Residential radon decay products are associated with cough and phlegm in patients with COPD

Abstract

Radon decay products attach to particulate matter (referred to as particle radioactivity, PR) has been shown to be potential to promote airway damage after inhalation. In this study, we investigated associations between PR with respiratory symptoms and health-related quality of life (HRQL) in patients with COPD. 141 male patients with COPD, former smokers, completed the St. George’s Respiratory Questionnaire (SGRQ) after up to four 1-week seasonal assessments (N=474) of indoor (home) and ambient (central site) particulate matter ≤ 2.5 μm in diameter (PM_2.5) and black carbon (BC). Indoor PR was measured as α-activity (radiation) on PM_2.5 filter samples. The ratio of indoor/ambient sulfur in PM_2.5 (a ventilation surrogate) was used to estimate α-PR from indoor radon decay. SGRQ responses assessed frequent cough, phlegm, shortness of breath, wheeze, and chest attacks in the past 3 months. Multivariable linear regression with generalized estimating equations accounting for repeated measures was used to explore associations, adjusting for potential confounders. Median (IQR) indoor α-PR was 1.22 (0.62) mBq/m³. We found that there were positive associations between α-PR with cough and phlegm. The strongest associations were with estimated α-PR of indoor origin for cough (31.1 % increase/IQR, 95 %CI: 8.8 %, 57.8 %), and was suggestive for phlegm (13.0 % increase/IQR, 95 %CI: −2.5 %, 31.0 %), similar adjusting for indoor BC or PM_2.5. α-PR of indoor origin was positively associated with an increase in SGRQ Symptoms score [1.2 units/IQR; 95 %CI: −0.3, 2.6] that did not meet conventional levels of statistical significance. Our results suggested that exposure to indoor radon decay products measured as particle radioactivity, a common indoor exposure, is associated with cough, and suggestively associated with phlegm and worse HRQL symptoms score in patients with COPD.

1. Introduction

Exposure to indoor sources of radon gas is common, and it is estimated that 1 in 15 homes in the United States have elevated levels (EPA, 2018). Radon originates from the decay of uranium in soil and rock. Indoor radon is mainly attributable to its infiltration through cracks, joints, and gaps in floors and walls, but small amounts may be emitted from stone or rock based building materials. Radon gas decays into radionuclides that attach to airborne particulate matter (PM) and emit α-, β-, and γ-radiation, referred to as particle radioactivity (PR), that has been associated with reduced pulmonary function, and increases in biomakers of oxidative stress and systemic inflammation in patients with chronic obstructive pulmonary disease (COPD) (Huang et al., 2020, 2021; Vieira et al., 2019; Wang et al., 2023). However, to the best of our knowledge, there has been no investigation about associations between PR with respiratory symptoms in patients with COPD.

COPD is the third leading cause of death worldwide, causing 3.23 million deaths in 2019 (WHO, 2021). In 2019, 4.6 % of adults in the US reported being diagnosed with COPD, which is responsible for millions of physician office visits and emergency department visits (Xu et al., 2022; CDC, 2012, 2015). Respiratory symptoms in patients with COPD are important clinically in determining treatment (GOLD, 2023). These symptoms are highly associated with a clinically meaningful reduction in health-related quality of life (HRQL), overall health status and prognosis (Wedzicha et al., 2014; Miravitlles and Ribera, 2017).

Therefore, in the present study we explore associations between indoor PR with common respiratory symptoms and HRQL scores based on the St. George’s Respiratory Questionnaire (SGRQ) in a cohort study that included 141 male patients with COPD from Eastern Massachusetts, USA. Since α-radiation is more harmful than β- and γ-radiation due to its high energy (ATSDR, 1999), and most of α-PR is from PM_2.5 (Liu et al., 2020), we measured α-radioactivity on PM_2.5 filter samples collected in the homes of the patients. We considered not only total indoor α-PR, but also estimated concentrations of indoor-originated and ambient-infiltrated α-PR and associations with respiratory symptoms and HRQL.

2. Material and methods

2.1. Study population

Recruitment of the cohort has previously been described (Busenkell et al., 2022; Wang et al., 2023). Individuals age ≥ 40 years old, not current smokers, with a post-bronchodilator forced expiratory volume in 1 second (FEV₁) to forced vital capacity (FVC) ratio (FEV₁/FVC) of less than 0.7 or emphysema based on a clinical CT scan report living in Eastern Massachusetts and vicinity were recruited at Veterans Affairs Boston Healthcare System between November 2012 and April 2017 by mail and phone from clinics, and by flyer. There were 224 individuals who consented for in-person screening. There were 49 excluded due to current smoking or regular in-home combustion sources (fireplace or wood stove use, burning candles, cigarette smoke, or other sources), 6 did not return after screening, and one person was excluded due to a new diagnosis of multiple sclerosis. We excluded 22 persons with <10 pack-years of smoking, and approach common in COPD treatment and epidemiologic studies (Albert et al., 2011). Participants were scheduled to return 4 times seasonally over one year, with visits approximately every 3 months. Prior to each visit, indoor PM_2.5 was collected in the participants home with a goal of a week of sampling. Only study visits where the home PM sampler was returned within 8 days before the study visit were included. At each visit, participants completed questionnaires (including the SGRQ) (Jones et al., 1991; Meguro et al., 2006), provided blood and urine samples, and completed spirometry before and after an inhaled short acting beta agonist bronchodilator (Wang et al., 2023). Visits were scheduled a minimum of two weeks after completion of antibiotics or steroids for a COPD exacerbation. There were 141 men and 3 women with eligible study visits. The protocol was approved by the Institutional Review Boards of VA Boston Healthcare System #2615 and Harvard Medical School #19–0656, and all participants provided written informed consent.

2.2. Health outcomes

Responses to questions regarding frequent cough, phlegm, shortness of breath, wheeze and chest attacks in the past 3 months were obtained from the SGRQ. There are 5 response options for symptoms and chest attacks (questions provided in Table S1 in the Supplement). Chest attack referred to the answer to the question “during the past 3 months how many severe or very unpleasant attacks of chest trouble have you had?”. We considered the symptom to be frequent when it was reported to be “most days a week” or “several days a week” and chest attacks frequent when reported as “more than 3 attacks” or “3 attacks”. The SGRQ symptoms component score, activity score, impacts score and total score were also calculated where 100 represents worst possible health status and 0 indicates best possible health status (Jones and Forde, 2009).

2.3. Exposure assessment

2.3.1. Assessment of indoor exposures

Indoor PM_2.5 was collected on a Teflon filter using a Harvard School of Public Health (HSPH) Micro-environmental Automated Particle Sampler (APS) in the main activity room of the house (excluding the kitchen) as previously described (Garshick et al., 2018; Busenkell et al., 2022). Indoor black carbon (BC) concentration was determined by measuring optical absorbance at 880 nm on the filters using a SootScan OT21 Transmissometer (Magee Scientific Company, model AE-16, Berkeley, CA). Indoor alpha radioactivity was measured from archived indoor PM_2.5 filters using a low-background proportional counter (Model LB4200; Canberra Industries, Inc., Meriden, CT). Alpha radioactivity on filters represents the decay of ²¹⁰Pb (the longest-lived radionuclide with a half-life period of 22 years) to alpha-emitter ²¹⁰Po (with a half-life period of 138 days) (EPA, 2018; Wang et al., 2023). After about one and a half years, alpha radioactivity of PM_2.5 from ²¹⁰Po decay is limited by the rate of ²¹⁰Pb decay. As alpha radioactivity was measured 3 years or more after sample collection, α-PR at the original time of sampling can be calculated based on the decay constant of ²¹⁰Po (Liu et al., 2020; Kang et al., 2020).

2.3.2. Assessment of ambient exposures

Daily ambient PM_2.5 samples were collected on Teflon filters at a central site located on the rooftop of the Francis A. Countway Library in Boston, MA, 6 stories above ground level (Huang et al., 2018; Kang et al., 2010) and α-PR measured on archived indoor and outdoor filter samples. Daily ambient α-PR was matched to and averaged over the indoor sampling dates.

2.3.3. Assessment of indoor exposures originated from indoor and ambient sources

We estimated residential α-PR of outdoor and indoor origin using the indoor-outdoor ratio of sulfur in PM_2.5 to represent the fraction of ambient particles (including outdoor PR) infiltrating indoors (Sarnat et al., 2002; Kang et al., 2010; Hsu et al., 2012). Ambient and indoor sulfur on PM_2.5 filters were measured using an energy dispersive X-ray fluorescence spectrometer (Model Epsilon 5; PANalytical, Netherlands), and daily ambient sulfur concentrations matched to and averaged over the indoor sampling dates. Indoor α-PR of outdoor origin was estimated by multiplying central site α-PR by the ratio of indoor/ambient sulfur (a ventilation surrogate). α-PR of indoor origin was calculated as total indoor α-PR minus α-PR of outdoor origin. Although these calculations resulted in some negative estimates, all values were included to preserve exposure distributions and rank ordering. To exclude homes with large indoor sulfur sources, we excluded observations with indoor-outdoor sulfur ratio larger than 1.2, as a higher sulfur ratio indicated indoor sulfur sources (Huang et al., 2018).

2.4. Covariates

At each visit we measured height and weight in kg (to calculate body mass index, BMI), and collected information about pulmonary medication use (long-acting bronchodilators and inhaled steroids), report of a respiratory illness in the last 2 weeks, and heart disease treated in the past 10 years. Age at each study visit was calculated from date of birth and seasonality assessed using a sine and cosine function of day of the year (Stolwijk et al., 1999). Percent predicted forced expiratory volume in one second (FEV₁) and forced vital capacity (FVC) were calculated using NHANES predicted values (Hankinson et al., 1999).

2.5. Statistical analysis

As we had repeated measurements, we adopted multivariable linear and logistic regression with generalized estimating equations accounting for clustering by individual to explore the associations between the target exposure with health outcomes (Package “geepack” in R software) (Halekoh et al., 2006). An exchangeable working correlation structure was assumed, as the QIC value was the lowest which represents that the accuracy was the highest with this structure. We examined associations between each symptom with measured indoor α-PR, and estimates of α-PR of indoor and outdoor origin, adjusting for age, race, a recent respiratory illness, long acting pulmonary medication use (including inhaled steroids), BMI, past smoking history (packyears), and seasonality. Results were expressed as %-increase per interquartile range (IQR) of α-PR. For the assessment of SGRQ scores, we adjusted for age, race, self-report of a respiratory illness in the last 2 weeks, seasonality, and included post-bronchodilator %-predicted FEV₁ to adjust for COPD severity, and heart disease as comorbidities potentially influencing HRQL, with results expressed as increase in each score per IQR. The covariates were selected via a causal directed acyclic graph (Figure S1). In sensitivity analyses we adjusted for indoor PM_2.5 or BC in separate models. All statistical analysis was performed using R 4.2.1 software. We were unable to include the 3 women in the cohort in the analysis as the analytic models did not converge as a result of the small numbers of women when sex was included a covariate.

3. Results

3.1. Descriptive characteristics

Of the 141 male participants, most were white, and the average age over all visits was mean (standard deviation, SD) 73.3 (8.2) years old (Table 1). Of 487 observations, 6 were excluded because they were outliers (measured PR greater than 3 SD from the mean), 5 observations did not have post bronchodilator pulmonary function values, and 2 had missing SGRQ scores, for a total of 474 observations. There were 80 participants with 4 visits, 37 with 3 visits, 19 with 2 visits, and 5 with 1 visit. The majority of participants were taking long-acting bronchodilators and inhaled steroids. The mean (SD) time between study visits was 3.4 (1.0) months and 97.3 % visits were completed within one year. Sampling was over a mean (SD) of 7.7 (0.8) calendar days (range = 4–14) and in 85 % of visits the sampler was returned within 2 days of the study visit. In 8.4 % of the observations (n = 40), a short-term indoor combustion source was reported, typically over 1–2 days (1–2 cigarettes by others, candle use of several hours, or fireplace or wood stove use). The percentages of frequent cough, phlegm, shortness of breath, wheeze and chest attacks were 51.9 %, 49.2 %, 59.7 %, 25.5 %, 29.7 % respectively (Table 1) and there was a wide range in SGRQ total score and the 3 component scores. The median (IQR) concentrations of α-PR were 1.22 (0.62) mBq/m³, α-PR of indoor origin was 0.11 (0.38) mBq/m³, and α-PR of outdoor origin was 1.11 (0.63) mBq/m³ (Table 2) with low median concentrations of indoor PM_2.5 and BC. Indoor and outdoor sulfur measured in the PM_2.5 samples were highly correlated (Spearman correlation coefficient = 0.73, p<0.001), consistent with the ratio of indoor/outdoor sulfur serving as a measure of infiltration of α-PR. As is shown in Table S2, we note that outdoor PM_2.5 and outdoor α-PR are highly correlated, also indicating sulfur in outdoor PM_2.5 may be used as a measure of infiltration.

Table 1.

Descriptive information for the 141 study participants.

	Mean (SD)	Range	N (%)
Descriptive characteristics
Age (yrs)	73.3 (8.2)	46.7-90.0
BMI (kg/m2)	30.3 (5.8)	17.0-50.8
Age categories	≤70 yrs		48 (34.0)
	>70 yrs and ≤80 yrs		60 (42.6)
	> 80 yrs		33 (23.4)
BMI categories	Under 24.9 kg/m2		22 (15.6)
	25 to 29.9 kg/m2		55 (39.0)
	30 kg/m2 or higher		64 (45.4)
Race	White		129 (91.5)
	Non-white		12 (8.5)
Report of a respiratory illness in the last 2 weeks	Yes		86 (18.1)
	No		388 (81.9)
Long-acting bronchodilators	Yes		372 (78.5)
	No		102 (21.5)
Inhaled steroids	Yes		355 (74.9)
	No		119 (25.1)
Heart disease in past 10 years	Yes		252 (53.2)
	No		222 (46.8)
Post-bronchodilator percent predicted FEV1	69.4 (22.4)	16.2-131.4
Post-bronchodilator percent predicted FVC	89.2 (20.7)	40.2-148.4
Post-bronchodilator FEV1 / FVC	0.56 (0.13)	0.21-0.87
Symptoms
Cough	Frequently		246 (51.9)
	Not frequently		228 (48.1)
Phlegm	Frequently		233 (49.2)
	Not frequently		241 (50.8)
Shortness of breath	Frequently		283 (59.7)
	Not frequently		191 (40.3)
Wheeze	Frequently		121 (25.5)
	Not frequently		353 (74.5)
Chest attacks	Frequently		141 (29.7)
	Not frequently		333 (70.3)
Health related quality of life (HRQL)
Symptoms score	43.2 (23.8)	0.0-100.0
Activity score	31.0 (14.4)	0.0-57.1
Impacts score	26.2 (18.9)	0.0-91.9
Total score	37.6 (19.1)	0.0-90.8

Table 2.

Concentrations of indoor α-particle radioactivity (PR), indoor PM_2.5 and BC, and estimates of indoor concentrations of α-PR of indoor and outdoor origin (N=474).

Exposure		Mean (SD)	Median (IQR)	Range
Measured indoor	α-PR (mBq/m3)	1.28 (0.50)	1.22 (0.62)	0.15–3.09
	PM2.5 (μg/m3)	8.34 (6.06)	6.66 (5.41)	0.17–45.88
	BC (μg/m3)	0.59 (0.25)	0.56 (0.26)	0.12–2.24
Ambient origin	α-PR (mBq/m3)	1.13 (0.48)	1.11 (0.63)	0.00-2.71
Indoor origin	α-PR (mBq/m3)	0.15 (0.36)	0.11 (0.38)	−1.25–1.42
Indoor/Outdoor sulfur		0.67 (0.22)	0.67 (0.28)	0.00–1.18

3.2. Associations between α-PR with symptoms

There was a positive association between total measured α-PR with an increased risk of frequent cough (28.1 % per IQR; 95 %CI: 2.4 %, 60.3 %) and with α-PR of indoor origin (31.1 % per IQR; 95 %CI: 8.8 %, 57.8 %) (Fig. 1a, Table S3). The association between total α-PR with an increased risk of frequent phlegm was weakly positive (3.0 % per IQR; 95 %CI: −14.4 %, 23.8 %) but was considerably stronger for α-PR of indoor origin (13.0 % per IQR; 95 %CI: −2.5 %, 31.0 %). The associations were similar and nearly unchanged adjusting for PM_2.5 and BC in separate models (Fig. 2, Table S5). There were no associations between α-PR of outdoor origin with frequent cough or phlegm (Fig. 1a, Table S3), and no associations between total α-PR, α-PR of indoor origin, or α-PR of outdoor origin with frequent shortness of breath, wheeze or chest attacks (Fig. 2, Tables S3–S6).

Fig. 1.

Change in symptoms and scores per IQR of indoor total α-PR, indoor α-PR of indoor origin, and indoor PR of outdoor origi. (a)%-change per interquartile range of total indoor α-PR, indoor origin α-PR, and outdoor origin α-PR, respectively. (a)Increase in HRQL scores per interquartile range of total indoor α-PR, indoor origin α-PR, and outdoor origin α-PR, respectively.

Fig. 2.

Change in symptoms per IQR of indoor origin α-PR using two-pollutant model adjusting for indoor PM_2.5 and indoor black carbon (BC).

3.3. Associations between α-PR with HRQL

The only overall positive associations for the SGRQ scores including the component scores were with α-PR of indoor origin (Fig. 1b, Table S3). The strongest association was with symptoms score, with an increase of 1.2 (95 %CI: −0.3, 2.6) per IQR, that was similar and nearly unchanged adjusting for PM_2.5 and BC in separate models (Fig. 3, Table S5). There were also positive associations between α-PR of indoor origin with impacts score and total SGRQ score (Fig. 3, Table S5). We also found negative associations between α-PR of outdoor origin with impacts score and total SGRQ score (Tables S3 and S6).

Fig. 3.

Change in SGQR scores per IQR of indoor origin α-PR using two-pollutant model adjusting for PM_2.5 and black carbon (BC).

3.4. Associations between α-PR with PM_2.5 and BC

There were no consistent associations with PM_2.5 and BC in one-pollutant models (Tables S7 and S8), or in two-pollutant models with total indoor α-PR, indoor origin α-PR, and outdoor α-PR in separate models (Figs. 2, 3; Tables S4–S6)

4. Discussion

We investigated the association between indoor α-PR with chronic respiratory symptoms, report of chest attacks, and HRQL assessed by the SGRQ in patients with COPD evaluated up to 4 times over a year, adjusting for indoor PM_2.5 or BC. The assessment of indoor α-PR included estimates of residential α-PR of indoor and outdoor origin. There were positive associations between total α-PR with frequent cough and phlegm that were stronger for α-PR of indoor origin and absent for α-PR of outdoor origin, and a suggestive positive association between α-PR of indoor origin and SGRQ Symptoms component score. The results of two-pollutant models adjusting for PM_2.5 or BC were almost the same with the results of the single-pollutant models, demonstrating the robustness of the associations with α-PR and α-PR of indoor origin. We found no associations with indoor PM_2.5 or BC, suggesting that the increased risk of cough and phlegm and increase in SGRQ Symptoms component score were attributable to α-PR and not particulate pollution.

Previous research regarding the biological mechanisms of the adverse effects of radioactivity mostly focused on high dose ionizing radiation (e.g., nuclear leakage or radiotherapy (Roy et al., 2021). The lung is one of several moderately radiosensitive organs, and radiation-induced lung injury at much higher doses than attributable to radon decay products may result in respiratory symptoms, including cough, shortness of breath, chest pain (Huang et al., 2017). There is experimental evidence that inhalation of radiation attributable to radon gas and α-emitting radionuclides at much lower doses could result in pulmonary airway damage from the development of inflammation and oxidative stress. Inhalation of α-emitting particles and of radon gas (resulting in radon decay product inhalation) in animal and in-vitro experiments have resulted in up-regulation of interleukin-6 (an inflammatory cytokine) in white blood cells (Li et al., 2007), evidence of airway and bronchoalveolar lavage (BAL) fluid inflammation (Li et al., 2007), increased oxidative stress biomarkers in urine and BAL fluid and in human fibroblasts (Narayanan et al., 1997), and up-regulation of gene pathways that include inflammatory and respiratory diseases (Chauhan et al.,. 2012). Our findings regarding respiratory symptoms are also consistent with our previous epidemiologic studies assessing associations between PR (measured as α-activity, γ-activity and β-activity) with reduced pulmonary function, and evidence of urinary oxidative stress and systemic inflammation (Huang et al., 2020, 2021; Li et al., 2018; Vieira et al., 2019; Nyhan et al., 2019; Wang et al., 2023). Most notably, in the current COPD cohort we found that radon decay product exposure measured by high indoor α-PR but low air infiltration was associated with a reduction in FEV₁ and FVC (Wang et al., 2023). This is consistent with the greater association of radon decay products of indoor origin with cough and phlegm in the current study.

Another novelty of this study is that we examined not only the effects of total indoor α-PR on COPD symptoms and HRQL, but also estimated the effects of indoor α-PR from indoor and outdoor sources that provided insight into the importance of exposure to radon decay products of indoor origin. In homes we previously studied in Massachusetts, we found that over half of the indoor α-PR on archived filters was attributable to an outdoor source (Matthaios et al., 2021). However, a limitation of the measurement of α-PR on archived filters is the inability to measure α-activity from short-lived radon decay products generated early in the radon decay chain. In a study where we measured residential radon decay products early (using real-time measurement methods) and late in the decay chain, we found that short-lived radon products emit considerably more α-activity (approximately 4 orders of magnitude greater) (Kang et al., 2020). This observation suggests that our estimation of α-PR of indoor origin in this study may serve as a surrogate for short-lived radon decay products of indoor origin that may account for its greater effect.

This study has some limitations. We had available only weekly exposures, which were not in the same period of symptoms assessed in the questionnaires (in the past 3 months). Our analysis suggests that the week-long exposures assessed seasonally may be surrogates for longer term exposures. Second, all the subjects of our study were predominantly white males living in the Northeastern US, which may restrict the generalizability of the study. Individuals with more severe COPD and who may be more sensitive to exposures were not included in our cohort, which may result in the underestimation of the effects of exposures. Although major COPD specific symptoms were assessed and the SGRQ is commonly used to assess COPD-related HRQL, it is possible that other respiratory health instruments may have provided different results. We also included covariates recognized to be important to the health of people with COPD, but there may have been also other unmeasured factors. We also note that a larger study cohort may have improved our ability to assess significant associations.

We note that our study is the first to investigate the association between indoor particle radioactivity with respiratory symptoms and HRQL in patients with COPD. In addition, our study distinguished associations with indoor PR from indoor and ambient sources. Although we did not directly study the effects of a reduction of indoor radon levels and PR, our results suggests that the control of indoor radon exposures may mitigate respiratory symptoms in patients with COPD.

5. Conclusions

Exposure to α-PR of indoor origin is positively associated with cough and phlegm and worse HRQL symptoms score in patients with COPD, associations which were similar adjusting for indoor BC and PM_2.5. Our results suggest that there may be benefits from radon mitigation that contributes to indoor PR in the homes of patients with COPD.

Supplementary Material

Supplementary data associated with this article can be found in the online version at doi:10.1016/j.ecoenv.2024.116693.

Abbreviations:

PR

particle radioactivity

HRQL

health-related quality of life

SGRQ

St. George’s Respiratory Questionnaire

BMI

body mass index

FEV₁

forced expiratory volume in one second

FVC

forced vital capacity

PM_2.5

particulate matter ≤ 2.5 μm in diameter

BC

black carbon

Data Availability

The authors do not have permission to share data.