damage was somewhat unexpected but may imply that changes in retinal microvasculature precede other signs of end-organ damage. Increased urinary albumin excretion in hypertensive individuals is thought to be an early marker of renal damage, and can be reversed and influenced by antihypertensive therapy, in particular by agents that inhibit the renin–angiotensin system. The majority of the hypertensive individuals used either angiotensin-converting-enzyme inhibitor or angiotensin receptor blocker and this might at least partially explain the lack of difference in albumin excretion between the blood pressure groups. Haematocrit as a marker of whole blood viscosity has been associated with hypertension and cardiovascular mortality.28,29 Whole blood viscosity has also been linked to components of the metabolic syndrome.30,31 In this study, we observed a strong relationship between haematocrit at baseline and CRAE and AVR. Thus, whole blood viscosity may be implicated in the development of hypertensive

organ damage in addition to being predictive of development of hypertension as suggested earlier.32 More than 90% of the study population presented with one or more components of the metabolic syndrome as defined by the International Diabetes Federation criteria.24 The individuals with the highest number of risk factors had the smallest diameter of the retinal arteries showing the relationship between hypertensive end-organ damage and the metabolic syndrome. This novel observation, linking the number of risk factors in the metabolic syndrome to hypertensive retinopathy, indicates that not only blood pressure but also other risk factors of the metabolic syndrome need to be considered. In the Rotterdam study,33 not only blood pressure but also cholesterol levels, atherosclerosis, high waist-to-hip ratio and inflammation markers affected the retinal vascular diameter. In our study, smoking influenced the diameter of the retinal veins, smokers having more dilated veins then nonsmokers. This was observed earlier in large population studies, the putative mechanism being that smoking causes retinal hypoxia, causing venular dilation.33–35 Smoking status did not influence the retinal arterial diameter in our study. This is consistent with the literature that states that smoking has little effect on arteriolar diameter or causes only mild arteriolar narrowing rather than widening.36,37 The strengths of the present investigation are the long observation period of 20 years and the unique homogeneity of the study population. Structural changes in the retinal vasculature are found both in hypertensive patients and in the normal ageing population. Different ages in a study population could therefore complicate interpretation of results. In our study, the men were all of the same age, making our findings easier to interpret, although the results may not be applicable to other populations. The quantitative measurement of the retinal vascular calibre and the fact that few of our participants had ungradable photographs strengthens our findings. The power of the study is limited by the small number of participants. The small number could, however, make the findings more clinically relevant. We could reexamine two-thirds of the original study population. Blood pressure at baseline did not differ between those who participated in the follow-up study and those who did not. It would therefore be fair to assume that selection errors were to a large extent avoided although a larger participation of the original group would have been preferable. High blood pressure is a preventable risk factor for cardiovascular morbidity and mortality. Subclinical organ damage is currently regarded as an intermediate stage in the continuum of vascular diseases and a strong determinant of total cardiovascular risk in both normotensive and hypertensive individuals. Fundus photography with quantitative measurement of retinal vascular diameter might enable more accurate stratification for cardiovascular risk in Journal of Human Hypertension


individuals with hypertension, ensuring optimal treatment and follow-up. The diversity in development of hypertensive organ damage with changes in retinal microvasculature preceding other signs of damage should encourage more liberal use of fundus photography. What is known about this topic K Screening for hypertensive organ damage is important in assessing cardiovascular risk in hypertensive individuals. K Previous studies have demonstrated a relationship between hypertensive retinopathy and current as well as past blood pressure values (1–4). What this study adds K This 20-year follow-up of middle-aged men showed that changes in the retinal microvasculature may precede other signs of hypertensive end-organ damage. K Fundus photography with quantitative measurement of retinal vascular diameter might enable more accurate stratification for cardiovascular risk in individuals with hypertension, ensuring optimal treatment and follow-up.

Conflict of interest The authors declare no conflict of interest.

Acknowledgements We thank the technical staff at the Cardiovascular and Renal research laboratory at the Oslo University Hospital, Ulleva˚l, Oslo for expert technical assistance. Special thanks to Ms A Skafjeld and to Ms Lise Bergengen for valuable help. This work was supported by The Norwegian Renal Association and research funding from the Department of Nephrology, Ulleva˚l University Hospital.

References 1 Leung H, Wang JJ, Rochtchina E, Wong TY, Klein R, Mitchell P et al. Impact of current and past blood pressure on retinal arteriolar diameter in an older population. J Hypertens 2004; 22(8): 1543–1549. 2 Sharrett AR, Hubbard LD, Cooper LS, Sorlie PD, Brothers RJ, Nieto FJ et al. Retinal arteriolar diameters and elevated blood pressure: the Atherosclerosis Risk in Communities Study. Am J Epidemiol 1999; 150(3): 263–270. 3 Wong TY, Klein R, Couper DJ, Cooper LS, Shahar E, Hubbard LD et al. Retinal microvascular abnormalities and incident stroke: the Atherosclerosis Risk in Communities Study. Lancet 2001; 358(9288): 1134–1140. 4 Wong TY, Hubbard LD, Klein R, Marino EK, Kronmal R, Sharrett AR et al. Retinal microvascular abnormalities and blood pressure in older people: the Cardiovascular Health Study. Br J Ophthalmol 2002; 86(9): 1007–1013. 5 de Margerie J, Boyd TA. A statistical investigation of the correlation of retinal arteriolar tortuosity with blood pressure and age. Trans Can Opthalmol Soc 1961; 24: 6–17. Journal of Human Hypertension

6 Wolffsohn JS, Napper GA, Ho SM, Jaworski A, Pollard TL. Improving the description of the retinal vasculature and patient history taking for monitoring systemic hypertension. Ophthalmic Physiol Opt 2001; 21(6): 441–449. 7 Chobanian AV, Bakris GL, Black HR, Cushman WC, Green LA, Izzo Jr JL et al. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: the JNC 7 report. JAMA 2003; 289(19): 2560–2572. 8 Mancia G, De Backer G, Dominiczak A, Cifkova R, Fagard R, Germano G et al. 2007 Guidelines for the Management of Arterial Hypertension: The Task Force for the Management of Arterial Hypertension of the European Society of Hypertension (ESH) and of the European Society of Cardiology (ESC). J Hypertens 2007; 25(6): 1105–1187. 9 Williams B, Poulter NR, Brown MJ, Davis M, McInnes GT, Potter JF et al. British Hypertension Society guidelines for hypertension management 2004 (BHSIV): summary. BMJ 2004; 328(7440): 634–640. 10 van den Born BJ, Williams B, Poulter NR, Brown MJ, Davis M, McInnes GT et al. Value of routine funduscopy in patients with hypertension: systematic review. BMJ 2005; 331(7508): 73. 11 Hubbard LD, Brothers RJ, King WN, Clegg LX, Klein R, Cooper LS et al. Methods for evaluation of retinal microvascular abnormalities associated with hypertension/sclerosis in the Atherosclerosis Risk in Communities Study. Ophthalmology 1999; 106(12): 2269–2280. 12 Leung H, Wang JJ, Rochtchina E, Tan AG, Wong TY, Klein R et al. Relationships between age, blood pressure, and retinal vessel diameters in an older population. Invest Ophthalmol Vis Sci 2003; 44(7): 2900–2904. 13 Taarnhøj NC, Larsen M, Sander B, Kyvik KO, Kessel L, Hougaard JL et al. Heritability of retinal vessel diameters and blood pressure: a twin study. Invest Ophthalmol Vis Sci 2006; 47(8): 3539–3544. 14 Foguet Q, Rodrı´guez A, Saez M, Ubieto A, Beltra´n M, Barcelo´ MA et al. Usefulness of optic fundus examination with retinography in initial evaluation of hypertensive patients. Am J Hypertens 2008; 21(4): 400–405. 15 Os I, Kjeldsen SE, Westheim A, Aakesson I, Norman N, Enger E et al. Decreased central dopaminergic activity in essential hypertension. J Hypertens 1987; 5(2): 191–197. 16 Os I, Kjeldsen SE, Westheim A, Lande K, Norman N, Hjermann I et al. Endocrine and haemodynamic responses to graded dopamine infusion in essential hypertension. Scand J Clin Lab Invest 1987; 47(4): 371–377. 17 Gudmundsdottir H, Strand AH, Høieggen A, Reims HM, Westheim AS, Eide IK et al. Do screening blood pressure and plasma catecholamines predict development of hypertension? Twenty-year follow-up of middle-aged men. Blood Press 2008; 17(2): 94–103. 18 Kass M, Witkin A, Terzopoulos D. Snakes: Active contour models. Int J Comput Vision 1988; 1(4): 321–331. 19 Knudtson MD, Lee KE, Hubbard LD, Wong TY, Klein R, Klein BE et al. Revised formulas for summarizing retinal vessel diameters. Curr Eye Res 2003; 27(3): 143–149.


20 Persson J, Stavenow L, Wikstrand J, Israelsson B, Formgren J, Berglund G et al. Noninvasive quantification of atherosclerotic lesions. Reproducibility of ultrasonographic measurement of arterial wall thickness and plaque size. Arterioscler Thromb 1992; 12(2): 261–266. 21 Strand AH, Gudmundsdottir H, Os I, Smith G, Westheim AS, Bjørnerheim R et al. Arterial plasma noradrenaline predicts left ventricular mass independently of blood pressure and body build in men who develop hypertension over 20 years. J Hypertens 2006; 24(5): 905–913. 22 Devereux RB, Alonso DR, Lutas EM, Gottlieb GJ, Campo E, Sachs I et al. Echocardiographic assessment of left ventricular hypertrophy: comparison to necropsy findings. Am J Cardiol 1986; 57(6): 450–458. 23 Wachtell K, Smith G, Gerdts E, Dahlo¨f B, Nieminen MS, Papademetriou V et al. Left ventricular filling patterns in patients with systemic hypertension and left ventricular hypertrophy (the LIFE study). Losartan Intervention For Endpoint. Am J Cardiol 2000; 85(4): 466–472. 24 Alberti KG, Zimmet P, Shaw J. Metabolic syndrome—a new world-wide definition. A Consensus Statement from the International Diabetes Federation. Diabet Med 2006; 23(5): 469–480. 25 Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron 1976; 16(1): 31–41. 26 Dodson PM, Lip GY, Eames SM, Gibson JM, Beevers DG et al. Hypertensive retinopathy: a review of existing classification systems and a suggestion for a simplified grading system. J Hum Hypertens 1996; 10(2): 93–98. 27 Wong TY, Klein R, Klein BE, Meuer SM, Hubbard LD et al. Retinal vessel diameters and their associations with age and blood pressure. Invest Ophthalmol Vis Sci 2003; 44(11): 4644–4650.

28 Gagnon DR, Zhang TJ, Brand FN, Kannel WB. Hematocrit and the risk of cardiovascular disease–the Framingham study: a 34-year follow-up. Am Heart J 1994; 127(3): 674–682. 29 Erikssen G, Thaulow E, Sandvik L, Stormorken H, Erikssen J et al. Haematocrit: a predictor of cardiovascular mortality? J Intern Med 1993; 234(5): 493–499. 30 Hoieggen A, Fossum E, Moan A, Enger E, Kjeldsen SE et al. Whole-blood viscosity and the insulin-resistance syndrome. J Hypertens 1998; 16(2): 203–210. 31 Fossum E, Høieggen A, Moan A, Nordby G, Velund TL, Kjeldsen SE et al. Whole blood viscosity, blood pressure and cardiovascular risk factors in healthy blood donors. Blood Press 1997; 6(3): 161–165. 32 Strand AH et al. Increased hematocrit before blood pressure in men who develop hypertension over 20 years. J Am Soc Hypertens 2007; 1(6): 400–406. 33 Ikram MK, de Jong FJ, Vingerling JR, Witteman JC, Hofman A, Breteler MM et al. Are retinal arteriolar or venular diameters associated with markers for cardiovascular disorders? The Rotterdam Study. Invest Ophthalmol Vis Sci 2004; 45(7): 2129–2134. 34 Klein R, Klein BE, Moss SE, Wong TY, Hubbard L, Cruickshanks KJ et al. Retinal vascular abnormalities in persons with type 1 diabetes: the Wisconsin Epidemiologic Study of Diabetic Retinopathy: XVIII. Ophthalmology 2003; 110(11): 2118–2125. 35 Wong TY, Kamineni A, Klein R, Sharrett AR, Klein BE, Siscovick DS et al. Quantitative retinal venular caliber and risk of cardiovascular disease in older persons: the cardiovascular health study. Arch Intern Med 2006; 166(21): 2388–2394. 36 Auerbach O, Carter HW, Garfinkel L, Hammond EC. Cigarette smoking and coronary artery disease. A macroscopic and microscopic study. Chest 1976; 70(6): 697–705. 37 Black HR, Zeevi GR, Silten RM, Walker Smith GJ. Effect of heavy cigarette smoking on renal and myocardial arterioles. Nephron 1983; 34(3): 173–179.

Journal of Human Hypertension