Abstract
Aim: The aim was to identify most relevant risk factors of high-altitude headache within a broad mountaineering population through a prospective, observational, rater-blinded study.
Methods: A total of 506 mountaineers were enrolled after their first overnight stay in one of seven alpine huts between 2200–3817 m. Structured interview including information on mountaineering histories, caffeine intake, smoking habits, alcohol consumption, intake of medication, rate of ascent, physical fitness, the level of exertion and the amount of fluids intake at the day of ascent were recorded along with a standardized medical examination.
Results: High-altitude headache occurred in 31% of study participants. Logistic regression analysis revealed a migraine history, low arterial oxygen saturation, high ratings of perceived exertion and fluid intake below 2 l to be independent risk factors for the development of high-altitude headache.
Conclusion: Given the high prevalence, high-altitude headache is a relevant medical condition and a better understanding of risk factors has important impact and may facilitate patient behaviour and future study design.
Introduction
Headache is by far the most frequent symptom experienced by high-altitude sojourners and constitutes the key symptom of acute mountain sickness (1–4). High-altitude headache has been defined by the International Headache Society as headache occurring within 24 h after ascent >2500 m which resolves within 8 h after descent (5). It has also been stated that 80% of people who ascend to high altitudes will develop high-altitude headache. Thus, high-altitude headache represents an important public health problem afflicting millions of high-altitude visitors annually (6,7). Nevertheless, the underlying pathophysiological mechanisms and even more the potential risk factors are still poorly understood. Hypobaric hypoxia has been proposed to elicit neurohumoral and haemodynamic responses resulting in overperfusion of microvascular beds and consequent brain swelling, the activation of the trigeminal vascular system and/or sensitization of intracranial pain receptors (2,3,6,7). The most likely reason for the existing uncertainty may be due to the reason that high-altitude headache only describes the symptom of possible different disease entities (1,4). Recently, Serrano-Dueñas (4) identified three types of headache occurring during acute exposure to high altitude: (i) headache as predominant symptom by the set of acute mountain sickness; (ii) headache independent of acute mountain sickness solely due to hypoxia; and (iii) altitude triggered migraine attacks. However, in contrast to hypoxia and the history of migraine, the set of acute mountain sickness might cover several risk factors. For example, exercise, acid-base, fluid and electrolyte disturbances and low blood glucose levels have been shown to provoke headache at low and high altitude as well (4,8–12). Thus, we hypothesised that high-altitude headache is the consequence of a complex interaction between the individual susceptibility and behaviour patterns, the high-altitude environment and exercise. A more comprehensive knowledge on the risk factors involved in high-altitude headache would be of importance for the prevention of high-altitude headache and may facilitate patient behaviour and future study design. Thus, the objective of this study was to identify the most relevant risk factors of high-altitude headache within a broad mountaineering population.
Subjects and methods
Prospective data collection on the incidence of high-altitude headache has been performed in seven alpine huts between 2200–3817 m in the Eastern and Western parts of the Alps over a period of 3 months.
Questionnaire and measurements
Structured questionnaires were used to obtain information on anthropometric data, altitude of permanent residence, caffeine intake (regular vs irregular), smoking habits (smoker vs non-smoker), alcohol consumption during the present ascent (type and amount converted in grams), pre-existing diseases including migraine and other primary headaches, intake of prophylactic and therapeutic medication, symptoms of acute mountain sickness on previous exposures, rate of ascent, previous exposures (annual and in the preceding 2 months), self-assessed physical condition (very good, good, moderate, poor), level of exertion when reaching the hut (very low, low, moderate, heavy) and the amount of fluid intake on the day of ascent. Regular migraine was defined as episodic migraine in accordance with the IHS guidelines with a minimum of one attack per month prior to enrolment. The mountaineering experience is expressed by the number of days spent above 2000 m annually.
The questionnaires were available in German, English and French, allowing the subjects to answer in their native language. Headache was scored on a 4-point scale (0 = none, 1 = mild, 2 = moderate, 3 = severe) as usually included in the acute mountain sickness assessment by the Lake Louise Score (2,13). Of course, our subjects taking NSAIDs because of headache were classified to have headache. Heart rate (HR) and oxygen saturation (SaO2) were measured with a pulse oximeter (Pulsox-3i, Minolta, Osaka, Japan) at rest in a sitting position while the subjects were filling out the questionnaires, the interviewer being blinded for the respective variables.
Subjects
All mountaineers who ascended to the huts and stayed overnight were asked to participate in the study. The survey took place in the morning before the mountaineers set off on their hike. Only 15% refused to participate. Mountaineers who agreed to participate underwent a short instruction and explanation of the purpose of the study. Finally, a total of 506 mountaineers were enrolled.
Statistical analysis
Statistical analyses were performed with the statistics software SPSS v15.0 for Windows. To examine risk factors associated with high-altitude headache, Student’s t-test was used to compare means of continuous data, the Mann–Whitney U-test was applied to evaluate differences between ordinal or not normally distributed data, and Chi-squared test for analysis of frequencies. Variables with a P-value of 0.20 or less were defined as potentially relevant risk factors for high-altitude headache. After dichotomizing, these variables were further subjected to a step-wise logistic regression analysis with high-altitude headache as the binary dependent variable. Differences were considered statistically significant at P < 0.05.
Results
Demographic data and risk factors in mountaineers with (i.e. headache 1, mild and >1, moderate to severe) and without (i.e. headache 0, none) high-altitude headache after ascending to high altitude
P-values for differences between patients without and with high-altitude headache, respectively.
Data are presented as mean (±SD) or frequencies (percentages).
AMS, acute mountain sickness; ASL, above sea level: SaO2, arterial oxygen saturation.
The use of medication with the potential to confound the interpretation of our results was low. Only 1.1% of the study participants used β-blockers, ACE-inhibitors or diuretics on a regular basis and 1.8% used acetazolamide for acute mountain sickness prophylaxis. Only the intake of pain medication differed significantly between subjects with (11.4%) and without headache (4.6%; P < 0.05).
Adjusted odds ratios (95% confidence intervals) regarding the prevalence of risk factors among mountaineers who developed high-altitude headache after ascending to high altitude compared to those without high-altitude headache
Nagelkerke R2 = 0.29.
SaO2, arterial oxygen saturation; RPE, rate of perceived exertion.
Discussion
High-altitude headache is a frequent symptom leading to significant burden to high-altitude sojourners occurring in more than 30% of our study cohort. This study revealed that high-altitude headache is associated with four independent risk factors, i.e. history of migraine, low SaO2 values, high rating of perceived exertion and low fluid intake. As we hypothesized, not all of these factors are necessarily related to high altitude. Only the low SaO2 values are direct consequences from the hypoxic environment whereas the other factors are known to be associated with headache at low and high altitude reflecting the complex interactions in the development of high-altitude headache.
Speculating on the risk factors of high-altitude headache, one has to keep in mind that the incidence of high-altitude headache increases when arterial oxygen saturation and associated oxygen partial pressure decline with increasing altitude (2,15,16). Thus, it might be assumed that a high ventilatory response to hypoxia and a diminished decrement of SaO2 should effectively prevent from high-altitude headache and probably acute mountain sickness if SaO2 desaturation is considered to be the only risk factor (15). Conversely, the larger desaturation during exercise at high altitude will contribute to the increase of the high-altitude headache incidence (16,17). The hypoxia-induced cerebral vasodilation and consequent brain swelling is considered the most likely mechanism responsible for headache development at high altitude (3,18). Besides, there is also substantial evidence that newly synthesised prostaglandins are involved in hypoxia-induced vasodilation and enhancement of nociception (19,20). Both mechanisms are probably involved in the development of high-altitude headache and acute mountain sickness, which promptly improve or can even be prevented by cyclooxygenase inhibition through NSAIDs (2,21). Several studies have shown that ambient hypoxia provokes migraine attacks in susceptible individuals (22,23), may change the character of migraine events (24), or results in a higher prevalence and severity of high altitude headache in subjects with migraine history (25). A multifactorial genetic ‘load’ has been postulated to determine the highly individual ‘migraine threshold’ of subjects, which is modulated by many factors including so called migraine triggers, e.g. hypoxia in the environment (26,27). Additionally, stresses like exercise may trigger migraine (28). An abnormal vascular response in individuals suffering from migraine has been suggested as the most important underlying pathophysiological mechanism. In migraineurs, cerebral vasodilation, e.g. in response to carbon dioxide and the dilator responses to nitric oxide donors, have been shown to be exaggerated between attacks (26,29,30). Taken together, the combination of hypoxia and exercise may largely explain the marked increased risk to develop high-altitude headache, including triggering a migraine attack, in mountaineers with a history of migraine.
Additionally, in the present study, a high rating of perceived exertion emerged as an independent risk factor for high-altitude headache. The perceived exertion becomes high when exercise intensity approximates the maximal individual performance capacity. Thus, the development of high-altitude headache seems rather to depend on the relative exercise intensity than exercise capacity per se. This is supported by the observation that the self-reported fitness level did not differ between mountaineers with and without high-altitude headache. Fluid loss is favoured by both, hypoxia and exercise and hypohydration may easily result in the case of inadequate rehydration. Hypohydration is a well known trigger for headache in normoxia and hypoxia (11,31) Besides, hypohydration may be accompanied by low energy intake leading to hypoglycaemia, which is also considered to be causative for headache development (12). Both, hypohydration and reduced blood glucose levels have been shown to trigger or worsen migraine attacks (32,33).
Taken together, among those risk factors found in our study the state of hydration, level of relative exertion and energy intake can be individually influenced and directed by susceptible subjects. Prevention of these respective risk factors might effectively reduce high-altitude headache burden, making this a recommendation to our patients.
Study limitations
There are at least two limitations that have to be addressed. First, due to the voluntary study participation, a selection bias cannot be ruled out. However, we believe this source of bias to be small as the percentage of subjects who refused was not higher than 15%. Second, the set of risk factors recorded may be incomplete or risk factor assessment may have been imprecise due to self-estimation, e.g. the state of acclimatization, the individual fitness level or the reported ratings of perceived exertion.
Conclusions
High-altitude headache is a frequent burden of high-altitude mountaineers. A history of migraine, hypobaric hypoxia, strenuous exercise and hypohydration are potent independent risk factors of headache at high altitude. They all may contribute to the development of high-altitude headache to a variable degree depending on the individual susceptibility and behaviour patterns at high altitude. Thus, high-altitude headache is not a uniform entity but seems rather to be comprised of several headache types depending on the predominance of risk factors. The reduction of each of these risk factors will help to reduce the probability of headache development when sojourning to high altitudes.