Impacts of improved cookstove interventions on personal exposure to carbon monoxide and particulate matter in Zambia

Eighty-four percent of sub-Saharan African households rely on polluting fuels (e.g., wood, charcoal) for cooking, leading to high levels of household air pollution (HAP). While switching to modern fuels/stoves could decrease HAP levels, they are not always available or affordable. Improved biomass cookstoves could provide an intermediate step supporting transitions from traditional biomass to clean burning fuels/stoves. We conducted two stove intervention trials in Lusaka, Zambia using targeted marketing/incentives to motivate participants to use improved biomass stoves, either the Mimi Moto (pellet) or the EcoZoom (charcoal). Before the intervention, 65% of participants exclusively used charcoal, while 27% relied on electricity to some extent for cooking. We measured 24-hour personal exposure to CO (n=747) and PM2.5 (n=90) of primary cooks. We implemented several statistical approaches to estimate the effects of interventions on exposure: household-specific endline minus baseline exposure, ranksum testing, difference-in-differences analyses, and cross-sectional analyses. We did not find that switching from traditional charcoal stoves to either intervention stove was associated with significantly reduced exposures. However, cooks using electric stoves independent of the intervention did have significantly lower CO exposures than those using traditional charcoal, with greater electric stove use corresponding to greater exposure reductions. Variability in exposure was dominated by seasonal, regional, and neighborhood differences rather than household stove/fuel choices. A focus on HAP exposure from cooking in urban settings is unlikely to yield expected exposure reductions. Policy makers should consider pollution reduction policies/interventions that target ambient air quality in tandem with HAP-mitigating strategies to address air pollution health burden.

Methods
Data:

We conducted two stove intervention trials in Lusaka, Zambia using targeted marketing and incentives to motivate traditional biomass stove users to switch to one of two improved biomass stoves, either the pellet burning Mimi Moto or the charcoal burning EcoZoom. Exposure data for this study was collected from primary cooks in Lusaka, Zambia in 2019 and 2021. Ambient data was collected using PurpleAir PM2.5 sensors available online at the time of the study (baseline, no longer available) and installed by researchers on the project (endline). Raw and processed data are included in CSVs. Refer to the descriptions below for what is included in each CSV file. Note that household ID (HHID) is used to differentiate households/primary cooks and match exposure data to questionnaire responses.

Hourly exposure data:

ZCCS_2019_CO_hourly_CSV - hourly CO exposure in ppm for all participants in 2019 where hour_0 = 00:00, hour_1 = 01:00, etc.

ZCCS_2019_PM_hourly_CSV - hourly PM2.5 exposure in μgm-3 for all participants in 2019 where hour_0 = 00:00, hour_1 = 01:00, etc.

ZCCS_2019_EM_temp_hourly_CSV - hourly temperature recorded by PM2.5 exposure monitors in °C for all participants in 2019 where hour_0 = 00:00, hour_1 = 01:00, etc.

ZCCS_2021_CO_hourly_CSV - hourly CO exposure in ppm for all participants in 2021 where hour_0 = 00:00, hour_1 = 01:00, etc.

ZCCS_2021_PM_hourly_CSV - hourly PM2.5 exposure in μgm-3 for all participants in 2021 where hour_0 = 00:00, hour_1 = 01:00, etc.

ZCCS_2021_EM_temp_hourly_CSV - hourly temperature recorded by PM2.5 exposure monitors in °C for all participants in 2021 where hour_0 = 00:00, hour_1 = 01:00, etc.

Processed exposure data, participant questionnaire results:

ZCCS_2019_ODK_HH_char_CSV - compilation of household and primary cook characteristics from questionnaires for 2019 participants used in Table 2 (e.g., primary cook education level, cook gender)

ZCCS_2019_2021_combined_EM_CSV - baseline (2019) and endline (2021) data matched for each participant; used for Figure 5.

ZCCS_2019_2021_Cross_Sectional_All_EM_CSV - compilation of all averaged exposure data and questionnaire data where participants for each year are treated as individual entries (as opposed to the same entry as in ZCCS_2019_2021_combined_EM_CSV); used for statistical analyses (Tables 3-4) and Figure 6.

PurpleAir raw data files:

Kabwe_1_Primary - raw data from Kabwe 1 PurpleAir sensor A (no longer available) from July to August 2019.

Kabwe_1_B_Primary - raw data from Kabwe 1 PurpleAir sensor B (no longer available) from July to August 2019.

Kabwe_2_Primary - raw data from Kabwe 2 PurpleAir sensor A (no longer available) from July to August 2019.

Kabwe_2_B_Primary - raw data from Kabwe 2 PurpleAir sensor B (no longer available) from July to August 2019.

CEEEZ_Primary - raw data from Lusaka CEEEZ PurpleAir sensor A from October to November 2021.

CEEEZ_B_Primary - raw data from Lusaka CEEEZ PurpleAir sensor B from October to November 2021.

SupaMoto_Primary - raw data from Lusaka SupaMoto PurpleAir sensor A from October to November 2021.

SupaMoto_B_Primary - raw data from Lusaka SupaMoto PurpleAir sensor B from October to November 2021.

Stove influenced (SI) hours:

ZCCS_2019_CO_SIperiods_7_20_2 - hourly mask variables for each HHID in 2019 where 0 means the primary cook was not experiencing 'stove-influenced' concentrations during that hour, while a 1 means they were; SI periods were estimated using CO exposure concentrations, n=7, alpha=20, and beta =2

ZCCS_2021_CO_SIperiods_7_20_2 - hourly mask variables for each HHID in 2021 where 0 means the primary cook was not experiencing 'stove-influenced' concentrations during that hour, while a 1 means they were; SI periods were estimated using CO exposure concentrations, n=7, alpha=20, and beta =2

ZCCS_2021_PM_SIperiods_7_20_2 - hourly mask variables for each HHID in 2021 where 0 means the primary cook was not experiencing 'stove-influenced' concentrations during that hour, while a 1 means they were; SI periods were estimated using PM2.5 exposure concentrations, n=7, alpha=20, and beta =2

PurpleAir raw data was corrected in the Jupyter Notebook file using the following literature corrections:

McFarlane, C., Isevulambire, P.K., Lumbuenamo, R.S., Ndinga, A.M.E., Dhammapala, R., Jin, X., McNeill, V.F., Malings, C., Subramanian, R., Westervelt, D.M., 2021a. First Measurements of Ambient PM2.5 in Kinshasa, Democratic Republic of Congo and Brazzaville, Republic of Congo Using Field-calibrated Low-cost Sensors. Aerosol Air Qual. Res. 21, 200619. https://doi.org/10.4209/aaqr.200619

McFarlane, C., Raheja, G., Malings, C., Appoh, E.K.E., Hughes, A.F., Westervelt, D.M., 2021. Application of Gaussian Mixture Regression for the Correction of Low Cost PM2.5 Monitoring Data in Accra, Ghana. ACS Earth Space Chem. 5, 2268–2279. https://doi.org/10.1021/acsearthspacechem.1c00217

Barkjohn, K.K., Gantt, B., Clements, A.L., 2021. Development and application of a United States-wide correction for PM2.5 data collected with the PurpleAir sensor. Atmospheric Measurement Techniques 14, 4617–4637. https://doi.org/10.5194/amt-14-4617-2021

Holder, A.L., Mebust, A.K., Maghran, L.A., McGown, M.R., Stewart, K.E., Vallano, D.M., Elleman, R.A., Baker, K.R., 2020. Field Evaluation of Low-Cost Particulate Matter Sensors for Measuring Wildfire Smoke. Sensors 20, 4796. https://doi.org/10.3390/s20174796

Magi, B.I., Cupini, C., Francis, J., Green, M., Hauser, C., 2020. Evaluation of PM2.5 measured in an urban setting using a low-cost optical particle counter and a Federal Equivalent Method Beta Attenuation Monitor. Aerosol Science and Technology 54, 147–159. https://doi.org/10.1080/02786826.2019.1619915

Software:

Data processing, analysis, and visualization was completed in a Jupyter Notebook available on GitHub (https://github.com/stephanieparsons14/Impacts-of-improved-cookstove-interventions-on-personal-exposure-to-CO-and-PM-in-Zambia) and preserved in Zenodo (https://zenodo.org/doi/10.5281/zenodo.13245015).

Code from the following website was referenced for the propensity score matching used for difference-in-differences analyses: https://www.r-bloggers.com/2022/04/propensity-score-matching/