Journal of Human Hypertension (2003) 17, 21–27 & 2003 Nature Publishing Group All rights reserved 0950-9240/03 $25.00 www.nature.com/jhh ORIGINAL ARTICLE Increased prevalence of pneumoniae but not Epstein–Barr antibodies in essential hypertensives

V Ch Pitiriga, V Th Kotsis, M-EK Alexandrou, VD Petrocheilou-Paschou, N Kokolakis, RN Zakopoulou and NA Zakopoulos Department of Clinical Therapeutics, Alexandra Hospital, University of Athens, National and Kapodistrial University, Athens, Greece

Conflicting data exist regarding the relationship be- without antihypertensive medication were defined as tween Chlamydophila pneunoniae (C. pneumoniae) and hypertensives. Controls were free of any history or hypertension. In this study, both C. pneumoniae IgG and clinical evidence of hypertension, cardiovascular or IgA titres and Epstein–Barr virus antibody levels were pulmonary disease. Of the total participants, 77 hyper- measured in 146 sustained hypertensives defined by tensives (52.7%) and 10 normotensives (18.5%) had IgA 24 h ambulatory blood pressure monitoring (ABPM) and titres X40 (crosstabs Po0.000), whereas 76 hyperten- 54 normotensives. C. pneumoniae antibodies were sives (52.1%) and 15 normotensives (27.8%) had IgG measured by microimmunofluorescence test. IgGX80 titres X80, (crosstabs Po0.002). No difference was and IgAX40 were defined as elevated antibody titres. found in Epstein–Barr antibodies, between hyperten- Epstein–Barr antibodies were measured in order to sives and normotensives. In conclusion, C. pneumo- investigate whether a possible association exists be- niae, but not Epstein–Barr, antibody levels were found tween hypertension and other, similarly widespread in significantly higher in sustained hypertensives, sug- the general population, intracellular microorganisms. All gesting high frequency of chronic C. pneumoniae, participants underwent casual blood pressure (BP) in this specific group of patients. readings and 24 h ABPM. Subjects having mean 24 h Journal of Human Hypertension (2003) 17, 21–27. systolic/diastolic ambulatory BP4125/80 mmHg, with or doi:10.1038/sj.jhh.1001501

Keywords: C. pneumoniae; hypertension; ; ; ABPM

Introduction tension.13,14 This discrepancy may be the result of performing different serological methods (enzyme- Chlamydophila pneumoniae is an intracellular linked immunosorbent assay (ELISA), microimmuno- Gram negative bacterium that is currently recog- fluorescence (MIF); it could be traced in the nized as a significant respiratory pathogen. It can examination technicalities used to define hyperten- cause acute and chronic pulmonary infections such sion in those studies, or in the sampling procedure. as , , sinusitis and 1 In order to clarify whether there is a real in both adults and children. Several studies have association between C. pneumoniae and essential documented an association between serologic evi- hypertension, C. pneumoniae IgA and IgG antibody dence of C. pneumoniae infection and various forms 2–4 levels were determined in a selected population of of vascular disease, such as coronary heart disease, sustained hypertensives and in healthy normoten- carotid atherosclerotic lesions5,6 and ischaemic 7 sives. As a result of the existing controversy on cerebrovascular disease. However, a number of whether white-coat hypertension is a benign condi- studies have failed to establish such an associa- 8–12 tion or a pathophysiological entity leading to similar tion. complications with sustained hypertension,15,16 Similarly, in prior research efforts, conflicting white-coat hypertensives were excluded from this findings exist regarding the relation between study, to avoid possible methodological errors. In C. pneumoniae antibodies and essential hyper- order to identify correctly sustained hypertensives, all participants underwent 24 h ambulatory blood pressure monitoring (ABPM). Correspondence: Dr N Zakopoulos, Department of Clinical Therapeutics, Alexandra Hospital, Vas. Sofias and Lourou str Antibodies against Epstein–Barr virus (EBV IgM, 11528, Athens, Greece. E-mail: [email protected] EBV IgG, EB nuclear (EBNA) IgG, EBV Received 2 June 2002; revised and accepted 11 October 2002 EA IgG were also measured to examine whether C. pneumoniae and Epstein–Barr infection in hypertension V Ch Pitiriga et al

22 a possible association between essential sustained number of years as a smoker times the average hypertension and infections concerns, except for number of smoked cigarettes. In regards to alcohol C. pneumoniae, and other intracellular micro- consumption, the case and control subjects were . It was chosen among others, based classified as abstainers and heavy drinkers on the fact that, like C. pneumoniae, infections (men460 g/day; women440 g/day alcohol intake, by EBV are frequently observed in the Hellenic with reference to the suggestions of the Italian population. Alcohol Society19). Blood glucose, cholesterol, tri- glycerides, HDL, LDL, C-reactive protein were determined. Materials and methods In all, 14 hypertensives refused to participate. Three hypertensives and one control were excluded Hypertensives were selected among patients con- because they did not meet the serologic criteria secutively referred to the Outpatient University (IgGX512). Eventually, 146 hypertensive patients Hypertension Center of the Department of Clinical Fusing antihypertensive medication or notFand Therapeutics at Alexandra University Hospital of 54 normotensive controls were included in the Athens. They were included in the sample if all of study. All participants gave informed consent before the following criteria were met: (a) mean 24 h entering the study. systolic/diastolic ambulatory BP4125/80 mmHg (in accordance with WHO 1999)17 and (b) history of hypertension (BP4140/90 mmHg, based on three sphygmomanometer measurements during each Clinic BP measurements examination, repeated in three different exami- BP was measured three times using a mercury nations in 1 month duration). The former criteria sphygmomanometer (standard cuff applied around for sustained hypertensives is in compliance with the nondominant arm and systolic and diastolic the existing literature, according to which the 24 h values identified from the first and fifth phase of ABPM is a highly reproducible procedure, adequate Korotkoff sounds). During the measurements, the to establish the long-term hypertensive status of an 17,18 study participants remained seated with the arm individual. placed at the heart level. The three initial BP values Controls were selected among healthy subjects were averaged to obtain a single systolic and proceeding to the hospital for routine medical diastolic ‘clinic’ value. check-up, including ECG, chest radiograph, echo- cardiography and ABPM. In the control subjects, the following criteria were met: (a) no history of hypertension, based on previous and past BP ABPM measurements (subjects who reported even once in their life BP values X140/90 mmHg were excluded), All participants underwent 24 h ABPM that was (b) no history of antihypertensive therapy, (c) BP performed using oscillometric Spacelabs 90209 measurements by sphygmomanometer at the out- equipment (Spacelabs, Redmond, Washing- patient unit o140/90 mmHg on the first and second ton, USA). The monitoring equipment was visit (on the next day) and mean 24 h ambulatory BP applied at the end of the medical visit. The o125/85 mmHg. device was set to obtain automatic BP readings at Finally, all participants were enrolled only if they 15-min intervals. The monitoring was always had no evidence (history, physical examination, performed on a working day and the subjects were ECG, chest radiograph, echocardiography) of other instructed to act and work as usual. Eventually, coexisting cardiovascular or pulmonary disease; no 80–96 pairs of systolic and diastolic BP values/ evidence (history, physical examination, previous patient/24 h values and times were fed into a PC/ laboratory tests provided by the patients) of second- Pentium II computer. No subject having fewer than ary hypertension (Cushing’s, Conn’s syndrome, three readings per hour, in any of the 24 h, was pheochromocytoma, etc) immunodeficiency, hyper- included in the study. The accuracy of the ABPM gammaglobulinaemia, autoimmune disease, diabetes system was tested against mercury sphygmoman- mellitus; normal results of serum urea, creatinine, ometer readings, both before each patient was electrolytes, plasma renin activity at rest and after leaving the hospital and on the following day when exercise, urinary excretion of catecholamines; he returned. and normal renal imaging studies (radionuclide renogram, intravenous pyelography or ultra- sound scan). Definition of white-coat hypertension Body weight and height were measured as body surface area (BSA). Occupation, residence, educa- It was defined as the combination of raised clinic tion, alcohol abuse, smoking habits, hyperlipidae- BP (4140 mmHg systolic and/or 90 mmHg diastolic mia, obesity and diabetes mellitus were recorded in BP) with average daytime ambulatory BPo135/ each subject. ‘Cigarette-years’ were calculated as the 85 mmHg.20

Journal of Human Hypertension C. pneumoniae and Epstein–Barr infection in hypertension V Ch Pitiriga et al

23 Serologic analyses reference tests are as follows: for EBV IgM ELISA test 97.4 and 98.3%, respectively; for EBV IgG Venus blood was drawn at the entry for the ELISA test 78.1 and 90.9%, respectively; for EBNA determination of C. pneumoniae IgG and IgA ELISA test 86.5 and 95.3%, respectively; for Early antibodies. Serum aliquots were kept frozen at ELISA test 90.1 and 96.5%, respectively. À201C, until analysed within a few days. The The C. pneumoniae MIF IgG assay sensitivity is presence of C. pneumoniae-specific IgG and IgA 65% (by sera) and the specificity is 96%. The antibodies in serum was determined by MIF tech- C. pneumoniae MIF IgA assay sensitivity is 72% nique. Elementary bodies of C. pneumoniae TW 183 (by sera) and the specificity is 94%. strain were used as antigen (BIOS Gmbh Postfach 1640 D-82166 Grafelfing, Munchen, Germany). In order to determine IgA titres, sera were absorbed Statistical analysis with Gullsorb (Gull Laboratories) at a dilution of Data are presented as mean 7 s.d. Analysis was 1/17 to remove all IgG antibodies.21 Serum dilutions performed using the SPSS/PS+ statistical package were incubated for 14–15 h at À41C before being (SPSS, Inc, Chicago IL, USA). In an effort to obtain applied to antigen slides containing elementary an indication about the appropriate sample size body preparations of C. pneumoniae. After the regarding both groups, Chebyshev’s inequality was procedure was performed, coverslips were mounted used. Specifically, at the confidence level of 95%, with buffered glycerol and slides were examined the appropriate sample size for the controls was 20 under a Zeiss UV microscope with a Â40 oil and 45 subjects for IgA and IgG, respectively, while immersion lens and a Â10 ocular lens. Control sera the appropriate sample size for the group of were included in every test run. All sera were hypertensives was 85 and 130 subjects for IgA and screened initially at three dilutions: 1/10, 1/80, IgG, respectively. Using the above procedure, it was 1/160, for IgG, and 1/20, 1/40, 1/160 for IgA assumed that 1 s.d. difference between the samples’ antibodies. Thereafter, positive sera were tested at mean and the population mean would suffice. higher dilutions 1/320 and 1/512. IgG titre X80 and Crosstabs Pearson w2 test was used to examine the IgAX40 have been defined as the criterion of frequency of C. pneumoniae seropositivity and elevated C. pneumoniae antibody levels. The same seronegativity between normotensives and hyper- investigator examined all slides without knowledge tensives, smokers and nonsmokers, males and of whether the samples belonged to patients or females. Independent-samples t-test was used to controls. Patients having IgG titres X512 were relate C. pneumoniae seropositivity with age, bio- excluded from the study, as they were considered chemical parameters, ABPM variables and clinic BP having acute infection or reinfection. values. Multivariable discriminant stepwise ana- The ELISA method (Gull Laboratories, Inc.) was lyses was used to estimate the possible influence of used in order to detect EBV IgM, EBV IgG, EBNA IgG risk factors for C. pneumoniae infection and hyper- and EBV EA IgG antibodies of EBV in human serum. tension such as age, obesity, gender, hyperlipidae- Blood was collected and centrifuged after clotting at mia, smoking habit. A P-value 0.05 was consi- 1500 g for 10 min. Serum was separated and stored o dered significant. at À201C. In order to perform the ELISA test, sera were loaded into cups, which were placed in the BIOS LAB unit and processed according to the Results manufacturer’s instructions. Antibody titres were expressed as activity units per millilitre (AcU/ml). Values of clinic BP and ambulatory monitoring The IgM response to viral capsid antigen (VCA) is derivatives, demographic characteristics and sero- usually present at the onset of the disease, peaking logic analyses in the two groups examined are listed within 4–6 weeks and become undetectable within in Table 1. The two groups did not differ in 2–3 months from onset of clinical symptoms. IgG socioeconomic status, age, male/female ratio by response is usually present at the onset of the selection, and in lipid profile, blood glucose level, disease and remains detectable for life. Antibodies CRP and smoking habit. Only two hypertensives and to EBNA gradually increase in titre, reaching plateau one normotensive were heavy drinkers. levels after 3–12 months. They persist for life like Of the total participants, 77 hypertensives (52.7%) IgG antibodies to VCA. Positive results for EBV IgG and 10 normotensives (18.5%) had IgA titres X40 and EBNA IgG were considered as best indication of (Po0.000). Of these, 76 hypertensives (52.1%) and past infection with EBV. VCA EBV IgG positive and 15 normotensives (27.8%) had IgG titres X80, EBNA IgG, negative cases were further differentiated (Po0.002). by determination of EBV EA (Early) IgG. All sera EBV IgM and EBV EA (Early) antibodies were not positive for EBV VCA IgM and EBV EA (Early) IgG detected in any of the hypertensive and normoten- were excluded, as indication of active EBV infec- sive subjects examined. No statistically significant tion.22 differences were found in EBV IgG and EBV EBNA The relative sensitivity and specificity of the antibody levels between hypertensives and control above serologic tests when compared with the group.

Journal of Human Hypertension C. pneumoniae and Epstein–Barr infection in hypertension V Ch Pitiriga et al

24 Table 1 Characteristics of hypertensives and normotensives

Variable Normotensives Hypertensives tP-value n=54 (Mean 7 s.d.) n=146 (Mean 7 s.d.)

Sex (Male/female) 20/34 58/88 F NS Age (years 7 s.d.) 46.46 7 13.18 44.62 7 8.55 1.15 NS Weight (kg) 69.3 7 13.52 78.32 7 16.12 À3.2 0.001 Height (cm) 166.38 7 8.6 168.35 7 10 À1.13 NS Body surface area (m2) 1.7 7 0.18 1.8 7 0.2 À2.9 0.004 Smoking (no/yes) 32/22 90/51 F NS Smoking, cigarattes-years 540 7 259 553 7 310 À0.3 NS Serum glucose (mg/dl) 93.7 7 15.94 93.44 7 17.10 0.1 NS Serum cholesterol (mg/dl) 213.89 7 36.8 220.5 7 38.9 À0.8 NS Serum triglycerides (mg/dl) 102 7 44.17 124.44 7 61.05 À2.26 0.01 Serum HDL(mg/dl) 55 7 11.16 50.92 7 13.68 1.8 NS Serum LDL (mg/dl) 134.31 7 34.4 142.08 7 47.5 À1.05 NS CRP (mg/L) 0.54 7 0.52 0.58 7 0.87 À0.38 NS Clinic SBP (mmHg) 118 7 12.12 148.76 7 18.07 À10.4 0.000 Clinic DBP (mmHg) 79.07 7 8.95 96.81 7 10.97 À9.6 0.000 Mean SBP 24 h (mmHg) 113.31 7 8.3 130.65 7 14.76 À7.5 0.000 Mean DBP 24 h (mmHg) 70.33 7 6.4 81.74 7 11.2 À6.3 0.000 EBV EBNA IgG (AcU/ml) 126 7 91.8 143.53 7 110.19 1.04 NS EBV IgG (AcU/ml) 37.01 7 28.12 37.51 7 31.2 0.50 NS

NS, not significant.

C. pneumoniae IgG and IgA seropositivity was not not examined the chlamydial IgA and IgG antibody associated with age and smoking habit. A statisti- titres in different levels of hypertension, so this cally significant difference in C. pneumoniae IgG study cannot provide further information on and IgA titres was found between men and women whether there is a connection between the severity (Po0.015 and Po0.016, respectively). of hypertension and the serological C. pneumoniae On multivariable discriminant stepwise analyses, antibody levels. It should be noted that the study IgA antibodies were in the first level associated with group consisted of patients who had hypertension hypertension (F-statistic ¼ 19.725, Po0.000), in the (therefore some degree of cardiovascular disease in second level with gender (F-statistic ¼ 11.966, its broader sense) and the controls who had no such Po0.000) and in the third level with age (F- history. As a result, it could be argued that a bias statistic ¼ 9.459, Po0.000). No significant associa- towards cardiovascular disease exists. However, tion was indicated between C. pneumoniae IgA hypertensive subjects in our study had mild hyper- antibodies and smoking habit and obesity. C. tension and were selected based on the absence of pneumoniae IgG antibodies were significantly asso- evidence of end organ disease. Consequently, the ciated with hypertension in the first level (F- possibility of vascular damage in this group was statistic ¼ 10.148, Po0.002) and gender in the minimized. second level (F-statistic ¼ 8.408, Po0.000). No sig- C. pneumoniae IgG and IgA seropositivity was not nificant relation was indicated between IgG anti- associated with age and smoking habit. A statisti- bodies and smoking habit, age and obesity. In the cally significant difference in C. pneumoniae IgG first level hypertension was related with IgA anti- and IgA titres was found between men and women, bodies (F-statistic ¼ 19.725, Po0.000) and in the in agreement with other reports.23 No association second level with obesity (F-statistic ¼ 12.594, was demonstrated regarding Epstein–Barr infection Po0.000), but not related to IgG antibodies, gender and subjects with essential hypertension. No sig- and age. nificant difference existed between normotensives and hypertensives in age, gender and smoking, risk factors for the prevalence of C. pneumoniae infec- Discussion tion. We further observed that between the two groups examined, body mass index is higher in the The characteristics of the two groups examined are group of hypertensives than in normotensives. listed in Table 1. Our findings show that hyper- However, this difference could not have possibly tensives had significantly higher C. pneumoniae IgG affected our results. Furthermore, there is no strong and IgA titres, comparing to normotensives. The evidence in the literature up to now connecting the association between hypertension and C. pneumo- prevalence of C. pneumoniae infection with the niae antibodies remained after the use of multi- body mass index. variable analyses, and we focused on an interesting We used the MIF test to define antibody titres. finding that hypertension was most strongly asso- Although it is the most currently used method for ciated with C. pneumoniae IgA antibodies. We have the serologic diagnosis of C. pneumoniae infection,

Journal of Human Hypertension C. pneumoniae and Epstein–Barr infection in hypertension V Ch Pitiriga et al

25 it is bedeviled by a high level of operator error,24 with white-coat hypertension.35 To avoid the inclu- and by cross-reactions with other chlamydophilae25 sion of patients with white-coat hypertension that and mycoplasma species.26,27 In addition, some might affect our results, we conducted this study, investigators support that the detection of C. using, in addition to the clinic BP measurements, pneumoniae antibodies may be a poor guide to 24 h ABPM. the presence of infection.28 As a consequence of the No association between Epstein–Barr infection above limitations, it is generally accepted and hypertension was demonstrated by our find- that, for C. pneumoniae, is a seriously ings. EBV is an intracellular agent, which, in the flawed tool, since it cannot provide high precision. past, has been investigated for implications in Additionally, the lack of the performance standardi- atherosclerosis,36–38 and it is also widespread in zations and the variety of the cutoff titres have the Hellenic community. Although C. pneumoniae made the comparison of data from different labora- is classified as a , it shares many tories quite difficult. We considered as positive, a similarities with viruses, as it is an obligate higher threshold (X40) than other studies (X8and intracellular organism that diverts the ’s X16),29,4,30 in order to minimize the percentage of energy production mechanisms and lyses the host false-positive results. IgG antibody titres X512 were cell at the end of the replication cycle.39 If EBV- excluded from our study, as evidence of acute C. elevated antibody levels were more frequently found pneumoniae infection or reinfection. Certainly, an in hypertensives, it would have led us to the acute infection cannot be associated with hyperten- speculation that, in general, they are more suscep- sion because of its short-lived effect on the human tible to infections. Thus, the presence of infections organism. in general, suggesting a low immunoresistance, may PJ Cook et al (1997)13 have reported an asso- be an important risk factor that leads to a chronic ciation between C. pneumoniae IgG antibody form of C. pneumoniae infection,40 and as a result, titres and severe essential hypertension. IgG the association between chronic C. pneumoniae antibodies were measured to define three groups infection and hypertension would have been less ‘acute’, ‘previous’ or ‘no chlamydial infection’. clear. Our results are in agreement with the above The mechanisms underlying this positive associa- findings. However, in this study, C. pneumoniae tion between C. pneumoniae infection and essential IgA antibodies were not examined. There is con- hypertension are not clarified. C. pneumoniae could siderable evidence that, of all C. pneumoniae either be an innocent bystander in atheromatous antibodies, the presence of IgA may be the best plaques resulting from high BP, or play a pathogenic marker of chronic C. pneumoniae infection.31,32 role in the progression of established plaques. Circulating IgA immunoglobulin is short-lived Nevertheless, it is possible that C. pneumoniae does (half-life 5.8 days); hence, its presence implies that contribute to the pathogenesis and if so, it could the antigen is still present, and the infection is be through a large number of potential mechanisms. chronic and persistent. Chronic chlamydial infec- Arterial baroreflex control of the BP, located tion could be a contributing factor in the develop- in specific regions of the arterial wall, is stimulated ment of hypertension, by triggering a chronic by vessel distensibility resulting from spontaneous inflammatory process. arterial pressure changes. It has been demonstrated Nishimura and co-workers9 found suggestive that atherosclerotic alteration in the architecture evidence of an inverse association between C. of the vessel wall, where baroreceptive endings pneumoniae and high BP in Japanese adults. terminate, is considered the predominant mechan- Specifically, C. pneumoniae antibodies were deter- ism responsible for the degreased baroreflex sensi- mined by a different method, the ELISA and clinic tivity, leading eventually to hypertension.41–44 measurementsFbut not 24 h ambulatory BP mea- Accepting the theory that C. pneumoniae contri- surementsFwere provided. Based on the above butes to the pathogenesis of atherosclerosis, the findings, our research team assumes that the inverse settlement of this microorganism in these vessel association could be explained by the genetic areas, by inducing a chronic immune response differences of the Japanese population (different orchestrated by cytokines, activation of monocytes Ethnic origin).33 Similarly, a number of studies and changes in lipid profile, contributes to structur- suggest that Japanese people have a higher inci- al injury of intima and media. Consequently, the dence of variant angina and induced coronary artery findings of the present study provide suggestive spasms after than Caucasian evidence that C. pneumoniae contributes as a risk patients, because of important racial differences.34 factorFcombined with otherFto the genesis of Of course, the previous claims require further hypertension. investigation. In our study, the sampling population Nevertheless, the specific underlying pathophy- had no ethnic variation. In extension, the classifica- siological mechanisms that link C. pneumoniae tion of normotensives and hypertensives by the with essential hypertension have not yet been sphygmomanometric technique, used in Nishi- defined. Therefore, further research, in this field, mura’s study, does not allow the identification of is required to reinforce the evidence presented in subjects with sustained hypertension and subjects this study.

Journal of Human Hypertension C. pneumoniae and Epstein–Barr infection in hypertension V Ch Pitiriga et al

26 References 20 Verdecchia P, Staessen JA, White WB, Imai Y, O’Brien ET. Properly defining white-coat hypertension. Eur 1 Grayston JT. Infection caused by pneumo- Heart J 2000; 23: 106–109. niae strain TWAR. Clin Infect Dis 1992; 5: 757–763. 21 Jauhianinen T et al. Interference of immunoglobulin G 2 Thom DH et al. Association of prior infection with antibodies in IgA antibody determinations for C. C. pneumoniae and angiographically demonstrated pneumoniae by microimmunofluorecence test. J Clin . JAMA 1992; 268: 68–72. Microbiol 1994; 32: 839–840. 3 Mendall MA et al. Chlamydia pneumoniae: risk factors 22 Bauer G. The rational basis for efficient Epstein–Barr for seropositivity and association with coronary heart virus (EBV) serology. Review. Clin Lab 1995; 41: disease. J Infect 1995; 30: 121–128. 623–634. 4 Saikku P et al. Serological evidence of an association of 23 Saikku P. The epidemiology and significance of Chla- a novel chlamydia TWAR with chronic coronary heart mydia pneumoniae. J Infect 1992; 25 (Suppl 1): 27–34. disease and acute myocardial infarctions. Lancet 1988; 24 Kutlin A et al. Evaluation of Chalamydia immunoglo- 2: 983–986. bulin M (IgM), IgG, and IgA rELISAs Medac for 5 Melnick LS et al. Past infection by Chlamydia diagnosis of Chalamydia pneumoniae infection. Clin pneumoniae strain TWAR and asymptomatic carotid Diag Lab Immunol 1997; 4: 213–216. atherosclerosis. Atherosclerosis Risk in Communities 25 Hughes C et al. Possible nosocomial transmission of (ARIC) Study investigators. Am J Med 1993; 95: . Infect Control Hosp Epidemiol 1997; 18: 499–504. 165–168. 6 Grayston JT et al. Chlamydia pneumoniae (TWAR) in 26 Ossewaarde JM, de Vries A, Bestebroer T, Angulo AF. the atherosclerosis of carotid artery. Circulation 1995; Application of a Mycoplasma group-specific PCR for 92: 3397–3400. monitoring decontamination of Mycoplasma-infected 7 Wimmer MLJ, Sandmann SR, Saikku P, Haberl RL. Chalamydia sp. strains. Appl Environ Microbiol 1996; Association of chlamydial infections with cerebrovas- 62: 328–331. cular disease. Stroke 1996; 67: 2207–2210. 27 Verkooyen RP et al. Widely used commercially avail- 8 Tavendale R et al. Serological markers of Chlamydia able Chlamydia pneumoniae antigen contaminated pneumoniae infection in men and women and sub- with mycoplasma. J Med Microbiol 1997; 46: 419–424. sequent coronary events; the Scottish Heart Health 28 Maass M et al. Endovascular presence of Chlamydia Study Cohort. Eur Heart J 2002; 23: 301–307. pneumoniae DNA is a generalized phenomenon in 9 Hirashima Y et al. Chlamydia pneumoniae infection is atherosclerotic vascular disease. Atherosclerosis 1998; not involved in carotid artery stenosis. Artherosclero- 140 (Suppl 1): S25–S30. sis 2002; 163: 165–168. 29 Del Piano M, Sessa R, Di Pietro M. Chlamydia 10 Johnson W, Moses J, Kipshidze N. Absence of pneumoniae infection and atherosclerotic coronary Chlamydia pneumoniae in surgical specimens of disease: a new possible etiopathogenic association. In: coronary and carotid arteries by polymerase chain Starling A (ed). Proceedings of the Third meeting of reaction. Cardiovasc Radiat Med 2001; 2: 221. the European society for Chlamydia Research. 1995, 11 Altman R, Rouvier J, Scazziota A, Gonzalez C. p 225. No causal association between inflammation and 30 Markus S et al. Chlamydia pneumoniae infection and Chlamydia pneumoniae in patients with chronic early asymptomatic carotid atherosclerosis. Circula- ischemic arterial disease. Inflammation 2002; 26: tion 1999; 100: 832–837. 25–30. 31 Saikku P. Epidemiology of Chlamydia pneumoniae in 12 Danesh J et al. Chlamydia pneumoniae IgA titres and atherosclerosis. Am Heart J 1999; 138 (Suppl 2): coronary heart disease; prospective study and meta- S500–S503. analysis. Eur Heart J 2002; 23: 371–375. 32 Saikku P et al. Chronic Chlamydia pneumoniae 13 Cook PJ et al. Chlamydia pneumoniae antibodies in infection as a risk factor for coronary heart disease in severe essential hypertension. Hypertension 1998; 31: the Helsinki Heart Study. Ann Intern Med 1992; 116: 589–594. 273–277. 14 Nishimura M et al. Inverse association of Chla- 33 Cook PJ, Davies P, Wise R, Honeybourne D. Chlamydia mydia pneumoniae infection with high blood pneumoniae infection and ethnic origin. Ethn Health pressure in Japanese adults. Am J Hypertens 2001; 1998; 3: 237–246. 14: 20–26. 34 Pristipino C et al. Major racial differences in coronary 15 Manning G, Rushton L, Millar-Craig MW. Clinical constrictor response between Japanese and Caucasians implications of white coat hypertension: an ambula- with recent myocardial infarction. Circulation 2000; tory blood pressure monitoring study. J Hum Hyper- 101 (10): 1102–1108. tens 1999; 13: 817–822. 35 Verdecchia P, Porcellati C, Schillaci G. Ambulatory 16 Pose-Reino A et al. Clinical implications of white coat blood pressure. An independent predictor of prognosis hypertension. Blood Press 1996; 5: 264–273. in essiential hypertension. Hypertension 1994; 24: 17 1999 World Health Organization–International Society 793–801. of Hypertension Guidelines for the management of 36 Yamashiroya M, Ghosh L, Yang R, Robertson L. hypertension. J Hypertens 1999; 17: 151–183. Herpesviridae in the coronary arteries and aorta of young 18 Zakopoulos NA et al. Reproducibility of ambulatory trauma victims. Am J Pathol 1988; 130 (1): 71–79. blood pressure measurements in essential hyperten- 37 Benditt PE, Barrett T, McDougall KJ. Viruses in the sion. Blood Press Monit 2001; 6: 41–45. etiology of atherosclerosis. Proc Natl Acad Sci USA 19 Galimberti L. Uso adeguato ed inadeguato degli 1983; 80: 6389–6389. alcoolici in patologie e problemi connessi all’ uso 38 Musiani M et al. Antibody patterns against Cytomega- inadeguato degli alcoolici. Giunta Regionale Venice lovirus and Epstein–Barr virus in human atherosclero- 1982; 10: 73–89. sis. Microbiologica 1990; 13: 35–41.

Journal of Human Hypertension C. pneumoniae and Epstein–Barr infection in hypertension V Ch Pitiriga et al

27 39 Alexander ER, Harrison HR. Chlamydial infections. 42 Vlachakis ND, Mendlowitz M, DGuia D, DGusman D. In: Evans AS, Brachman PS (eds). Bacterial Infections Diminished baroreceptors sensitivity in elderly hyper- of Humans: Epidemiology and Control, 2nd tensives. Possible role of atherosclerosis. Atherosclero- edn. Plenum Medical Book Co: New York, 1991, sis 1976; 24: 243–249. pp 179–206. 43 Sanderson JE, Billingham JD, Floras J. Baroreceptor 40 Leinonen M, Saikku P. Interaction of Chlamydia function in the hypertensive black African. Clin Exp pneumoniae infection with other risk factors of athero- Hypertens A 1983; 5: 339–351. sclerosis. Am Heart J 1999; 138 (Supp 1): S504–S506. 44 Li Z, Mao ZH, Abboud M, Chapleau WM. Oxygen- 41 Chapleau WM et al. Structural versus functional derived free radicals contribute to baroreceptor dys- modulation of the arterial baroreflex. Hypertension function in atherosclerotic rabbits. Circ Res 1996; 79: 1995; 26: 341–347. 802–811.

Journal of Human Hypertension