Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Feb 10.
Published in final edited form as: Environ Adv. 2022 Jun 29;9:100257. doi: 10.1016/j.envadv.2022.100257

In-utero personal exposure to PM2.5 impacted by indoor and outdoor sources and birthweight in the MADRES cohort

Karl O’Sharkey a, Yan Xu b, Thomas Chavez a, Mark Johnson a, Jane Cabison a, Marisela Rosales a, Brendan Grubbs a, Claudia M Toledo-Corral a,c, Shohreh F Farzan a, Theresa Bastain a, Carrie V Breton a, Rima Habre a,b,*
PMCID: PMC9912940  NIHMSID: NIHMS1853765  PMID: 36778968

Abstract

Background:

In-utero exposure to outdoor particulate matter with aerodynamic diameter less than 2.5 μm (PM2.5) is linked with low birthweight. However, previous results are mixed, likely due to measurement error introduced by estimating personal exposure from ambient data. This study investigated the effect of total personal PM2.5 exposure on birthweight and whether it differed when it was more heavily impacted by sources of indoor vs outdoor origin in the MADRES cohort study.

Methods:

Personal PM2.5 exposure was measured in 205 pregnant women in the 3rd trimester using 48 h integrated, filter-based sampling. Linear regression was used to test the association between personal PM2.5 exposure and birthweight, adjusting for key covariates. Interactions of PM2.5 with variables representing indoor sources of PM2.5, home ventilation, or time spent indoors tested whether the effect of total PM2.5 on birthweight varied when it was more impacted by sources of indoor vs outdoor origin.

Results:

In a sample of largely Hispanic (81%) pregnant women, total personal PM2.5 was not significantly associated with birthweight (β = 38.6 per 1SD increase in PM2.5; 95% CI:−21.1, 98.2). This association however, differed by home type (single family home: 156.9 (26.9, 287.0), 2–4 attached units:−16.6 (−111.9, 78.7), 5+ units:−62.6 (−184.9, 59.6), missing: 145.4 (−4.1, 294.9), interaction p = 0.028) and by household air conditioner use (none of the time: −27.6 (−101.5, 46.3) vs. some of the time: 139.9 (42.9, 237.0), interaction p = 0.008) Additionally, the effect of personal PM2.5 on birthweight varied by time spent indoors (none or little of the time: − 45.1 (−208.3, 118.1) vs. most or all of the time: 57.1 (−7.3, 121.6), interaction p = 0.255).

Conclusions:

While no significant association between total personal PM2.5 exposure and birthweight was found, there was evidence that multi-unit housing (vs. single-family homes), candle and/or incense smoke, and greater outdoor source contributions to personal PM2.5 were more strongly associated with lower birthweight.

Keywords: Air pollution, PM2.5, Birthweight, Prenatal exposure, Pregnancy, Personal monitoring

1. Introduction

In the United States (U.S.), an estimated 8.3% of newborns are born with low birthweight (LBW) (Martin et al., 2019); defined as below 2500 grams (g). The impact of LBW is far reaching, with research showing it is associated with infant mortality (Vilanova et al., 2019; Watkins et al., 2016) and later life obesity (Jornayvaz et al., 2016), type-2 diabetes (Mi et al., 2017), cardiovascular disease (Risnes et al., 2011; Smith et al., 2016; Umer et al., 2020), and impaired cognitive development (Upadhyay et al., 2019; Whitaker et al., 2006). Within the U.S., such health outcomes are often disproportionate with regard to race/ethnicity, with obesity and type-2 diabetes prevalence highest in Hispanic and Black populations across the lifetime (Petersen, 2019; Rossen, 2014).

In the past decade, several epidemiological studies have established a relationship between outdoor air pollution and birthweight and/or LBW (Lamichhane et al., 2015; Li et al., 2017; Pedersen et al., 2013). Within the US, these studies primarily focused on federally-regulated criteria air pollutants. One of those is particulate matter (PM) with an aerodynamic diameter less than 2.5 μm (PM2.5) (Huang et al., 2015; Rich et al., 2015; Schembari et al., 2015). In-utero PM2.5 exposure is hypothesized to create a hostile intrauterine environment likely resulting from oxidative stress, DNA methylation changes, mitochondrial DNA content alteration, and endocrine disruption (Clemente et al., 2016; Li et al., 2019). Such mechanistic alterations may lead to health risks in later life such as the development of visceral adiposity and altered glucose homeostasis (Barnes and Ozanne, 2011; Morrison et al., 2010; Visentin et al., 2014).

While several reviews have concluded a weak to moderate association between outdoor PM2.5 and several birth outcomes, including a decrease in birthweight and an increased risk of LBW (Li et al., 2017; Stieb et al., 2016; Sun et al., 2016), the literature remains inconsistent. Reductions in birthweight due to outdoor PM2.5 exposure also vary by race/ethnicity (Basu et al., 2014), possibly due to Hispanic and Black mothers experiencing the greatest burden of air pollution exposure (Bell and Ebisu, 2012; Mikati et al., 2018). Effect estimates also differ depending on the exposure window under study, with the 3rd trimester showing the most consistent evidence of greater risk of LBW (Dadvand et al., 2014; Schembari et al., 2015; Zhu et al., 2015). Additionally, most health studies to date estimated an individual’s exposure to outdoor PM2.5 at the residential level using models that typically incorporate ambient monitoring, remote sensing, and/or geospatial data (Ebisu et al., 2014; Gray et al., 2010; Harris et al., 2014). While these models are increasingly capable of capturing spatial variability in outdoor air pollution, they inherently suffer from exposure measurement error in terms of estimating personal exposure to air pollution of outdoor origin, which might bias effect estimates and attenuate power to detect health effects (Carroll, 2005; Kioumourtzoglou et al., 2014; Zeger et al., 2000). This is because individuals spend the majority of their time indoors, and their ‘true’ personal exposure to PM2.5 of outdoor origin is a result of the infiltration efficiency of PM2.5 indoors and time-activity patterns, most accurately captured by personal monitoring (Gray et al., 2011). Finally, there is currently very little research into the effect of total personal exposure to PM2.5 prenatally on birthweight. Total personal PM2.5 is impacted by multiple sources including personal activity, indoor sources, and outdoor sources (or PM2.5 of outdoor origin, which may only represent a small fraction of an individual’s total personal PM2.5 exposure) (Habre et al., 2014). Therefore, quantifying the influence of total personal PM2.5 on birthweight is also an important question that has not yet been thoroughly investigated.

Personal monitoring of air pollution is a sophisticated, yet often expensive and burdensome method of exposure assessment and as such, only a small number of studies have used it, the majority of which have focused on toxic polyaromatic hydrocarbons (PAHs) (Choi et al., 2012, 2008; Rundle et al., 2012). One study found an inverse association with personal PM2.5 and birthweight (Jedrychowski et al., 2009). However, very few studies have been conducted in a health disparities population with potentially greater exposure to PM2.5 of outdoor origin and greater vulnerability or susceptibility to its effects (Morello-Frosch et al., 2011), particularly in the 3rd trimester where most fetal weight gain occurs (Kiserud et al., 2018). Additionally, there is a pressing need to evaluate the effects of PM2.5 impacted by sources of indoor vs outdoor origin (hereinafter referred to as indoor vs outdoor sources for simplicity) due to the differences in their chemical composition and thus potential toxicity, and the fact that only ambient PM2.5 concentrations are regulated.

Therefore, the purpose of this present study was to bridge these gaps in knowledge by evaluating the role of 3rd trimester personal PM2.5 exposure on birth weight in a health disparities population in Los Angeles, CA. In addition, this study investigated whether the effect of total personal PM2.5 on birthweight was different when it was more impacted by indoor vs outdoor sources. To accomplish this, this study tested interactions with questionnaire-based variables that correlate directly with greater indoor (e.g., indoor burning of candles or incense) or outdoor (e.g., time spent outdoors) contributions to total personal PM2.5.

2. Material and methods

2.1. Study population

The Maternal and Developmental Risks from Environmental and Social Stressors (MADRES) study is an ongoing prospective cohort study of ~900 pregnant, primarily Hispanic, low-income mothers in Los Angeles County, motivated to investigate the cumulative impact of environmental pollutants and psychosocial, behavioral, and built environmental risk factors on maternal and infant health outcomes (Bastain et al., 2019). Pregnant women were enrolled via partnerships with four prenatal care providers in Los Angeles beginning November 2015, including one county hospital clinic, two non-profit community health clinics, and a private obstetrics and gynecology practice.

Participant eligibility included: (1) at least 18 years old, (2) fluency in either Spanish or English, and (3) less than 30-weeks gestation at recruitment. Exclusion criteria for the study included: (1) multiple gestation, (2) current incarceration, (3) HIV positive, and (4) having a physical, mental, or cognitive disability that would prevent the participant from providing informed consent.

The current analysis leverages data collected as part of a 214-participant personal PM2.5 exposure monitoring study nested within the MADRES cohort. Women were asked to wear a crossbody sampling purse with a personal monitoring apparatus for a 48 h monitoring period in the 3rd trimester. This subset was comparable to the larger MADRES cohort in terms of demographics, birth outcomes, and outdoor air pollution metrics.

2.2. Personal PM2.5 exposure monitoring

Total, 48 h integrated personal PM2.5 exposure was measured in the 3rd trimester using a custom sampling design on a subset of 214 women recruited from the larger cohort between October 2016 and February 2020. A trained, bilingual study staff member recruited participants during one of their 3rd trimester study visit at the University of Southern California (USC) clinic. Participants were provided with a personal sampling crossbody purse containing a Gilian Plus Datalogging Pump (Sensidyne Inc., Clearwater, FL), which was programmed to start at midnight (the following day) and actively sample at a 50% cycle and flow rate of 1.8 liters per minute (LPM). The pump was connected to a Harvard PM2.5 personal environmental monitor (PEM) with a 37mm Pall Teflo filter. Staff members provided instructions regarding proper use and demonstrated how to wear the sampling bag, with the sampling inlet located on the purse strap in the shoulder area around the breathing zone.

Participants were instructed to wear the sampling device during all waking hours while going about their normal daily activities. Exceptions to this requirement included while performing potentially dangerous activities (e.g., driving), showering, sleeping, or otherwise unable to. Participants were asked to protect the sampling device from water, high humidity (such as showering or sauna), heat, pets, and from children. When they could not wear the monitor continuously, such as when sleeping or driving, they were asked to place it on a bed side table or besides them on the passenger seat, away from surfaces as much as possible and unobstructed. Additionally, when not wearing the monitor, individuals were asked to keep the monitor elevated from the ground and away from the walls due to sampling artifacts that could result from resuspended dust or removal on surfaces, respectively.

The sampling pump was programmed to shut down after the 48 h sampling period was completed, and study staff coordinated device pickup and conducted a brief exit survey with participants the following day. When the sampling devices arrived at the USC Exposure Analytics lab, they were handled by trained staff. Pump data were downloaded, checked for errors, and securely stored. Filters were removed from the PEMs, allowed to equilibrate within a dedicated chamber and gravimetrically weighed in temperature and relative humidity-controlled glove box using an MT-5 calibrated microbalance to obtain PM2.5 mass concentration.

2.3. Birthweight

Birthweight in grams was abstracted from electronic medical record (EMR) for 210 mothers. Four mothers did not have birthweight recorded, possibly due to being lost to follow-up, and were removed from the analysis. Birth weight-for-gestational age z-scores were obtained for each participant using methods described in Aris et al. (2019).

2.4. Questionnaire and other covariate data

A priori covariates assessed in this analysis included factors related to maternal demographics, pregnancy and birth outcomes, meteorology, and study design variables, including recruitment site. Covariate data was collected during follow-up within the MADRES cohort from a series of in-person and telephone staff-administered questionnaires in either English or Spanish, ascertained throughout the study period up until date of infant birth. Anthropometric measurements were conducted via regular clinic visits. Data from the 3rd trimester visit was primarily used to coincide with the exposure period being studied. Additional covariate data came from the participants’ 1st visit, such as race/ethnicity, pre- pregnancy Body Mass Index (BMI, kg/m2), etc. and from pregnancy outcome data, including infant sex.

Maternal demographic variables analyzed for potential confounding included: age at baseline (years), pre-pregnancy BMI (continuous), education level (completed < 12th grade, completed high school, some college, completed college), household income (less than $15,000, $15,000–29,999, $30,000–49,999, $50,000+, Do not know), personal smoking status during pregnancy (yes/no), smoking status (ever/never), diabetes status (no diabetes, glucose intolerant, gestational diabetes, chronic diabetes), preeclampsia status (no hypertension, preeclampsia, chronic hypertension, chronic hypertension w/preeclampsia, gestational hypertension), and total weight gain (kg) during pregnancy. Diabetes and preeclampsia status were ascertained from EMR, while pre-pregnancy BMI was calculated using self-reported pre-pregnancy weight and standing height measured by MADRES staff at the first study visit via stadiometer (Perspectives Enterprises model PE-AIM-101, Portage, MI), or height from EMR if missing from first visit data. Self-reported pre- pregnancy weight was used because initial study visits ranged in terms of participants’ gestation. Race was recategorized from the NIH categories to a three-level variable containing Hispanic, Black non-Hispanic, and Other non-Hispanic. This was conducted to save degrees of freedom in the later regression analysis and because this sample is composed of largely Hispanic women (81%), followed by a smaller subset of Black non-Hispanic women (11%), and with non-Hispanic Whites, Asians, and Others combined making up just 8%.

Pregnancy and birth outcome-related potential covariates included: sex of infant (male/female), parity (defined as 1 or more previous births), and gestational age (GA; weeks). Infant sex was obtained via EMR, or if missing, through interviewer-administered questionnaires at the 7–14 day post-pregnancy follow-up. A missing category was created for 6 participants with missing parity. Gestational age at birth was estimated using a hierarchy of methods including, the preferred ultrasound measurement of crown-rump length at < 14 weeks gestation (60%), ultrasound measurement of fetal biparietal diameter at < 28 weeks gestation (30%), and from physicians’ clinical estimate from EMR (10%).

Meteorological parameters included ambient air temperature (Celsius) (calculated as average of minimum and maximum air temperature) and relative humidity (%), both integrated over the 48 h sampling period and estimated at the residential location based on a high-resolution (4 km x 4 km) gridded surface meteorological dataset (Abatzoglou, 2011). Season was categorized as Cool (Winter), Warm (Summer), and Transition (Spring and Autumn).

Finally, variables describing home ventilation, time-activity patterns, and presence of indoor sources of PM2.5 came from two different questionnaires. The first was from the 3rd trimester visit that asked questions related to the past month (or since the last visit in the 2nd trimester), while the second was from the exit survey administered after completing the 48 h personal monitoring period. These variables were chosen since they correlate with the potential of outdoor PM2.5 infiltration into the indoor home environment where participants likely spend most of their time, exposure to outdoor PM2.5 by spending time outdoors, or exposure to PM2.5 generated indoors from sources like cooking or candle use, respectively.

To describe these potential relationships in more detail, greater time spent indoors generally corresponds to greater exposure to indoor PM2.5, which in turn is predominantly composed of PM2.5 from indoor sources (or of indoor origin) and PM2.5 from outdoor origin (infiltrated indoors). The degree to which PM2.5 originating outdoors infiltrates into the indoor home environment depends on several factors including home ventilation (e.g., AC use, window opening, etc.) (Breen et al., 2014; Habre et al., 2014). Overall, greater time participants spend outdoors corresponds to potentially greater contribution of outdoor PM2.5 to their personal exposures (and vice versa). Additionally, several studies reported air tightness can be lower (higher leakiness) and air exchange rates can be higher in multi-unit residences (compared to single homes), which could mean greater potential for PM2.5 of outdoor origin or from neighboring units (e.g., secondhand smoke) to infiltrate indoors (King et al., 2010; Price et al., 2006; Rosofsky et al., 2019), but this likely varies across different contexts. AC use in the home can also remove indoor PM2.5 or correlate with lower infiltration of outdoor PM2.5 (due to more time with windows and doors closed and greater home sealing to the outdoors).

The final list of variables included: home type (building type/number of attached units), home ventilation (e.g., AC use, window opening time), time-activity patterns (e.g., time spent indoors, time spent outdoors), and indoor sources (e.g., cooking smoke, candle and incense smoke). All variables in this final list were available in both the exit survey and 3rd trimester questionnaire, apart from home type, which was only available from the 3rd trimester questionnaire, and candle smoke exposure, which was only asked in the exit survey. Several of these variables were re-categorized, when necessary, based on the distribution of the variable (Table S1).

2.5. Statistical analysis

2.5.1. Descriptive statistics

Descriptive statistics for birthweight and total personal PM2.5 were calculated by sample population characteristics. This preliminary bivariate analysis was also used to elicit potential confounders in this analysis. The distribution of PM2.5 exposure and birthweight were assessed to identify any deviations from normality and potential outliers. Differences in birthweight and total personal PM2.5 by the categorical sample characteristics were evaluated using analysis of variance (ANOVA) tests. Pearson’s correlation coefficients were calculated between continuous population characteristics and birthweight and total personal PM2.5 separately. Next, a correlation analysis was conducted to assess whether potential covariates were related to each other to examine collinearity and inform covariate inclusion in the models. Finally, a chi-square test was conducted to determine how well the two questionnaire measures correlated with one another for similar variables thereby providing a consistency check for differently worded questions, or questions that were asked at different points in time and referred to somewhat different time windows (e.g., past 48 h monitoring period versus the last month in the 3rd trimester).

2.5.2. Personal vs. outdoor residential PM2.5 exposure

To assess the relationship between total personal and outdoor PM2.5 exposure, daily outdoor residential PM2.5 concentration was estimated using inverse distance-weighted spatial interpolation from regulatory monitoring data. Daily estimates were averaged to correspond to the 48 h monitoring period and to the 3rd trimester of pregnancy. Descriptive statistics were obtained for the same 48 h monitoring period and for the 3rd trimester, and Pearson’s correlation coefficients were used to evaluate the relationship between personal and outdoor residential PM2.5.

2.5.3. Multiple linear regression models

Multiple linear regression models were used to investigate the association between in-utero exposure to PM2.5 and the continuous outcome birthweight. All parameter estimates for continuous variables were reported per 1 SD increase in personal PM2.5 concentrations, which is equivalent to 17.1 μg/m3 as shown in Table 1. Maternal age and race/ethnicity were included in all models due to their importance and inclusion in prior research. Additionally, due to the design of MADRES, recruitment site was also assessed in this analysis but did not impact findings, so was not included. A list of potential covariates based on the previous literature into the effect of air pollution and birth outcomes, and from the bivariate analysis conducted within this analysis, were assessed for inclusion into the model one-by-one based on evidence of confounding. Confounding was observed by gestational age, parity, diabetes status, infant sex, and smoking status. Pre-pregnancy BMI and total weight gain during pregnancy also introduced confounding; however, they were highly correlated with each other and with diabetes status. Each of these variables were assessed one at a time with the other included covariates and diabetes status was finally chosen to remain as it impacted the personal PM2.5 effect estimate the largest of the three.

Table 1.

Descriptive statistics of study participants (N = 205).

Variable Mean (SD) or n (%) Variable Mean (SD) or n (%)
Personal PM2.5 (μg/m3) 22.3 (17.1) Pre-Pregnancy BMI (kg/m2) 28.8 (6.8)
Birthweight (g) 3291.2 (485.1) Normal 62 (30.2%)
Total weight gain (kg) 10.9 (6.9) Overweight 64 (31.2%)
Gestational age (weeks) 39.1 (1.5) Obese 79 (38.5%)
Maternal age (years) 28.2 (6.0) Parity
Gender Yes 130 (63.4%)
 Female 105 (51.2%) No 69 (33.7%)
 Male 100 (48.8%) Missing 6 (2.9%)
Race Maternal Income
 Hispanic 166 (81.0%) $50,000–$99,999 14 (6.8%)
 Black, Non-Hispanic 23 (11.2%) $30,000–$49,999 29 (14.1%)
 Other, Non-Hispanic 16 (7.8%) $15,000–$29,999 46 (22.4%)
Education Less than $15,000 41 (20.0%)
 < 12th grade 49 (23.9%) Do not know 75 (36.6%)
 Completed high school 64 (31.2%) Smoking
 Some college 62 (30.2%) Ever 43 (21.0%)
 Completed college 30 (14.6%) Never 162 (79.0%)
Diabetes Temperature (°C) 19.3 (4.2)
 Normal 136 (66.3%) Season
 Glucose intolerant 46 (22.4%) Warm 40 (19.5%)
 Gestational diabetes 10 (4.9%) Cool 64 (31.2%)
 Chronic diabetes 13 (6.3%) Transition 101 (49.3%)
Open in a new tab

Notes: PM2.5 = particulate matter with aerodynamic diameter less than 2.5 μm; BMI = body mass index; SD = standard deviation; g = grams; kg = kilograms; transition = spring and autumn.

The final fully adjusted model included the following covariates: GA at birth, maternal age, race/ethnicity, infant sex, parity, diabetes status, temperature, and personal smoking history. This model was used to (1) evaluate the effect of total personal PM2.5 exposure on birthweight, (2) evaluate whether the effect of total personal PM2.5 exposure on birthweight was modified by the degree of which indoor vs outdoor sources contributed to or impacted personal PM2.5 exposures (broadly derived using questionnaire variables). The a priori significance level for the adjusted main exposure/outcome analysis was an alpha of 0.05. Model diagnostics were conducted to ensure they satisfied modeling assumptions and were not affected by multi-collinearity or influential points. Non-linear PM2.5 effects were evaluated using graphical plots and by adding polynomials into the model and evaluating statistical significance compared to linear terms. Due to birthweight and gestational age being closely linked, birthweight-for-gestational age z-scores were evaluated with personal PM2.5, however, results were not included as they did not reveal any additional information about the relationship between personal PM2.5 and birthweight. The analysis was conducted using SAS v9.4 (SAS Institute, Inc., Cary, NC, USA.).

2.5.4. Effect modification analyses for PM2.5 impacted by indoor vs outdoor sources

As described earlier, the second aim was to evaluate how the effects of total personal PM2.5 exposure differed when the contribution of outdoor sources (or PM2.5 of outdoor origin) was higher compared to indoor sources. Indoor vs outdoor origin of PM2.5 was approximated using interaction terms with variables that correspond to time-activity patterns (e.g., time spent indoors vs outdoors), indoor sources (e.g., cooking, candle use), home ventilation (e.g., AC use, window use), and home type (e.g., building type/number of attached units). This study investigated effect modification by adding an interaction term to the fully adjusted model, using an a prior significance level of 0.10 for the interaction.

2.5.5. Sensitivity analysis

Several analyses were conducted to evaluate the sensitivity of results to various inclusions. First, in the fully adjusted model, this study examined associations only among full-term births (37 weeks or older gestation) to assess whether the pre-term births impacted the associations seen in the full sample. Additionally, due to concerns regarding bias introduced by adjusting for GA, namely, that gestational age may be a mediator (Wilcox et al., 2011), this study performed the analysis without adjustment for GA. Finally, a model was run without the inclusion of the highest 4 personal PM2.5 concentrations, determined by concentrations being 2 SDs greater than the mean, to elicit their leverage on results.

3. Results

3.1. Descriptive statistics

Of the 214 mothers who participated in the personal exposure monitoring study, nine participants were removed due to incomplete or erroneous personal PM2.5 exposure data or birth outcomes data, resulting in a final analytical sample of 205 mother-infant dyads (Table 1). The women in the study were predominantly Hispanic (81%) and lower income, with over 55% of women reporting a household income less than $50,000 a year. Additionally, around 67% of participants were overweight or obese prior to pregnancy and most had at least one prior pregnancy (63%). One participant indicated that they had smoked cigarettes, cigars, or pipes during the 48 h sampling period, while all participants indicated not smoking during pregnancy on the 3rd trimester questionnaire (results not shown). Birthweight was normally distributed with a mean (SD) of 3291.2 (485.1) g Total personal PM2.5 exposure was right skewed with a mean (SD) of 22.3 (17.1) μg/m3 and median (IQR) of 18.2 (14.3) μg/m3. The participants had a mean (SD) age of 28.2 (6.0) years, delivered at a mean gestational age at birth of 39.1 (1.5) weeks, and gained on average 10.9 (6.9) kilograms throughout pregnancy.

3.2. Sociodemographic and household characteristics in relation to birthweight

Mothers who spent most or all of the time indoors had infants with significantly higher birthweight (most and all of the time: 3332.0 g vs. none and a little of the time: 3065.3 g; p = 0.005), while those who answered yes to using AC during the sampling period had infants that were about 190 g greater in birthweight than mothers that did not use AC (p = 0.013). Participants who had at least one child prior to this pregnancy had infants with higher birthweight (yes: 3338.8 g vs. no: 3180.6 g; p = 0.042). Infants of women who have completed at least college, had gestational or chronic diabetes, or were in the non-Hispanic Other category, had higher birthweight compared to their counterparts, however, none of these differences met statistical significance (Table S1).

Birthweight displayed a positive correlation with gestational age (Pearson r = 0.43; p < 0.001) and total weight gain throughout pregnancy (r = 0.29; p < 0.001), while maternal age showed no correlation (r = 0.02; p = 0.738; Table S2).

3.3. Sociodemographic and household characteristics in relation to total personal PM2.5 exposure

A statistically significant difference in personal PM2.5 was observed by maternal income, however, no obvious pattern emerged, with the highest and lowest income groups having the highest personal PM2.5 exposure (p = 0.025; Table S1). Participants who opened their windows none or a little of the time during the sampling period had slightly higher personal PM2.5 exposure (24.7 μg/m3) vs. most and all of the time (20.2μg/m3), which was marginally significant (p = 0.058). Personal PM2.5 differed by season of sampling (warm: 19.3 μg/m3 vs. transition: 21.0 μg/m3 vs. cool: 26.3; p = 0.072). Additionally, personal PM2.5 was significantly negatively associated with average 3rd trimester temperature (r = −0.15; p = 0.038; Table S2). Next, women who spent most or all of the time inside during the sampling period had lower personal PM2.5 exposure (21.4 vs 25.3 μg/m3 for women who spent none or a little of the time indoors; p = 0.249). Using the monitoring time-aligned exit survey question on cooking smoke exposure, there was no significant difference between those reporting being near cooking smoke and those that did not (none of the time: 22.1μg/m3 vs. a little, most, or all of the time 21.8μg/m3; p = 0.884).

3.4. Relationship Between Personal and Outdoor PM2.5 Exposure

The mean (SD) personal PM2.5 was 22.3 (17.1) μg/m3, while the outdoor residential estimate had a mean of 11.9 (5.5) μg/m3 for the same 48 h monitoring period, and 12.0 (2.3) μg/m3 for the 3rd trimester. Fig. 1. depicts these relationships between total personal and outdoor residential PM2.5. During the monitoring period, there was statistically significant yet weak correlation between total personal PM2.5 and ambient PM2.5 (r = 0.19; p = 0.006). A weak, positive and non-significant correlation between total personal PM2.5 and 3rd trimester ambient PM2.5 was also observed (r = 0.11; p = 0.110).

Fig. 1.

Fig. 1.

Open in a new tab

Relationship of personal PM2.5 and outdoor PM2.5 in (a) the 48 h monitoring period and (b) the third trimester of pregnancy.

3.5. Association of total personal PM2.5 exposure with birthweight

This study found no significant association between PM2.5 and birthweight (β = 37.4; 95% CI: −29.6, 104.3; p =0.273, per 1 SD increase in PM2.5) in the crude (unadjusted) regression model. Results remained similar in the fully-adjusted model (with maternal age, GA, maternal race/ethnicity, infant sex, parity, diabetes status, smoking status, and 3rd trimester average temperature, (β = 38.6; 95% CI: −21.1, 98.2; p = 0.204), as shown in Table 2. In the fully adjusted model, a one week increase in GA was associated with a 180.3g increase in birthweight (p < 0.001), females were on average 124.8 g lighter than males (p = 0.033), and participants that had not had a pregnancy before had on average 323.4 g lighter babies compared to those that had (p < 0.001). Finally, diabetes status was also an important predictor of birthweight, with participants with chronic diabetes (379.4 g; p = 0.003) and gestational diabetes (300.9 g; p = 0.028) having higher birthweight infants compared to those without diabetes.

Table 2.

Regression results for base model of PM2.5 and birthweight (N = 205).

Variable β 95% CI Lower 95% CI Upper p-value
Intercept −3452.3 −5096.7 −1807.9 < 0.001
Personal PM2.5 (μg/m3)a 38.6 −21.1 98.2 0.204
Gestational age (weeks) 180.3 141.1 219.5 < 0.001
Maternal age (years) −8.8 −19.8 2.3 0.121
Temperature (°C) ^ 11.3 −5.8 28.4 0.194
Race/ethnicity
 Hispanic −135.3 −356.9 86.3 0.230
 Black, non-Hispanic −248.7 −520.0 22.6 0.072
 Other, non-Hispanic Ref.
Sex of infant
 Female −124.8 −239.2 −10.3 0.033
 Male Ref.
Parity
 Missing 176.4 −166.0 518.7 0.311
 No −323.4 −459.4 −187.4 < 0.001
 Yes Ref.
Diabetes
 Chronic diabetes 379.4 129.7 629.1 0.003
 Gestational diabetes mellitus 300.9 33.1 568.7 0.028
 Glucose intolerant 111.3 −26.9 249.5 0.114
 Normal Ref.
Smoking
 Ever smoker −175.8 −319.7 −31.8 0.017
 Never smoker Ref.
Open in a new tab

Notes:

a

Per 1 SD increase in personal PM2.5; PM2.5 = particulate matter with aerodynamic diameter less than 2.5μm; CI = confidence interval

^ -

third trimester average temperature in degrees Celsius; Ref. = reference level.

3.6. Effect modification of total personal PM2.5 by contribution of indoor vs outdoor sources

While total personal PM2.5 was not associated with birthweight in the first aim of this study, this association differed significantly by several factors (Table 3). Home type was a significant effect modifier of personal PM2.5 exposure on birthweight (interaction p = 0.028, Fig. 2b). Participants living in a “house with no joining walls” (β =156.9; 95% CI: 26.9, 287.0) had a positive association with birthweight; while a negative association was observed as the number of units in the housing building increased (2–4 units: β = −16.6; 95% CI:−111.9, 78.7; 5+ units: β = −62.6; 95% CI:−184.9, 59.6). Additionally, the effect of PM2.5 on birthweight was significantly different by AC use (interaction p = 0.008), with more negative associations for participants that reported no AC use on the exit survey (β = −27.6; 95% CI:−101.5, 46.3), compared to any AC use during the 48 h monitoring period (β = 139.9; 95% CI: 42.9, 237.0) (Fig. 2d). A similar significant interaction and pattern was observed for AC use reported in the 3rd trimester (Table 3).

Table 3.

Estimated change in birthweight (g) per 1 SD increase in personal PM2.5 from interaction analyses (N = 204).

Time Activity Pattern β 95% CI Lower 95% CI Upper p-value
Time spent indoors
How much of the time did you spend indoors (at your residence, or someone else’s residence)?a 0.255
 None and a little of the time −45.1 −208.3 118.1
 Most and all of the time 57.1 −7.3 121.6
Thinking back to a typical weekday in this past week, approximately how many hours (out of 24 h in total) did you spend indoors?b 0.383
 ≤ 16 h −25.6 −184.8 133.6
 > 16 h 50.0 −13.9 113.9
Time spent outdoors
How much of the time did you spend outdoors (not commuting in a car, bus or train)?a 0.402
 None and a little of the time 59.5 −13.6 132.6
 Most and all of the time 6.8 −94.2 107.7
Thinking back to a typical weekday in this past week, approximately how many hours (out of 24 h in total) did you spend outdoors?b 0.411
 < 8 h 46.0 −16.5 108.5
 ≥ 8 h −33.5 −215.2 148.3
Home characteristics and ventilation
Home type
Which best describes the home in which you currently live most of the time?b 0.028
 A single-family house (no joining wall) 156.9 26.9 287.0
 A building with 2–4 attached Units −16.6 −111.9 78.7
 A building with 5+ attached Units −62.6 −184.9 59.6
Missing 145.4 −4.1 294.9
Time with windows open
How much of the time were windows (or porch/balcony doors if applicable) open in your home, when you were there with the sampler?a 0.936
 None and a little of the time 33.8 −35.3 102.9
 Most and all of the time 28.2 −89.3 145.8
On average, how much of the time were the windows open in your home this past week?b 0.230
 None and a little of the time 53.9 −14.2 122.0
 Most and all of the time −30.2 −151.8 91.5
Air conditioner use
How much of the time was the air conditioner used in your home, when you were there with the sampler?a 0.008
 None of the time −27.6 −101.5 46.3
 A little, most, and all of the time 139.9 42.9 237.0
Do you use air conditioning in your home?b 0.044
 No −24.3 −107.1 58.5
 Yes 99.4 13.5 185.3
Indoor sources
Cooking
How much of the time were you close to smoke or fumes from cooking (yourself, or nearby cooking by someone else) e.g., burnt toast, barbeque, stir fry, etc.?a 0.153
 None of the time 75.5 −3.5 154.5
 A little, most, and all of the time −10.6 −100.0 78.8
Since we last saw you/spoke to you in your first/second trimester, on average, how many times a week do you cook (using the stove/range/oven, not microwave)?b 0.085
 Never 158.0 9.2 306.8
 1 or more times a week 15.2 −50.2 80.5
Candle or incense
How much of the time were you close to smoke from candles or incense burning nearby?a 0.004
 None of the time 81.2 15.9 146.6
 A little, most, and all of the time −144.7 −282.7 −6.8
Open in a new tab

Notes: All interactions are adjusted for gestational age, gender, parity, race/ethnicity, maternal age, diabetes status, smoking, and temperature; PM2.5 = particulate matter with aerodynamic diameter less than 2.5 μm; CI = confidence interval

a

From exit questionnaire administered to participants after completing the 48 h personal exposure monitoring period

b

From 3rd trimester survey; bolded = statistically significant at p-value < 0.1.

Fig. 2.

Fig. 2.

Open in a new tab

Predicted relationship of personal PM2.5 exposure on birthweight (a) overall, (b) by type of home, (c) time spent indoors, and (d) air conditioner use at home, for each level of the interaction variable where applicable.

Participants who reported any exposure to smoke from candles or incense (only assessed in exit survey) had a negative association between PM2.5 and birthweight (β = −144.7; 95% CI: −282.7, −6.8), vs those who reported no exposure (β = 81.2; 95% CI: 15.9, 146.6). This interaction was statistically significant (p = 0.004). There was no significant interaction with cooking smoke exposure in the 48 h monitoring period (none: β = 75.5; 95% CI: −3.5, 154.5 vs any: β = −10.6; 95% CI: −100.0, 78.8; interaction p = 0.153). Results were similar when using the 3rd trimester questionnaire (Table 3).

There were also consistent, observable differences in the effect of PM2.5 on birthweight for variables related to time-activity patterns, although these interactions were not statistically significant (Table 3). Fig. 2c depicts when participants spent most and all of their time indoors during the 48 h monitoring period (at their residence or someone else’s), PM2.5 was positively associated with birthweight (β = 57.1; 95% CI: −7.3, 121.6), as compared to participants who spent none and a little of their time indoors (β = −45.1; 95% CI: −208.3, 118.1). The 3rd trimester questionnaire revealed a similar pattern, with a positive effect of PM2.5 on birthweight when participants spent greater than 16 h inside per day (β = 50.0; 95% CI: −13.9, 114.0) compared to a slight negative association for participants who spent less than or equal to 16 h inside (β = −25.6; 95% CI: −184.8, 133.6).

3.7. Results of sensitivity analyses with various inclusions

When the fully adjusted model was restricted to just those participants that had a full-term birth (≥ 37 weeks gestation; n = 182), the effect of total personal PM2.5 on birthweight increased slightly (β =55.0; 95% CI: −6.2, 116.1), compared to the base model used in aim 1 (β = 38.6; 95% CI: −21.1, 98.2; p = 0.204). In non-diabetics only (n=181), no association between total personal PM2.5 on birthweight was observed (β =19.2; 95% CI: −44.7, 83.2). When GA was not adjusted for, there was a 24% attenuation in the effect estimate for total personal PM2.5 on birthweight (β = 29.3; 95% CI: −41.8, 100.5). Finally, after excluding three observations that had high leverage and particularly high PM2.5 (>95 μg/m3) from the model, the effect of total personal PM2.5 on birthweight changed direction but remained non-significant (β = −40.1; 95% CI: −122.3, 42.1).

4. Discussion

Using data from the MADRES in-utero personal exposure monitoring study, this study evaluated the effect of total personal PM2.5 in the 3rd trimester on birthweight in a largely lower income, Hispanic population in Los Angeles, CA. According to a thorough review of existing literature, this is the first time this has been attempted in a health disparities population. This study finds that total personal PM2.5 was not statistically significantly associated with birthweight. Most studies of outdoor air pollution generally found a slight negative association with birthweight (Stieb et al., 2012); however, those studies were aimed at investigating personal exposure to PM2.5 of outdoor origin rather than total personal PM2.5 and generally relied on outdoor estimates of PM2.5 as its surrogate. Albeit a different question to the effect of total personal PM2.5 on birthweight, these outdoor estimates generally fail to account for time-activity patterns and infiltration of outdoor pollution into the home, thereby likely suffer from measurement error.

One study did look at the effect of total personal PM2.5 on birthweight in the 2nd trimester using personal monitoring over a 48 h period in a cohort of non-smoking women in Poland. They found an increase of ~30 μg/m3 in PM2.5 was associated with a 97.2g (95% CI: − 201.0, 6.6) decrease in birth weight (Jedrychowski et al., 2009). Despite the Poland findings not being statistically significant, a possible hypothesis for the difference in findings could be related to the differences between the women participating in the two studies. For example, compared to this study, participants were free from chronic diseases including diabetes, which this current study found to be a significant predictor of higher birthweight. Additionally, sources and chemical components of PM2.5 exposure in Poland may be different compared to this present study area in urban Los Angeles, CA. Studies have shown that PM2.5 sources and chemical components can vary in their effect on birthweight, with differences observed across regions (Basu et al., 2014; Bell et al., 2007), and across race/ethnic groups, especially in California (Bell and Ebisu, 2012). This study participants were largely Hispanic from Los Angeles County, while the Polish study population was predominantly non-Hispanic Whites (Jedrychowski et al., 2009). This highlights the importance of treating PM2.5 as a mixture in health analyses, with variable contributions from a wide range of indoor and outdoor sources with potentially differing physiochemical properties, components, and effects on birthweight. Comparing outdoor PM2.5 effects across regions, or outdoor to total personal PM2.5 effects, does not necessarily take into account this complexity or heterogeneity.

While there was no association between total personal PM2.5 and birthweight, this study did find evidence that home characteristics, such as home type and AC use, as well as exposure to candle or incense smoke, modified this association. Mothers residing in multi-unit buildings had a negative association of personal PM2.5 with birthweight, compared to a strong positive association for those who reside in a single-family home. One possible reason for this is that individuals in multi-unit homes may have greater secondhand smoke infiltration into their home from neighboring units (King et al., 2010; Price et al., 2006), and secondhand smoke has been shown to be negatively associated with birthweight (Ghosh et al., 2013; Wahabi et al., 2013). This study also considered whether single-family home type could be acting as a proxy for higher income. However, maternal income and educational attainment did not correlate with home type in this sample (results not shown). Additionally, despite having similar total personal PM2.5 exposure, the effect of PM2.5 on birthweight was significantly lower for participants that did not use AC compared to those that did. Using AC at home likely correlates with greater sealing of the home or closing of windows and doors to operate the AC unit(s), which also correlates with less infiltration outdoor PM2.5 indoors (and thus less exposure to PM2.5 of outdoor origin).

Although this study did not have complete information on all the possible indoor sources of PM2.5, this study saw evidence of significantly more negative or potentially harmful effects of candle and/or incense burning indoors on birthweight. Previous literature has shown that prenatal incense burning was associated with lower birthweight (Chen and Ho, 2016). One possible explanation is candle and incense burning emit black carbon (BC) and other chemicals indoors (Habre et al., 2014; Stabile et al., 2012). Several studies have reported an association between BC concentrations in PM2.5 and low birthweight; however, they were using outdoor BC as a surrogate or marker of outdoor, traffic-related air pollution (Lakshmanan et al., 2015; Slama et al., 2007). Bové et al. (2019) found BC accumulated on the fetal side of the human placenta, representing a potential mechanism for negative health effects. However, without further information on chemical composition and properties of the personal PM2.5 exposure mixture in these studies, it is difficult to conclude whether the candles and incense burning mixture as a whole or any particular component of it, such as BC, is driving these adverse effects.

This study did not find consistent effect modification results for exposure to PM2.5 from cooking, as another potentially important indoor source in this population. Most cooking smoke exposure and birthweight studies have concentrated on solid fuel sources (e.g., coal, wood), often in low- and middle- income countries, but generally find a negative association with birthweight (Wylie et al., 2017; Zhang et al., 2016). These findings may not be a suitable comparison with this study sample considering the participants were based in Los Angeles, CA where solid fuel cooking is not common, and where the composition or mixture of cooking related exposures may be different due to most participants using gas stoves (data not shown). It is also possible that cooking emits particles in the ultrafine size range (< 0.1 μm in aerodynamic diameter) which do not contribute significantly to PM2.5 mass concentrations, and thus, the measurements may not be sensitive enough to differentiate their contribution (as compared to particle number concentrations for example, which were not available in this study).

These analyses also revealed a consistent pattern where personal PM2.5 exposure with greater influence or contribution of outdoor sources was generally more strongly associated with lower birthweight, despite these interactions not reaching statistical significance. This was true for greater time spent outside (and less time spent inside), and greater time with open windows. However, these associations should be explored further, as the effect modification of time with windows open reported on the exit survey was less pronounced. While the present study may be underpowered to tease apart these differences, the results are consistent with prior studies assessing the impact of specific outdoor sources of PM2.5 or their surrogates such as on road gasoline and on road diesel, or residing closer to major roadways, respectively, which were associated with greater risk of LBW compared to PM2.5 as a whole (Bell et al., 2010; Laurent et al., 2016).

There are several strengths of this analysis. The first is the use of personal monitoring that provides a unique opportunity to examine personal exposure to PM2.5 and disentangle its impact on birthweight when it was more impacted by outdoor sources. This approach drastically reduces exposure measurement error as compared to using outdoor estimates of PM2.5 despite it being limited to a small sample (Gray et al., 2010). Next, this study was able to evaluate total personal PM2.5 exposure in the 3rd trimester, which may be particularly important for birthweight, given that most fetal growth occurs late in pregnancy. Most studies on the effect of PM2.5 on birthweight have used ambient monitoring data to estimate personal exposure to PM2.5 of outdoor origin, rather than the total personal PM2.5 to which individuals are exposed. Understanding the effects of outdoor PM2.5 on health is certainly an important question to evaluate, due to this being the fraction of PM2.5 that is regulated, but it is not the same question as the effect of total personal PM2.5 which takes into account all indoor, outdoor, and personal activity related sources that contribute to personal exposure as a result of realistic day-to-day behaviors and time-activity patterns. However, this study was also able to indirectly investigate whether the effects of personal PM2.5 differed when it was more impacted by indoor vs outdoor sources using interaction analyses with detailed, time-aligned questionnaire variables. It is also important to note that the chemical composition and size distribution of outdoor PM2.5 changes as it infiltrates indoors, which further highlights the importance of deciphering the independent effects of personal exposure to PM2.5 of indoor versus outdoor origin (Meng et al., 2007).

Furthermore, the MADRES cohort study is a well characterized prospective study in a health disparities population, with a host of covariates available, making this an ideal study to assess the research question at hand. This study also had the advantage of using two questionnaire data sources that differed in their time coverage and alignment (a longer-term 3rd trimester questionnaire vs an exit survey immediately following the 48 h monitoring period). This also allowed us to shed light on whether the 48 h sampling period reasonably represented behaviors, time-activity patterns, and 3rd trimester exposures in general.

The sample size of this study is a potential limitation with a final working sample of 205 participants, which while small for population-based studies is actually reasonably large for personal exposure monitoring studies (Dadvand et al., 2012; Sarnat et al., 2000; Suh and Zanobetti, 2010). Despite this limitation, this study was still able to observe differences in the influence of personal PM2.5 on birthweight by factors that drive indoor/outdoor source contributions, and most significantly for AC use and home type. Finally, participation bias may be a factor regarding who from the MADRES cohort chose to participate in the personal exposure monitoring study, however, participants who chose to participate were not noticeably different than the larger MADRES cohort study apart from being slightly more likely to have had a prior child (data not presented).

5. Conclusion

Overall, the results of this study did not find a significant association between total personal PM2.5 exposure and birthweight, however, there was evidence that multi-unit housing (vs. single family homes), candle and/or incense smoke exposure, and greater outdoor source contributions to personal PM2.5 were more strongly associated with lower birthweight. This highlights the importance of disentangling the mixture and apportioning PM2.5 by sources for health analyses, including potentially a more refined or chemically speciated approach to apportion indoor from outdoor source contributions to personal PM2.5.

Supplementary Material

1

Acknowledgments

The authors gratefully acknowledge the contributions of Lisa Valencia, Eleanne van Vliet, and the larger MADRES team. We also thank the MADRES participants and their families and our clinic partners for their time and effort.

Funding

The study was supported by NIEHS R01ES027409, NIEHS P30ES007048 pilot funding, the MADRES Center (NIH P50ES026086 and P50MD015705, and EPA 83615801) and NIMHD 5R01MD011698.

Footnotes

CRediT authorship contribution statement

Karl O’Sharkey: Formal analysis, Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing. Yan Xu: Data curation, Writing – review & editing. Thomas Chavez: Data curation, Writing – review & editing. Mark Johnson: Data curation, Writing – review & editing. Jane Cabison: Data curation, Investigation, Methodology, Project administration, Writing – review & editing. Marisela Rosales: Data curation, Investigation, Project administration, Writing – review & editing. Brendan Grubbs: Investigation, Writing – review & editing. Claudia M. Toledo-Corral: Investigation, Writing – review & editing. Shohreh F. Farzan: Conceptualization, Data curation, Investigation, Writing – review & editing. Theresa Bastain: Conceptualization, Data curation, Funding acquisition, Investigation, Project administration, Resources, Writing – review & editing. Carrie V. Breton: Conceptualization, Data curation, Funding acquisition, Investigation, Project administration, Resources, Writing – review & editing. Rima Habre: Conceptualization, Data curation, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Visualization, Writing – review & editing.

Supplementary materials

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.envadv.2022.100257.

References

  1. Abatzoglou JT, 2011. Development of gridded surface meteorological data for ecological applications and modelling. Abatzoglou - 2013 - International Journal of Climatology - Wiley Online Library Int. J. Climatol. 33, 121–131. [Google Scholar]
  2. Aris IM, Kleinman KP, Belfort MB, Kaimal A, Oken E, 2019. A 2017 US reference for singleton birth weight percentiles using obstetric estimates of gestation. Pediatrics 144, e20190076. 10.1542/peds.2019-0076. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Barnes SK, Ozanne SE, 2011. Pathways linking the early environment to long-term health and lifespan. Prog. Biophy. Mol. Biol. Dev. Program. Phenotype Divers. Epigenet. Dis. Risk 106, 323–336. 10.1016/j.pbiomolbio.2010.12.005. [DOI] [PubMed] [Google Scholar]
  4. Bastain TM, Chavez T, Habre R, Girguis MS, Grubbs B, Toledo-Corral C, Amadeus M, Farzan SF, Al-Marayati L, Lerner D, Noya D, Quimby A, Twogood S, Wilson M, Chatzi L, Cousineau M, Berhane K, Eckel SP, Lurmann F, Johnston J, Dunton GF, Gilliland F, Breton C, 2019. Study design, protocol and profile of the maternal and developmental risks from environmental and social stressors (MADRES) pregnancy cohort: a prospective cohort study in predominantly low-income hispanic women in Urban Los Angeles. BMC Pregnancy Childbirth 19. 10.1186/s12884-019-2330-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Basu R, Harris M, Sie L, Malig B, Broadwin R, Green R, 2014. Effects of fine particulate matter and its constituents on low birth weight among full-term infants in California. Environ. Res. 128, 42–51. 10.1016/j.envres.2013.10.008. [DOI] [PubMed] [Google Scholar]
  6. Bell ML, Belanger K, Ebisu K, Gent JF, Lee HJ, Koutrakis P, Leaderer BP, 2010. Prenatal exposure to fine particulate matter and birth weight. Epidemiology 21, 884–891. 10.1097/EDE.0b013e3181f2f405. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bell ML, Dominici F, Ebisu K, Zeger SL, Samet JM, 2007. Spatial and temporal variation in PM2.5 chemical composition in the united states for health effects studies. Environ. Health Perspect. 115, 989–995. 10.1289/ehp.9621. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Bell ML, Ebisu K, 2012. Environmental inequality in exposures to airborne particulate matter components in the United States. Environ. Health Perspect. 120, 1699–1704. 10.1289/ehp.1205201. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Bové H, Bongaerts E, Slenders E, et al. , 2019. Ambient black carbon particles reach the fetal side of human placenta. Nat. Commun. 10, 3866. 10.1038/s41467-019-11654-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Breen MS, Schultz BD, Sohn MD, Long T, Langstaff J, Williams R, Isaacs K, Meng QY, Stallings C, Smith L, 2014. A review of air exchange rate models for air pollution exposure assessments. J. Expo. Sci. Environ. Epidemiol. 24, 555–563. 10.1038/jes.2013.30. [DOI] [PubMed] [Google Scholar]
  11. Carroll RJ, 2005. Measurement error in epidemiologic studies. Encyclopedia of Biostatistics. American Cancer Society. 10.1002/0470011815.b2a03082. [DOI] [Google Scholar]
  12. Chen LY, Ho C, 2016. Incense burning during pregnancy and birth weight and head circumference among term births: the taiwan birth cohort study. Environ. Health Perspect. 124, 1487–1492. 10.1289/ehp.1509922. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Choi H, Rauh V, Garfinkel R, Tu Y, Perera FP, 2008. Prenatal exposure to airborne polycyclic aromatic hydrocarbons and risk of intrauterine growth restriction. Environ. Health Perspect. 116, 658–665. 10.1289/ehp.10958. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Choi H, Wang L, Lin X, Spengler JD, Perera FP, 2012. Fetal window of vulnerability to airborne polycyclic aromatic hydrocarbons on proportional intrauterine growth restriction. PLoS One 7, e35464. 10.1371/journal.pone.0035464. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Clemente DBP, Casas M, Vilahur N, Begiristain H, Bustamante M, Carsin A-E, Fernández MF, Fierens F, Gyselaers W, Iñiguez C, Janssen BG, Lefebvre W, Llop S, Olea N, Pedersen M, Pieters N, Santa Marina L, Souto A, Tardón A, Vanpoucke C, Vrijheid M, Sunyer J, Nawrot TS, 2016. Prenatal ambient air pollution, placental mitochondrial DNA content, and birth weight in the INMA (Spain) and ENVIRONAGE (Belgium) birth cohorts. Environ. Health Perspect. 124, 659–665. 10.1289/ehp.1408981. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Dadvand P, de NA, Triguero, -Mas Margarita, Schembari A, Cirach M, Amoly E, Figueras F, Basaga ña X., Ostro B, Nieuwenhuijsen M, 2012. Surrounding greenness and exposure to air pollution during pregnancy: an analysis of personal monitoring data. Environ. Health Perspect. 120, 1286–1290. 10.1289/ehp.1104609. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Dadvand P, Ostro B, Figueras F, Foraster M, Basagaña X, Valentín A, Martinez D, Beelen R, Cirach M, Hoek G, Jerrett M, Brunekreef B, Nieuwenhuijsen MJ, 2014. Residential proximity to major roads and term low birth weight: the roles of air pollution, heat, noise, and road-adjacent trees. Epidemiology 25, 518–525. 10.1097/EDE.0000000000000107. [DOI] [PubMed] [Google Scholar]
  18. Ebisu K, Belanger K, Bell ML, 2014. The Association between Airborne PM2.5 chemical constituents and birth weight–implication of buffer exposure assignment. Environ. Res. Lett. 9 10.1088/1748-9326/9/8/084007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Ghosh JKC, Wilhelm M, Ritz B, 2013. Effects of residential indoor air quality and household ventilation on preterm birth and term low birth weight in Los Angeles County, California. Am. J. Public Health 103, 686–694. 10.2105/AJPH.2012.300987. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Gray SC, Edwards SE, Miranda ML, 2010. Assessing exposure metrics for PM and birthweight models. J. Expo. Sci. Environ. Epidemiol. 20, 469–477. 10.1038/jes.2009.52. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Gray SC, Gelfand AE, Miranda ML, 2011. Hierarchical spatial modeling of uncertainty in air pollution and birth weight study. Stat. Med. 30, 2187–2198. 10.1002/sim.4234. [DOI] [PubMed] [Google Scholar]
  22. Habre R, Coull B, Moshier E, Godbold J, Grunin A, Nath A, Castro W, Schachter N, Rohr A, Kattan M, Spengler J, Koutrakis P, 2014. Sources of indoor air pollution in New York City residences of asthmatic children. J. Expo. Sci. Environ. Epidemiol. 24, 269–278. 10.1038/jes.2013.74. [DOI] [PubMed] [Google Scholar]
  23. Harris G, Thompson WD, Fitzgerald E, Wartenberg D, 2014. The association of PM2.5 with full term low birth weight at different spatial scales. Environ. Res. 134, 427–434. 10.1016/j.envres.2014.05.034. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Huang C, Nichols C, Liu Y, Zhang Y, Liu X, Gao S, Li Z, Ren A, 2015. Ambient air pollution and adverse birth outcomes: a natural experiment study. Popul. Health Metr. 13 10.1186/s12963-015-0050-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Jedrychowski W, Perera F, Mrozek-Budzyn D, Mroz E, Flak E, Spengler JD, Edwards S, Jacek R, Kaim I, Skolicki Z, 2009. Gender differences in fetal growth of newborns exposed prenatally to airborne fine particulate matter. Environ. Res. 109, 447–456. 10.1016/j.envres.2009.01.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Jornayvaz FR, Vollenweider P, Bochud M, Mooser V, Waeber G, Marques- Vidal P, 2016. Low birth weight leads to obesity, diabetes and increased leptin levels in adults: the CoLaus study. Cardiovasc. Diabetol. 15 10.1186/s12933-016-0389-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. King BA, Travers MJ, Cummings KM, Mahoney MC, Hyland AJ, 2010. Secondhand smoke transfer in multiunit housing. Nicotine Tob. Res. 12, 1133–1141. 10.1093/ntr/ntq162. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Kioumourtzoglou MA, Spiegelman D, Szpiro AA, Sheppard L, Kaufman JD, Yanosky JD, Williams R, Laden F, Hong B, Suh H, 2014. Exposure measurement error in PM2.5 health effects studies: a pooled analysis of eight personal exposure validation studies. Environ. Health 13, 2. 10.1186/1476-069X-13-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Kiserud T, Benachi A, Hecher K, Perez RG, Carvalho J, Piaggio G, Platt LD, 2018. The world health organization fetal growth charts: concept, findings, interpretation, and application. Am. J. Obstet. Gynecol. 218, S619–S629. 10.1016/j.ajog.2017.12.010. [DOI] [PubMed] [Google Scholar]
  30. Lakshmanan A, Chiu YHM, Coull BA, Just AC, Maxwell SL, Schwartz J, Gryparis A, Kloog I, Wright RJ, Wright RO, 2015. Associations between prenatal traffic-related air pollution exposure and birth weight: modification by sex and maternal pre-pregnancy body mass index. Environ. Res. 137, 268–277. 10.1016/j.envres.2014.10.035. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Lamichhane DK, Leem JH, Lee JY, Kim HC, 2015. A meta-analysis of exposure to particulate matter and adverse birth outcomes. Environ. Health Toxicol. 30 10.5620/eht.e2015011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Laurent O, Hu J, Li L, Kleeman MJ, Bartell SM, Cockburn M, Escobedo L, Wu J, 2016. Low birth weight and air pollution in California: which sources and components drive the risk? Environ. Int. 92–93, 471–477. 10.1016/j.envint.2016.04.034. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Li X, Huang S, Jiao A, Yang X, Yun J, Wang Y, Xue X, Chu Y, Liu F, Liu Y, Ren M, Chen X, Li N, Lu Y, Mao Z, Tian L, Xiang H, 2017. Association between ambient fine particulate matter and preterm birth or term low birth weight: an updated systematic review and meta-analysis. Environ. Pollut. 227, 596–605. 10.1016/j.envpol.2017.03.055. [DOI] [PubMed] [Google Scholar]
  34. Li Zhou, Tang Y, Song X, Lazar L, Li Zhen, Zhao J, 2019. Impact of ambient PM2.5 on adverse birth outcome and potential molecular mechanism. Ecotoxicol. Environ. Saf. 169, 248–254. 10.1016/j.ecoenv.2018.10.109. [DOI] [PubMed] [Google Scholar]
  35. Meng QY, Turpin BJ, Lee JH, Polidori A, Weisel CP, Morandi M, Colome S, Zhang J, Stock T, Winer A, 2007. How does infiltration behavior modify the composition of ambient PM2.5 in indoor spaces? An analysis of RIOPA data. Environ. Sci. Technol. 41, 7315–7321. 10.1021/es070037k. [DOI] [PubMed] [Google Scholar]
  36. Mi D, Fang H, Zhao Y, Zhong L, 2017. Birth weight and type 2 diabetes: a meta-analysis. Exp. Ther. Med. 14, 5313–5320. 10.3892/etm.2017.5234. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Mikati I, Benson AF, Luben TJ, Sacks JD, Richmond-Bryant J, 2018. Disparities in distribution of particulate matter emission sources by race and poverty status. Am. J. Public Health 108, 480–485. 10.2105/AJPH.2017.304297. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Morello-Frosch R, Zuk M, Jerrett M, Shamasunder B, Kyle AD, 2011. Understanding the cumulative impacts of inequalities in environmental health: implications for policy. Health Aff. 30, 879–887. 10.1377/hlthaff.2011.0153. [DOI] [PubMed] [Google Scholar]
  39. Morrison JL, Duffield JA, Muhlhausler BS, Gentili S, McMillen IC, 2010. Fetal growth restriction, catch-up growth and the early origins of insulin resistance and visceral obesity. Pediatr. Nephrol. 25, 669–677. 10.1007/s00467-009-1407-3. [DOI] [PubMed] [Google Scholar]
  40. Pedersen M, Giorgis-Allemand L, Bernard C, Aguilera I, Andersen A-MN, Ballester F, Beelen RMJ, Chatzi L, Cirach M, Danileviciute A, Dedele A, Eijsden M van, Estarlich M, Fernández-Somoano A, Fernández MF, Forastiere F, Gehring U, Grazuleviciene R, Gruzieva O, Heude B, Hoek G, Hoogh K.de, van den Hooven EH, Håberg SE, Jaddoe VWV, Klümper C, Korek M, Kramer U, Lerchundi A, Lepeule J, Nafstad P, Nystad W, Patelarou E, Porta D, Postma D, Raaschou-Nielsen O, Rudnai P, Sunyer J, Stephanou E, Sørensen M, Thiering E, Tuffnell D, Varró MJ, Vrijkotte TGM, Wijga A, Wilhelm M, Wright J, Nieuwenhuijsen MJ, Pershagen G, Brunekreef B, Kogevinas M, Slama R, 2013. Ambient air pollution and low birthweight: a European cohort study (ESCAPE). Lancet Respir. Med. 1, 695–704. 10.1016/S2213-2600(13)70192-9. [DOI] [PubMed] [Google Scholar]
  41. Petersen R, 2019. Racial and ethnic disparities in adult obesity in the United States: CDC’s tracking to inform state and local action. Prev. Chronic Dis. 16 10.5888/pcd16.180579. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Martin JA, Hamilton BE, Osterman MJK, Driscoll AK Births: Final data for 2018. National Vital Statistics Reports; vol 68, no 13. Hyatsville, MD: National Center for Health Statistics. 2019. Accessed 29 June 2022 from https://www.cdc.gov/nchs/data/nvsr/nvsr68/nvsr68_13-508.pdf. [PubMed] [Google Scholar]
  43. Price PN, Shehabi A, Chan RW, Gadgil AJ Indoor-Outdoor Air Leakage of Apartments and Commercial Buildings. United States: N. p, 2006. Web. doi: 10.2172/1010623. [DOI] [Google Scholar]
  44. Rich DQ, Liu K, Zhang Jinliang, Thurston SW, Stevens TP, Pan Y, Kane C, Weinberger B, Ohman-Strickland P, Woodruff TJ, Duan X, Assibey-Mensah V, Zhang Junfeng, 2015. Differences in birth weight associated with the 2008 Beijing olympics air pollution reduction: results from a natural experiment. Environ. Health Perspect. 123, 880–887. 10.1289/ehp.1408795. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Risnes KR, Vatten LJ, Baker JL, Jameson K, Sovio U, Kajantie E, Osler M, Morley R, Jokela M, Painter RC, Sundh V, Jacobsen GW, Eriksson JG, Sørensen TIA, Bracken MB, 2011. Birthweight and mortality in adulthood: a systematic review and meta-analysis. Int. J. Epidemiol. 40, 647–661. 10.1093/ije/dyq267. [DOI] [PubMed] [Google Scholar]
  46. Rosofsky A, Levy JI, Breen MS, Zanobetti A, Fabian MP, 2019. The impact of air exchange rate on ambient air pollution exposure and inequalities across all residential parcels in Massachusetts. J. Expo. Sci. Environ. Epidemiol. 29, 520–530. 10.1038/s41370-018-0068-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Rossen LM, 2014. Neighbourhood economic deprivation explains racial/ethnic disparities in overweight and obesity among children and adolescents in the USA. J. Epidemiol. Commun. Health 68, 123–129. 10.1136/jech-2012-202245. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Rundle A, Hoepner L, Hassoun A, Oberfield S, Freyer G, Holmes D, Reyes M, Quinn J, Camann D, Perera F, Whyatt R, 2012. Association of childhood obesity with maternal exposure to ambient air polycyclic aromatic hydrocarbons during pregnancy. Am. J. Epidemiol. 175, 1163–1172. 10.1093/aje/kwr455. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Sarnat JA, Koutrakis P, Suh HH, 2000. Assessing the relationship between personal particulate and gaseous exposures of senior citizens living in Baltimore, MD. J. Air Waste Manag. Assoc. 50, 1184–1198. 10.1080/10473289.2000.10464165. [DOI] [PubMed] [Google Scholar]
  50. Schembari A, de Hoogh K, Pedersen M, Dadvand P, Martinez D, Hoek G, Petherick ES, Wright J, Nieuwenhuijsen MJ, 2015. Ambient air pollution and newborn size and adiposity at birth: differences by maternal ethnicity (the born in bradford study cohort). Environ. Health Perspect. 123, 1208–1215. 10.1289/ehp.1408675. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Slama R, Morgenstern V, Cyrys J, Zutavern A, Herbarth O, Wichmann H-E, Heinrich J, 2007. Traffic-related atmospheric pollutants levels during pregnancy and offspring’s term birth weight: a study relying on a land-use regression exposure model. Environ. Health Perspect. 115, 1283–1292. 10.1289/ehp.10047. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Smith CJ, Ryckman KK, Barnabei VM, Howard BV, Isasi CR, Sarto GE, Tom SE, Van Horn LV, Wallace RB, Robinson JG, 2016. The impact of birth weight on cardiovascular disease risk in the Women’s Health Initiative. Nutr. Metab. Cardiovasc. Dis. 26, 239–245. 10.1016/j.numecd.2015.10.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Stabile L, Fuoco FC, Buonanno G, 2012. Characteristics of particles and black carbon emitted by combustion of incenses, candles and anti-mosquito products. Build. Environ. 56, 184–191. 10.1016/j.buildenv.2012.03.005. [DOI] [Google Scholar]
  54. Stieb DM, Chen L, Eshoul M, Judek S, 2012. Ambient air pollution, birth weight and preterm birth: a systematic review and meta-analysis. Environ. Res. 117, 100–111. 10.1016/j.envres.2012.05.007. [DOI] [PubMed] [Google Scholar]
  55. Stieb DM, Chen L, Hystad P, Beckerman BS, Jerrett M, Tjepkema M, Crouse DL, Omariba DW, Peters PA, van Donkelaar A, Martin RV, Burnett RT, Liu S, Smith-Doiron M, Dugandzic RM, 2016. A national study of the association between traffic-related air pollution and adverse pregnancy outcomes in Canada, 1999–2008. Environ. Res. 148, 513–526. 10.1016/j.envres.2016.04.025. [DOI] [PubMed] [Google Scholar]
  56. Suh HH, Zanobetti A, 2010. Exposure error masks the relationship between traffic-related air pollution and heart rate variability (HRV). J. Occup. Environ. Med. 52, 685–692. 10.1097/JOM.0b013e3181e8071f. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Sun X, Luo X, Zhao C, Zhang B, Tao J, Yang Z, Ma W, Liu T, 2016. The associations between birth weight and exposure to fine particulate matter (PM2.5) and its chemical constituents during pregnancy: a meta-analysis. Environ. Pollut. 211, 38–47. 10.1016/j.envpol.2015.12.022. [DOI] [PubMed] [Google Scholar]
  58. Umer A, Hamilton C, Cottrell L, Giacobbi P, Innes K, Kelley GA, Neal W, John C, Lilly C, 2020. Association between birth weight and childhood cardiovascular disease risk factors in West Virginia. J. Dev. Orig. Health Dis. 11, 86–95. 10.1017/S204017441900045X. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Upadhyay RP, Naik G, Choudhary TS, Chowdhury R, Taneja S, Bhandari N, Martines JC, Bahl R, Bhan MK, 2019. Cognitive and motor outcomes in children born low birth weight: a systematic review and meta-analysis of studies from South Asia. BMC Pediatr. 19, 35. 10.1186/s12887-019-1408-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Vilanova CS, Hirakata VN, de Souza Buriol VC, Nunes M, Goldani MZ, da Silva CH, 2019. The relationship between the different low birth weight strata of newborns with infant mortality and the influence of the main health determinants in the extreme south of Brazil. Popul. Health Metr. 17, 15. 10.1186/s12963-019-0195-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Visentin S, Grumolato F, Nardelli GB, Di Camillo B, Grisan E, Cosmi E, 2014. Early origins of adult disease: low birth weight and vascular remodeling. Atherosclerosis 237, 391–399. 10.1016/j.atherosclerosis.2014.09.027. [DOI] [PubMed] [Google Scholar]
  62. Wahabi HA, Alzeidan RA, Fayed AA, Mandil A, Al-Shaikh G, Esmaeil SA, 2013. Effects of secondhand smoke on the birth weight of term infants and the demographic profile of Saudi exposed women. BMC Public Health 13, 341. 10.1186/1471-2458-13-341. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Watkins WJ, Kotecha SJ, Kotecha S, 2016. All-Cause mortality of low birthweight infants in infancy, childhood, and adolescence: population study of england and wales. PLoS Med. 13 10.1371/journal.pmed.1002018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Whitaker AH, Feldman JF, Lorenz JM, Shen S, McNicholas F, Nieto M, McCulloch D, Pinto-Martin JA, Paneth N, 2006. Motor and cognitive outcomes in nondisabled low-birth-weight adolescents: early determinants. Arch. Pediatr. Adolesc. Med. 160, 1040. 10.1001/archpedi.160.10.1040. [DOI] [PubMed] [Google Scholar]
  65. Wilcox AJ, Weinberg CR, Basso O, 2011. On the pitfalls of adjusting for gestational age at birth. Am. J. Epidemiol. 174, 1062–1068. 10.1093/aje/kwr230. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Wylie BJ, Kishashu Y, Matechi E, Zhou Z, Coull B, Abioye AI, Dionisio KL, Mugusi F, Premji Z, Fawzi W, Hauser R, Ezzati M, 2017. Maternal exposure to carbon monoxide and fine particulate matter during pregnancy in an urban Tanzanian cohort. Indoor Air 27, 136–146. 10.1111/ina.12289. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Zeger SL, Thomas D, Dominici F, Samet JM, Schwartz J, Dockery D, Cohen A, 2000. Exposure measurement error in time-series studies of air pollution: concepts and consequences. Environ. Health Perspect. 108, 419–426. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Zhang Yaqun, Cao S, Xu X, Qiu J, Chen M, Wang D, Guan D, Wang C, Wang X, Dong B, Huang H, Zhao N, Jin L, Bai Y, Duan X, Liu Q, Zhang, Yawei, 2016. Metals compositions of indoor PM2.5, health risk assessment, and birth outcomes in Lanzhou, China. Environ. Monit. Assess. 188, 325. 10.1007/s10661-016-5319-y. [DOI] [PubMed] [Google Scholar]
  69. Zhu X, Liu Y, Chen Y, Yao C, Che Z, Cao J, 2015. Maternal exposure to fine particulate matter (PM2.5) and pregnancy outcomes: a meta-analysis. Environ. Sci. Pollut. Res. 22, 3383–3396. 10.1007/s11356-014-3458-7. [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

1
NIHMS1853765-supplement-1.pdf (202.3KB, pdf)

RESOURCES