Introduction

Anemia is a major global health concern, particularly among children and women of reproductive age in low- and middle-income countries (LMICs)1,2. Approximately one-third of women of reproductive age worldwide are affected, with prevalence in South Asia ranging from 40 to 52%2,3. Anemia has been associated with numerous adverse maternal and neonatal outcomes, including maternal mortality, preterm birth, impaired cognitive function, and cardiovascular disorders4,5,6,7. Identifying modifiable risk factors and implementing timely interventions for women is therefore a critical global health priority8.

Air pollution represents another major public health threat, contributing to an estimated 6.67 million deaths worldwide9, with women being particularly susceptible10. Emerging evidence suggests that long-term exposure to fine particulate matter (PM2.5) is linked to adverse hematological outcomes11,12,13,14,15. Studies from China, India, and the United States report that PM2.5 exposure increases the odds of anemia by 7–33%11,15,16,17. Proposed mechanisms include systemic inflammation, oxidative stress, altered DNA methylation, and disrupted iron metabolism18,19,20. Elevated oxidative stress and inflammation can impair hematopoietic function, disrupt the production of red blood cells and reduce hemoglobin (Hb) levels, ultimately contributing to anemia17,20,21. However, most existing studies focus on children16,22, specific groups such as pregnant women and older adults12,17,23, and indoor PM2.5 exposure20,24, leaving a gap in our understanding of its impact on women of reproductive age in LMICs.

In South Asia, ambient air pollution, particularly PM2.5, is a growing environmental concern16,25. In Nepal, air pollution affects both urban and rural populations26, yet existing studies are limited to respiratory, cardiovascular, or mental health conditions27,28,29,30. Rapid urbanization, increased vehicular traffic, and forest fires have worsened air quality, and a newborn in Nepal may lose up to two years of life expectancy due to air pollution exposure9,31. Limited monitoring infrastructure and the country’s topography, which traps pollutants, further exacerbate health risks27,32. Major cities frequently exceed WHO air quality guidelines by more than fivefold30,31, highlighting the need to understand environmental contributions to anemia in this population. To date, only one study has explored the link between air pollution and anemia, and it focused solely on children under five22. In addition, despite national efforts, anemia prevalence among women of reproductive age in Nepal has remained stagnant at approximately 35% over the past decade33.

In the present study, we show that increased exposure to PM2.5 is associated with lower blood Hb levels and an increased risk of anemia. The risk is more pronounced among women with lower educational attainment and those residing in mountainous regions. These findings highlight the need for targeted public health actions and improved air quality management to mitigate anemia risk in vulnerable populations. The purpose of the present study was to examine the association between PM2.5 exposure, blood Hb levels, and anemia among women aged 15–49 years using nationally representative samples from the 2022 Nepal Demographic and Health Survey (NDHS).

Methods

Study settings and population

This study used nationally representative data from the 2022 NDHS. The NDHS 2022 employs an updated sampling framework derived from the 2011 Housing and Population Census. The survey, administered by the Ministry of Health and Population (MoHP), was conducted over a period from January 5 to June 22, 2022. To ensure national representativeness, each of Nepal’s provinces was stratified into urban and rural areas, resulting in 14 sampling strata. A total of 476 primary sampling units (PSUs) were selected using a probability-proportional-to-size method, 248 from urban areas and 228 from rural areas. From each PSU, 30 households were randomly selected, yielding a total sample of 14,280 households (7440 urban and 6840 rural). Among the surveyed households, 15,238 women aged 15–49 were identified as eligible for the NDHS, and 14,845 were successfully interviewed, yielding a response rate of 97.42%. Further details on the NDHS 2022 procedure are available elsewhere34. Of these 14,845 women, after screening, we selected a weighted sample of 4133 women aged 15–49 years who met all of the following inclusion criteria: 1) available data on Hb levels, anthropometric measurements (height and weight), and systolic and diastolic blood pressure (SBP and DBP); 2) residency in the same province for at least five years; and 3) available data on study covariates (Supplementary Fig. S1).

Outcome measures

The primary outcomes of the study were blood Hb levels and anemia. In the NDHS, blood samples were collected via finger prick, and Hb levels were measured on-site using a HemoCue analyzer. Additionally, Hb levels in the NDHS dataset were adjusted for smoking status and altitude35. Anemia was defined based on the Hb levels of each participant, with values < 12 g/dl for non-pregnant women and <11 g/dl for pregnant women, following the WHO guidelines36.

Air pollution exposure

We obtained monthly average concentrations of particulate matter <2.5 μm in diameter (PM2.5) provided by air quality monitoring stations located in each province nationwide26. These measurements adhered to the standard reference protocol outlined in the National Ambient Air Quality Monitoring Standard, Government of Nepal. The Department of Environment has installed 27 real-time air quality monitoring stations across all provinces, equipped with EDM Grimm 180+ devices to measure PM2.526. These monitoring stations are distributed across seven provinces, with the number of stations varying by province (Fig. 1). Using this data, we calculated the yearly average concentrations from August 2021 to July 2022 by aggregating data from all monitoring stations within each province while accounting for the number of stations in each region. This one-year exposure period was selected because annual PM levels provide a more stable and representative measure of long-term exposure than short-term values. This approach aligns with prior studies across multiple health outcomes that have used annual exposure levels even when surveys were conducted over shorter time frames37,38, and it better captures the exposure window relevant for conditions like anemia, which develop over weeks to months. Given that we did not have access to participants’ exact home addresses, we employed a semi-ecological approach, assigning pollutant exposure levels to all participants who had lived in the same domiciles for over five years. Consequently, exposure levels were matched using the annual mean concentration for the administrative division corresponding to each participant’s provincial residence. However, assigning province-level exposure to individuals may lead to potential non-differential exposure misclassification. Due to the lack of more granular PM exposure data and the absence of exact residential addresses for participants in NDHS, we were unable to assign exposure at a finer spatial resolution than the provincial level.

Fig. 1: Mean PM2.5 levels during the study period and the geographical distribution of air quality monitoring stations across Nepal.
Fig. 1: Mean PM2.5 levels during the study period and the geographical distribution of air quality monitoring stations across Nepal.The alternative text for this image may have been generated using AI.
Full size image

Black triangle points indicate the geographic location of an individual station.

Covariates

We incorporated a range of covariates, including age (years), education level (none, primary, secondary, higher), marital status (yes, no), currently pregnancy (yes, no), wealth index (poorest, poorer, middle, richer, richest), residency area (urban, rural), ecological region (hill, mountain, terai), smoking status (yes, no), body mass index (BMI, kg/m²), mid-upper arm circumference (MUAC), obesity (yes, no), hypertension (yes, no), temperature, and relative humidity. These variables were selected based on previous studies12,16,17,23,39,40,41,42,43. The wealth index was calculated using principal component analysis based on the wealth quintiles of household assets34,44. In the NDHS, it was determined following Demographic and Health Surveys (DHS) guidelines44, which consider factors such as home ownership, construction materials, cooking fuel, sanitation, water access, handwashing facilities, and asset ownership. The resulting wealth index is divided into five quintiles—poorest, poorer, middle, richer, and richest—which are available in the NDHS datasets used for this study34. Prior studies suggest that factors such as BMI, obesity, pregnancy status, and hypertension are associated with anemia41,45,46 and may therefore influence the relationship between PM2.5 exposure and anemia23,47,48. In NDHS, BP was monitored using a multi-user upper arm BP monitor with automatic inflation and pressure release, along with an appropriately sized cuff34. The final BP value was the average of the second and third readings, which were then used to classify hypertension. Hypertension was defined as systolic pressure ≥140 mmHg and/or diastolic pressure ≥90 mmHg, per WHO guidelines49. Participants who reported taking any antihypertensive medications were also classified as hypertensive, following the National Guidelines for the Management of Hypertension in Nepal and previous studies50,51,52. BMI (kg/m2) was calculated based on the measured height (cm) and weight (kg) of the participants. Obesity was defined as a BMI of ≥25 kg/m², following WHO guidelines for Asian populations53. The NDHS also measured women’s MUAC, with undernutrition classified as MUAC < 23 cm for this study54. Meteorological factors, including temperature (°C) and relative humidity (%), are associated with and potentially influence the relationship between air pollution exposure and health outcomes13,27,40. We obtained meteorological data, including minimum and maximum temperatures and humidity, from the Department of Hydrology and Meteorology for the same period as the PM2.5 data55.

Statistical analyses

We conducted a weighted sample analysis to account for the complex survey design of the NDHS 2022. In the NDHS, sampling weights are applied to each case to adjust for differences in the probability of selection and interview. These weights are crucial for obtaining accurate case numbers for the study variables. To apply the correct weights, we utilized the weight variable for women in the dataset, following the DHS sampling weight method34. Baseline information was presented by anemia status using descriptive statistics, including means (SD), percentages, and frequencies. Between-group differences were assessed using the Chi-square test for categorical data and the Student’s t-test for continuous data. The normality of the data was evaluated using the Shapiro-Wilk test and Q-Q plots. Given the minimal missing PM2.5 values (<3% of total observations) and randomly distributed missingness, we applied multiple imputation with predictive mean matching, generating 10 imputed datasets56. Considering prior studies suggesting a non-linear relationship between pollution exposure and outcomes such as Hb levels or anemia12,14, we incorporated quadratic polynomial regression into the analysis models57 to assess the association between PM2.5 exposure and health outcomes. The quadratic polynomial regression model was selected based on its lower AIC value compared with models using other polynomial degrees. The beta coefficients (β) and odds ratios (ORs) with 95% confidence intervals (CIs) for a 10 µg/m³ increase in PM2.5 were estimated using three adjusted models. The analysis models were developed as follows: The basic model (Model 1) was adjusted for age, pregnancy status, BMI, humidity, and temperature. The intermediate model (Model 2) further adjusted for education status, marital status, wealth index, residency area, and ecological region in addition to variables from Model 1. The full model (Model 3) included additional adjustments for smoking status and hypertension status, alongside the variables included in Models 1 and 2. Model fit was assessed using the Akaike Information Criterion (AIC) and Nagelkerke’s R². We also conducted stratified analyses by age groups, marital status, education level, pregnancy status, residency area, ecological region, wealth index, smoking status, MUAC, obesity, and hypertension status to estimate the odds of anemia associated with PM2.5 exposure. Multicollinearity was evaluated using the variance inflation factor. In addition, we performed sensitivity analyses to assess the robustness of the findings. First, we restricted the analysis to non-pregnant women11,23. Second, we estimated the association between PM2.5, Hb levels or anemia among individuals residing in urban and rural areas58. The model is considered robust if the associations remain unchanged from the main findings. All statistical analyses were performed using R (Version 4.4.3), and a two-sided p-value of <0.05 was considered statistically significant for all analyses.

Ethical declarations

We obtained approval from the DHS to access and use the dataset for this study (no. 2024-09-02). All datasets used in this study are secondary and publicly accessible through the official website of the DHS. The 2022 NDHS received ethical approval from the Institutional Review Board of ICF International in the United States (180657.0.001.NP.DHS.01). This study adhered to the principles outlined in the Declaration of Helsinki and followed the STROBE guidelines. During the NDHS 2022, written informed consent was obtained from all adult participants. For participants under the age of 18, consent was obtained from their parents or legal guardians, along with assent from the minors themselves.

Results

A total of 4133 women (weighted sample) aged 15–49 years were included after applying the inclusion criteria. The mean age of the participants was 29.78 ± 9.61 years, and the prevalence of anemia was 35.2% (95% CI: 34.0–36.9%). The baseline characteristics of the participants according to anemia status are summarized in Table 1. Significant differences across baseline characteristics, including age, marital status, pregnancy status, wealth index, ecological region, smoking status, Hb levels, BMI, and MUAC, were observed between women with and without anemia (all p < 0.05). Women with anemia exhibited lower rates of obesity (23.3% vs. 33.1%, p < 0.001) and hypertension (5.4% vs. 9.0%, p < 0.001) compared to those without anemia. The average concentration of PM2.5 across Nepal during the study period is illustrated in Fig. 1. The minimum and maximum levels of PM2.5 were 28.56 µg/m³ and 58.82 µg/m³, respectively, indicating that these levels exceeded the WHO threshold for health-exposure risk to the general population in all provinces59.

Table 1 Baseline characteristics by anemic status in women of reproductive age (n = 4133)
Full size table

The associations between PM2.5 exposure, Hb levels, and anemia—analyzed using adjusted linear and logistic regression models (basic, intermediate, and full)—are presented in Fig. 2 and Fig. 3. Each 10 µg/m³ increment in PM2.5 levels was inversely associated with Hb levels across all three models: Model 1 (β, 95% CI: −0.093, −0.181 to −0.036), Model 2 (β, 95% CI: −0.139, −0.207 to −0.102), and Model 3 (β, 95% CI: −0.161, −0.228 to −0.099). Similarly, each 10 µg/m³ increase in PM2.5 exposure was positively associated with anemia across all three models: Model 1 (OR, 95% CI: 1.08, 1.03–1.19), Model 2 (OR, 95% CI: 1.23, 1.11–1.35), and Model 3 (OR, 95% CI: 1.29, 1.14–1.42). Notably, Hb levels significantly decreased, and the odds of anemia significantly increased as additional covariates were adjusted for in the analysis. (Fig. 2 and Fig. 3)

Fig. 2: Association between PM2.5 exposure and blood hemoglobin levels.
Fig. 2: Association between PM2.5 exposure and blood hemoglobin levels.The alternative text for this image may have been generated using AI.
Full size image

β represents the coefficients for each 10 µg/m3 increase in PM2.5 estimated using a quadratic polynomial linear regression model adjusted for covariates. Model 1 (basic): adjusted for age, pregnancy status, BMI, humidity, and temperature. Model 2 (intermediate): adjusted for age, pregnancy status, BMI, humidity, temperature, education status, marital status, wealth index, residency area, and ecological region. Model 3 (full): adjusted for Model 1 + Model 2 + MUAC, smoking, and hypertension status. Total sample size (n = 4133) is indicated for the analysis. A two-sided p-value of <0.05 was considered statistically significant for the analysis. The p-values for Model 1, Model 2, and Model 3 were 0.006, <0.001, and <0.001. CI confidence interval.

Fig. 3: Association between PM2.5 exposure and anemia.
Fig. 3: Association between PM2.5 exposure and anemia.The alternative text for this image may have been generated using AI.
Full size image

ORs represent the odds ratios for each 10 µg/m3 increase in PM2.5, estimated using a quadratic polynomial logistic regression model adjusted for covariates. Model 1 (basic): adjusted for age, pregnancy status, BMI, humidity, and temperature. Model 2 (intermediate): adjusted for age, pregnancy status, BMI, humidity, temperature, education status, marital status, wealth index, residency area, and ecological region. Model 3 (full): adjusted for Model 1 + Model 2 + MUAC, smoking, and hypertension status. Total sample size (n = 4133) is indicated for the analysis. A two-sided p-value of <0.05 was considered statistically significant for the analysis. The p-values for Model 1, Model 2, and Model 3 were 0.027, <0.001, and <0.001. CI, confidence interval.

We further conducted stratified analyses to estimate the risk of anemia associated with PM2.5 exposure across baseline variables (Fig. 4). Women with higher education levels exhibited lower odds of anemia associated with PM2.5 exposure than those without any formal education (OR, 95% CI: 0.96, 0.94–0.98, p < 0.009). Women residing in the hilly region were at a lower risk of anemia due to PM2.5 exposure compared to those in the mountain region (OR, 95% CI: 0.97, 0.93–0.99, p < 0.037). Additionally, the sensitivity analyses indicate that the associations between PM2.5 exposure, Hb levels or anemia are robust, even after restricting the analysis to non-pregnant women and to participants residing in urban and rural areas (Supplementary Table S1).

Fig. 4: Stratified analysis of the association between PM2.5 exposure and anemia.
Fig. 4: Stratified analysis of the association between PM2.5 exposure and anemia.The alternative text for this image may have been generated using AI.
Full size image

OR odds ratios for each 10 µg/m3 increase in PM2.5; CI, confidence interval. Wealth index (reference: poorest), ecological region (reference: mountain), and education (reference: none). Total sample size (n = 4133) is indicated for the analysis. A two-sided p-value of <0.05 was considered statistically significant for the analysis.

Discussion

This study aimed to investigate the association between long-term exposure to PM2.5 and anemia in women of reproductive age in Nepal. Our findings demonstrate that prolonged exposure to PM2.5 is significantly associated with decreased Hb levels and an increased risk of anemia. In the fully adjusted model, each 10 μg/m³ increase in PM2.5 exposure was associated with a 29% increase in anemia risk. Given the baseline prevalence of anemia was 35%, this corresponds to a substantial increase, underscoring the notable impact of air pollution on women’s health.

Prior epidemiological studies have suggested a significant association between ambient air pollutants and anemia across various population groups11,12,13,14,15,17,39, a finding supported by our results. For instance, studies among American17 and Chinese adults12 reported that long-term exposure to PM2.5 was positively associated with anemia prevalence. However, these prior studies primarily focused on specific subgroups such as the elderly, children, or pregnant women14,16,18,22, whereas our study extends these findings to a broader population of women. Furthermore, our research provides new evidence specific to Nepalese women and suggests a possible dose–response relationship between PM2.5 exposure and anemia risk. However, longitudinal observational studies are needed to confirm this association. Nevertheless, it is particularly concerning, as anemia during the reproductive years can lead to multiple adverse outcomes, including stillbirth, low birth weight, and preterm delivery, thereby negatively affecting both maternal and child health5,6.

Several possible biological mechanisms have been documented linking air pollution—particularly PM2.5 exposure—to reduced Hb levels and anemia through systemic inflammation and oxidative stress13,20,21,24. PM2.5 can trigger chronic inflammation, leading to impaired erythropoietin function in the kidneys and increased resistance to this essential hormone, which is critical for red blood cell production17,21. Additionally, inhalation of PM2.5 may elevate pro-inflammatory cytokine levels, which stimulate hepcidin production. Elevated hepcidin disrupts iron homeostasis by impairing intestinal iron absorption and increasing iron sequestration in macrophages, ultimately inhibiting Hb synthesis and contributing to anemia14,24. However, these mechanisms may vary across populations and geographic regions due to genetic and biological differences60,61. Therefore, large-scale clinical studies that incorporate gene–environment interactions are warranted to more comprehensively elucidate the pathways through which PM2.5 exposure influences Hb levels and contributes to anemia in the general women population.

South Asian regions, including Nepal, exhibit a heightened prevalence of anemia among women of reproductive age1, with multiple sociodemographic factors contributing to this burden3,62. Our findings highlight air pollution as a significant additional contributor, further informing health initiatives aimed at reducing anemia in the region. Notably, we observed a significant association between education level and anemia risk among women exposed to PM2.5. Education level has been recognized as a key determinant of anemia among Nepalese women and in other LMICs33,62. Higher education is often linked to better health literacy, enabling women to make healthier dietary choices and adhere to preventive health practices, including iron supplementation. Educated women may also have higher socioeconomic status, allowing for improved housing quality, cleaner cooking fuels, and reduced exposure to indoor air pollution62,63, which can mitigate the impact of PM2.5 on anemia risk. Additionally, educated individual might better access to health information, which may further protect against anemia. Moreover, geographic variation in vulnerability was evident, with women residing in hilly regions exhibiting a lower risk compared to those in mountainous areas. This disparity may be attributed to differences in PM2.5 composition, pollution dispersion, meteorological conditions, or lifestyle factors across ecological regions28,29,40,64. Residents at higher altitudes adapt with increased hemoglobin levels65, which may reduce apparent anemia prevalence, while geographic variation in reliance on biomass fuels and healthcare practices may influence the impact of air pollution on health outcomes66. In particular, mountainous terrain constrains wind flow and enhances the retention of air pollutants, heightening exposure risks for nearby communities67. These findings highlight the need for further clinical investigation into the susceptibility of specific subgroups of women to air pollution exposure, with these groups prioritized in targeted interventions.

Our findings have significant public health implications, particularly given that ambient air pollution contributes substantially to the global burden of disease, as reported by the State of Global Air9. The levels of PM2.5 concentration observed across Nepal exceed the WHO guideline values59, placing the entire female population at risk of adverse health outcomes. Given the stagnation in anemia prevalence over the past decade in Nepal, these findings underscore the urgent need for integrated approaches that address both traditional risk factors, such as nutritional deficiencies68 and environmental determinants like air pollution32. Evidence also suggests that vitamin B supplementation may attenuate the epigenetic effects of PM2.519. Future research should explore whether similar nutritional interventions could mitigate the impact of PM2.5 exposure on anemia risk among women. Additionally, a study from Korea found that reducing PM concentrations using air purifiers had a positive effect on anemia-related blood indicators24. Such interventions could be particularly beneficial for high-risk groups identified in our study, such as women with lower education levels and those residing in mountainous regions. Therefore, incorporating these factors into the development of health policies and environmental strategies may be crucial for effectively addressing the burden of anemia in women of reproductive age.

Our study presents some notable strengths. This study investigates the association between PM2.5 exposure and anemia among women of reproductive age in Nepal using nationally representative large-scale data. The analysis employed rigorous statistical approaches, including adjusted regression models and stratified analyses, to address potential confounding and identify subpopulations at greater risk. Additionally, the findings expand the current understanding of air pollution’s health impacts by highlighting its potential role in anemia, thereby extending the evidence base beyond established links with respiratory, cardiovascular, and mental health outcomes27,28,29. However, there are several limitations that should be acknowledged. First, the cross-sectional design restricts the ability to establish causal relationships between exposure and health outcomes39. Therefore, future longitudinal studies are needed to determine the temporal association between PM2.5 exposure and anemia. Second, assigning exposure based solely on provincial residential addresses may not accurately reflect individuals’ true exposure levels and could result in exposure misclassification. This limitation is particularly relevant because individual exposure can vary substantially due to factors such as time spent in different occupational settings, daily commuting patterns, and varying levels of indoor air pollution. Third, reliance on self-reported data may introduce biases, and the use of secondary data from the NDHS 2022 limits the authors’ control over the study design. Fourth, this study focused solely on PM, as these are the primary pollutants monitored in Nepal, and data on other pollutants remain limited. However, several air pollutants, including NO2, CO, SO2, and O3, have demonstrated a significant impact on Hb levels and anemia, both individually and through their interactions12,13,23,39,69. Despite this limitation, the key findings of the study emphasize the need for enhanced air quality monitoring in Nepal. It calls on the government and policymakers to invest in additional monitoring stations to track a wider array of pollutants and evaluate their potential health impacts. Such an initiative would promote future research in Nepal, expanding the scope of studies to include various pollutants and providing a more comprehensive understanding of the relationship between air pollution exposure and multiple health outcomes. Fifth, the study did not account for dietary or nutrition patterns, such as iron supplementation, physical activity, or mental health conditions—all of which may influence the outcomes64,68,70. Sixth, the majority of households (>70%) in Nepal rely on solid or unclean cooking fuels, and women, who spend substantial time on household activities, including cooking, are disproportionately exposed71. Household indoor air pollution resulting from these fuels has been associated with anemia23,72, but was not accounted for in this study, which may have influenced the observed associations. Therefore, interventions focusing on improving air quality through clean cooking programs may help mitigate the risk of anemia. Seventh, while this study adjusted for meteorological factors such as temperature and humidity, other seasonal variations, including specific seasons and monsoon patterns67,73, were not accounted for and may still influence the study findings. Finally, we did not clinically characterize anemia types (iron-deficiency, megaloblastic, or others), which may exhibit distinct relationships with environmental exposures74. Therefore, addressing these limitations in future research is crucial for advancing our understanding of the complex relationship between air pollution and anemia among women of reproductive age.

Conclusions

The findings indicate that long-term exposure to PM2.5 reduces blood Hb levels, thereby increasing the risk of anemia among women of reproductive age. Women with lower education levels and those residing in mountainous regions appear particularly vulnerable to the associated risks. Given the high prevalence of anemia and elevated PM2.5 exposure levels exceeding WHO guidelines across Nepal, these findings underscore the urgent need of integrating air pollution awareness into anemia prevention and control initiatives. As exposure data constraints are a major issue in Nepal, enhancing PM monitoring in rural areas—where a large proportion of the population relies on unclean cooking fuels—is needed to accurately assess exposure and design effective interventions. Future longitudinal studies are warranted to investigate the causal relationship and to explore potential interventions aimed at mitigating these adverse health effects.