Abstract
purpose. The present study investigated whether age, blood pressure (BP), and baseline vessel diameter influence the retinal arterial response to flicker light.
method. Thirty healthy subjects (mean age, 46.3; range, 22–73 years) and 15 patients with untreated essential arterial hypertension (mean, 50.9; range, 26–69 years) were examined. The diameter of the retinal arterioles was measured by a Retinal Vessel Analyzer (RVA; Imedos, Weimar, Germany). Each examination consisted of a 100-second baseline measurement and five 20-second periods of flicker stimulation, followed by an 80-second observation period. The five stimulation periods were then averaged. The rectangular luminance flicker operated at 12.5 Hz at a wavelength of 530 to 600 nm. The baseline-corrected flicker response (bFR) was defined as the difference between the peak dilatation and subsequent constriction after flicker stimulation minus the fluctuation of the baseline. The BP was measured at 1-minute intervals during the examination.
results. In 26 subjects with normal BP, flicker light induced a bFR of +6.4% ± 2.7%. The bFR decreased nonsignificantly in healthy subjects with increasing age (y = 8.48 − 0.048x; r = 0.26). The baseline diameter did not influence the amplitude of the flicker response over a range of 70 to 140 measuring units. The hypertensive patients reacted with a bFR of +2.2% ± 2.5% (P < 0.001). Four hitherto healthy subjects with elevated BP during the examination were excluded from analysis.
conclusions. A significant correlation of age and bFR was not found in the small sample examined. Untreated arterial hypertension appeared to be associated with a reduced flicker response. The value of such functional vessel properties in the screening of vasosclerosis and in diagnostics in arterial hypertension should be examined in further studies.
Screening and early diagnosis of sclerotic and hypertensive vessel changes are gaining increasing importance in industrialized countries because of lifestyle factors and increasing life expectancy.
1 The ophthalmoscopic assessment of retinal vessels has long been a routine element of diagnostic testing and follow-up for arterial hypertension. Today’s clinical routine examination consists of observation of the fundus and a qualitative classification
2 of vessel features or parenchymal changes. The diameter of a vessel (i.e., the red blood cell column formed in the vascular lumen without administration of a dye) is an important quantitative parameter for describing vessels. A correlation of arterial retinal diameter
3 4 or the arteriole-to-venule ratio
4 5 with systemic blood pressure (BP) was found in extensive epidemiologic studies. In these studies the vessel diameter was measured semiautomatically in single fundus images. For some time, automated, objective, and continuous recording of the diameter of a vessel segment has been possible with the Retinal Vessel Analyzer (RVA; Imedos, Weimar, Germany).
6 7 8 Thus, in addition to the aforementioned parameters, local (segmental) fluctuations of the vessel diameter and small periodic changes (e.g., pulse waves) have been observed together for the first time.
6
However, the parameters of the uninfluenced vessel diameter exhibit significant interindividual scatter. They are therefore suitable for comparative observations in groups
9 10 but not as an individual diagnostic criterion. In the presence of intact autoregulation, the retinal microcirculation attempts to compensate for disturbances and to ensure a blood supply commensurate with retinal requirements. Thus, artificial disturbances (e.g., provocations) provide a means for the functional diagnosis of retinal vessels. Slow vessel diameter reactions caused by changes in the breathing air or increasing BP can be measured by single fundus images.
11 12 13 14 Moreover, the influence of the cardiac cycle on the retinal vessel diameter is well known
15 and can be equalized in single fundus images only by electrocardiogram (ECG) triggering or by means of a larger number of images randomized during the heart cycle. The RVA solves this problem by providing continuous, automated vessel analysis combined with an on-line correction of small eye movements. Provocation methods that have so far been used in conjunction with retinal vessel analysis include BP increase due to muscular exertion,
7 8 15 drug-induced reduction of BP,
7 oxygen inhalation,
13 16 17 18 carbogen or CO
2 inhalation,
6 and elevation of intraocular pressure with a suction cup.
19 20 Light and periodic changes of its color and intensity are original stimuli to the eye. Flicker light stimulation has a number of advantages over other methods—for example, direct interference with the target organ (the retina), standardization of the stimulus intensity, simple execution and a high level of acceptance. Recently several ocular perfusion parameters have been investigated by using flicker provocation.
21 22 23 24 25 26 27 28 This includes assessing the retinal vessel diameter response to flicker light in single images
21 and by means of the RVA.
22 The flicker response is very fast, especially at the beginning and after the end of the flicker period.
22 Hence, measurement with the retinal vessel analyzer appears more favorable than with single images.
29
Knowledge is required of as many factors as possible that affect vessel response, to measure them in clinical studies or diagnostic tests or to keep them constant. Age-dependency of the arterial response to oxygen inhalation has already been demonstrated
18 and is also expected in connection with the flicker response. BP changes themselves induce vessel diameter changes,
3 4 5 and it is therefore very likely that systemic BP influences other provoked vessel responses. Earlier investigations suggest a more marked diameter response in small vessels than in large vessels.
30 Polak et al.,
22 by contrast, found no differences in the flicker response in peripapillary retinal arteries before and after the first bifurcation.
The present study seeks to clarify the following questions regarding the retinal artery flicker response:
-
Does an age dependency exist?
-
Do large arteries respond differently than small arteries?
-
Is a different response evident in subjects with elevated BP or known hypertension?
In a clinical study, one eye of each of 45 subjects was examined. Thirty healthy subjects (mean age, 46.3 years; range, 22–73; 17 women) had no systemic disease and took no medications (exception: hormonal contraception in some cases). Fifteen patients (mean age, 50.9 years; range, 26–69; eight women) were referred by their general practitioners for fundus assessment due to the initial diagnosis of untreated essential arterial hypertension. Exclusion criteria were clouding of the refractory media, visual acuity less than 0.5, astigmatism more than 2.0 D, myopia more than 7.0 D, ocular disease, history of an ocular operation or injury in the examination eye, wearing of contact lenses within the previous 24 hours, acute infection, known diabetes mellitus, pregnancy, and nursing.
At the start of the study a clinical ophthalmic examination was performed (measurement of visual acuity, objective refraction, slit lamp microscopy, applanation tonometry, funduscopy), after which the pupil was dilated with tropicamide. The preparation for the examination took at least 30 minutes, allowing enough time for the volunteer to acclimatize and for the BP to normalize. The vessel diameter was measured on-line with an RVA. Details of this device and its process of vessel diameter measurement have been described elsewhere.
6 7 8 9 An optoelectronic shutter was inserted in the retina camera in place of an additional optical filter. The shutter interrupted the observation light (530–600 nm, irradiance at the fundus approximately 1.96 × 10
−4 W/cm
2; measured by the Institute of Biomedical Technique, Technical University Ilmenau, see RC_illumination.pdf at http://www.iovs.org/cgi/content/full/45/5/1486/DC1) over the entire 30° visual field of the retinal camera and worked with a bright-to-dark ratio of 25:1. The chosen frequency of 12.5 Hz of rectangular light interruption provided a sequence of one normal illuminated and one dark single frame at a video frequency of 25 Hz
1 . This frequency lies in the range of the maximally exciting flicker frequency.
22 23 The optoelectronic shutter was controlled by a special program running on the RVA computer.
Measurement of the baseline vessel diameter for 100 seconds in continuous light was followed by five cycles of 20-second flicker provocation and 80-second observation.
An arterial segment of approximately 1.5 mm in length was evaluated in each eye. Selection criteria for the segment were location within a circular area of two disc diameters, no crossing or bifurcation in the measuring segment, curvature of not more than 30°, a distance to neighboring vessels of at least one vessel diameter and sufficient contrast to the surrounding fundus. The selected vessel segment was placed in the middle part of the fundus image by eye movement, by using the inner fixation of the camera before the vessel measurement began
2 . The vessel segment was scanned 25 times a second in the measurement window under optimum conditions. The position of the vessel edges, the vessel course, the vessel diameter, and correction for ocular movements were calculated automatically on-line. Because the image scale of each eye was unknown, the measured values were expressed in relative units (RU). These units correspond to micrometers if the examination eye has the dimensions of the normal Gullstrand eye. The enormous amount of diameter values for all sections of the vessel segment were averaged to the mean diameter of the vessel segment during 1 second. A theoretical available account of 600 diameter values during 10 minutes was not possible in practice, owing to faults caused by blinks, extensive eye movements, or other events. The period from −30 to −5 seconds before every flicker provocation was taken as the baseline, to which the subsequent diameter response was normalized. From each baseline-related curve (%), the baseline-corrected flicker response (bFR) was calculated as the difference between the peak dilatation after provocation (dil%) and the minimum of the subsequent reactive constriction (constr%) and the width of the baseline amplitude (width
BL%;
3 ).
\[\mathrm{bFR}\%\ {=}\ \mathrm{dil}\%\ {-}\ \mathrm{constr}\%\ {-}\ \mathrm{width}_{\mathrm{BL}}\%\]
A comparison of the diameter response of small and large vessels requires a higher resolution because an image of the vessel cross section over at least six pixels of the CCD camera matrix is needed to calculate the diameter. This was achieved by using the 20° angle of the retina camera together with the same CCD matrix. It is typical of this camera (FF 450plus; Carl Zeiss Meditec, Jena, Germany) that a change in angle from 30° to 20° modifies the observation but not the illumination of the fundus. This fact (1) produces an unchanged illumination area on the fundus in the 20° and 30° modus (i.e., the same flicker area); and (2) requires a higher intensity of light in the 20° mode to sufficiently illuminate each CCD pixel by the spread reflected fundus light.
Six eyes of six volunteers were again examined on another day. The examination protocol was the same as previously described except for the observation angle and the intensity of the illumination light (3.08 × 10−4 W/cm2, see RC_illumination.pdf at http://www.iovs.org/cgi/content/full/45/5/1486/DC1). The same large arterial segment was placed by eye movement by the inner fixation of the camera at an equivalent image position and was measured on-line. A smaller arterial segment after the second bifurcation was selected according the aforementioned criteria and was analyzed off-line from the videotape. A magnification factor was inserted in the software to compensate for the higher resolution. This equipment (RVA; Imedos) can measure vessel diameters more than 55 μm. The 20° flicker response is not immediately comparable to the 30° response because of the higher light intensity during the observation and flicker phases.
Automatic BP measurement was obtained with the an intensive care monitor (Cardiocup II; Datex Ohmeda, Louisville, CO), and the retina camera was adjusted. The BP measurement took nearly 33 seconds and was repeated at 1-minute intervals. The mean arterial systemic BP was derived from the systolic and diastolic BP measured at 1-minute intervals during the examination [BPmean = BPdiast + ([frax1;3]BPsys − BPdiast]).
The Mann-Whitney test, the Wilcoxon test, and regression analysis (Excel 2000, WinStat; Microsoft Corp., Redmond, WA) were applied for statistical analysis. The study was performed in accordance with the guidelines set forth in the Declaration of Helsinki and was approved by the Ethics Committee of the Thuringia State Medical Board. All the participants had given their informed consent in writing.
In the present study, flicker light generated by periodic interruption of the fundus illumination light of the RVA was a suitable stimulus for provoking an arterial diameter response. This involves only minor manipulation of the illumination system and produces a luminance flicker in the green wavelength range with a contrast of 25:1. Every second video frame during the flicker period is dark, and no vessel diameter or eye movement can be detected. The local changes in the next normal illuminated frame are greater than in a continuous video sequence. This explains cases with some missing data during the flicker period. This problem is augmented in the 20° measurement because of the magnification, not only of the fundus image but also of the eye movements. The higher magnification is useful for detailed examination but appears unsuitable for the clinical routine.
It appears expedient to include the entire individual response curve in the qualitative analysis. A similar approach has been adopted in electrophysiology, where the measurement curve must be observed to validate the ERG amplitude. The constriction response to values below the baseline after flicker-induced dilatation has been observed in other studies
22 29 and results from an overshooting regulatory response. No defined relationship is evident in our material between dilatation and constriction (i.e., slight dilatation is not always associated with slight constriction). The amplitude of the arterial flicker response therefore appears to be a better parameter for describing the reaction than the maximum dilation. Large fluctuations of the baseline vessel diameter are evidence of a spontaneous vessel diameter change that is not flicker induced and/or of poor patient cooperation. These variations are superimposed in some circumstances on the flicker response. The evaluation of flicker-induced dilation with respect to the baseline of the same measurement is a possible way to approximate original flicker reaction. Taking the difference of the flicker response and the baseline fluctuation as the “baseline-corrected flicker response” can result in (1) a bFR greater than, equal to, or smaller than the dilative maximum; (2) a bFR of 0 or less; the flicker response is equal to or smaller than the baseline fluctuations. This can then be interpreted as an undetectable flicker-induced diameter change.
In our study, we also tried a measuring protocol calling for a number of sequential provocational cycles to average the vessel response. This method is well known to yield stable results. Repetition of a measuring cycle in a clinically practical manner leads to a shortened provocational period. The flicker response is unusually fast, and after 20 seconds the arterial diameter dilates only slightly.
22 It is for this reason that we selected a flicker time of 20 seconds and five repetitions. The resultant measuring time of 10 minutes was as well tolerated by the volunteers.
The age versus bFR scattergram
5 shows a slightly decreasing flicker response and increasing dispersion of the measured values in subjects of middle to advanced age. The small coefficient of correlation (
r = 0.26) indicates a weak correlation between the two parameters. The decrease in bFR of approximately 0.5% per decade of life for the analyzed age range of 20 to 70 years is not significant in the present data. This result is surprising, because a decrease of vessel reactivity with increasing age was expected. Reasons for this could be the small sample size and the tendency toward greater dispersion in the age range above 40 years. This gives rise to the following hypotheses:
-
A confirmation of the independence of age and flicker response in extended clinical studies could simplify the judgment of the flicker response as “normal” or “abnormal.”
-
Responses with a bFR of less than 2% in the group of healthy subjects provide additional diagnostic information, possibly relating to unknown or preclinical vessel pathologies or cardiovascular disease. These can be clarified by means of extended diagnostic or clinical observation. These cases then have to be separated from the group of healthy subjects. The result would be a mathematically smaller reduction of the bFR with increasing age and a stronger coefficient of correlation.
No correlation between the flicker response and the baseline arterial diameter could be found over a range of 70 RU or more (equal to micrometers in the normal Gullstrand eye). This contradicts former opinions
30 and model microcirculation calculations (Berthold K, et al.
IOVS 2002;43:ARVO E-Abstract 229), according to which smaller vessels undergo larger diameter changes. It is conceivable that these model expectations apply only to smaller arterioles of 20 to 30 μm in diameter, which probably serve the hemodynamic function of precapillary sphincters in the retina and may respond differently. However, our measurements as well as those by Polak et al.
22 rule out the baseline diameter as a factor influencing the flicker response of large retinal arteries.
Four volunteers with hypertension were excluded from further examinations, because it was not clear whether they had undiagnosed hypertension or the elevated BP was caused by experimental stress. We advised them to contact a general practitioner for circulatory diagnostics. The flicker response was significantly reduced in untreated hypertensive patients in comparison to healthy subjects. The difference of the bFR is notable and, even in the small sample of eyes examined in our study, is greater than a single standard deviation. The range in which the bFR of healthy and hypertensive subjects overlap may be interesting for the vascular diagnosis of patients with known hypertension. In cases of hypertension but where the bFR is still intact, there may be no permanent vascular wall changes. In contrast, a small flicker response could indicate vascular risks for patients with normal or low BP. It is currently unclear what causes the diminished flicker response. Possible explanations are endothelial cell damage or sclerotic vascular wall changes that render a diameter change impossible. Normalization of the flicker reaction under antihypertensive therapy may indicate improved microvascular function. The suitability of the flicker-induced diameter response to demonstrate effects of vasoactive substances has already been demonstrated for dopamine
28 and sildenafil.
31 The flicker-induced retinal vessel response may be valuable as a method of investigating the effects of vasoactive drugs on the arteries of the central microcirculatory system in the diameter range of approximately 100 μm.
To date, no clinical tests have been established for examining functional properties of the arterioles of the central microcirculation. The flicker response of retinal vessel diameter could be a new and noninvasive method for detecting early functional damage. The flicker-induced diameter response of retinal arterioles should be investigated in further clinical studies for its suitability as a screening method for vascular diseases, as a further vascular diagnostic test for hypertensive patients, and as a parameter for therapy monitoring.
Supported by German Federation of Education and Research Grant 13N7999.
Submitted for publication June 29, 2003; revised October 20 and November 24, 2003; accepted December 4, 2003.
Disclosure:
E. Nagel, Imedos (C);
W. Vilser, Imedos (I, E, P);
I. Lanzl, None
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “
advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Corresponding author: Edgar Nagel, Consultant Ophthalmologist in Practice, Anton-Sommer-Strasse 55, D-07407 Rudolstadt, Germany;
e_a_nagel@t-online.de.
Table 1. Characteristics of the Study Groups
Table 1. Characteristics of the Study Groups
Table 2. Parameters of the Flicker Response
Table 2. Parameters of the Flicker Response
Table 3. Comparison of the bFR in Healthy Subjects and Hypertensive Patients Aged ≥ 40 Years or More
Table 3. Comparison of the bFR in Healthy Subjects and Hypertensive Patients Aged ≥ 40 Years or More
Wolf-Maier K, Cooper RS, Banegas JR, et al. Hypertension prevalence and blood pressure levels in 6 European countries, Canada and the United States. JAMA. 2003;289:2363–2369.
Naumann GOH. Spezielle Pathologie der Retina. Gefässerkrankungen der Retina. Naumann GOH eds. Pathologie des Auges. 1997;2:1000. Springer Auflage. Berlin.
Wong TY, Hubbard LD, Klein R, et al. Retinal microvascular abnormalities and blood pressure in older people: the Cardiovascular Health Study. Br J Ophthalmol. 2002;86:10007–10013.
Leung H, Wang JJ, Rochtchina E, et al. Relationship between age, blood pressure, and retinal vessel diameters in an older population. Invest Ophthalmol Vis Sci. 2003;44:2900–2904.
Harrett AR, Hubbard LD, Cooper LS, et al. Retinal arteriolar diameter and elevated blood pressure: the Arteriosclerosis Risk in Communities Study. Am J Epidemiol. ;150:263–270.19999
Vilser W, Nagel E, Fuhrmann G, Riemer T. Retinale Gefässanalyse: Neue Möglichkeiten zur Untersuchungen von Netzhautgefässen. Schmidt KG Pillunat LE eds. Fortbildung Glaukom. 2000;73–91. Enke Verlag Stuttgart, Germany. Funktionsdiagnostik und pathogenetische Konzepte; bd. 3.
Polak K, Dorner G, Kiss B, et al. Evaluation of the Zeiss retinal vessel analyser. Br J Ophthalmol. 2000;84:1285–1290.
Blum M, Bachmann K, Wintzer D, Riemer T, Vilser W, Strobel J. Noninvasive measurement of the Bayliss effect in retinal autoregulation. Graefes Arch Clin Exp Ophthalmol. 1999;237:296–300.
Blum M, Bachmann K, Pietscher S, Bräuer-Burchhardt C, Vilser W, Strobel J. Online-Messung retinaler Arterienäste bei Typ-2-Diabetikern: Erste klinische Versuche vor und nach fokaler Laserkoagulation. Ophthalmologe. 1997;94:724–727.
Nagel E, Vilser W, Fuhrmann G, Lanzl IM, Lang GE. Influence of topical dorzolamide on retinal branch vessel diameter. Ophthalmology. 2001;21:05.http://www.onjoph.com/Artikel/nagel_body.htm
Tachibana H, Gotoh F, Ishikawa Y. Retinal autoregulation in normal subjects. Stroke. 1982;13:149–155.
Rassam SM, Patel V, Kohner EM. The effect of experimental hypertension on retinal vascular autoregulation in humans: a mechanism for the progression of diabetic retinopathy. Exp Physiol. 1995;80:53–68.
Grunwald JE, Riva CE, Petrig BL, Sinclair SH, Brucker AJ. Effect of pure O2-breathing on retinal blood flow in normals and in patients with background diabetic retinopathy. Curr Eye Res. 1984;3:239–241.
Brinchmann-Hansen O, Myhre K. Vascular response of retinal arteries and veins to acute hypoxia of 8,000, 10,000, 12,500, and 15,000 feet of simulated altitude. Aviat Space Environ Med. 1990;61:112–116.
Chen HC, Patel V, Wiek J, et al. Vessel diameter changes during the cardiac cycle. Eye. 1994;8:97–103.
Blum M, Kubetschka U, Hunger-Dathe W, Bachmann K, Müller UA, Strobel J. Autoregulation retinaler Arteriolen bei Patienten mit Diabetes mellitus und Normalprobanden. Klin Monatsbl Augenheilkd. 2000;216:40–44.
Lanzl IM, Witta B, Kotliar K, Vilser W. Reaktion retinaler Gefäßdurchmesser auf 100% O2-Atmung: funktionelle Messung mit dem Retinal Vessel Analyzer an 10 Probanden. Klin Monatsbl Augenheilkd. 2000;217:231–235.
Blum M, Scherf C, Bachmann K, Strobel J. Alterskorrelierte Kontraktilität retinaler Arteriolen bei Sauerstoffatmung. Ophthalmolge. 2001;98:265–268.
Nagel E, Vilser D, Fuhrmann G, Vilser W, Lang GE. Dilatation grosser Netzhautgefässe nach Intraokulardrucksteigerung. Ophthalmologe. 2000;97:742–746.
Nagel E, Vilser W, Lanzl I. Functional analysis of retinal vessel diameter reaction to artificially raised intraocular pressure in glaucoma patients with and without dorzolamide therapy. Vasa. 2002;31:230–234.
Formaz F, Riva CE, Geiser M. Diffuse luminance flicker increases retinal vessel diameter in humans. Curr Eye Res. 1997;16:1252–1257.
Polak K, Schmetterer L, Riva Ce. Influence of flicker frequency on flicker-induced changes of retinal vessel diameter. Invest Ophthalmol Vis Sci. 2002;43:2721–2726.
Riva CE, Falsini B, Logean E. Flicker-evoked response of human optic nerve head blood flow: luminance versus chromatic modulation. Invest Ophthalmol Vis Sci. 2001;42:756–762.
Falsini B, Riva CE, Logean E. Flicker-evoked changes in human optic nerve blood flow: relationship with retinal neural activity. Invest Ophthalmol Vis Sci. 2002;43:2309–2316.
Falsini B, Riva CE, Logean E. Relationship of blood flow changes of the human optic nerve with neural retinal activity: a new approach to the study of neuro-ophthalmic disorders. Klin Monatsbl Augenheilkd. 2992;219:296–298.
Michelson G, Patzelt A, Harazny J. Flickering light increases retinal blood flow. Retina. 2002;22:336–343.
Garhofer G, Huemer KH, Zawinka C, Schmetterer L, Dorner GT. Influence of diffuse luminance flicker on choroidal and optic nerve head blood flow. Curr Eye Res. 2002;24:109–113.
Huemer KH, Garhofer G, Zawinka C, et al. Effects of dopamine on human retinal vessel diameter and its modulation during flicker stimulation. Am J Physiol. 2003;284:H358–H363.
Nagel E, Vilser W, Lanzl I. Flicker observation light induces diameter response in retinal arterioles. a clinical methodological study. Br J Ophthalmol. 2004;88:54–56.
Dollery CT. Physiological principles. Wise GN Dollery CT Henkind P eds. The Retinal Circulation. 1971;102–105. Harper & Row New York.
Polak K, Wimpissinger B, Berisha F, Georgopoulos M, Schmetterer L. Effects of sildenafil on retinal blood flow and flicker-induced retinal vasodilatation in healthy subjects. Invest Ophthalmol Vis Sci. 2003;44:4872–4876.