Alcohol and HypertensionAn UpdateLawrence J. Beilin, Ian B. PuddeyGiventhesocial significanceofalcohol worldwideit isnot surprising that there is continuing strong interest intherelationbetweenalcohol andhypertension. Recent re-search continues to address unresolved questions concerningthe balance between the medical hazards and the cardiovas-cular benefits of alcohol, the possible significance of differenttypes of beverage, the role of different drinking patternsfor cardiovascular morbidity and mortality, and mechanismsunderlyingthepressor effects of ethanol. Theseissues allneedtobe consideredinthe context of social aspects ofdrinking and effects on noncardiovascular morbidity andmortality. Epidemiologic data relating the type or quantity ofalcohol consumption to blood pressure or cardiovasculardisease needs to be viewed with circumspection fromanumber of viewpoints. First, heavier drinkers or problemsdrinkers are far less likely to participate in surveys thanothers. Second, alcohol consumption past or present is noto-riously underreported. Third, drinking patterns are difficult toquantify and often not reported at all. The type, quantity, andpatternofdrinkingareall highlycorrelatedwithsocioeco-nomic and other lifestyle behaviors,1many of which may notbe measured or not measured accurately enough for adequateadjustments in statistical models. A relation between averageweeklyalcohol consumption, bloodpressurelevel, andhy-pertension prevalence has been consistent worldwide andcontinues to be studied in different populations to evaluate itscontributioninrelationtoother riskfactors. However, be-causeof theaboveissues somecaveats shouldbeput oninterpretation of data where authors emphatically implycausal relationships from statistical associations.The more recent cross-sectional studies have concentratedon the effects of pattern of drinking and the consumption ofalcohol with or without food, beverage type, and the relativeeffects of alcohol on hypertension subtypes.Pattern of Drinking, Beverage Type, andEffects on Hypertension SubtypesThe relationship between the pattern of alcohol drinking andtheriskofhypertensionwasaddressedinacross-sectionalstudyof2609NewYorkersfreefromothercardiovasculardisease2and considered further in an accompanying editorial3in which some of the key issues surrounding the topic werediscussed. Compared with lifetime abstainers, those whoreporteddrinkingonadailybasisorapartfromfoodhadasignificantly higher risk of hypertension, but this effectdisappeared after accounting for the amount of alcoholconsumedintheprevious30days. Whencurrent drinkersonlywerestudiedandadjustment madefor theamount ofalcohol consumed in the previous 30 days, the risk of beinghypertensiveincreased(64%) onlyinthosewhosaidtheydrank without food. Preference for any one type of beveragedidnotinfluencetheassociationbuttheimportantpotentialconfounderofdietaryhabitswasnotassessed.Neverthelessthisfindingwasconsistent withanearlier studyof Italianwine drinkers coming from a different cultural background,4andafollow-upinthatpopulationshoweddrinkingalcoholoutsideof meals was associatedwithhigher riskof deathfrom all causes and cardiovascular disease. Similar relation-ships were seen between drinking outside of meals or snacksandincreasedriskofmyocardialinfarctioninarecentcasecontrol study in men.5Thepossibleeffectsofindividualalcoholicbeveragesonblood pressure continue to arouse interest. However, thepreviouslycitedstudyfromwesternNewYork3foundnoconsistent beverage-specificassociations withhypertensionriskinNorthAmericansdrinkingbeer, wine, or spirits. Across-sectional study in Chinese men that examined theassociations between alcohol intake and isolated systolic,combined systolic and diastolic, and isolated diastolic hyper-tensionfoundthat thoseinthehighest alcohol intakecate-gory (30 drinks/week) were twice as likely as nondrinkerstohaveanyofthesehypertensivesubtypeswithpopulationattributable risks of 13.9%, 13.4% and 12.0%, respectively.6In this Chinese study,6liquor drinking was associated with ahigheroddsratioofisolatedsystolichypertension, but thisfindingwasprobablybecausetheliquordrinkersgenerallydrankmorealcohol. AnotherstudygroupedJapanesemaleworkers7on the basis of their total consumption of beer, sake,shochu (traditional Japanese spirits), whiskey, or wine. Bloodpressurewas highest intheshochugroupbut ananalysisadjusting for total alcohol consumption resulted in disappear-ance of this difference. The interest in the effects of specificReceived January 18, 2006; first decision January 26, 2006; revision accepted March 8, 2006.From the Royal Perth Hospital Unit (L.J.B.), School of Medicine & Pharmacology, and the Faculty of Medicine and Dentistry and Health Sciences(I.B.P.), University of Western Australia, Australia.CorrespondencetoLawrenceJ.Beilin,ProfessorofMedicine,RoyalPerthHospitalUnit,SchoolofMedicine&Pharmacology,UniversityofWesternAustralia, Level 4, Medical Research Foundation Bldg, Rear, 50 Murray St, Perth, Western Australia, Australia. E-mail Lawrie.Beilin@uwa.edu.au(Hypertension. 2006;47:1-4.) 2006 American Heart Association, Inc.Hypertension is available at http://www.hypertensionaha.org DOI: 10.1161/01.HYP.0000218586.21932.3c1beverageshasbeenevokedinpart bytheso-calledFrenchParadox of a relatively low incidence of coronary disease inFrance despite a high intake of saturated fat, a phenomenonthathasbeenattributedtotheconsumptionofredwine.8Ithas also been suggested that wine drinkers may be protectedfrom the blood pressureraising effects of regular moderate toheavyalcoholconsumption, perhapsbecauseofantioxidantand vasodilator effects of polyphenolic flavonoids improvingendothelial function.9In our opinion, these suggestions havebeen laid to rest in a randomized crossover trial10thatconfirmed suggestions from population studies that moderatealcohol consumption raises blood pressure regardless ofsource. In that study, normotensive men showed similarelevationsofawakeambulatorysystolicbloodpressureandheart rateafter 4weeks of either beer or redwine(40-gethanol equivalent per day) compared with a control-abstinence period. De-alcoholized red wine had no effect onblood pressure, and neither this beverage nor alcohol contain-ingredwinehadanyeffect onfloworglyceryl trinitratemediateddilation. Itwasconcludedthatresultsfrompopu-lation studies suggesting differential effects of red winecompared with other beverages on blood pressure were mostlikelybecauseofconfoundinglifestyledifferencesinwinedrinkers. Twenty-fourhour endothelin-1excretionwasin-creased with beer and wine drinking, leading to the sugges-tionthat thismight reflect increasedvascular endothelin-1productionasatleastacontributortothepressoreffectsofalcohol.10Asystematic review11of alcohol intervention studiesconfirmed the previous findings of an initial meta-analysis byXinet al12withsimilar estimatesfor theeffect of alcoholrestriction to reduce systolic and diastolic blood pressure by2.7mm Hgand1.4mm Hg, respectively. However, theseauthorsalsocompareddatafromstudiesthat usedconven-tional clinic or office review of blood pressure with those thatincorporated ambulatory or home blood pressure monitoring,highlightingbiphasiceffects of alcohol onbloodpressurewith an early presumably vasodilator effect of alcohol leadingto a reduction in blood pressure (in the immediate hours afterexposure)andalatereffect (thenext day)ofraisingbloodpressure.Effects on Large Vessel Structureand FunctionHow might alcohol exert its coronary protective effects overand above influences on high-density lipoprotein (HDL)cholesterol, platelet function, and fibrinogen and in the faceof adverse effects on blood pressure and homocysteinemetabolism? Cross-sectional data from Holland described aninverse or J-shaped relation between alcohol intake andmeasures of aortic stiffness in middle-aged and older men andpre- and postmenopausal women13,14but not in youngermen.15They recognized that these results might be con-foundedbyotherlifestylefactorsbutsuggestedthatiftheywere causallylinkeddirect effects of alcohol toimprovevascular compliance might contribute to any cardioprotectiveeffect. As discussed in an accompanying editorial,16thefindings are not easytoreconcile withthe associationofalcohol with isolated systolic hypertension6nor with a 9-yearlongitudinal studyshowinganassociationbetweenalcoholconsumptionandincreasedaorticstiffness inmiddle-agedJapanese men who were initially free of aortic stiffness.17Alcohol, the Metabolic Syndrome,and DiabetesThe nature of the association between liver enzymes, body fatdistribution, alcoholconsumption, andtheriskofhyperten-sionhas beenexploredfurther intheWesternNewYorklongitudinal studyof 1455menandwomen.18It hadbeenassumed that the increased glutamyltransferase (GT)levels seen in hypertensive subjects were caused alcohol. Inthis 6-year longitudinal study,GT levels within the normalrange were associated with incident hypertension in bothdrinkers and nondrinkers, but only in participants who wereabovemedianmeasuresoffatness. Theauthorsinterpretedthese findings to indicate that serumGT may predicthypertensionamongindividuals withincreasedcentral fatdistribution, with fatty liver representing an important under-lying mechanism for the association. A closely related area ofinterest concerns the possible link between alcohol consump-tion, diabetes, and the metabolic syndrome. Alcohol has beenlinked not only to an increase in blood pressure but to severalother elements of this syndrome, in particular the increase intriglyceridelevels,centraladiposity,andelevateduricacid.However, alcohol simultaneously acts to increase HDL-cho-lesterol levels, so whether it makes any significant contribu-tion to the metabolic syndrome has remained controversial. Inthe 1998 Korean National Health and Nutrition ExaminationSurvey19the consumption of 30 g alcohol/d was associatedwithanincrease inbloodpressure inmen, a highbloodglucose in women, and higher triglycerides in both men andwomen, whereas for both sexes and across all alcoholconsumptioncategoriestherewasasignificant increaseinHDL-cholesterol. Despite these contrasting effects on differ-ent components, overall there was a doseresponse relation-ship between increasing alcohol intake and the odds of havingthe metabolic syndrome. In contrast, a report on 4510 whiteparticipants fromthe National Heart and Blood InstituteFamily Heart Study in the United States, after careful adjust-ment for confounders including education, diet, and physicalactivity,20actuallyfoundasubstantiallyreducedprevalenceof thesyndromeacross all beveragetypes comparedwithnever drinkers (odds ratio down to 0.32 for wine drinkersonly). DatafromtheThirdNational HealthandNutritionExaminationSurvey21alsosuggestedalcohol consumptionwas inversely associated with the prevalence of severalcomponentsof thesyndrome, low-serumHDLcholesterol,elevatedserumtriglycerides, highwaistcircumference, andhyperinsulinemia, a finding that was strongest among whitesand among beer and wine drinkers. These contrasting resultsfromseveral different populationstudies suggest that anyoverall effects of alcohol onthe metabolic syndrome areprobably dictated by a number of competing and confoundinginfluences, suchasvolumeandtypeof alcohol consumed,gender, race, and ethnicity. Therefore, an effect of alcohol toinduce hypertension and the metabolic syndrome by impair-inginsulinresistance is doubtful. This is supportedbyarandomizedcontrolledalcoholinterventiontrialinvolvinga2 Hypertension June 2006reductionof alcohol consumptionby80%for 4weeks inregular moderate drinkers,22which was unable to detect anyeffect of changing alcohol on glucose or insulin homeostasis.Safe Levels of Drinking for HypertensivePeople in the Context of Alcohol and theGlobal Burden of Hypertensive DiseaseOn the basis of coronary protective effects of red wine,Bulpitt posed the question How many alcoholic drinks mightbenefit an older person with hypertension?23Largely usingmeta-analyses from general population studies, he concludedthat hypertensivepatientsaged60yearswhodrank16drinks a week should be advised to reduce their consumption.This is broadly in accord with international guidelines on themanagement of hypertension. Some of the pitfalls of limitingthefocus tocoronaryarterydiseasewerediscussedinanaccompanyingeditorial.24Care alsoneeds tobe takeninextrapolating data quantitatively across cultures with widelydiffering coronary rates and social circumstances. Connor etal25showed howsome ethnic communities living withinWesternizedpopulations, suchasMaorisinNewZealand,show a far higher burden from death, disease, and disabilityfromalcohol. InJapanstroke remains the most commoncause of cardiovascular death, coronary deaths are relativelylow, and there is a linear relation between alcohol consump-tion and hemorrhagic stroke. Youthfulness carries an in-creasedshareofhazardsassociatedwithalcohol causedbydrinking patterns, risky behaviors, and greater periods ofexposure to alcohol-related liver disease and cancers, and theagedependenceofalcoholrisksandbenefitshasbeenwellillustrated by Jackson and Beaglehole.26Furtherdataontherelativebenefitsandrisksoflight tomoderatealcoholconsumptioninhypertensivepatientswasprovided from a study of total and cardiovascular mortality ina population of 14 125 men derived fromthe PhysiciansHealthStudycohortandidentifiedwithahistoryofpastorcurrent treatment for hypertension.27There were 1018 deathsduring the study period, and about half of these werecardiovascular. Compared with individuals who reported thatthey rarely or never drank alcohol they found that those whoreported monthly, weekly, or daily consumption had increas-inglysignificanttrendsforreducedtotalandcardiovascularmortality. The beneficial effects of light to moderate drinkingwereseenregardlessofwhetherbloodpressurelevelswereaboveor below140/90mm Hg. However, thefindingthatconsumptionofaslittleasasinglealcoholicdrinkmonthlycouldreduce overall cardiovascular riskby18%stronglysuggests confounding from an unmeasured effect modifier. Inthisregard,nondrinkershaverecentlybeencharacterizedasan inappropriate comparison group for such studies thatoverestimateanybeneficial effect of alcohol as aresult.28Similarconfoundingwasprobablyalsooperativeinanewstudy from France29that involved a 13- to 21-year follow-upof 36 583 initially healthy middle-aged men. Moderate winedrinkers (their definition 60 g alcohol/d) had lower risks ofdeaths from all causes at all levels of systolic blood pressure.Nosignificant reductioninall-causemortalitywasseeninheavier drinkersor inthosewhoconsumedpredominantlybeer or spirits. Unmeasured confounding may have includeddietarydifferencesthat differentiatebeer, wine, andspiritsdrinkers as well as markedlifestyledifferences, includingpatterns of alcohol consumption, that are likely to haveimpactedonmanydisease processes contributingtototalmortality.30The most recent study measuring cardiovascularoutcomes in drinking hypertensive people was in the LosartanInterventionForEndpointreductioninhypertension(LIFE)studycohort,31whichfoundthat indrinkers therewas nodecrease in composite cardiovascular risk when being treatedwithlosartancomparedwithatenololbecauseadecreaseintheincidenceof myocardial infarctioninthedrinkerswasoffset by an increase in the risk of stroke.In terms of the overall significance of the effects of alcoholtoelevate bloodpressure, ananalysis fromthe landmarkWorldHealthOrganizationGlobalBurdenofDisease2000ComparativeRiskAnalysis study32assessedtherisks andbenefits of alcohol by region and then globally and attributed16%of all hypertensivediseasetoalcohol. Toaddtothecontroversy concerning safe levels of drinking, Jackson etal33concludedthat issues of bidirectional confoundinginpopulationstudies hadbeenunderestimatedinrelationtoalcoholandcoronaryheartdiseaseandthatanybenefitsofmoderate alcohol consumption on coronary disease werelikelytobeoutweighedbyharmful effects (ie, therewasprobably no free lunch).References1. Ruidavets JB, Bataille V, Dallongeville J, Simon C, BinghamA,AmouyelP,ArveilerD,DucimetiereP,FerrieresJ.Alcoholintakeanddiet inFrance, theprominent roleof lifestyle. EurHeart J. 2004;25:11531162.2. Stranges S, Wu T, Dorn JM, Freudenheim JL, Muti P, Farinaro E, RussellM, Nochajski TH, Trevisan M. Relationship of alcohol drinking pattern toriskofhypertension: apopulation-basedstudy. Hypertension. 2004;44:813819.3. Klatsky AL. Alcohol-associated hypertension: when one drinks makes adifference. Hypertension. 2004;44:805806.4. TrevisanM, SchistermanE, Mennotti A, Farchi G, Conti S. Drinkingpattern and mortality: the Italian Risk Factor and Life Expectancy Poolingproject. Ann Epidemiol. 2001;11:312319.5. Trevisan M, Dorn J, Falkner K, Russell M, Ram M, Muti P, FreudenheimJL, Nochajaski T, Hovey K. Drinking pattern and risk of non-fatalmyocardial infarction: a population-based case-control study. Addiction.2004;99:313322.6. WildmanRP, GuDF, Muntner P, HuangG, ChenJ, DuanX, HeJ.Alcohol intake and hypertension subtypes in Chinese men. J Hypertens.2005;23:737743.7. Okamura T, Tanaka T, Yoshita K, Chiba N, Takebayashi T, Kikuchi Y,Tamaki J, TamuraU, Minai J, Kadowaki T, MiuraK, NakagawaH,TaniharaS, OkayamaA, UeshimaH. Specificalcoholicbeverageandblood pressure in a middle-aged Japanese population: the High-risk andPopulationStrategyfor Occupational HealthPromotion(HIPOP-OHP)Study. J Hum Hypertens. 2004;18:916.8. Fuchs FD. Vascular effects of alcoholic beveragesis it only alcohol thatmatters? Hypertension. 2005;45:851852.9. Hashimoto M, Kim S, Eto M, Iijima K, Ako J, Yoshizumi M, Akishita M,Kondo K, Itakura H, Hosoda K, Toba K, Ouchi Y. Effect of acute intakeof red wine on flow-mediated vasodilatation of the brachial artery. Am JCardiol. 2001;88:14571460.10. Zilkens RR, Burke V, Hodgson JM, Barden A, Beilin LJ, Puddey IB. Redwine and beer elevate blood pressure in normotensive men. Hypertension.2005;45:874879.11. McFaddenCB, Brensinger CM, BerlinJA, TownsendRR. Systematicreviewof theeffect of dailyalcohol intakeonbloodpressure. AmJHypertens. 2005;18:276286.12. Xin X, He J, Frontini MG, Ogden LG, Motsamai OI, Whelton PK. Effectsofalcohol reductiononbloodpressure: ameta-analysisofrandomizedcontrolled trials. Hypertension. 2001;38:11121117.Beilin and Puddey Alcohol and Hypertension 313. van den Elzen AP, Sierksma A, Oren A, Vos LE, Witteman JG, GrobbeeDE, HendriksHF, Uiterwaal CS, BotsML. Alcohol intakeandaorticstiffness in young men and women. J Hypertens. 2005;23:731735.14. SierksmaA, LebrunCE, vanderSchouwYT, GrobbeeDE, LambertsSW,HendriksHF,BotsML.Alcoholconsumptioninrelationtoaorticstiffnessandaorticwavereflections: across-sectional studyinhealthypostmenopausal women. Arterioscler Thromb Vasc Biol. 2004;24:342348.15. Sierksma A, Muller M, van der Schouw YT, Grobbee DE, Hendriks HF,Bots ML. Alcohol consumption and arterial stiffness in men. J Hypertens.2004;22:357362.16. BeilinL. Alcohol andcardiovascular disease: possible protectionviaeffects on aortic stiffness. J Hypertens. 2005;23:703705.17. Nakanishi N, Kawashimo H, Nakamura K, Suzuki K, Yoshida H, UzuraS, Tatara K. Association of alcohol consumption with increase in aorticstiffness:a9-yearlongitudinalstudyinmiddle-agedJapanesemen.IndHealth. 2001;39:2428.18. Stranges S, Trevisan M, Dorn JM, Dmochowski J, Donahue RP. Body fatdistribution, liver enzymes, and risk of hypertension: evidence from theWestern New York Study. Hypertension. 2005;46:11861193.19. Yoon YS, Oh SW, Baik HW, Park HS, Kim WY. Alcohol consumption andthe metabolic syndrome in Korean adults: the 1998 Korean National Healthand Nutrition Examination Survey. Am J Clin Nutr. 2004;80:217224.20. DjousseL, ArnettDK, EckfeldtJH, ProvinceMA, SingerMR, EllisonRC. Alcohol consumptionandmetabolicsyndrome: does thetypeofbeverage matter? Obesity Research. 2004;12:13751385.21. FreibergMS, Cabral HJ, HeerenTC, VasanRS, EllisonRC. Alcoholconsumption and the prevalence of the metabolic syndrome in the USacross-sectional analysis of data fromthe Third National Health andNutrition Examination Survey. Diab Care. 2004;27:29542959.22. Zilkens RR, BurkeV, Watts G, BeilinLJ, PuddeyIB. Theeffect ofalcohol intake on insulin sensitivity in men: a randomized controlled trial.Diab Care. 2003;26:608612.23. Bulpitt CJ. How many alcoholic drinks might benefit an older person withhypertension? J Hypertens. 2005;23:19471951.24. BeilinL. Alcohol andhypertension: balancingtherisks andbenefits.J Hypertens. 2005;23:19531955.25. Connor J, Broad J, Rehm J, Vander HS, Jackson R. The burden of death,disease, anddisabilitydue toalcohol inNewZealand. NZMedJ.2005;118:U1412.26. Jackson R, Beaglehole R. Alcohol consumption guidelines: relative safetyvs absolute risks and benefits. Lancet. 1995;346:716.27. Malinski MK, SessoHD, Lopez-JimenezF, BuringJE, GazianoJM.Alcohol consumption and cardiovascular disease mortality in hyper-tensive men. Arch Int Med. 2004;164:623628.28. Naimi TS, Brown DW, Brewer RD, Giles WH, Mensah G, Serdula MK,MokdadAH, HungerfordDW, LandoJ, Naimi S, StroupDF. Cardio-vascular risk factors and confounders among nondrinking and moderate-drinking U.S. adults. Am J Prev Med. 2005;28:369373.29. RenaudSC, GueguenR, ConardP, Lanzmann-PetithoryD, OrgogozoJM, HenryO. Moderatewinedrinkershavelowerhypertension-relatedmortality: aprospectivecohort studyinFrenchmen. AmJClinNutr.2004;80:621625.30. Tjonneland A, Gronbaek M, Stripp C, Overvad K. Wine intake and dietin a random sample of 48763 Danish men and women. Am J Clin Nutr.1999;69:4954.31. Reims HM, Kjeldsen SE, Brady WE, Dahlof B, Devereux RB, Julius S,Beevers G, De Faire U, Fyhrquist F, Ibsen H, Kristianson K, Lederballe-Pedersen O, Lindholm LH, Nieminen MS, Omvik P, Oparil S, Wedel H.Alcoholconsumptionandcardiovascularriskinhypertensiveswithleftventricularhypertrophy:theLIFEstudy.JHumanHypertens.2004;18:381389.32. Rehm J, Room R, Monteiro M, Gmel G, Graham K, Rehn N, Sempos CT,JerniganD. Alcohol asariskfactorforglobal burdenofdisease. EurAddict Res. 2003;9:157164.33. JacksonR, BroadJ, Connor J, WellsS. Alcohol andischaemicheartdisease: probably no free lunch. Lancet. 2005;366:19111912.4 Hypertension June 2006