Evaluating Pulsatile Ocular Blood FlowAnalysis in Normal and Treated

Glaucomatous Eyes

ALI AYDIN, MD, GADI WOLLSTEIN, MD, LORI LYN PRICE, MS,AND JOEL S. SCHUMAN, MD

● PURPOSE: To evaluate pulsatile ocular blood flow(POBF) analysis in normal subjects and glaucoma pa-tients by comparison of POBF measurements with func-tional (as determined by visual field [VF]) and structural(as determined by optical coherence tomography [OCT])measures.● DESIGN: Prospective, cross-sectional study.● METHODS: Forty-one eyes of 24 consecutive glaucomapatients and 20 eyes of 10 healthy subjects were studied;POBF analysis was performed on all subjects at the samevisit as VF testing and OCT retinal nerve fiber layer(NFL) thickness measurement. The mean results ofnormal and glaucomatous eyes were compared for eachmethod. Correlation between measurements obtainedwith each modality and the discriminating power usingreceiver operator characteristic curves was tested.● RESULTS: The mean POBF (standard deviation [SD])in the normal group was 1,010.4 (292.8) �l/min and989.3 (305.5) �l/min in the glaucoma group (P � .90).Significant differences between groups were found forVF mean deviation and pattern standard deviation (P �.02, P � .004, respectively) and OCT mean NFLthickness (P < .0001). No correlation was found be-tween POBF parameters and intraocular pressure, VF, orOCT variables except for intraocular pressure in glau-coma group (r � �.43, P � .003). The area under thereceiver operator characteristic curves was higher for VF

indexes and OCT mean NFL thickness than POBFparameters for distinguishing between normal and glau-comatous eyes.● CONCLUSIONS: The wide range of normal values andthe low discriminating power of POBF between normaland glaucomatous eyes limits the clinical use of thedevice for glaucoma patients. (Am J Ophthalmol 2003;136:448–453. © 2003 by Elsevier Inc. All rightsreserved.)

O CULAR BLOOD FLOW IMPAIRMENT MAY HAVE A

detrimental role in the pathophysiology of severalocular diseases including diabetic retinopathy,

age-related macular degeneration, retinitis pigmentosa,myopia, and glaucoma.1–8 Blood flow abnormalities inglaucoma have been reported using different techniques,such as fluorescein angiography,9 color Doppler imaging,8laser Doppler flowmetry,10 and pulsatile ocular blood flow(POBF) measurements.11 Although the presence of bloodflow abnormalities in glaucoma has been detected in thesestudies, it is not yet clear whether this disturbance of bloodcirculation is among the causes of the disease or a result ofthe glaucomatous process.

The POBF analysis is used for assessing ocular vascularpathologies. This noninvasive assessment of choroidalblood flow is based on the analysis of the intraocularpressure (IOP) pulse quantified by Perkins utilizing amodified applanation prism with distensible film at thecontact surface.12 Langham used the pneumotonometer toprovide a wave form of the ocular pulse, and he derivedvalues for POBF by assessing this wave form, its amplitude,and heart rate.13 The method principally measures pulsa-tile blood flow in the choroid and the anterior optic nervehead supplied by the posterior ciliary vessels, which isthought to account for 80% to 90% of total ocular bloodflow.13,14 Langham and coworkers13 suggested that POBFanalysis provided reliable and reproducible measurementsof the choroidal circulation.15

A reduced POBF has been shown in patients with bothchronic open-angle glaucoma and normal tension glau-

Accepted for publication Feb 19, 2003.InternetAdvance publication at ajo.com April 28, 2003.From the New England Eye Center, Tufts-New England Medical

Center, Tufts School of Medicine, Boston, Massachusetts (A.A., G.W.,J.S.S.); Biostatistics Research Center, Tufts-New England MedicalCenter, Boston, Massachusetts (L.L.P.).

The first two authors, A.A. and G.W., share equal part in thepreparation of this manuscript.

Supported, in part, by National Institute of Health grants RO1-EY13178, and P30-EY13078, Bethesda, Maryland, by a grant from theMassachusetts Lions Eye Research Fund Inc., and by Research to PreventBlindness. Dr. Schuman is an inventor of optical coherence tomography.

Inquiries to Joel S. Schuman, MD, New England Eye Center, Tufts-New England Medical Center, Tufts University School of Medicine, 750Washington St. Box 450, Boston, MA 02111; fax: 617-636 4866; e-mail:jss2002@attbi.com

© 2003 BY ELSEVIER INC. ALL RIGHTS RESERVED.448 0002-9394/03/$30.00doi:10.1016/S0002-9394(03)00237-X

coma compared with normal16,17 and ocular hypertensivesubjects.11 The purpose of this study was to assess the roleof POBF analysis in glaucoma practice by comparing thePOBF parameters in normal and glaucomatous eyes and toinvestigate the relationship between POBF parameters,visual field (VF) test indexes, and mean circumpapillarynerve fiber layer (NFL) thickness as measured by opticalcoherence tomography (OCT).

METHODS

SUBJECTS ENROLLED IN THE STUDY WERE NORMAL,

healthy volunteers and consecutive glaucoma patientsattending the glaucoma service at New England EyeCenter, Tufts-New England Medical Center, Boston, Mas-sachusetts. Informed consent approved by New EnglandMedical Center Human Investigation Review Committeewas signed by all subjects. This study was carried out inaccordance with the principles of the Declaration ofHelsinki.

All subjects underwent a complete ophthalmic exami-nation, including medical history, undilated and dilatedbiomicroscopy, IOP measurement using a Goldmann ap-planation tonometer (Haag-Streit, Bern, Switzerland), VF,POBF testing and OCT scanning. Subjects receiving anysystemic vasoactive medication such as beta-blockers,nitrates, or calcium channel blockers were excluded fromthe study. Eyes were excluded if they exhibited significantmedia opacity or signs of ocular pathology other than thoseattributed to glaucoma. Both eyes were included in thestudy if they were found to be eligible.

Normal subjects were defined as having IOP � 22 mmHg, normal optic nerve head appearance, normal VF, nohistory of any systemic or ocular disease other thanrefractive abnormalities, no history of glaucoma in a firstdegree relative, and no history of retinal pathology, lasertherapy, or intraocular surgery. Glaucomatous eyes weredefined as those that exhibited reproducible glaucomatousVF defects on at least two consecutive visits, as defined inthe following text, and/or IOP � 35 mm Hg despite a fullVF in the presence of marked optic nerve head cuppingand/or a definite nerve fiber layer defect on stereo biomi-croscopy.

● VISUAL FIELD TESTING: All subjects had VF testingwith either Humphrey full-threshold 24-2 achromaticperimetry or Swedish Interactive Thresholding Algorithm(SITA) standard 24-2 perimetry. A reliable VF test wasconsidered as one with less than 30% fixation losses, falsepositive or false negative responses. Glaucomatous VFdefects were defined as typical arcuate, or paracentralscotomata and/or a nasal step consisting of a cluster ofthree or more adjacent points depressed more than 5dB or

two adjacent points depressed more than 10dB. The VFmean deviation (MD) and pattern standard deviation(PSD) were used for the analysis.

● PULSATILE OCULAR BLOOD FLOW ASSESSMENT: ThePOBF assessment was performed using the Blood FlowAnalyzer (BFA) (Dicon Diagnostic Division, ParadigmMedical Industries, Inc., San Diego, California, USA).Detailed descriptions of the principles of the system havebeen previously published.13,15,18 Briefly, the pulsatile com-ponent of the choroidal circulation is calculated from theocular pulse wave produced by the bolus of blood enteringthe eye during cardiac systole. Because the eye is a closedcompartment, the volumetric blood changes are trans-ferred into a pressure gradient that is recorded by thepneumotonometer linked to the blood flow analyzer de-vice; POBF measurements using this system have beenfound to give reproducible results.19,20

A single operator (A.A.) performed all POBF measure-ments with subjects in the sitting position after topicalinstillation of proparacaine hydrochloride 0.5% drops. TheBFA automatically recorded five consecutive pulse wavesof equivalent amplitude and calculated mean IOP, pulserate, pulse amplitude, pulse volume, and POBF.

● OPTICAL COHERENCE TOMOGRAPHY SCANNING:

The principles of OCT have been described previously.21,22

Briefly, OCT is the optical equivalent of ultrasonographyfor high-resolution measurements and cross-sectional im-aging of the human retina, NFL, and the optic nerve head.Due to different optical properties of the various retinallayers, OCT is able to measure NFL thickness in anoninvasive and reproducible manner.23

All participants in this study were scanned with OCT 2(Carl Zeiss Meditech, Dublin, California, USA. Softwareversion A6). The axial resolution of the OCT 2 system wasexperimentally determined to be 14 �m in air and 10 �min the eye.21 Pupils were dilated for scanning with 1%tropicamide and 2.5% phenylephrine hydrochloride. Eacheye had three circular scans centered at the optic disk witha diameter of 3.4 mm. The average of qualified scans wasused to calculate the mean circumpapillary NFL thickness.Poor-quality OCT scans were defined as scans with signal-to- noise ratio � 50 or the presence of overt misalignmentof the surface detection algorithm of at least 15 consecu-tive pixels or 20 additive pixels.

● STATISTICAL ANALYSIS: Because many subjects hadtwo eyes participating in the study, mixed models statisti-cal analysis was used for comparisons between normallydistributed variables. The Wilcoxon rank-sum test wasused for variables that were skewed. The P value was basedon the variable averaged over eyes. Descriptive statisticsare reported on an eye level. The Pearson correlationcoefficient for normally distributed variables and theSpearman correlation coefficient for skewed variables are

POBF IN NORMAL AND GLAUCOMATOUS EYESVOL. 136, NO. 3 449

reported at the eye level. P values from mixed models wereused to assess the relationship between variables for Pear-son correlation. The Spearman correlation P value is basedon the variable averaged over eyes. The area under thereceiver operator characteristic curve (AROC) was calcu-lated for each modality to measure the capability ofdistinguishing between the groups. All analyses were per-formed using SAS software (version 8.2, SAS InstituteInc., Cary, North Carolina, USA). P values � .05 wereconsidered to be statistically significant.

RESULTS

FORTY-ONE EYES OF 24 GLAUCOMA PATIENTS AND 20 EYES

of 10 normal subjects qualified for this study. The meanages were 46.4 (standard deviation [SD], 18.2 years) in thenormal group and 70.6 years (13.2 years) in the glaucomagroup (P � .007). The normal group included five men andfive women; eight were white. The glaucoma group in-cluded 14 men and 10 women; 19 were white, two AfricanAmerican, and three were of other ethnicities. The glau-coma group included the following clinical diagnoses:primary open-angle glaucoma (32 eyes), pigmentary glau-coma (4 eyes), pseudoexfoliation glaucoma (2 eyes), andmixed mechanism glaucoma (3 eyes). The mean refractiveerror was �1.5 � 3.8 diopters in the normal group and�1.5 � 1.5 diopters in the glaucoma group. The visualacuity was 20/50 or better in all subjects. Intraocularpressures were 14.6 mm Hg (1.8 mm Hg) and 15.8 mm Hg(4.0 mm Hg) in the normal and the glaucoma groups,respectively. Age was the only demographic characteristic

that demonstrated a significant difference between groups.All glaucoma patients except one were under topicalantiglaucoma treatment; 19 of 41 eyes (46.4%) weretreated with a topical prostaglandin analog, 26 (63.4%)with beta-blockers, and 15 (36.5%) with an alpha-agonist.

Due to the large SD of POBF measurements (300�l/min), we had a power of only 9% to detect a differenceof 50 �l/min with alpha � .05. Table 1 summarizes thePOBF parameters measured by BFA, VF test indexes, andmean circumpapillary NFL thickness measured by OCT.The mean POBF (SD) was 1,010.4 (292.8) �l/min in thenormal group and 989.3 (305.5) �l/min in the glaucomagroup (P � .90); VF mean deviation (MD), patternstandard deviation (PSD), and OCT mean NFL thicknesswere significantly different between the two groups (P �.01, P � .001 and P � .0001, respectively).

The power to detect a significant correlation of at least.80 for the current study size was 95%. The results of thecorrelation between POBF parameters and IOP, age, re-fractive error, VF indexes, and mean NFL thickness arepresented in Table 2. There was no significant correlationbetween the POBF parameters and any of the othervariables except for the POBF parameter and IOP in theglaucoma group only. A significant correlation was foundin both groups between IOP as measured by applanationtonometry and average IOP, which was calculated byblood flow analyzer (BFA) (r � .88, P � .0001 forglaucoma group, and r � .74, r � .008 for normal group).There was also a significant correlation between OCTmean NFL thickness and VF indexes in glaucoma grouponly (r � .58, P � .004 for MD, r � �.53, P � .002 forPSD).

TABLE 1. Comparison of Mean (SD) Results of POBF Measurements Obtained by BFA, VFGlobal Indexes, and Mean NFL Thickness Measured by OCT Between Glaucoma and

Normal Groups

Normal Group

Glaucoma

Group P Value*

POBF parameters: 20 eyes 41 eyes

Average IOP (mm Hg) 14.0 (2.1) 13.6 (4.5) .96

Pulse amplitude (mm Hg) 3.0 (0.8) 2.7 (0.9) .42

Pulse volume (�l) 6.7 (2.1) 6.9 (2.1) .93

Pulse rate (/min) 71.4 (6.8) 68.9 (9.4) .48

POBF (�l/min) 1,010.4 (292.8) 989.3 (305.5) .90†

VF parameters 20 eyes 35 eyes

MD �1.0 (0.8) �6.4 (6.5) .02

PSD 1.4 (0.3) 5.1 (3.8) .004†

OCT parameters 20 eyes 41 eyes

NFL thickness (�m) 116.1 (7.6) 79.9 (19.7) �.0001

BFA � blood flow analyzer; IOP � intraocular pressure; MD � mean deviation; NFL � nerve fiber

layer; OCT � optical coherence tomography; POBF � pulsatile ocular blood flow; PSD � pattern

standard deviation; SD � standard deviation; VF � visual field.

*P values obtained using mixed models; †Wilcoxon rank-sum test.

AMERICAN JOURNAL OF OPHTHALMOLOGY450 SEPTEMBER 2003

The AROC was calculated for evaluating the capabilityof each modality to distinguish between groups (Table 3).The AROC was higher for VF indexes and OCT meanNFL thickness than POBF parameters.

DISCUSSION

THIS STUDY EVALUATED THE MEASUREMENTS OBTAINED

by the POBF device in normal and glaucomatous eyes. Wecompared the ocular blood flow measurements to func-tional and structural measures as determined by VF andOCT. No significant difference was found for any of thePOBF parameters between normal and glaucomatous eyes(Table 1). The AROC curves were profoundly lower forPOBF than for VF or OCT (Table 3).

The glaucoma group in this study included predomi-nantly mild-to-moderate glaucoma patients (Table 1). The

mean age of the normal group was significantly youngerthan the glaucoma group. This difference might introducea bias in favor of the discriminating ability of POBF,because a previous study found negative correlation withage.24 Because the glaucoma group in our study wassignificantly older than the normal group, it would beexpected to have better discrimination between groupsutilizing the POBF; however, no significant difference wasfound for mean POBF measurements between the groups(Table 1). Moreover, no correlation was found in our studybetween age and POBF in either of our two groups (Table2).

Previous studies have reported a wide range of POBFmeasurements both in normal and glaucomatous eyes.Langham, who had described POBF analysis, has reporteda mean POBF of 724 �l/min in the normal population.13 Instudies in which POBF analysis has been used, the range ofmean POBF in normal eyes is 428 to 1,198 �l/min,17,20

whereas, in glaucomatous eyes, is 301 to 1,164 �l/min.17,25

The range given for BFA that was used in this study was600 to 1,380 �l/min in males, 660 to 1,740 �l/min infemales, according to the user’s manual.26 The measure-ments in our study are within this range (Table 2). Thedifference between groups was nonsignificant for all POBFparameters; VF indexes and OCT mean NFL thicknesswere significantly different. Due to the large SD of thePOBF measurements, the power to detect a difference ofPOBF in the current study size is inadequate to draw adefinitive conclusion. The large variability of mean POBFmeasures has been attributed to the relationship with anumber of variables, including age, gender, heart rate,vascular status, pregnancy, body posture, and axiallength.24,27–32 Our finding is in contradiction to Kerr et al.and Trew and Smith studies,11,16,31 where the authorsfound a significant difference between POBF measure-ments of glaucoma patients and normal or ocular hyper-

TABLE 2. Correlation Between POBF Parameters and IOP, Age, VF Indexes and OCT Mean NFL Thickness Values in Glaucomaand Normal Groups

Correlated Variables

Glaucoma Group Normal Group

Correlation P Value Correlation P Value

Avg. IOP IOP (apl. t.) .88 �.0001 .74* .008

POBF IOP (apl. t.) �.43 .003 �.27* .80

POBF Age .28* .15 �.36 .24

POBF Refractive error .09 .22 .52* .07

POBF MD �.16 .31 .17* .33

POBF PSD .21* .18 �.27* .34

POBF Mean NFL .32 .21 �.009* .99

Mean NFL MD .58 .004 .43* .15

Mean NFL PSD �.53* .002 �.45* .07

apl.t. � applanation tonometry; Avg � average; IOP � intraocular pressure; MD � mean deviation; NFL � nerve fiber layer thickness; OCT

� optical coherence tomography; POBF � pulsatile ocular blood flow; PSD � pattern standard deviation; VF � visual field.

*Spearmann correlation.

TABLE 3. AROC for POBF, VF Global Indexes and OCTMean NFL Thickness for Discriminating Between Normal

and Glaucomatous Eyes

Variables AROC

POBF .49

Pulse amplitude .59

Pulse volume .54

VF-MD .88

VF-PSD .88

OCT mean NFL .99

AROC � area under the receiver operating characteristic

curve; MD � mean deviation; NFL � nerve fiber layer thickness;

OCT � optical coherence tomography; POBF � pulsatile ocular

blood flow; PSD � pattern standard deviation; VF � visual field.

POBF IN NORMAL AND GLAUCOMATOUS EYESVOL. 136, NO. 3 451

tensive subjects. Significantly lower POBF measurementswere also reported in normal tension glaucoma patientswhen compared with normal subjects either during topicaltreatment33 or 2 weeks after withdrawal of medication.17

Our glaucoma patients were all (except for one patient)under topical treatment; 19 of 41 eyes (46.4%) weretreated with a topical prostaglandin analog, 26 (63.4%)with beta-blockers, and 15 (36.5%) with an alpha-agonist.Previous studies reported that latanoprost increases ocularblood flow, and timolol has no effect upon it.25,34–37

Controversial findings were reported in regard to the effectof brimonidine and different beta-blockers.25,34,35,37,38 Thehigh POBF measurements in our glaucoma group mightalso be due to the high ratio of prostaglandin analogadministration; however, these findings highlight the lim-itations of the use of this modality in clinical practice,where a large percentage of the glaucoma patients receivechronic medical therapy.

The mean IOP by applanation tonometry in the glau-coma group was 15.8 mm Hg, and 14.6 mm Hg in thenormal group (P � .25). A negative correlation betweenPOBF and IOP was found only in the glaucoma group (r ��.43, P � .003), in agreement with the findings of Harriset al.39

The VF global indexes and NFL thickness as measuredby OCT were highly correlated in the glaucoma group. Onthe other hand, no correlation was found between anyPOBF measurement and either VF indexes or OCT meanNFL thickness (Table 2). This finding is in contrast withprevious studies where significant correlations were foundbetween ocular blood flow and visual function.11,33,40–45

Taking into account that nonsignificant difference wasfound in POBF measurements between normal and glau-comatous eyes, the low discriminating power to differen-tiate between these eyes and the poor correlation witheither VF or OCT findings, the usefulness of the POBF inthe clinical setting is questionable. This finding should beevaluated with caution due to the limited power of thisstudy. The device might be useful for evaluating theindividual vascular response to treatment; however, thisstudy did not investigate that possibility.

REFERENCES

1. Quigley M, Cohen S. A new pressure attenuation index toevaluate retinal circulation. A link to protective factors indiabetic retinopathy. Arch Ophthalmol 1999;117:84–89.

2. Ciulla TA, Harris A, Chung S, et al. Color Doppler imagingdiscloses reduced ocular blood flow velocities in non-exuda-tive age related macular degeneration. Am J Ophthalmol1999;128:75–80.

3. Friedman E. A hemodynamic model of the pathogenesis ofage-related macular degeneration. Am J Ophthalmol 1997;115:786–791.

4. Cellini M, Lodi R, Passati GL, Sbrocea M, Pelle D, GiubileiN. Color Doppler ultrasonography in retinitis pigmentosa:preliminary study. J Fr Ophtalmol 1997;20:659–663.

5. Ravalico G, Pastori G, Croce M, Toffoli G. Pulsatile ocularblood flow variation with axial length and refractive error.Ophthalmologica 1997;211:271–273.

6. Ohno-Matsui K, Morishima N, Ito M, et al. Indocyaninegreen angiography of retrobulbar vascular structures in severemyopia. Am J Ophthalmol 1997;123:494–505.

7. Chung HS, Harris A, Evans DW, et al. Vascular aspects inthe pathophysiology of glaucomatous optic neuropathy. SurvOphthalmol 1999;43:s43–50.

8. Butt Z, O’Brien C, McKillop G, Aspinall P, Allan P. ColorDoppler imaging in untreated high- and normal-pressureglaucoma. Invest Ophthalmol Vis Sci 1997;38:690–696.

9. Schwartz B, Rieser JC, Fishbein SL. Fluorescein angiographicdefects of the optic disc in glaucoma. Arch Ophthalmol1977;95:1961–1974.

10. Nicolela MT, Hnik P, Drance SM. Scanning laser Dopplerflowmeter study of retinal and optic disc blood flow inglaucomatous patients. Am J Ophthalmol 1996;122:775–783.

11. Kerr J, Nelson P, O’Brien C. A comparison of ocular bloodflow in untreated primary open-angle glaucoma and ocularhypertension. Am J Ophthalmol 1998;126:42–51.

12. Perkins ES. The ocular pulse and intraocular pressure as ascreening test for carotid artery stenosis. Br J Ophthalmol1985;69:676–680.

13. Langham ME, Farell MA, O’Brien V, Silver DM, Schilder P.Blood flow in the human eye. Acta Ophthalmol 1989;67(Suppl 191):9–12.

14. Hitchings R. The ocular pulse. Br J Ophthalmol 1991;75:65.15. Silver DM, Farrel RA, Langham ME, O’Brien V, Schilder P.

Estimation of pulsatile ocular blood flow from intraocularpressure. Acta Ophthalmol 1989;191:25–29.

16. Trew DR, Smith SE. Postural studies in pulsatile ocular bloodflow. II. Chronic open angle glaucoma. Br J Ophthalmol1991;75:71–75.

17. James CB, Smith SE. Pulsatile ocular blood flow in patientswith low tension glaucoma. Br J Ophthalmol 1991;75:466–470.

18. Krakau CE. Calculation of the pulsatile ocular blood flow.Invest Ophthalmol Vis Sci 1992;33:2754–2756.

19. Butt Z, O’Brien C. Reproducibility of ocular blood flowmeasurements. J Glaucoma 1995;4:214–8.

20. Morgan A, Hosking S. Ocular blood flow tonometer repro-ducibility: the effect of operator experience and mode ofapplication. Ophthalmic Physiol Opt 2001;21:401–406.

21. Huang D, Swanson EA, Lin CP, et al. Optical coherencetomography. Science 1991;254:1178–1181.

22. Schuman JS, Hee MR, Puliafito CA, et al. Quantification ofnerve fiber layer thickness in normal and glaucomatous eyesusing optical coherence tomography. Arch Ophthalmol1995;113:586–596.

23. Schuman JS, Pedut-Kloizman T, Hertzmark E, et al. Repro-ducibility of nerve fiber layer thickness measurements usingoptical coherence tomography. Ophthalmology 1996;103:1889–1898.

24. Ravalico G, Toffoli G, Pastori G, Groce M, Calderini S.Age-related ocular blood flow changes. Invest OphthalmolVis Sci 1996;37:2645–2650.

25. Sponsel WE, Paris G, Trigo Y, et al. Latanoprost andbrimonidine: therapeutic and physiologic assessment beforeand after oral nonsteroidal anti-inflammatory therapy. Am JOphthalmol 2002;133:11–8.

26. Banfiel R. Interpreting results. In: Banfiel R, editor. DiconDiagnostics Blood Flow Analyzer User’s Manual. California:Paradigm Medical Industries, Inc., 2001:7.1.

27. Yang YC, Hulbert MF, Batterbury M, Clearkin LG. Pulsatile

AMERICAN JOURNAL OF OPHTHALMOLOGY452 SEPTEMBER 2003

ocular blood flow measurements in healthy eyes: reproduc-ibility and reference values. J Glaucoma 1997;6:175–179.

28. Trew DR, James CB, Thomas SH, Sutton R, Smith SE.Factors influencing the ocular pulse—the heart rate. GraefesArch Clin Exp Ophthalmol 1991;229:553–556.

29. Schilder P. Ocular blood flow responses to pathology of thecarotid and cerebral circulations. Surv Ophthalmol 1994;38:S52–S58.

30. Centofanti M, Migliardi R, Bonini S, et al. Pulsatile ocularblood flow during pregnancy. Eur J Ophthalmol 2002;12:276–80.

31. Trew DR, Smith SE. Postural studies in pulsatile ocular bloodflow. I. Ocular hypertension and normotension. Br J Oph-thalmol 1991;75:66–70.

32. Mori F, Konno S, Hikichi T, Yamaguchi Y, Ishiko S, YoshidaA. Factors affecting pulsatile ocular blood flow in normalsubjects. Br J Ophthalmol 2001;85:529–530.

33. Fontana L, Poinoosawmy D, Bunce CV, O’Brien C, Hitch-ings RA. Pulsatile ocular blood flow investigation in asymet-ric normal tension glaucoma and normal subjects. Br JOphthalmol 1998;82:731–736.

34. Harris A, Spaeth GL, Sergott RC, Katz LJ, Cantor LB,Martin BJ. Retrobulbar arterial hemodynamic effects ofbetaxolol and timolol in normal-tension glaucoma. Am JOphthalmol 1995;120:168–175.

35. Morsman CD, Bosem ME, Lusky M, Weinreb RN. The effectof topical beta-adrenoceptor blocking agents on pulsatileocular blood flow. Eye 1995;9:344–347.

36. Sponsel WE, Mensah J, Kiel JW, et al. Effects of latanoprostand timolol-XE on hydrodynamics in the normal eye. Am JOphthalmol 2000;130:151–159.

37. Poinoosawmy D, Indar A, Bunce C, Garway-Heath DF,Hitchings RA. Effect of treatment by medicine or surgery onintraocular pressure and pulsatile ocular blood flow in nor-

mal-pressure glaucoma. Graefes Arch Clin Exp Ophthalmol2002;240:721–726.

38. Vetrungo M, Maino A, Cantatore F, Ruggeri G, Cardia L.Acute and chronic effects of brimonidine 0.2% on intraoc-ular pressure and pulsatile ocular blood flow in patients withprimary open angle glaucoma: an open-label, uncontrolled,prospective study. Clin Ther 2001;23:1519–1528.

39. Harris A, Joos K, Kay M, et al. Acute IOP elevation withscleral suction: effects on retrobulbar haemodynamics. Br JOphthalmol 1996;80:1055–1059.

40. Sponsel WE, DePaul KL, Kaufman PL. Correlation of visualfunction and retinal leukocyte velocity in glaucoma. Am JOphthalmol 1990;109:49–54.

41. Michelson G, Groh MJ, Groh ME, Grundler A. Advancedprimary open-angle glaucoma is associated with decreasedophthalmic artery blood-flow velocity. Br J Ophthalmol1994;78:818–822.

42. Nicolela MT, Walman BE, Buckley AR, Drance SM. Ocularhypertension and primary open-angle glaucoma: a compara-tive study of their retrobulbar blood flow velocity. J Glau-coma 1996;5:308–310.

43. Sponsel WE, Kaufman PL, Blumm FG. Association of retinalcapillary perfusion with visual status during chronic glau-coma therapy. Ophthalmology 1997;104:1026–1032.

44. Liu CJ, Chiou HJ, Chiang SC, Chou JC, Chou YH, Liu JH.Variations in ocular hemodynamics in patients with earlyand late glaucoma. Acta Ophthalmol Scand 1999;77:658–662.

45. Kondo Y, Niwa Y, Yamamoto T, Sawada A, Harris A,Kitazawa Y. Retrobulbar hemodynamics in normal-tensionglaucoma with asymmetric visual field change and asymmet-ric ocular perfusion pressure. Am J Ophthalmol 2000;130:454–560.

POBF IN NORMAL AND GLAUCOMATOUS EYESVOL. 136, NO. 3 453