lable at ScienceDirect

Atherosclerosis 235 (2014) 189e195

Contents lists avai

Atherosclerosis

journal homepage: www.elsevier .com/locate/atherosclerosis

Review

Self-reported habitual snoring and risk of cardiovascular diseaseand all-cause mortality

Dongmei Li*, Debao Liu, Xiuming Wang, Dongdong HeDepartment of Air Force Service, The Chinese People’s Liberation Army 463 Hospital, Number 46, Xiaoheyan Road, Dadong District, Shenyang 110000, China

a r t i c l e i n f o

Article history:Received 22 March 2014Received in revised form16 April 2014Accepted 27 April 2014Available online 6 May 2014

Keywords:Habitual snoringCardiovascular diseaseStrokeAll-cause mortalityMeta-analysis

* Corresponding author. Tel./fax: þ86 024 2884537E-mail address: lidongmeishy@163.com (D. Li).

http://dx.doi.org/10.1016/j.atherosclerosis.2014.04.0310021-9150/� 2014 Elsevier Ireland Ltd. All rights rese

a b s t r a c t

Objective: Inconsistent findings have reported the association between self-reported habitual snoringand risk of cardiovascular disease (CVD) and all-cause mortality. We conducted a meta-analysis toinvestigate whether self-reported habitual snoring was an independent predictor for CVD and all-causemortality using prospective observational studies.Methods: Electronic literature databases (PubMed, Medline, Embase, Cochrane Library, Wanfang data-base, and China National Knowledge Infrastructure) were searched for publications prior to September2013. Only prospective studies evaluating baseline habitual snoring and subsequent risk of CVD and all-cause mortality were selected. Pooled adjust hazard risk (HR) and corresponding 95% confidence in-tervals (CI) were calculated for categorical risk estimates.Results: Eight studies with 65,037 subjects were analyzed. Pooled adjust HR was 1.26 (95% CI 0.98e1.62)for CVD, 1.15 (95% CI 1.05e1.27) for coronary heart disease (CHD), and 1.26 (95% CI 1.11e1.43) for strokecomparing habitual snoring to non-snorers. Pooled adjust HR was 0.98 (95% CI 0.78e1.23) for all-causemortality in a random effect model comparing habitual snoring to non-snorers. Habitual snoringappeared to increase greater stroke risk among men (HR 1.54; 95% CI: 1.09e2.17) than those in women(HR 1.22; 95% CI: 1.05e1.41).Conclusions: Self-reported habitual snoring is a mild but statistically significant risk factor for stroke andCHD, but not for CVD and all-cause mortality. However, whether the risk is attributable to obstructivesleep apnea syndrome or snoring alone remains controversial.

� 2014 Elsevier Ireland Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12. Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

Conflict of interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6Supplementary data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

1. Introduction

Habitual snoring is a much more common disorder, with prev-alence 20e40% in adult population [1,2]. Frequency of the

4.

rved.

comorbidities was presented in patients with obstructive sleepapnea syndrome (OSAS) and simple snoring [3]. Although snoringis considered a symptom of OSAS [4], many snore persons do nothave OSAS. Snoring without OSAS has long been considered a socialnuisance. However, the cumulative evidences showed that snoringmight have important adverse health implications.

Many studies [5e12] have assessed the association betweenself-reported habitual snoring and cardiovascular disease (CVD)

D. Li et al. / Atherosclerosis 235 (2014) 189e195190

and mortality. Habitual snoring may help clinicians identify in-dividuals at higher risk for CVD. However, conflicting result [13] hasbeen reported. Whether this reflects confounding factors orwhether the association is attributable to OSAS remains contro-versial. Severe OSAS was an independent risk factor for CVD [14].Habitual snoring could be identified as an early marker of OSAS.Therefore, it is important to evaluate the association between self-reported habitual snoring and clinical outcomes.

To the best of our knowledge, the magnitude of the associationbetween self-reported habitual snoring and risk of CVD or all-causemortality has not been quantitatively evaluated. Given these rea-sons, a meta-analysis may help clarify this issue. The objective ofour meta-analysis was to quantitatively evaluate findings fromprospective observational studies on self-reported habitual snoringand risk of CVD or all-cause mortality, and determine whether self-reported habitual snoring is an independent risk factor of CVD orall-cause mortality.

2. Methods

We conducted a PubMed, Medline, Embase, Cochrane Library,Wanfang database, and China National Knowledge Infrastructuresearch prior to September 2013 for studies assessing the associationbetween self-reported habitual snoring and future risk of CVD orall-cause mortality. Potentially relevant studies were identifiedusing the word ‘snoring’, ‘snorer’ ‘self-reported snoring’ plus atleast one of the following terms (Supplementary Text S1): cardio-vascular disease, ischemic heart disease, coronary heart disease,stroke, death and mortality, prospective, follow-up study. In addi-tion, we also manually searched the reference lists to detect addi-tional eligible studies.

Studies satisfying the following criteria were selected: 1) pro-spective observational study reporting self-reported snoring andrisk of CVD or all-cause mortality in a general population; 2) sub-jects were initially free from CVD at baseline or excluded in the finalstatistical analysis; and 3) providing adjusted hazard risk (HR) orodds ratio (OR) and 95% confidence interval (CI) comparinghabitual snoring to the non-snorers. The question of snoring wasphrased: ‘Do you now or have you ever been told that you snoreduring the night?’ Habitual snoring was defined by the individualstudy, the rest as non-snorers. Studies were excluded if 1) not aprospective design; 2) reporting unadjusted HR or OR; and 3)snoring was not measured by questionnaire or interview.

Outcome measures included incident CHD, stroke, total CVD,and all-cause mortality. Outcome assessment is defined in theincluded studies based on medical diagnostic codes and clinicalcriteria. CHD is defined as definite and probable myocardialinfarction, CHD death, coronary revascularization, and angina. CVDincluded CHD (ICD-8: 410e414; ICD-9: 410e414; ICD-10: I20e25)and stroke (ICD-8: 431e438; ICD-9: 430e438; ICD-10: I60e68,G45). Death was obtained from the medical records, or from officialdeath certificates.

Two reviewers (DM Li and XM Wang) independently extractedthe data from each study.We extracted the following items: author;year of publication; the location of study; sample size, gender (%),and the mean age or age range of participants; outcome assess-ment; adjusted HR and 95% CI; follow-up duration; and adjust-ments for confounding. Where discrepancies were identified,reviewers resolved these by discussion.

Quality assessment was performed with consideration for thefollowing aspects followed the Meta-analysis of ObservationalStudies in Epidemiology guidelines [15]: clear inclusion andexclusion criteria; documentation of the loss to follow-up rate;clear definition of outcome and outcome assessment; sufficientduration of follow-up; appropriate statistical analysis; and

important confounded and prognostic factors identified. All itemshad the following answer options: yes/no/too little information toanswer the question.

Data analyses used most fully adjusted HR and 95% CI. Wepooled the separate HR for the different items and compared thehabitual snoring to the non-snorers. Before pooling the data,adjusted HR was converted to their logHR to stabilize the variancesand to normalize the distributions. Homogeneity of HR acrossstudies was assessed using Cochran’s Q-test and the I2 statistic. AP > 0.10 or I2<50% were taken as indicators of the same scale ofoutcomes using a fixed-effect model; otherwise using a randomeffect model [16].

Publication bias was assessed by both the Begg’s rank correla-tion test [17] and Egger linear regression test (P < 0.10) [18].Sensitivity analysis was performed by sequentially omitting onestudy at each turn with the metaninf algorithm in STATA toinvestigate the influence of a single study on the overall risk esti-mate. All analyses were performed using Revman software fromthe Cochrane Collaboration (version 5.1, Oxford, UK) and STATAstatistical software (version 12.0; STATA Corp LP, College Station,TX, USA). P < 0.05 was considered as statistically significant.

3. Results

A total of 168 potentially relevant citations were identified inour initial literature search. Of these, 8 studies [5e12] with 65,037subjects met the inclusion criteria (Fig. 1). The detailed character-istics of the included studies are listed in Table 1. Quality of theincluded studies is shown in the Supplement Table S1. In general,these studies had a relatively high quality.

Five studies [5,8,10e12] reported CVD. The total number ofsubjects included in this meta-analysis was 59,372, with 4842reporting CVD. As shown in Fig. 2A, habitual snoring was associatedwith an increase risk of CVD (HR 1.26; 95% CI 0.98e1.62) in arandom effect model compared with the non-snorers; however,this positive association was not statistically significant. Significantheterogeneity was observed (I2 ¼ 71%; P ¼ 0.004). Evidence ofpublication bias for studies reporting adjusted HR of CVD was notfound by the Begg’s rank correlation test (P ¼ 0.707) and Egger’slinear regression test (P ¼ 0.377).

Six studies [5e9,12] reported data on CHD. The total number ofsubjects included in this meta-analysis was 53,186, with 3677reporting CHD. As shown in Fig. 2B, habitual snoring was associatedwith an increase risk of CHD (HR 1.15; 95% CI 1.05e1.27) in a fixed-effect model compared with the non-snorers No obvious hetero-geneity was observed (I2 ¼ 11%; P¼ 0.35). There was no evidence ofpublication bias for studies reporting adjusted HR of CHD, as sug-gested by the Begg’s rank correlation test (P ¼ 0.764) and Egger’slinear regression test (P ¼ 0.689).

Six studies [5,6,8e10,12] reported stroke. The total number ofsubjects included in this meta-analysis was 55,587, with 1676reporting stroke. As shown in Fig. 2C, habitual snoring was asso-ciatedwith an increase risk of stroke (HR 1.26; 95% CI 1.11e1.43) in afixed-effect model compared with the non-snorers. No heteroge-neity was observed (I2 ¼ 0%; P ¼ 0.86). There was no evidence ofpublication bias for studies reporting adjusted HR of stroke, assuggested by the Begg’s rank correlation test (P ¼ 1.000) andEgger’s linear regression test (P ¼ 0.139).

Four studies [5e7,10] reported data on all-cause mortality. Thetotal number of participants included in this meta-analysis was13,467, with 748 reporting all-cause mortality. As shown in Fig. 3,habitual snoring was not associated with an increase risk of all-cause mortality (HR 0.98; 95% CI 0.78e1.23) in a random effectmodel compared with the non-snorers. The heterogeneity wasobvious (I2 ¼ 50%; P ¼ 0.09). There was no evidence of publication

Fig. 1. Flow chart of study selection process for meta-analysis.

D. Li et al. / Atherosclerosis 235 (2014) 189e195 191

bias for studies reporting adjusted HR of stroke, as suggested by theBegg’s rank correlation test (P ¼ 0.806) and Egger’s linear regres-sion test (P ¼ 0.594).

Subgroup analysis (Fig. 4A) showed that habitual snoring wasassociated with an increase risk of CHD among women (HR 1.16;95% CI: 1.04e1.30), but not for men (HR 1.16; 95% CI 0.94e1.43).Habitual snoring appeared to increase greater risk of stroke amongmen (HR 1.54; 95% CI: 1.09e2.17) than those in women (HR 1.22;95% CI: 1.05e1.41) (Fig. 4B). Sensitivity analyses were performedbased on CHD and stroke. In the sensitivity analysis, there was littleinfluence in the quantitative pooled measure of HR or 95% CI whenomission of anyone studies (Data not shown).

4. Discussion

This meta-analysis provided evidence that self-reportedhabitual snoring was a mild but statistically significant risk factorfor stroke and CHD, but not for CVD and all-cause mortality. Sub-jects with habitual snoring led to 26% greater risk of stroke and 15%greater risk of CHD. However, there is likely substantial misclassi-fication in the use of a self-reported measure of snoring. The resultsof this meta-analysis may overestimate the total impact of habitualsnoring on cardiovascular disorder. The included studies did not

assess OSAS using polysomnography; therefore, the apparentexcess risk might be explained by the consequences of OSAS ratherthan snoring itself.

Snoring affects both sexes. A male predominance in snoring isfound in the general population [19] and the gender differencemight be related to the differences in pharyngeal collapsibility andcentral respiratory drive [20]. Subgroup analyses based on genderindicated that men with habitual snoring increased risk of stroke(HR 1.54); in contrast, women with habitual snoring were at arelative lower risk (HR 1.22). Habitual snoring appeared to increase16% greater risk of CHD in women; whereas, there was no statis-tically significant risk of CHD in men (HR 1.16; 95% CI: 0.94e1.43).However, due to the small number of studies, the results of sub-group analyses based on gender might be not reliable. Therefore,whether men or women with habitual snoring have more risk ofCVD remains unclear. Age appears to have a significant influence onhabitual snoring and risk of CVD [7]. Due to the limited studiesincluded, we could not conduct further subgroup analysis by age.

A few studies that did not satisfy the inclusion criteria for themeta-analysis also discussed these issues. In a case-controlledstudy [21] involving 400 patients admitted to hospital withstroke, 3.2-fold greater of stroke was observed in regular snorersthan those without snoring. Habitual snoring increased 2.13e2.9-

Table 1Summary of clinical studies included in the meta-analysis.

Study/year Country Type of study Subjects (%men) agerange

Assessment of habitualsnorers

Follow-up (year) Outcome assessment HR or OR (95% CI)/outcome/events number

Adjustment for covariates

Koskenvuoetal.[5]1987

Finnish Prospectivestudy

4388 (100) 40e69 year Questionnaire answer:never; sometimes; often;almost always; and couldnot say. Almostalways þ often snorers vs.never.

3 ICD-8 IHD 410e 414;Stroke: 431e 438. Deathcertificate

2.72 (1.24e5.97) IHD 1493.65 (1.83e7.29) CVD 1871.66 (0.82e3.38) Strokea421.29 (0.92e1.81) Totaldeath 105

Age, BMI, history ofhypertension, smoking, andalcohol. Adjustment age formortality.

Jennum et al.[6]1994

Denmark Prospectivecohort study

804 (51.2) > 70 year Questionnaire answer:never; rarely or hardly;ever or sometimes; alwaysor almost always.Always þ almost alwayssnorers vs. rest

6 ICD-8 IHD 410e 411,413;Stroke: 431e 436. Deathcertificate

0.91 (0.56e1.60) IHD 881.26 (0.70e2.29) Stroke 601.01 (0.73e1.42) Totaldeath 180

Gender, tobacco, alcohol,BMI, social class, physicalactivity, DBP, HDL andcholesterol.

Jennum et al.[7]1995

Denmark Prospectivecohort study

2937 (100) 54e74 year Questionnaire answer:rarely, hardly ever, often,always. Often þ alwayssnorers vs. the rest

6 ICD-8 IHD 410e 414; StrokeICD 431e 438. Deathcertificate

1.0 (0.6e1.6) 54e63 yearsIHD 791.0 (0.7e1.6) 64e74yearsIHD 1030.8 (0.5e1.2) 54e63yearsTotal death 920.7 (0.5e1.0) 64e74yearsTotal death 182

Age, tobacco use, alcoholconsumption, and BMI.

Hu et al. [8]2000

USA Prospectivecohort study

939 (0) 40e65 year Questionnaire answer:regularly, occasionally;never. Regularly snorers vs.never

8 CVD was defined asnonfatal MI, CHD, andstroke. National DeathIndex

1.33 (1.06e 1.67) CVD 10421.33 (1.00e 1.77) CHD 6441.35 (0.91e 1.99) stroke398

Age, smoking, alcohol,physical activity, BMI,menopausal status,parental history ofpremature MI, DB, sleepingduration,hypercholesterolemia andposition.

Elwood et al.[9]2006

UK Prospectivecohort study

1874 (100) 55e 69 year Questionnaire answer:none, mild or infrequent,frequent. Frequent snorersvs. none

10 ICD-10 IHD 121e 5 andstrokes ICD 163e 4

1.53 (1.01e2.23) Stroke 1071.13 (0.84e1.54) IHD 213

Age, social class, smoking,alcohol consumption, BMI,and neck circumference

Yeboah et al.[10]2011

USA Prospectivecohort study

5338 (50.5) 45e 84 year Questionnaire answer:Habitual snorers:snoring� 3e 5 days/week.Snoring� 3e 5 days/weekvs. rest

7.5 CVD is defined asresuscitated cardiac arrest,MI, angina, stroke, stroke,CHD or other CVD death.Death certificate

1.17 (0.81e 1.68) Totaldeath 1891.24 (0.73e 2.13) Stroke 760.91 (0.69e 1.20) CVD 310

Age, gender, race/ethnicity,BMI, smoking, DB, SBP,triglycerides, totalcholesterol, HDL, BP,alcohol, benzodiazepineand statin, use.

Nagayoshi et al.[11]2012

Japan Prospectivecommunity-based study

6513 (63.9) 40e69 year Questionnaire answer:almost every day,sometimes, never, andunknown. Snoring daily vs.never

6 CVD is defined asMI, anginapectoris, and suddencardiac death and stroke.

0.9 (0.4e 1.9) Men CVD 561.9 (0.8e 4.9) Women CVD41

Age, alcohol consumption,cigarette smoking,community; for women,menopausal status, BMI,SBP, antihypertensivemedication, DB, andhypercholesterolemia.

Sands et al.[12]2013

USA Prospectivecohort study

42,244 (0) 50e79 year Questionnaire answer:nonsnorers, occasionalsnorers (< 1e4 times/week), and frequentsnorers (� 5 times/week).Frequent snorers vs.nonsnorers

14.8 CHD is defined as MI, CHDdeath, coronaryrevascularization, andhospitalized angina. CVD isdefined as CHD andischemic stroke.

1.12 (1.01e 1.24) CVD 32471.19 (1.02e 1.40) Stroke9931.14 (1.01e1.28) CHD 2401

Age, race, education,income, smoking,depression, DB, BP, BMI,physical activity, alcoholintake, waist-to-hip ratio,and hyperlipidemia.

Abbreviations: BMI, body mass index; HR, hazard risk; OR, odds ratio; BMI, body mass index; BP, blood pressure; SBP, systolic blood pressure; DBP, diastolic blood pressure; MI, myocardial infarction; DB, diabetes mellitus; CVD,cardiovascular disease; IHD, ischemic heart disease; CHD, coronary heart disease; HDL, high density lipoprotein.

a Frequent snorers vs occasional snorers þ nonsnorers for stroke.

D.Li

etal./

Atherosclerosis

235(2014)

189e195

192

Fig. 2. HR and 95% CI from the included studies of habitual snoring with cardiovascular disease (A), coronary heart disease (B), and stroke (C) comparing habitual snoring to thenon-snorers.

D. Li et al. / Atherosclerosis 235 (2014) 189e195 193

fold greater risk of brain infarction in the caseecontrol study[22,23]. In a prospective study [24] involving 3100 men, self-reported snoring without excessive daytime sleepiness did notincrease risk of mortality. However, ascertaining of snoring byquestionnaire is likely to induce various reporting biases. In aprospective community-based cohort which recruited 397 subjects,snoring (excluding other causes of sleep disturbance by a homesleep apnea monitoring device) was not associated with all-causemortality, incident CVD, or stroke over 17 years follow-up [13].After excluded patients with sleep apnea, these studies found nosuch association.

Exact mechanisms linking habitual snoring to the developmentof cardiovascular disorder are not completely elucidated. A possibleexplanation is that snorers consisted of high proportion of subjectwith OSAS or unmeasured OSAS; whereas severe OSASwas a strongindependent predictor for future CVD [14]. Heavy snoring has beenlinked to an increased carotid intima media thickness and plaque[25,26] and atherosclerosis [27,28]. Therefore, snoring might in-crease the risk of CVD through atherosclerosis. Snoring was alsolinked to hypertension [29,30], type 2 diabetes mellitus [31], andmetabolic syndrome [32,33], all of which may increase greater risk

of cardiovascular disorder. In addition, habitual snoring was asso-ciated with excessive body mass index [34], and the obesity mightpartly mediate the association on snoring and the risk of cardio-vascular disorder.

There are several potential limitations in this study. First, apotential limitation is the reliance of self-reported snoring ratherthan objective measurements. Snoring status was obtained froma self-reported questionnaire or the use of invalidated question-naire. So lack of awareness of snoring or the absence of acohabiting partner could have resulted in misclassification.However, findings suggest self-reported snoring seemed to bereliable tools [35]. Second, a major limitation was the lack of anobjective measure of snoring and OSAS. Snoring might be anearly marker of unmeasured OSAS. However, none of the studieslinked habitual snoring and cardiovascular disorder using poly-somnography or other objective measure to exclude OSAS. Thus,we were unable to determine whether snoring without OSASmight increase the risk of CVD outcomes. Misclassification couldpotentially inflate the risk estimates for snoring. Third, anotherlimitation was the possibility of uncontrolled confounding; theindividual studies did not adjust for potential risk factors in a

Fig. 3. HR and 95% CI from the included studies of habitual snoring with all-cause mortality comparing habitual snoring to the non-snorers.

D. Li et al. / Atherosclerosis 235 (2014) 189e195194

consistent way. The lack of adjustment for these potential con-founding might have resulted in a slight overestimation of therisk estimate.

Snoring is a very common source of complaints from partnersand neighbors. A total of 32.4% of habitual snorers were found tohave OSAS in a middle-aged urban Indian population [36]. Asmall part of snoring peoples might have undetected OSAS. Se-vere OSAS was a definitive risk factor for future CVD and mor-tality [14]. Self-reported habitual snoring could serve as a usefulscreening tool for OSAS. Based on the high prevalence of habitualsnoring in the general population, even a relatively mild

Fig. 4. Subgroup analysis of habitual snoring and risk of cor

independent association with the development of CVD wouldresult in an important risk to the population as a whole [37]. Thecurrent study provided evidence for clinicians that habitualsnoring should be considered in the prevention of CHD andstroke. From this point of view, for persons with cardiovasculardisorders, particularly those with CHD and stroke, asking thequestion ‘have you ever been told that you snore’ may achievemore benefit in the clinical practice. Moreover, further evaluationof the possibility of OSAS is warranted in these high risk pop-ulations. Early detection and intervention of OSAS may help toreduce CHD or stroke risks.

onary heart disease (A) and stroke (B) based on gender.

D. Li et al. / Atherosclerosis 235 (2014) 189e195 195

5. Conclusions

In conclusion, self-reported habitual snoring is a mild but sta-tistically significant risk factor for stroke and CHD, but not for CVDand all-cause mortality. However, whether the risk is attributableto OSAS or habitual snoring alone remains controversial. Morewell-designed prospective studies are required using the objectivemeasurements of snoring and OSAS to confirm whether habitualsnoring is an independent risk factor.

Conflict of interest

The authors report no relationships that could be construed as aconflict of interest.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.atherosclerosis.2014.04.031.

References

[1] Stradling JR, Crosby JH. Predictors and prevalence of obstructive sleep apnoeaand snoring in 1001 middle aged men. Thorax 1991;46:85e90.

[2] Hoffstein V. Apnea and snoring: state of the art and future directions. ActaOtorhinolaryngol Belg 2002;56:205e36.

[3] Hizli O, Ozcan M, Unal A. Evaluation of comorbidities in patients with OSASand simple snoring. ScientificWorldJournal 2013;2013:709292.

[4] Young T, Peppard PE, Gottlieb DJ. Epidemiology of obstructive sleep apnea:a population health perspective. Am J Respir Crit Care Med 2002;165:1217e39.

[5] Koskenvuo M, Kaprio J, Telakivi T, Partinen M, Heikkila K, Sarna S. Snoring as arisk factor for ischaemic heart disease and stroke in men. Br Med J (ClinRes Ed) 1987;294:16e9.

[6] Jennum P, Schultz-Larsen K, Davidsen M, Christensen NJ. Snoring and risk ofstroke and ischaemic heart disease in a 70 year old population. A 6-yearfollow-up study. Int J Epidemiol 1994;23:1159e64.

[7] Jennum P, Hein HO, Suadicani P, Gyntelberg F. Risk of ischemic heart diseasein self-reported snorers. A prospective study of 2,937 men aged 54 to 74years: the Copenhagen Male Study. Chest 1995;108:138e42.

[8] Hu FB, Willett WC, Manson JE, Colditz GA, Rimm EB, Speizer FE, et al. Snoringand risk of cardiovascular disease inwomen. J AmColl Cardiol 2000;35:308e13.

[9] Elwood P, Hack M, Pickering J, Hughes J, Gallacher J. Sleep disturbance, stroke,and heart disease events: evidence from the caerphilly cohort. J EpidemiolCommunity Health 2006;60:69e73.

[10] Yeboah J, Redline S, Johnson C, Tracy R, Ouyang P, Blumenthal RS, et al. Associ-ationbetween sleep apnea, snoring, incident cardiovascular events and all-causemortality in an adult population: MESA. Atherosclerosis 2011;219:963e8.

[11] Nagayoshi M, Tanigawa T, Yamagishi K, Sakurai S, Kitamura A, Kiyama M, et al.Self-reported snoring frequency and incidence of cardiovascular disease: theCirculatory Risk in Communities Study (CIRCS). J Epidemiol 2012;22:295e301.

[12] Sands M, Loucks EB, Lu B, Carskadon MA, Sharkey K, Stefanick M, et al. Self-reported snoring and risk of cardiovascular disease among postmenopausalwomen (from the Women’s Health Initiative). Am J Cardiol 2013;111:540e6.

[13] Marshall NS, Wong KK, Cullen SR, Knuiman MW, Grunstein RR. Snoring is notassociated with all-cause mortality, incident cardiovascular disease, or strokein the Busselton Health Study. Sleep 2012;35:1235e40.

[14] Dong JY, Zhang YH, Qin LQ. Obstructive sleep apnea and cardiovascular risk:meta-analysis of prospective cohort studies. Atherosclerosis 2013;229:489e95.

[15] Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting.Meta-analysis of Observational Studies in Epidemiology (MOOSE) group. J AmMed Assoc 2000;283:2008e12.

[16] Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency inmeta-analyses. Br Med J 2003;327:557e60.

[17] Begg CB, Mazumdar M. Operating characteristics of a rank correlation test forpublication bias. Biometrics 1994;50:1088e101.

[18] Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysisdetected by a simple, graphical test. Br Med J 1997;315:629e34.

[19] Chan CH, Wong BM, Tang JL, Ng DK. Gender difference in snoring and how itchanges with age: systematic review and meta-regression. Sleep Breath2012;16:977e86.

[20] Valipour A. Gender-related differences in the obstructive sleep apnea syn-drome. Pneumologie 2012;66:584e8.

[21] Spriggs DA, French JM, Murdy JM, Curless RH, Bates D, James OF. Snoringincreases the risk of stroke and adversely affects prognosis. Q J Med 1992;83:555e62.

[22] Palomaki H. Snoring and the risk of ischemic brain infarction. Stroke 1991;22:1021e5.

[23] Neau JP, Meurice JC, Paquereau J, Chavagnat JJ, Ingrand P, Gil R. Habitualsnoring as a risk factor for brain infarction. Acta Neurol Scand 1995;92:63e8.

[24] Lindberg E, Janson C, Svardsudd K, Gislason T, Hetta J, Boman G. Increasedmortality among sleepy snorers: a prospective population based study. Tho-rax 1998;53:631e7.

[25] Lee SA, Amis TC, Byth K, Larcos G, Kairaitis K, Robinson TD, et al. Heavysnoring as a cause of carotid artery atherosclerosis. Sleep 2008;31:1207e13.

[26] Li Y, Liu J, Wang W, Yong Q, Zhou G, Wang M, et al. Association of self-reported snoring with carotid artery intima-media thickness and plaque.J Sleep Res 2011;21:87e93.

[27] Drager LF, Lorenzi-Filho G. Heavy snoring and carotid atherosclerosis: is theremore than an association? Sleep 2008;31:1335. discussion 1337.

[28] Leineweber C, Kecklund G, Janszky I, Akerstedt T, Orth-Gomer K. Snoring andprogression of coronary artery disease: the Stockholm Female coronaryAngiography Study. Sleep 2004;27:1344e9.

[29] Young T, Finn L, Hla KM, Morgan B, Palta M. Snoring as part of a dose-responserelationship between sleep-disordered breathing and blood pressure. Sleep1996;19:S202e5.

[30] Kim J, Yi H, Shin KR, Kim JH, Jung KH, Shin C. Snoring as an independent riskfactor for hypertension in the nonobese population: the Korean Health andGenome Study. Am J Hypertens 2007;20:819e24.

[31] Al-Delaimy WK, Manson JE, Willett WC, Stampfer MJ, Hu FB. Snoring as a riskfactor for type II diabetes mellitus: Stockholm Female coronary AngiographyStudy. Am J Epidemiol 2002;155:387e93.

[32] Leineweber C, Kecklund G, Akerstedt T, Janszky I, Orth-Gomer K. Snoring andthe metabolic syndrome in women. Sleep Med 2003;4:531e6.

[33] Troxel WM, Buysse DJ, Matthews KA, Kip KE, Strollo PJ, Hall M, et al. Sleepsymptoms predict the development of the metabolic syndrome. Sleep2010;33:1633e40.

[34] Svensson M, Lindberg E, Naessen T, Janson C. Risk factors associated withsnoring in women with special emphasis on body mass index: a population-based study. Chest 2006;129:933e41.

[35] Telakivi T, Partinen M, Koskenvuo M, Salmi T, Kaprio J. Periodic breathing andhypoxia in snorers and controls: validation of snoring history and associationwith blood pressure and obesity. Acta Neurol Scand 1987;76:69e75.

[36] Reddy EV, Kadhiravan T, Mishra HK, Sreenivas V, Handa KK, Sinha S, et al.Prevalence and risk factors of obstructive sleep apnea among middle-agedurban Indians: a community-based study. Sleep Med 2009;10:913e8.

[37] Chang JL, Kezirian EJ. What are the health risks of untreated snoring withoutobstructive sleep apnea? Laryngoscope 2013;123:1321e2.