The Effect of Ocular Warming on Ocular Circulation in Healthy Humans | Medical Devices and Equipment | JAMA Ophthalmology | JAMA Network
[Skip to Navigation]
Sign In
Figure 1. 
The change in retinal circulationinduced by ocular warming in the retinal artery (A) and vein (B). All dataare expressed as a percentage of the prewarming level. Symbols and error barsrepresent the mean ± SE. Asterisks indicate P<.05compared with prewarming values by repeated-measures analysis of variancefollowed by Dunnett procedure.

The change in retinal circulationinduced by ocular warming in the retinal artery (A) and vein (B). All dataare expressed as a percentage of the prewarming level. Symbols and error barsrepresent the mean ± SE. Asterisks indicate P<.05compared with prewarming values by repeated-measures analysis of variancefollowed by Dunnett procedure.

Figure 2. 
The change in choroidal circulationin the foveal region induced by ocular warming. All data are expressed asa percentage of the prewarming level. Asterisks indicate P<.05 compared with prewarming values by repeated-measures analysisof variance followed by Dunnett procedure.

The change in choroidal circulationin the foveal region induced by ocular warming. All data are expressed asa percentage of the prewarming level. Asterisks indicate P<.05 compared with prewarming values by repeated-measures analysisof variance followed by Dunnett procedure.

Table 1. 
Systemic and Ocular Variables at Baseline*
Systemic and Ocular Variables at Baseline*
Table 2. 
Retinal and Choroidal Circulation at Baseline*
Retinal and Choroidal Circulation at Baseline*
Table 3. 
Changes in Ocular Temperature*
Changes in Ocular Temperature*
Clinical Sciences
October 2004

The Effect of Ocular Warming on Ocular Circulation in Healthy Humans

Author Affiliations

From the Department of Ophthalmology, Asahikawa Medical College, Asahikawa,Japan. The authors have no relevant financial interest in this article.

Arch Ophthalmol. 2004;122(10):1477-1481. doi:10.1001/archopht.122.10.1477

Objective  To examine the effect of ocular warming on retinal blood flow (RBF)and subfoveal choroidal blood flow (CBF) in humans.

Methods  Ocular warming was induced in 10 healthy volunteers using an ocularwarming lamp for 10 minutes. The ocular surface temperature was measured beforeand after warming. The RBF in the retinal artery and vein and the CBF in thefoveal region were examined with a retinal laser Doppler velocimetry systemand a laser Doppler flowmeter, respectively. Ocular blood flow measurementswere performed before and 3, 6, and 9 minutes after warming.

Results  The ocular surface temperature significantly increased just after warmingand returned to baseline 10 minutes later. Three minutes after warming, themean ± SE RBF significantly increased in the retinal artery (14.2%± 3.5%, P = .01) and vein (15.8% ±3.6%, P = .006). Six minutes after warming, the RBFreturned to baseline in the artery and vein. Three and 6 minutes after warming,the mean ± SE CBF significantly decreased 16.6% ± 4.2% and 24.2%± 4.7%, respectively (P = .001 for both).Nine minutes after warming, the measurements returned to baseline.

Conclusions  The RBF increased and the CBF decreased in the foveal region after cessationof ocular warming in healthy young volunteers. The CBF in the foveal regionmay contribute to maintaining a constant retinal temperature in response toocular warming.

Some studies1-3 recentlyhave reported that patients with age-related macular degeneration (AMD) havechoroidal perfusion anomalies. In addition, a recent study4 usingcolor Doppler imaging reported that transpupillary thermotherapy, which hasbeen used successfully to treat occult subfoveal choroidal neovascular membranesin patients with AMD, could cause alterations in the choroidal circulation.Although those authors speculated that the decrease in choroidal blood flow(CBF) may be caused by occlusion of the choroidal neovascular membrane orthe choriocapillaris, leading to increased vascular resistance, the exactmechanism of the alteration of choroidal circulation induced by transpupillarythermotherapy is unclear. One possibility is that the increased ocular tissuetemperature may be associated with alteration of the choroidal circulation.When the effect of transpupillary thermotherapy on ocular blood flow in patientswith AMD is considered, it is also necessary to consider how ocular bloodflow changes in response to the rise in ocular tissue temperature.

Few studies have examined the relation between ocular temperature andocular blood flow. In previous investigations of the relation between choroidalcirculation and retinal temperature in monkeys5-7 andhumans,8 Parver et al suggested that the choroidalcirculation maintains a stable temperature in the outer retinal layer. However,they measured the ocular tissue temperature as an index of ocular blood flow.Therefore, it remains to be examined how the ocular blood flow changes inresponse to ocular warming in humans. We investigated the physiologic responseof ocular circulation to ocular warming in healthy humans. For this purpose,we used a laser Doppler velocimetry system and laser Doppler flowmetry toevaluate the retinal blood flow (RBF) and CBF, respectively.


Measurements were obtained from 10 eyes of 10 healthy young male volunteers(mean ± SE age, 27.3 ± 1.4 years; range, 21-36 years). The procedurefollowed the tenets of the Declaration of Helsinki. Written informed consentwas obtained from all subjects after the study was fully explained. All subjectshad corrected visual acuity better than 20/20, clear media, and no historyof ocular or systemic disease or therapy. The temperature in the examinationroom was maintained within a constant range from 22°C to 24°C. Thesubjects abstained from drinking coffee and smoking for at least 2 hours beforethe test. To prepare for the test, each subject rested for 10 to 15 minutesin a quiet room before the test.

Measurement of ocular surface temperature

We measured the temperature changes of the ocular surface using noncontactinfrared radiation thermography (LAIRD-S270ME; Nikon, Tokyo, Japan), whichis sensitive to temperatures between −40°C and 160°C, with aresolution of 0.1°C. This technique has been described elsewhere usinga prototype device.9,10 Briefly,the temperature data were rapidly transformed into a color-coded image thatwas displayed on the monitor. A region of interest was electrically outlinedand the mean temperature computed, as described previously.9 Inthe present study, we defined the mean value of the bilateral scleral regionas the ocular temperature.

Rbf measurement

In the present study, a retinal laser Doppler velocimetry system (CanonLaser Blood Flowmeter, model CLBF 100; Canon, Tokyo) was used to estimatethe RBF. This device has recently been described.11,12 Briefly,it allows noninvasive measurement of the absolute values of the red bloodcells flowing in the center line of the vessel, based on the bidirectionallaser Doppler velocimetry.13 This device alsocontains a vessel diameter measurement system and a vessel tracking system.Laser Doppler measurements were obtained from a temporal retinal artery andadjacent vein in 1 eye of each subject. The arteries chosen for measurementhad relatively straight segments that were sufficiently distant from adjacentvessels. Measurement sites were generally between the disc margin and thefirst bifurcation in the supratemporal retinal vessels. The location of themeasurement site was recorded on a color fundus photograph. As previouslydescribed, the RBF in the retinal artery was calculated as RBF = V(area/2),where V is the mean of the center line blood speed during the cardiac cycle,and area is the cross-sectional area of the retinal artery at the laser Dopplermeasurement site.14 The area was calculatedfrom the arterial diameter, assuming a circular cross section. The factorof 2 in the formula for the blood flow arises from the assumption of Poiseuilleflow.15

Measurement of cbf in the foveal region

Determinations of relative foveolar choroidal blood velocity, bloodvolume, and CBF were obtained using a method based on laser Doppler flowmetrytechnique.16 Detailed descriptions of the methodhave been published previously.17 Velocityis expressed in hertz, and volume and flux are expressed in arbitrary units.In a previous study,18 the mean coefficientsof variability of choroidal blood volume, choroidal blood velocity, and CBFin 5 healthy subjects were 12.7%, 10.0%, and 6.8%, respectively. Subjectsfixated on the probing laser beam to determine the foveolar CBF. During bloodflow measurements, proper fixation was ascertained by direct observationsof the foveola through the fundus camera. All measurements were performedwith the subjects seated in a dark room.

The choroidal circulation was measured continuously in each participantfor about 30 seconds. Measurements were performed twice in each subject. Allflow variables were then averaged over 2 periods of 30 seconds each. Beforethe data were analyzed, spikes due to micromovements and blinks were removed,as described previously.19 Data analysis wasperformed by a masked observer using a computer (NeXT Computer, Redwood City,Calif) with software specifically developed to analyze Doppler signals fromocular tissues.20

Study protocol

A traditional ocular warming lamp (Hand Ophlar; Handaya Co, Tokyo) wasused to induce ocular warming. This instrument, which is held in front ofthe eye, heats the eyelid and ocular tissue with an electric bulb. This devicehas long been used to treat hordeolum, meibomian gland obstruction, and acuteconjunctivitis in Asia.10 The participants'eyes were warmed for 10 minutes through closed eyelids. The study eyes werechosen randomly. Care was taken not to press the eye during warming. The oculartemperature was measured before warming, immediately after, and 10 minutesafter cessation of the warming. After ocular warming, the eyes were examinedat the slitlamp, and the mean arterial blood pressure and heart rate wereestimated by electronic sphygmomanometer (model EP-88Si; Colin, Tokyo) atthe same time the ocular temperature was measured.

The procedure was performed in a masked fashion. The ocular circulationwas evaluated before ocular warming and 3, 6, and 9 minutes after cessationof ocular warming. Before ocular warming, 5 measurements of each variablewere obtained every minute, and the mean value was defined as the baselinevalue. Measurements of variables were performed for 30 seconds before eachtime point after cessation of warming. Ocular warming was performed twicefor all subjects on 2 separate days, because the RBF and CBF were measuredon another day in the same manner and at the same time of day. The measurementof RBF or CBF was performed randomly in each subject.

Statistical analysis

All values are expressed as mean ± SE. For statistical analysis,we used the repeated measures analysis of variance followed by post hoc comparisonwith Dunnett procedure. P<.05 was considered statisticallysignificant.

Systemic and ocular variables at baseline

Systemic circulatory variables and intraocular pressure (IOP) on thefirst and second trial days at baseline are shown in Table 1 and Table 2.There were no differences between the 2 trial days in any variables.

Changes in ocular surface temperature

Just after ocular warming, the ocular temperature significantly increasedfrom 34.5°C ± 0.2°C to 37.8°C ± 0.3°C in thefirst trial and from 34.6°C ± 0.3°C to 37.9°C ±0.2°C in the second trial (P<.001) (Table 3). The ocular temperature then returnedto the prewarming level 10 minutes after cessation of warming. In addition,no significant changes in systolic blood pressure, diastolic blood pressure,mean arterial blood pressure, and heart rate were induced by ocular warming.

Changes in rbf

Three minutes after cessation of ocular warming, the RBF in the retinalartery and vein significantly increased by 14.2% ± 3.5% (P = .01) and 15.8% ± 3.6% (P = .006),respectively, compared with baseline (Figure1). The transient increase in RBF then returned to the prewarminglevel 6 and 9 minutes, respectively, after warming in the artery and vein.In the retinal artery, although the vessel diameter did not significantlychange in response to ocular warming, 3 minutes after cessation of ocularwarming the blood velocity significantly increased by 9.3% ± 1.9% byrepeated-measures analysis of variance (P = .03).In the retinal vein, the vessel diameter significantly increased by 4.8% ±1.7% (P = .004) and 4.1% ± 1.7% (P = .01) over baseline 3 and 6 minutes, respectively, after cessationof ocular warming, whereas the change in blood velocity was not significant.

Changes in cbf in the foveal region

In contrast to the change in retinal circulation, the mean CBF for thegroup significantly decreased by 16.6% ± 4.2% and 24.2% ± 4.7%(P = .001 for both), 3 and 6 minutes, respectively,after warming (Figure 2) by repeated-measuresanalysis of variance followed by post hoc comparison with Dunnett procedure.Nine minutes after warming, the CBF returned to the baseline value. Althoughthe mean choroidal velocity for the group did not significantly change afterocular warming, 6 minutes after warming the choroidal volume significantlydecreased by 23.7% ± 4.8% (P = .002), suggestingthat the decrease in CBF was mainly caused by the decrease in choroidal volume.


In the present study, we report for the first time (to our knowledge)increases in RBF and decreases in CBF in the foveal region after cessationof ocular warming in healthy young volunteers.

The choroidal circulation is regulated to prevent retinal damage inresponse to ocular hyperthermia, which can affect the transport of substanceswithin the eye,21 produce coagulation of intracellularproteins,22 cause retinal edema,23 andbreak down the blood retinal barrier.24 Thisfunction of the choroidal circulation is especially true for the macular region.5 Parver et al6 suggestedthat the CBF acts as a heat source when the temperature of the choroid ishigher than the systemic temperature. In the present study, the CBF actedas a heat source, as the temperature of the choroid was higher than the systemictemperature as a result of ocular warming (Table 1). In addition, those authors reported that a decrease inCBF produced by increasing IOP reduced the ocular tissue temperature.6 Therefore, our results that the CBF decreased in responseto ocular warming suggest that the CBF in the foveal region may contributeto maintaining the temperature of the retinal tissue in response to ocularwarming. Our finding that the decrease in CBF was mainly caused by the decreasein choroidal blood volume (Figure 1)suggests that the choroidal vasculature, mainly choriocapillaris, may constrictin response to the increase in ocular temperature. These changes in choroidalcirculation observed in the present study might be considered thermal autoregulationof the CBF.

In normal conditions, the retinal-choroidal tissue temperature is regulatedto the lower level by the cooler anterior segment.6 Inour experiment, the temperature of the anterior segment significantly increasedby ocular warming (Table 3). Insuch a situation, although we could not measure the tissue temperature ofthe posterior segment, it is reasonable to believe that the ocular warmingincreased the retinal-choroidal temperature.

Parver et al7 reported that light-inducedocular warming caused the increased CBF in monkeys, indicating the abilityof the choroidal circulation to control the retinal thermal environment. Ourfinding that the CBF in the foveal region decreased in response to ocularwarming seems inconsistent with their results. One explanation may be thedifference between the methods used to increase the ocular temperature. Light-inducedwarming, used in their investigation, would mainly increase the temperatureat the photoreceptor level, whereas the method we used in the present studywarmed the entire eye, probably including the choroid and sclera (Table 3). It is likely that the CBF actsas a heat source in the former but as a heat sink in the latter. Therefore,we speculate that the CBF in the foveal region, which acted as a heat sourcein the present study, decreased to maintain the retinal temperature. Anotherpossible explanation is that the CBF may increase during the warming phaseand then decrease after cessation of warming. We could not examine this becauseit is not possible to measure the CBF during the warming phase.

In contrast to the choroidal circulation, the RBF increased in responseto ocular warming in the artery and vein (Figure 2). To our knowledge, no study has examined how retinal vesselsreact to an increase in ocular temperature. The underlying mechanisms causingincreased RBF in response to ocular warming are incompletely understood. However,based on the present data that velocity significantly increased in the arterybut that the vessel diameter increased in the vein, dilation of the retinalvessel produced by ocular warming might occur downstream of the measured retinalartery, especially in the mural cells. Because the downstream arterioles possessthe metabolic mechanism of microvascular regulation,25,26 wespeculate that the change in RBF may be associated with the metabolic changeinduced by ocular warming. Moreover, a possible explanation of the differencesbetween the retinal and choroidal circulation in response to ocular warmingmay be the differences in neurologic control, which affects choroidal circulationbut not retinal circulation.27

Recently, it has been reported that patients with AMD have impairedCBF.2,3 Grunwald et al3 reported that the CBF in the foveal region in patientswith AMD was 37% lower than in control subjects, using a laser Doppler flowmeter,which is the same technique used in the present study. In addition, with colorDoppler imaging, it was recently reported that transpupillary thermotherapycould lead to alterations in choroidal circulation.4 Thesefindings indicate that the change in ocular blood flow and the response toincreased temperature may play some role in the development and progressionof AMD. Although it may be difficult to measure ocular blood flow using laserDoppler velocimetry or laser Doppler flowmetry in older subjects, becausethese methods require good fixation, we believe that the examination of ocularblood flow in response to ocular warming may be a useful test of ocular vascularresponsiveness in older patients with macular disease, especially early-stageAMD.

It has been reported that increased ocular temperature induced a decreasein IOP.6 In the present study, we did not measurethe IOP during the measurement of ocular blood flow. In a preliminary study,18 the change in IOP in response to ocular warming in5 healthy men was evaluated, and the IOP did not change significantly during10 minutes of ocular warming, using the same method as in the present study.Therefore, we believe that the change in IOP induced by ocular warming haslittle effect on the present results.

In the present study, we used an ocular warming lamp, which producesheat in front of the eyelid, to increase the temperature of the entire eye.It is possible that heat damage to the cornea may have affected our results.In 5 healthy subjects under the same conditions, we determined by standardizedpachymetry that the induction of ocular warming did not change the total cornealthickness, with no significant change in total corneal thickness before (687± 18 µm) and just after (686 ± 18 µm) ocular warming.

In conclusion, our results suggest the presence of an active mechanismthat modulates subfoveal CBF to maintain retinal temperature in response toa warming lamp. We believe that the ability of the choroidal circulation inthe foveal region to modulate the temperature of the macula should promptfurther study of the role of temperature in macular disease. Additional studiesin older subjects are needed to elucidate the possibility that the responseof the CBF to ocular warming may be an effective marker of macular disease,especially AMD.

Correspondence and reprints: Taiji Nagaoka, MD, PhD, Department ofOphthalmology, Asahikawa Medical College, Midorigaoka Higashi 2-1-1-1, Asahikawa078-8510, Japan (

Submitted for publication April 28, 2003; final revision received December2, 2003; accepted March 5, 2004.

Ciulla  TAHarris  AChung  HS  et al.  Color Doppler imaging discloses reduced ocular blood flow velocitiesin nonexudative age-related macular degeneration.  Am J Ophthalmol. 1999;12875- 80PubMedGoogle ScholarCrossref
Mori  FKonno  SHikichi  TYamaguchi  YIshiko  SYoshida  A Pulsatile ocular blood flow study.  Br J Ophthalmol. 2001;85531- 533PubMedGoogle ScholarCrossref
Grunwald  JEHariprasad  SMDuPont  J  et al.  Foveolar choroidal blood flow in age-related macular degeneration.  Invest Ophthalmol Vis Sci. 1998;39385- 390PubMedGoogle Scholar
Ciulla  TAHarris  AKagemann  L  et al.  Transpupillary thermotherapy for subfoveal occult choroidal neovascularization:effect on ocular perfusion.  Invest Ophthalmol Vis Sci. 2001;423337- 3340PubMedGoogle Scholar
Parver  LMAuker  CCarpenter  DO Choroidal blood flow as a heat dissipating mechanism in the macula.  Am J Ophthalmol. 1980;89641- 646PubMedGoogle Scholar
Parver  LMAuker  CRCarpenter  DO The stabilizing effect of the choroidal circulation on the temperatureenvironment of the macula.  Retina. 1982;2117- 120PubMedGoogle ScholarCrossref
Parver  LMAuker  CRCarpenter  DODoyle  T Choroidal blood flow, II: reflexive control in the monkey.  Arch Ophthalmol. 1982;1001327- 1330PubMedGoogle ScholarCrossref
Parver  LMAuker  CRCarpenter  DO Choroidal blood flow, III: reflexive control in human eyes.  Arch Ophthalmol. 1983;1011604- 1606PubMedGoogle ScholarCrossref
Mori  AOguchi  YOkusawa  YOno  MFujishima  HTsubota  K Use of high-speed, high-resolution thermography to evaluate the tearfilm layer.  Am J Ophthalmol. 1997;124729- 735PubMedGoogle Scholar
Mori  AOguchi  YGoto  E  et al.  Efficacy and safety of infrared warming of the eyelids.  Cornea. 1999;18188- 193PubMedGoogle ScholarCrossref
Nagaoka  TMori  FYoshida  A Retinal artery response to acute systemic blood pressure increase duringcold pressor test in humans.  Invest Ophthalmol Vis Sci. 2002;431941- 1945PubMedGoogle Scholar
Yoshida  AFeke  GTMori  F  et al.  Reproducibility and clinical application of a newly developed stabilizedretinal laser Doppler instrument.  Am J Ophthalmol. 2003;135356- 361PubMedGoogle ScholarCrossref
Riva  CEFeke  GTEberli  B Bidirectional LDV system for absolute measurement of blood speed inretinal vessels.  Appl Opt. 1979;182301- 2306Google ScholarCrossref
Feke  GTTagawa  HDeupree  DMGoger  DGSebag  JWeiter  JJ Blood flow in the normal human retina.  Invest Ophthalmol Vis Sci. 1989;3058- 65PubMedGoogle Scholar
Feke  GTRiva  CE Laser Doppler measurements of blood velocity in human retinal vessels.  J Opt Soc Am. 1978;68526- 531PubMedGoogle ScholarCrossref
Riva  CECranstoun  SDMann  RMBarnes  GE Local choroidal blood flow in the cat by laser Doppler flowmetry.  Invest Ophthalmol Vis Sci. 1994;35608- 618PubMedGoogle Scholar
Riva  CECranstoun  SDGrunwald  JEPetrig  BL Choroidal blood flow in the foveal region of the human ocular fundus.  Invest Ophthalmol Vis Sci. 1994;354273- 4281PubMedGoogle Scholar
Yoshida  A New examination methods for macular disorders: application of diagnosisand treatment [in Japanese].  Nippon Ganka Gakkai Zasshi. 2000;104899- 942PubMedGoogle Scholar
Movaffaghy  AChamot  SRPetrig  BLRiva  CE Blood flow in the human optic nerve head during isometric exercise.  Exp Eye Res. 1998;67561- 568PubMedGoogle ScholarCrossref
Petrig  BLRiva  CE Optic Nerve Head Laser Doppler Flowmetry: Principlesand Computer Analysis.  Basel, Switzerland S Karger AG1996;
Stiehl  WLGonzalez-Lima  FCarrera  ACuebas  LMDiaz  RE Active defense of retinal temperature during hypothermia of the eyein cats.  J Physiol Paris. 1986;8126- 33PubMedGoogle Scholar
Parver  LM Temperature modulating action of choroidal blood flow.  Eye. 1991;5 (2) 181- 185PubMedGoogle ScholarCrossref
Riedel  KGSvitra  PPSeddon  JM  et al.  Proton beam irradiation and hyperthermia.  Arch Ophthalmol. 1985;1031862- 1869PubMedGoogle ScholarCrossref
Kiryu  JOgura  YMoritera  TYoshimura  NHonda  Y Breakdown of the blood-retinal barrier after radiofrequency-inducedocular hyperthermia.  Ophthalmologica. 1993;206107- 110PubMedGoogle ScholarCrossref
Kuo  LDavis  MJChilian  WM Endothelial modulation of arteriolar tone.  News Physiol Sci. 1992;75- 9Google Scholar
Nagaoka  TSakamoto  TMori  FSato  EYoshida  A The effect of nitric oxide on retinal blood flow during hypoxia incats.  Invest Ophthalmol Vis Sci. 2002;433037- 3044PubMedGoogle Scholar
Alm  ABill  A Ocular circulation. Hart  WMJed Adler's Physiology of the Eye. 9th St Louis, Mo Mosby– Year Book Inc1992;198- 227Google Scholar