Meu espaço Começar anamnese
Biblioteca Piantare · metanálise · 2018

Álcool e apneia do sono: revisão sistemática e metanálise

Alcohol and the risk of sleep apnoea: a systematic review and meta-analysis

Autores Simou E, Britton J, Leonardi-Bee JRevista Sleep medicineAno 2018DOI 10.1016/j.sleep.2017.12.005PubMed 29458744
AEvidência forte. Metanálise, revisão sistemática ou diretriz: o resultado se repete em muitos estudos. Como classificamos
Resumo da Piantare

O que o estudo mostra

A metanálise reuniu 21 estudos epidemiológicos que compararam o consumo de álcool e a ocorrência de apneia do sono, as pausas na respiração durante a noite, em adultos. Consumo maior de álcool se associou a 25% mais chance de apneia do sono. Os autores sugerem que reduzir o álcool pode ter valor preventivo e como apoio ao cuidado da apneia. O limite é a grande variação entre os estudos, com heterogeneidade alta, e o fato de serem observacionais, o que mostra associação, mas não prova causa. Os estudos também mediram consumo e apneia de formas diferentes.

Resumo escrito pelo Time Piantare a partir do estudo, para leitura em português. Não substitui o artigo nem uma consulta. Conteúdo educativo, não é diagnóstico.

Ler o estudo completoLer em portuguêsVer na fonte original

Onde a Piantare cita este estudo

Texto completo

Abstract

Objective

A systematic review and meta-analysis of the association between alcohol consumption and risk of sleep apnoea in adults.

Methods

We searched Medline, EMBASE and Web of Science databases from 1985 to 2015 for comparative epidemiological studies assessing the relation between alcohol consumption and sleep apnoea. Two authors independently screened and extracted data. Random effects meta-analysis was used to estimate pooled effect sizes with 95% confidence intervals (CI). Heterogeneity was quantified using I2 and explored using subgroup analyses based on study exposure and outcome measures, quality, design, adjustment for confounders and geographical location. Publication bias was assessed using a funnel plot and Egger's test.

Results

We identified 21 studies from which estimates of relative risk could be obtained. Meta-analysis of these estimates demonstrated that higher levels of alcohol consumption increased the risk of sleep apnoea by 25% (RR 1.25, 95%CI 1.13–1.38, I2 = 82%, p < 0.0001). This estimate's differences were robust in alcohol consumption and sleep apnoea definitions, study design, and quality but was greater in Low and Middle Income Country locations. We detected evidence of publication bias (p = 0.001). A further eight included studies reported average alcohol consumption in people with and without sleep apnoea. Meta-analysis revealed that mean alcohol intake was two units/week higher in those with sleep apnoea, but this difference was not statistically significant (p = 0.41).

Conclusion

These findings suggest that alcohol consumption is associated with a higher risk of sleep apnoea, further supporting evidence that reducing alcohol intake is of potential therapeutic and preventive value in this condition.

Introduction

Obstructive sleep apnoea (OSA) is a disorder characterised by repeated episodes of partial or complete upper airway collapse or obstruction resulting in hypopnoea or apnoea during sleep [1], [2]. A diagnosis of OSA is confirmed by demonstration of apnoea's (10-s interruptions of breathing) on overnight polysomnography [3], [4], or of apnoea or hypopnoea resulting in a 3% reduction in oxygen saturation on overnight oximetry [1], - occurring five or more times per hour [1], [5]. The frequency of these events [the apnoea index (AI) or apnoea/hypopnoea index (AHI)] is also used to grade the severity of OSA: AHI ≥ 5 to <15 episodes being graded as mild, AHI ≥ 15 to <30 as moderate, and over AHI ≥ 30 as severe [6], [7].

The prevalence of OSA in the USA has been estimated at around 15% in men and 5% in women for people aged 30–70 years [7]. OSA is present in the 41% patients with a body mass index greater than 28 [3]. OSA is also more common in older aged people, current smokers, and in those with coronary artery disease, stroke, hypertension, and diabetes mellitus [4]. OSA is thought to be more common among people who consume alcohol, possibly because alcohol increases upper airway collapsibility [4], [8], [9] and also because alcohol intake can contribute to a higher body mass index. However, evidence on the effect of alcohol consumption on the risk of OSA remains mixed, with some studies reporting an increased risk in those who consume alcohol [5], [10], [11], and others finding a decreased risk [12], [13].

To clarify and quantify the association between alcohol consumption and the risk of OSA we have therefore carried out a systematic review and meta-analysis of comparative epidemiological studies including data on alcohol consumption affects and OSA in adults.

Methods

The systematic review and meta-analysis was conducted in adherence with the PRISMA [14] and MOOSE [15] guidelines. The protocol was published in the National Institute for Health Research International prospective register of systematic reviews (PROSPERO) under the registration number: 42015029910.

Inclusion criteria and search strategy

We included all longitudinal, cohort, case control, and cross sectional studies assessing the association between alcohol consumption and sleep apnoea in adults aged 18 years and over identified from comprehensive searches of Medline (accessed via Ovid), EMBASE (accessed via Ovid), and Web of Science databases between December 1985 and December 2015. We used search filters for observational study designs [16] and search terms for both outcome and exposure developed from relevant Cochrane Reviews groups [17]. We also searched the reference lists of included studies. No language restrictions were imposed; and foreign language papers were translated into English as necessary. The full search term strategy is presented in Supplementary Table 1. Two authors (ES and JL-B) independently screened the titles and abstracts to identify potentially relevant studies. The full text of these studies was then screened independently by the same authors. Any disagreements were resolved through discussion. Studies of populations with HIV, hepatitis B and hepatitis C viruses were excluded, as these represent selected populations, thus may not be generalizable to the general population.

Measures of outcome and exposure

Objective measures of OSA severity were assessed using the Respiratory Distress Index (RDI), where applicable, or Apnoea-Hypopnoea Index (AHI), in which a higher score indicated increased severity. Where possible, an AHI ≥ 5 or RDI ≥ 5 was used to define the presence of OSA, and a score <5 to define absence of OSA.

For categorical measures of alcohol consumption, where possible we defined these using the dichotomy of any alcohol versus no alcohol (reference group). However, for studies which did not report data for those who consumed no alcohol, we used the lowest exposed group as the reference group. Continuous measures of alcohol consumption were standardised, into units per week using recognised definitions: one drink being defined as 0.6 ounces, 14.0 g or 1.2 tablespoons of pure alcohol [18] or using the UK guidelines we defined one unit as eight grams of ethanol [19].

For additional analyses, categorical measures of alcohol consumption were further defined as levels of consumption: light/moderate/heavy drinking; alcohol dependence; or drinks/grams of ethanol per day, week or year. For continuous data we standardized alcohol to grams of ethanol per week. Where alcohol intake was expressed in daily terms, we multiplied these by seven to obtain a per-week figure. Heavy drinking was defined as weekly consumption of 15 or more drinks for men, and eight or more for women, whereas binge drinking either as five or more drinks during a single occasion for men, and four or more for women. Excessive drinking was defined as the presence of either binge or heavy drinking [18]. Also, according to Dietary Guidelines for Americans, moderate alcohol drinking defined as consuming less than one drink to two drinks per day for men and women respectively [20].

Data extraction

Data extraction was carried out in duplicate by ES and JL-B using a previously piloted form. Key data elements extracted included: study design, definition of exposure (alcohol) and outcome (OSA), geographic location, reference population, setting, number of people recruited, demographic of study population, finding, and identified limitations.

Quality assessment

Assessment of methodological quality was carried out using the Newcastle-Ottawa Quality Scale [21], with separate criteria for longitudinal/cohort, case control, and cross sectional studies. The maximum attainable score for longitudinal/cohort and case control studies was nine stars, and for cross sectional studies seven stars. A high quality study was deemed to be identified by a score of at least six. Methodological quality was independently assessed by ES and JL-B, with any disagreement resolved through discussion. Where the same population was used in multiple publications, we included the publication with the highest methodological quality. Due to lack of information, we did not assess the quality of studies published only in abstract forms.

Statistical analysis

Results were extracted from the individual studies as either adjusted measures of effect, crude measures of effect, or using raw data. Effect estimates adjusted for smoking and other factors were used in preference. Binary effect measures were extracted as odds ratios (OR), hazard ratios (HR) or risk ratios, with 95% confidence intervals (CI). Where effect measures were not presented in the paper, we estimated crude RRs for cohort and cross sectional studies and ORs for case control studies. Where exposure to alcohol was reported using quantiles or categories, we extracted adjusted effect measures relating to a comparison of the highest to the lowest exposure group. Continuous effect measures were estimated as mean differences (MD) with 95% CI. Where more than two categories of exposure to alcohol were presented within a study, we combined the categories using standard formula to estimate pooled means and the pooled standard errors [17].

Pooled measures of effect across studies were estimated using random effects meta-analysis using the generic inverse variance method to weight the studies. A random effects model was deemed appropriate due to anticipated differences in effect measures related to the inherent biases within the observational study designs included in the review. We pooled ORs and risk ratios together to estimate pooled Relative Risks (RRs) since the OSA was not assumed to be common (<5%); however HRs were not pooled with other effect measures.

Heterogeneity between the studies was quantified using I2 [22]. As systematic reviews combine studies that are diverse both clinically and methodologically, and it is likely that the strength of association between alcohol and OSA varies by participant characteristics; heterogeneity is anticipated [22]. Where high levels of heterogeneity were seen (I2>50%), subgroup analysis was used according to study quality, study design, adjustment for confounders, ascertainment of alcohol levels (excessive vs low/moderate) and OSA (objective versus subjective) and geographical location of the study population. We also carried out subgroup analyses of studies in which control groups comprised individuals who never consumed alcohol, and those in which the controls were individuals with lower alcohol intake. To assess whether the association between alcohol and OSA was independent of Body Mass Index (BMI), we performed a sensitivity analysis restricted to studies which provided BMI-adjusted measures of effect. We also performed sensitivity analyses restricted to studies which provided smoking-adjusted measures of effect. Funnel plots were used to visually assess evidence of publication bias, where at least ten studies were included in the meta-analysis. Also, Egger's statistical test was used for the assessment of publication bias. Review Manager (version 5.3) and STATA (version 14) were used to perform analyses. P values < 0.05 were taken to represent statistical significance.

Results

Overview of included studies

From an initial 4378 studies identified from the literature searches, 3639 had potentially eligible titles and abstracts, and 178 had potentially eligible full texts. After the full text screening and the exclusion of non-eligible papers, 31 studies met our criteria for inclusion in the systematic review (Fig. 1).

Fig. 1

Flow chart of studies.

Twenty of the included studies used a cross sectional design [10], [11], [13], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37], [38], [39], nine were case controlled [5], [12], [40], [41], [42], [43], [44], [45], [46], and two were cohort studies [47], [48]. Eleven were conducted in Asia (five studies in Korea, two studies in India, one study in Israel, one study in Taiwan, one study in Japan, and one study in Pakistan) [5], [11], [24], [25], [28], [29], [30], [38], [39], [40], [42]; ten in US [13], [23], [27], [34], [35], [36], [37], [44], [47], [48], four in Australia [33], [43], [45], [46]; five in Europe (Denmark, United Kingdom, Ireland, Norway, Germany) [10], [12], [26], [32], [41], and one in Africa [31].

Most studies assessed alcohol consumption by self-report using a standardized questionnaire; one study measured alcohol dependence used International Classification of Diseases (ICD-9) codes [13]. Fifteen studies reported the quantity of alcohol as a continuous measure (number of drinks, units of alcohol or grams of ethanol consumed) and 11 studies used categories [5], [11], [13], [27], [29], [32], [34], [36], [37], [38], [40] including; current drinking [5], [34], [37]; alcohol abuse [40]; alcohol dependence [13], harmful alcohol [30], excessive [32], binge drinking [36], habitual drinking [11], [38], and regular drinking [31]. The remaining five studies did not provide additional information about how alcohol was quantified [25], [28], [29], [42], [46]. Ten of the 31 included studies reported alcohol consumption as a binary exposure, comparing any intake with no intake [5], [24], [25], [27], [28], [29], [34], [39], [42], [47]. The results presented in Table 1.

Results (cont.)

Table 1 Characteristics of the included studies.
Study& YearStudy designGeographical locationPopulationAlcohol ascertainmentAlcohol definitionSleep apnoea ascertainmentSleep apnoea definitionAdjustment
Baik, 2014 [5]Case controlAsia/KoreaGeneral populationSelf-reportCurrent drinkers vs no drinkerspolysomnographyAHI ≥ 15 vs AHI < 5NRG1 genotype, age, sex, BMI, neck circumference
Coughlin, 2004 [12]Case controlEurope/United KingdomPeople recruited from clinic and general publicSelf-reportUnits/week 0–4 units vs 4–50 unitspolysomnographyAHI ≥ 15 vs AHI < 5
Enright, 2001 [23]Cross sectionalNorth America/USElderly populationSelf-report≥25 drinks/weekSelf-report questionnaire–Chronic bronchitis, clinic care, marital status
Fredheim, 2011 [26]Cross sectionalEurope/NorwayPopulation recruited from Obesity/CentreSelf-reportUnits/weekPolysomnographyAHI ≥ 5 vs AHI < 5
Gilat, 2014 [40]Case controlAsia/IsraelGeneral populationSelf-reportAlcohol abusepolysomnography–Matched for: age and sex
Heiskel, 2002 [41]Case controlEurope/ GermanHospital sleep laboratory populationSelf-report7glasses/week vs 8 to>21glasses/ weekPolysomnography–
Hussain, 2009 [24]Cross sectionalAsia/PakistanGeneral populationSelf-reportAlcohol yes vs noSelf-report–Sex, age, BMI, collar size, shift work, nasal blockage, current smoker, family history of snoring
Hwang, 2010 [28]Cross sectionalAsia/TaiwanHospital volunteer helpersSelf-reportAlcohol yes vs alcohol noPolysomnographyAHI ≥ 5 vs AHI < 5
Jennum, 1994 [10]Cross sectionalEurope/DenmarkGeneral populationSelf-reportBeverage: 10grams/day of ethanolRespiratory distress indexRDI ≥ 5 vs RDI < 5
K.Kang, 2013 [42]Case controlAsia/KoreaGeneral populationSelf-reportAlcohol yes vs alcohol noSelf-report (Berlin questionnaire)(High risk vs low risk of OSA)
K.Kang, 2014 [30]Cross sectionalAsia/KoreaGeneral populationSelf-reportHarmful yes vs no harmful alcoholSelf-report Berlin questionnaireHigh risk of OSASex, age, educational level, occupation, shift work, smoking status, exercise, musculoskeletal diseases, diabetes, hypertension
Kang, 2014 [29]Cross sectionalAsia/Koreasleep clinic patientsSelf-reportAlcohol yes vs alcohol noPolysomnographyAHI ≥ 5 vs AHI < 5
Kim, 2006 [25]Cross sectionalAsia/KoreaGeneral population (40–69 years)Self-reportAlcohol yes vs noPolysomnographyAHI ≥ 5 vs AHI<5Age, sex, smoking, BMI, hypertension
Marshall, 2008 [33]Cross sectionalOceania/AustraliaGeneral populationSelf-reportGrams/weekRespiratory distress indexRDI ≥ 5 vs RDI < 5
McArdle, 2006 [43]Case controlOceania/AustraliaSleep clinic of a hospitalSelf-reportGrams/weekPolysomnographyAHI ≥ 15 vs AHI < 5Waist circumference, age
Mc Cague, 2014 [32]Cross sectionalEurope/IrelandhospitalSelf-reportExcessive vs no excessive drinkingPolysomnographyAHI > 30 vs AHI < 5Adjusted (not known the variables)
Ngahan, 2010 [31]Cross sectionalAfrica/AfricaGeneral populationSelf-reportRegular drinkingSelf-report (Berlin questionnaire)(High risk vs low risk of OSA)Waist circumference, snoring
Nieto, 2000 [27]Cross sectionalNorth America/USMiddle aged and elderly populationSelf-reportAlcohol no vs >7drinks/weekPolysomnographyAHI ≥ 5 vs AHI < 5
Yue, 2014 [34]Cross sectionalNorth America/USGeneral populationSelf-reportCurrent alcohol use(within the past 12 months) vs neverSelf-report questionnaire–Gender, age, race, marital status, income, education, asthma, diabetes, hypertension
Peppard, 2000 [35]Cross sectionalNorth America/USGeneral populationSelf-reportDrinks/weekPolysomnographyAHI ≥ 5 vs AHI < 5
Peppard, 2006 [47]CohortNorth America/USGeneral populationSelf-reportDrinks/day 0 drinks vs 3 drinksPolysomnographyAHI ≥ 5 vs AHI < 5Age, BMI, neck and waist circumference, current cigarette smoking antidepressant, antihypertensive, cholesterol-lowering medication
Popovici, 2013 [36]Cross sectionalNorth America/USGeneral populationSelf-reportBinge drinkingSelf-report–Age, race/ethnicity, years of schooling, current employment and marital status, number of persons in the household, labor market income in the past year, and being born outside the U.S, smoking, diabetes, hypertension, depression, panic disorder, psychological or emotional abuse
Shamsuzzaman, 2014 [44]Case ControlNorth America/USSleep laboratory patientsSelf-reportDrinks/weekPolysomnographyAHI ≥ 5 vs AHI < 5
Sharafkhaneh, 2005 [13]Cross sectionalNorth America/USGeneral populationICD 9 codesAlcohol dependenceICD 9 codes–Age, sex, and ethnicity
Sharma, 2006 [11]Cross sectionalAsia/IndiaGeneral populationSelf-reportHabitual drinking (>100gr/day) yes vs no habitualPolysomnographyAHI ≥ 5 vs AHI < 5
Simpson, 2015 [45]Case controlOceania/AustraliaCases form sleep study/Controls from populationSelf-reportGrams/weekPolysomnographyAHI ≥ 15 vs AHI < 15
Udwadi, 2003 [38]Cross sectionalAsia/IndiaGeneral populationSelf-reportHabitual vs no habitual drinkingPolysomnographyAHI ≥ 5 vs AHI < 5
Wetter, 2004 [48]CohortNorth America/USGeneral populationSelf-reportDrinks/weekPolysomnographyAHI ≥ 5 vs AHI < 5
Worsnop 1998 [46]Case controlOceania/AustraliaGeneral populationSelf-report–polysomnographyAHI
Yaggi, 2005 [37]Cross sectionalNorth America/USMiddle aged population from sleep centreSelf-reportCurrent vs no current alcoholpolysomnographyAHI ≥ 5 vs AHI < 5
Zenda, 2014 [39]Cross sectionalAsia/JapanGeneral populationSelf-reportAlcohol yes (150 grams/week) vs alcohol noPolysomnographyAHI ≥ 5 vs AHI < 5

Twenty three studies ascertained sleep apnoea using polysomnography either measuring its presence and severity using the Apnoea-Hypopnoea Index (AHI) [5], [11], [12], [25], [26], [27], [28], [29], [32], [35], [37], [38], [39], [40], [41], [43], [44], [45], [46], [47], [48] or the Respiratory distress index (RDI) [10], [33]. Seven studies relied on self-report [23], [24], [30], [31], [34], [36], [42] and in the remaining study, sleep apnoea was ascertained using ICD codes [13].

The methodological quality using the Newcastle-Ottawa Assessment Scale (Table 2) showed that quality of the included studies ranged from two to seven, with only one study deemed to be of high quality (overall score ≥ 6) [47]. Thirteen studies reported effect estimates adjusted for confounders [5], [13], [23], [24], [25], [30], [31], [32], [34], [36], [40], [43], [47]; with six adjusted for smoking [5], [24], [25], [30], [35], [47]. Lower scores were primarily due to lack of adjustment for confounding factors, instead using selective populations or a self-report assessment of alcohol exposure.

Results (cont.)

Table 2 Critical appraisal of included studies using Newcastle Ottawa scale.
Study, YearStars number
SelectionbComparabilitycExposured
Baik, 2014 [5]311
Coughlin, 2004 [12]201
Enright, 2001 [23]212
Fredheim, 2011 [26]102
Gilat, 2014 [40]311
Heiskel, 2002 [41]202
Hussain, 2009 [24]112
Hwang, 2010 [28]101
Jennum, 1994 [10]202
K.Kang, 2013 [42]302
K.Kang, 2014 [30]122
Kang, 2014 [29]102
Kim, 2006 [25]121
Marshall, 2008 [33]201
McArdle, 2006 [43]311
Mc Cague, 2014a[32]–––
Ngahane, 2010 [31]–––
Nieto, 2000 [27]
Pan, 2014 [34]112
Peppard, 2000 [35]
Peppard, 2006 [47]222
Popovici, 2013 [36]
Shamsuzzaman, 2014 [44]301
Sharafkhaneh, 2005 [13]212
Sharma, 2006 [11]202
Simpson, 2015 [45]200
Udwadi, 2003 [38]102
Wetter, 2004 [48]201
Worsnop 1998 [46]301
Yaggi, 2005 [37]102
Zenda, 2014 [39]201

Meta-analysis findings

A pooled analysis of the 21 studies from which relative risks could be estimated found the overall risk of OSA to be increased by 25% in people who consumed alcohol at all, or in higher amounts, than those who consumed no, or lower amounts of alcohol, respectively (pooled RR 1.25, 95% CI 1.13 to 1.38, I2 = 82%; Fig. 2). A stratified analysis based on the effect measure used found very similar results for studies reporting ORs (1.26, 95% CI 1.04 to 1.53, I2 = 79, 14 studies) and those reported risk ratios (1.27, 95% CI 1.08 to 1.49, I2 = 86%, seven studies).

Fig. 2

Forest plot of the association between alcohol consumption versus non-alcohol/lower alcohol consumption and the risk of OSA. *A pooled analysis of 22 comparisons from 21 studies; one study reported estimates separately for men and women [13].

A further eight studies presented average levels of consumption of alcohol in people with and without OSA [10], [26], [33], [35], [43], [44], [45], [48]. A meta-analysis of these studies found that people with OSA consumed, on average, approximately two units alcohol per week more than people without OSA (pooled mean difference 1.93, 95% CI -2.62 to 6.49, I2 = 66%, Fig. 3); however, the effect was not significant (p = 0.41). Data from the remaining two studies [36], [46] were reported in insufficient detail for inclusion in either of the above meta-analyses: one of these studies found a 9% increased risk of OSA in those who consumed alcohol compared to those with no alcohol consumption [36] while the other reported no significant association (p = 0.82) [46].

Fig. 3

Forest plot of the association between levels of alcohol consumption in people with and without OSA.

Among the 21 studies included in the pooled analysis of relative risks a sensitivity analysis showed that relative to people who did not consume any alcohol, those who consumed alcohol had 25% higher risk in OSA (pooled RR 1.25, 95%CI 1.08 to 1.45, I2 = 69%, 10 studies). However, in studies comparing any alcohol intake to low intake the risk of OSA was increased by 29% (pooled RR 1.29, 95%CI 1.07 to 1.55, I2 = 87%, 11 studies) (Fig. 2). A further sensitivity analysis restricted to the four studies which provided BMI-adjusted estimates [5], [24], [25], [47] found similar results to the main analysis, with a 41% increased risk of OSA in people who consume alcohol at all or in higher amounts (pooled RR 1.41, 95% CI 1.13 to 1.75, I2 = 0%, four studies). Moreover, studies which provided smoking-adjusted estimates [5], [24], [25], [47] found a marginally larger magnitude of effect compared to the main analysis (pooled RR 1.56, 95% CI 1.38 to 1.76, I2 = 0%, four studies).

Subgroup analyses of the 21 studies found that methodological quality (high versus low; p = 0.69), study design (cross-sectional, case-control, longitudinal/cohort; p = 0.25), adjustment for confounders (p = 0.40), level of alcohol consumption (excessive vs moderate/low; p = 0.65) and ascertainment of sleep apnoea (subjective vs objective; p = 0.26) did not explain high level of heterogeneity (Supplementary Table 2). Furthermore, the geographic location of the study was found to explain heterogeneity between the studies, with a quantitative interaction seen between High and Low/Middle Income Countries (p-value for subgroup differences = 0.01, Fig. 4) relating to a larger effect size in Low/Middle Income Countries (pooled RR 1.47, 95% CI: 1.17–1.85) compared to High Income Countries (pooled RR 1.07, 95% CI: 0.98–1.17). Evidence of publication bias was detected in the meta-analysis of 21 studies (Egger's test, p = 0.001).

Fig. 4

Forest plot of alcohol consumption and the risk of OSA; subgroup analysis based on high vs low/middle income countries.

There were insufficient data to combine study results in an exposure-response analysis, although one study estimated that for every drink consumed per day there was a 25% increased odds of at least mild sleep disordered breathing in men (p for trend = 0.006) [47]. In contrast, there was no significant exposure-response relation in women (p for trend = 0.73).

Discussion

Alcohol consumption is associated with a range of important health consequences, and a history of alcohol consumption is common among people with OSA. It is plausible that alcohol increases the risk of OSA because alcohol consumption reduces genioglossal muscle tone, predisposing patients to upper airway collapse [49] and generally increasing upper airway resistance [50]. High alcohol intakes also contribute to dietary energy intake, and hence in some cases a high body mass index, which is itself a risk factor for OSA [4]. However the extensive scientific literature of studies of the association between OSA and alcohol intake provides mixed evidence on the qualitative and quantitative nature of this association, making systematic review and meta-analysis likely to be particularly helpful in this context.

To our knowledge, this paper reports the first meta-analysis combining all available worldwide literature assessing the association between alcohol consumption and the risk of OSA, and demonstrates that people who consume alcohol, either in relation to no exposure or in those with relatively high intakes compared to those with low intakes, are approximately 25% more likely to have OSA.

Inevitably, our review findings are limited by publication bias and the low methodologic quality of the included studies, but they at least provide a comprehensive and systematic overview and synthesis of the available worldwide evidence. Another limitation of our study is the fact that only 13 of the 31 included studies adjusted for confounders; however, no appreciable difference was seen between studies which did and did not adjust for confounders. Our analysis also highlighted a high level of heterogeneity between study findings, which subgroup analysis indicated was related to economic development, with the association between alcohol and OSA appearing to be stronger in countries of low or middle income. The explanation for this finding is not clear, but confounding by obesity or other lifestyle factors is a potential factor since alcohol consumption is more likely to be limited to the most affluent in low and middle income settings. Although we expected the association between alcohol and OSA to be confounded by obesity, our restricted analysis of studies presenting BMI-adjusted estimates confirmed an independent effect of alcohol which was in fact slightly stronger than the unadjusted estimate, indicating that obesity is not the main mechanism responsible for the higher risk of OSA among those who consume alcohol.

A previous narrative review published in 2005 also found that sleep disordered breathing was likely to be more common in relation to alcohol consumption, with heavy drinkers being at particularly high risk of OSA [51]. However that study also highlighted that whilst two or three drinks before bedtime can promote sleep, this effect is not sustained after a few days of regular consumption. With more sustained regular alcohol intake, and with alcohol intake earlier in the day, alcohol consumption can be associated with insomnia [51]. It therefore appears likely that the timing and regularity of alcohol consumption are both important to the effect of alcohol on OSA, since airway muscle relaxation and reduced sensitivity to apnoea are both likely to be greatest when alcohol levels are rising, as for example after bedtime consumption [52], [53]. Our data do not enable us to estimate the nature of the exposure-response relation for either alcohol at bedtime, or alcohol in general (data being limited to single study [47]), on the occurrence or severity of OSA, however, these questions could in principle be resolved by randomised trials. We also searched for randomised controlled trials on alcohol and sleep apnoea but found none. Therefore, there is a need for the conduction of future clinical trials in order to address this important question.

Our findings thus provide confirmation that alcohol consumption may cause or exacerbate OSA, and the wider literature suggests that this may especially be the case with alcohol consumed shortly before bedtime. Whilst our findings do not confirm or refute the causality of this association, health care professionals might consider advising against bedtime alcohol among people with, or at risk of, OSA.

Referências do artigo (53)
  1. 1 Flemons W. Buysse D. Redline S. Sleep-related breathing disorders in adults Sleep 22 5 1999 667 689 10450601
  2. 2 McNicholas W. Public health and medicolegal implications of sleep apnoea Eur Respir J 20 6 2002 1594 1609 12503722 10.1183/09031936.02.00404502
  3. 3 Park J.G. Ramar K. Olson E.J. Updates on definition, consequences, and management of obstructive sleep apnea Mayo Clin Proc 86 6 2011 Jun 30 549 554 21628617 10.4065/mcp.2010.0810 PMC3104914
  4. 4 Punjabi N.M. The epidemiology of adult obstructive sleep apnea Proc Am Thorac Soc 5 2 2008 136 143 18250205 10.1513/pats.200709-155MG PMC2645248
  5. 5 Baik I. Seo H.S. Yoon D. Associations of sleep apnea, NRG1 polymorphisms, alcohol consumption, and cerebral white matter hyperintensities: analysis with genome-wide association data Sleep 38 7 2015 1137 1143 25325441 10.5665/sleep.4830 PMC4481000
  6. 6 Epstein L.J. Kristo D. Strollo P.J. Jr. Clinical guideline for the evaluation, management and long-term care of obstructive sleep apnea in adults J Clin Sleep Med 5 3 2009 263 276 19960649 PMC2699173
  7. 7 Jonas D.E. Amick H.R. Feltner C. Screening for obstructive sleep apnea in adults: evidence report and systematic review for the US preventive services task force Jama 317 4 2017 415 433 28118460 10.1001/jama.2016.19635
  8. 8 Franklin K.A. Lindberg E. Obstructive sleep apnea is a common disorder in the population—a review on the epidemiology of sleep apnea J Thorac Dis 7 8 2015 1311 26380759 10.3978/j.issn.2072-1439.2015.06.11 PMC4561280
  9. 9 Mitler M.M. Dawson A. Henriksen S.J. Bedtime ethanol increases resistance of upper airways and produces sleep apneas in asymptomatic snorers Alcohol Clin Exp Res 12 6 1988 801 805 3064641 10.1111/j.1530-0277.1988.tb01349.x PMC2336897
  10. 10 Poul Jennum A.S. Self- assessed cognitive function in snorers and sleep apneics Eur Neurol 34 4 1994 204 208 8082678 10.1159/000117039
  11. 11 Sharma S.K. Kumpawat S. Banga A. Prevalence and risk factors of obstructive sleep apnea syndrome in a population of Delhi, India Chest 130 1 2006 149 156 16840395 10.1378/chest.130.1.149
  12. 12 Coughlin S.R. Mawdsley L. Mugarza J.A. Obstructive sleep apnoea is independently associated with an increased prevalence of metabolic syndrome Eur Heart J 25 9 2004 735 741 15120883 10.1016/j.ehj.2004.02.021
  13. 13 Sharafkhaneh A. Giray N. Richardson P. Association of psychiatric disorders and sleep apnea in a large cohort Sleep 28 11 2005 1405 1411 16335330 10.1093/sleep/28.11.1405
  14. 14 Moher D. Liberati A. Tetzlaff J. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement Ann Intern Med 151 4 2009 264 269 19622511 10.7326/0003-4819-151-4-200908180-00135
  15. 15 Stroup D.F. Berlin J.A. Morton S.C. Meta-analysis of observational studies in epidemiology: a proposal for reporting J Am Med Assoc 283 15 2000 2008 2012 10.1001/jama.283.15.2008 10789670
  16. 16 SIGN. Search Filters, Obseravtional studies. Available from: http://www.sign.ac.uk/methodology/filters.html#obs [Accessed 4 December 2015].
  17. 17 Cochrane Library. Available from: http://www.cochranelibrary.com/ [Accessed 4 December 2015].
  18. 18 Centers for Disease Control and Prevention (CDC). Alcohol and Public Health, Fact Sheets-Alcohol Use and Your Health. https://www.cdc.gov/alcohol/fact-sheets/alcohol-use.htm [Accessed 10 December 2015].
  19. 19 Drinkware, what is an alcohol unit. Available from: https://www.drinkaware.co.uk/alcohol-facts/alcoholic-drinks-units/what-is-an-alcohol-unit/ [Accessed 15 December 2015].
  20. 20 U.S. Department of Health and Human Services and U.S. Department of Agriculture 2015 – 2020. Dietary Guidelines for Americans 8th ed. 2015 Washington, DC https://health.gov/dietaryguidelines/2015/resources/2015-2020_Dietary_Guidelines.pdf
  21. 21 Wells G, Shea B, O’connell D, et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. Available from: http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp [Accessed 10 February 2016].
  22. 22 Higgins J.P. Thompson S.G. Deeks J.J. Measuring inconsistency in meta-analyses Bmj 327 7414 2003 557 560 12958120 10.1136/bmj.327.7414.557 PMC192859
  23. 23 Enright P.L. Newman A.B. Wahl P.W. Prevalence and correlates of snoring and observed apneas in 5,201 older adults Sleep 19 7 1996 531 538 8899931 10.1093/sleep/19.7.531
  24. 24 Hussain S.F. Cloonan Y.K. Islam M. Prevalence and associated risk factors of sleep-disordered breathing symptoms in young and middle-aged Pakistani employed adults Sleep Breath 14 2 2010 137 144 19760442 10.1007/s11325-009-0299-6
  25. 25 In K. Kim J. You S. Prevalence of sleep-disordered breathing in middle-aged Korean men and women Am J Respir Crit Care Med 170 10 2004 1108 1113 15347562 10.1164/rccm.200404-519OC
  26. 26 Fredheim JM Rollheim J Omland T Type 2 diabetes and pre-diabetes are associated with obstructive sleep apnea in extremely obese subjects: a cross-sectional study Cardiovasc Diabetol 10 2011 2 9 21943153 10.1186/1475-2840-10-84 PMC3206416
  27. 27 Nieto F.J. Young T.B. Lind B.K. Association of sleep-disordered breathing, sleep apnea, and hypertension in a large community-based study JAMA 283 14 2000 Apr 12 1829 1836 10770144 10.1001/jama.283.14.1829
  28. 28 Hwang JH Chen JC Hsu CJ Association of obstructive sleep apnea and auditory dysfunctions in older subjects Otolaryngol Head and Neck Surg 144 2010 114 119 10.1177/0194599810390859 21493399
  29. 29 Kang H.H. Kang J.Y. Ha J.H. The associations between anthropometric indices and obstructive sleep apnea in a Korean population PLoS One 9 12 2014 e114463 10.1371/journal.pone.0114463 PMC4256422 25474257
  30. 30 Kang K. Seo J.G. Seo S.H. Prevalence and related factors for high-risk of obstructive sleep apnea in a large Korean population: results of a questionnaire-based study J Clin Neurol 10 1 2014 42 49 24465262 10.3988/jcn.2014.10.1.42 PMC3896648
  31. 31 Mbatchou Ngahane B. Dzudie A. Nganda M. Prevalence and factors associated with high risk of sleep apnea in hypertensive patients in a sub-saharan africa setting Am J Resp Crit Care Med 191 2015 A5043
  32. 32 Mc Cague S. Kent B.D. Nolan G. The relationship of alcohol consumption to disease severity in a sleep apnoea cohort Ir J Med Sci 181 2012 S448-S
  33. 33 Marshall NS Wong KK Liu PY Sleep apnea as an independent risk factor for all-cause mortality: the Busselton health study Sleep 31 8 2008 1079 1085 18714779 PMC2542953
  34. 34 Pan Y. Wang W. Wang K.S. Associations of alcohol consumption and chronic diseases with sleep apnea among US adults Int J High Risk Behav Addiction 3 2 2014 10.5812/ijhrba.19088 PMC4080510 25032163
  35. 35 Peppard P.E. Young T. Palta M. Prospective study of the association between sleep-disordered breathing and hypertension N Engl J Med 342 19 2000 1378 1384 10805822 10.1056/NEJM200005113421901
  36. 36 Popovici I. French M.T. Binge drinking and sleep problems among young adults Drug Alcohol Depend 132 1–2 2013 207 215 23466223 10.1016/j.drugalcdep.2013.02.001 PMC3748176
  37. 37 Yaggi H.K. Concato J. Kernan W.N. Obstructive sleep apnea as a risk factor for stroke and death N Engl J Med 353 19 2005 2034 2041 16282178 10.1056/NEJMoa043104
  38. 38 Udwadia AVD Zarir F. Lonkar Sharmila G. Singh Chandrajeet I. Prevalence of Sleep-disordered breathing and sleep apnea in middle-aged urban indian men Am J Respir Crit Care Med 169 2003 168 173 14604837 10.1164/rccm.200302-265OC
  39. 39 Zenda T. Hamazaki K. Oka R. Endoscopic assessment of reflux esophagitis concurrent with hiatal hernia in male Japanese patients with obstructive sleep apnea Scand J Gastroenterol 49 9 2014 1035 1043 25048181 10.3109/00365521.2014.926984
  40. 40 Gilat H. Vinker S. Buda I. Obstructive sleep apnea and cardiovascular comorbidities: a large epidemiologic study Medicine (United States) 93 9 2014 e45 10.1097/MD.0000000000000045 PMC4602425 25144324
  41. 41 Heiskel H. Gunzenhauser D. Seidler A. Sleep apnea and occupational exposure to solvents Scand J Work Environ Health 28 4 2002 249 255 12199426 10.5271/sjweh.672
  42. 42 Kang K. Park K.S. Kim J.E. Usefulness of the Berlin Questionnaire to identify patients at high risk for obstructive sleep apnea: a population-based door-to-door study Sleep Breath 17 2 2013 May 1 803 810 23054593 10.1007/s11325-012-0767-2
  43. 43 McArdle N. Hillman D. Beilin L. Metabolic risk factors for vascular disease in obstructive sleep apnea - a matched controlled study Am J Respir Crit Care Med 175 2 2007 190 195 17068329 10.1164/rccm.200602-270OC
  44. 44 Shamsuzzaman A. Amin R.S. Calvin A.D. Severity of obstructive sleep apnea is associated with elevated plasma fibrinogen in otherwise healthy patients Sleep Breath 18 4 2014 761 766 24510497 10.1007/s11325-014-0938-4 PMC4384427
  45. 45 Simpson L. McArdle N. Eastwood P.R. Physical inactivity is associated with moderate-severe obstructive sleep apnea J Clin Sleep Med 11 10 2015 1091 1099A 26285117 10.5664/jcsm.5078 PMC4582050
  46. 46 Worsnop C.J. Naughton M.T. Barter C.E. The prevalence of obstructive sleep apnea in hypertensives Am J Respir Crit Care Med 157 1 1998 111 115 9445287 10.1164/ajrccm.157.1.9609063
  47. 47 Peppard P.E. Austin D. Brown R.L. Association of alcohol consumption and sleep disordered breathing in men and women J Clin Sleep Med 3 3 2007 265 270 17561593 PMC2564771
  48. 48 Wetter D.W. Young T.B. Bidwell T.R. Smoking as a risk factor for sleep-disordered breathing Arch Intern Med 154 19 1994 2219 2224 7944843
  49. 49 Krol R. Knuth S. Bartlett D. Jr. Selective reduction of genioglossal muscle activity by alcohol in normal human subjects Am Rev Respir Dis 129 2 1984 247 250 6421210
  50. 50 Robinson R.W. White D.P. Zwillich C.W. Moderate alcohol ingestion increases upper airway resistance in normal subjects 1–3 Am Rev Respir Dis 132 6 1985 1238 1241 4073664 10.1164/arrd.1985.132.6.1238
  51. 51 Stein M.D. Friedmann P.D. Disturbed sleep and its relationship to alcohol use Subst Abuse 26 1 2006 1 13 10.1300/j465v26n01_01 PMC2775419 16492658
  52. 52 Dawson A. Lehr P. Bigby B.G. Effect of bedtime ethanol on total inspiratory resistance and respiratory drive in normal nonsnoring men Alcohol Clin Exp Res 17 2 1993 256 262 8488964 10.1111/j.1530-0277.1993.tb00759.x PMC2426749
  53. 53 Scrima L. Broudy M. Nay K.N. Increased severity of obstructive sleep apnea after bedtime alcohol ingestion: diagnostic potential and proposed mechanism of action Sleep 5 4 1982 Sep 1 318 328 7163721 10.1093/sleep/5.4.318

Simou E et al.. Alcohol and the risk of sleep apnoea: a systematic review and meta-analysis. Sleep medicine, 2018. Publicado em acesso aberto sob a licença CC BY 4.0; original em doi.org/10.1016/j.sleep.2017.12.005. A Piantare reproduz o texto com formatação adaptada, sem mudar o conteúdo; a tradução para o português é uma adaptação automática. A reprodução não indica endosso dos autores à Piantare.

Da ciência para o seu dia

Quer entender o que isso muda para você?

A anamnese da Piantare leva 6 minutos e mostra por onde começar. Se ficou uma dúvida, o Time Piantare responde.

Começar a anamnesePergunte ao EspecialistaVoltar à Biblioteca