Final Paper

8 pages

Please download to get full document.

View again

of 8
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
British Journal of Clinical Pharmacology DOI:10.1111/j.1365-2125.2005.02358.x Khat chewing is a risk factor for acute myocardial infarction: a case-control study A. Al-Motarreb,1 S. Briancon,2 N. Al-Jaber,1 B. Al-Adhi,1 F. Al-Jailani,1 M S. Salek3 & K J. Broadley3 1 Al Thawrah Teaching Hospital, Sana’a, Yemen, 2School of Public Heath, Clinical Epidemiology and Evaluation CHU, Nancy and School of Public Health, Nancy I Henri Poincaré University, France, 3Welsh School of Pharmacy, Cardiff Univer
   Br J Clin Pharmacol   59   :5574–581574© 2005 Blackwell Publishing Ltd  British Journal of Clinical Pharmacology  DOI:10.1111/j.1365-2125.2005.02358.x    Correspondence  Dr Sam Salek,  Centre forSocioeconomic Research, WelshSchool of Pharmacy, Cardiff University,Cardiff CF10 3XF, UK. E-mail:   Keywords  khat, myocardial infarction,case-control study, smoking Received 2 April 2004 Accepted 29 November 2004  Khat chewing is a risk factor for acute myocardialinfarction: a case-control study   A. Al-Motarreb,   1  S. Briancon,   2  N. Al-Jaber,   1  B. Al-Adhi,   1  F. Al-Jailani,   1  M S. Salek    3  & K J. Broadley    3   1   Al Thawrah Teaching Hospital, Sana’a, Yemen, 2  School of Public Heath, Clinical Epidemiology and Evaluation CHU, Nancy and School of Public Health, Nancy I Henri Poincaré University, France, 3  Welsh School of Pharmacy, Cardiff University, UK    Aim   Khat chewing is a common habit in Yemen and east African countries. Millions of people chew khat leaves daily for its euphoric and energetic effects and to increasealertness. Cathinone, the main active substance in fresh khat leaves, has sympatho-mimetic effects which increase heart rate and blood pressure. The aim was to examinethe hypothesis that khat chewing is a risk factor for acute myocardial infarction (AMI)using a hospital-based matched case-control study.   Method   Between 1997 and 1999, we selected 100 patients admitted to the Al-Thawrateaching hospital Sana’a ICU, Yemen with acute myocardial infarction. 100 controlsubjects, matched to cases for sex and age, were recruited from the outpatients clinicsof the same hospital. A questionnaire was completed for case and control groupscovering personal history of khat chewing, smoking, hypertension, diabetes and anyfamily history of myocardial infarction. A blood sample was collected for performinglipid profiles. Cases and controls were compared by analysis conducted using condi-tional logistic regression which corrected for baseline imbalances leading to lessbiased estimations of odds ratio (OR). The risk associated with each classical factorand khat chewing habits was then investigated. OR values greater than 2.5 indicateda significant risk factor.   Results   Khat chewing was significantly higher among the AMI case group than control group(OR =   5.0, 95% CI 1.9–13.1). A dose–response relationship was observed, the heavykhat chewers having a 39-fold increased risk of AMI.   Conclusion   This study indicates that khat chewing is associated with AMI and is an independentdose-related risk factor for the development of myocardial infarction.  Introduction  Fresh leaves of the khat (or qat) tree (  Catha edulis  Forsk.) are widely chewed in Yemen and east Africancountries. People chew khat for its central nervous sys-tem stimulant effects such as euphoria, energy, alertnessand for social purposes [1,2]. Yemenis chew khatheavily on a daily basis. The habit of khat chewing isdeeply rooted in Yemeni society [3] and has sharplyincreased in recent years. Air transport has removed themain obstacle to its distribution outside Yemen, namelyits perishability, so that khat is now available to Yemeni,Somali and east African communities living in the USAand UK. Seven tonnes of khat passes through Heathrowairport each week; smaller amounts are also imported   Khat chewing and acute myocardial infarction    Br J Clin Pharmacol   59   :5575  through other airports. Some of this amount is re-exported to other European countries [4].The pharmacological effect of khat chewing is mainlydue to the cathinone present in the fresh leaves [1,2,5,6].Cathinone is regarded as an indirectly acting sympath-omimetic alkaloid having catecholamine-releasing prop-erties at both central dopaminergic [7] and serotonergic[8] synapses as well as at peripheral noradrenaline stor-age sites [9]. It operates through the same mechanismas amphetamine which explains the observed centralnervous system stimulant effects in the khat chewer [1].Cardiovascular effects of khat chewing in humansinclude elevated blood pressure and increases in heartrate [3,6,10]. Increases in blood pressure, heart rate andcardiac contractile force in anaesthetized dogs [11] andpositive inotropic and chronotropic actions in isolatedatria [12] have been reported after the administration of the active ingredient, cathinone.In a controlled study, khat chewing has been shownto increase blood pressure and heart rate which arerelated to the presence of cathinone in the plasma [13].Recently, we have reported an increased incidence of acute myocardial infarction (AMI) in Yemen, whichwas associated with khat chewing. There was also adifference in the diurnal pattern of AMI presentationbetween khat users and non-khat users. In non-khatusers the peak presentation of AMI is in the early hours,whereas in khat users it is shifted to the late afternoonand evening, which coincided with the khat chewingsession [14]. Thus, we undertook a hospital-basedmatched case-control study in order to address the issueof a possible increased risk of AMI among khat chew-ers, controlling for other risk factors and exploringchewing habits.  Methods   Subjects  100 male or female patients admitted with acute myo-cardial infarction to the intensive care unit (ICU) of Al-Thawra Teaching Hospital, Sana’a, Yemen wereenrolled between 1997 and 1999. Recruitment wassequential and followed the myocardial infarction. Allcases recruited took part in the study. The diagnosis of the AMI was based on clinical symptoms, recent ECGchanges and elevated cardiac enzymes; creatine phos-phokinase (CPK) and lactate dehydrogenase (LDH).Normal values of CPK and LDH were 24–190 IuL   -   1  and230–460 IuL   -   1  , respectively, and a doubling of the upperlimits was associated with AMI. The criteria for positivediagnosis according the ECG changes were; pathologi-cal Q waves or 1 mm ST-segment elevation in two ormore contiguous leads or a new left bundle branch block or new persisting ST-T changes.The control group was selected from the dermatology,neurology or internal medicine outpatient clinics of Al-Thawra Hospital, where it was considered the incidenceof cardiovascular disease would be low. All controlpatients were new referrals and had to be free of anyischaemic heart disease, that is, no history of typicalangina pain or history of previous attack of unstableangina or AMI and their 12-leads ECG had to be normal.In all case subjects, there was also no previous historyof an AMI. Each subject of the case group was individ-ually matched to a subject of a control group for sex andage within 5 years’ difference. The decision was madeat the outset not to match control and case subjects forsmoking. Confounding factors, such as cigarette smok-ing are neutralized by application of tests that makeadjustment for any effects as a result of nonmatching.Since matching was for age and sex, which are alsorelated to smoking and khat chewing, matching on cig-arette smoking was not essential. To have selected equalnumbers of nonsmokers in case and control groupswould have resulted in a substantial loss of subjectsfrom the study.  Data collection  A doctor interviewed each patient and completed aquestionnaire in English for each case and controlsubject. The questionnaire was completed at the firstvisit of control subjects and during days 2–4 postinf-arct in the case group. The questionnaire wasdesigned to cover the personal history of khat chew-ing (amount of khat per day, number of hours perday, number of days per week and number of years),smoking (cigarette and water pipe), hypertension, dia-betes and any family history of myocardial infarction(MI). Hypertension was not assessed since in the casegroup, post-MI blood pressure would not have indi-cated pre-existing hypertension. The criterion forbeing overweight was according to ideal weight esti-mation (Broca’s Index; ideal weight (kg) = height(cm) - 100).Blood samples were collected in fasted subjects in themorning before breakfast for performing lipid profiles.In cases, this was in the ICU period between 2 and4 days post-MI and in controls at entry into the study.Triglyceride and total cholesterol were determined in5 m l aliquotes of serum by standard coupled enzymaticprocedures After 10 min incubation of serum with theassay reagents, measurements were made at 700 nm ina Beckman Synchro CXS Decision Calibrant (BeckmanInstruments, High Wycombe, Buckinghamshire, UK)   A. Al-Motarreb et al. 576   59   :5    Br J Clin Pharmacol  and linearly related to a standard triglyceride orcholesterol curves. The laboratory reference ranges fortriglyceride and total cholesterol were up to 2.3 and6.5 mmol l   -   1  , respectively.   Statistical analysis  Data were recorded on EPI Info® files (WHO, Geneva,Switzerland; from the Centres for Disease Control andAevention (CDC), USA). Statistical analysis was per-formed using SAS® Statistical Software version 8.1(SAS system, SAS Institute Inc., Cary, NC, USA).Cases and controls were compared through conditionallogistic regression which allowed nonbiased estimationsof odds ratio (OR). After a comparison of the two groupsfor general characteristics, the risk associated with eachclassical factor and khat chewing habits was investi-gated. Final models were built on with and withoutexploring dose–response relationship. The significancelevel for variables in the multivariate models was set at0.05. The models provided the regression coefficients of independent variables, antilogs of which gave oddsratios expressing the effect on the probability of beinga case when changing one unit of the independent vari-able, holding other variables in the equation model con-stant. Odds ratio values are given with their 95%confidence limits as an additional indicator of clinicalsignificance.The inclusion of 100 cases and controls had beenplanned to permit a detection of OR > 2.5, giving prob-ability of committing type 1 error of 5% ( P   = 0.05) andpower of study of 80% and an approximated 65% khatchewing frequency among control (N-Query advisorsoftware®, Statcon, Witzenhausen, Germany). Thus,OR values > 2.5 indicate a significant risk factor.  Results  The mean age of the acute myocardial infarction groupwas 48.6 ± 9.0 years, 67% of patients were less than50 years of age and 92% were men. Three cases out of four came from Sana’a province, because Al-Thawrahospital is the referral and main hospital in Yemen. 53%of the patients had acute extensive myocardial infarction,while 14% had anterior and 32% had inferior MI. Asexpected from the matching of patients, cases and controlgroups were completely compared for age and sex. Res-idence and occupation were also comparable. A signifi-cant difference occurred in the occupational status wherecivil servants were less represented in the case group(Table 1). Data collected for this study were consideredby the hospital as part of the routine clinical managementof the patient. All patients gave verbal informed consentprior to taking part in the study. No formal hospital orLocal Ethics Committee operates at Al-Thawrah Teach-ing Hospital, Sana’a. However, approval for the studywas granted by the Clinical Directorate on review of thestudy protocol.Comparisons for classical risk factors are shown inTable 2. As expected, we found that cigarette smokingwas an important risk factor for AMI with a linearincrease of risk with doses (number of cigarettes perday). Water-pipe smoking, traditionally linked to khatchewing session, was nonsignificantly related to AMIoccurrence. High blood pressure was less frequentamong cases than controls, although nonsignificantly,while there was no difference in the incidence of diabetes.Overweight patients were more frequent among controlsthan in the case group (OR = 0.35 P   < 0.03), however,these 18% of subjects were reported as being overweightand not classical obese. Since both cases and controlswere matched for age and that 50% were under 50 yearsof age, these were a relatively young population whichwould make them less prone to classical risk factors suchas being overweight.When treating the total cholesterol as a quantitativevariable, we did not find any relationship with AMI. Butwhen creating three categories, low, normal and high,we found a significant effect suggesting a nonlinear rela-tionship: higher risk among normal level of cholesterolsubjects as compared with low-level subjects (OR = 4.7 P   < 0.02), but non significant increased risk among highlevel subjects (OR = 2.7).Results for khat chewing habits are shown in Table 3.Khat chewing was significantly more frequent amongcases than control groups (89% vs. 69%) leading to arisk estimation of 5.0. A significant dose–response rela-tionship was evident with the four indicators of quantityused, each showing a threshold effect. Considering themean estimated quantity of khat chewed per session,mild chewers were not shown to be at risk, while mod-erate khat chewers were shown to be at high risk (OR = 7.62) and heavy khat chewers at even higher risk (OR = 22.28). Another indicator of the quantity of khatis the number of hours per session; the risk of occur-rence of AMI increased when the duration of sessionwas more than 4 h per session. Investigating the regu-larity of the consumption, only daily khat chewers weresignificantly more at risk (OR = 6.4). Finally, consider-ing the number of years duration of the khat chewinghabit, we have shown that only recent and chronic khatchewers had higher risk (OR = 22 and 5, respectively).Table 4 shows the results of the two-multivariatemodels. The first model did not include khat and tobaccodoses and only three variables remained significant:occupational status, khat chewing and cigarette smok-   Khat chewing and acute myocardial infarction    Br J Clin Pharmacol   59   :5577  CaseControlProportionMeanSDProportionMeanSD–291130–39161640–49353550–59333360–69121270–7933SexMale9292   NS  Female88Area of residenceSANAA7479   NS  IBB68DHAMAR54ALBAIDA32Daily44HAJJA10ALMAHWEET20MAREB21HADRAMOUTH21ALJOUF10Occupational statusSenior manager2114Manual worker3624Civil servant1650*Small trader189Unemployed/Retired93  NS Non significant   *  P  <   0.001.Numbers of subjects are percentages since there are 100 subjects per group.  ing. The level of risk for the last two (i.e. khat chewingand cigarette smoking) was close to each other(OR = 5.8 and 7.02, respectively), the risk being slightlyhigher for smoking. The second model included doses of khat and tobacco and confirmed the increased risk withincreasing doses of both smoking and khat chewing. Adirect relationship appeared to exist between the dura-tion of khat chewing sessions and its risk. Subjects chew-ing khat for more than 6 h were at high risk (OR = 39).The other indicators of khat quantity were nonsignificantpredictors of AMI occurrence.Restricting the analysis to AMI occurring during thekhat effective period, i.e. the usual time of the khat chew-ing session and for several hours afterwards (14.00–24.00 h) and in spite of the loss of power (only 66 pairswere retained), the risk from khat chewing was increasedto 9, which was almost double that of smoking(OR = 4.84).Finally, we did not find evidence of any interactionbetween smoking and khat chewing, indicating that eachwas an independent risk factor for AMI.  Discussion  In this case control study, khat chewing was significantlyand independently associated with AMI in a dose–response relationship. As far as we know, this is the firstepidemiological study on this subject. Classically, case-control studies especially hospital-based ones are subjectto bias, particularly in patient selection and measure-ments. A selection bias is always possible and could beevoked according to the difference in occupational statusin the two groups. Nevertheless, the risk of AMI is not  Table 1  Socio-demographic characteristics of caseand control groups
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks