Human Ocular Vasodynamic Changes in Light and Darkness

Abstract

purpose. To determine whether changes in the retinal blood flow in light and darkness occur in humans.

methods. The systolic and diastolic flow velocities were measured by color Doppler in the ophthalmic and the central retinal arteries in 12 healthy individuals in light and darkness.

results. In the ophthalmic artery there was a trend toward lower systolic velocity in darkness compared with that in the light, but there was no change in diastolic velocity. In the central retinal artery the systolic and the diastolic flow velocities were markedly increased in darkness. After re-exposure to light the systolic flow velocity decreased.

conclusions. Darkness is associated with increased blood flow velocity in the central retinal artery, probably reflecting increased retinal metabolic demands by the photoreceptors.

It is well known that even slight hypoxemia reduces dark adaptation, probably by interfering with the high energy demands of the photoreceptor complex. Likewise, it has been shown in animal experiments that the retinal consumption of oxygen and glucose are increased in darkness compared with that in the light. ^{1 2 3} Whether this increase in retinal metabolism is associated with significant changes in retinal blood flow is not clear. ⁴ In the few human studies of retinal blood flow in light and darkness that have been performed, the results have been inconsistent. ^{5 6 7} To see if changes in the retinal blood flow in light and darkness occur in humans, the arterial flow velocities within the central retinal and ophthalmic arteries were measured by color Doppler technique in healthy subjects in standardized light and in darkness. 

Methods

Subjects

Twelve healthy individuals were examined by color Doppler imaging (5 men and 7 women; 24–51 years of age; mean, 33 years). All participating individuals were anamnestically free from cardiovascular and neurologic diseases, took no medications, and were nonsmokers. They were ophthalmologically healthy (i.e., had normal vision, intraocular pressure [IOP], dark adaptation, and pupillary reactions). 

Test Procedure

All subjects were identically examined by color Doppler imaging (CDI) in supine position, shielded from the monitor. The measurements were performed by an experienced laboratory technician. Blood flow velocities in the ophthalmic and central retinal arteries were measured in each pair of eyes seven times under standardized conditions of light and darkness (Fig. 1) . The right eye was always examined before the left. The subject kept the eye not being examined open, when the room was in light. The initial recording was done in standardized light (120 lux in the position of the eyes; Fig. 1 ). Then the subject looked with both eyes into a 55-W lamp for 5 minutes from a distance of 20 cm (1350 lux in the position of the eyes), after which the right and the left eyes were examined. Consequently, the left eye was exposed to light somewhat longer than the right eye before examination (Fig. 1) . Then the monitor and the control panel were covered with red plastic film (Lee Filters, 027 medium red, BS 3944: Part 1, 1992; Light transmission: 0% [600 nm], >50% [650 nm], >80% [>700 nm]). All leaks of light were eradicated. The examination room was in complete darkness except for the red light emitted from the monitor and control panel. The subject kept both eyes closed during the period of darkness. Measurements in the right and left eyes were done after 5, 15, and 25 minutes of darkness. The light in the room was then turned on, the red filter was removed, and the sixth measurement of flow velocities in the right and left eyes was performed. Finally, the subject with both eyes open was once again exposed to the light from the 55-W lamp for 5 minutes, whereafter the right and the left eyes were reexamined. 

CDI

All examinations were done with an Acuson XP 128 (Acuson, Mountain View, CA) equipped with a 7-MHz linear-array real-time B-mode scanner, including a 5-MHz pulsed and color Doppler. The eye was insonated from the front with the long axis of the transducer approximately horizontal to the eye. The transducer was applied very gently to the closed eyelid, using conventional ultrasound coupling gel. The estimated in-situ spatial peak time-average intensity of the ultrasound signals was constantly kept below 25 mW/cm². Primarily the shadow of the optic nerve was localized, after which the central retinal and ophthalmic arteries were searched for using CDI. The velocity scale of the CDI was constantly kept below 9 cm/s. 

Blood flow velocities were measured with pulsed Doppler, with a sample volume gate of 1.5 mm. Registrations from the central retinal artery were performed at a depth of approximately 25 to 30 mm (i.e., 3–4 mm behind the optic disc). Because the angle between the ultrasound signal and flow direction was close to 0° no angle correction was used in examination of this artery. The ophthalmic artery was insonated at a depth of approximately 30 to 35 mm, close to where the artery crosses the optic nerve. Correction for the angle between blood flow and Doppler signal was carefully performed. 

From each registration the peak systolic velocity (V _S ) and end-distolic velocity (V _D ) were measured, ^{8 9} and the resistive index (RI) was subsequently calculated ¹⁰ as RI = (V _S − V _D )/V _S . 

Statistical Methods

The results are expressed as mean ± SEM. In all calculations the right and left sides were treated separately. 

Paired t-tests were used when comparing blood flow velocities (systolic or diastolic) or resistive index under different conditions of light and darkness in the central retinal artery or the ophthalmic artery. The basic level of significance was chosen as P ≤ 0.05. Because multiple t-tests were done, we used the Bonferroni method to minimize the influence of multiple comparisons. The five first registrations were regarded as one experiment, and the three last as another. After Bonferroni adjustment the levels of significance were P ≤ 0.005 (registrations 1–5) and P ≤ 0.017 (registrations 5–7). 

The regression coefficients were calculated from registrations 2 to 5 for systolic and diastolic flow velocities and resistive index in the central retinal artery and the ophthalmic artery. The regression coefficients were tested against zero (hypothesis tested: b = 0). 

The study was approved by the Ethics Committee at the University of Lund. The subjects gave informed consent to participate in the study. All experimental procedures conformed to the tenets of the Declaration of Helsinki. 

Results

Ophthalmic Artery

In the ophthalmic arteries the mean predarkness systolic velocities (range, 42.8–45.1 cm/s) tended to be higher than those in darkness (Fig. 2) . The maximal reduction of systolic velocity in darkness was 5.6 cm/s (12%; P = 0.09) in the right eye and 5.0 cm/s (11%; P = 0.06) in the left eye. After 25 minutes in darkness the mean systolic velocities in the left eye showed an increase to the predarkness levels both immediately after the light was turned on and after 5 minutes in bright light (5.1 and 5.1 cm/s; P = 0.06 and 0.02). The corresponding changes in the right eye were notably less marked and not significant. 

The diastolic flow velocities in the ophthalmic arteries (range, 8.8–11.9 cm/s) did not consistently change from light to darkness. The resistive indices in the ophthalmic arteries did not show any conclusive changes. 

Central Retinal Artery

During the light-darkness-light exposures (Fig. 1) there were marked changes in the systolic and diastolic flow velocities in the central retinal artery (Fig. 3) . The mean systolic flow velocity, which was 7.4 to 7.7 cm/s in standard and bright light, increased with the duration in darkness and exceeded after 25 minutes the velocity in light by approximately 2 cm/s (range, 9.5–9.8 cm/s). This increase was highly significant on both sides at all three registrations in darkness compared with standard light and corresponded after 25 minutes to a 25% to 32% increase in systolic flow velocity. 

With re-exposure to standard light there was a reduction in the systolic flow velocity by 1.5 cm/s in the right eye (15%; P = 0.004) and by 1.3 cm/s in the left eye (14%; P = 0.02). The mean decrease in flow velocity was slightly more pronounced after bright light, 1.7 cm/s on both sides (right eye: 17%; P = 0.002; left eye: 18%; P = 0.0005). 

Similar changes were seen in the diastolic flow velocities in darkness in the central retinal arteries (Fig. 3) . The mean velocity in standard and bright light predarkness ranged between 1.9 and 2.2 cm/s. The increases reached statistical significance after 15 and 25 minutes in darkness compared with standard light. The maximal increase was 1.3 cm/s in the right eye after 25 minutes (68%; P = 0.001) and in the left eye 1.0 cm/s after 15 minutes (53%; P = 0.001). With re-exposure to light there were no consistently significant changes. 

The resistive index of the central retinal artery in the initial standard light was compared with the values after 15 and 25 minutes in darkness. The resistive index in darkness showed a trend toward reduction (Table 1) . 

The four consecutive registrations (after bright light and after 5, 15, and 25 minutes in darkness) were also used to calculate the regression coefficients for the changes in velocity per minute. This was done for the systolic and diastolic velocities and the resistive index of the right and left ophthalmic artery and central retinal artery. The regression coefficients were tested against zero. For the central retinal artery the differences were highly significant for the systolic velocity on both sides (right eye: b = 0.074, P < 0.005; left eye: b = 0.068, P < 0.005). For the diastolic velocity there was also statistical significance for both eyes (right eye: b = 0.034, P < 0.025; left eye: b = 0.026, P < 0.05). The other calculated differences were nonsignificant. 

Discussion

This study was performed in a group of selected healthy subjects (nonsmokers), and the results should therefore reflect the physiological changes in ocular blood flow that occur in the transition from light to darkness and vice versa. 

In humans information about ocular blood flow changes in darkness is sparse. ^{5 6 7} In 1983 Feke et al. ⁵ and Riva et al. ⁶ each reported results from similar studies in three healthy subjects, examined by bidirectional fundus laser Doppler velocimetry. They used a helium–neon laser and reported increased retinal blood flow velocity after the transition from light to dark, interpreted as a consequence of an increased retinal oxygen consumption in darkness. However, one considerable disadvantage with the helium–neon laser Doppler velocimeter is that the laser beam has a light-adapting effect. ⁶ Because blood flow velocity cannot be measured in complete darkness by this method it was assumed that the average velocity measured within 15 to 20 seconds after turning on the laser beam would indicate blood velocity during darkness. ⁷ Consequently, it may be concluded that the helium–neon laser is not ideal for examining the effects of light and dark on the retinal circulation. ⁷  

To avoid this problem, in 1987 Riva et al. ⁷ used near infrared laser Doppler velocimetry in a single subject to examine retinal blood velocity in darkness. Surprisingly, in this individual the blood velocity decreased slightly during darkness, although not significantly. After illumination there was a rapid (within 10–20 seconds) and marked increase in retinal blood flow velocity. They concluded that blood velocity does not increase in darkness and that the sudden transition from dark to light caused the previously observed increase ^{5 6} rather than the prolonged state of dark adaptation. 

In the present study we used CDI. There are certain advantages with CDI compared with laser Doppler velocimetry when examining ocular flow velocity in light and darkness. The problem that the laser beam causes (illumination of the retina) is avoided. The examination can be performed in complete darkness with the subject’s eyes closed (with the weak light sources in the room covered by a red filter). Repeated examinations can be done. There is no need for strict visual fixation by the subject. One disadvantage with the CDI technique may be the possibility of increasing IOP by the gentle pressure of the probe on the globe. This influence is minimized when the examination is performed by an experienced ultrasonographer. 

Although there were no significant changes in the blood flow velocities in the ophthalmic artery between light and darkness, there may be a trend toward a lower systolic flow velocity in darkness (Fig. 2) . The reason for this may be that according to the normal vascular anatomy of the orbita, the flow velocity in the ophthalmic artery was measured distal to the branching off of the central retinal artery. Furthermore, the ophthalmic artery has a larger lumen diameter and considerably higher systolic and diastolic flow velocities than the central retinal artery. Consequently, even marked changes in the flow in the central retinal artery (see below) may not be accompanied by statistically significant flow velocity changes in the ophthalmic artery. It may be added that the reduced choroidal flow in darkness, secondary to a reduced need for retinal “cooling,” ¹¹ may contribute to a lowered flow velocity in the ophthalmic artery. 

The main finding in this study was that the systolic and diastolic blood flow velocities in the central retinal artery were markedly increased in darkness and that the systolic velocity was reduced after re-exposure to light (Fig. 3) . Somewhat unexpectedly, the diastolic velocities in both eyes did not return to baseline values after re-exposure to light. This may reflect a rather slow desadaptation process, for which the time course of the possibly accompanying change in flow is unknown. ¹²  

What may be the reason for the increase in flow velocity in the central retinal artery in darkness? Factors of relevance for flow velocity are the mean arterial blood pressure, IOP, blood viscosity, lumen diameter of the vessel, and peripheral resistance. 

An increased mean arterial blood pressure in darkness is theoretically a possible explanation. However, the mean arterial blood pressures in supine position in six healthy individuals did not change significantly (authors’ unpublished observations) during dark adaptation compared with light adaptation (see “Methods”). Furthermore, in 1991 Guthoff et al. ¹³ did not find any correlation between the systolic flow velocity in the central retinal artery and the systolic brachial artery blood pressure in a group of normal individuals. Consequently, the observed changes in flow velocities can hardly be ascribed to variations in systemic blood pressure. 

Another possible explanation for the increase in flow velocity in the central retinal artery in darkness might be a decrease in IOP. All subjects in this study had normal IOPs. It is known that normally there is an increase in IOP amounting to a few millimeters of mercury in darkness. ^{14 15} This increase may be due to changes in vasoregulation, but it is not caused by any change of the pupillary diameter. ^{14 15} It is also known that artificial elevation of IOP by suction cup dynamometry in normal eyes will result in progressive diminution of velocity with increasing IOP. ^{13 16} However, in normal eyes Guthoff et al. ¹³ (in 1991) found no correlation between the systolic flow velocity in the central retinal artery and the IOP. Consequently, our finding of an increased flow velocity in the central retinal artery in darkness cannot be explained by any expected change in the IOP in darkness. 

Among the remaining factors governing flow velocity, changes in viscosity should reasonably be excluded, considering the short duration of the experiments. 

Changes in the lumen diameter of the examined vessel affect blood flow velocity, provided that all other parameters (e.g., perfusion pressure, blood viscosity) remain constant. The possible variation, between light and darkness in the lumen caliber of the central retinal artery at the position of repeated color Doppler measurements, is not known. Information about changes in the caliber of the retinal arteries in light and darkness is sparse in the literature. Feke et al. ⁵ reported a negligible dilatation of larger branch retinal arteries in humans, amounting to 2% to 3% in darkness, whereas Hill and Houseman ⁴ did not find any conclusive changes in the caliber of retinal arterioles between light and darkness in the cat. 

Assuming that luminal changes in the central retinal artery between light and darkness are minor, the only possible interpretation of our findings is that there is a decrease in peripheral retinal vascular resistance in darkness. The fall in resistive index (Table 1) , derived to reflect variations in the peripheral vascular resistance, supports such an interpretation. Furthermore, Feke et al. ⁵ and Riva et al. ⁶ (1983) reported an increase in the diameter of large retinal veins, amounting to 5% to 8% after 20 to 30 minutes of dark adaptation, a finding that may be compatible with such an interpretation. 

Having considered and excluded various other possible explanations, the progressive increases in systolic and diastolic flow velocities in the central retinal artery with time spent in darkness and subsequently improving dark vision (Fig. 3) may possibly reflect an increased metabolic demand by the photoreceptors. Experimental studies in animals have shown that darkness is accompanied by a markedly increased oxidative metabolism in the retinal photoreceptors. ^{1 2 3} Consequently, dark vision is critically dependent on a sufficient oxygen supply. 

What may then be the explanation for the increase in flow velocity in the central retinal artery? The retina has a dual vascular supply: the retinal arteries and the choroidal circulation. The contribution of oxygen from each of these sources at different retinal depths depends on conditions of light and darkness. ^{2 3} The oxygen derived from the choroid reaches the inner segments of the photoreceptors independent of light conditions. In darkness the photoreceptors have a markedly raised oxidative metabolism, and, consequently, the amount of oxygen diffusing from the choroid into the retina beyond the photoreceptors is reduced. Because the inner retina maintains its oxidative metabolism irrespective of light and dark, ^{1 3} the oxygen tension in the inner retina will fall in darkness and trigger an increased retinal blood flow by mechanisms of autoregulation. 

 Reprint requests: Ulf Havelius, Department of Ophthalmology, University Hospital MAS, S-205 02 Malmö, Sweden.

 Supported by grants from the Herman Järnhardt Foundation and the Frederico Hecht Foundation.

 Submitted for publication April 7, 1998; revised January 5, 1999; accepted February 24, 1999.

 Proprietary interest category: N.

 

Figure 1.

 

Schematic illustration of the test procedure. D1, D2, etc., the seven consecutive Doppler registrations.

Figure 2.

 

The peak systolic and end-diastolic flow velocities (mean ± SEM) in the ophthalmic artery (OA) of the right and left eyes of 12 healthy subjects under different conditions of light and darkness. The seven consecutive Doppler registrations (1–7) are indicated on the abscissa. •, right eye; ○, left eye.

Figure 3.

 

The peak systolic and end-diastolic flow velocities (mean ± SEM) in the central retinal artery (CRA) of the right and left eyes of 12 healthy subjects under different conditions of light and darkness. The velocity values in standard light are compared with the values after 5, 15, and 25 minutes in darkness. The probability value of the eye with the lowest significance level is given. After re-exposure to standard light there is a decrease in systolic flow rate. •, right eye; ○, left eye.

Table 1.

 

Table 1.

 

Comparison of Resistive Index (mean ± SEM) in the Central Retinal Artery in 12 Subjects

Eye Examined	Standard Light	Darkness 15 Min	Darkness 25 Min
Right	0.74 ± 0.02	0.68 ± 0.02	0.68 ± 0.02
P		0.11	0.17
Left	0.76 ± 0.03	0.70 ± 0.02	0.71 ± 0.02
P		0.02	0.02

The authors thank Anette Holmén and Elzbieta Krolikowska, Department of Clinical Physiology, University Hospital, Malmö, Sweden, for performing the Doppler measurements.