High-Altitude Erythrocytosis: Mechanisms of Adaptive and Maladaptive Responses

Abstract

Erythrocytosis, or increased production of red blood cells, is one of the most well-documented physiological traits that varies within and among in high-altitude populations. Although a modest increase in blood O₂-carrying capacity may be beneficial for life in highland environments, erythrocytosis can also become excessive and lead to maladaptive syndromes such as chronic mountain sickness (CMS).

Introduction

Increased production of red blood cells (RBCs), or erythrocytosis, has been considered a hallmark response of acclimatization in lowlanders at high altitude since its first description by Viault in 1890 (1). Hematocrit, the ratio of RBC volume to the total volume of blood generally reaches a moderately high steady-state value after a few weeks and remains stable during the length of altitude exposure. In high-altitude populations, hematocrit and hemoglobin concentrations ([Hb]) have become oversimplified markers of successful adaptation to highland environments around the world; however, the ability to offset challenges of decreased oxygen (O₂) availability depends on an integrated system of O₂ transport and cellular responses to hypoxia (2, 3). One hypothesis regarding increased RBC production at altitude is that the erythropoietic response to hypoxia arose in evolution to correct anemia (i.e., hemorrhage), compensating for decreased blood O₂-carrying capacity and correcting tissue hypoxia (4–7). However, at high altitude, tissue hypoxia cannot be fully corrected by simply increasing O₂-carrying capacity due to low inspired Po₂. Questions regarding the beneficial effects of increased RBCs for life at high altitude are still debatable due to the occurrence of some maladaptive outcomes (8–12).

Populations with generations of high-altitude ancestry differ greatly in their hematological characteristics, features of O₂ transport, and genetic factors that likely contribute to distinct physiological responses to hypobaric hypoxia. For example, erythrocytosis can become excessive in some highlanders and give rise to maladaptive syndromes such as chronic mountain sickness (CMS) or Monge’s disease, which is more common among Andeans than Tibetans (12). [Hb] has a genetic basis in Tibetans (13–15), and CMS is associated with key genetic factors in Andeans (16–19). In addition, epigenetic changes could be associated with key traits that remain to be comprehensively examined and are likely further compounded by age, sex, and multiple systemic factors (Yu et al., unpublished observations). While these findings suggest populations with relatively lower [Hb] at high altitude may be better adapted, the differences in genetic backgrounds, environmental and epigenetic factors, and feedback between erythrocytosis and O₂ transport and delivery remain unclear.

Efforts to understand the cause-and-effect relationship between the hypoxic stimulus and the erythropoietic response at a given altitude are active areas of study. For instance, the point at which hematocrit and [Hb] are truly “excessive,” and how best to define this with regard to other factors, are yet to be determined. Furthermore, whether this designation should be determined based on altitude or set at a physiological, altitude-independent threshold remains to be established. How these values impact the O₂ transport cascade and adaptation and maladaptation at individual and population levels requires additional systematic investigation. This review will address mechanisms of high-altitude erythrocytosis, genetic evidence for associations with [Hb] that provide clues into adaptive and maladaptive pathways, and detailed consideration of thresholds commonly used for excessive erythrocytosis and CMS scores.

Mechanisms of High-Altitude Erythrocytosis

Hypoxemia, or low blood Po₂, stimulates erythropoiesis. Although increased hematocrit and [Hb] values are observed in high-altitude populations, “normal” values vary greatly and depend on unique adaptations as well as the altitude of residence. Many Tibetans exhibit relatively lower [Hb] and CMS prevalence relative to Andeans at comparable altitudes and are often categorized as a population well adapted to the highland environment. [Hb] in healthy Tibetan women and men typically falls within ∼14–16 g/dl at ∼4,000 m compared with ∼17–19 g/dl in Andeans (20, 21). At the systemic level, compared with Andean highlanders, Tibetans tend to exhibit elevated pulmonary ventilation and lower end-tidal Pco₂ (a proxy for higher alveolar ventilation) and a small alveolar-arterial Po₂ difference (22–24), suggesting a high lung diffusing capacity due to greater total lung volumes that may be attributed to developmental factors (25, 26). The combination of these physiological characteristics suggests that Tibetans might have an advantage in defending arterial Po₂ and O₂ saturation from dropping dramatically. However, despite these characteristics, Po₂ in arterial blood in Tibetans at ∼3,700 m tends to be slightly lower compared with Andean high-altitude natives at the same altitude (24, 27), and [Hb] in arterial blood is less saturated with O₂ among Tibetans relative to their Andean counterparts (28, 29). Rightward as well as leftward shifts of the Hb-O₂ dissociation curve have been reported in native highlanders, but at present, there is limited consensus across studies. Even with very similar arterial Po₂, arterial blood O₂ saturation (SaO₂) has generally been reported to be higher in Andeans than Tibetans (20, 30–34).

Given these findings, it is likely other mechanisms could underlie suitable oxygenation in Tibetans to sustain normal aerobic metabolism or improved O₂ utilization. This may include increased blood flow, which may be attributed to changes in nitric oxide (NO) metabolism (20, 35, 36), and O₂ diffusion from the bloodstream to cells. Having a denser capillary network could potentially improve perfusion and O₂ delivery, with each capillary supplying a smaller area of tissue and a shorter O₂ diffusion distance. Tibetans have higher capillary density in muscles as compared with Andean high-altitude natives (37); thus this adaptive feature may help overcome low arterial O₂ content with lower [Hb] as it allows a high rate of diffusion to tissues. When [Hb] is reduced, the time to diffusive equilibration between microcirculatory vessels and cells is shortened (38). Piiper and Scheid (39) showed that the compound constant D/(β · Q) determined the degree of diffusion equilibration to be expected in muscle (D is muscle diffusing capacity, β is the average slope of the O₂-Hb dissociation curve, and Q is blood flow). Since β must decrease when total [Hb] decreases, D/(β · Q) must rise, assuming no change, or even a modest increase in Q, diffusion equilibration would take less time. As a result, tissue O₂ extraction increases when [Hb] is lower, and O₂ delivery is improved.

Ethiopians generally exhibit [Hb] comparable to Tibetans at high altitude (37, 40, 41) and are also believed to be well adapted to the high-altitude environment. This group does not show elevated pulmonary ventilation but maintains a relatively high SaO₂ at similar altitudes compared with Tibetans and Andeans (42, 43). Although various physiological features in Ethiopian highlanders have yet to be studied, it seems low [Hb] does not enhance convective O₂ transport but prevents an increase in blood viscosity and the impairment of vascular and hemodynamic function. What underlies low [Hb] values and whether it is the result of blunted erythropoiesis or other physiological adjustments remain to be determined.

In Tibetans, lower [Hb] is associated with polymorphisms in genes involved in O₂ sensing and response. Genome-wide analyses revealed positive selection in key hypoxia-inducible factor (HIF) pathway genes such as EPAS1 (13–15) and EGLN1 (14, 15), as well as PPARA (14), which encode HIF-2α, HIF-prolyl-hydroxylase 2 (PHD2), and peroxisome proliferator-activated receptor-α, respectively. Variants within these putatively adaptive gene regions are also associated with decreased [Hb] in Tibetans (reviewed in Ref. 44) and the latter with more efficient O₂ utilization (45, 46). High-frequency missense mutations identified in the EGLN1 gene encoding the oxygen sensor PHD2 results in a lower K_m value for O₂, suggesting increased HIF degradation under hypoxic conditions (47) but a loss-of-function in other contexts (48). Interestingly, variants in the EPAS1 locus in Tibetans are most similar to archaic DNA sequence relative to other populations, suggesting adaptive introgression at this genomic region (49, 50). These findings highlight the importance of considering distinct evolutionary histories among highland populations and their effects on genetic background and, therefore, physiology. While these findings provide a step forward in understanding associations, further studies are needed to determine their functional relevance and if [Hb] is the direct target of selection or secondary to other adaptive factors (44, 51).

Recent studies suggest decreased erythropoiesis might not be the only mechanism involved in maintaining relatively low [Hb] values at a given altitude. Hematocrit and [Hb] are also determined by plasma volume as well as RBC destruction. A comparison between Sherpas and Andeans indicates that lower [Hb] in the former is the result of considerably larger plasma volume and a modestly increased packed red blood cell mass (52); therefore, Sherpas have an equivalent blood volume to Andeans but a lower [Hb]. This difference is critically important, as it means Sherpa are able to benefit from the increased O₂-carrying capacity that comes from an expansion of their Hb mass but are not restricted by an increase in blood viscosity that affects vascular function and hemodynamics (53–55). Our recent studies indicate Tibetans with relatively lower [Hb] also exhibit slightly elevated endogenous carbon monoxide (CO) (Moya et al., unpublished observations; Gu et al., unpublished observations), which is a natural by-product of hemolysis and could underlie a reduction in [Hb]. Increased CO levels are also noted among Andeans with excessive erythrocytosis (EE) (56); however, this is likely due to the excessive RBCs produced and then destroyed as part of the RBC lifecycle (56). Thus these findings suggest that hematocrit in high-altitude populations with relatively low [Hb] is not regulated only through the control of erythropoiesis, but perhaps through water volume regulation via a feedback loop based on renal Po₂ (52) and potentially differences in RBC destruction and/or life span versus production.

Pathology and Risk Factors of High-Altitude Erythrocytosis

The most prominent manifestation of the overproduction of RBCs is excessive erythrocytosis (EE), the hallmark feature of CMS, a highly prevalent and incapacitating syndrome in Andeans and other high-altitude populations across the world (9, 57). EE coincides with severe hypoxemia, neurological deficits, and sleep disorders (12) and is often associated with pulmonary hypertension, myocardial infarction, and stroke owing to blood hyperviscosity predisposing to thrombophilia (8–10). It is estimated that 5–10% of the world’s population living at high-altitude may develop EE (9), and its prevalence increases with altitude and age (58–65). Above 4,300 m in the central Andes of Peru, more than 30% of highlanders by their mid-50s develop EE (59–61, 66). The prevalence of EE varies considerably within particular populations and also between men and women. Our group has shown that while hematocrit increases uniformly with age in men, this is further compounded in women owing in part to a more pronounced increase after menopause due to the combined reduction in progesterone and estradiol concentrations (59, 66). Female hormones protect women against the development of EE through their stimulatory effect on pulmonary ventilation and through the inhibition of erythropoiesis (18, 66). In addition, menstruation acts as a regular natural phlebotomy that prevents hematocrit from rising excessively.

In the absence of chronic pulmonary diseases (pulmonary emphysema, chronic bronchitis, bronchiectasis, cystic fibrosis, lung cancer, etc.) or other underlying chronic medical conditions that worsen hypoxemia, potential risk factors for the development of EE include perinatal adverse events (67, 68), the natural age-dependent reduction in pulmonary ventilation, the consequent accentuation of arterial hypoxemia (63), as well as an increased occurrence of sleep-disordered breathing with bouts of intermittent hypoxia and nocturnal hypoxemia (69–72) (FIGURE 1). Higher central and peripheral chemoreflex set points might lead to hypoventilation at a given Pco₂, resulting in lower arterial O₂ saturation and increased erythropoietic response. Frequently, individuals with CMS have lower ventilatory sensitivities to CO₂ compared with non-CMS Andean controls (73, 74). Also, ventilatory sensitivities to O₂ and CO₂ play a key role in sleep-disordered breathing in highlanders (70, 75), and low ventilatory sensitivity to hypoxia or CO₂ can lead to more severe desaturation during sleep and/or prolonged desaturation periods (76). Recent studies have shown that sleep-disordered breathing is more prevalent in Peruvian highlanders than in lowlanders at sea level (71), and nocturnal hypoxemia and sleep apnea events are independently associated with EE. We have recently shown that lower pulse O₂ saturation (SpO₂) during sleep and during the day are associated with higher hematocrit in Andean men and women. When adjusting for age and SpO₂, obstructive apnea index and apnea-hypopnea index also predict higher hematocrit and CMS scores in highlander men. While the hypoxic ventilatory response is blunted in Andeans with and without EE, lower hypoxic chemosensitivity is associated with lower daytime SpO₂ and may therefore play a role in the development of EE (69).

FIGURE 1.

Schematic diagram showing the expression of the adaptive and maladaptive erythropoietic phenotypes as the result of genetic background, epigenomic processes, and physiological responses to chronic hypoxia across the life span In recent years, an increasing amount of evidence has accumulated to support the genetic basis of the adaptive and maladaptive erythropoietic responses to chronic hypoxia. The genetic background of an individual will ultimately determine the development of a moderate or an excessive erythrocytosis (EE) in response to the different stimuli, epigenomic processes, and physiological events throughout the life span. In utero and perinatal events can condition the erythropoietic response in later life. As individuals age, they go through several physiological processes during adulthood such as menopause and the age-dependent decline in ventilation, and can also develop conditions such as obesity; all recognized risk factors for the development of EE. Thus, what determines that some individuals end up taking the adaptive or maladaptive path? Our genetic background, together with epigenetic modifications accumulated through life will determine the impact of these physiological or pathophysiological changes on predetermined characteristics such as ventilatory control or erythropoietic sensitivity. Individuals with an adaptive genetic background and favorable epigenetic changes for life at high altitude will have a preserved central ventilatory sensitivity to CO₂, hence maintaining adequate ventilation during day and night; will have fewer and/or less prolonged respiratory events during sleep, therefore maintaining proper oxygenation; and will have a regular erythropoietic sensitivity of erythroid progenitors. All these characteristics assure maintaining a moderate hemoglobin concentration ([Hb]) at a sufficient level for their altitude of residence. On the other hand, individuals with a maladaptive genetic background may express maladaptive traits early in life and develop EE. These individuals may show adequate responses during young age but reach a tipping point at some stage in their lives in which maladaptive responses become manifest (i.e., decreased ventilatory central sensitivity to CO₂, increased sleep disorders, and severe hypoxemia) resulting in a stronger stimulus for erythropoiesis. Alternatively, they may have increased erythropoietic sensitivity, even with “normal” hypoxemia (usual SpO₂ range for a given altitude in healthy individuals), and without any respiratory or sleep alterations, they will develop EE over time. RBC, red blood cell. Image created with BioRender.com and used with permission.

While various physiological mechanisms are potentially involved in the etiopathogenesis of EE, once EE has developed, excess RBCs might impair oxygenation. We can say that at some point during life at high altitude, chronic hypoxemia initiates a vicious cycle in which hematocrit starts rising, causing abnormal blood rheology and imposing a burden on vasculature. This in turn could contribute to impairments in the distribution of pulmonary blood flow, ventilation/perfusion ratio, and O₂ diffusion. Impaired pulmonary gas exchange further augments the degree of arterial hypoxemia (6, 77, 78), which would in the end stimulate further erythropoiesis (FIGURE 1).

Although at present the pathophysiology of EE is sufficiently described to provide plausible mechanistic explanations, it is necessary to recognize the paradoxes between what is actually known about this high-altitude disease and how that knowledge is interpreted. It is commonly believed that severe hypoxemia and/or increased erythropoietin (EPO) concentration underlies an exacerbated erythropoietic response in highlanders. However, this is not always the case. Although the relative risk (RR) of having EE with severe hypoxemia [SpO₂ <83% (60)] at 4,340 m in Cerro de Pasco, Peru is 2.66 [95% confidence interval (CI), 2.29–3.08, P < 0.0001, n = 965; Villafuerte FC, unpublished observations)], ∼27% of highlanders with SpO₂ values above 83% have EE. On the other hand, 28% of highlanders with SpO₂ below 83% have [Hb] within the standard range relative to the altitude of residence (FIGURE 2). Highlanders with serum EPO > 1 standard deviation (SD) of the average value of healthy individuals at the same altitude (79) show a RR of 1.63 of having EE (95% CI: 1.23–2.17, P = 0.0007, n =146; Villafuerte FC, unpublished observations). Important to note, however, is that 47% of individuals without EE have high serum EPO (79). Therefore, this variety of phenotypes shows that severe hypoxemia or high EPO levels are not always necessary to develop EE. These observations highlight the individual variability and the potential role for the modulation of EPO signaling and the hypoxia- or EPO-sensitivity of erythroid progenitors. We have shown that EE is associated with decreased soluble EPO receptor (sEPOR) levels and with a higher EPO-to-sEPOR ratio during day and sleep, which implies a stronger erythropoietic stimulus at similar EPO concentrations (80). Our results suggest that sEPOR could act as an extracellular regulator of erythropoiesis and mechanistic link for the development of EE. We have also recently shown that erythroid progenitors derived from highlanders with EE show an increased proliferative response under hypoxic conditions and upregulated expression of proerythropoietic genes (17). Thus, besides any extracellular modulation of the erythropoietic signal, it can be hypothesized that erythroid progenitors derived from highlanders with CMS are genetically determined for exacerbated proliferation, which could be further enhanced by poor systemic oxygenation during daytime or sleep as a consequence of altered respiratory control or sleep-disordered breathing, or by normal physiological processes such as menopause (18, 70, 75, 81).

FIGURE 2.

Relationship between Hb concentration ([Hb]) and blood oxygenation Shown is the inverse relationship between [Hb] and SpO₂ in male highlanders from Cerro de Pasco, Peru at 4,340 m (r = −0.55, P < 0.0001, n = 965; Villafuerte FC, unpublished observations). The dotted lines represent the thresholds of [Hb] to determine excessive erythrocytosis (EE; ≥21 g/dl) and of SpO₂ (<83%) to determine severe hypoxemia. The different phenotypes on these variables can be observed from the 4 quadrants. Although it is expected that individuals with severe hypoxemia develop EE (relative risk of 2.66, 95% confidence interval of 2.29–3.08, P < 0.0001), about one-third of highlanders with SpO₂ values >83% have EE. On the other hand, also a third of highlanders with SpO₂ <83% have [Hb] within the usual range relative to the altitude of residence.

In Andeans, genetic variants, at least partly, underlie the regulation of erythropoiesis. Whole-genome and genotyping studies have shown an association between the sentrin-specific protease 1 (SENP1) allelic variant rs7963934 and the EE phenotype among Andeans (16, 82, 83), yet the precise functional variant(s) remain unknown. Recent studies suggest SENP1, which regulates the activity of transcription factors such as HIF and GATA-binding factor 1 (GATA1), plays a central role in the excessive production of RBCs, possibly by modulating different stages of erythropoiesis, including steps of the EPO signaling pathway and apoptosis in erythroid progenitors (84–86) (FIGURE 3). Under in vitro hypoxic conditions, CMS erythroid progenitors derived from human-induced pluripotent stem cells, obtained from skin fibroblasts of Andean highlanders, show increased proliferation and upregulation of SENP1 expression, which is associated with stabilization and upregulation of GATA1 and GATA1-responsive genes such as the mitochondrial anti-apoptotic factor Bcl-xL (84). These findings suggest a crucial role for SENP1 in erythroid proliferation seen in CMS. We have confirmed the upregulation of SENP1, GATA1, and EPOR expression in erythroid progenitors derived from peripheral blood mononuclear cells (PBMCs) of highlanders with EE (17). We have also recently shown, through a genome-wide association study (89), associations between CMS and a variant of the AEBP2 gene, an epigenetic regulator for neural crest cells (90), the calpastatin gene CAST, which has an inhibiting effect on calpain and has been associated with the progression of pulmonary hypertension (91), and MCTP2, which has been associated with body fat levels and obesity in other populations (92).

FIGURE 3.

Role of sentrin-specific protease 1 (SENP1) and GATA-binding factor 1 (GATA1) in the maladaptive and adaptive erythropoietic response to chronic hypoxia Whole-genome sequencing analysis revealed single nucleotide polymorphisms in the SENP1 gene in Andean highlanders (16) that explain part of the maladaptive and adaptive erythopoietic responses to life at high altitude. Under normoxia, the hypoxia-inducible factor-α (HIFα) subunit is hydroxylated by prolyl hydroxylase enzymes (PHDs), targeted for ubiquitylation (UB) by the von Hippel-Lindau protein (pVHL), and subsequently degraded in the proteasome (87). Under hypoxia, the activity of PHDs decreases due to low Po₂, and HIFα escapes hydroxylation and undergoes SUMOylation. SENP1-dependent deSUMOylation rescues HIFα from ubiquitylation/degradation and allows its association with nuclear HIFβ. The heterodimer binds to a core consensus sequence at the promoters of HIF-responsive genes and, upon binding to coactivators, initiates transcription. SENP1 plays a critical role in the deSUMOylation and stabilization of HIFα and GATA1 by preventing their ubiquitylation and subsequent degradation, and enhancing their transcriptional activity (85, 86). Individuals with a reduced frequency of the SENP1 gene allelic variant rs7963934, showed increased expression of the SENP1 protein under hypoxic conditions and a maladaptive erythropoietic response (16, 82). The overexpression of SENP1 increases the deSUMOylation activity and favors the stabilization of HIF and GATA1, resulting in an increased expression of hypoxia responsive genes and genes that regulate the erythropoietic process. These, in turn, promote an exaggerated proliferation of erythroid cells (84). On the other hand, individuals with an increased frequency of the rs7963934 variant show an adaptive erythropoietic response to hypoxia, which favors the ubiquitination and degradation of SUMOylated HIFα and GATA1, keeping the response to hypoxia and the proliferation of erythroid cells under a normal range relative to the altitude of residence. Estrogen also plays an adaptive role as a protective factor against excessive erythrocytosis by binding to its β-receptor, causing a decrease in the expression of the HIF complex and GATA1, and increasing apoptosis in erythroid cells (18). After menopause, however, estrogen levels decrease (59, 66), which allows an increase in the transcriptional activity of HIF and GATA1, which leads to an exaggerated proliferation of erythroid cells. This eventually might lead to a maladaptive outcome such as excessive erythrocytosis in women. Also, recently, ARID1B, a chromatin remodeling factor, has been suggested to act as an adaptive regulator of erythropoiesis in Andean highlanders by modulating the expression of GATA1 and maintaining erythroid cell proliferation within the normal high-altitude range (88). ER, estrogen receptor; EPO, erythropoietin; EPOR, EPO receptor; Bcl-xL, B-cell lymphoma-extra large anti-apoptotic factor; ALA-D, delta-aminolevulinic acid dehydratase; ALA-S, delta-aminolevulinic acid synthase. Image created with BioRender.com and used with permission.

In addition to these genes, other key pathway and HIF-regulated genes have been identified as potential targets of selection in Andeans, including those related to cardiovascular function (93). While limited data are available regarding genotype-phenotype relationships, our recent work suggests HIF-related genes, similar to those reported in Tibetans (94, 95), are under selection in Andeans with likely distinct functional variants (96). Recent studies also indicate differential methylation at the EGLN1 locus in Andeans with and without EE (97) and differential methylation patterns at various sites across Andean genomes, including reduced methylation at EPAS1 (98, 99). These findings suggest epigenetic mechanisms could underlie developmental plasticity and long-term effects at high-altitude that warrant additional investigation.

Thus these recent findings provide insight into the multiple risk factors and potential mechanisms of EE in terms of adaptive and maladaptive outcomes. However, longitudinal studies with integrative translational approaches should be conducted to obtain a deep appreciation of the natural history of the disorder on an individual basis.

Definition of Excessive Erythrocytosis and the Qinghai CMS Score

The presence and severity of CMS are assessed through the Qinghai Score, a system based on the presence of EE ([Hb] ≥21 g/dl in men and ≥19 g/dl in women) and a group of signs and symptoms agreed by consensus at the International Working Group for the study of CMS that formed in 1998 and established the diagnostic guide principles for this syndrome (9) (Table 1). According to the sum of individual scores, the presence and severity of primary CMS are determined. It is important to point out that EE is the defining feature of CMS. Thus it is possible to have EE and not have CMS, but it is not possible to have CMS without EE. Therefore, [Hb] must be measured as the first step in diagnosing CMS, as a high score without EE rules out the presence of the syndrome.

TABLE 1.

Assessments through the Qinghai Score

Points	Sign or Symptom
Hemoglobin concentration
Men
0	[Hb] <21 g/dl
3	Excessive erythrocytosis, [Hb] ≥21 g/dl
Women
0	[Hb] <19 g/dl
3	Excessive erythrocytosis, [Hb] ≥19 g/dl
Breathlessness and/or palpitations
0	No breathlessness/palpitations
1	Mild breathlessness/palpitations
2	Moderate breathlessness/palpitations
3	Severe breathlessness/palpitations
Sleep disturbance
0	Slept as well as usual
1	Did not sleep as well as usual
2	Woke many times, poor night’s sleep
3	Could not sleep at all
Cyanosis
0	No cyanosis
1	Mild cyanosis
2	Moderate cyanosis
3	Severe cyanosis
Dilatation of veins
0	No dilatation of veins
1	Mild dilatation of veins
2	Moderate dilatation of veins
3	Severe dilatation of veins
Paresthesia
0	No paresthesia
1	Mild paresthesia
2	Moderate paresthesia
3	Severe paresthesia
Headache
0	No headache
1	Mild headache symptoms
2	Moderate headache
3	Severe headache, incapacitating
Tinnitus
0	No tinnitus
1	Mild tinnitus
2	Moderate tinnitus
3	Severe tinnitus

The Qinghai score (9) has been designed to assess the presence and severity of chronic mountain sickness (CMS) based on hemoglobin concentration ([Hb]) and a group of signs and symptoms at the altitude of residence. The addition of points for each sign or symptom category results in the total CMS score. CMS absent: score = 0–5; mild: score = 6–10; moderate: score =11–14; and severe: score >15.

There is an ongoing controversy and discussion regarding the use of a single threshold value for determining EE and the use of the scoring system. One issue is that the [Hb] threshold value for diagnosing EE is altitude independent, and thus the same value would be used to define EE either at 3,400, 4,400, or 5,100 m. This single value was obtained from the largest epidemiological study in the early 1990s conducted in Cerro de Pasco, Peru, at 4,340 m. [Hb] was determined to be excessive when it was higher than 2 SD above the mean [Hb] value of young healthy highlanders at that altitude (58, 61). At the time, Cerro de Pasco was the largest and highest demographically stable population in the world and, therefore, if an individual was determined to have EE at that altitude, it was assumed definitively excessive at lower altitudes. If we accept that the “normal” [Hb] value increases with altitude, having a threshold based on “normal” [Hb] at 4,340 m would result in an underestimation of the prevalence of EE if we apply this value to a population living at a lower altitude. Conversely, now that large semistable populations are inhabiting higher altitudes (i.e., La Rinconada, Peru, 5,100 m) (100, 101), applying this value would result in an overestimation of EE prevalence. Accordingly, recent studies have shown that at La Rinconada, the presence of EE does not necessarily correlate with a high CMS score and that sometimes a high CMS score is observed in highlanders without EE (100, 102). It would be straightforward to think that if [Hb] increases with altitude, the current [Hb] threshold at 5,100 m would no longer define EE. Thus 21 g/dl or 19 g/dl in men and women, respectively, would not necessarily be excessive. This has been wrongly interpreted as EE not being the main sign of CMS at extreme altitudes and led to questioning of the validity of the Qinghai CMS score.

In cases of threshold-based diagnosis of EE and low CMS scores (i.e., <6 points; absence of CMS), it is necessary to consider the age of individuals, the time of residence at a given altitude, the age of EE onset, and the frequency of migration to and from lower altitudes. For example, the longer amount of time an individual has had an excessive hematocrit, the more severe its effects on vasculature. Hence, it is possible that some young or middle-aged individuals with EE and relatively few years living at an extreme altitude, and whose migratory patterns to low altitude are frequent, do not present marked symptoms. A recent study at 5,100 m (102) showed a significant number of highlanders with high hematocrit values and relatively young age (64–76%, 36–47 yr, respectively) but apparently without symptoms of CMS. In fact, although these high-altitude natives showed a low CMS score, the prognosis for this high [Hb] level over time may be a higher score and the occurrence of thromboses and hemorrhages, as it has been observed in the Cerro de Pasco population, where the prevalence of stroke is associated with a high prevalence of EE (103).

In cases of the absence of EE but a high CMS score (i.e., ≥6 points) it is necessary to consider the typical clinical picture of the syndrome, first described by Carlos Monge-M in 1925 (104–107), where the excessive count of RBCs is central in the definition of CMS. In other words, there is no CMS without EE. Therefore, a high CMS score (≥6 points) without EE is just an addition of nonspecific symptoms. Bloodletting and hemodilution studies at high altitude, as well regular phlebotomies (common practice of highlanders with EE to manage the condition), have shown that once hematocrit is reduced, CMS signs and symptoms disappear rapidly, which confirms that these are secondary to EE (Refs. 6, 108–110; Anza-Ramírez C, et al., unpublished observations).

Thus, from an epidemiological point of view, a critical reappraisal of the consensus criteria might consider population studies at different altitudes to estimate normal and excessive [Hb] values and either increase or decrease the [Hb] threshold value to define EE. However, from the physiological point of view, it is possible that above certain value, [Hb] becomes detrimental irrespective of the altitude of residence. Recent evidence from our group has provided a molecular basis to explain the hemodynamic impairments observed in EE owing to the associated exponential increases in blood viscosity and shear stress which can collectively induce vascular endothelial dysfunction (11, 54, 55, 111). Impaired vascular endothelial function has been associated with a free radical-mediated induction of inflammation and scavenging of vascular NO bioavailability, a metabolic cascade collectively termed oxidative-inflammatory-nitrosative stress (OXINOS) (53, 111, 112). Bailey et al have consistently demonstrated systemic OXINOS to be selectively elevated in CMS patients compared with age-matched non-CMS controls (53, 112). Elevated OXINOS was associated with more pronounced impairments in systemic and cerebrovascular endothelial function in the form of elevated arterial stiffness, cerebral hypoperfusion, blunted cerebral vasoreactivity, cognitive impairment, and depression (53, 112, 113). Thus physiologically, it is also possible to establish a threshold [Hb] value based on longitudinal studies of cerebrovascular function as a function of increasing [Hb], and its relationship with the neurological symptoms associated with EE, which defines the full clinical picture of CMS. This information coupled with an individual’s life history and tracking of biomarkers across the life span (or at least part of it) would provide much needed insight into the development of EE and CMS and combinations thereof that develop in some but not all individuals.

Clinical Significance of Research in the Field

EE and CMS are major public health problems in high-altitude populations around the world not only because their incapacitating nature, but also because their association with cardiovascular and cerebrovascular disease. We and others have shown that EE associates with increased cardiovascular disease risk and cardiometabolic disorders (114–117) in various high-altitude regions across the world (118–121). Importantly, we have identified independent associations between EE and 24-h ambulatory hypertension including systolic-diastolic and isolated diastolic hypertension (114, 122) and a significant proportion of masked hypertension, the latter linked to increased cardiovascular morbidity and mortality in lowlanders (123). Our data also show that the incidence of ambulatory hypertension in highlanders with EE is greater compared with non-EE individuals across age groups. We have identified independent associations between EE, insulin resistance, hyperglycemia, and dyslipidemia (114) and shown relationships between EE, control of breathing, and sleep phenotypes (69), which have major impacts on cardiometabolic health. These findings agree with other studies at high altitude in the Peruvian Andes that have identified independent relationships between EE and hypertension, hypertriglyceridemia (103, 116, 118), and metabolic syndrome (115). Importantly, EE may have a more adverse impact on cardiovascular and cerebrovascular disease risk in women, especially following menopause.

Conclusions

Values of [Hb] and hematocrit are commonly used to characterize an individual’s physiological response to chronic high-altitude hypoxia. However, mechanisms underlying differences in this response are multifaceted and should be evaluated in the context of other O₂ transport components that assess physiological impact, genetic and life history information, and altitude of residence. Thresholds used to define EE and CMS are currently under discussion and would benefit from thorough considerations of these factors, as well as much-needed, carefully planned, longitudinal assessments that aim to determine the effects of elevated [Hb] at the individual and population levels.