Missing correlation of retinal vessel diameter with high-altitude headache.

The most common altitude-related symptom, high-altitude headache (HAH), has recently been suggested to originate from restricted cerebral venous drainage in the presence of increased inflow caused by hypoxia. In support of this novel hypothesis, retinal venous distension was shown to correlate with the degree of HAH. We quantified for the first time retinal vessel diameter changes at 4559 m using infrared fundus images obtained from a state of the art Spectralis™ HRA+OCT with a semiautomatic VesselMap 1® software. High-altitude exposure resulted in altered arterial and venous diameter changes at high altitude, however, independent of headache burden.

Introduction

The cause of high-altitude headache (HAH) is unknown. The idiosyncratic nature of HAH during high-altitude exposure and consequently in acute mountain sickness (AMS) was first hypothesized to result from an increase in intracranial pressure (ICP) in the form of cerebral edema secondary to high-altitude hypoxemia.1 However, subsequent studies investigating HAH in AMS and the presence of increased ICP have found little evidence to support this hypothesis.2–4 Furthermore, recent studies suggest that the degree of hypoxia in AMS is not able to cause sufficient cerebral edema resulting in respective symptoms typical of AMS including headache.2,5

Although it is well known that high-altitude exposure leads to a marked increase in cerebral blood flow,6 cerebral –and subsequently retinal – venous physiology has until recently been thought to be of little consequence with regard to disease pathophysiology of HAH. However, as various similarities between clinical syndromes resulting from restricted cerebral venous outflow and HAH may exist,7,8 recent studies have started to focus on gaining a better understanding of the cerebral inflow/outflow mismatch at high altitude with regard to HAH.

Wilson et al.9,10 have postulated and first shown that distension of large cerebral veins due to restricted cerebral venous outflow in the presence of increased inflow may account for HAH. This finding was supported by a positive correlation between retinal venous distension and the severity of HAH on preceding days during trekking to an altitude of 5300 m.

Our study investigated retinal vessel diameter in correlation with HAH burden and AMS in order to investigate the proposed pathophysiological mechanisms of HAH using – for the first time – infrared fundus imaging and semiautomatic assessment of vascular diameters in a well-established setting for studying AMS.11

Methods

Study design and clinical assessment of high-altitude exposure

Eighteen healthy subjects (14 volunteers and four investigators; seven females, 11 males; age 25–54 years) were exposed to the research facility of the Capanna Margherita at 4559 m (CM; Valais Alps, Italy) to assess retinal vessel changes with regard to AMS/HAH as described previously.12,13 Ascent to CM was completed within 24 h with one overnight stay at the Capanna Gnifetti 3647 m. All subjects spent three nights at the research facility at high altitude before descending back to low altitudes (341 m) on day 4. Before and after high-altitude exposure, subjects were asked to spend at least 2 weeks below 2000 m to exclude confounding effects due to previous altitude exposure (baseline = before and baseline 2 = after descent from high altitude). Ophthalmological exclusion criteria included current or positive history of conditions affecting the posterior chamber (e.g., diabetic retinopathy, retinal artery or vein occlusion, age-related macular degeneration, history of retinal surgery or glaucoma) on either eye and/or optical opacities limiting imaging quality.

Assessment of AMS, HAH, and clinical parameters

For assessment of AMS and intensity of HAH, the Lake-Louise (LL) and the AMS-cerebral scores (AMS-C) of the Environmental Symptom Questionnaire (ESQ III) were used once at each baseline examination and twice daily at high altitude. For both scores, subjects answered questions about the severity of symptoms of AMS including HAH; for LL symptoms are graded 0 = no, 1 = mild, 2 = moderate, and 3 = severe symptoms, and for the AMS-C Score symptoms are graded from 0 = no to 5 = extremely severe symptoms, respectively. Consistent with previous literature, AMS was assumed when both scores met the cut-off criteria set for LL ≥ 5 and AMS-C ≥ 0.70 in the presence of headache.2,12,14 Measurements of oxygen saturation (SpO2) and heart rate (HR) were performed once at each baseline and at high altitude in the morning before getting up and after >5 min at rest in the evening with a finger pulse oxymeter (oxy control 4c®; Geratherm Medical AG, Geschwenda, Germany). Values were recorded after 1 min of steady measurement.

Retinal vessel analysis

Infrared fundus images were obtained using two identical Spectralis™ HRA+OCT (Heidelberg Engineering, Heidelberg, Germany) devices at baseline and high altitude in the morning after arrival as described previously.12,15 Careful calibration of the laser light source of the Spectralis™ HRA+OCT device at high altitude was performed after helicopter transport (Air Zermatt AG Heliport, Zermatt, Switzerland). Assessment of retinal artery and vein diameters was performed using the semiautomatic VesselMap 1® software (Imedos Systems, Jena, Germany) to quantify vascular changes in one disk diameter peripheral of the optic disk margin between baseline and high altitude (Fig.1A).

Figure 1

Vessels diameter increase at high altitude. (A) Infrared fundus images obtained from a Spectralis® (Heidelberg Engineering, Germany) at baseline (left) and at high altitude (right). VesselMap 1 analyzer (Imedos Systems, Germany) was used to quantify the change in retinal vessel diameter one disk diameter from the optic disk margin in 18 healthy subjects for arteries and veins. (B) Intraindividual differences in vessel diameter for arteries (top) and veins (bottom) in the morning after arrival at high altitude compared to baseline recordings before exposure revealed a significant increase in vessel diameter. Recordings after high-altitude exposure revealed no significant differences in vessel diameter compared to preexposure values.

Statistical analysis

A longitudinal analysis between altitude and baseline measurements was performed by multivariate analysis of variance (MANOVA) for repeated measures (significance level of P < 0.05). Pearson’s correlation coefficient assessed a possible correlation between retinal vessel diameter changes and scores of HAH and AMS as assessed by LL and AMS-C as well as clinical parameters of SpO2 and HR the morning after arrival at high altitude.16 Data are shown in terms of intraindividual differences (value during altitude exposure – value at baseline) with 95% confidence limits at each time point.

The study was performed after volunteers were informed about the nature of the study protocol and after written informed consent in accordance with the tenets of the Declaration of Helsinki 1975 (1983 revision). It was approved by the local IRB (Ethik-Kommission der Medizinischen Fakultät/Universitätsklinikum Tübingen, IEC project number: 258/2010B01).

Results

Clinical parameters, AMS, and HAH

Incidence of AMS (LL ≥ 5 and AMS-C ≥ 0.7) was 56% (n = 10/8) in the morning after arrival at high altitude. The overall scores for AMS were LL = 5.2 ± 2.6 (mean ± SD) and AMS-C = 1.1 ± 0.8; mean LL for subjects with AMS (AMS+) = 6.9 ± 1.5 and without (AMS−) = 3.1 ± 1.7 and the average AMS-C for AMS+ = 1.7 ± 0.5 and AMS− = 0.4 ± 0.3, respectively. Average headache scores in the morning after arrival at high altitude were LLheadache = 1.3 ± 0.6 and AMS-Cheadache = 2.6 ± 1.2; mean LLheadache for AMS+ = 1.7 ± 0.5 and AMS− = 0.9 ± 0.4; and AMS-Cheadache for AMS+ = 3.3 ± 0.8 and AMS− = 1.6 ± 0.9, respectively. Clinical parameters for peripheral oxygen saturation decreased to SpO2 = 73 ± 6% in all subjects in the morning after arrival at high altitude compared to baseline = 98 ± 1.%; mean SpO2 for AMS+ = 72 ± 6% and AMS− = 75 ± 6% at high altitude (P > 0.05); HR increased to 83 ± 10/min compared to baseline = 60 ± 7/min; mean HR for AMS+ = 85 ± 9/min and AMS− = 80 ± 10/min in the morning after arrival at high altitude (P > 0.05). All changes in clinical parameters were completely reversible after descent to low altitudes.

Retinal vessel analysis and correlation with AMS and HAH

Mean arterial (AD) and venous diameters (VD) at baseline before exposure were ADbaseline = 122.7 ± 14.8 μm and VDbaseline = 148.0 ± 15.3 μm. The mean of ADaltitude significantly increased by 13.6 ± 19.8 to 136.4 μm and VDaltitude by 26.7 ± 22.1 to 171.7 μm in the morning after arrival at high altitude (P < 0.05; Fig.1B). Average ADaltitude in AMS+ = 133.4 ± 18.9 μm and in AMS− = 140.0 ± 21.7 μm; average VDaltitude in AMS+ = 168.2 ± 17.3 μm and in AMS− = 176.2 ± 27.6 μm. All changes were completely reversible upon descent as shown at baseline 2 (P > 0.05; Fig.1B). As anticipated by the values listed above, Pearson’s coefficient showed no correlation between increased mean AD or VD for both LL and AMS-C scores of AMS (Fig.2A and B) or HAH (Fig.2C and D). In addition, changes in mean AD and VD between baseline and high altitude did not correlate with clinical parameters of SpO2 and HR (only a nonsignificant trend between vessel diameter and SpO2 was seen in the form of a negative correlation (AD vs. SpO2: r = −0.39, P = 0.11; VD vs. SpO2: r = −0.26, P = 0.29).

Figure 2

Missing correlation of vessel diameter changes to high-altitude headache and acute mountains sickness. Pearson’s correlations between arterial (top panel) and venous (bottom panel) vessel diameters during high-altitude exposure and scores for acute mountain sickness (A and B) and its respective headache scores (C and D). None of the correlation reached statistical significance. AMS-C, acute mountain sickness cerebral score; LL, Lake Louise score; AMS+, acute mountain sick; AMS−, healthy subjects.

Interpretation

A significant increase in central retinal vessel diameter for both arteries and veins occurred in response to high-altitude exposure in healthy subjects. Consistent with previous literature this increase was greater in veins compared to arteries.17 Our findings demonstrate the increase in cerebral blood flow due to hypoxia as a physiological response to high-altitude exposure in nonacclimatized humans. However, in contrast to the recently published data by Wilson et al.,9,10 who reported a correlation between HAH or AMS scores during ascent and retinal vessel diameters measured 2–3 days after arrival at 5300 m, we found no such correlation when assessing retinal vessel diameters and HAH or AMS scores simultaneously at 4559 m. In our opinion, the type of exposure protocol with immediate assessment of HAH/AMS scores and respective vessel analysis is of great importance with regard to a valid outcome. In addition, the assessment of AMS and HAH remains highly subjective presenting a potential limitation of the study. Nevertheless, our findings do not support the concept of restricted cerebral venous outflow in the presence of increased inflow accounting for HAH or AMS when assuming that central retinal veins may reflect gross intracerebral venous changes because of draining into the sinus cavernosus.