This hospital-based study conducted in Lhasa showed low Hb concentration among pregnant women. The average hemoglobin concentration was 127.6 g/L, lower than the 137.1 g/L among those residing at about the same altitude (3600 m) in Bolivia [14
]. The Hb level for most of the participants in our study was lower than the WHO criteria values related to long-term high-altitude exposure [1
Our study showed that pregnant Tibetans had lower Hb levels than non-Tibetans. The Tibetans are the aboriginal inhabitants in Tibet with a long history [15
]. Some physical adaptations have induced specific changes in Hb with increasing altitude. For example, Tibetans had higher vital capacity, thus can increase capacity to move more oxygen through the lungs which might be expected to result in more oxygen in the bloodstream [16
]. Another important physiological feature found particularly among Tibetans was a denser capillary network [18
]. It could make more thorough exchange of oxygen. As per the above adaptations, we suggested that Tibetans have an improved capability for accommodating the low-oxygen circumstances without having an increased concentration of hemoglobin. However, most of the non-Tibetans in the region were immigrants whom from low-lying provinces and their physiological adaptation for environment of high altitude might be different from the Tibetan, which could contribute to the ethnic differences in Hb level and prevalence of anemia to some extent. This needs further investigation.
A previous study has shown a U-shaped relationship between Hb and gestational age [19
], which is generally attributed to physiological hemodilution. Hb levels appear to be lowest when dilution is at its maximum (at seven to eight months' gestation). This study showed that Hb level among pregnant women living in Tibet decreased with increase in gestational age. The Hb of urban pregnant women was lower than that of rural pregnant women. Similar finding was found in a large study in China in 1992 [20
], but the reasons for such a finding are unclear.
According to Bessman et al.'s study [22
], iron deficiency anemia cases show specific erythrocyte changes, such as an MCV decrease and RDW-CV increase. In our study, although the prevalence of anemia was 70.0% (CDC method's), but neither MCV nor RDW-CV showed any change. As such, we could not conclude that the population in our study suffered iron deficiency anemia. MCH indicates the average hemoglobin concentration in each erythrocyte. The increase of MCH may suggest a physiological adaptation differences in this population.
Using three different altitude correction procedures (CDC method, Dirren et al. method, and Dallman et al. method), we estimated the prevalence rate of anemia being 70.0%, 77.9% or 41.3%, respectively, in contrast with 19.5%, 13.6% and 28.9% for pregnant women in Beijing [23
], Shanghai [24
] and other part of China [21
]. Although the anemia was quite high among the pregnant women in our study, most of anemic women were not with typical clinical symptoms of anemia. For example, in all 380 women included in this study, only 14% had frequent dizziness, 18% reported lack of physical strength, and about 8% were found to have pallid palpebral conjunctiva. Therfore, we concluded that these three methods were probably not suitable for correcting Hb for altitude for pregnant women living in the highlands of Tibet. The most important reason for this may be that the altitude ranges for those three methods were all between 0 to 3200 meters, while the participants in this study were living in the altitude of 3680 m. So extrapolations of these three methods may lead to unreliable results. The second reason may be that different methods used different fitting curves. The CDC's is conic, while that for the above-3000 m results is steep. Dirren et al. is an exponential curve. Although it has a gentle slope, it had a small sample size at each altitude. In comparison, Dallman et al. is a straight line. Other previous studies have shown that Hb did not increase linearly with higher altitudes [14
]. However, so far it is still unclear what relationship between Hb and altitude for Tibetan women is.
Using a survey dataset on nutritional status of younger children living on the Qinghai-Tibet Plateau, Dang et al showed a positive association between Hb concentrations and altitude, but that the results did not contrast dramatically with these from the CDC, Dirren et al. and Dallman et al. methods [25
]. The clear relationship between altitude and Hb concentration has still not been established yet in pregnant women living in the Qinghai-Tibet Plateau. As such, we applied three available methods to assess the prevalence of anemia but found very inconsistent prevalence rates. By comparing the consistency among these three methods, we concluded that the CDC method might be slightly better until a new hemoglobin-altitude formula for people living at this altitude can be established.
This is the first study that described the pattern of Hb concentration and prevalence of anemia among pregnant women living in the highlands of Tibet. The hospital-based cross-sectional design and selection bias of the sample in this study may underestimate the real prevalence of anemia and may not provide direct epidemiological inference for causality. However, our study serves as useful information and indicates a need for further anemia studies and interventions for pregnant women in the highlands of Tibet. A smaller sample size among the first-trimester pregnancies, due to some participants' restrictive traditional beliefs, may have introduced bias into this study. Lack of measurements of serum indexes such as serum ferritin and transferrin receptor may have limited our understanding of the true prevalence of anemia. In addition, the results from this hospital-based study may be subject to unobserved confounding factors, which need further investigation.