Excessive Erythrocytosis and Chronic Mountain Sickness in the Highest City in the World

FOR EDITORIAL COMMENT, SEE PAGE 1136Chronic mountain sickness (CMS), a specific condition affecting 5% to 33% of high-altitude dwellers, is characterized by severe hypoxemia, excessive erythrocytosis (EE), and a range of associated clinical symptoms.1León-Velarde F. Maggiorini M. Reeves J.T. et al.Consensus statement on chronic and subacute high altitude diseases.High Alt Med Biol. 2005; 6: 147-157Crossref PubMed Scopus (374) Google Scholar,2Villafuerte F.C. Corante N. Chronic mountain sickness: clinical aspects, etiology, management, and treatment.High Altitude Medicine & Biology. 2016; 17: 61-69Crossref PubMed Scopus (103) Google Scholar Previous studies showed that the prevalence of EE and CMS increases with age.3León-Velarde F. Arregui A. Monge C. Ruiz HR y Aging at high altitudes and the risk of chronic mountain sickness.J Wilderness Med. 1993; 4: 183-188Abstract Full Text PDF Scopus (44) Google Scholar, 4Whittembury J. Monge C. High altitude, haematocrit and age.Nature. 1972; 238: 278-279Crossref PubMed Scopus (25) Google Scholar, 5Sime F. Monge C. Whittembury J. Age as a cause of chronic mountain sickness (Monge's disease).Int J Biometeorol. 1975; 19: 93-98Crossref PubMed Scopus (42) Google Scholar, 6Monge C. León-Velarde F. Arregui A. Increasing prevalence of excessive erythrocytosis with age among healthy high-altitude miners.N Engl J Med. 1989; 321: 1271-1272Crossref PubMed Google Scholar However, because of the lack of longitudinal studies, whether ageing per se or the duration of high-altitude residency is promoting EE and CMS remains unclear.2Villafuerte F.C. Corante N. Chronic mountain sickness: clinical aspects, etiology, management, and treatment.High Altitude Medicine & Biology. 2016; 17: 61-69Crossref PubMed Scopus (103) Google Scholar,3León-Velarde F. Arregui A. Monge C. Ruiz HR y Aging at high altitudes and the risk of chronic mountain sickness.J Wilderness Med. 1993; 4: 183-188Abstract Full Text PDF Scopus (44) Google Scholar,5Sime F. Monge C. Whittembury J. Age as a cause of chronic mountain sickness (Monge's disease).Int J Biometeorol. 1975; 19: 93-98Crossref PubMed Scopus (42) Google Scholar Furthermore, we and others7Hancco I. Bailly S. Baillieul S. et al.Excessive erythrocytosis and chronic mountain sickness in dwellers of the highest city in the world.Front Physiol. 2020; 11: 773Crossref PubMed Scopus (9) Google Scholar,8Gonzales G.F. Rubio J. Gasco M. Chronic mountain sickness score was related with health status score but not with hemoglobin levels at high altitudes.Respir Physiol Neurobiol. 2013; 188: 152-160Crossref PubMed Scopus (16) Google Scholar previously challenged in cross-sectional studies the relationship between EE and clinical symptoms of CMS, with both high hematocrits and CMS symptoms showing minimal correlation. However, these studies were not designed to assess the time course of EE and CMS symptom development, and longitudinal studies are still lacking to better understand the relationship between the development of EE, CMS symptoms, and the duration of high-altitude residency. Thus, this longitudinal study aimed to prospectively investigate EE, CMS symptoms, and changes in physiologic variables in dwellers from the highest city in the world. Between 2005 and 2019, we conducted a longitudinal follow-up of 90 male dwellers from La Rinconada (5,100-5,300 m), a gold mining town located in southeastern Peru and considered the highest city in the world.9Enserink M. Hypoxia city.Science. 2019; 365: 1098-1103Crossref PubMed Scopus (6) Google Scholar Participants (all born at 3,500-4,500 m and permanently living in La Rinconada) were recruited during a medical consultation for families of miners and had at least 1 year of residency in La Rinconada at the time of inclusion.7Hancco I. Bailly S. Baillieul S. et al.Excessive erythrocytosis and chronic mountain sickness in dwellers of the highest city in the world.Front Physiol. 2020; 11: 773Crossref PubMed Scopus (9) Google Scholar The study was approved by the ethics committees of Inter-région Rhône-Alpes-Auvergne (IRB-5891) and Universidad Nacional Mayor de San Marcos (CIEI-2019-002). Each highlander was examined at least once a year during the study period. During each visit, the seven clinical symptoms of CMS (Table 1) were collected and scored according to the Qinghai CMS score (ranging from 0/no symptom to 3/severe symptom) and reported further as the CMS clinical score8Gonzales G.F. Rubio J. Gasco M. Chronic mountain sickness score was related with health status score but not with hemoglobin levels at high altitudes.Respir Physiol Neurobiol. 2013; 188: 152-160Crossref PubMed Scopus (16) Google Scholar; Hematocrit was measured from capillary blood samples using a microhematocrit centrifugation, and EE was defined as a hematocrit ≥ 63%, based on the current international consensus; CMS diagnosis was made in case of a total Qinghai CMS score > 5, including the presence of EE (adding 3 points to the CMS clinical score).1León-Velarde F. Maggiorini M. Reeves J.T. et al.Consensus statement on chronic and subacute high altitude diseases.High Alt Med Biol. 2005; 6: 147-157Crossref PubMed Scopus (374) Google Scholar Oxygen saturation (Spo2) was measured by a finger sensor (NELLCOR OxiMaxN-65, TycoHealthcare) after a 5-minute rest period in a sitting position.Table 1Physiological Variables, CMS Symptoms, and CMS Clinical and Total Scores Changes Over the 14 Years of Follow-upVariablesYear 0Year 2Year 4Year 6Year 8Year 10Year 12Year 14PaFor simplicity, data are presented every 2 years, but P value for year's effect refers to years of follow-up as a continuous variable. For clarity, data regarding to CMS symptoms are presented as categorical by degree of severity but were analyzed as continuous variables, considering a binomial generalized linear mixed model or a Poisson regression as appropriate.Spo2, %82.6 ± 2.182.7 ± 2.082.6 ± 2.181.9 ± 1.3bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.82.0 ± 1.5bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.81.9 ± 1.3bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.81.9 ± 1.3bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.81.9 ± 1.3bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.< .001Hematocrit, %66.5 ± 6.967.1 ± 6.4bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.67.5 ± 6.2bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.67.7 ± 6.2bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.68.0 ± 6.0bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.68.3 ± 5.8bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.68.5 ± 5.9bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.68.4 ± 5.7bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.< .001Breathlessness/Palpitations Absent (0)75 (83%)80 (89%)80 (89%)81 (90%)75 (83%)72 (80%)73 (90%)55 (81%)NS Mild (1)15 (17%)10 (11%)10 (11%)9 (10%)15 (17%)18 (20%)8 (10%)13 (19%) Moderate (2)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%) Severe (3)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)Sleep disturbances Absent (0)84 (93%)86 (96%)86 (96%)86 (96%)71 (79%)bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.67 (74%)bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.61 (75%)bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.57 (84%)bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.< .001 Mild (1)6 (7%)4 (4%)4 (4%)4 (4%)19 (21%)23 (26%)20 (25%)11 (16%) Moderate (2)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%) Severe (3)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)Cyanosis Absent (0)24 (27%)24 (27%)23 (26%)23 (26%)23 (26%)21 (23%)21 (26%)15 (22%)NS Mild (1)38 (42%)42 (46%)47 (52%)45 (50%)45 (50%)45 (50%)42 (52%)37 (54%) Moderate (2)28 (31%)24 (27%)20 (22%)22 (24%)22 (24%)24 (27%)18 (22%)16 (24%) Severe (3)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)Dilation of veins Absent (0)89 (99%)89 (99%)89 (99%)89 (99%)88 (98%)78 (87%)bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.79 (98%)bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.65 (96%)bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.< .001 Mild (1)1 (1%)1 (1%)1 (1%)1 (1%)2 (2%)12 (13%)2 (2%)3 (4%) Moderate (2)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%) Severe (3)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)Tinnitus Absent (0)87 (97%)90 (100%)90 (100%)90 (100%)90 (100%)90 (100%)81 (100%)68 (100%)NS Mild (1)3 (3%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%) Moderate (2)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%) Severe (3)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)Paresthesia Absent (0)84 (93%)85 (94%)85 (94%)84 (93%)71 (79%)bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.51 (57%)bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.50 (62%)bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.46 (68%)bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.< .001 Mild (1)6 (7%)5 (6%)5 (6%)6 (7%)19 (21%)39 (43%)31 (38%)22 (32%) Moderate (2)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%) Severe (3)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)Headache Absent (0)32 (36%)47 (52%)51 (57%)47 (52%)38 (42%)28 (31%)35 (43%)31 (46%)NS Mild (1)47 (52%)42 (47%)39 (43%)43 (48%)52 (58%)60 (67%)43 (53%)37 (54%) Moderate (2)11 (12%)1 (1%)0 (0%)0 (0%)0 (0%)2 (2%)3 (4%)0 (0%) Severe (3)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)0 (0%)CMS clinical score1.9 ± 1.31.7 ± 1.2bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.1.6 ± 1.2bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.1.6 ± 1.1bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.2.1 ± 1.2bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.2.4 ± 1.2bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.2.3 ± 1.3bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.2.2 ± 1.3bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.< .001Total CMS score3.9 ± 2.14.0 ± 1.94.0 ± 1.94.0 ± 1.94.5 ± 1.8bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.4.8 ± 1.8bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.4.8 ± 1.9bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.4.6 ± 1.8bSignificant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable.< .001N = 90, except for year 12 (n = 81) and year 14 (n = 68, except for hematocrit, n= 69) because of loss to follow-up.Data are expressed as mean ± SD or No. (%). Because participants have been assessed 2.7 ± 1.1 times a year, individual data were averaged by years. Each CMS symptom was scored on a 0-3 severity scale; CMS clinical score represented the sum of the seven CMS symptoms. Total CMS score was calculated according to current recommendations (see Methods for more details). CMS = chronic mountain sickness; NS = nonsignificant; Spo2 = pulse oxygen saturation.a For simplicity, data are presented every 2 years, but P value for year's effect refers to years of follow-up as a continuous variable. For clarity, data regarding to CMS symptoms are presented as categorical by degree of severity but were analyzed as continuous variables, considering a binomial generalized linear mixed model or a Poisson regression as appropriate.b Significant difference (P < .01) compared with the baseline value at year 0 (2005) and obtained from mixed models considering years of follow-up as a categorial variable. Open table in a new tab N = 90, except for year 12 (n = 81) and year 14 (n = 68, except for hematocrit, n= 69) because of loss to follow-up. Data are expressed as mean ± SD or No. (%). Because participants have been assessed 2.7 ± 1.1 times a year, individual data were averaged by years. Each CMS symptom was scored on a 0-3 severity scale; CMS clinical score represented the sum of the seven CMS symptoms. Total CMS score was calculated according to current recommendations (see Methods for more details). CMS = chronic mountain sickness; NS = nonsignificant; Spo2 = pulse oxygen saturation. Data were expressed as mean ± SD or median [25th-75th percentiles]. Prevalence of EE and CMS at inclusion were reported with their 95% CI, as well as the crude incidence of EE and CMS during the study period, among the disease-free highlanders at baseline. To assess the longitudinal changes of the continuous variables, generalized linear mixed models were computed, with years as fixed effect and participant as random effect. Progressive changes in EE and CMS incidences over the years were visualized using cumulative incidence curves. Exploratory linear mixed models were performed to investigate the association of potential a priori independents predictors available (Spo2, CMS clinical score, age at inclusion, years of residency in La Rinconada, and years of follow-up),2Villafuerte F.C. Corante N. Chronic mountain sickness: clinical aspects, etiology, management, and treatment.High Altitude Medicine & Biology. 2016; 17: 61-69Crossref PubMed Scopus (103) Google Scholar,7Hancco I. Bailly S. Baillieul S. et al.Excessive erythrocytosis and chronic mountain sickness in dwellers of the highest city in the world.Front Physiol. 2020; 11: 773Crossref PubMed Scopus (9) Google Scholar with changes in both hematocrit and total CMS score over the time. Among the 90 highlanders (29 [24-41] years old on inclusion; previous residency time in La Rinconada, 2 [1-4] years) included in 2005, prevalence of EE and CMS at baseline was 76% (68/90; 95% CI, 67%-84%) and 31% (28/90; 95% CI, 22%-41%), respectively. None of them took medication (eg, acetazolamide) against EE or CMS. Among the disease-free highlanders at inclusion in 2005 (ie, 22 [24%] and 62 [69%] highlanders for EE and CMS, respectively), the crude incidence rate of EE and CMS during the total 14-years' follow-up was 6.3 (95% CI, 5.0-7.8) and 4.4 (95% CI, 3.4-5.7) cases by person-years, respectively, with a continuous and progressive rise of both incidences over the years (Fig 1). Mean hematocrit increased early during the follow-up period, whereas the changes in CMS total score (increase), Spo2 (decrease), and CMS clinical score (slightly decrease then increase) were delayed (Table 1). Linear mixed model with hematocrit as dependent variable revealed that Spo2 (coef −0.12 [95% CI, −0.08 to −0.16]; P < .001), CMS clinical score (coef 0.15 [95% CI, 0.09-0.22]; P < .001), and the years of follow-up (coef 0.12 [95% CI: 0.11-0.13], P < .001), but not age on inclusion, were significantly associated with the increase in hematocrit over time. In a similar model (not including the CMS clinical score variable), with CMS total score as dependent variable, Spo2 (coef −0.08 [95% CI, −0.10 to −0.05]; P < .001) and the years of follow-up (coef 0.07 [95% CI, 0.06-0.08]; P < .001) remained significantly associated with the increase in CMS total score over time. This longitudinal study showed that in relatively young highlanders, both EE and CMS occurrence increase over the years spent at 5,100 m, with an independent relationship between the increase in hematocrit and CMS total score and the duration of residency at high altitude. To our knowledge, this is the first longitudinal study assessing hematocrit and CMS symptoms changes in highlanders. The fact that the increase in hematocrit and CMS total score was independent of the age at inclusion suggests that the time spent at high altitude is a major risk factor for EE and CMS occurrence. Spo2 also appears as an independent factor associated with both EE and CMS, suggesting that both the duration and severity of hypoxic exposure underlie the development of these chronic high-altitude diseases. Nevertheless, other important findings from this study should also be underlined. First, the high initial prevalence of EE (76%) in this population permanently living above 5,000 m raises the pertinence of a unique threshold to define EE (based on the mean ± 2 SD value measured at 4,300 m1León-Velarde F. Maggiorini M. Reeves J.T. et al.Consensus statement on chronic and subacute high altitude diseases.High Alt Med Biol. 2005; 6: 147-157Crossref PubMed Scopus (374) Google Scholar,3León-Velarde F. Arregui A. Monge C. Ruiz HR y Aging at high altitudes and the risk of chronic mountain sickness.J Wilderness Med. 1993; 4: 183-188Abstract Full Text PDF Scopus (44) Google Scholar,6Monge C. León-Velarde F. Arregui A. Increasing prevalence of excessive erythrocytosis with age among healthy high-altitude miners.N Engl J Med. 1989; 321: 1271-1272Crossref PubMed Google Scholar) rather than thresholds corrected for the altitude level.7Hancco I. Bailly S. Baillieul S. et al.Excessive erythrocytosis and chronic mountain sickness in dwellers of the highest city in the world.Front Physiol. 2020; 11: 773Crossref PubMed Scopus (9) Google Scholar Second, the modest decrease in Spo2 (albeit statistically significant), delayed compared with the early rise in hematocrit (Table 1), may question the relationship between hypoxemia and EE. A common hypothesis about EE includes an exacerbation of erythrocytosis caused by severe hypoxemia in some highlanders induced by a relative hypoventilation.2Villafuerte F.C. Corante N. Chronic mountain sickness: clinical aspects, etiology, management, and treatment.High Altitude Medicine & Biology. 2016; 17: 61-69Crossref PubMed Scopus (103) Google Scholar However, it also has been suggested that EE may, by itself, impair pulmonary ventilation/perfusion ratio, and thus worsen hypoxemia, which could further exacerbate erythropoiesis.10Cruz J.C. Diaz C. Marticorena E. Hilario V. Phlebotomy improves pulmonary gas exchange in chronic mountain polycythemia.Respiration. 1979; 38: 305-313Crossref PubMed Scopus (27) Google Scholar The exact timeframe and interaction between hypoxemia and EE remain to be investigated. Third, the delayed increase in CMS symptom severity (reflected by the CMS clinical score) suggests that several years are probably required before EE leads to clinical symptoms.7Hancco I. Bailly S. Baillieul S. et al.Excessive erythrocytosis and chronic mountain sickness in dwellers of the highest city in the world.Front Physiol. 2020; 11: 773Crossref PubMed Scopus (9) Google Scholar These findings may explain at least in part the lack of relationship between EE and CMS symptoms highlighted in previous cross-sectional studies.7Hancco I. Bailly S. Baillieul S. et al.Excessive erythrocytosis and chronic mountain sickness in dwellers of the highest city in the world.Front Physiol. 2020; 11: 773Crossref PubMed Scopus (9) Google Scholar,8Gonzales G.F. Rubio J. Gasco M. Chronic mountain sickness score was related with health status score but not with hemoglobin levels at high altitudes.Respir Physiol Neurobiol. 2013; 188: 152-160Crossref PubMed Scopus (16) Google Scholar To conclude, despite some limitations (eg, the lack of additional physiological data over the follow-up period), this study provides unique insights into the association between chronic severe hypoxic exposure and the occurrence of EE and CMS in high-altitude dwellers. Further studies are required for confirming a causality relationship. Role of sponsors: The sponsor had no role in the design of the study, the collection and analysis of the data, or the preparation of the manuscript. Chronic Mountain Sickness Evolving Over Time: New Data From on HighCHESTVol. 161Issue 5PreviewChronic mountain sickness (CMS) is characterized by excessive erythrocytosis, above the usual degree of polycythemia that develops in people living at high altitude, and symptoms arising from an increased red cell mass. A consensus panel has defined excessive erythrocytosis (EE), independent of any altitudinal gradient, as a hemoglobin concentration of > 21 g/dL in men and > 19 g/dL in women.1 CMS is defined by EE and a myriad of symptoms and signs that are suggestive of maladaptation, in part caused by EE, as quantified by the Qinghai CMS score. Full-Text PDF