Prevalence of anemia among high-altitude pregnant women in Shannan, China, and evaluation of altitude-adjustment methods: A cross-sectional study.

OBJECTIVE

To quantify how five altitude-adjustment algorithms affect anemia prevalence at 3560 m and to evaluate diagnostic agreement.

METHODS

We conducted a retrospective, registry-based cross-sectional study of 10 660 pregnant women receiving antenatal care in Shannan, China (3560 m) from January 2022 to December 2024. Each woman contributed one hemoglobin (Hb) measurement, which was assigned to a trimester by gestational week at sampling. Observed Hb values were adjusted using five algorithms (WHO 2024, WHO 2011, Centers for Disease Control and Prevention [CDC], Dirren, Dallman). Anemia thresholds were <110 g/L for early and late and <105 g/L for mid-pregnancy under WHO-2024; for the other algorithms, <110 g/L across trimesters. Outcomes included overall and trimester-specific prevalence and diagnostic agreement quantified by Cohen's κ.

RESULTS

Mean observed Hb decreased from 141.1 g/L in early pregnancy to 117.7 g/L in late pregnancy. Unadjusted anemia prevalence was 20.7%. After altitude adjustment, prevalence estimates were 69.3% (WHO 2024), 78.2% (WHO 2011/CDC), 52.9% (Dallman), and 82.9% (Dirren). Across algorithms, prevalence increased with advancing gestation. Agreement was perfect for WHO 2011 versus CDC (κ = 1.000), moderate for WHO 2024 versus CDC (κ = 0.773), and fair for Dirren versus Dallman (κ = 0.376).

CONCLUSION

Altitude adjustment increased estimated anemia prevalence by up to four-fold at 3560 m in China. Algorithms derived from non-high-altitude populations may misclassify anemia in these settings. Population-tailored standards that incorporate altitude and trimester are warranted to improve screening accuracy and guide interventions.