Journal of Epidemiology and Global Health (2015) xxx xxxndashxxx

http wwwelsev ier com locate

Hypertension in the Lebanese adultsImpact on health related quality of life

httpdxdoiorg101016jjegh2015020032210-6006ordf 2015 Published by Elsevier Ltd on behalf of Ministry of Health Saudi Arabia

Corresponding author at Clinical and Epidemiological Re-search Laboratory Faculty of Pharmacy Lebanese UniversityHadath Campus Lebanon Tel +961 5 463365

E-mail addresses pascalesalameh1hotmailcom psalamehuledulb (P Salameh) amalkehhotmailcom (AA Hajje)sanaa_awadayahoocom (S Awada) samarrachidioutlookcom (S Rachidi) wbawabhotmailcom wbawabuledulb(W Bawab)

Please cite this article in press as Khalifeh M et al Hypertension in the Lebanese adults Impact on health related quality of life J EpidemiolHealth (2015) httpdxdoiorg101016jjegh201502003

jegh

Malak Khalifeh Pascale Salameh Amal Al Hajje Sanaa AwadaSamar Rachidi Wafa Bawab

Laboratory of Epidemiological and Clinical Research Lebanese University Beirut Lebanon

Received 29 December 2014 received in revised form 4 February 2015 accepted 11 February 2015

KEYWORDSHypertensionQuality of lifeSF-8Risk factorsLebanese population

Abstract Cardiovascular disease is a major cause of morbidity and mortality world-wide hypertension being one of their most prevalent risk factors Information onhealth related quality of life (QOL) of hypertensive individuals in Lebanon is lackingOur objectives were to evaluate QOL of hypertensive patients compared with non-hypertensive subjects and to suggest possible predictors of QOL in Lebanon We con-ducted a case control study among individuals visiting outpatient clinics Quality oflife was assessed using the eight item (SF-8) questionnaire administered face to faceto the study population applied to hypertensive (N = 224) and non-hypertensivecontrol (N = 448) groups Hypertensive patients presented lower QOL scores in alldomains particularly in case of high administration frequency and occurrence ofdrug related side effects Among hypertensive patients QOL was significantlydecreased with the presence of comorbidities (b = 13865 p = 0054) daily fre-quency of antihypertensive medications (b = 8196 p lt 0001) presence of drugside-effects (b = 19262 p = 0031) older age (b = 0548 p lt 0001) female gen-der (b = 21363 p = 005) lower education (b = 22949 p = 0006) and cigar-ettes smoked daily (b = 0726 p lt 0001) regular sport activity (b = 2315p lt 0001) significantly increased quality of life These findings indicate the neces-sity for health professionals to take these factors into account when treating hyper-tensive patients and to tackle special subgroups with attention to theirdeteriorated QOLordf 2015 Published by Elsevier Ltd on behalf of Ministry of Health Saudi Arabia

1 Introduction

Cardiovascular disease (CVD) is a major cause ofmorbidity and mortality worldwide [12] Globallythey account for approximately 17 million deaths a

Global

2 M Khalifeh et al

year [3] Hypertension is considered a major cardio-vascular risk factor with high prevalence in almostall countries with a slow and silent progression[4] Complications of hypertension are highly preva-lent and account for 94 million deaths worldwideevery year [5] On another hand the dramaticincrease of longevity has brought attention that itshould be accompanied with improvements inhealth-related quality of life (QOL) The WorldHealth Organization (WHO) has summarized theseconcerns stating that lsquolsquoadding years to life is anempty victory without adding life to yearsrsquorsquo [6]Regardless of the condition type patients QOL isadversely affected by chronic illness [7] this is animportant consideration in the management ofasymptomatic conditions such as hypertension [8]

With this regard studies assessing the impact ofhypertension on QOL had been carried out in vari-ous countries For example Roca-Cusachs et alassessed the impact of QOL clinical variables andreported that hypertensive patients experienceda significant reduction in QOL when compared withnormotensive individuals [9] Li et al studied theChinese population and observed that QOL washigher in normotensive individuals [10] Bardageand Isacson found that QOL scores were poorer inhypertensive patients than in general population[11]

As far as Lebanon is concerned a populationbased study conducted in 2005 revealed that theoverall prevalence of declared hypertension was231 [13] Compared to other Mediterraneancountries Lebanon has a lower prevalence ofhypertension however compared to other coun-tries [1415] Lebanon still has a high prevalenceof hypertension Nevertheless no study has beenperformed regarding the health related quality oflife of hypertensive patients and its correlatesThe objective of this study was to evaluate QOLof hypertensive patients compared to non-hyper-tensive subjects in Lebanon and to suggest possiblepredictors of their QOL

2 Material and methods

21 Study design

This is an observational case control study carriedout in hypertensive patients undergoing treatmentand non-hypertensive individuals Patients wererandomly selected from those visiting outclinicsof two tertiary care hospitals (Hamoud and RaeeHospitals) The Institutional Review Board of bothinstitutions waived the need for an official approvalto perform this observational study provided it

Please cite this article in press as Khalifeh M et al Hypertension in the LebHealth (2015) httpdxdoiorg101016jjegh201502003

respected patients autonomy and confidentialityThe patients were informed about the objectivesof the study and asked to give an oral consentOnly those who gave their consent to participatein the study were enrolled Data were collectedfrom the 2nd of March till the 1st of June 2014

22 Study population

Patients enrolled were of both genders aged18 years and older and divided into two groupsHypertensive group consisting of patients diag-nosed with hypertension and taking antihyperten-sive medication for at least three months whilethe control group consisted of non-hypertensiveindividuals (blood pressure lt 12090 mmHg) thelatter should be healthy or consulting for acutehealth related problems such as urinary tractinfections or otitis Exclusion criteria were thepresence of chronic or disabling diseases includingdecompensated diabetes cancer pregnancysequelae of stroke decompensated heart failurechronic kidney disease liver failure acute myocar-dial infarction within the past six months and theuse of antipsychotic drugs

23 Data collection

Data were acquired through a structured face toface interview The questionnaire was administeredin Arabic and included different sections that eval-uated socio-demographic data including sex ageeducational level and marital status monthlyincome bodymass index and lifestyle data includingsmoking status involvement in sports activitiesalcohol caffeine consumption and salt intakeDetailed health history was also assessed in relationwith the onset of the hypertension the presence ofassociated complications and co-morbidity Clinicaldata related to hypertension classification (con-trolled or uncontrolled) pharmacological manage-ment (class and number of antihypertensive drugsused) as well as any drug side-effects experiencedwere obtained Blood pressure was recordedAccordingly controlled blood pressure was definedas systolic blood pressure 6 140 mmHg anddiastolic 6 90 mmHg [16] We also assessed thepatients attitudes and knowledge about the sever-ity of disease and the use of medication

24 Quality of life measurement

We assessed the QOL using the SF-8 questionnaireWhile one of the most widely used tools is the SF-36 generic QOL questionnaire its derivative (SF-8)is an additional easier instrument used in

anese adults Impact on health related quality of life J Epidemiol Global

Hypertension in the Lebanese adultsImpact on health related quality of life 3

population-based studies andwould provide a candi-date instrument for measuring QOL in hypertensivepatients [12] This questionnaire offers the advan-tage that scores can be readily compared withresults from the original SF-36 version Moreoverit represents a major advance in measuring QOLfor purposes of achieving both brevity and compre-hensiveness in population health surveys [12]

SF-8 has eight items relying on a single item tomeasure each of the eight domains of health in theSF-36 survey physical functioning physical rolebodily pain general health vitality social func-tioning emotional role and mental healthPhysical functioning covers limitations in daily lifedue to health problems The role physical scalemeasures role limitations due to physical healthproblems The bodily pain scale assesses pain fre-quency and pain interference with usual rolesThe general health scale measures individual per-ceptions of general health The vitality scaleassesses energy levels and fatigue The social func-tioning scale measures the extent to which illhealth interferes with social activities The roleemotional scale assesses role limitations due toemotional problems and the mental health scalemeasures psychological distress [12]

The SF-8 was thus translated into Arabic by abilingual Lebanese researcher the main languageof Lebanon It is then back translated into Englishby another translator who has no knowledge ofthe English version Forward translation intoArabic was conducted The Arabic version was thenpilot-tested for understanding in a small patientgroup before it was used Its psychometric proper-ties were also assessed in the Lebanese population

25 Sample size

Sample size was calculated using Epi info7 assum-ing a type I error of 5 and a study power of 80and a confidence interval of 95 In the absenceof baseline data the exposure probability to hyper-tension was considered to be equal to 50 inhealthy population The minimal sample sizenecessary to show a twofold increase in the riskof hypertension consisted of 336 subjects Wedecided to study about twice this number in a ratiocasecontrol = 12 to have an adequate power forstudying subgroups

26 Statistical analysis

Data were entered and analyzed using StatisticalPackage for Social Sciences version 19 No missingdata were obtained A p-value 6 005 was consid-ered significant in all tests Quality control of data

Please cite this article in press as Khalifeh M et al Hypertension in the LebHealth (2015) httpdxdoiorg101016jjegh201502003

entry and cleaning were done Descriptive statisticsincluding frequency mean and standard deviationwere used to summarize patients characteristics

After data cleaning items quality was assessedby analyzing the distribution of item responses ofthe SF-8 using aggregate endorsement frequenciesFor instruments with around a 5 point responserange such as the SF-8 any item with 2 or moreadjacent response points showing less than 10 ofthe response on aggregate are problematic [17]

The QOL score was also validated by testing forinter-item correlation The internal consistency ofthe SF-8 was assessed using Cronbachs reliabilitycoefficient alpha Internal consistency of SF-8items with the underlying constructs of physicalcomponent summary (PCS) and mental componentsummary (MCS) was assessed using Pearson cor-relation test [18] It was hypothesized that therewould exist strong correlations between the PCSand items 1ndash5 and strong correlations betweenMCS and items 6ndash8 assuming that items more clo-sely related to a common dimension would showa correlation of P 050

The major dependent variable was PCS MCSand overall QOL which is the sum of all individualdomains of SF-8 items Each domain was scoredover 100 using a norm-based scoring method [19]When PCS and MCS measures were calculatedhigher summary scores indicated better healthScores above and below 50 were considered aboveand below the average based upon results recordedby the developers of the SF-8 [12]

An appropriate bivariate analysis was done forevery explanatory variable For continuous vari-ables used in comparison we used Student testor ANOVA for variables with adequate normal dis-tribution to compare between means For non-nor-mally distributed continuous variables theKruskalndashWallis test was used For categorical vari-ables Chi2 and Fisher exact tests were used tocompare percentages

Multivariable analysis using linear regression wascarried out to compare measures between groupsof comparison taking into account potential con-founding variables that had a p lt 020 in bivariateanalysis The analysis sought to explore factorsassociated with QOL using regression analysisThe final model was accepted after ensuring ade-quacy of data

3 Results

During the study period the number of subjectsthat agreed to participate in this study was 672(out of 785 856) among them 224 patients

anese adults Impact on health related quality of life J Epidemiol Global

4 M Khalifeh et al

(333) were included in the hypertension groupand 448 (667) in the control group Patientswho refused to participate did not deliver anyinformation to the researcher

31 SF-8 data quality internal consistencyand reliability

The response distributions of each item for thesensitivity aggregate endorsement frequencyshowed all 8 items performing well Table 1 showsthat no item of the SF-8 questionnaire has 2 ormore adjacent responses showing lt 10 ofresponses Internal consistency measured byPearson correlation is also presented in Table 1These show a generally strong level of consistency(P060) of items 1ndash5 with PCS including generalhealth (r = 0744) physical functioning(r = 0858) role physical limitation (r = 0884)bodily pain (r = 0765) and vitality (r = 0765)Also the results show strong correlation items 6ndash8 with MCS including social functioning (r = 05)mental health (r = 0914) emotional role(r = 0852)

Reliability Cronbachs alpha of the model (SF-8)was 0868 with average intraclass correlation coef-ficient (ICC) = 0868 (95CI = 0853ndash0883p lt 0001) Cronbachs alpha of each of the twosummaries PCS was 0882 and that of MCS was0763 which is very high (gt07) Moreover theremoval of any item of the score was accompaniedby a decrease in the models Cronbachs alpha

32 Socio-demographic characteristics ofthe study population

The comparison of characteristic differencesbetween hypertension and control groups is shownin Table 2 Both groups were homogenous regard-ing gender (p = 0557) Participants were mostly

Table 1 Distribution of frequencies amp internal consistency of

QOL items Response option frequencies in N ()

1 2 3

General health 41 (61) 167 (249) 233 (347Physical functioning 205 (305) 192 (286) 148 (22)Role physicallimitation

199 (296) 184 (274) 165 (246

Bodily pain 236 (351) 88 (131) 103 (153Vitality 69 (103) 227 (338) 273 (406Social functioning 296 (44) 153 (228) 127 (189Mental health 206 (307) 107 (159) 172 (256Role emotionallimitation

316 (47) 142 (211) 124 (185

Please cite this article in press as Khalifeh M et al Hypertension in the LebHealth (2015) httpdxdoiorg101016jjegh201502003

females in hypertension and control groups(5900 versus 6227) Hypertension individualswere older less educated widowed or divorcedretired or never working have less monthly incomeand more financial support and were more over-weight and obese (p lt 0001)

33 Health related quality of life

Controls had higher significant scores (p lt 005)than hypertensive individuals in all domains ofthe QOL score Controls showed better QOL in gen-eral health physical functioning role physical lim-itation bodily pain vitality and PCS with(p lt 0001) They also showed better health perfor-mance in mental health related items includingsocial functioning (p = 0033) mental health(p = 0011) emotional limitation role (p lt 0001)and MCS (p = 0026) (Fig 1)

34 Multivariable analyses

In the multivariable linear regression model(Table 3) we found that the presence of hyperten-sion (b = 17241 p = 0009 95CI [301664317]) and mood disorder (b = 17241p lt 0001 95CI [58011 16661]) were nega-tively associated to QOL Increasing age(b = 0388 p lt 0001 95CI [752 024])and female gender (b = 14485 p = 003795CI [23323 5648]) also significantlyaffected QOL where females had lower QOL thanmales Practicing regular sport (b = 15937 95CI[6362 25513]) and higher education level(b = 15937 95CI [8451 2768]) were positivelyassociated to QOL (p lt 0001) Moreover the pres-ence of more comorbidities was adversely corre-lated the overall QOL (b = 8083 p = 000995CI [14184 1982]) The frequency of med-ication intakeday was the most significant risk

SF-8 (N = 672)

Internalconsistency

4 5 6 r PCS r MCS

) 150 (223) 69 (103) 12 (18) 0744 0315113 (168) 14 (21) 0858 0194

) 102 (152) 22 (33) 0884 0226

) 156 (232) 81 (121) 8 (12) 0765 0286) 95 (141) 8 (12) 0636 0407) 93 (138) 3 (04) 0409 0500) 164 (244) 23 (34) 0153 0914) 83 (124) 7 (1) 0196 0852

anese adults Impact on health related quality of life J Epidemiol Global

Table 2 Socio-demographic characteristics of the study population

Characteristics Hypertension group N = 224 Control group N = 448 P-value

Age in years mean (SD) 5953 (1208) 4057(1323) lt0001

Sex n ()Males 90 (4017) 169 (3772) 0557Females 134 (59) 279 (6227)

Educational level n ()Illiterate 30 (1339) 12 (267) lt0001Elementary 73 (3225) 54 (1205)Intermediate 56 (2500) 100 (2232)Secondary 25 (1116) 92 (2053)University 40 (1780) 190 (4241)

Marital status n ()Single 11 (491) 87 (1941) lt0001Married 183 (8169) 354 (7901)Divorcedwidowed 30 (1339) 7 (156)

Employment n ()Unemployed 138 (6161) 195 (4352) lt0001Employed 60 (2678) 239 (5334)Retired 26 (1161) 14 (312)

Working hours n ()None 160 (7142) 207 (4621) lt0001[3ndash8] h 29 (1294) 85 (1227)[7ndash16] h 35 (1562) 156 (3482)

Monthly income n ()[500000ndash1000000 LL] 98 (4375) 116 (2589) lt0001[1000000ndash2000000 LL] 94 (4196) 239 (5334)[gt2000000 LL] 32 (1428) 93 (2075)

Living n ()Alone 37 (1651) 13 (29) lt0001With family 187 (8348) 435 (9709)

Financial support n ()Yes 146 (657) 89 (1986) lt0001No 78 (3482) 359 (8013)

BMI mean (SD) 28324 (4604) 26113 (3724) lt0001Underweight n () 0 (0) 4 (089)Normal n () 53 (2366) 178 (3973)Overweight n () 109 (4866) 202 (4508)Obese n () 62 (2767) 64 (1428)

Data presented as mean (SD) and number () were performed using T test and Chi2 respectively and a p-value lt 005 is consideredsignificant

Body mass index

Hypertension in the Lebanese adultsImpact on health related quality of life 5

factor affecting overall QOL (b = 11602p = 0001 95CI [18673 4531]) wherepatients with lower frequency of medication hadbetter QOL

In the multivariable linear regression modelestimating the predictors of PCS (Table 3) wefound that the frequency of medicationsday wasthe most significant inverse correlate(b = 11602 p = 0001 95CI [3139 101])

Please cite this article in press as Khalifeh M et al Hypertension in the LebHealth (2015) httpdxdoiorg101016jjegh201502003

Age (b = 0117 95CI [0173 0061]) andgender (b = 2555 95CI [4115 0995]) werealso inversely correlated with PCS (p lt 0001)Regular sport (b = 3461 p lt 0001 95CI [20024921]) and education (b = 2415 p = 000195CI [0934 3897]) were positively correlatedwith PCS while working effect was not statisticallysignificant (p = 0076) Moreover comorbiditieswere adversely associated with PCS (b = 3631

anese adults Impact on health related quality of life J Epidemiol Global

0

10

20

30

40

50

60

Hypertension

Control

Fig 1 Mean scores of each SF-8 domain comparing hypertension and control groups Data are presented as meansstatistical significance was tested using T test all p-values were lt 005

Table 3 Multivariable Analysis of QOL score and its components

Predictors Un stand b Stand b 95CI lower upper p-Value

Linear regression dependent variable QOL overallHypertension 17241 0128 30166 4317 0009Gender 14485 0111 23323 5648 0037Age 0388 0095 0752 0024 lt0001Education (University) 18065 0134 8451 2768 lt0001Regular sport 15937 0109 6362 25513 0001Number of comorbidities 8083 0120 14184 1982 0009Mood disorder 17241 0121 58011 16661 lt0001Frequency of drugsday 11602 0186 18673 4531 0001

Linear regression dependent variable physical component (PCS)Hypertension 2074 0144 5292 1332 0001Gender 2555 0114 4115 0995 0001Age 0117 0168 0173 0061 lt0001Education (University) 2415 0105 0934 3897 0001Regular sport 3461 0139 2002 4921 lt0001Working hours 1830 0760 0192 3467 0076No cigaretteday 0053 0066 1020 0005 0031Presence of comorbidity 3631 0166 5385 1878 lt0001Mood disorder 3250 0062 6435 0660 0045Frequency of drugsday 2074 0195 3139 1010 lt0001

Linear regression dependent variable mental component (MCS)Regular sport 1932 0077 0091 3774 004Smoking waterpipe 2555 0081 4908 0202 0033Drinking coffee 3084 0094 5498 0671 0012Mood disorder 7582 0143 11502 3663 lt0001Frequency of drugsday 1940 0181 2746 1134 lt0001

CI confidence interval p lt 005 is considered significant b regression coefficient

6 M Khalifeh et al

Please cite this article in press as Khalifeh M et al Hypertension in the Lebanese adults Impact on health related quality of life J Epidemiol GlobalHealth (2015) httpdxdoiorg101016jjegh201502003

Hypertension in the Lebanese adultsImpact on health related quality of life 7

p lt 0001 95CI [5385 1878]) whereashypertension (b = 2074 95CI [52921332]) and mood disorders (b = 17241 95CI[6435 066]) were negatively associated withPCS (p lt 0001) In addition higher daily cigarettesreduced PCS significantly (b = 0053 p = 003195CI [102 0005])

In the multivariable linear regression modelestimating the predictors of MCS (Table 3) regularsport was positively associated with MCS(b = 1932 p lt 005 95CI [0091 3774]) Thefrequency of medicationsday was the most signifi-cant risk factor to be negatively correlated withMCS (b = 194 p lt 0001 95CI [27461134]) The presence of mood disorders signifi-cantly lowered MCS (b = 7582 p lt 0001 95CI[11502 3663]) Moreover smoking waterpipe(b = 2555 p = 0033 95CI [4908 0202])and drinking coffee (b = 3084 p = 0012 95CI[5498 0671]) were associated with lowerMCS scores

35 Intragroup analysis of QOLhypertensive individuals

Female gender and age were significantly associ-ated with lower QOL (b = 21363 p = 000195CI [34353 8374] and b = 0548p = 0044 95CI [108 0016] respectively)Higher education level (b = 22949 p = 000695CI [6798 38101]) and regular sport(b = 2315 p lt 0001 95CI [9969 3633]) werepositively associated with QOL The presence ofdrugs side effects (b = 19262 p = 0031 95CI[36702 1822]) and of comorbidities(b = 13865 p = 0054 95CI [27992 262])were negatively associated with overall QOLMoreover the increase in daily cigarettes(b = 0726 p lt 0001 95CI [1127 0324])and salt consumption (b = 15728 p = 004995CI [31418 0038]) was adversely corre-lated with overall QOL The frequency of med-icationsday (b = 8193 p = 0058 95CI[16669 0282]) also reduced the overall QOL(Table 4)

In the multivariable linear regression modelestimating the predictors of PCS (Table 4) femalegender (b = 4308 p = 0006 95CI [343538374]) and older age (b = 0241 p lt 000195CI [108 016]) were strongly and inverselyassociated with PCS Practicing regular sport(b = 2315 p = 0017 95CI [9969 3633]) andincreasing working hours (b = 4493 p = 002295CI [0647 8339]) were positively correlatedwith PCS Moreover the number of cigarettesday(b = 0101 p = 0016 95CI [0184 0019])

Please cite this article in press as Khalifeh M et al Hypertension in the LebHealth (2015) httpdxdoiorg101016jjegh201502003

as well as the frequency of medication taken day were negatively associated with PCS(b = 1595 but p = 0079) Dyslipidemia negativelyaffected PCS (b = 4493 p = 0001 95CI[7259 2009]) In addition angiotensin con-verting enzyme inhibitorsangiotensin receptorsblockers (b = 3663 p = 0049 95CI [00125314]) and combination drugs had positive sta-tistically significant correlation with PCS(b = 2998 p = 003 95CI [0285 5712])However understanding the severity of the diseasewas associated with significant negative associa-tion on PCS (b = 5019 p = 0037 95CI[9735 0303])

In the multivariable linear regression modelestimating the predictors of MCS (Table 4) regularsport was positively correlated with MCS(b = 1932 p = 003 95CI [034 6752])Increasing the number of cigaretteday was themost significant risk factor negatively associatedwith MCS (b = 0117 p = 0015 95CI [02110032]) Drugs side effects were negatively corre-lated with MCS (b = 4584 p = 0003 95CI[8799 0369])

4 Discussion

The SF-8 health survey was internally consistentand a reliable tool for the assessment of the qualityof life of patients in Lebanon Hypertensivepatients had a poor QOL with lower scores in alldomains of the SF-8 in comparison with non-hyper-tensive individuals in both physical and mentaldomains taking into account differences in othercharacteristics Although this study contradictsthe traditional concept that claims hypertensionas an asymptomatic disease it is consistent withstudies that demonstrate the impact of hyperten-sion on QOL [9ndash11] The clinical significance ofthese differences in QOL is established [12] giventhat differences found between hypertensive andcontrol individuals were well above 2 for all physi-cal components items (11 points on PCS) andbetween 2 and 3 for mental component itemshypertension seems to affect mainly affect physi-cal health in the Lebanese population

Moreover the present study showed QOL wasaffected by several factors hypertensive malesand younger individuals had better QOL which cor-roborates the data obtained by Bardage and Isacson[11] For men the QOL score was better in alldomains The facts that men had better QOL scoresmay be possibly due to them being more tolerant tochronic diseases thus less emotionally affected bythem when compared to women [11] With regard

anese adults Impact on health related quality of life J Epidemiol Global

Table 4 QOL predictors among hypertensive patients

Predictors Un stand b Stand b 95CI lower upper p-Value

Linear regression (in hypertension group) dependent variable QOL overallGender 21363 0196 34353 8374 0001Age 0548 0123 1080 0016 0044Education (University) 22949 0164 6798 38101 0006Regular sport 23150 0199 9969 3633 lt0001Drugs side effects 19262 0125 36702 1822 0031Presence of comorbidity 13865 0117 27992 0262 0054Frequency of drugsday 8193 012 16669 0282 0058High Salt consumption 15728 0112 31418 0038 0049

Linear regression (in hypertension group) dependent variable physical component (PCS)Gender 4308 0185 7286 1330 0006Age 0241 0241 0353 0130 lt0001Regular sport 4200 0170 1474 7926 0017Working hours 4493 0143 0647 8339 0022Number of cigarettesday 0101 0138 0184 0019 0016Frequency of drugsday 1595 0110 3374 0185 0079Dyslipidemia 4493 0203 7259 2009 0001Angiotensin related drugs 3663 0112 0012 5314 0049Combination 2998 0121 0285 5712 003Understand severity of hypertension 5019 0117 9735 0303 0037

Linear regression (in hypertension group) dependent variable mental component (MCS)Regular sport 3546 0144 0340 6752 0030Number of cigarettesday 0117 0159 0211 0032 0015Drugs side effect 4584 0141 8799 0369 0033

CI confidence interval p lt 005 is considered significant b regression coefficient

8 M Khalifeh et al

to age younger hypertensive patients had betterQOL in fact during the aging process health haz-ards may arise as a result of physiological changesmaking the individual more prone to chronic dis-eases which can affect QOL [10] Patients with ahigher level of education had higher QOL scoresin overall QOL and PCS The level of educationinfluences the ability to understand informationwhich favors knowledge on the disease and treat-ment adherence Therefore patients with a higherlevel of education may have better QOL scores[20] Individuals with employment status andhigher working hours also had higher QOL scoresin the physical domain and this finding is in agree-ment with other study [21]

Furthermore the dimensions of the QOL scorewere found to be significantly associated withmany lifestyle factors Smoking was significantlyand negatively associated with QOL in the selectedpopulation as well as the magnitude of this associa-tion was determined by the number of cigarettessmoked Waterpipe smoking was also significantlyassociated with poorer mental score outcomespossibly associated with a number of deleterioushealth outcomes due to the withdrawal effect ofhigh nicotine concentrations This finding is similar

Please cite this article in press as Khalifeh M et al Hypertension in the LebHealth (2015) httpdxdoiorg101016jjegh201502003

to those reported by Wilson et al [22] Howeverregular sport was significantly and positively corre-lated with QOL in all domains Remaining physicallyactive can enhance functional capacity and mentalwell-being to maintain QOL and independenceBesides caffeine was negatively associated withMCS score Caffeine is known to elevate systolicblood pressure and diastolic blood pressure at restand during mental and exercise stress [23] In addi-tion high salt consumption was shown to reduceQOL significantly in hypertensive group A strongassociation between increased sodium intake andblood pressure has been shown to be correlatedwith QOL [24] This leads to stronger enthusiasmfor strict control of sodium intake for hypertensivepeople

As expected the presence of comorbidities wasassociated with lower QOL in hypertensivepatients This finding is consistent with that ofWang et al [19] showing a greater decline in func-tioning than those with only one conditionDepression and anxiety were significantly moreassociated with lower score of QOL This is due tothe negative effect of depression on mood perfor-mance and eventually on the functional capacityAlso we observed that dyslipidemia significantly

anese adults Impact on health related quality of life J Epidemiol Global

Hypertension in the Lebanese adultsImpact on health related quality of life 9

reduced QOL in hypertensive patients This is dueto the negative effect of dyslipidemia on the physi-cal performance These findings suggest that effec-tive treatment and prevention of chronic diseasesmay be important to preserve the QOL the lattersuggestion remains to be established in adequateprospective studies

As for medications the use of angiotensin con-verting enzyme inhibitors angiotensin receptorblockers and drug combinations improved the physi-cal score This finding is in concordance with a pre-vious study that analyzed selected aspects of theQOL such as physical performance [25] Howeverb blockers may actually worsen the QOL [26] whichis consistent with our results Of the drugs evaluatedin our study calcium channel blockers and diureticsappeared to have non-significant effects on the QOLand were associated only with a decline in physicalhealth The present study also revealed a significantunfavorable influence of drug side-effects on qualityof life particularly in overall QOL and MCS Studiesshow that there is a deficit in the QOL of hyperten-sive patients because of the adverse effects of thedrugs used in the antihypertensive treatment [9]It is therefore essential for clinicians to select thedrug best tolerated by individual patients to balancetheir needs with the best quality of life Moreoverthe increase in the dose frequency per day and theincreased number of prescribed medications werethe primary factors associated with lower QOL inhypertensive patients This is consistent withanother study suggesting that the overuse of med-ications decrease the QOL [19] The use of over-the-counter medications was also associated withlower QOL probably because of the perception ofovermedication or the consequences of adverseeffects on hypertensive patients Knowledge of thehypertension severity was significantly correlatedwith PCS perhaps because patients felt stigmatizedafter discovering that they had the diseaseMoreover neither drug cost per patient nor thepresence of a medical insurance could predict theQOL This finding is supported by evidence fromliterature [27] Finally when evaluating the influ-ence of the number of antihypertensive drugs thecontrolled status of BP and the duration of hyperten-sion on QOL no significant difference were found inany of the domains studied However another studyverified that hypertensive patients who were undertreatment and whose BP levels were under controlhad higher QOL than those with uncontrolled bloodpressure [10] This discrepancy may be explainedby the low subsample size

Although our study has several points ofstrength since this is the first study that formally

Please cite this article in press as Khalifeh M et al Hypertension in the LebHealth (2015) httpdxdoiorg101016jjegh201502003

assesses the QOL of hypertensive patients com-pared with general individuals in Lebanon it hasseveral limitations To begin with it is a cross-sec-tional study therefore it does not allow causalityto be attributed to the observed associations Thebaseline characteristics differ between the hyper-tension and control group this may introducepotential confounding factors that were taken intoaccount during multivariable analyses howeverresidual confounding may still be possible due toconfounders that we may have not adequatelymeasured Moreover there could also be a possibil-ity of respondent and information bias since theresults of our study were based on a face to facequestionnaire Another study limitation was theincapacity to assess the influence of the antihyper-tensive drug associations on QOL of hypertensiveindividuals due to the large number of combina-tions In addition the lower than expected preva-lence of common diseases (eg peripheral arterydisease neuronal diseases chronic venous insuffi-ciency) may suggest that polymorbidity was under-reported However since the associations wefound were all consistent with the literature andof biological plausibility we have no reason tobelieve that our results suffer from seriousmethodological flaws

5 Conclusion

In Lebanon hypertensive patients presented alower QOL scores in all domains of SF-8 than non-hypertensive individuals Therefore it is essentialto find medical and social alternatives that have afavorable influence on QOL this can be achievedby the development of drugs that provide bloodpressure control and favorable effects on QOL withminimal side effects Hence effective healthinterventions should ensure maintenance of desir-able QOL in order to prevent or reduce comorbidi-ties of hypertension There is a necessity for healthprofessionals to take these factors into accountwhen treating hypertensive patients and to tacklespecial subgroups with attention to their deterio-rated QOL

Conflict of interest

No potential conflicts of interest were disclosed

Acknowledgement

The authors thank the Lebanese University for fundingthis project

anese adults Impact on health related quality of life J Epidemiol Global

10 M Khalifeh et al

References

[1] Nogueira D Faerstein E Coeli CM Chor D Lopes CSWerneck GL Reconhecimento tratamento e controle dahipertensao arterial Estudo Pro-Saude Brasil Rev PanamSalud Publica 201027(2)103ndash9

[2] Rosini N Machado MJ Xavier HT Estudo da prevalencia emultiplicidade de fatores de risco cardiovascular emhipertensos do municıpio de Brusque SC Arq Bras Cardiol200686(3)219ndash22

[3] World Health Organization Causes of Death Geneva 2008Available at httpwwwwhointmediacentrefact-sheetsfs317en

[4] Sociedade Brasileira de Cardiologia Sociedade Brasileira deHipertensao Sociedade Brasileira de Nefrologia VIDiretrizes brasileiras de hipertensao Arq Bras Cardiol201095(1)1ndash51 (Suppl 1)

[5] Lim SS Vos T Flaxman AD Danaei G et al A comparativerisk assessment of burden of disease and injury attributableto 67 risk factors and risk factor clusters in 21 regions1990ndash2010 a systematic analysis for the Global Burden ofDisease Study 2010 Lancet 2012380(9859)2224ndash60

[6] World Health Organization WHO (1998) Global burden ofdiabetes Press Release WHO63 September 14 1998Available at httpwwwwhointinf-pr-1998enpr98-63html

[7] Stewart AL et al Functional status and well-being ofpatients with chronic conditions Results from the MedicalOutcomes Study J Am Med Assoc 1989262(7)907ndash13

[8] Plaisted CS et al The effects of dietary patterns on qualityof life a substudy of the Dietary Approaches to StopHypertension trial J Am Diet Assoc 199999(Suppl 8)S84ndash9

[9] Roca-Cusachs A Dalfo A Badia X Aristegui I Roset MRelation between clinical and therapeutic variables andquality of life in hypertension J Hypertens200119(10)1913ndash9

[10] Li W Liu L Puente JG Li Y Jiang X Jin S et alHypertension and health related quality of life an epi-demiological study in patients attending hospital clinics inChina J Hypertens 200523(9)1667ndash76

[11] Bardage M Isacson DGL Hypertension and health relatedquality of life epidemiological studies in Sweden J ClinEpidemiol 200154172ndash81

[12] Ware JE Kosinki M Dewey J Gandek B How to Score andInterpret Single-Item Health Status Measures A Manual forUsers of the SF-8 Health Survey Boston QualityMetric Inc2001 4ndash8

[13] Tohme RA Jurjus AR Estephan A The prevalence ofhypertension and its association with other cardiovasculardisease risk factors in a representative sample of theLebanese population J Hum Hypertens 200519(11)861ndash8

Please cite this article in press as Khalifeh M et al Hypertension in the LebHealth (2015) httpdxdoiorg101016jjegh201502003

[14] Mafawzy M Mokhtar N Wan Mohammad WB Hypertensionand associated cardiovascular risk factors in Kelantan MedJ Malaysia 200358(4)556ndash64

[15] Tanuseputro P et al Canadian cardiovascular outcomesresearch team risk factors for cardiovascular disease inCanada Can J Cardiol 200319(11)1249ndash59

[16] Wood S JNC8 at last Guidelines ease up on BP thresholdsdrug choices Heartwire [serial online] Assessed December30 2013 Available at http wwwmedscapecomviewarticle817991

[17] The World Health Organization Quality of Life Assessment(WHOQOL) development and general psychometric proper-ties Soc Sci Med 1998 46(12) 1569ndash5

[18] Lohr KN Aaronson NK Alonso J Burnam MA Patrick DLPerrin EB Roberts JS Evaluating quality-of-life and healthstatus instruments development of scientific reviewcriteria Clin Ther 199618(5)979ndash92

[19] SF-8TM Health Survey Scoring Demonstration available athttpwwwsf-36orgdemosSF-8html and httpwwwqualitymetriccomWhatWeDoGenericHealthSurveysSF8HealthSurveytabid187Defaultaspx

[20] Brunye T Mahoney C Rapp D Ditman T Taylor H Caffeineenhances real-world language processing evidence from aproofreading task J Exp Psychol Appl 20121895ndash108

[21] Wang R Zhao Y He X Ma X Yan X Sun Y et al Impact ofhypertension on health-related quality of life in a pop-ulation-based study in Shangai China Public Health2009123(8)534ndash9

[22] Wilson D Parsons J Wakefield M The health-relatedquality-of-life of never smokers ex-smokers and lightmoderate and heavy smokers Prev Med 199929139ndash44

[23] Terry Hartley R Sung Bong Hee Pincomb Gwendolyn AWhitsett Thomas L Wilson Michael F Lovallo William RHypertension risk status and effect of caffeine on bloodpressure Hypertension 200036137ndash41

[24] Nancy J Aburto Anna Ziolkovska Lee Hooper Paul ElliottFrancesco P Cappuccio Joerg J Meerpohl Effect of lowersodium intake on health systematic review and meta-analyses BMJ 2013 346 doi httpdxdoiorg101136bmjf1326 (Published 4 April 2013)

[25] Tedesco MA Ratti G Mennella S et al Comparison oflosartan and hydrochlorothiazide on cognitive function andquality of life in hypertensive patients Am J Hypertens1999121130ndash4

[26] Kawecka-Jaszcz K Klocek M Tobiasz-Adamczyk B Qualityof life in patients with arterial hypertension In Hhjh Meditor Quality of Life in Cardiovascular Diseases TermediaWydawnictwo Medyczne 2006 p 122

[27] Rolnick J Pawloski P et al Patient characteristics associ-ated with medication adherence Clin Med Res201311(2)54ndash65

ScienceDirectAvailable online at wwwsciencedirectcom

anese adults Impact on health related quality of life J Epidemiol Global