Retinal Vessel Diameters in Relation to Hematocrit Variation during Acclimatization of Highlanders to Sea Level Altitude

Abstract

purpose. To examine variations in retinal vessel diameters during acclimatization of native highlanders to normobaric normoxia at sea level.

methods. Fifteen healthy residents of the greater La Paz region in Bolivia (3600 m above sea level) were examined thrice over a 72-day period, after having traveled by airplane to Copenhagen, Denmark, near sea level.

results. In the study subjects, hematocrit decreased from 49.6% (day 2) to 45.9% (P = 0.0066, day 23) and 41.7% (P < 0.0001, day 72); from days 2 to 23, retinal vein diameter increased by 2.68% (P = 0.0079); whereas retinal artery and vein diameters were indistinguishable from baseline after 72 days. No funduscopic signs of retinopathy were observed. Arterial blood pressure remained stable throughout the study.

conclusions. Although a 16% reduction in hematocrit occurred between days 2 and 72 after arrival at sea level, the only significant excursion observed was that the diameter of the veins was larger at day 23 than at days 2 and 72. Retinal vessel diameters demonstrated a wide homeostatic range during acclimatization-driven hematocrit variation.

Changes in ambient oxygen tension have profound immediate effects on retinal blood flow and retinal vessel diameters. ¹ Oxygen supply is fundamental for retinal function, yet the middle layers of the retina are physiologically hypoxic and acidic, which is presumably evidence of lactate production and hence a significant component of glycolytic energy production. ^{2 3} Because retinal vessels have no innervation, retinal vasodynamics are driven by autoregulation. The retinal vessels also appear to be influenced by the dense choriocapillaris network of blood vessels on the outside of the retina, from which oxygen diffuses into the retina. ^{4 5}  

Autoregulation balances perfusion with stromal activity. Autoregulation of retinal blood flow can be demonstrated during hypoxia and high altitude simulation, where an acute increase in retinal vessel diameters and in macular blood flow can be demonstrated. ^{6 7} Correspondingly, normobaric hyperoxia is followed by acute retinal vessel contraction and decreased retinal blood flow. ^{1 7 8 9 10 11 12 13 14} Information based on short-term experiments examining acute physiological responses are of limited value, however, in relation to the pathophysiology of the most prevalent retinal diseases, because most of these diseases are chronic. 

Electroretinography studies indicate that the retina can adapt to a change in its systemic metabolic environment—for instance, a chronic change in glycemia, by altering its workload. ^{15 16} The mechanisms behind retinal adaptation to metabolic challenges are of interest because ischemia and hypoxia in various proportions is believed to be involved in the pathogenesis of retinal diseases, such as diabetic retinopathy and glaucoma. ^{17 18}  

While retinal vasodilation and hemorrhage can follow acutely after ascent to high altitude, healthy residents at high altitude do not display such characteristics, presumably because high hematocrit levels and other physiological characteristics have adjusted to hypobaric hypoxia. ^{19 20 21 22} Our recent electrophysiological observations in a subgroup of the study population reported on herein suggest that acclimatization also involves intrinsic retinal mechanisms. ²³  

In the present study, we have examined the effect on retinal vessel diameters of a long-lasting change in the metabolic environment of the retina in the form of descent from high altitude to sea level. This transition and the subsequent acclimatization has no reported adverse effect on the retina, suggesting that it provides an excellent opportunity to study the relation between retinal vessel characteristics, ambient oxygen tension, and key cardiovascular and hematologic parameters. 

Materials and Methods

Subjects

The study included 15 healthy subjects, all native residents of the Bolivian highland (altitude, 3100–4600 m), all in good systemic and ocular health, all in good nutritional condition without being obese, and all with a better than average socioeconomic background. No subject used any systemic or ocular medications. Subjects were examined 2, 23, and 72 days after traveling to Copenhagen, near sea level, on a two-leg flight from La Paz. The purpose of the travel was educational and the participants traveled together. The days of examination were the earliest after arrival, the latest before departure, and the one intervening day the subjects were available for examination. The study participants returned to Bolivia shortly after the 72-day examination and were unavailable for follow-up in Bolivia. The subjects remained in or around Copenhagen (altitude, 1–20 m) during the entire study period. Additional examinations included ocular lens autofluorescence ²⁴ and retinal electrophysiology. ²³ Written and oral informed consent was obtained from the participants after full explanation of the nature and possible consequences of the study. The study was approved by the medical ethics committee of Copenhagen County and adhered to the tenets of the Declaration of Helsinki. 

Methods

Digital fundus photography and computerized determination of retinal trunk vessel diameters were performed as previously described. ^{25 26} For each subject, a series of optic disc-centered, red-free 50° photographs (VISUPAC and FF 450plus Fundus Camera; Carl Zeiss Meditec AG, Jena, Germany) of the subject’s right eye were obtained after pupil dilation to a diameter of ≥7 mm with 10% phenylephrine hydrochloride and 1% tropicamide. At each visit, no less than four fundus images for each subject were recorded and used to determine vessel diameters with a custom-made computer program, with the results averaged for a minimum of four photographs from the same day to reduce the effect of pulsatile variation in retinal vessel diameters. ²⁷ A grid was placed on the 50° digital image. Only the retinal vessels crossing the annular belt spanning a distance of 0.5 to 1.0 disc diameters from the margin of the optic disc were analyzed. For every artery or vein crossing the annular belt, a grader chose the vessel segment that was deemed the most suitable for analysis, based on image quality, contrast, straightness of the vessel, absence of branching, and absence of vessel crossings. When possible, the full length of the vessel from 0.5 to 1.0 disc diameters was selected. When bifurcation or branching occurred within the annular belt, the trunk was preferred to its branches, unless the trunk segment was shorter than 80 μm. The program identified the six largest arteries and the six largest veins and summarized these values by calculating a central retinal artery equivalent (CRAE) diameter and a central retinal vein equivalent (CRVE) diameter according to formulas described elsewhere. ²⁸ The same six arteries and veins were analyzed when the procedure was repeated at a later date. Images were considered ungradable if the image was of poor quality (low contrast), as judged by the grader. Calibration of the image was performed assuming a uniform vertical optic disc head diameter of 1800 μm. Reference intervals (mean ± 2 SD) in healthy Danish subjects aged 20 to 46 years are CRAE, 136 to 196 μm, and CRVE, 211 to 282 μm. ²⁶  

Additional examinations included best corrected visual acuity, intraocular pressure, slit lamp biomicroscopy, ophthalmoscopy, arterial blood pressure manometry, and venous blood analyses for erythrocyte and hemoglobin concentrations, hematocrit, and erythropoietin concentration. Erythropoietin concentration was analyzed in eight subjects. Peripheral venous blood values were normalized for the effect of sex by scaling to the values of males. ²⁹ Mean arterial blood pressure (MAP) was calculated as follows: MAP = pressure_diastolic + ⅓(pressure_systolic − pressure_diastolic). Ocular perfusion pressure (OPP) was calculated as OPP = ⅔MAP − intraocular pressure. All subjects had best corrected visual acuity of 20/20 or better in both eyes and no detectable malformation, eye disease, or sequela of eye disease (Table 1) . 

Statistical analyses were performed with commercial software (SAS 9.1 software for Windows; SAS Institute, Inc., Cary, NC). Data were analyzed by standard t-tests, an initial ANOVA being used to test whether an overall significant change was found between visits, before a post hoc analysis was performed. The level of statistical significance was set at P < 0.05. 

Results

When examined 2 days after arrival at sea level, calculated central retinal vessel diameters (CRAE and CRVE) in the 15 native highlanders (Table 2)were found to be comparable with those of healthy Danish lowlanders, ²⁶ both in terms of artery (CRAE in highlanders, 164 ± 7.2 μm [mean ± SD], and in lowlanders, 165.8 ± 14.9 μm) and vein diameters (CRVE in highlanders, 244 ± 9.2 μm, and in lowlanders, 246.2 ± 17.7 μm). The highlanders’ vessel diameters remained comparable with those of the lowlanders throughout the study. 

On day 2 at sea level, measures of red blood cell concentration in the highlanders were found to be in the upper reference range of the healthy lowlanders: hemoglobin, 16.7 ± 0.5 g/dL (mean ± SD), and reference, 13.8–17.2 g/dL ²⁹ ; erythrocyte count, 5.4 ± 0.2 million cells/μL, and reference, 4.7–6.1 million cells/μL; and hematocrit, 49.6 ± 1.7% and reference 40.7–50.3%. Erythropoietin at 5.8 ± 1.3 mU/mL was within the normal range for lowlanders (0–19 mU/mL). From days 2 to 23, highlander hemoglobin decreased by 7.6% (P = 0.0004), hematocrit by 7.3% (P = 0.0066), and erythrocyte concentration by 6.9% (P = 0.0049). From days 2 to 72 the reduction in hemoglobin was 12.0% (P < 0.0001), in hematocrit was 16.0% (P < 0.0001), and in erythrocyte concentration was 15.6% (P < 0.0001). Compared with day 2, erythropoietin was 54.2% higher on day 23 (P = 0.014) and 50.7% higher on day 72 (P = 0.068; Table 3 ). 

Compared with day 2 after arrival, the mean retinal vessel diameters in highlanders on day 23 had increased by a nominal 2.70% for CRAE (P = 0.085) and a significant 2.68% for CRVE (P = 0.0079; Table 2 ). The nominal change in mean retinal vessel diameters from days 2 to 72 was an increase in CRAE of 1.05% (P = 0.56) and a reduction in CRVE of 0.78% (P = 0.49; Table 2 ). 

Ocular perfusion pressure did not change significantly during the observation period (P = 0.34 and P = 0.48 for comparison of day 32 vs. day 2 and day 72 vs. day 2; Table 2 ). Blood glucose, blood pressure, and intraocular pressure remained comparable with day 2 values throughout the study. No significant correlation was found between CRAE and CRVE changes during the period of observation and age, sex, blood glucose, blood pressure, intraocular pressure, hemoglobin, erythrocyte concentration, hematocrit, or erythropoietin. No retinal hemorrhage, thrombosis, cotton–wool spot or other sign of microangiopathy was observed. 

Discussion

The present study involved 15 healthy adult subjects who were examined shortly after moving from their normal habitat in hypobaric hypoxia to an environment of normobaric normoxia where they remained for 10 weeks, a sufficient period to undergo normalization of their erythrocyte concentration parameters. Despite a mean 16% reduction in hematocrit from days 2 to 72, no significant change was seen in retinal vessel diameters between these two observations. At day 23 CRVE was 2.68% bigger than at days 2 and 72, which was the only detectable sign of an imbalance in retinal vessel diameter homeostasis. 

This study is, to the best of our knowledge, the first to follow retinal vessel diameters and retinal morphology during an extended period of altitude acclimatization. Observations of the acute effect of ascent to high altitude and studies of acute experimental hypoxia indicate that major adaptive phenomena are involved in overcoming the adverse effects of hypobaric hypoxia and mountain sickness. Although such adaptation may involve numerous cardiovascular and pulmonary mechanisms as well as intrinsic retinal mechanisms, ²³ their net effect on retinal vessels has hitherto been unknown. The present study concerned descent, which appeared to be less challenging for the retina than ascent, one advantage being that effects such as intraretinal hemorrhage, which may invoke independent vasotropic mechanisms, were avoided. ^{6 8 20 30} Our results support that in the retina, a surplus of circulating erythrocytes within the range observed in this study, does not have adverse effects on healthy subjects. This finding is of interest for the management of polycythemia and anemia, including the therapeutic use of erythropoietin for bone marrow stimulation. 

In the present study, we did not address the immediate effects of the transition from high altitude to sea level or the accompanying increase in ambient oxygen tension. Acutely increased ambient oxygen tension leads to instant retinal vasoconstriction and decreased retinal blood flow, ^{1 7 9 10 11 12 13 14} but it is unknown whether this response persists for as long as the 2 days that elapsed between arrival in Copenhagen and the first examination of our study subjects. They did show relatively low erythropoietin levels when first seen 2 days after arrival at sea level, but this is consistent with previous studies showing a short half-life of circulating erythropoietin. ³¹  

The balance between dilating and contracting stimuli, and hence the net effect on CRAE and CRVE, was not different at days 2 and 72. One may speculate that reduction in oxygen transportation capacity that followed from the decrease in hematocrit was compensated by a concomitant increased blood flow that did not result in vasodilation because blood viscosity had decreased together with the hematocrit, but the demonstration of supernormal and gradually increasing electroretinographic function during acclimatization adds further to the complexity of the process. ²³ In mice exposed to hypobaric hypoxia for 3 weeks, a shift was found to occur in brain cortex metabolism toward lower aerobic and higher anaerobic enzymatic activity. ³² It appears likely that similar changes may occur in the retina. In the present study, where the subjects were observed after descent, a reduction in retinal glycolysis could be expected to occur, ^{3 32} which may explain the apparent involvement of the retinal stroma in acclimatization. 

When empiric formulas derived from laser Doppler velocimetry studies were used to describe the relation between retinal vessel diameters and retinal volumetric blood flow, ³³ the 2.68% retinal vein dilation from days 2 to 23 corresponded to an increase in retinal blood flow by a factor of 1.0268^2.84 = 1.078 (veins). This result is close to the concomitant 7.3% decrease in hematocrit, suggesting that from days 2 to 23 retinal oxygen consumption was stable, whereas blood flow increased sufficiently to compensate for the decrease in hematocrit. In contrast, the change observed from days 23 to 72 was qualitatively inverse, in that a reduction in oxygen binding capacity was followed by reduction of retinal vein diameter and presumably, a reduction in retinal blood flow. This finding could be evidence of an increase in retinal metabolic efficacy during this period. It should be noted however, that our estimates fail to take into account the changes in blood viscosity that follow the change in hematocrit. The mechanisms involved in retinal vasodynamics are of considerable interest, one reason being that they seem to fail in anemia in the presence of diabetes, where anemia is a documented risk factor for diabetic retinopathy. ^{34 35}  

Previous studies in which lowlanders underwent examination of retinal vessel diameter responses to a physical exercise test before and after spending 7 weeks in the Himalayas at up to 5850 m above sea level demonstrated postaltitude changes lasting at least 18 days, ⁸ supporting that acclimatization to high altitude has effects that last several weeks after return to near sea level. 

Our observations on CRAE and CRVE in relation to hematocrit reduction during acclimatization support that studies of acclimatization after altitude change should extend considerably longer than 10 weeks, to cover the entire physiological response to the change in altitude and oxygenation. 

 Supported by the Center for Biomedical Optics and New Laser Systems (www.biop.dk) under the Danish National Advance Technology Foundation; the Lundbeck Foundation Center for Neurovascular Signaling, Patient-Oriented Diabetes Research Career Award 8-2000-130 from the Juvenile Diabetes Research Foundation (ML), and the Velux Foundation.

 Submitted for publication November 25, 2008; revised February 19 and March 16, 2009; accepted June 18, 2009.

 Disclosure: P.K. Kofoed, None; B. Sander, None; G. Zubieta-Calleja, None; L. Kessel, None; M. Larsen, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Clinical Characteristics of Study and Reference Population

	Highlanders	References*	P
Sex (men/women)	10/5	94/116	0.10
Age (y)	41.3 (9.4)	35.3 (7.2)	0.011
MAP (mm Hg)	83.9 (8.7)	85.2 (8.6)	0.31
Blood glucose (mmol/L)	6.1 (0.7)	4.9 (0.5)	0.17
Altitude (m)	3633 (296)	Sea level
Intraocular pressure (mm Hg)	11.7 (2.8)	NA
Visual acuity (Snellen)	1.0–1.2	NA

 Data are presented as the mean (SD).

*  Reference values in healthy lowlanders. ²⁶

Table 2.

 

Retinal Vessel Diameters in Highlanders during Sea Level Acclimatization

	Day 2	Day 23	Day 72
CRAE (μm), mean (95% CI)	164.3 (157.1–171.5)	168.7 (160.8–176.6)	166.6 (159.3–174.0)
(Δ%), mean (95% CI)		2.70 (−0.40–5.80)	1.05 (−2.50–1.10)
P		0.085	0.56
CRVE (μm), mean (95% CI)	244.3 (235.1–253.6)	249.9 (241.1–258.7)	242.5 (234.4–250.6)
(Δ%), mean (95% CI)		2.68 (0.80–4.50)	−0.78 (−2.80–1.50)
P		0.0079	0.49

  P, paired t-test, data for days 23 and 72 versus those for day 2.

Table 3.

 

Hematology and Ocular Perfusion Pressure during Sea Level Acclimatization

	Day 2	Day 23	Day 72
Hemoglobin
Mean (g/dL)	16.7	15.4	14.7
95% CI	(16.2–17.2)	(14.9–15.9)	(14.1–15.3)
Δ%		−7.5	−12.0
P		0.0004	<0.0001
Erythrocytes
Mean (million cells/μL)	5.4	5.0	4.6
95% CI	(5.2–5.6)	(4.9–5.2)	(4.4–4.7)
Δ%		−6.9	−15.6
P		0.0049	<0.0001
Hematocrit
Mean (%)	49.6	45.9	41.7
95% CI	(47.9–51.3)	(44.4–47.5)	(40.1–43.2)
Δ%		−7.3	−16.0
P		0.0066	<0.0001
Erythropoietin
Mean (mU/mL)	5.8	8.9	8.7
95% CI	(4.5–7.1)	(7.6–10.2)	(6.8–10.7)
Δ%		54.2	50.7
P		0.014	0.068
Ocular perfusion pressure
Mean (mm Hg)	43.0	45.2	44.3
95% CI	(40.0–46.1)	(39.9–50.4)	(41.0–47.7)
Δ%		5.0	3.0
P		0.34	0.48

  P, paired t-test, data for days 23 and 72 versus day 2 data. Reference values in healthy lowlanders: hemoglobin (13.8–17.2 g/dL), erythrocyte count (4.7–6.1 million cells/μL), hematocrit (40.7%–50.3%), and erythropoietin (0–19 mU/mL). ²⁹

The authors thank Gustavo Zubieta-Castillo, Sr, and Poul-Erik Paulev for valuable comments and suggestions.