Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Jun 26.
Published in final edited form as: Int J Hyg Environ Health. 2025 Sep 1;270:114649. doi: 10.1016/j.ijheh.2025.114649

Household air pollution exposure in adult women is associated with increased carotid intima-media thickness: A cross-sectional study of the Household Air Pollution Intervention Network trial

Lindsay J Underhill a, Lisa de las Fuentes a,b, Laura Nicolaou c, Shakir Hossen d, Anaite Diaz-Artiga e, Ajay Pillarisetti f, Aris T Papageorghiou g, Florien Ndagijimana h, Ghislaine Rosa i, Gurusamy Thangavel j, John P McCracken n, Kalpana Balakrishnan j, Krishnendu Mukhopadhyay j, Kyle Steenland k, Lisa M Thompson k,l, Lance A Waller m, Maggie L Clark o, Michael A Johnson p, Sarada Garg j, Sankar Sambandam j, Suzanne M Simkovich q,r, Vigneswari Aravindalochanan j, Kendra N Williams c, Wenlu Ye s, Jennifer L Peel o, Thomas F Clasen k, William Checkley c, Victor G Davila-Roman a,t,*; for the HAPIN Investigators
PMCID: PMC13295014  NIHMSID: NIHMS2165640  PMID: 40897110

Abstract

Background:

Cardiovascular disease is the leading cause of global morbidity and mortality, disproportionately affecting people in low- and middle-income countries (LMICs). Biomass fuels used for cooking in LMICs contribute significantly to household air pollution (HAP), which has been associated with inflammation, oxidative stress, and other pathways linked to atherosclerosis. We evaluate the association between HAP exposure and atherosclerosis by use of carotid artery ultrasound.

Methods:

An exposure-response analysis was conducted using cross-sectional baseline data from 397 women aged 40–79 years from the Household Air Pollution Intervention Network (HAPIN) trial in Guatemala, India, Peru, and Rwanda. Participants underwent ultrasound evaluation of their carotid arteries to measure intima-media thickness (CIMT) and atherosclerotic plaques. Additionally, 24-h personal exposures to particulate matter (PM2.5), carbon monoxide (CO), and black carbon (BC) were assessed.

Findings:

Mean 24-h PM2.5 exposure was 119 μg/m3 (range 10–803), BC was 13 μg/m3 (range 1.1–72), and CO was 2.3 ppm (range 0–39). Mean and maximal unadjusted CIMT measurements were 0.64 ± 0.13 mm and 0.75 ± 0.14 mm, respectively. Prevalence of atherosclerotic plaques was 7.1 % (range: 0.8 %–11.6 % by country). In adjusted linear models, each 10 μg/m3 increase in PM2.5 was associated with a 0.001 mm increase in mean CIMT (95 % CI: 0 to 0.002) and a 0.002 mm increase in maximal CIMT (95 % CI: 0 to 0.003). For CO, each 10 ppm increase was associated with a 0.04 mm increase in maximal CIMT (95 % CI: 0 to 0.08), with the highest quartile of CO exposure having 0.036 mm and 0.05 mm greater mean and maximal CIMT (95 % CI: 0.002 to 0.07; 0.01, 0.09), respectively, than the lowest quartile. No significant associations were found between BC and CIMT or between any exposures and carotid atherosclerotic plaque.

Interpretation:

In this cross-sectional study, higher personal exposures to PM2.5 and CO were associated with greater mean and maximal CIMT, a well-established biomarker of atherosclerosis, further supporting the association between HAP and cardiovascular disease.

Keywords: Household air pollution, Atherosclerosis, Carotid artery ultrasound

1. Background

Cardiovascular disease is the leading cause of global morbidity and mortality, disproportionately affecting people living in low- and middle-income countries (LMICs). A major underlying cause of CVD is atherosclerosis, a process characterized by the accumulation of fatty and fibrous material in the innermost layer of the arterial wall (the intima) (Libby, 2019). This accumulation results from long-term exposure to risk factors such as high cholesterol, atherogenic lipoproteins, and inflammatory cytokines, as well as traditional cardiometabolic risk factors like diabetes, hypertension, and obesity. Environmental risk factors such as air pollution have also been associated with cardiovascular events such as myocardial infarction, heart failure, cardiac arrhythmias, and stroke. Household air pollution (HAP), generated by the burning of biomass, other solid fuels, and kerosene for cooking, heating, and lighting, affects an estimated 3.6 billion people worldwide and has also been associated with CVD (IHME). HAP exposure is high in LMICs, where inefficient stoves and inadequate ventilation lead to the indoor accumulation of HAP pollutants, such as fine particulate matter (PM2.5), carbon monoxide (CO), and black carbon (BC). These pollutants may drive systemic inflammation and oxidative stress, key mechanisms in the development of atherosclerosis and CVD. Despite the significant burden of HAP exposure in LMICs, its association with CVD remains poorly studied and thus represents an important public health gap.

Carotid intima-media thickness (CIMT) is an ultrasound measure of carotid arterial wall thickness and a well-established imaging biomarker of atherosclerosis. Increased CIMT is a strong predictor of CVD risk, including myocardial infarction and stroke (Willeit et al., 2020). Studies from our group and others have shown an association between increased CIMT and chronic exposure to both HAP and ambient air pollution (Ofori et al., 2018; Ranzani et al., 2019; Painschab et al., 2013; Provost et al., 2015). Several studies have observed higher CIMT among women who self-reported HAP exposure from wood-burning cookstoves in Nigeria, South India, and Peru (Ofori et al., 2018; Ranzani et al., 2019; Painschab et al., 2013). For ambient air pollution, associations have been documented primarily at low-to-medium pollution levels in high-income countries. For example, exposure-response analyses have demonstrated that exposure to ambient and traffic-related PM2.5 is positively associated with CIMT, although exposure measurements were ecological, based on local area sampling, or on modeled exposures (Ranzani et al., 2019; Liu et al., 2015; Sommar et al., 2022).

While an association between HAP and CIMT has been previously demonstrated, the specific exposure-response relationship measured by personal HAP monitors and CVD biomarkers, particularly CIMT and atherosclerotic plaque, has not been well-characterized. The current study aims to address this gap by measuring the association of CIMT and plaque presence with 24-h personal exposures to PM2.5, BC and CO from 397 women aged 40–79 years in Guatemala, India, Peru, and Rwanda participating in the Household Air Pollution Intervention Network (HAPIN) trial. The hypothesis of the study is that higher personal exposures will be associated with thicker CIMT and carotid artery atherosclerotic plaques.

2. Methods

2.1. Study design

The current analysis uses baseline data collected during the HAPIN trial, a multi-country randomized controlled trial (RCT) of a clean cooking intervention in rural communities in Guatemala, India, Peru, and Rwanda. HAPIN enrolled 3200 households (800 per country) with pregnant women (18–34 years old at 9–19 weeks gestation) who exclusively or primarily used biomass fuels for cooking. An additional 418 older adult women (40–79 years old) residing in the same household were also enrolled and constitute the population for the current analysis. Households were randomized 1:1 to receive either the HAPIN intervention package—including an LPG stove, continuous fuel supply, and behavioral messaging to promote exclusive LPG use—or to continue using biomass fuels. Prior to randomization and intervention delivery, baseline data was collected via sociodemographic and health history surveys, 24-h personal exposure monitoring for PM2.5, BC, and CO was collected, and clinical health assessments were conducted, including carotid artery ultrasound imaging for CIMT and plaque identification. Full trial details and protocol has been described previously (Clasen et al., 2020).

The study protocol was approved by all participating institutional review boards (IRBs) and/or ethics committees from Emory University (00089799), Johns Hopkins University (00007403), Washington University in St. Louis (WashU) (201611159), Sri Ramachandra Institute of Higher Education and Research (IEC-N1/16/JUL/54/49), the Indian Council of Medical Research—Health Ministry Screening Committee (5/8/4–30/(Env)/Indo-US/2016-NCD-I), Universidad del Valle de Guatemala (146–08-2016), Guatemalan Ministry of Health National Ethics Committee (11–2016), Asociación Benefica PRISMA (CE2981.17), the London School of Hygiene and Tropical Medicine (11664–3), and the Rwandan National Ethics Committee (No.016/RNEC/2018).

2.2. Carotid artery ultrasound methods

Carotid artery ultrasound imaging was performed by certified sonographers using a portable ultrasound system (SonoSite Edge, Fujifilm-SonoSite Inc, Bothell, WA, USA) with a 7–15 MHz linear transducer. Ultrasound system settings, transducer placement, and other details have been described previously (Dávila-Román et al., 2021). Briefly, ECG-gated video and still frame images of the posterior wall of the distal 1 cm of the right and left carotid arteries of the common carotid artery (just proximal to the carotid artery bulb) were obtained as previously described (Stein et al., 2009). After identification of the 1-cm region of interest, the semi-automated edge detection software algorithm provided 100 separate dimensional measurements. From these 100 measurements, CIMT values were derived: mean CIMT (average thickness of the artery intima-media layer across the 100 measurements) and maximal CIMT (highest thickness value observed in the segment). Atherosclerotic plaques were defined as a focal intima-media thickness >1.5 mm or focal wall thickening that protruded into the lumen >0.5 mm or >50 % of the surrounding common carotid, bulb and internal carotid segments (Stein et al., 2009; Lin et al., 2018). Complete studies included approximately 8 images. All images were obtained by trained sonographers (see below) with acquisition taking approximately 20–30 min. Case report forms were completed online by site sonographers and linked to each ultrasound study. Site sonographers (three per site) underwent a 2-week in-person standardized training workshop at WashU.

Completed studies were uploaded to a cloud-based image server (Trice Imaging Inc., Del Mar, CA, USA) by site sonographers for subsequent offline analysis at the Ultrasound Core Laboratory (UCL) at WashU. For CIMT, automatic edge detected values for the mean and maximal CIMT are presented as the average of three each of the right and left common carotid arteries. Uploaded studies were reviewed by UCL sonographers blinded to study site and subject identity to assess image quality and adequacy of measurements (Dávila-Román et al., 2021).

2.3. Household air pollution personal exposure assessment

The HAPIN exposure assessment protocol has been described in detail previously (Johnson et al., 2020). Prior to randomization and intervention delivery, participants wore lightweight monitors for 24-h to measure their baseline personal exposure to PM2.5, BC, and CO. Participants carried the monitors in custom-designed aprons or vests that were developed in collaboration with participants to ensure appropriateness and comfort. During activities such as bathing and sleeping, participants were instructed to hang the aprons in a safe location nearby. Although limited to a single 24-hr period, this approach was designed to capture total indoor and outdoor exposure during participants’ routine daily activities. This can provide a more precise individual-level estimate than modeled or measured area-based measurements that are prone to exposure misclassification.

PM2.5 and BC exposures were measured using the Enhanced Children’s MicroPEM (ECM), a validated, lightweight particulate sampler (RTI Inc, Research Triangle Park, NC, USA). The ECM uses a built-in internal 0.3 L/min pump to draw air through an impactor attached to a cassette containing 15-mm Teflon® filters (PT15-AN-PF02; MTL Corporation, Minneapolis, MN), onto which particulates deposit. Gravimetric measurements of PM2.5 mass concentration were based on the pre- and post-weights of the filters, adjusted by sampling time and flow rate. Real-time, continuous PM2.5 exposure was collected by an internal light-scattering sensor. The ECM also monitored temperature, relative humidity, and filter-pressure drop. When a gravimetric sample was deemed invalid, due to missing or damaged filters or flow faults, nephelometer data were used to estimate PM2.5 concentrations. BC was estimated post-sampling on the ECM filters using the SootScan Model OT-21 Optical Transmissometer (Magee Scientific, Berkeley, CA, USA). Real-time, continuous monitoring of CO was conducted using the Lascar EL-USB-300 monitor (Lascar Electronics, Erie, PA), a lightweight, data-logging monitor. The Lascar instrument runs on batteries, has a sensing range between 0 and 300 ppm, and uses an electrochemical cell to measure CO.

2.4. Anthropometric measurements and surveys

Body Mass Index (BMI) was calculated for each participant, using the formula of weight in kilograms divided by the height in meters squared (kg/m2). Blood pressure (BP) was measured in triplicate on the right arm using an automatic digital BP monitor (OMRON HEM907XL) as detailed in the study protocol (Clasen et al., 2020). Surveys were administered to participants by trained field staff proficient in the local language to determine demographic and socioeconomic status (SES), health history, household characteristics, exposure to environmental tobacco, and other characteristics. SES was determined using an index constructed through principal components analysis (PCA) based on household assets, water and sanitation quality, food insecurity, education level, and housing materials, as described in detail previously (Nicolaou et al., 2022). To help ensure the validity and reliability of the collected data, all questionnaires were piloted prior to implementation.

2.5. Statistical methods

Single- and multi-variable linear regression models were run to test the association between personal exposures (PM2.5, BC or CO) and mean CIMT, maximal CIMT, and atherosclerotic plaque. Covariates were selected a priori using a directed acyclic graph (DAG) to identify a minimally sufficient set of factors for adjustment, to address potential confounders, and to ensure chosen covariates did not function as colliders (Fig. S1). The final models included the following covariates: study site (Guatemala, India, Peru, Rwanda), SES index (continuous: higher is worse), secondhand smoke (categorical: yes or no), and age (continuous). Separate models for PM2.5, BC, and CO were run due to moderate-to-high correlations between pollutants (PM2.5 and BC: Spearman’s ρ = 0.49; PM2.5 and CO: Spearman’s ρ = 0.52; CO and BC: Spearman’s ρ = 0.53) (Fig. S2).

Linear and log-linear models were tested, where pollutant exposures (PM2.5, BC, CO) were natural log-transformed (Fig. S3). To assess potential non-linear exposure-response relationships, exposures were categorized into quartiles (Fig. S4). Non-linear associations were further examined using generalized additive models (GAMs) with smoothing splines and tensor-product smooths (Fig. S5). Model performance and fit was evaluated using visualizations and Akaike Information Criterion (AIC), where lower AIC values indicate a better fit (Figs. S3–S5).

3. Results

3.1. Baseline characteristics of the study population

Of the 418 older women enrolled in HAPIN, 21 were missing baseline CIMT measurements and excluded from the current analysis. Accordingly, the study population consisted of 397 older adult women with CIMT measurements across the four study sites (Table 1). The mean age (52 ± 8 years) was similar across countries. BMI was highest in Peru and lowest in India. Second-hand smoke exposure (self-reported) was highest in India (28 %) followed by Rwanda (7 %) and Guatemala (5 %), with no secondhand smoking reported in Peru. Only 12 % and <1 % of participants in Rwanda and Peru, respectively, reported smoking previously; no participants reported smoking currently. Reported clinical history of high blood pressure (5.6 %), high cholesterol (1.2 %), and other heart disease (1.9 %) was low overall. More than half (57 %) were employed outside the household. The SES index was highest (indicating worse off) in Rwanda, followed by Guatemala, Peru, and India.

Table 1.

Baseline characteristics of all study participants who had CIMT measurements, by study site.

Variable Overall Guatemala India Peru Rwanda
N 397 133 93 128 43
Age (mean yrs ± SD) 52 ± 8 54 ± 8 49 ± 7 52 ±
8		 53 ± 9
Household size (mean ± SD) 6 ± 2.6 8 ± 2.9 4 ±
1.4		 6 ± 2 6 ± 2
SES Index (±SD) (higher is worse) 0.2 ± 1 −0.2 ± 0.8 1 ± 0.7 1 ± 0.5 −1.2 ± 0.7
BMI (mean kg/m2 ± SD) 25 ± 5 26 ± 4 21 ± 3 29 ± 5 23 ± 4
SBP (mean mmHg ± SD) 116 ± 18 120 ± 20 122 ± 14 108 ± 12 119 ± 17
DBP (mean mmHg ± SD) 70 ± 11 71 ± 11 77 ± 10 64 ± 9 73 ± 9
Self-report health history:
Hypertension (%) 6 10 0 5 9
Diabetes (%) 3 4 4 2 2
High Cholesterol (%) 1 2 0 2 0
Heart Disease (%) 2 3 0 1 5
Smoking History (%) 2 0 0 1 12
Secondhand Smoke – Current (%) 9 5 28 0 7
Hypertension medication: current (%) 2 2 0 2 7
Open in a new tab

3.2. Distribution of personal HAP exposures

Twenty-four-hour personal exposure measurements were available as follows: PM2.5: 91 % (362 participants); BC: 79 % (313 participants) and CO: 83 % (329 participants) (Table 2; Fig. S6). Across sites, the median 24-h personal exposure to PM2.5 was 87 μg/m3 (range: 60 μg/m3 in Peru to 110.3 μg/m3 in Guatemala). Notably, 81 % of participants experienced personal PM2.5 exposures exceeding the World Health Organization annual interim target-1 guideline of 35 μg/m3. Across sites, overall median 24-h exposure to BC was 11 μg/m3 (range: 10 μg/m3 in Peru to 12 μg/m3 in Guatemala), and 24-h median CO exposure was 1.4 ppb (range: 0.6 ppb in Rwanda to 2.3 ppb in Peru).

Table 2.

Personal 24-Hour exposures to PM2.5 (μg/m3), BC (μg/m3), and CO (ppm), by study site, among those with CIMT measurements.

All Guatemala India Peru Rwanda
PM2.5 (μg/m 3 ) N = 362 N = 129 N = 82 N = 110 N = 41
Mean ± SD 119 ± 121.2 144 ± 127.9 114 (127.8) 106 (119.9) 93 (69)
Median [IQR] 87 [44, 139] 110 [71, 180] 74 [44, 125] 60 [29, 123] 75 [49, 118]
Range 10, 803.4 14, 803 10, 760 13, 698 15, 312
BC (μg/m 3 ) N = 313 N = 115 N = 79 N = 91 N = 28
Mean ± SD 13 ± 9.8 13 ± 6.3 14 ± 10.9 13 ± 13 10 ± 5
Median [IQR] 11 [7, 16] 12 [9, 16] 11 [7, 19] 10 [3, 16] 10 [7, 13]
Range 1.1, 72 1.1, 47 1.6, 69 1.3, 72 2.8, 20
CO (ppb) N = 329 N = 121 N = 79 N = 91 N = 38
Mean ± SD 2.3 ± 3.5 1.8 ± 1.8 1.6 ± 3 3.9 ± 5.2 1.3 ± 1.8
Median [IQR] 1.4 [0.5, 3] 1.4 [0.6, 2.6] 0.7 [0.2, 2.1] 2.3 [1, 5.3] 0.6 [0.3, 1.6]
Range 0.001, 39 0.001, 9 0.001, 21 0.00067, 39 0.0317, 9
Open in a new tab
*

Population restricted to participants with valid exposure and CIMT measurements.

3.3. Distribution of CIMT and carotid artery atherosclerotic plaques

The average mean and maximal CIMT were 0.64 ± 0.13 mm and 0.75 ± 0.14 mm, respectively (Table 3). Both mean and maximal CIMT were highest in Rwanda, followed by Guatemala, India, and Peru. The overall prevalence of atherosclerotic plaques was 7.1 %, with the highest prevalence in Guatemala (11.6 %) and lowest in Peru (<1 %). Overall, mean and maximal CIMT were positively correlated with age (Spearman’s ρ = 0.45, p < 0.001 and ρ = 0.42, p < 0.001, respectively) and SBP (ρ = 0.44, p < 0.001 and ρ = 0.46, p < 0.001, respectively) (Fig. S7).

Table 3.

Mean and maximum CIMT and prevalence of carotid artery atherosclerotic plaque, by study site.

Overall
N = 397		 Guatemala
N = 133		 India
N = 93		 Peru
N = 128		 Rwanda N = 43
Mean CIMT (mm)
Mean ± SD 0.64 ± 0.13 0.69 ± 0.12 0.61 ± 0.13 0.59 ± 0.10 0.72 ± 0.13
Range [0.4, 1.38] [0.46, 1.04] [0.42, 1.38] [0.4, 0.99] [0.47, 1.01]
Maximum CIMT (mm)
Mean ± SD 0.75 ± 0.14 0.79 ± 0.14 0.73 ± 0.17 0.71 ± 0.11 0.83 ± 0.14
Range [0.49, 1.7] [0.52, 1.23] [0.51, 1.7] [0.49, 1.11] [0.56, 1.12]
Carotid artery atherosclerotic plaques
N = 393 N = 129 N = 93 N = 128 N = 43
Prevalence (%) 7.1 % 12 % 9.7 % 0.8 % 6.7 %
Open in a new tab

3.4. Associations between personal exposure to HAP pollutants and CIMT and carotid artery atherosclerotic plaques

Statistical models assessing the relationship between 24-h personal exposures and CIMT or atherosclerotic plaque were restricted to participants who had both relevant exposure and outcome data and included adjustments for the a priori covariates country, age, secondhand smoke exposure, and SES. In adjusted linear models, each 10 μg/m3 increase in PM2.5 was associated with a 0.001 mm increase in mean CIMT (p = 0.008; 95 % CI: 0 to 0.002) and a 0.002 mm increase in maximal CIMT (p = 0.005; 95 % CI: 0 to 0.003) (Table 4). For CO, each 10 ppm increase was associated with a 0.04 mm increase in maximal CIMT (p = 0.48; 95 % CI: 0 to 0.08). Additionally, on average, those in the fourth quartile of CO exposure had a 0.036 mm greater mean CIMT (95 % CI: 0.002, 0.071; p = 0.04) and 0.05 mm greater maximal CIMT (p = 0.02; 95 % CI: 0.01 to 0.09) compared to those in the first quartile. The association between CO and mean CIMT was positive but did not reach statistical significance at the α = 0.05 level. No significant exposure-response relationships were observed between black carbon and mean or maximal CIMT (Table 4) or between atherosclerotic plaques and any pollutant exposures (Table S1).

Table 4.

Associations between personal 24-h exposure to PM2.5 (μg/m3), BC (μg/m3), and CO (ppm) and mean and maximum CIMT (mm) across linear, log-linear, and quartile-based models.

Pollutant Model Type Estimate 95 % CI p-value AIC
Mean CIMT PM2.5 Linear 0.001 (0, 0.002) 0.008 * −595.2
Log linear 0.026 (−0.002, 0.057) 0.075 −591.2
Quartile 2 −0.005 (−0.036, 0.027) 0.771 −587.9
Quartile 3 0.002 (−0.031, 0.034) 0.928
Quartile 4 0.024 (−0.008, 0.056) 0.144
BC Linear 0.003 (−0.01, 0.02) 0.668 −497.0
Log linear 0.002 (−0.034, 0.038) 0.939 −496.8
Quartile 2 0.012 (−0.024, 0.049) 0.499 −494.9
Quartile 3 −0.008 (−0.044, 0.028) 0.667
Quartile 4 −0.009 (−0.044, 0.026) 0.612
CO Linear 0.03 (−0.001, 0.06) 0.140 −535.3
Log linear 0.016 (−0.002, 0.033) 0.074 −536.3
Quartile 2 0.011 (−0.022, 0.044) 0.501 −536.5
Quartile 3 −0.006 (−0.040, 0.028) 0.738
Quartile 4 0.036 (0.002, 0.071) 0.041 *
Max CIMT PM2.5 Linear 0.002 (0, 0.003) 0.005 * −477.4
Log linear 0.030 (−0.005, 0.065) 0.081 −472.6
Quartile 2 −0.008 (−0.045, 0.029) 0.674 −469.3
Quartile 3 −0.001 (−0.039, 0.037) 0.949
Quartile 4 0.026 (−0.012, 0.063) 0.181
BC Linear 0.003 (−0.01, 0.02) 0.690 −389.7
Log linear −0.002 (−0.044, 0.041) 0.958 −389.6
Quartile 2 0.007 (−0.036, 0.050) 0.746 −346.3
Quartile 3 −0.014 (−0.057, 0.029) 0.522
Quartile 4 −0.016 (−0.058, 0.026) 0.448
CO Linear 0.040 (0, 0.08) 0.048 * −420.1
Log linear 0.023 (0.002, 0.044) 0.025 * −421.2
Quartile 2 0.017 (−0.022, 0.056) 0.395 −420.2
Quartile 3 0.001 (−0.040, 0.042) 0.964
Quartile 4 0.05 (0.009, 0.092) 0.018 *
Open in a new tab
*

Estimates for linear and log-linear models represent the change in mean CIMT (mm) per 10-unit increase in pollutant exposure (e.g., 10 μg/m3 for PM2.5 and BC, 10 ppm for CO). For quartile-based models, estimates represent the difference in mean CIMT (mm) for participants in quartiles 2, 3, and 4 of exposure compared to quartile 1 (reference group). Significant associations (p < 0.05) are bolded and indicated with an asterisk ().

4. Discussion

In this cross-sectional study of older adult women living in households that primarily used biomass fuel for cooking in four LMICs, 24-h personal exposures to PM2.5, BC, and CO and carotid artery ultrasound (mean CIMT, maximal CIMT and atherosclerotic plaques) were measured. Higher personal exposure to PM2.5 was associated with greater mean and maximal CIMT, and higher CO exposure was associated with maximal CIMT. Additionally, women in the highest quartile of CO exposure had significantly greater maximal CIMT compared to those in the lowest quartile. To our knowledge, this is the first multi-country study to measure associations between directly measured personal exposures to PM2.5, BC and CO among women chronically exposed to household biomass smoke and CIMT, a well-validated biomarker of atherosclerosis and CVD (Ruijter et al., 2012; Lorenz et al., 2007, 2012; van den Oord, 2013; O’Leary et al., 1999). These findings address a critical gap in understanding the potential impacts of HAP exposure, which is widespread globally and particularly in LMICs, on the development of CVD, the most common cause of death and disability worldwide.

The current study contributes to the limited existing literature on HAP and atherosclerosis. Previous studies have primarily relied on indirect proxies of HAP exposure, such as self-reported biomass stove use or geographic location. For example, an earlier study in Peru from our group reported that rural participants (who primarily used biomass fuels) exhibited significantly thicker mean and maximal CIMT and had a higher prevalence of atherosclerotic plaque compared to urban women (Painschab et al., 2013). Similarly, a study of rural dwelling women in southern Nigeria found that biomass fuel users had increased CIMT compared to clean fuel users (Ofori et al., 2018). Another study in peri-urban southern India found a positive but non-significant association between predicted personal PM2.5 exposure and CIMT (Ranzani et al., 2020). Notably, the average PM2.5 exposure (61 μg/m3, IQR = 12.9) was approximately half of that observed in the present study. Furthermore, predicted personal exposures derived from modeling a subset of measurements were used in the Ranzani et al. study, rather than direct personal monitoring. Differences in self-reported biomass use (60 % in their population) may further explain differences in findings. Additionally, a study of adults (40–79 years old) from China reported a borderline significant positive association between directly measured personal PM2.5 exposure and CIMT. Together, these studies suggest a probable association between HAP exposure and vascular thickening; however, most relied on proxy measures of HAP exposure. The present study strengthens this evidence base by using direct 24-h personal exposure measurements across diverse LMIC settings and demonstrating significant positive associations between CIMT and both PM2.5 and CO, suggesting a potential pollutant-specific relationship.

Our finding that PM2.5 is positively associated with CIMT is biologically plausible as PM2.5 can induce systemic inflammation and oxidative stress, two key pathways in the development of the vascular thickening (measured by increased CIMT) that precedes atherosclerosis. Our findings also generally align with epidemiological evidence from higher-income settings with lower exposure levels. For example, a 2015 meta-analysis of pooled data from 9 studies conducted in Europe and North America reported that each 5 μg/m3 increase in ambient PM2.5 was associated with a 12.1 μm increase in CIMT (Provost et al., 2015). Subsequent studies have generally supported a positive association between PM2.5 and CIMT, though with some heterogeneity across studies (Jilani et al., 2020; Kim et al., 2024). While our results are broadly consistent with this literature, direct comparisons should be made with caution. The studies from high-income countries typically relied on measured or modeled ambient exposures at much lower concentrations, whereas our study measured total 24-h personal exposure—including both indoor and outdoor sources—at substantially higher levels in rural and peri-urban LMIC settings. The smaller effect sizes observed in our study may reflect attenuation at higher exposures, non-linear exposure-response relationships, or physiological adaptation to chronic biomass smoke. Nevertheless, findings from these ambient and personal exposure studies are complementary and together contribute to a more complete understanding of the vascular impacts of air pollution.

The present study found an association between CO exposure and maximal CIMT, particularly among those participants in the highest exposure quartile. While the acute effects of CO poisoning and cardiopulmonary disease are well established (e.g., myocardial ischemia, arrhythmias, pulmonary edema, cardiac dysfunction), the effects of chronic CO exposure on cardiopulmonary disease and/or development of atherosclerosis are less well understood. Epidemiologic studies have shown associations between short-term exposure to ambient CO with CVD morbidity and mortality (Chen et al., 2007, 2011; Bell et al., 2009; Tian et al., 2015). A recent study of 272 cities in China reported strong evidence of an association between short-term exposure to ambient CO and increased CVD mortality, and particularly coronary heart disease mortality (Chen et al., 2011). CO contributes to vascular damage through mechanisms like hypoxia and oxidative stress, leading to endothelial dysfunction and atherosclerosis. A case-control study in Turkey among 40 barbecue workers with high CO exposure vs. 48 controls reported an association between higher carboxyhemoglobin levels, a biomarker of CO exposure, and increased CIMT. In the current study, a significant association between CO exposure and maximal CIMT (but not mean CIMT) was found. While mean CIMT is more widely used, maximal CIMT is considered a more sensitive biomarker of atherosclerosis progression as it captures focal thickening of the arterial wall, a precursor to atherosclerotic plaque development (Bots et al., 2003; O’Brien et al., 2024). Together, results from prior studies and our findings suggest a potential role of chronic CO exposure in CVD development.

This is the first study to measure associations between CIMT and directly measured personal exposure to BC. While a small positive association between BC and CIMT was found, it did not reach statistical significance. In contrast, Wilker et al. showed a significant association between long-term BC exposure and increased CIMT in a senior population in Massachusetts (Wilker et al., 2013). Similarly, Kim et al. found a positive association between BC and CIMT in middle-aged adults from four US cities, after adjustment for demographic, behavioral, socioeconomic, and comorbidity factors, and in subgroup analyses by race and sex (Kim et al., 2024). BC can stimulate systemic inflammation and oxidative stress pathways implicated in the development of cardiovascular disease (Kim et al., 2024; Alexeeff et al., 2011; Jiang et al., 2020). The lack of a significant association between BC and CIMT in the present study may be explained by differences in exposure assessment. Unlike previous studies, which relied on ambient air pollution or modeled BC exposures from traffic-related pollution in urban US settings, the present study used personal exposure measurements in populations primarily exposed to HAP from biomass combustion. The composition and source of PM2.5, including BC, may differ between traffic-related and biomass-related pollution. It is possible that other components of PM2.5, rather than BC, are the primary drivers of the observed associations in the present study.

Finally, a study of 606 asymptomatic low cardiovascular risk adults from Australia found that PM2.5 ambient air pollution was associated with the degree of coronary artery calcification, another imaging biomarker of atherosclerosis, independent of other risk factors (Huynh et al., 2020). Collectively, evidence from prior studies and ours using imaging biomarkers of vascular remodeling and atherosclerosis indicate that air pollution may play a role in residual CVD risk beyond traditional risk factors.

The present cross-sectional study did not find a significant association between 24-h personal exposure to PM2.5, BC, or CO and atherosclerotic plaques. Few studies in high exposure settings exist for comparison and results are mixed. Similar to our study, Kanagasabai et al., reported positive, albeit insignificant associations between total area of atherosclerotic plaques and self-reported solid fuel use and PM2.5 exposure (Kanagasabai et al., 2021). In contrast, in Peru, our group reported that greater self-reported use of biomass fuels for cooking was associated with a higher prevalence of atherosclerotic plaques (Painschab et al., 2013). Differences in overall atherosclerotic plaque prevalence (26 % in Painschab et al. vs. 7 % in the current study) and population characteristics (inclusion of both sexes vs. only women in the current study) may also explain the discrepant findings. In a review of high-income, low-exposure settings, Julani et al. reported mostly positive but some weak associations between PM2.5 and atherosclerotic plaques (Jilani et al., 2020). Conversely, a cross-sectional study in a Swedish cohort did not find an association between annual mean PM2.5 and atherosclerotic plaques (Hasslöf et al., 2020).

4.1. Limitations

The present study is cross-sectional, which limits causal inference between personal exposures and CIMT outcomes. Additionally, baseline personal exposure assessment occurred over a single 24-h period, which may not capture intra-individual variability related to seasonal or weekly routines or reflect long-term exposures patterns. However, the advantage of this approach is that it captures total exposure across all indoor and outdoor environments during the limited sampling period, thereby providing a more precise individual-level estimate than modeled or measured area-based samples that are prone to exposure misclassification. Additionally, since all participants were still exclusive or primary biomass users at the time of baseline measurement, and stove and fuel type are strong drivers of exposure, we expect historical fuel use and exposure levels to have been relatively stable and the one-time measurement to be a reasonable estimate of habitual exposure. Additional limitations include reliance on participant adherence to wearing the monitors during the 24-h exposure assessment period. To mitigate this, we piloted the monitoring protocol using context-appropriate vests and provided training to ensure comfort and compliance. Furthermore, the participant population was concentrated at the higher end of the exposure distribution, which may have limited our ability to detect associations at lower exposure levels. Finally, observed differences in exposure levels and CIMT across study sites may reflect regional clustering of fuel types, cooking practices, and underlying cardiovascular risk profiles. Despite adjustment for study site in our models, unmeasured heterogeneity of these individual factors may have reduced our ability to detect additional associations.

4.2. Conclusions

In this cross-sectional study of adult women living in households that primarily cook with biomass fuels, higher personal exposures to PM2.5 and CO were associated with greater mean and maximal CIMT, a well-established biomarker of atherosclerosis, further supporting the association between HAP and cardiovascular disease.

Supplementary Material

Underhill_IJHEH_2025_SI

Acknowledgments

The investigators would like to thank the members of the advisory committee – Drs. Patrick Breysse, Donna Spiegelman, and Joel Kaufman - for their valuable insight and guidance throughout the implementation of the trial. We also wish to acknowledge all research staff and study participants for their dedication to and participation in this important trial.

A multidisciplinary, independent Data and Safety Monitoring Board (DSMB) appointed by the National Heart, Lung, and Blood Institute (NHLBI) monitored the quality of the data and protected the safety of patients enrolled in the HAPIN trial. The DSMB consisted of: Catherine Karr (Chair), Nancy R. Cook, Stephen Hecht, Joseph Millum, Nalini Sathiakumar (deceased), Paul K. Whelton, and Gail Weinmann and Thomas Croxton (Executive Secretaries). Program Coordination: Gail Rodgers, Bill & Melinda Gates Foundation; Claudia L. Thompson, National Institute of Environmental Health Sciences; Mark J. Parascandola, National Cancer Institute; Marion Koso-Thomas, Eunice Kennedy Shriver National Institute of Child Health and Human Development; Joshua P. Rosenthal, Fogarty International Center; Concepcion R. Nierras, NIH Office of Strategic Coordination – The Common Fund; Katherine Kavounis, Dong-Yun Kim, Barry S. Schmetter (deceased), and Antonello Punturieri, NHLBI.

The following are the HAPIN Investigators (in addition to all authors): Gloriose Bankundiye, Dana Boyd Barr, Vanessa Burrowes, Alejandra Bussalleu, Devan Campbell, Eduardo Canuz, Adly Castañaza, Howard H. Chang, Yunyun Chen, Marilú Chiang, Carmen Lucia Contreras, Rachel Craik, Oscar De León, Priya D’Souza, Ephrem Dusabimana, Lisa Elon, Juan Gabriel Espinoza, Irma Sayury Pineda Fuentes, Sarada S. Garg, Ahana Ghosh, Dina Goodman-Palmer, Savannah Gupton, Sarah Hamid, Steven A. Harvey, Mayari Hengstermann, Ian Hennessee, Phabiola Herrera, Marjorie Howard, Penelope P. Howards, Shirin Jabbarzadeh, Katherine Kearns, Miles A. Kirby, Jacob Kremer, Margaret A. Laws, Grace Lee, Pattie Lenzen, Jiawen Liao, Amy E. Lovvorn, Jane Mbabazi, Eric D. McCollum, John P. McCracken, Julia N. McPeek, Rachel Meyers, J. Jaime Miranda, Erick Mollinedo, Libny Monroy, Alexie Mukeshimana, Krishnendu Mukhopadhyay, Moses Mutabazi, Bernard Mutariyani, Luke P. Naeher, Durairaj Natesan, Azhar Nizam, Jean de Dieu Ntivuguruzwa, Parinya Panuwet, Ricardo Piedrahita, Naveen Puttaswamy, Elisa Puzzolo, Karthikeyan Dharmapuri Rajamani, Sarah Rajkumar, Usha Ramakrishnan, Rengaraj Ramasami, Alexander Ramirez, Joshua Rosenthal, P. Barry Ryan, Sudhakar Saidam, Sankar Sambandam, Sheela S. Sinharoy, Kirk R. Smith, Kyle Steenland, Damien Swearing, Gurusamy Thangavel, Lisa M. Thompson, Ashley Toenjes, Viviane Valdes, Amit Verma, Jiantong Wang, Megan Warnock, Bonnie N. Young, Ashley Younger.

Funding

The HAPIN trial was funded by the U.S. National Institutes of Health (cooperative agreement 1UM1HL134590) in collaboration with the Bill & Melinda Gates Foundation (OPP1131279).

Role of the funding source

The funders of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.ijheh.2025.114649.

Footnotes

CRediT authorship contribution statement

Lindsay J. Underhill: Writing – review & editing, Writing – original draft, Visualization, Validation, Software, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Lisa de las Fuentes: Writing – review & editing, Writing – original draft, Supervision, Methodology, Formal analysis, Conceptualization. Laura Nicolaou: Writing – review & editing, Investigation. Shakir Hossen: Writing – review & editing, Methodology, Investigation, Formal analysis. Anaite Diaz-Artiga: Writing – review & editing, Supervision, Resources, Project administration, Methodology, Investigation. Ajay Pillarisetti: Writing – review & editing, Validation, Methodology, Formal analysis. Aris T. Papageorghiou: Writing – review & editing, Methodology, Investigation, Conceptualization. Florien Ndagijimana: Writing – review & editing, Methodology, Investigation. Ghislaine Rosa: Writing – review & editing, Methodology, Investigation, Funding acquisition, Conceptualization. Gurusamy Thangavel: Writing – review & editing, Project administration, Methodology, Investigation. John P. McCracken: Writing – review & editing, Methodology, Investigation. Kalpana Balakrishnan: Writing – review & editing, Project administration, Methodology, Investigation, Conceptualization. Krishnendu Mukhopadhyay: Writing – review & editing, Investigation. Kyle Steenland: Writing – review & editing, Methodology, Investigation, Conceptualization. Lisa M. Thompson: Writing – review & editing, Project administration, Methodology, Investigation. Lance A. Waller: Writing – review & editing, Methodology, Investigation, Data curation, Conceptualization. Maggie L. Clark: Writing – review & editing, Writing – original draft, Validation, Supervision, Methodology, Data curation. Michael A. Johnson: Writing – review & editing, Project administration, Methodology, Investigation, Conceptualization. Sarada Garg: Writing – review & editing, Project administration, Methodology, Investigation. Sankar Sambandam: Writing – review & editing, Project administration, Methodology, Investigation. Suzanne M. Simkovich: Writing – review & editing, Methodology, Investigation. Vigneswari Aravindalochanan: Writing – review & editing, Investigation. Kendra N. Williams: Writing – review & editing, Methodology, Investigation. Wenlu Ye: Writing – review & editing, Investigation. Jennifer L. Peel: Writing – review & editing, Methodology, Investigation, Funding acquisition, Conceptualization. Thomas F. Clasen: Writing – review & editing, Writing – original draft, Investigation, Funding acquisition, Conceptualization. William Checkley: Writing – review & editing, Methodology, Investigation, Funding acquisition, Conceptualization. Victor G. Davila-Roman: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Conceptualization.

Declaration of interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data sharing

Data will be available upon request. To protect the confidentiality of study participants, identifiable information will be excluded from all datasets that will be made available. These will be accompanied by documentation necessary to understand the content (such as data dictionaries or metadata descriptions). All source datasets will be made available through the corresponding authors, subject to ethical, data protection, and other obligations being addressed. Data will also be published to a registry.

References

  1. Alexeeff SE, et al. , 2011. Medium-term exposure to traffic-related air pollution and markers of inflammation and endothelial function. Environ. Health Perspect. 119, 481–486. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bell ML, Peng RD, Dominici F, Samet JM, 2009. Emergency hospital admissions for cardiovascular diseases and ambient levels of carbon monoxide. Circulation 120, 949–955. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bots ML, Evans GW, Riley WA, Grobbee DE, 2003. Carotid intima-media thickness measurements in intervention studies. Stroke 34, 2985–2994. [DOI] [PubMed] [Google Scholar]
  4. Chen T-M, Kuschner WG, Gokhale J, Shofer S, 2007. Outdoor air pollution: nitrogen dioxide, sulfur dioxide, and carbon monoxide health effects. Am. J. Med. Sci. 333, 249–256. [DOI] [PubMed] [Google Scholar]
  5. Chen R, et al. , 2011. Ambient carbon monoxide and daily mortality in three Chinese cities: the China Air Pollution and Health Effects Study (CAPES). Sci. Total Environ. 409, 4923–4928. [DOI] [PubMed] [Google Scholar]
  6. Clasen T, et al. , 2020. Design and rationale of the HAPIN study: a multicountry randomized controlled trial to assess the effect of liquefied petroleum gas stove and continuous fuel distribution. Environ. Health Perspect. 128, 47008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Dávila-Román VG, et al. , 2021. Ultrasound Core laboratory for the household air pollution intervention Network trial: standardized training and image management for field studies using portable ultrasound in fetal, Lung, and vascular evaluations. Ultrasound Med. Biol. 47, 1506–1513. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Hasslöf H, et al. , 2020. Long-term exposure to air pollution and atherosclerosis in the carotid arteries in the Malmö diet and cancer cohort. Environ. Res. 191, 110095. [DOI] [PubMed] [Google Scholar]
  9. Huynh Q, et al. , 2020. Long-term exposure to ambient air pollution is associated with coronary artery calcification among asymptomatic adults. Eur. Heart J. Cardiovasc. Imaging 22, 922–929. [DOI] [PubMed] [Google Scholar]
  10. IHME. Air pollution. https://www.healthdata.org/research-analysis/health-risks-issues/air-pollution. [Google Scholar]
  11. Jiang N, et al. , 2020. Low-dose combined exposure of carboxylated black carbon and heavy metal lead induced potentiation of oxidative stress, DNA damage, inflammation, and apoptosis in BEAS-2B cells. Ecotoxicol. Environ. Saf. 206, 111388. [DOI] [PubMed] [Google Scholar]
  12. Jilani MH, et al. , 2020. Associations between particulate matter air pollution, presence and progression of subclinical coronary and carotid atherosclerosis: a systematic review. Atherosclerosis 306, 22–32. [DOI] [PubMed] [Google Scholar]
  13. Johnson MA, et al. , 2020. Air pollutant exposure and stove use assessment methods for the household air pollution intervention Network (HAPIN) trial. Environ. Health Perspect. 128, 047009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Kanagasabai T, et al. , 2021. Household air pollution and blood pressure, vascular damage, and subclinical indicators of cardiovascular disease in older Chinese adults. Am. J. Hypertens. 35, 121–131. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Kim K, et al. , 2024. Associations of exposure to PM2.5 and its compounds with carotid intima-media thickness among middle-aged adults. Sci. Total Environ. 955, 177098. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Libby P, et al. , 2019. Atherosclerosis. Nat. Rev. Dis. Prim. 5, 56. [DOI] [PubMed] [Google Scholar]
  17. Lin X, et al. , 2018. Endogenous cholesterol excretion is negatively associated with carotid intima–media thickness in humans. Arterioscler. Thromb. Vasc. Biol. 37, 2364–2369. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Liu X, et al. , 2015. Association of exposure to particular matter and carotid intima-media thickness: a systematic review and meta-analysis. Int. J. Environ. Res. Publ. Health 12, 12924–12940. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Lorenz MW, Markus HS, Bots ML, Rosvall M, Sitzer M, 2007. Prediction of clinical cardiovascular events with carotid intima-media thickness. Circulation 115, 459–467. [DOI] [PubMed] [Google Scholar]
  20. Lorenz MW, et al. , 2012. Carotid intima-media thickness progression to predict cardiovascular events in the general population (the PROG-IMT collaborative project): a meta-analysis of individual participant data. Lancet 379, 2053–2062. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Nicolaou L, et al. , 2022. Cross-sectional analysis of the association between personal exposure to household air pollution and blood pressure in adult women: evidence from the multi-country Household Air Pollution Intervention Network (HAPIN) trial. Environ. Res. 214, 114121. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Ofori SN, Fobil JN, Odia OJ, 2018. Household biomass fuel use, blood pressure and carotid intima media thickness; a cross sectional study of rural dwelling women in Southern Nigeria. Environ. Pollut. 242, 390–397. [DOI] [PubMed] [Google Scholar]
  23. O’Brien MW, Kimmerly DS, Theou O, 2024. Sex-specific frailty and chronological age normative carotid artery intima-media thickness values using the Canadian longitudinal study of aging. Vascular 32, 579–588. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. O’Leary DH, et al. , 1999. Carotid-artery intima and media thickness as a risk factor for myocardial infarction and stroke in older adults. N. Engl. J. Med. 340, 14–22. [DOI] [PubMed] [Google Scholar]
  25. Painschab MS, et al. , 2013. Chronic exposure to biomass fuel is associated with increased carotid artery intima-media thickness and a higher prevalence of atherosclerotic plaque. Heart 99, 984–991. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Provost EB, Madhloum N, Panis LI, Boever PD, Nawrot TS, 2015. Carotid intima-media thickness, a marker of subclinical atherosclerosis, and particulate air pollution exposure: the meta-analytical evidence. PLoS One 10, e0127014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Ranzani OT, et al. , 2019. Association between ambient and household air pollution with carotid intima-media thickness in peri-urban South India: CHAI-Project. Int. J. Epidemiol. 49, 69–79. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Ranzani OT, et al. , 2020. Personal exposure to particulate air pollution and vascular damage in peri-urban South India. Environ. Int. 139, 105734. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Ruijter HMD, et al. , 2012. Common carotid intima-media thickness measurements in cardiovascular risk prediction: a meta-analysis. JAMA 308, 796–803. [DOI] [PubMed] [Google Scholar]
  30. Sommar JN, et al. , 2022. Long-term exposure to particulate air pollution and presence and progression of carotid artery plaques - a northern Sweden VIPVIZA cohort study. Environ. Res. 211, 113061. [DOI] [PubMed] [Google Scholar]
  31. Stein JH, Korcarz CE, Post WS, 2009. Use of carotid ultrasound to identify subclinical vascular disease and evaluate cardiovascular disease risk: summary and discussion of the American society of echocardiography consensus statement. Prev. Cardiol. 12, 34–38. [DOI] [PubMed] [Google Scholar]
  32. Tian L, et al. , 2015. Carbon monoxide and stroke: a time series study of ambient air pollution and emergency hospitalizations. Int. J. Cardiol. 201, 4–9. [DOI] [PubMed] [Google Scholar]
  33. van den Oord SC, et al. , 2013. Carotid intima-media thickness for cardiovascular risk assessment: systematic review and meta-analysis. Atherosclerosis 228, 1–11. [DOI] [PubMed] [Google Scholar]
  34. Wilker EH, et al. , 2013. Long-term exposure to black carbon and carotid intima-media thickness: the normative aging study. Environ. Health Perspect. 121, 1061–1067. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Willeit P, et al. , 2020. Carotid intima-media thickness progression as surrogate marker for cardiovascular risk. Circulation 142, 621–642. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Underhill_IJHEH_2025_SI
NIHMS2165640-supplement-Underhill_IJHEH_2025_SI.docx (12.7MB, docx)

Data Availability Statement

Data will be available upon request. To protect the confidentiality of study participants, identifiable information will be excluded from all datasets that will be made available. These will be accompanied by documentation necessary to understand the content (such as data dictionaries or metadata descriptions). All source datasets will be made available through the corresponding authors, subject to ethical, data protection, and other obligations being addressed. Data will also be published to a registry.

RESOURCES