As a burden of disease specific to a particular region, this study focused on CMS in the highland immigrant population. To our knowledge, this work is the first to introduce the concept of DALYs to highland medicine. In this study, we established the local DWs for CMS according to seven main symptoms, which have not been considered in previous global or regional BOD studies. This study is also the first to estimate the burden of CMS among the young male Chinese immigrant population in Tibet and it was based on the data from two recent annual surveys.
Most BOD studies use the selection of diseases states from the International Classification of Disease, Injuries, and Causes of Death, Tenth Revision (ICD-10) [30
]. In the ICD-10, CMS (namely, secondary polycythemia due to high altitude) is coded D75.1 [33
], but there are no further subdivided consequence categories. Considering the multiple symptoms characteristic of CMS, we did not merely establish a single DW for CMS because this illness is not characterized by a single health problem. In addition, it is difficult to accurately measure the DALYs with a single DW because the condition has a variable presentation that includes several different symptoms, each with its own severity level. Thus, we derived DWs for the detailed CMS symptoms and 21 potential health states that can result from those symptoms at varying degrees of severity. This division made the DWs more detailed and easier to handle. The calculation of the DWs in this study was based on the PTO methodology and used a panel approach to elicit valuations of the different CMS health states. The panel included representatives from four types of specialties, which may reflect different specialties.
In this study, the overall disease burden caused by CMS among the young male Chinese immigrant population was 35.54-37.09 DALYs/1,000 in the sample region. Although there are no comparable DALYs data specific to CMS from previous studies, we compared our data with other non-communicable diseases as a reference. According to the WHO disease and injury regional estimates for 2004 [34
], the DALYs/1,000 due to CMS in our study were lower than those attributable to neuropsychiatric conditions in China’s 15- to 44-year-old male population, which caused 40.66 DALYs/1,000. However, the CMS DALYs/1,000 were higher than those for any other type of non-communicable condition. Unfortunately, a comparison between this work and the GBD study of non-communicable conditions is only partly possible because of differences in the data collection strategies and in the populations studied. The sample in the present study did not represent an entire nation or region, but rather was a specific immigrant population with high susceptibility to CMS. Nevertheless, the burden of CMS in the immigrant population may be considered to be high because approximately 3 months of healthy life is lost per CMS patient. This loss equates to approximately 2 weeks per immigrant when considered in relation to the population as a whole.
In previous reports, CMS has been found to affect 5-15% of the population at and above 3,200 m in Andean countries [35
]. In a 4,300-m area in Peru, the prevalence was 6.8-15.4% in a miner population between the ages of 20 and 39. In Tibet, an overall prevalence of 5.6% in immigrant civilians was reported; in Lhasa immigrant civilians, the CMS rate was 2.2-8.7% [17
]. However, despite some differences in diagnostic criteria, the results of the different studies are partially comparable because of the variance in environment, population characteristics, and geographical factors. In a similar young soldier population in Tibet, the CMS prevalence was reported to range from 9.3% (in Lhasa at 3,650 m) to 30.4% (in an area at 5,000 m) [36
], which is similar to our findings (2.1-37.4%), although, our figures were lower for lower-altitude areas at 3,500-3,999 m above sea level (2.1-3.2%). The main reason for this discrepancy may be that their criteria were slightly lower (Hb
200 g/L and SaO2
85%). In another report from Peru, in which the criteria were Hb
213 g/L and SaO2
83%, a male population living at 4,300 m had a CMS prevalence of 15.6% [37
], which is similar to our findings (4,000-4,499 m: 13.8-14.8%).
In this study, the prevalence of CMS significantly increased with ascent. When the altitude increased from approximately 3,500 m to 5,500 m, the prevalence increased approximately ten fold. Similarly, both the individual DALYs and the DALYs/1,000 also increased significantly with altitude. This result indicates that higher altitude not only increases the prevalence of CMS but also contributes to an aggravated CMS disease burden. For example, in the 3,500-3,999 m area for the total immigrant population, every immigrant lost nearly one-and-a-half days of healthy life annually because of CMS. In contrast, in the 5,000-5,400 m area, they lost one-and-a-half months of healthy life each year. When considering only the CMS patient, in the 3,500-3,999 m area, every CMS patient lost nearly two months of healthy life annually; in the 5,000-5,400-m area, they lost nearly three-and-a-half months of healthy life annually.
The statistical analyses indicated that some factors, including age, HSYs, BP, HR, smoking, and occupation, not only significantly differed between the CMS and non-CMS groups but also associated with the individual disease burden in the CMS population. The CMS population had a higher age, HSYs, BP, HR, smoking rate, and proportion of E&C occupations. These factors will now be discussed in relation to previous studies.
The study from Peru suggested that CMS is a clinical manifestation of aging at high altitudes in native highlands [38
]. Other studies have also reported that high-altitude dwellers show earlier cardiovascular degenerative changes with aging [39
]. The results of the present study indicate that the CMS group was older than the non-CMS group and had a longer highland service history. Age and HSYs were also positively associated with the individual DALYs of the CMS patients. In the sample population, the men always joined the army at the ages of 17–19 and were sent to the highland for service after 0.5 to 1 year of training; therefore, subjects of the same age had similar highland service years. Thus, the effects of age and HSYs cannot be discussed independently because both variables represent aging in the highland. Aging in the highland causes not only a greater number of CMS patients but also a heavier CMS disease burden, i.e., prolonged time spent in the highland could cause a healthy person to develop CMS [17
] and could aggravate the severity of CMS symptoms [7
]. Thus, redeploying soldiers seriously affected by CMS to a lower-altitude service area would help to decrease the health lost for the highland service population.
A previous study found that subjects with CMS showed reductions in the response to the stimulation of the peripheral chemoreflexes and in baroreflex control of heart rate and blood pressure, which correlated with increases in CMS scores and in hemoglobin levels [41
]. It is also known that vasodilator action of hypoxia at the microcapillary level may be incapable of decreasing systemic hypertension [42
]. In this study, we found that, compared with the non-CMS group, members of the CMS group had higher BP (both SBP and DBP) and HR; this result is consistent with the findings of other reports [43
]. We also found that BP (especially DBP) and HR were also significantly positively associated with individual DALYs in CMS patients. This result implies a progressive impairment of cardiovascular regulation among the CMS patients. Accordingly, improving cardiovascular regulation may help to alleviate the disease burden of CMS.
As some scholars have suggested, smoking is a risk factor for altitude disease [9
]. Previous research in the male Han population has suggested that the prevalence of CMS is three times higher in smokers than in non-smokers [9
]. In the present study, the CMS group had a higher smoking rate than the non-CMS group. Moreover, smoking was also associated with higher individual DALYs in the CMS patients. The mechanism causing this effect could be that cigarette smoking worsens hypoxia, produces a lower oxygen-carrying capacity, causes centrilobular emphysema, reduces alveolar ventilation, and thus impairs cardiopulmonary function. Accordingly, smoking control among the highland immigrant population would decrease both the prevalence and the disease burden of CMS; the authorities should consider this possibility.
Patients with E&C occupations were more likely to develop CMS and had heavier disease burdens. The possible reason for this result is that these occupations are associated with a relatively heavy physical burden. Physical exertion leads to greater oxygen consumption, which aggravates hypoxia [47
]. These occupations may also be more likely to bring people into contact with dust, waste gas, and cold weather, which potentially increase the risk of CMS. Thus, immigrants with E&C occupations in the highland should have more protection for CMS prevention and more compensation for health damages.
Regarding symptom categories, the greatest BOD due to CMS is associated with headaches, sleep disturbances, and breathlessness/palpitations, which made up almost three fourths of the total BOD. The high contribution of these symptoms to BOD is due to both their high prevalence and their high DW. These three symptoms may play an important role in the course of CMS, and they should be considered in measures to improve the quality of life for immigrant highlanders.
Strengths and limitations
There are several strengths of this study. First, the validity of the CMS data obtained from the military medical surveys is likely to be high. The data were collected by trained investigators using consistent criteria; the investigators were able to adequately assess the young men’s health status in the vast majority of the cases. Second, this study improves on some earlier studies in which the comorbidity of DWs for different health states was determined by simple addition; this procedure leads to the DW exceeding 1 at the individual level. Comorbidity was handled in the DW and DALYs calculations in our work, which makes the results more reliable. Third, this study described the BOD based on data from recent years; therefore, the DALYs may represent the actual disease burden in the population. We attempted to improve the DALYs measure’s sensitivity by comparing two sequential years; the results of the two data sets revealed no significant differences. A further strength of this study is that the authors have made modest and sensible health improvement recommendations relating to the specific symptoms that were identified as problematic, thus avoiding the criticism made of many studies of this general type that, for a variety of reasons, the measures used are not suitable for use as a basis for resource allocation decisions [49
However, the results of this study should be interpreted with caution for the following reasons.
First, in the study of DW, cultural differences may affect the determination of PTO values, and different cultural backgrounds might account for variations in the DW ratings. Because of the relative localization of the expert panel, the DWs derived from this study are limited and should be considered regional DWs. However, because our primary aim was to focus on Tibet and because most highland immigrants reside in this territory, the cultural difference problem may not have a great impact on this study.
Second, The PTO method has been questioned for having a lower test-retest reliability than the time-trade-off technique, which is another common technique used in DALYs studies [50
]; however, this may be less of a problem when using the Delphi method.
Third, a case-based approach may cause a substantial increase in the amount of data required, whereas an integrated health registration system has not been used for the entire immigrant population; therefore, the authors selected a population of young servicemen as a sample. Consequently, the data were collected from a specialized male population in which age, diet, and behaviors were similar; this specialized population may limit the generalizability of the results to other populations.
Fourth, this study handled DALYs in the same way that we handled YLD. However, whether some people may have died from CMS is difficult to determine. Because of the confounding effects of other diseases and the difficulty attributing each case to CMS, some potential symptoms not listed in the Qinghai-CMS-questionnaire may have not been considered.
Moreover, the assumption that the duration of CMS is an entire year and the averaging of its severity over the course of a year are not precise; some symptoms may occur intermittently, and the severity may change over time. Nevertheless, a disease that is chronic in nature may not change too much on average over a year, and the purpose of this study was to investigate the population level, not individual level, CMS burden, therefore this assumption was acceptable in this study.
Finally, BOD studies can detect disease-specific and overall trends in BOD within a region or in a population over time [26
]. This study, however, did not analyze the tendency of CMS to vary within an area; we investigated the BOD only within two neighborhoods in two recent years using a simple cross-sectional investigation. Thus, future studies may be necessary.