What Happens to Intraocular Pressure at High Altitude?

Abstract

purpose. To investigate changes of intraocular pressure on ascent to high altitude.

methods. The Apex 2 medical research expedition provided the opportunity to measure intraocular pressure (IOP) and central corneal thickness (CCT) in 76 healthy lowlanders. They all arrived in La Paz, Bolivia (altitude, 3700 m), where they spent 4 days before being driven more than 2 hours to the Cosmic Physics Laboratory at Chacaltaya (5200 m) where they stayed for 7 days. IOP and CCT were measured with a hand-held tonometer and ultrasound pachymetry on the first, third, and seventh days at 5200 m. Pre- and postexpedition CCT and postexpedition IOP readings at sea-level were also measured.

results. IOP increased significantly from baseline after acute exposure to altitude before returning to baseline with time. IOP at baseline, change in IOP from baseline, and IOP at altitude did not predict symptoms of acute mountain sickness (AMS) or development of high-altitude retinopathy (HAR).

conclusions. Acute exposure to altitude caused a statistically significant but clinically insignificant increase in IOP. This finding may be partially explained by the change in CCT. IOP returned to baseline levels and possibly lower with prolonged exposure to altitude. Changes in IOP at altitude are not predictive of symptoms of acute mountain sickness (AMS) or development of high-altitude retinopathy (HAR).

Intraocular pressure (IOP) at high altitude has been the subject of controversy for many years. In 1918 Wilmer and Berens ¹ measured IOP in 14 aviators in a hypobaric chamber but found no significant changes. More recently, some groups have found decreased IOP, ² others have found increased IOP, ^{3 4} normal IOP, ^{5 6} and even a reduction in IOP that occurred within hours of ascent and recovered during acclimatization. ^{7 8} The mechanism of these changes remains mysterious. However, it has been suggested that aqueous humor dynamics and the controversial relationship between intraocular and intracranial pressures are involved. ^{9 10} Aqueous is produced by both active and passive processes in the nonpigmented ciliary epithelium. Ultrafiltration and diffusion are passive processes influenced by systemic factors such as BP, IOP, and the venous osmotic pressure. Aqueous drainage is partially an active process but is also influenced by episcleral venous pressure. Ascent to altitude produces a characteristic physiologic response which may alter IOP by altering one or several of these variables with possible culprits including vasogenic cerebral edema, ¹¹ vascular endothelial damage, ¹² respiratory alkalosis, ⁷ altered cerebrovascular autoregulation, ¹³ sustained vasodilatation, and increased cerebral capillary pressure. ¹⁴  

IOP measurement can be complicated by anatomic, instrumental, and physiological sources of error. The effect of altitude on IOP and its significance may be masked by several factors. Failure to correct for corneal thickness (which increases significantly at altitude ¹⁵ ) can artificially inflate IOP readings. Furthermore, it is difficult to generalize from studies with a small number of subjects, measured in difficult circumstances by different methods, with the added problem of the confounding effects of exercise and cold exposure on different expeditions. 

The Apex 2 expedition to Chacaltaya in Bolivia was undertaken to address these problems. It had a large number of subjects and a fast, nonexertional ascent profile, to reduce the effect of exercise. The cosmic physics laboratory provided a controlled, indoor environment for a field study to reduce the effect that environmental wind and cold could have on readings. Corneal thickness was measured in addition to IOP. 

Methods

One hundred and four healthy subjects (56 men, 48 women) with a median age of 21 years (range, 18–60) took part in the expedition. The tenets of the Declaration of Helsinki were adhered to and the research was approved by the Lothian Research ethics committee. At a pre-expedition meeting, informed consent was gained from subjects. No participant had a history of ocular disease. Measurements of IOP and CCT were always taken in the early afternoon, to minimize the effects of diurnal variation. Measurements were taken in one eye per subject under topical anesthesia with benoxinate. A hand-held tonometer (Tono-Pen XL; Medtronic-Solan, Jacksonville, FL) was used to measure IOP, because of its proven efficacy in field trials. ¹⁶ Corneal thickness was measured with an ultrasound pachymeter (SP-2000; Tomey, Nagoya, Japan). Both the tonometer and pachymeter were calibrated to ensure accuracy before each testing session. IOP measurements were repeated three times at each testing session (three averages of three) and CCT measurements were taken nine times at each testing session. Subjects were flown to La Paz (altitude, 3700 m), Bolivia, where they acclimatized for 4 days before being driven over 2 hours to the Cosmic Physics Laboratory at Chacaltaya (5200 m). Testing took place at sea level before and at least 1 month after all subjects had returned to sea level. Testing on the expedition took place on the first, third, and seventh days at the laboratory. After the seventh day, the subjects were driven back to La Paz (the ascent profile is shown in Fig. 1 ). In La Paz (3700 m), subjects were advised to avoid any strenuous activity and were not allowed to descend below 2000 m. At the laboratory (5200 m), subjects were not formally restricted in their activity but were advised to rest for the first 2 days. Afterward, they were allowed to exercise moderately if they wished but were not allowed to descend below 5000 m. Cigarette smoking and alcohol consumption were prohibited for the duration of the expedition. Acute mountain sickness (AMS) scores were recorded using the Lake Louise scale. ¹⁷ Systolic and diastolic blood pressures (BP) were measured by one observer with a mercury sphygmomanometer. The subjects also took part in a double-blind randomized controlled trial to compare the efficacy of sildenafil and antioxidants versus placebo in preventing AMS. No other drugs were allowed unless one of the expedition doctors had given permission. None of the subjects was allowed to take acetazolamide, which is known to lower IOP. Twenty-four subjects did not consent to take part and 4 did not complete any testing sessions at altitude due to severe AMS, which left 76 subjects (40 men and 36 women) with a mean age of 22 years (range, 18–60, SD 5). Seventy-six (100%) completed the first altitude testing session; 70 (92%) completed both the first and second altitude testing sessions, and 63 (83%) completed all three altitude testing sessions. Data in each testing session was collected in only 53 subjects. Reasons for missing testing sessions are given in Table 1 . 

Results

Analysis of the three subgroups in the drug trial showed that neither sildenafil nor the antioxidant preparation had any effect on IOP or CCT when compared with placebo. 

The AMS incidence was 72.4% on the first, 34% on the second, and 19% on the third altitude testing sessions, respectively. The incidence of HAR was 2.6% on the first, 5.7% on the second and 12.7% on the third altitude testing sessions respectively, with an overall incidence over the course of the expedition of 14.5%. 

There was no significant change in logMAR (logarithm of the minimum angle of resolution) visual acuity between sea level and 5200 m. 

There was an increase in recorded IOP from sea level to altitude. The mean increased from a sea level pressure of 11.35 ± 2.9 to 12.42 ± 3.2 mm Hg on day 1 at 5200 m. IOP then fell to 12.07 ± 2.7 mm Hg on day 3 and 10.79 ± 2.9 mm Hg on day 7 at 5200 m (Fig. 2for the illustration of this trend). 

These differences were analyzed by repeated-measures ANOVA. As Mauchly’s test of sphericity was significant (W(5) = 0.785, P = 0.031), the Greenhouse-Geisser correction factor was applied and the differences were not significant (F(3136) = 2.8, P = 0.052). 

The analysis was rerun correcting the IOP for change in CCT from baseline, using two different correction factors, as recommended in the literature. ^{18 19} After the first (0.10 mm Hg/10 μm of CCT) and second (0.31 mmHg/10 μm of CCT) correction factors were applied, slightly lower values were obtained (Table 2) . When the first correction factor was used, the analysis (Greenhouse-Geisser) produced similar results (F(3136) = 2.67, P = 0.058). The results using the second correction factor were significant (F(2125) = 4.41, P = 0.001). Post hoc pair-wise comparisons (least significant difference) were used to find where these differences lay. There was a significant difference between the following comparisons: day 1 at altitude and day 7 at altitude (P = 0.016), day 3 at altitude and day 7 at altitude (P = 0.012), and sea level and day 7 at altitude (P = 0.005). 

Only 53 subjects completed testing on all days and were therefore included in the ANOVA analysis (listwise analysis). (For reasons for withdrawal see Table 1 ). This reduced the power and may have resulted in a type 2 error. To maximize power given the borderline significance of the ANOVAs, we used paired t-tests to compare the data at each measurement point, as this increased the subject numbers included in the analysis (Table 3) . These tests indicated that there was a significant difference between the following measurements: day 1 at altitude and sea level (t(68) = 2.52, P = 0.014); day 1 at altitude and day 7 at altitude (t(57) = 2.18, P = 0.034); day 3 at altitude and sea level (t(63) = 2.27, P = 0.026); and day 3 at altitude and day 7 at altitude (t(59) = 2.43, P = 0.018). Application of correction factor 1 only altered this trend by removing the significant difference between day 3 at altitude and sea level. Application of correction factor 2, however, made the difference between days 1 and 3 at altitude and sea level nonsignificant, while making the day 7 at altitude reading significantly different from all other readings. There was no correlation between change in CCT and change in IOP at altitude (Pearson bivariate correlation). 

There was an increase in recorded systolic and diastolic BP from sea level to altitude that was sustained at altitude. There was no correlation between change in either systolic or diastolic BP and change in IOP at altitude (Pearson bivariate correlation). 

An analysis was performed to assess whether there is a relationship between IOP and acute mountain sickness (AMS) using the Kruskal-Wallis test. The Lake Louise AMS scale was split into two categories: no AMS (scale scores of less than 3) and AMS (scores ≥3). Change in IOP was not related to dichotomized AMS scores and therefore is unlikely to be a significant predictor of AMS. 

A similar analysis was performed (also using the Kruskal-Wallis test) to assess whether there is a relationship between IOP and HAR. There was no difference in IOP at any level between those affected by HAR and those unaffected. 

Discussion

This study is the largest trial to date on the changes of IOP at altitude and demonstrated some statistically significant changes in IOP, suggestive of a definite effect of altitude. However, this effect is small and does not appear to be of clinical significance. The significant changes in IOP only became obvious once paired t-tests were used instead of a repeated-measures ANOVA, which excluded the subjects list-wise, thereby reducing sample sizes. Correcting the IOP for changes in CCT removed the significance of increases in IOP and revealed a significant decrease in IOP at day 7 at altitude. 

The Fick equation states that IOP measurement is dependent on corneal thickness. ²⁰ As CCT increases at altitude ¹⁵ alterations in measured IOP may be artifactual. Previous studies have shown that a thickened cornea will give artificially high IOP readings and a thin cornea will cause artificially low IOP readings. ²¹ A recent study appears to confirm these results by measuring central corneal thickness and IOP both by applanation tonometry and directly with an intracameral probe. ²² This is still a subject of controversy and the optimum correction algorithm is still in doubt. ²³ Nevertheless, having corrected our IOP measurements for CCT according to a range of values suggested by the literature ^{18 19} a significant increase in IOP at altitude was no longer detected (using paired t-tests) but a significant decrease with time was unmasked (in both a repeated-measures ANOVA and paired t-tests). It therefore appears that CCT is an important variable when measuring IOP at altitude even if there are no clinically significant changes in either variable. CCT affects different tonometry techniques to a variable degree and may be an even more significant factor in studies using different techniques. 

Previous confusion over IOP at altitude may stem from the change in exercise levels from baseline or because of environmental conditions under which testing took place (cold air is known to decrease IOP ²⁴ ). Exercise has been shown to reduce IOP as part of a hypotensive and sympathetic response. ²⁵ Exercise conditioning can also significantly reduce baseline IOP. ²⁶ This factor may have confounded previous studies which reported a reduction in IOP at altitude, ^{2 7 8} as subjects at altitude often exercise more than normal and become increasingly fit during a sojourn at high altitude. Our subjects were driven to altitude and therefore exercise and exertion were less likely to be factors affecting their IOP. This may explain why we noted an initial increase in IOP and then a decline once activity levels increased after prolonged exposure. Our findings contradict the finding of a nonsignificant reduction in IOP reported on fast, nonexertional ascent to an altitude of 3048 m in a plane, ⁶ but are consistent with a reported increase noted in a hypobaric chamber study simulating an altitude of 9144 m. ³ This increase may only occur at higher altitudes. The increase observed in this study is in direct contrast to a decrease in IOP noted during exposure to hyperbaric oxygen therapy at 2.5 atmospheres where a significant decrease in IOP was noted. ²⁷  

The hand-held tonometer (Tono-Pen; Medtronic-Solan) proved that it is a good tool for field trials ¹⁶ and concern about its tendency to underestimate IOP above 30 mm Hg and overestimate IOP under 9 mm Hg ²⁸ is unlikely to have affected this study, as IOP values tended to be within the normal range of 11 to 21 mm Hg. The ease of measuring IOP at altitude has led to the suggestion that it may be a useful noninvasive screening test in high-altitude climbers to inform about acclimatization and alert one to the risk of AMS and its severe consequences. ^{8 29} This study failed to demonstrate a predictive relationship between IOP and AMS or HAR in a large number of subjects. Eight of these needed to be evacuated due to the severity of their AMS and 14.5% had signs of HAR. 

This study suggests that the initial increase in IOP observed may be due to the increase in corneal thickness as correcting for this resulted in the disappearance of this trend. However, there was no correlation between change in CCT and change in IOP. Although blood pressure is known to affect IOP, ³⁰ this study failed to show any relationship between acute changes in BP at altitude and change in IOP. The reason for the observed increase in IOP may include both changes in CCT and exercise but a complete understanding of the mechanisms behind these changes remains distant. Although IOP appeared ineffective at predicting severity of AMS or HAR, further research is necessary to evaluate changes in these parameters in the most severely affected and to assess whether these changes are of significance when advising those who have ocular hypertension and glaucoma about traveling to regions of high altitude. 

 Supported by student travel prizes from The Wellcome Trust and The Royal College of Ophthalmologists.

 Submitted for publication October 13, 2006; accepted January 18, 2007.

 Disclosure: J.E.A. Somner, None; D.S. Morris, None; K.M. Scott, None; I.J.C. MacCormick, None; P. Aspinall, None; B. Dhillon, 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.

 

Figure 1.

 

Ascent profile of volunteers.

Table 1.

 

Numbers of Subjects Missing Different Testing Sessions and Reasons

Measurement Point	Expedition at 5200 m					Baseline Sea Level
		Day 1	Day 3	Day 7
Failure to attend	3	1	2	4
Evacuated for AMS	0	5	4	0
Evacuated for diarrhea and vomiting	0	1	1	0
Voluntary descent	0	0	2	0

Figure 2.

 

The changes in recorded IOP ± SEM.

Table 2.

 

The Changes in Recorded IOP on the Expedition ± SD

Measurement Point	Expedition at 5200 m					Baseline Sea Level
		Day 1	Day 3	Day 7
Mean IOP	12.42	12.07	10.79	11.35
SD IOP	3.16	2.73	2.86	2.93
Mean IOP (corrected for CCT1)	12.23	11.86	10.55	—
SD IOP (corrected for CCT1)	3.16	2.76	2.89	—
Mean IOP (corrected for CCT2)	11.83	11.42	9.82	—
SD IOP (corrected for CCT2)	3.21	2.81	3.23	—

Table 3.

 

Paired Comparisons between Measurement Points both Corrected and Uncorrected

Paired Comparisons	Original Data			Correction Factor 1			Correction Factor 2
		t	df	t	df	t	df
IOP 5200 m day 1 vs. IOP sea level	2.524*	68	−2.002*	68	−0.906	68
IOP 5200 m day 3 vs. IOP sea level	2.273*	63	−1.711	63	−0.542	63
IOP 5200 m day 7 vs. IOP sea level	−0.868	55	1.555	55	3.131, **	55
IOP 5200 m day 1 vs. IOP 5200 m day 3	0.644	65	0.696	65	0.796	65
IOP 5200 m day 1 vs. IOP 5200 m day 7	2.177*	57	2.3*	57	2.736, **	57
IOP 5200 m day 3 vs. IOP 5200 m day 7	2.428*	59	2.397*	59	2.744, **	59

*   P < 0.05.

**   P < 0.01.

The authors thank the volunteers and researchers who took part in the Apex 2 expedition: Instituto de Investigaciones Fisicas, Universidad Mayor de San Andres; and the Instituto Boliviano de Biología de Altura, La Paz, Bolivia; Roger Thompson for organizing the expedition; and Alistair Simpson, Anna Thompson, Shaima Elnour, David Cavanagh, and Jill Inglis for numerous contributions.