Retinal Blood Flow and Vascular Reactivity in Chronic Smokers

Abstract

Purpose.: The aim of this study was to investigate the impact of cigarette smoking in otherwise healthy young individuals on retinal blood flow (RBF) and vascular reactivity (RVR).

Methods.: An automated gas flow controller (RespirAct) was used to achieve normoxic hypercapnia in 10 nonsmokers (mean age 28.9; SD 4.6 years) and nine smokers (mean age 27.55; SD 4.7 years). Retinal blood flow measurements were obtained using a prototype Doppler spectral-domain optical coherence tomographer (SD-OCT) and bidirectional laser Doppler velocimetry and simultaneous vessel densitometry during baseline, normoxic hypercapnia, and recovery. Group mean P_ETCO₂ (end-tidal partial pressure of CO₂) was increased by 15.9% in the nonsmoking group and by 15.7% in the smoking group, with a concomitant increase in P_ETO₂ (end-tidal partial pressure of O₂) by approximately 1.5% to 2% in both groups.

Results.: In nonsmokers, retinal arteriolar diameter (P < 0.0001), centerline velocity (P = 0.0004), and blood flow (P < 0.0001) significantly increased during normoxic hypercapnia. Similarly, the venous area (P = 0.0418), venous velocity (P = 0.0068), and total venous RBF (P < 0.0001), as measured by the prototype Doppler SD-OCT, significantly increased. In smokers, normoxic hypercapnia resulted in a significant increase in velocity (P = 0.0019), flow (P = 0.0029), and total venous RBF (P = 0.002). Comparing smokers and nonsmokers, the percentage change in arteriolar diameter (P = 0.0379) and blood flow (P = 0.0101) was significantly lower in the smoking group. There was no significant difference in baseline P_ETCO₂ level between smokers and nonsmokers.

Conclusions.: Retinal vascular reactivity in response to normoxic hypercapnia is significantly reduced in young, healthy smokers compared with nonsmokers.

Introduction

Cigarette smoking is considered to be one of the major risk factors for several ocular diseases. In particular, it is an independent, but avoidable risk factor for age-related macular degeneration (AMD). Cigarette smoke contains more than 4000 chemicals, 200 poisonous gases, and a number of unidentified chemical components. ¹ Nicotine and carbon monoxide (CO) are identified as two major constituents of cigarette smoke. ² The cytotoxic effects of free radicals found in cigarette smoke may be the main contributing factor in accelerating smoking-induced disease mechanisms. ³  

Smoking enhances the generation of free radicals and decreases the level of antioxidants in the blood circulation. This places the eye at increased risk to damage due to elevated oxidative stress and the site of this damage is focused on the RPE-photoreceptor interface where metabolic activity is at its highest. ⁴ A causal relationship between smoking and AMD ^5,6 has been clearly elucidated by many studies; especially the Beaver Dam Eye Study, ⁷ Rotterdam Study ⁸ and Blue Mountains Eye Study, ⁹ all of which report a strong association between current smoking and exudative AMD. Thus, the detrimental effect of cigarette smoking on retinal physiology is of highly relevant clinical interest. 

Autoregulation in the purest sense of the word is defined as the ability of the vasculature to maintain blood flow at relatively constant levels despite moderate variations in perfusion pressure. ¹⁰ On the other hand, metabolic autoregulation is the ability of the vasculature to modulate blood flow in order to maintain metabolic components, including tissue oxygen (O₂) and carbon dioxide (CO₂), at relatively constant levels. Metabolic autoregulation is otherwise known as “vascular reactivity.” ¹¹ The impact of smoking on retinal autoregulation, vascular reactivity and other ocular retinal hemodynamic parameters in otherwise healthy individuals is controversial. Retinal hemodynamic alteration in smokers deserves more attention because these changes might predispose an individual to sight-threatening AMD, essentially a disease with a clear vascular component in its pathophysiology. 

In the past, provocations with various mixtures of inhaled O₂ and CO₂ have been undertaken extensively by our research group to assess retinal vascular reactivity (RVR) in healthy ^12–15 and diseased cohorts. ^16,17 Studies have shown that both retinal and the cerebral vessels react similarly by constricting to O₂ and by dilating to CO₂. ^13,14,18,19 Vascular reactivity to CO₂ has long been used to test retinal ^19–21 and cerebrovascular hemodynamics. It has been shown that increase in arterial blood partial pressure of CO₂ above normal resting values, termed “hypercapnia,” results in vasodilation of retinal vessels, thereby increasing retinal arteriolar blood flow. The loss of RVR to CO₂ provocation may provide insight into disease mechanisms and potential treatments. ^17,22,23  

Previous studies have shown that smoking induces a marked decrease in retinal blood flow (RBF) and a reduced ability of retinal vessels to autoregulate immediately after smoking. ²⁴ A few studies have reported reduced reactivity in retinal vessels to hyperoxia in smokers, ^25,26 whereas the response to hypercapnia was shown to be altered in the choroid, but not in the retina. ²⁷ Smoking has also been shown to markedly increase the basal cerebral blood flow by approximately 25% in humans. ²⁸ Clearly, the vascular reactivity response of the retina to increase in P_ETCO₂ in chronic smokers needs to be clarified. 

The hypothesis for this study is that the RVR to inhaled carbon dioxide will be reduced in otherwise healthy subjects who smoke when compared with nonsmoking, age-matched controls. The overall aim of this study was to investigate the impact of cigarette smoking in otherwise healthy young individuals on RBF and RVR. The impact of both acute (after smoking) and chronic (before smoking) cigarette smoking on total RBF was also investigated. 

Materials and Methods

Sample

This study was approved by the University of Waterloo, Office of Research Ethics, Waterloo (Ontario, Canada), and by the University Health Network Research Ethics Board (Toronto, Canada). The sample comprised 10 nonsmokers (mean age 28.9; SD 4.6 years; four female) and nine smokers (mean age 27.55; SD 4.7 years; four female). Previous studies have shown that smoking reduces the magnitude of vascular reactivity by approximately 50%. ^25,26 Based on the results of Venkataraman and coworkers, ¹⁵ the effect of hypercapnia on mean retinal arteriolar blood flow in control subjects is an increase of 24.9% (SD 7%). The standardized effect size is calculated using  

 

Assuming an alpha of 0.05 and a beta of 0.1 (i.e., power), a sample size of 7 was sufficient for each group. One eye of each subject was randomly chosen for this study. All subjects had a logMAR visual acuity of 0.0 or better. Smokers with a smoking history of at least 2 years and who regularly smoked 15 to 25 cigarettes per day were included. All the nonsmoking participants had no previous history of smoking. Exclusion criteria included any refractive error greater than ±6.00-diopters (D) sphere and/or ±1.50-D cylinder, IOP greater than 21 mm Hg, obesity, treatable respiratory disorders (e.g., asthma), systemic hypertension, cardiovascular disease, diabetes, endocrine disorders, medications with known effects on blood flow (e.g., antihypertensive, medications with activity at autonomic receptors, smooth muscles, or medications that affect nitric oxide release), family history of glaucoma, or a personal history of any ocular disease. All the participants were asked to abstain from caffeine, red meat, and alcohol for 12 hours, and avoid rigorous exercise approximately 1 hour prior to their study visit. Informed consent was obtained from each subject after a thorough explanation of the nature of the study and its possible consequences, according to the tenets of the Declaration of Helsinki.  

The study was performed during a single visit for all participants. However, nonsmokers attended for a single session, whereas smokers had two sessions in a single day so that we could measure both the acute (after smoking) and chronic (after nonsmoking period of approximately 6 hours) effect of cigarette smoking. Each session was approximately 1.5 hours in length. 

Instrumentation

Doppler Spectral Domain-Optical Coherence Tomography (SD-OCT).

The prototype Doppler SD-OCT, as incorporated into the commercially available Optovue RTVue OCT (Optovue, Inc., Fremont, CA, USA), is a novel technique for the noninvasive quantitative measurement of total retinal blood flow (TRBF) in absolute units. ^29–31 Doppler SD-OCT uses the principle of the Doppler effect to determine the velocity of moving red blood cells. This spectrometer-based OCT system contains a superluminescent diode with a center wavelength of 841 nm and a bandwidth of 49 nm. The axial resolution is 5.6 μm in tissue and transverse resolution is 20 μm as limited by optical diffraction of the eye. The composite signal produced by the interference between the sample and reference arm light in the fiber coupler is detected by a custom spectrometer, consisting of a 1024-pixel line-scan camera. Fourier transformation of the resulting interference pattern gives the sample depth profile or A-scan through the neural layers of the retina, each of which has differing reflectance properties (Fig. 1). The phase difference between sequential axial scans at each pixel is used to determine the Doppler shift. ^29,30,32  

Figure 1

 

Schematics of a spectrometer-based Fourier-domain OCT system. (Reprinted with permission from Huang D; from the article OCT terminology demystified. Ophthalmol Manage. 2009;13:62–64. A PentaVision publication.)

To derive TRBF, the Doppler scans are acquired using the RTVue SD-OCT (Optovue, Inc.) and a double circular scan protocol (DCSP). ³¹ The DCSP consist of two concentric circles of 3.4- and 3.75-mm diameters, respectively. These two circles transect all retinal arterioles and venules located near the optic nerve head in two locations. A total of six dual circular scan frames were obtained and averaged for each ring. From the measured Doppler shift with in the vessel and Doppler angle estimation from the vessel center depth difference between two concentric rings, volumetric flow is derived using a semi-automated software algorithm named Doppler optical coherence of retinal circulation (DOCTORC). ³³ Total retinal blood flow is calculated by summing up flows from all branch veins at one point in time. ³⁰  

Canon Laser Blood Flowmeter.

The Canon Laser Blood Flowmeter (CLBF; Model 100; Canon, Tokyo, Japan) measures RBF (μL/min) in absolute units. The CLBF uses the Doppler shift to identify the fastest moving blood particles in the center of a retinal blood vessel and then calculates the average velocity. The technique has been described in detail from previous publications from our lab. ^34,35 It has the major advantage of being able to measure absolute blood velocity and also benefits from an eye tracking system that not only measures vessel diameter but also stabilizes the Doppler and vessel diameter lasers during physiological eye movement. ^36–39  

Gas Delivery System.

The gas challenges were manipulated using a sequential gas delivery (SGD) breathing circuit (Hi-Ox⁸⁰; VIASYS Health care, Yorba Linda, CA, USA). The SGD circuit has been described in previous publications. ^13,15 The subject's minute CO₂ production and O₂ consumption, gas flow, and composition entering the SGD breathing circuit will be attained and stabilized using an automated gas flow controller (RespirAct; Thornhill Research, Inc., Toronto, Canada). The distinguishing feature of the SGD and automated gas flow controller over other gas delivery systems is the efficacy in ‘targeting' end-tidal gas concentrations, such that the software enables the user to repeatedly target P_ETCO₂ and P_ETO₂, independent of each other and independent of minute ventilation. ^40,41  

Procedures

All subjects completed a standard Fagerstrom Tolerance Questionnaire (FTQ). The FTQ quantitates the nicotine dependency of a smoker. ^42,43 The questionnaire has a total of seven questions regarding the smoking characteristics of an individual. The FTQ has a scoring range of 0 to 11 points. A score of 0 indicates minimum nicotine dependence, whereas a score of 11 indicates maximum nicotine dependence. Under normal smoking conditions, sampled CO was found to be highly correlated with self-reported FTQ scores. ^42,44 The FTQ scoring scale to quantitate nicotine dependence was reported to be valid, reliable, and applicable to adolescent smokers. ^45–47  

Carbon monoxide levels in the breath were analyzed for all participants using CO Sleuth carbon monoxide monitor (Breathe E-Z Systems, Inc., Leawood, KS, USA). The CO Sleuth is a simple portable instrument, which enables a noninvasive and direct measure of breath carbon monoxide level. The subject breathes out into a disposable mouth piece after holding the breath for 20 seconds. The end-tidal breath sample enters through a nonreturn valve, and reaches an electrochemical sensor. This novel sensor detects alveolar breath CO concentration, thus enabling the monitor to measure CO level in parts per million (ppm). The CO concentration in alveolar breath was found to be highly correlated with blood carboxyhaemoglobin (CO-Hb) levels. ⁴⁸  

LogMAR visual acuity and IOP (using the Goldmann Applanation Tonometer; Haag-Streit, Koniz, Switzerland) were recorded for both eyes. One eye was randomly selected for the study and dilated with 1.0% Mydriacyl, Tropicamide (Alcon, Mississauga, Canada). Subjects were rested in sitting position for 10 minutes to stabilize baseline cardiovascular and respiratory parameters. Following that, participants were fitted with a face mask connected distally to the partial rebreathing circuit and the gas sequencer. The automated gas flow monitor connected to the rebreathing circuit was used to sample inspired and expired gases. Retinal blood flow measurements were obtained sequentially using Doppler SD-OCT and CLBF, with a duration of 45 minutes for each of these methodologies under the following conditions lasting for 15 minutes each. 

Initially, baseline RBF (while breathing room air to establish baseline cardiovascular parameters). Then, RBF was assessed during normoxic hypercapnia (∼15% increase in end-tidal CO₂ relative to the baseline). Finally, RBF was assessed during postnormoxic hypercapnia (while breathing room air to establish recovery from hypercapnia). 

Blood pressure, pulse rate, and oxygen saturation were monitored continuously using a rapid response critical care gas analyzer (Cardiocap 5; Datex-Ohmeda, Helsinki, Finland). Six CLBF measurements of a relatively straight segment of the superior arteriole, away from bifurcations, were acquired during each provocation stage. For TRBF measurement using Doppler SD-OCT, a total of six scans (three inferior nasal and three superior nasal) using ‘double circular scan protocol' were acquired during each breathing condition. The instrument order was randomized between subjects. 

For smokers, following completion of session 1 (∼1.5 hours), a break of 15 minutes was allocated, during which the examiner asked the subject to smoke one full cigarette. After the break, the subject was seated for 10 minutes during which the breath CO level was analyzed and the subject was fitted again with a face mask and seated comfortably to begin RBF measurements for session 2. It took almost 30 minutes to start obtaining RBF measurements after smoking. The order of instruments remained the same as with the previous session. Following completion of the study sessions, a color fundus photograph of optic disc (Canon nonmydriatic fundus camera, CR-DGI; Canon) of the study eye was taken for all participants. 

Statistical Analysis

After ensuring the data was drawn from a normally distributed population using Shapiro-Wilk test, a repeated measures analysis of variance (reANOVA) was undertaken to determine any significant change in RBF and RVR between smokers and nonsmokers in CLBF and Doppler SD-OCT measurements. In addition, a separate reANOVA was performed within each group (i.e., smokers and nonsmokers) to determine the significance of any change of each dependent variable during baseline, normoxic hypercapnia, and recovery. If a significant result was achieved using reANOVA, then post hoc testing was performed using Tukey's honestly significant difference (HSD) test. Significance of change of any systemic and gas parameters including P_ETCO₂, P_ETO₂, respiration rate, systolic and diastolic blood pressure, pulse rate, and oxygen saturation were also analyzed using reANOVA. 

To determine the significance in TRBF measurements before and after smoking (n = 7), a Wilcoxon rank sum test was performed. The CO level in nonsmokers and smokers (before and after smoking) was compared using one-way ANOVA. Statistica software (version 11.0; StatSoft, Inc., Tulsa, OK, USA) was used for analyzing the data. A change was considered significant if P was less than 0.05. 

Results

The mean age, IOP, and mean ocular perfusion pressure (MOPP) for all participants are shown in Table 1. MOPP was calculated using the equation shown below (BP; blood pressure):  

 

 

Table 1

 

Age, IOP, and MOPP in Nonsmokers and Smokers

Parameter	Nonsmokers	Smokers
Age, y	28.9 ± 4.6	27.55 ± 4.7
IOP, mm Hg	16.6 ± 2.4	14.11 ± 3.2
MOPP, mm Hg	40.7 ± 5.6	39.8 ± 7.7

Retinal Vascular Reactivity and Blood Flow in Nonsmokers

There was a significant increase in retinal arteriolar diameter, blood velocity, and flow by +4.1% (SD 2.8, P < 0.0001), +16.7% (SD 14.6, P = 0.0004), and +29.6% (SD 12.5, P < 0.0001) during normoxic hypercapnia (reANOVA, Fig. 2). The study revealed an increase in diameter during hypercapnia relative to baseline (P = 0.0014) and a reduction in recovery (P = 0.0002) relative to effect (Tukey's HSD). Similarly, hypercapnia resulted in an increase in velocity measurements relative to baseline (P = 0.0027) and a reduction in recovery (P = 0.0006) relative to effect. Retinal arteriolar flow during hypercapnic provocation was increased relative to baseline (P = 0.0001) and recovery (P = 0.0001) was reduced relative to effect. For all three hemodynamic parameters, baseline, and recovery values were unchanged. 

Figure 2

 

Box plots represent retinal arteriolar diameter (upper left), blood velocity (upper right), and blood flow (center) at baseline, during normoxic hypercapnia and at recovery in nonsmokers. The legend in the middle of the box represents the median value, the upper and lower extremes of the box represent 25th and 75th percentiles, and the error bars represent the nonoutlier range. *P < 0.001, **P < 0.01.

Venous area, venous velocity, and TRBF significantly increased during normoxic hypercapnia by +7% (SD 8.6, P = 0.0418), +18.1% (SD 20.8, P = 0.0068), and +26% (SD 22.9, P < 0.0001), respectively (reANOVA, Fig. 3). The study revealed an increase in TRBF during hypercapnia relative to baseline (P = 0.0001) and a decrease in TRBF (P = 0.0001) relative to recovery. Similarly, hypercapnia resulted in an increase in venous velocity relative to baseline (P = 0.0102) and a reduction in velocity (P = 0.0204) relative to recovery. A significant increase in venous area during hypercapnia (P = 0.0418) was found, however, post hoc testing (Tukey's HSD test) did not show a significant difference in pairwise comparison between baseline and hypercapnia (P = 0.0832) as well as between hypercapnia and recovery (P = 0.0571). For all three hemodynamic parameters, baseline, and recovery values were unchanged. 

Figure 3

 

Box plots represent venous area (upper left), blood velocity (upper right), and total venous retinal blood flow (center) at baseline, during normoxic hypercapnia and at recovery in nonsmokers. The legend in the middle of the box represents the median value, the upper and lower extremes of the box represent 25th and 75th percentiles, the error bars represent the nonoutlier range, and the circles represent outliers. *P < 0.001, ***P < 0.05.

Retinal Vascular Reactivity and Blood Flow in Smokers

There was a significant increase in velocity of +12.0% (SD 6.2, P = 0.0019) and flow of +14.6% (SD 9.5, P = 0.0029), while diameter was unchanged (+1.7%, SD 1.7, P = 0.2616; Fig. 4). The study showed an increase in blood velocity compared with baseline (P = 0.0194) and a decrease in velocity (P = 0.0019) relative to recovery. Similarly, there was an increase in retinal arteriolar blood flow during normoxic hypercapnia compared with baseline (P = 0.0076) and a decrease in blood flow relative to recovery (P = 0.0058). For all three hemodynamic parameters, baseline, and recovery values were unchanged. 

Figure 4

 

Box plots represent retinal arteriolar diameter (upper left), blood velocity (upper right), and blood flow (center) at baseline, during normoxic hypercapnia and at recovery in smokers. The legend in the middle of the box represents the median value, the upper and lower extremes of the box represent 25th and 75th percentiles, the error bars represent the nonoutlier range, and the circles represent outliers. **P < 0.01, ***P < 0.05.

There was a significant increase in TRBF by +19.3% (SD 18.4, P = 0.002) during normoxic hypercapnia (reANOVA, Fig. 5). There was an increase in TRBF during hypercapnia compared with baseline (P = 0.0035) as well as a significant decrease in TRBF (P = 0.0087) relative to recovery. For TRBF, baseline and recovery values were unchanged. There was no difference in venous area (P = 0.3322) and venous velocity measurements (P = 0.1185) during hypercapnia compared with baseline and recovery (Fig. 5). 

Figure 5

 

Box plots represent venous area (upper left), blood velocity (upper right), and total venous retinal blood flow (center) at baseline, during normoxic hypercapnia and at recovery in smokers. The legend in the middle of the box represents the median value, the upper and lower extremes of the box represent 25th and 75th percentiles, the error bars represent the nonoutlier range, and the circles represent outliers. Star legend: extreme value. **P < 0.01.

Comparison of Retinal Vascular Reactivity and Blood Flow Between Smokers and Nonsmokers

The percentage change in retinal arteriolar diameter to normoxic hypercapnia in smokers was significantly reduced compared with nonsmokers (P = 0.0379). Similarly, the percentage change in flow was also reduced in smokers (P = 0.0101; Fig. 6) compared with nonsmokers. However, blood velocity was not significantly different between the groups (P = 0.3851). Total retinal blood flow (P = 0.3624), venous area (P = 0.5669), and venous velocity (P = 0.5189) were not different between the groups. 

Figure 6

 

Percentage change from baseline in group mean arteriolar diameter (upper left), blood velocity (upper right), and blood flow (center) in response to normoxic hypercapnia in nonsmokers and smokers. Level of significance was set to P < 0.05.

Comparison of Retinal Vascular Reactivity and Blood Flow Before and After Smoking

The percentage change in retinal arteriolar diameter, velocity, and flow in response to normoxic hypercapnia after smoking was not significantly different when compared with before smoking (Wilcoxon rank sum test). In addition, the percentage change in vascular reactivity in response to normoxic hypercapnia was also unchanged in terms of TRBF, venous area, and venous velocity measurements before and after smoking. 

Systemic and Gas Parameters

In the nonsmoking group, P_ETCO₂ was increased by +16% (SD 1.7, P < 0.001) relative to homeostatic baseline with a concomitant nonsignificant increase in mean P_ETO₂ by +0.9% (SD 2.1, P = 0.29) in session 1a. There was an increase in mean P_ETCO₂ by +15.9% (SD 2.9, P < 0.001) relative to baseline with a concomitant increase in mean P_ETO₂ by +1.8% (SD 0.5, P = 0.005) in session 1b. There was no change in other systemic parameters measured during different breathing conditions across both the study sessions (Table 2). 

Table 2

 

Group Mean (±SD) for Gas and Systemic Parameters Across Different Breathing Conditions (i.e., Baseline, Normoxic Hypercapnia, and Recovery) in Nonsmokers During Sessions 1a and 1b

Gas & Systemic Parameters, Nonsmokers	Baseline	Hypercapnia	Recovery	P Value, reANOVA
Session 1a
PETCO2, mm Hg	33.6 ± 3.1	39.0 ± 3.7	33.3 ± 3.4	P < 0.0001
PETO2, mm Hg	119.1 ± 5.9	120.2 ± 6.6	119.8 ± 7.1	NS
Respiration rate, breaths/min	16.9 ± 2.9	18.0 ± 3.1	17.4 ± 2.7	NS
Pulse rate, beats/min	71.5 ± 8.9	71.6 ± 10.5	73.0 ± 9.8	NS
Systolic blood pressure, mm Hg	111.2 ± 10	114.4 ± 8.2	111.3 ± 10.7	NS
Diastolic blood pressure, mm Hg	73.4 ± 5.3	74.1 ± 3.9	72.7 ± 4.7	NS
O2 saturation, %	97.3 ± 1.5	97.7 ± 1.3	97.8 ± 1.1	NS
Session 1b
PETCO2, mm Hg	33.4 ± 3.6	38.7 ± 3.7	33.5 ± 3.3	P = 0.009
PETO2, mm Hg	120.4 ± 7.1	121.8 ± 7.2	120.4 ± 7.1	P = 0.005
Respiration rate, breaths/min	17.6 ± 3.2	18.2 ± 2.9	16.6 ± 1.5	NS
Pulse rate, beats/min	71.6 ± 8.8	72.2 ± 8.8	70.3 ± 9.3	NS
Systolic blood pressure, mm Hg	111.7 ± 8.2	113.4 ± 6.8	113.3 ± 6.6	NS
Diastolic blood pressure, mm Hg	76.5 ± 5.0	77.4 ± 4.4	76.6 ± 4.1	NS
O2 saturation, %	97.8 ± 1.0	97.6 ± 1.2	97.6 ± 1.3	NS

 Level of significance was set to P < 0.05. NS, not significant; O₂ saturation, oxygen saturation.

In smokers, there was a +15.7% (SD 2.2, P < 0.0001) increase in P_ETCO₂ with a concomitant nonsignificant increase in P_ETO₂ by +1.7% (SD 1.4, P = 0.3135) in session 1a. There was an increase in P_ETCO₂ by +15.7% (SD 2.6, P < 0.0001), but there was no change in P_ETO₂ (+1.4%, SD 4.2, P = 0.2168) relative to baseline in session 1b (Table 3). Also, session 2a results show an increase in pulse rate after smoking during baseline (P = 0.0279), hypercapnia (P = 0.0179), and recovery (P = 0.0464) compared with before smoking. Similarly, baseline respiration rate was higher (P = 0.0267) as compared with before smoking, but showed no change during hypercapnia and recovery. 

Table 3

 

Group Mean (±SD) for Gas and Systemic Parameters Across Different Breathing Conditions (i.e., Baseline, Normoxic Hypercapnia, and Recovery) in Smokers During Sessions 1a and 1b

Gas & Systemic Parameters, Smokers: Before Smoking	Baseline	Hypercapnia	Recovery	P Value, reANOVA
Session 1a
PETCO2, mm Hg	35.2 ± 5.1	40.7 ± 5.7	35.1 ± 5.3	P < 0.0001
PETO2, mm Hg	113.1 ± 10.7	114.7 ± 10.1	112.8 ± 13.0	NS
Respiration rate, breaths/min	16.3 ± 2.2	16.2 ± 2.5	16.3 ± 2.5	NS
Pulse rate, beats/min	65.2 ± 9.4	66.4 ± 7.9	67.0 ± 6.3	NS
Systolic blood pressure, mm Hg	109.0 ± 15.6	110.3 ± 14.3	109.3 ± 14.8	NS
Diastolic blood pressure, mm Hg	66.8 ± 11.0	70.6 ± 11.5	68.6 ± 11.8	NS
O2 saturation, %	97.3 ± 1.4	98.1 ± 0.8	97.7 ± 1.0	NS
Session 1b
PETCO2, mm Hg	35.0 ± 5.2	40.5 ± 5.6	34.7 ± 5.4	P < 0.0001
PETO2, mm Hg	112.9 ± 12.9	114.1 ± 11.4	114.7 ± 12.0	NS
Respiration rate, breaths/min	16.4 ± 2.5	17.0 ± 2.9	16.7 ± 2.8	NS
Pulse rate, beats/min	66.4 ± 8.3	67.1 ± 7.6	67.4 ± 8.0	NS
Systolic blood pressure, mm Hg	108.2 ± 15.6	111.6 ± 14.0	110.1 ± 15.4	NS
Diastolic blood pressure, mm Hg	68.3 ± 11.2	71.3 ± 10.7	71.3 ± 16.9	NS
O2 saturation, %	97.5 ± 1.4	98.0 ± 0.9	97.7 ± 1.0	NS

 Level of significance was set to P < 0.05.

Breath CO

The CO level in exhaled breath for nonsmokers, smokers before smoking, and smokers after smoking was 0.3 ppm (SD 0.4), 3.33 ppm (SD 2.3), and 8.42 ppm (SD 6.6), respectively. The CO level in the smoking group after smoking was significantly increased (P = 0.0305). 

FTQ Questionnaire

The FTQ score for nonsmokers and smokers were 0 ± 0 and 3.33 ± 2.2, respectively. 

Discussion

This study showed that RVR to inhaled CO₂ was reduced in otherwise healthy subjects who smoke when compared with nonsmoking age-matched controls. Cigarette smoking reduced RVR in young healthy smokers. This finding was confirmed using the established CLBF technique. The prototype Doppler SD-OCT technique of the RTVue, however, showed no significant difference between smokers and nonsmokers, although trends for similar findings were present in the Doppler SD-OCT data. Retinal vascular reactivity has been widely studied using various gas provocation challenges including hyperoxia, hypercapnia, and carbogen (gas mixture of 95% O₂ and 5% CO₂). An increase in arterial CO₂ concentration was shown to vasodilate retinal vessels. ^15,49 Conversely, an increase in arterial O₂ concentration constricts the retinal vessels. ^12,13,50 In the present study, a sustained, stable, and standardized hypercapnic stimulus was used for gas provocation in young healthy individuals. Studies from our own lab previously showed that, a 15% increase in P_ETCO₂ from standardized baseline, while simultaneously maintaining isoxia increased the retinal arteriolar vessel diameter, blood velocity, and flow by 3.3%, 16.9%, and 24.9%, respectively. ¹⁵ The magnitude of change in hemodynamic parameters in this study was similar to those of Venkataraman and coworkers. ¹⁵  

Various other studies have assessed the RVR to hypercapnic gas provocation in healthy individuals. Dorner and coworkers ¹⁸ reported that, to a 21% increase in P_ETCO₂, retinal arteriolar and venular diameter increased by 4.2% and 3.2%, respectively. Sponsel and coworkers ²⁰ found a 26% increase in perimacular leukocyte velocity using blue field entoptic technique following inhalation of 5% CO₂. For a 10% increase in CO₂ retinal arteriolar and venular diameter increased by 2% and 1%, respectively, as measured using fundus photography. ⁵¹ Venkataraman and coworkers ¹⁴ reported 3%, 26%, and 35% increase in retinal arteriolar diameter, blood velocity, and flow, respectively, in response to 12% increase in P_ETCO₂ using the CLBF. In the present study retinal arteriolar diameter, blood velocity, and flow increased in response to normoxic hypercapnia by 4.1%, 16.7%, and 29.6%, respectively, using the CLBF. The magnitudes of change in hemodynamic parameters in this study were similar to those of previous studies. Studies done by our research group confirm that the magnitude of normoxic (i.e., “clamping” of PaO₂) hypercapnia induced using the computer-controlled gas blender are highly repeatable. ^14,15,17,49 The variability associated with the velocity measurement is virtually always greater than that of diameter or flow (Fig. 6), which explains why only diameter and flow showed significant differences between the groups. 

In this study, we used a prototype SD-OCT Doppler technology to measure the TRBF, venous area, and venous velocity changes in response to normoxic hypercapnia, in addition to CLBF measurements. In nonsmokers, normoxic hypercapnia increased the venous area, venous velocity, and total retinal blood flow by 7%, 18.1%, and 26%, respectively. This compares with values for the CLBF of +4.1%, +16.7%, and +29.6%, respectively. In interpreting the results from these two techniques, one needs to consider that they differ in terms of instrumental set-up as well as site of measurement of blood flow. The prototype Doppler OCT scans all branch retinal arterioles and venules using two concentric scans of 3.4- and 3.75-mm diameters around the ONH, and in terms of TRBF it uses only venular measurements, whereas the CLBF measures the blood velocity and diameter at a single location, in this case along the superior temporal arteriole, after the first bifurcation. For this reason, no correlation was found between the two techniques at the baseline (r ² = 0.004, P = 0.86), hypercapnia (r ² = 0.002, P = 0.89), and recovery (r ² = 0.05, P = 0.56); this emphasizes the fact that the two blood flow assessment techniques are fundamentally different. Nevertheless, the two methodologies proved to show remarkably similar results in the magnitude of vascular reactivity. The fact that a significant result was found using reANOVA, but then was not significant using Tukey's HSD can be explained by the exaggerated variability of the venous area measurement and the conservative correction of the Tukey's test for the possibility of Type 1 experimental error. 

In smokers, the CO levels were significantly elevated after smoking. Carbon monoxide and nicotine are two major constituents of cigarette smoke. Carbon monoxide has approximately 200 to 250 times greater affinity to hemoglobin than oxygen. The presence of CO-Hb reduces the O₂ carrying capacity of the erythrocytes. The CO-Hb dissociates very slowly in the blood due to the tight bonding of CO to hemoglobin, thus having a half-life of approximately 3 to 4 hours. ⁵² Nicotine, on the other hand, initiates catecholamine release through the activation of the sympathetic nervous system resulting in increased heart rate, blood pressure, and vasoconstriction. ⁵³  

Retinal vascular reactivity to flicker stimuli ^54,55 as well as to gas provocation ^24,25 was previously shown to be reduced in smokers. Several human and animal studies report that cigarette smoking is capable of inducing morphologic alterations to the vascular endothelium, ⁵⁶ and also alters the production of endothelial derived constricting and dilating factors. ^57–59 Our study reports that hypercapnia induced vasodilation is reduced in smokers. The enzyme nitric oxide synthase (eNOS) found in the vascular endothelium produces nitric oxide, which is thought to be a mediator for hypercapnia induced vasodilation. Nitric oxide binds to the iron atom of heme in guanylate cyclase and thereby increases intracellular cyclic guanosine monophosphate levels, in turn leading to a decrease in intracellular calcium levels, and hence vasorelaxation. Inhibition of eNOS activity has been shown to result in impaired hypercapnia induced vasodilation. ⁶⁰ Studies report that cigarette smoking impairs eNOS activity, thereby reducing the bio availability of nitric oxide resulting in impaired endothelial-dependent relaxation. ⁵⁸  

The limitations of the current study in terms of blood flow measurement technique is that, the prototype Doppler signal strength for the Optovue RTvue tended to be more variable compared with that of the CLBF. Another limitation in comparing RVR before (session 1) and after (session 2) smoking was that a few subjects demonstrated fatigue for session 2. Therefore, out of six CLBF measurements only three good quality measurements were considered for the analysis. 

In summary, this study used a novel gas provocation technique to investigate the retinal vascular reactivity in smokers and nonsmokers using CLBF and a prototype Doppler SD-OCT. Cigarette smoking reduced RVR in smokers. Total RBF and venous area measurements showed only a reducing trend in terms of RVR; however, the retinal arteriolar diameter and flow response was significantly reduced to normoxic hypercapnia in smokers. Whether or not reduced RVR reported in smokers reflects impaired endothelial function needs further investigation. 

Acknowledgments

The authors thanks Joseph Fisher (Department of Anesthesiology, University of Toronto, Ontario, Canada) for his advice on the use of the carbon monoxide monitor used in this study, and Richard Hughson (Department of Kinesiology, University of Waterloo, Ontario, Canada) for his overall advice and expertise as an advisory committee member of Kalpana Rose. 

Supported by grants from the Ontario Research Fund, Research Excellence award, Vision Science Research Program, University of Toronto, and an anonymous donor. 

Disclosure: K. Rose, None; J.G. Flanagan, Optovue, Inc. (F); S.R. Patel, None; R. Cheng, None; C. Hudson, Optovue, Inc. (F), Thornhill Research, Inc. (I), P