Phthalate biomarkers and associations with respiratory symptoms and healthcare utilization among low-income urban children with asthma

Abstract

Background:

Phthalates are synthetic chemicals present in building materials, personal care products and other consumer goods. Limited studies link phthalates to pediatric asthma incidence; however, their effects on respiratory-related outcomes among those with pre-existing asthma remains unclear.

Objective:

We examined associations between phthalates and asthma symptoms, healthcare use, lung function, and lung inflammation among children with asthma.

Methods:

We collected repeated measures of urinary biomarkers for select phthalates and phthalate replacements (MBzP, MCINP, MCIOP, MCPP, MECPTP, MEHHTP, molar sum of DEHP biomarkers [MECPP, MEHHP, MEHP, MEOHP], MEP, MiBP, MnBP) and asthma symptoms, healthcare utilization, lung function, and inflammation among 148 predominantly low-income Black children (5–17 years) with persistent asthma every 3 months for one year. We used generalized estimating equations to assess associations between biomarker concentrations and asthma-related measures adjusting for age, sex, race/ethnicity, caregiver’s education level, presence of smokers in the home, and season. We also considered co-exposures to other contaminants previously associated with asthma morbidity.

Results:

We observed consistent positive associations with individual DEHP biomarkers, the molar sum of DEHP, and BBzP with increased odds of asthma symptoms and with healthcare utilization (adjusted Odds Ratio for general asthma symptoms: ΣDEHP:1.49,95% Confidence Interval, CI:1.08–2.07; BBzP:1.34, CI:1.04–1.73). We observed similar associations between the DEHP phthalate replacement biomarker MEHHTP and most asthma symptoms evaluated; and with select low molecular weight phthalates (DiBP, DBP) and healthcare utilization. Results were similar when controlling for other environmental exposures (e.g., PM_2.5, BPA). No associations were observed with lung function or inflammation, and overall, we did not observe consistent evidence of sexually dimorphic effects.

Conclusion:

In the present study, we found evidence to suggest that exposure to select phthalates may be associated with asthma symptoms and healthcare utilization. These findings warrant confirmation given the high asthma burden and widespread and disparate phthalate exposures reported among select populations of color.

1. Introduction

Asthma is a chronic pediatric disease, impacting over 6 million U.S. children (~7% of U.S. children) and disproportionately affecting Black and Latinx children (Assari and Moghani Lankarani, 2018; Centers for Disease Control and Prevention, 2020; Forno and Celedón, 2009). Emerging evidence suggests that exposure to environmental agents with endocrine disrupting properties like phthalates may increase risk of asthma development and alteration of immune function (Benjamin et al., 2017; Li et al., 2017; North et al., 2014; Robinson and Miller, 2015). Phthalates are synthetic chemicals widely used as plasticizers to impart flexibility and durability and also used as solvents in many consumer products. Phthalates may be grouped into low and high molecular weight. Beyond their chemical properties these two groups generally reflect distinct exposure sources. Low molecular weight phthalates are mostly used in a variety of personal hygiene and cosmetic products, such as fragrances as scent stabilizers; while high molecular weight phthalates are commonly used as plasticizers, including in toys, medical equipment, polyvinyl chloride (PVC) building materials, and as excipients in medications (North et al., 2014; Wang et al., 2019). The primary route of exposure among school-aged children, particularly to high molecular weight phthalates, is believed to be dietary ingestion (foods and drinks packaged in plastic); other exposure routes include inhalation and dermal absorption (Benjamin et al., 2017; Wang et al., 2019).

Although the underlying mechanisms by which phthalates could impact respiratory health have not been clearly elucidated, prior reviews on experimental studies indicate that phthalates can increase allergic immune response, contribute to airway remodeling (Kimber and Dearman, 2010), and possibly induce asthma (Bølling et al., 2020; North et al., 2014; Tsai et al., 2012). Experimental studies also report inconsistent inflammatory pathways based on exposure route and dose, (Kimber and Dearman, 2010), but overall suggest that phthalates may induce oxidative stress, and initiate and augment Th2-dominated airway inflammation potentially leading to adverse respiratory effects (Alfardan et al., 2018; Benjamin et al., 2017; Bølling et al., 2020; Robinson and Miller, 2015). Increasing epidemiologic evidence also continues to support the possible association of phthalates with the development of allergic conditions and asthma (Benjamin et al., 2017; North et al., 2014; Robinson and Miller, 2015; Wang et al., 2019). For example, cross-sectional studies in the U.S. have reported positive associations between phthalate biomarker concentrations and current asthma in children and adults (Hoppin et al., 2013; Odebeatu et al., 2019). Phthalates have also been associated with asthma incidence and allergy symptoms in studies conducted in other countries around the world including Taiwan (Hsu et al., 2012), Sweden (Bornehag et al., 2004), and China (Shi et al., 2018). At present, most studies have focused on asthma incidence in adult populations or in understanding the role of prenatal phthalate exposures on pediatric asthma development; however, studies on the role of phthalates as environmental triggers among those with respiratory conditions remain limited (Benjamin et al., 2017; Li et al., 2017; Wang et al., 2019). For example, the role of phthalate exposure in asthma-related symptoms among those with chronic asthma remains unclear despite ubiquitous and disparate human exposure to phthalates and emerging links with asthma incidence and development (Benjamin et al., 2017; Odebeatu et al., 2019). Furthermore, due to emerging evidence of health concerns linked to phthalates, newer phthalate replacements (such as Di-2-ethylhexyl terephthalate, DEHTP, and di (isononyl)cyclohexane-1,2-dicarboxylate, DINCH) are being widely used even though very limited information about their health impacts exists (Silva et al., 2013, 2019).

In the present study, we sought to examine associations between urinary phthalate biomarker concentrations and concurrent asthma symptoms, healthcare utilization, lung function, and lung inflammation among children with asthma in a low-income urban setting. We also include newer phthalate replacements introduced to the market in recent years given limited exposure and epidemiologic data on these chemical substitutes. We hypothesized that exposure to phthalates and newer replacements would be associated with increased asthma symptoms and healthcare utilization, altered pulmonary function and increased lung inflammation.

2. Methods

2.1. Study population

We used data and samples collected from 148 children with asthma participating in the Mouse Allergen and Asthma Cohort Study (MAACS), a 12-month prospective study in Baltimore, Maryland, USA. MAACS aimed to evaluate the impact of indoor allergens and air pollutants on clinical markers of asthma morbidity (Ahluwalia et al., 2013; Matsui et al., 2013; Torjusen et al., 2013). Briefly, children 5–17 years of age were recruited between April 2007 and June 2009 from the Johns Hopkins Hospital Emergency Department (ED), health fairs, and word of mouth. Eligible children had received an asthma physician diagnosis at least one year prior to study enrollment, experienced more than one exacerbation in the prior year, and had a controller medication prescription or met the National Asthma Education and Prevention Program criteria for persistent asthma (National Heart, Lung, and Blood Institute (US), 2007). An exacerbation was defined as asthma symptoms requiring an emergency medical visit or an oral corticosteroid burst. Enrollment was limited to non-smokers as verified with a rapid urine cotinine screening and to non-pregnant females. Rolling enrollment was used allowing for sample collection and outcomes to be measured over several seasons throughout the 12-month study period. The Johns Hopkins School of Medicine Institutional Review Board Study approved all study protocols and informed consent was obtained from children’s caregivers prior to study enrollment.

2.2. Study procedures and clinical assessments

Children participated in up to five clinic visits, including a baseline visit and four additional visits occurring at three-month intervals (Figure E1). At the baseline visit, study staff collected demographic information and administered an allergy skin prick test with the MultiTest II device (Lincoln Diagnostics, Decatur, IL). Atopy was defined as ≥1 positive skin test response, with a wheal size of ≥2 mm greater than the negative control.

At each clinic visit, we collected information on number of days with symptoms and rescue medication use in the prior two weeks, and asthma-related healthcare utilization in the prior three months. Asthma-related symptoms included: general symptoms (defined as chest tightness, wheeze, cough), nocturnal wakening with symptoms, exercise-related symptoms (defined as coughing or chest tightness when running/going upstairs), coughing without a cold, and slowed activity due to asthma (defined as the child having to slow down or stop activities while at home or playing with other children because of asthma, wheezing tightness in the chest or cough). A “maximal symptom days” variable was generated and defined as the largest value reported among the following symptom variables: the maximum number of days contributed by general symptoms, nocturnal wakening with symptoms, and slowed activity due to asthma in the prior two weeks (Ahmed et al., 2020; Quirós-Alcalá et al., 2021). Healthcare utilization included acute care visits (unscheduled visits for asthma-related symptoms), emergency department visits, unscheduled physician visits, and hospitalization for asthma-related symptoms. The questionnaires used to capture information on asthma-related symptoms have been previously validated (Juniper et al., 1999, 2010; Wu et al., 2019) and extensively used in pediatric asthma studies, including those conducted in urban settings (Gruchalla et al., 2005; Morgan et al., 2004; Phipatanakul et al., 2000; Pongracic et al., 2008).

We also collected lung function metrics (forced expiratory volume in the first second, FEV₁, and forced vital capacity, FVC) with spirometry measurements collected at each clinic visit with a KoKo spirometer (nSpire Health, Longmont, Colorado). We estimated percent predicted values of lung function measures using the Global Lung Initiative multi-ethnic, age-specific reference equations for individuals three years and older (Quanjer et al., 2012). Bronchodilator reversibility was defined as a ≥12% increase in FEV₁ 15 min after administration of two puffs of short-acting β-agonist. Lung inflammation was assessed by measuring the fraction of exhaled nitric oxide (FeNO) levels with a NIOX Mino (Aerocrine, Solana, Sweden) following American Thoracic Society/European Respiratory Society guidelines (American Thoracic Society; European Respiratory Society, 2005).

2.3. Exposure assessment of phthalates and phthalate replacements

Spot urine samples (n = 650) were collected at each clinic visit in polypropylene phthalate-free urine collection cups, aliquoted, and stored at −80 °C until shipment for urinary phthalate biomarker analysis. To adjust for sample dilution, we measured specific gravity in each sample using a digital refractometer (ATAGO^™3741, Tokyo, Japan). The 16 biomarkers analyzed, their respective parent compounds, and common sources are presented in Table E1 [monoethyl phthalate (MEP), mono-n-butylphthalate (MnBP), mono-isobutyl phthalate (MiBP), mono (3-carboxypropyl) phthalate (MCPP), mono (2-ethylhexyl) phthalate (MEHP), monobenzyl phthalate (MBzP), monoisononyl phthalate (MNP), mono (2-ethyl-5-oxohexyl) phthalate (MEOHP), mono (2-ethyl-5-hydroxyhexyl) phthalate (MEHHP), mono (2-ethyl-5-carboxypentyl) phthalate (MECPP), monocarboxyoctyl phthalate (MCOP), monocarboxynonyl phthalate (MCNP), and cyclohexane-1,2-dicarboxylic acid-mono (hydroxy-isononyl) ester (MHNCH)]. Analysis of urinary biomarkers was performed at NSF International (Ann Arbor, MI, USA) using liquid chromatography tandem mass spectrometry method described in detail elsewhere that was developed to replicate the Centers for Disease Control and Prevention (CDC) Phthalate biomarkers in Urine Method No: 6306.03 and was evaluated against similar acceptance criteria established within the CDC method (Centers for Disease Control and Prevention, 2010; Lewis et al., 2013; Silva et al., 2007). Briefly, samples underwent enzymatic deconjugation of glucuronidated species, online solid phase extraction, and analysis with a Thermo Scientific (Waltham, MA, USA) Vantage triple quadrupole mass spectrometer in using multiple reaction monitoring in negative ionization mode. The validated analyte calibration curve correlation coefficient (R²) range was 0.985–1.000. The method accuracy (% nominal concentration) and precision (%RSD) were determined through six replicate analyses of analytes spiked at four different concentrations in aqueous and human urine across validation runs (Lewis et al., 2013). Biomarker concentrations were specific gravity-corrected to normalize for urine dilution (Adibi et al., 2008; Hauser et al., 2004; Hines et al., 2009) using the following formula: C_sg = [(C*1.024–1)/(SG-1)], where C_sg is the specific-gravity corrected phthalate biomarker concentration (ng/mL), C is the observed phthalate biomarker concentration (ng/mL), 1.024 is the median specific gravity for our study population, and SG is the specific gravity for each urine sample (Boeniger et al., 1993). For concentrations below the limit of detection (LOD), we used the instrument values when a signal was detected and replaced values with the LOD/sqrt (2) when no signal was detected (Hornung and Reed, 1990; Lubin Jay et al., 2004).

2.4. Data analyses

We calculated descriptive statistics to summarize baseline participant characteristics and phthalate biomarker concentrations for all visits. Concentrations for biomarkers with a detection frequency (DF) ≥70% were right-skewed and log₁₀-transformed in subsequent analyses. Biomarkers included in the analyses were modeled as continuous variables (no differences were observed when MCINP was modeled as a dichotomous or continuous variable, DF = 69%). We assessed variability of biomarker concentrations by calculating intraclass correlation coefficients (ICC) using mixed-effects models. We evaluated individual DEHP biomarkers (MEHP, MEHHP, MEOHP, MECPP) as the concentrations of individual phthalate biomarkers as well as their combined molar sum (ΣDEHP) as follows: ΣDEHP (μmol/L) = C_MEHP (ng/mL) * (1/278.34) + C_MEHHP (ng/mL) * (1/294.34) + C_MEOHP (ng/mL) * (1/292.33) + C_MECPP (ng/mL) * (1/308.33); where C_MEHP (ng/mL), for example, is the individual’s urinary concentration of MEHP in ng/mL, and each value in the denominator represents the molecular weight in g/mol for each respective DEHP biomarker (MEHP, MEHHP, MEOHP, MECPP, respectively).

To evaluate associations between phthalates and our target outcomes, we used generalized estimating equations (GEE) to account for the panel study design. We conducted both, crude and multivariable binomial regression models to examine associations between repeated phthalate exposures and binary outcome measures (e.g., symptoms, medication use, healthcare utilization) and multivariable linear regressions for continuous measures (e.g., FEV1/FVC%, FeNO).

We identified covariates based on the literature and using a directed acyclic graph, and included confounders associated with the exposure and with pediatric asthma morbidity that were not on the causal pathway. Main model covariates included: age (in years), sex, race/ethnicity (Black vs. Other), caregiver’s education level (less than high school, high school graduate, and some college or more), presence of smokers in the home, and season. Caregiver’s education level was included as a measure of socioeconomic status (SES) as information on household income was missing on 9% of study participants (inclusion of income level as a covariate in models did not alter our results). Lastly, because many phthalates are known endocrine disrupting compounds and prior studies have reported sexually dimorphic effects in the relationship between phthalates and asthma (Meeker, 2012; Wang et al., 2019) we assessed effect measure modification by sex by including an interaction term and stratifying by sex in separate models.

2.5. Sensitivity analyses

To determine the robustness of our results, we conducted additional sensitivity analyses by including select covariates in separate models. First, we accounted for co-exposure to bisphenol A (BPA) by controlling for BPA biomarker concentrations in our models since we previously observed significant positive associations with asthma symptoms in the present cohort (Quirós-Alcalá et al., 2021). We also considered indoor concentrations of nitrogen dioxide (NO₂) and fine particulate matter (PM_2.5, particulate matter with an aerodynamic diameter <2.5 μm), measured in the participants’ homes within ±2 weeks of each clinic visit. In addition, because obesity has been linked to phthalate exposure and is reported to play a role in asthma control and exacerbations (Meeker, 2012; Wang et al., 2019), we re-ran main models including children’s age- and sex-standardized body mass index (BMI) percentiles (US Centers for Disease Control and Prevention, 2021) modeled as a continuous covariate. Additionally, we included nebulizer use as a covariate since one study reported that nebulizer medications may contain low levels of phthalates (Kwak et al., 2010). We ran models further controlling for degree of atopy, a recognized risk factor for asthma morbidity (Arasi et al., 2019) by including the total number of positive skin prick tests as a continuous covariate to reflect degree of allergic sensitization. Lastly, we ran additional models controlling for mouse allergen sensitization (coded as yes/no to indicate if the child was allergic to mice) as we previously identified this to be the allergen that drives asthma outcomes in this population (Ahluwalia et al., 2013; Torjusen et al., 2013).

Analyses were performed with Stata/IC 14.2 software (StataCorp, College Station, Texas); and statistical significance criteria was set at p < 0.05 and p < 0.10 for main effects and effect measure modification, respectively.

Given the exploratory nature of our study, we did not include a multiple comparison adjustment method in our analyses. Instead, we focused on examining the outcome measures for the emergence of consistent patterns and interpreting p-values for statistical significance in the context of all results (Althouse, 2016; Rothman, 1990).

3. Results

3.1. Study population characteristics

The study participants had a mean age of 11.2 (SD:4.0) years and approximately 57% were male (Table 1). Children were predominantly Black (91%), most had an average annual household income <$35,000 (69%) and were receiving government-based health insurance (86%). More than half of the caregivers (63%) reported not having a college education. Most of the children (91%) had more than one positive skin prick test response (Table 2) and reported having an acute health care visit for asthma in the previous three months (93%); with 68% reporting an ED visit for asthma during this period. Additional clinical characteristics are presented in Table 2. When we evaluated whether baseline and clinical characteristics differed by sex, we found that, on average, girls were slightly older (mean age 12, SD 4 for females vs 11, SD 4 years for males) and had a higher BMI compared to boys (mean BMI index 76, SD 29 for females vs 62, SD 31 for males).

Table 1.

Baseline demographic characteristics for MAACS children ages 5–17 years (n=148).^a

Demographic Characteristic
Sex
	Male, n (%)	85 (57)
	Female, n (%)	63 (43)
Age b	mean (SD)	11 (4)
	5–7 years, n (%)	41 (28)
	8–10 years, n (%)	35 (24)
	11–13 years, n (%)	26 (18)
	14–17 years, n (%)	46 (31)
Child’s race
	Black/African American, n (%)	135 (91)
	Other c, n (%)	13 (9)
Body mass index (BMI)	mean (SD)	22 (8)
Underweight (<5th percentile), n (%)		6 (4)
Normal Weight (5th–<85th percentile), n (%)		77 (53)
Overweight (85th–<95th percentile), n (%)		22 (15)
Obese (>95th percentile), n (%)		41 (28)
Caregiver’s Education Level
	Less than high school, n (%)	42 (28)
	High school graduate, n (%)	51 (35)
	Some college or more, n (%)	55 (37)
Annual household income
	<$35,000, n (%)	92 (69)
	>$35,000, n (%)	42 (31)
Insurance
Private/self-pay, n (%)		20 (14)
Government-based (public), n (%)		126 (86)
Smoker lives in child’s home
	No, n (%)	64 (43)
	Yes, n (%)	84 (57)

^aInformation reported is for 148 MAACs children with biomarker exposure data. Information was missing for select demographic characteristics at baseline (BMI: n = 2; annual household income: n = 14; health insurance: n = 2).

^bThe mean (SD) age of MAACS children was 11.2 (4.0) years.

^c“Other” race category includes White (n = 6); American Indian or Alaska Native (n = 5); Other Pacific Islander (n = 1); and Unknown (n = 1).

Table 2.

Baseline clinical characteristics for MAACS children ages 5–17 years (n = 148).^a

Clinical Characteristic
Allergic sensitization characteristics	n (%)
Atopic (≥ 1 Positive skin prick test response)	134 (91)
Skin test sensitivities
Cat	96 (65)
Cockroach	91 (62)
Dust mite	85 (58)
Mouse	78 (53)
Dog	26 (18)
Total IgE (kU/L) a Median (p25, p75)	190 (55.7, 458)
Asthma symptoms in the prior 2 weeks	n (%)
Number of days with symptoms	[0–4]	[5–9]	[10–14]
Coughing, wheezing, or chest tightness	114 (77)	18 (12)	16 (11)
Nocturnal wakening with symptoms	126 (88)	14 (10)	4 (3)
Exercise-related symptoms	129 (88)	8 (5)	9 (6)
Cough without a cold	118 (85)	8 (6)	13 (9)
Slowed activity due to asthma b	124 (84)	13 (9)	11 (7)
Maximal symptom days	103 (70)	23 (16)	22 (15)
Short-acting Beta Agonist (SABA) use	101 (68)	17 (11)	30 (20)
Slowed speech due to asthma	147 (99)	1 (1)	0 (0)
Lung function and inflammation a
FEV1 (% predicted) Mean (SD)	91.3 (15.6)
FVC (% predicted) Mean (SD)	100.2 (13.5)
FEV1/FVC (%) Mean (SD)	80.6 (9.6)
Bronchodilator reversibility c n/total (%)	37/130 (28.5)
FeNO (ppb) Median (p25, p75)	33 (16.0, 62.0)
Asthma-related health care utilization in the prior 3 months	n (%)
Acute health care visit d			93 (63)
ED visit	68 (46)
Hospitalization	14 (10)
Unscheduled doctor’s office visit	35 (24)
Controller medication use e	106 (72)

Abbreviations: FEV1: Forced exhaled volume in the first second; FVC: Forced vital capacity; FeNO: fractional exhaled nitric oxide; ED visit: Emergency department visit for asthma; p25: 25th percentile; p75: 75th percentile.

^aInformation was missing on some study participants (atopic status and skin sensitivities: n = 1; IgE: n = 3; FVC % predicted: n = 16; FEV1% predicted: n = 16; FEV1/FVC%: n = 16; bronchodilator reversibility: n = 18; FeNO: n = 17).

^bSlowed activity due to asthma: child had to slow down or stop activities while at home or playing with other children because of asthma, wheezing, tightness in the chest or cough.

^cBronchodilator reversibility is defined as an increase in FEV1 of >12% following treatment with a short-acting β-agonist or SABA.

^dAcute health care visits consists of a composite measure of any unscheduled healthcare visit for asthma-related symptoms (e.g., ED visits, hospitalization, and/or unscheduled doctor’s visits for asthma).

^eInhaled corticosteroid or leukotriene inhibitor.

3.2. Phthalate biomarker concentrations

As shown in Table 3, 12 phthalate biomarkers were detected above their respective limits of detections in >80% of the samples (MECPP, MEHHP, MEHP, MEOHP, MBzP, MCIOP, MCPP, MECPTP, MEHHTP, MEP, MiBP, MnBP) (summary statistics for uncorrected concentrations are presented in the Supplemental Information, Table E1). MCINP was detected in 69% of the urine samples and MINP was detected in <12% of samples. The phthalate replacements MHNCH and MCOCH were detected in 38% and 1.7% of the samples, respectively. Overall, biomarker concentrations varied greatly within children over the follow-up period (ICCs: 0.11–0.45). Girls had significantly higher MEP biomarker concentrations compared to boys; no other sex differences were noted.

Table 3.

Summary statistics and variability measures for specific gravity-corrected phthalate biomarker concentrations and phthalate replacements in ng/mL based on the total number of urine samples collected on 148 MAACS participants (n = 650 samples).

Biomarker	Specific gravity-corrected biomarker concentrations, ng/mL								Measure of biomarker variabilityb
	LOD (ng/mL)	%>LOD	GMa (GSD)	Min	Median (p25, p75)	Max	Boys GMa (GSD)	Girls GMa (GSD)	ICC (95%CI)
Σ DEHPc		–	0.5 (2.7)	0.003	0.4 (0.2, 0.8)	11.8	0.4 (2.6)	0.5 (2.7)	0.18 (0.12, 0.27)
MECPPd	0.2	99.8	40.3 (2.5)	<LOD	37.3 (22.5, 67)	983.2	37.6 (2.5)	44.6 (2.5)	0.21 (0.14, 0.30)
MEHHPd	0.1	99.8	72.8 (2.8)	0.1	69.7 (39.6, 130.6)	2036.1	68.8 (2.8)	79.3 (2.8)	0.18 (0.11, 0.27)
MEHP	1	80.2	3 (3.3)	<LOD	2.9 (1.5, 6)	179.0	2.7 (3.2)	3.6 (3.4)	0.29 (0.21, 0.38)
MEOHPd	0.1	99.8	15.4 (2.8)	0.1	14.8 (8.4, 27.3)	442.1	14.5 (2.8)	16.8 (2.9)	0.18 (0.11, 0.27)
MBzPd	0.2	99.8	22.9 (3.3)	<LOD	21.4 (10.3, 48.8)	1573.3	23.9 (3.4)	21.3 (3.2)	0.45 (0.37, 0.53)
MCINP	0.2	69.3	0.4 (2.6)	<LOD	0.3 (0.2, 0.6)	17.0	0.4 (2.7)	0.4 (2.6)	0.13 (0.07, 0.22)
MCIOP	0.2	97.4	1.9 (2.7)	<LOD	1.7 (1, 3.4)	70.4	1.9 (2.8)	2 (2.5)	0.11 (0.06, 0.20)
MCPP	0.2	99.2	2.8 (2.3)	<LOD	2.6 (1.7, 4.7)	37.8	2.8 (2.3)	2.9 (2.2)	0.22 (0.15, 0.32)
MECPTPe	0.2	96.5	1.4 (2.8)	<LOD	1.2 (0.7, 2.5)	226.3	1.3 (2.8)	1.5 (2.7)	0.27 (0.19, 0.36)
MEHHTPe	0.2	81.6	0.6 (2.8)	<LOD	0.6 (0.3, 1.2)	14.6	0.6 (2.8)	0.7 (2.8)	0.15 (0.09, 0.24)
MHNCHe	0.2	38.1	0.2 (3.7)	<LOD	0.2 (<LOD, 0.4)	39.8	0.2 (3.6)	0.2 (3.7)	–
MCOCHe	0.2	11.1	<LOD	<LOD	<LOD	13.3	< LOD	<LOD	–
MINP	0.5	1.7	<LOD	<LOD	<LOD	6.1	0.4 (1.8)	0.4 (1.6)	–
MEPd	1	99.8	103.2 (3.1)	<LOD	92.3 (44.5, 216.3)	4876.8	85.1 (3)f	138 (3)f	0.24 (0.17, 0.34)
MiBP	0.1	99.8	17.5 (2.4)	0.1	18.8 (10.3, 30.1)	210.8	16.2 (2.3)	19.6 (2.4)	0.21 (0.14, 0.30)
MnBP	0.5	99.8	31.8 (2.4)	<LOD	33.5 (19.4, 54.8)	395.4	30 (2.3)	34.7 (2.4)	0.24 (0.16, 0.33)

Abbreviations: LOD-Limit of detection; %LOD-Percent of samples with concentrations above the limit of detection for each specified analyte; GM-geometric mean; GSD: geometric standard deviation; 95%CI: 95% confidence interval; ICC: Intraclass correlation coefficient.

^aGeometric Mean (GM) values reported are based on detectable concentrations, including machine read values, for each biomarker.

^bICC values and respective 95% confidence intervals are based on data from children who provided 2 or more samples during the 12-month follow-up period. ICCs are not reported for MHNCH, MCOCH and MINP due to low detection frequency (<40%).

^cDEHP was estimated as the molar sum of DEHP-devolving biomarkers in nMol/mL: Σ DEHP = [(MEHP*(1/278.34)) + (MEHHP*(1/294.34)) + (MEOHP*(1/292.33)) + (MECPP*(1/308.33))]. Values in the denominators reflect individual DEHP-devolving biomarker molecular weights.

^dSome analyses were repeated with diluted sample, LOD for diluted samples: MBzP 5 samples (1%); LOD: 1 ng/mL; MECPP 3 samples (0.5%); LODs: 1 ng/mL, 2 ng/mL; MEHHP 12 samples (2%); LODs: 0.5 ng/mL, 1 ng/mL; MEOHP 2 samples (0.3%); LODs: 0.5 ng/mL, 1 ng/mL; MEP 3 samples (0.5%); LOD: 5 ng/mL.

^eParent compounds used as replacements of DEHP (MECPTP and MEHHTP are biomarkers of DEHTP; MHNCH and MCOCH are biomarkers of DINCH).

^fConcentrations were statistically significantly higher among girls compared to boys (p < 0.001; Student t-test).

3.3. High molecular weight phthalates, asthma symptoms and healthcare utilization

We observed consistent positive significant associations of individual DEHP biomarkers, the molar sum of DEHP (ΣDEHP), and MBzP (biomarker of BBzP) with increased odds of asthma symptoms and healthcare utilization. Multivariable model results for asthma outcome measures are presented in Figs. 1 and 2 (details and comparisons on the crude and adjusted effect estimates are available in Tables E3–E6 in this article’s Online Repository). Specifically, we observed increased odds of general symptoms (coughing, wheezing or chest tightness) with ΣDEHP, and MBzP (adjusted Odds Ratio, aOR, ΣDEHP: 1.49, 95% Confidence Interval, CI: 1.08–2.07; MBzP: 1.34, CI: 1.04–1.73). Consistent and significant positive associations were also observed for maximal symptom days (aOR ΣDEHP: 1.51, CI: 1.11–2.07; MBzP: 1.32, CI: 1.01–1.72), slowed activity due to asthma (aOR ΣDEHP: 1.64, CI: 1.11–2.44; MBzP: 1.25, CI: 0.94–1.67), and exercise related symptoms (aOR ΣDEHP: 1.59, CI: 1.03–2.46; MBzP: 1.24, CI: 0.92–1.66). We observed consistent positive associations for ΣDEHP, individual DEHP biomarkers, and MBzP with cough without a cold and slowed speech due to asthma (Fig. 1 and Table E3).

Fig. 1.

Forest plots of adjusted associations between repeated urinary high molecular weight phthalate biomarkers and asthma symptoms among 148 children 5–17 years old participating in MAACS (n = 650 samples).

Multivariable models were adjusted for age, sex, race, caregiver education, season, and presence of smokers in the home as a proxy for environmental tobacco smoke exposure (n = 148). Asthma symptom variables were treated as dichotomous outcomes using binomial regression models with generalized estimating equations and log link to obtain ORs; phthalate biomarker concentrations were modeled as continuous variables (log-10 transformed concentrations). DEHP was estimated as the molar sum of DEHP-devolving biomarkers in nMol/mL: Σ DEHP = [(MEHP*(1/278.34)) + (MEHHP*(1/294.34)) + (MEOHP*(1/292.33)) + (MECPP*(1/308.33))]. Abbreviations: OR=Odds ratio; aOR = Adjusted OR; 95%CI: 95% Confidence Interval.

Fig. 2.

Forest plots of adjusted associations between repeated urinary high molecular weight phthalate biomarkers and healthcare utilization among 148 children 5–17 years old participating in MAACS (n = 650 samples).

Consistent with findings observed for asthma symptoms, we observed positive associations with a 10-fold increase in ΣDEHP and MBzP biomarker concentrations, and health care utilization in the previous three months (Fig. 2 and Table E4). Specifically, a 10-fold increase in ΣDEHP concentrations was associated with 32% increased emergency department visits (aOR ΣDEHP 1.32, CI: 0.87–1.99), 47% increased odds in acute care visits (aOR ΣDEHP 1.47, CI: 1.02–2.11) and 51% increased odds in unscheduled doctor visits (aOR ΣDEHP 1.51, CI: 0.94–2.43). A 10-fold increase in MBzP concentrations was also associated with increased odds of emergency department visits, increased acute care visits and unscheduled doctor visits (aOR 1.24, CI: 0.83–1.86; aOR 1.26, CI: 0.88–1.81; aOR 1.41, CI: 0.95–2.10, respectively).

Additionally, we observed significant positive associations with concentrations of MEHHTP, a biomarker of DEHTP, which is a structural isomer or replacement of DEHP, and most asthma symptoms evaluated (Fig. 1). Specifically, we observed associations for MEHHTP with general symptoms (aOR: 1.39, CI: 1.02–1.90); maximal symptom days (aOR: 1.42, CI: 1.04–1.94); slowed activity due to asthma (aOR: 1.62, CI: 1.12–2.35) and exercise related symptoms (aOR: 1.56, CI: 1.06–2.28). Consistent positive associations were observed for MEHHTP and additional symptom variables evaluated (Fig. 1 and Table E3). We did not observe consistent associations with other high molecular weight phthalate biomarkers.

3.4. Low molecular weight phthalates asthma symptoms and healthcare utilization

Biomarker concentrations of MiBP and MnBP (biomarkers of DiBP and DBP) were positively associated with general asthma symptoms (coughing, wheezing or chest tightness), albeit only among boys (coughing, wheezing or chest tightness: MiBP: aOR 1.47, CI:1–2.16, p-value of interactions 0.013, MnBP: aOR 1.54, CI:1.05–2.25, p-value of interactions 0.01; Table E5). We also observed positive associations between MiBP and MnBP concentrations and emergency department visits (MiBP: 1.73, CI:1.01–2.96; MnBP 1.6, CI: 0.94–2.73; Table E6). The effect estimates of these associations were generally larger for boys compared with girls (Table E7).

3.5. Associations with lung function and lung inflammation

Overall, we did not observe associations with lung function measures (FEV1, FVC percent predicted, FEV1/FVC%) or FeNO (Tables E8–E11).

3.6. Sex differences and sensitivity analyses

Overall, we did not observe clear consistent patterns indicating sex differences across most outcomes (Tables E7, E9, E12–E14). However, some associations had larger effect estimates among boys than girls for select phthalates (specifically MBzP, MiBP, MnBP) and most outcomes evaluated (e.g., asthma symptom variables, ED visits, acute care visits, and unscheduled doctor visits). We confirmed that the results in the main models were robust to controlling for other factors previously shown to be associated with our target exposures and/or outcomes, including BPA, PM_2.5, NO₂, BMI, degree of atopy, and mouse allergen sensitization (results not shown). Lastly, when additionally controlling our main models for nebulizer use, some attenuation of effect estimates was observed (change in effect estimates <10%; results not shown); however, the positive associations observed persisted.

4. Discussion

In the present study, we examined associations between exposures to phthalates and newer phthalate replacements and pediatric asthma morbidity, including respiratory symptoms, healthcare utilization, as well as pulmonary function and inflammation in a cohort of predominantly Black children with established asthma. Overall, we observed strong and consistent positive associations with two high molecular weight phthalates (specifically, DEHP biomarkers/molar sum, and MBzP) with asthma symptoms and healthcare utilization. We also observed consistent associations with asthma symptoms and MEHHTP (a biomarker of DEHTP, a commonly used DEHP replacement). Among low molecular weight phthalates, we observed positive associations with MiBP and MnBP with select healthcare utilization variables. While no associations were observed with hospitalizations potentially due to the low frequency (<10%) of children reporting hospitalizations, we still observed consistent associations with emergency department visits, acute care visits, and unscheduled doctor’s visits. We also observed suggestive evidence of stronger associations in boys for select biomarkers (MBzP, MiBP, MnBP), overall, we did not observe strong consistent evidence of sexually dimorphic effects across phthalate and phthalate replacement biomarkers and asthma symptoms.

Consistent with our findings, other studies also report higher concentrations of phthalate biomarkers among Black children compared to children in the U.S. general population (Silva et al., 2004; US EPA, 2015). Geometric mean concentrations for high molecular weight phthalate biomarker concentrations in our predominantly low-income cohort were generally 2–4 times higher than among children in the U. S. general population in urine samples collected around the same time period (2007–2012 National Health and Nutrition Examination Survey, NHANES, cycle years) (Odebeatu et al., 2019). Differences across race/ethnicity may be due to different dietary consumption patterns and consumer product uses. For example, children’s exposure to high molecular weight phthalates, especially DEHP, in the U.S. is primarily thought to originate from diet (foods and drinks packaged in plastic) and materials made of PVC (Hammel et al., 2019; Smith et al., 2021; Varshavsky et al., 2018). High levels of DEHP biomarkers have also been detected in individuals who report frequently consuming fast foods; this association has been reported to be more pronounced in populations of color in the U.S. (Zota Ami et al., 2016). Consistent with prior studies (Berger et al., 2019), MEP concentrations were higher in girls compared to boys likely due to higher personal care product use among girls. MEP is a biomarker for DEP which is primarily used in personal care products (PCPs) and higher urinary MEP concentrations have been associated with greater use of cosmetics and PCPs, particularly among adolescent females (Berger et al., 2019; Lewis et al., 2013). MBzP, the primary biomarker of BBzP, is used in toys, PVC materials, personal care products, and food packaging (Benjamin et al., 2017; Braun et al., 2013; Odebeatu et al., 2019).

Leaching of phthalates from medications and from plastics in medical delivery devices may increase phthalate exposures, but their contribution is suspected to be low. A previous study that evaluated phthalate exposures in children with asthma reported that asthma medication use in the prior 14 days (leukotriene agonists, β-2 agonists, and inhaled corticosteroids) was not associated with urinary phthalate biomarkers (Bertelsen et al., 2013), suggesting that asthma medication plays a minimal role compared to other sources for these chemicals. Although detectable, phthalate concentrations in several albuterol aerosol medications have been reported to be very low (Kwak et al., 2010). Consistent with this observation, when controlling for nebulizer use we found that although effect estimates became attenuated, positive associations persisted. Still, it remains unclear whether these exposure levels play an appreciable role in total phthalate exposure or in increasing the risk of asthma outcomes. However, we cannot rule out the possibility that asthma symptoms may cause increased use of nebulizers which may contaminate medications with low levels of phthalates.

While the current study focused on asthma morbidity, including asthma-related symptoms and healthcare utilization among those with existing asthma, it is worth noting that prior studies on asthma prevalence and incidence have reported associations with the same high molecular weight phthalate biomarkers for which we observed significant associations, DEHP and MBzP. High molecular weight phthalates have been previously associated with prevalence of pediatric allergic disease including asthma (Benjamin et al., 2017; Braun et al., 2013; Li et al., 2017; Wang et al., 2019). A cross-sectional study using the NHANES data also reported a positive association between MBzP and current asthma in children ages 6–17 years (Odebeatu et al., 2019), while another NHANES study reported a positive association between exposure to MBzP and current asthma, wheeze, hay fever, and rhinitis in adults (>18 years), but not among children (aged 6–17 years, n = 779) (Hoppin et al., 2013). Our study overcomes several of the limitations that studies using NHANES data face such as their cross-sectional nature, their inability to capture in-depth information on asthma symptoms and healthcare utilization using comprehensive questionnaires, and lack of measures on key co-exposures known to impact asthma such as air pollution. DEHP and MBzP have been reported to be associated with asthma incidence and allergy symptoms in children in several locations around the world including Taiwan (Hsu et al., 2012), Sweden (Bornehag et al., 2004), China (Shi et al., 2018), and the U.S. (Odebeatu et al., 2019). These biomarkers have also been linked with allergic diseases, including asthma and eczema in children when exposure occurred prenatally (Braun et al., 2013; Li et al., 2017). Nonetheless, as in any epidemiologic study, we cannot rule out the possibility of unmeasured confounding in prior studies as well as the present study.

While we did not observe consistent associations with lung function, a study conducted by Kim et al. reported associations between longitudinal measures of phthalate biomarkers (MEHHP, MEOHP, and MnBP) with lung function (peak expiratory flow rate) among 56 Korean children with asthma (Kim et al., 2018). Authors reported that phthalate exposure might have a delayed effect on adverse pulmonary function of one day (Kim et al., 2018); however, we assessed lung function the same day of urine sample collection limiting our ability to assess short-term lagged effects in our study. Notably, concentrations among children in Korea are reported to exceed concentrations among children from the U. S. general population (Kim et al., 2018; Song et al., 2013).

We also did not observe clear associations between phthalate and phthalate replacement biomarkers and eosinophilic pulmonary inflammation as measured by FeNO (Vestergaard et al., 2015), suggesting that asthma exacerbations in the study population might not occur through eosinophilic inflammation. Although two previous studies have reported positive associations with FeNO in children, phthalate biomarkers implicated were inconsistent across these two studies. Kim et al. reported associations of lung inflammation with DEHP biomarkers (MEHHP, MEOHP), but not MnBP. Contrary to this finding, a larger study conducted among 244 U.S. children ages 6–9 years with seroatopy or wheeze reported that MnBP and MBzP biomarker concentrations were positively associated with lung inflammation, but not with DEHP biomarkers (Just et al., 2012). Although results from animal models assessing the effect of phthalates on airways suggest that eosinophilic inflammation may pay a role, other studies have implicated other cell and cytokine profiles suggesting other potential pathways (Bølling et al., 2020). The mechanism through which phthalates are associated with asthma symptoms remains unclear and existing evidence is limited regarding the effects of phthalates other than DEHP (Bølling et al., 2020). Further studies evaluating a wide range of phthalates are needed to clarify the impact of their exposure on asthma symptoms and elucidate mechanistic pathways.

In response to growing health concerns with select phthalates, including DEHP, replacement phthalates are increasingly being used in consumer products (e.g., DINCH, DEHTP). Despite our sample collection being conducted between 2007 and 2009 when concentrations of phthalate replacement biomarkers were lower than those reported in more recent studies (Silva et al., 2019), we still observed significant positive associations with MEHHTP, biomarker for DEHTP, and asthma symptoms. Coincidentally, DEHTP is used in consumer products to replace phthalates including DEHP for which we observed consistent positive associations with the target outcomes in the present study. Further studies are warranted to identify the impact of replacement phthalates on respiratory health, especially given that concentrations in the study population do not likely reflect current exposures to these newer phthalates.

Because phthalates can disrupt endocrine processes, for example, by mimicking sex hormones, their health effects may differ by sex (Chen et al., 2014; Grindler et al., 2018). Although we found stronger associations among boys with asthma symptoms and select phthalate biomarkers (MBzP, MiBP, MnBP), we did not find clear evidence across outcomes for sexually dimorphic effects. However, previous studies have reported sexually dimorphic effects with stronger associations among boys and asthma for MEP and DEHP biomarkers and asthma development (Ku et al., 2015; Odebeatu et al., 2019). Our study, however, may be underpowered to examine such effects. It has also been observed that pubertal status can affect asthma exacerbation patterns (Fuchs et al., 2017). Thus, it is plausible that results could also differ by pubertal stage given that phthalates are endocrine disrupting compounds and that asthma prevalence is reported to differ by pubertal stage. Our small sample size and lack of objective pubertal status measures limited our ability to further stratify by pubertal status among all children and by sex. Although the evidence is mixed regarding sexually dimorphic effects of phthalates on asthma, several potential mechanisms of such effects have been proposed. Differences in phthalate-asthma relationships by sex could simply reflect well described sex differences in asthma prevalence and phenotype (Carey et al., 2007; Chang and Mitzner, 2007). It is also postulated that gene and environment interactions with phthalate exposure may result in sex-specific differences and susceptibility to asthma respiratory-related symptoms among boys (Odebeatu et al., 2019). Further studies are warranted to assess whether sex differences influence the associations between phthalates and asthma respiratory-related outcomes and to further examine how chemical exposures impact asthma based on pubertal stage.

The present study has several limitations. First, the sample size of the study may have limited our ability to identify sex differences within pubertal stages on the impact of phthalates. Still, our findings with DEHP and MBzP biomarkers are consistent with other respiratory health studies (Benjamin et al., 2017; Braun et al., 2013; Li et al., 2017; Wang et al., 2019). We were also not able to account for and prospectively collect information on potential phthalate or phthalate replacement sources in the study population given that the analyses were conducted after all data collection concluded using stored biospecimens. While the overall aim of the study was to examine associations between phthalate exposures and asthma symptoms and healthcare utilization, it will be important to understand the major exposure sources of these chemicals (e.g., fast food consumption, food packaging, flooring materials, personal care product use) in populations that suffer disproportionally high phthalate exposures and asthma morbidity burden to inform targets for intervention. We were also not able to assess lagged effects of the exposures on lung function since samples and spirometry measurements were collected during the same study visit. We can also not dismiss the possibility of unmeasured confounding or reverse causation. For example, we cannot dismiss the plausibility that increased use of nebulizers may contaminate medications with low levels of phthalates. However, when controlling for nebulizer use, we found that positive associations persisted. Still, it remains unclear whether these exposures play an appreciable role in the risk of exacerbations. Another limitation is potential generalizability of our findings to other pediatric populations of other racial/demographic groups. However, we note that these results may still be relevant to Black children with asthma living in urban settings with similar exposures. Additionally, consistent findings with the same phthalate parent compounds across other studies of respiratory related outcomes suggests that our findings may also be relevant to other pediatric populations with asthma. Phthalates also have a relatively short biological half-life. To overcome this limitation and account for potential variability within individuals, we collected repeated urine samples. Nonetheless, it is plausible that exposures measured may not truly represent those occurring during periods reflected by the measured outcomes (e.g., symptoms in the prior two weeks and healthcare utilization in the prior three months) leading to nondifferential misclassification and attenuated effect estimates. Finally, some results might have been influenced by chance alone, however we still identified consistent patterns in our findings across different exposure/outcome relationships beyond just statistical significance across the same biomarkers and consistency with other studies.

Despite the limitations noted, the study has several strengths. First, the study was among the first conducted to evaluate asthma symptoms and healthcare utilization on a predominantly low-income, Black pediatric cohort, an underrepresented and understudied population disproportionally affected by phthalate exposures and asthma. Second, to our knowledge, only one other study has examined the role of phthalates on asthma symptoms. The repeated measures panel design is another major study strength that allowed for improved exposure and outcome characterization, relevant for chemicals with relatively short biological half-lives such as phthalates. We also measured several phthalate biomarkers, including newer phthalate replacements, and identified consistency of associations between several respiratory-related measures and several phthalate biomarkers. The consistency of our associations across several of our outcomes and the same biomarkers increases our confidence that the observed results may not due to chance. This study is also among the first to characterize temporal variability of phthalate concentrations in children. Additionally, we assessed the robustness of the findings by controlling for important confounders (e.g., race, education, season, smoking exposure), co-exposures to other contaminants (e.g., PM_2.5, NO₂, BPA) and other potential recognized outcome risk factors (e.g., BMI, mouse sensitization), allowing us to assess the independent effects of phthalate biomarkers on several outcome measures.

5. Conclusions

In summary, our results suggest that exposure to select high molecular weight phthalates (particularly DEHP, BBzP and DEHTP, a DEHP phthalate replacement), and select low molecular weight phthalates (DiBP and DBP) among a low-income cohort of predominantly Black children with asthma may be associated with increased asthma symptoms and health care utilization. Given the widespread use of phthalates in consumer products and disparate exposures reported by race/ethnicity, findings warrant replication in larger multiracial/multiethnic cohorts given the potential implications for asthma disparities and interventions to reduce disparities. The growing epidemiologic evidence implicating phthalate exposure, especially to high molecular weight phthalates, in incident asthma, asthma symptoms and healthcare utilization, as well as lack of information on phthalate replacements supports the need for further research. Specifically, cohort studies, mechanistic studies, and animal model studies, which contribute to an assessment of the likelihood that these relationships are causal will be important for advancing our understanding of the potential impact of phthalates and their replacements on respiratory health.

Abbreviations

aOR

Adjusted odds ratio

BMI

Body mass index

BPA

Bisphenol A

CDC

Centers for Disease Control and Prevention

CI

Confidence Interval

DF

Detection frequency

ED

visit Emergency department visit

FEV1

Forced exhaled volume in the first second

FeNO

Fractional exhaled nitric oxide

FVC

Forced vital capacity

GEE

Generalized estimating equation

GM

Geometric mean

GSD

Geometric standard deviation

HMW

High molecular weight

ICC

Intraclass correlation coefficient

LMW

Low molecular weight

LOD

Limit of detection

MAACS

Mouse Allergen and Asthma Cohort Study

NHANES

National Health and Nutrition Examination Survey

OR

Odds ratio

PCPs

Personal care products

PVC

Polyvinyl chloride

SABA

Short-acting beta agonist

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.envres.2022.113239.