Secondhand Smoke Exposure is Associated with Abnormal P-Wave Axis

Abstract

Objectives:

Secondhand smoke exposure (SHSE) is associated with increased risk of cerebrovascular accident (CVA). Abnormal p-wave axis (aPWA) is a marker for atriopathy that is also associated with CVA risk. We hypothesized that SHSE is associated with aPWA.

Methods:

This analysis included 5,986 nonsmokers (age 61.7±13.8, 45.8% men, 77.4% whites) from the Third National Health and Nutrition Examination Survey. SHSE was defined as serum cotinine ≥1ng/mL. aPWA was defined as any p-wave axis outside of 0–75 degrees. Multivariable logistic regression was used to examine the association between SHSE and aPWA, overall and among subgroups stratified by demographics and comorbidities.

Results:

About 18.5% (n=1109) of the participants had SHSE. aPWA was more prevalent among those with SHSE than those without (23.9% vs. 19.8%, respectively, p-value = 0.003). In a model adjusted for sociodemographic and potential confounders, presence (vs. absence) of SHSE was associated with increased odds of aPWA (odds ratio (95% CI): 1.28 (1.09, 1.50); p-value=0.003). This association was stronger among whites vs. non-whites (interaction p-value=0.04) and non-obese vs. obese (interaction p-value=0.04). Higher levels of serum cotinine were associated with increased odds of aPWA. Compared to serum cotinine level <1 ng/mL, serum cotinine ≥3ng/mL and ≥6ng/mL were associated with 35% (p-value=0.002) and 38% (p-value=0.002) increased odds of aPWA, respectively.

Conclusions:

SHSE is associated with abnormal atrial conduction, measured as aPWA, with possible effect modification by ethnicity and obesity. These findings underscore the harmful effects of SHSE on cardiovascular health which merits a personalized risk assessment when counseling patients on SHSE.

Introduction:

Secondhand smoke exposure (SHSE) is a public health concern responsible for ~1% of global mortality^{1, 2}. The association of both active smoking and SHSE with atrial fibrillation (AF)/cerebrovascular accident (CVA) is well-established^{3, 4}. Nicotine exposure has direct effects on atrial structural remodeling^{5, 6} with subsequent impacts on autonomic function⁷, both of which are implicated in the pathogenesis of AF and CVA. In addition, there is some evidence that SHSE is actually more dangerous than active smoking due to highly toxic side stream smoke⁸. Hence, early detection of underlying atrial myopathy may potentially help improve risk stratification for those with nicotine exposure, whether passive or active.

Abnormal P-wave axis (aPWA), a marker of atrial myopathy, is independently associated with AF and CVA of ischemic etiology^9–11. At the molecular level, atrial myopathy causes gap junction remodeling including changes in distribution, intercellular orientation, and expression of gap junction proteins resulting in electrophysiological and structural stimuli for sustained AF^{12, 13}. Further, atrial myopathy predisposes the left atrium to a thrombotic milieu via fulfilment of Virchow’s Triad¹⁴. We hypothesized that SHSE would be associated with abnormal aPWA in a cross-sectional cohort study.

Methods:

The Third National Health and Nutrition Examination Survey (NHANES-III) is a survey of the United States population which estimates disease prevalence and health status¹⁵. Data in NHANES-III were collected from 1988–1994 through in-home interviews and appointments examination centers.

We only included participants with ECG data in our analysis. Participants with poor quality ECGs or major intraventricular conduction delay were excluded. Any participants with missing data for any of the variables examined were also excluded. No participants <18 years of age were included. Age, sex, race (white, black, and other), smoking status, and history of CVD (stroke or heart attack) were defined by self-report. Diabetes was defined as hemoglobin A1c ≥6.5%, fasting serum glucose ≥126 mg/dL, or use of a glucose-lowering medication. Hypertension was defined as use of an antihypertensive medication, systolic blood pressure ≥130 mmHg, or diastolic blood pressure ≥80 mmHg according to American Heart Association/American College of Cardiology guidelines¹⁶. Hyperlipidemia was defined as total cholesterol ≥200 mg/dL, serum triglycerides ≥150 mg/dL, or use of antihyperlipidemic medications. Obesity was defined as body mass index ≥30 kg/m².

Serum cotinine was measured via isotope-dilution-high-performance liquid chromatography atmospheric pressure chemical isonization-tandem mass spectrometry¹⁷.

Twelve-lead ECGs were obtained by trained technicians with a Marquette MAC 12 system (Marquette Medical Systems, Milwaukee, WI) during examination center visits. P-wave axis was obtained by computerized analysis which included selective averaging to obtain amplitudes and durations of all ECG waveform components. aPWA was defined as any p-wave axis outside of 0–75 degrees¹⁰ and modeled as a binary variable.

Population characteristics were compared between those with SHSE and those without. SHSE was defined as serum cotinine ≥1 ng/mL which is considered the cut-off for “heavy” SHSE by the Centers for Disease Control¹⁸. Continuous variables were reported as mean ± standard deviation. Categorical variables were reported as frequency and percentage. A chi-square test was used to compare categorical variables, and a student’s t-test was used to compare continuous variables.

Multivariable logistic regression was used to examine the association between aPWA and SHSE. Two models were used: model 1 was adjusted for age, sex, and race and model 2 was adjusted for model 1 plus hypertension, diabetes, hyperlipidemia, obesity, and history of CVD.

The association between SHSE and aPWA was also evaluated when stratified by race, sex, hypertension, obesity, diabetes, hyperlipidemia, and history of CVD. Interaction was tested for using similar variables to model 2 with the interaction term between SHSE status and subgroup stratification.

All statistical analyses were conducted using R version 4.0.3 and p-values were considered significant if <0.05.

Results:

Table 1 shows population characteristics stratified by SHSE. Participants with SHSE were more likely to be male, black, and have aPWA. Age and prevalence of hypertension, obesity, diabetes, hyperlipidemia, and history of CVD were similar between groups.

Table 1:

Population Characteristics

Characteristics	Secondhand Smoke Exposure		p-value
	Present	Absent
(Mean ± SD or n (%))	1109 (18.5%)	4877 (81.5%)
Age (years)	61.3 ± 13.7	61.8 ± 13.8	0.22
Men	646 (58.3%)	2096 (43.0%)	<0.001
Race
White	737 (66.5%)	3896 (79.9%)	<0.001
Black	345 (31.1%)	846 (17.3%)
Other	27 (2.4%)	135 (2.8%)
Hypertension	761 (68.6%)	3261 (66.9%)	0.27
Obesity	315 (28.4%)	1472 (30.2%)	0.26
Diabetes	217 (19.6%)	838 (17.2%)	0.07
Hyperlipidemia	826 (74.5%)	3677 (75.4%)	0.55
History of CVD	122 (11.0%)	494 (10.1%)	0.42
aPWA	265 (23.9%)	968 (19.8%)	0.003

Secondhand Smoke Exposure = serum cotinine ≥1ng/mL

CVD = cardiovascular disease (myocardial infarction or stroke)

aPWA = abnormal p-wave axis

The regression results among nonsmokers with SHSE is shown in Table 2. When using a cotinine cut-point of ≥1ng/mL, SHSE was associated with 28% increased odds of aPWA [Odds Ratio (95% Confidence Interval): 1.28 (1.09–1.50), p=0.003]. When using a cotinine cut-point of ≥3ng/mL, SHSE was associated with 35% increased odds of aPWA [Odds Ratio (95% Confidence Interval): 1.35 (1.11–1.63), p=0.002]. When using a cotinine cut-point of ≥6ng/mL, SHSE was associated with 38% increased odds of aPWA [Odds Ratio (95% Confidence Interval): 1.38 (1.13–1.69), p=0.002].

Table 2:

Association of Secondhand Smoke Exposure and Abnormal P-Wave Axis

Serum Cotinine Level	Participants	Reference Level	Model 1		Model 2
			Odds Ratio (95% CI)	p-value	Odds Ratio (95% CI)	p-value
≥1 ng/mL	1109	<1 ng/mL	1.28 (1.09–1.50)	0.002	1.28 (1.09–1.50)	0.003
≥3 ng/mL	707	<1 ng/mL	1.39 (1.15–1.68)	<0.001	1.35 (1.11–1.63)	0.002
≥6 ng/mL	603	<1 ng/mL	1.44 (1.18–1.75)	<0.001	1.38 (1.13–1.69)	0.002

Model 1 adjusted for age, sex, and race

Model 2 adjusted for model 1 plus hypertension, diabetes, obesity, hyperlipidemia, and history of cardiovascular disease (myocardial infarction or stroke)

Table 3 shows the association of SHSE and aPWA when stratified among sub-groups using a serum cotinine cut-point of ≥1ng/mL. The association was stronger among those of white ethnicity (interaction p-value=0.04) and among those without obesity (interaction p-value=0.04). Results were consistent among the remainder of sub-groups analyzed.

Table 3:

Association of Secondhand Smoke and Abnormal P-Wave Axis Among Sub-groups

Sub-group		Model 1†	Model 2†
		Odds Ratio (95% CI)	Odds Ratio (95% CI)	Interaction p-value*
Race	Non-white	1.12 (0.84–1.49)	1.17 (0.87–1.56)	0.04
	White	1.38 (1.14–1.67)	1.36 (1.12–1.66)
Sex	Men	1.44 (1.16–1.79)	1.43 (1.15–1.78)	0.10
	Women	1.11 (0.88–1.41)	1.12 (0.88–1.42)
Hypertension	Present	1.21 (0.99–1.48)	1.21 (0.98–1.48)	0.21
	Absent	1.45 (1.12–1.89)	1.45 (1.11–1.89)
Obesity	Present	0.85 (0.58–1.26)	0.85 (0.58–1.26)	0.04
	Absent	1.38 (1.16–1.65)	1.39 (1.17–1.67)
Diabetes	Present	1.26 (0.84–1.87)	1.25 (0.84–1.88)	0.74
	Absent	1.30 (1.09–1.54)	1.30 (1.09–1.55)
Hyperlipidemia	Present	1.34 (1.11–1.61)	1.31 (1.08–1.59)	0.59
	Absent	1.16 (0.86–1.57)	1.21 (0.89–1.65)
History of CVD	Present	1.49 (0.93–2.36)	1.46 (0.91–2.34)	0.56
	Absent	1.26 (1.07–1.50)	1.26 (1.06–1.50)
Tobacco Smoking History	Never	1.08 (0.86–1.37)	1.13 (0.89–1.43)	0.26
	Former	1.44 (1.16–1.79)	1.37 (1.10–1.72)

95% CI = 95% Confidence Interval

Model 1 adjusted for age, sex, and race

Model 2 adjusted for model 1 plus hypertension, obesity, diabetes, hyperlipidemia, and cardiovascular disease (myocardial infarction or stroke)

^†Serum cotinine ≥1ng/mL

^*Interaction p-value calculated from model 2

CVD = cardiovascular disease (myocardial infarction or stroke)

Discussion:

In this large, ethnically diverse cohort, both active smoking and SHSE were significantly associated with aPWA. Effect modification by ethnicity and obesity was present among those with SHSE.

Cotinine is a nicotine metabolite with a half-life of ~18–20 hours¹⁹. Significant reporting bias exists when examining SHSE exposure by self-report only. A prior study of the US workforce showed that complete denial of SHSE was only ~28% accurate²⁰. An advantage of using serum cotinine is that it allows for an objective measure of the degree of SHSE. It has been used as a biomarker for SHSE because of its long half-life^{21, 22}. Per the Centers for Disease Control and Prevention (CDC) cotinine biomonitoring summary, typical SHS exposure usually confers a serum cotinine of ≤1 ng/mL and heavy exposure is >10 ng/mL^{18, 23}. Therefore, we examined the association between SHSE with aPWA using a primary cut-point of 1 ng/mL. We also included cut-points of 3 ng/mL and 6 ng/mL to measure possible dose-response.

Mechanistically, nicotine accelerates atrial collagen accumulation in a concentration-dependent fashion via expression of collagen III mRNA expression leading to symptomatic atrial fibrosis²⁴. In canine models, nicotine also downregulates miR-133 and miR-590 predisposing a profibrotic state in the atrium²⁵. Beyond the direct inflammatory effects, nicotine also indirectly impacts atrial fibrosis via its effects on the autonomic nervous system. Sympatho-vagal imbalances lead to direct atrial alterations with downstream atrial myopathy²⁶. One well-known example is the imbalance nicotine causes in the renin-angiotensin system²⁷. In particular, angiotensin II is a well-characterized profibrotic molecule and may cause atrial fibrosis and dilation²⁸. Nicotine has been shown to exacerbate cardiovascular remodeling induced by angiotensin II in mouse-models²⁹. Similar to direct tobacco exposure, SHSE also leads to sympathoexcitation^{30, 31}.

In our sub-group analysis, the effect of SHSE on aPWA was stronger in Caucasians. This finding corroborates findings from the Cardiovascular Heart Study demonstrating that white men on average have a larger left atrium than their African-American counterparts³² with an increased risk of AF in this same population³³. Our findings also lend support to an analysis from the Atherosclerosis Risk in Communities Study cohort showing increased incidence of AF in whites compared to African-Americans, despite risk factors being more prevalent in the latter population³⁴. Notably, the effect modification of obesity seemed to serve a “protective role” for co-existence of aPWA in SHSE. Interestingly, this protective effect has been well-described in other CV disease processes and has been termed the “obesity paradox”, so this may explain these paradoxical results³⁵.

Our study links SHSE to aPWA. aPWA is a known marker for atrial myopathy, which is independently associated with the development of ischemic stroke, particularly of cardioembolic etiology¹⁰. It is currently thought that aPWA reflects underlying atrial remodeling³⁶, which includes chamber enlargement and myocardial fibrosis, a recognized risk factor for AF^{37, 38} and a critical component for thrombogenesis in the atrium^{14, 39}. Further, incorporation of aPWA into the CHA₂DS₂-VASc (congestive heart failure, hypertension, age, diabetes mellitus, stroke, vascular disease, sex) score has shown to improve ischemic stroke prediction in AF⁴⁰.

Unlike other P-wave indices, P-wave axis is a routine measure reported on 12-lead ECGs that reflects the net vector component of atrial depolarization in the frontal pane and does not require digitization or specialized analytic software. Although the mechanism underlying the development of AF and CVA in those with active and passive exposure is well-described, our study is important because it allows clinicians to readily identify individuals in this subgroup who have an increased predisposition to these pathologies.

Our study has numerous limitations. Cross-sectional studies are subject to temporality bias and residual confounders. Smoking status was defined by self-report and is subject to significant reporting bias. The CDC cotinine biomonitoring summary reports that active smokers usually have serum cotinine levels >10 ng/mL and often >500 ng/mL^{18, 23}, so it is possible that our cohort suffered from misclassification bias of smoking status. On the other hand, elevated serum cotinine levels do not exclude heavy SHS exposure among nonsmokers due to individual metabolic and environmental differences⁴¹. Several studies have examined a specific cotinine cut-point to determine active versus passive smoking with inconsistent results, so this alone is an unreliable metric to resolve “true” smoking status⁴¹. Finally, it is unclear whether or not former smokers have persistent aPWA, so this is another area that needs further examination.

Conclusions:

SHSE was associated with aPWA with possible effect modification by ethnicity and obesity. This is a novel finding that highlights the harmful effects of SHSE on cardiovascular health and necessitates a personalized risk assessment when counseling patients on environmental tobacco smoke exposure. Further prohibition of smoking in public areas is needed, especially in areas with less-stringent public health policies.