Relation of high cytomegalovirus antibody titres to blood pressure and brachial artery flow-mediated dilation in young men: the Cardiovascular Risk in Young Finns Study

Authors


Dr A. Haarala, Department of Microbiology and Immunology, University of Tampere, Medical School, Finn-Medi 1, Biokatu 6, FIN-33014 Tampere, Finland. E-mail: atte.haarala@uta.fi

Summary

Human cytomegalovirus (CMV) infection is associated with a higher risk of cardiovascular disease in immunocompromised organ transplant patients. It has been linked with the pathogenesis of elevated arterial blood pressure. However, controversy exists as to whether CMV infection is associated with endothelial function, and little is known about its role as a potential risk factor for early atherosclerosis development at a young age. We aimed to discover if CMV antibody titres are associated with early vascular changes (carotid intima-media thickness, carotid artery distensibility and brachial artery flow-mediated dilation), blood pressure elevation or other traditional cardiovascular risk factors. CMV antibody titres were measured in 1074 women and 857 men (aged 24–39 years) taking part in the Cardiovascular Risk in Young Finns study. CMV antibody titres were significantly higher in women compared to men. In men, high CMV antibody titres were associated directly with age (P < 0·001) and systolic (P = 0·053) and diastolic (P = 0·002) blood pressure elevation, and associated inversely with flow-mediated dilation (P = 0·014). In women, CMV antibody titres did not associate with any of the analysed parameters. In a multivariate regression model, which included traditional atherosclerotic risk factors, CMV antibody titres were independent determinants for systolic (P = 0·029) and diastolic (P = 0·004) blood pressure elevation and flow-mediated dilation (P = 0·014) in men. High CMV antibody titres are associated independently with blood pressure and brachial artery flow-mediated dilation in young men. This association supports the hypothesis that common CMV infection and/or an immune response to CMV may lead to impaired vascular function at a young age.

Introduction

Human cytomegalovirus (CMV) is a member of the herpes virus family. It is a widely spread virus, and up to 40% of people acquire the infection during their first year of life [1]. The prevalence increases progressively with increasing age, and more than 90% of senior citizens are CMV seropositive [2]. Women are known to have higher CMV seroprevalence than men [2]. Additionally, the seroprevalence is higher among people with low socio-economic status (SES) [2]. In most immunocompetent people, the primary CMV infection is mild or even asymptomatic. CMV infection is typically a clinical problem in immunocompromised people, including patients with congenital infections or immunodeficiency syndrome, as well as patients receiving organ transplants [1].

CMV infection is associated with various chronic inflammatory diseases, including cardiovascular disease (CVD), autoimmune diseases and certain cancers [1]. The role of CMV as a risk factor for CVD is controversial. CMV infection has been shown to increase the risk of cardiac allograft vascular disease and graft rejection in immunocompromised organ transplant patients [3]. The use of anti-viral treatment with ganciclovir has been shown to decrease the risk of allograft vascular disease in transplant patients [4,5]. In immunocompetent people, the association between CMV infection and an increased risk of CVD is less clear. Significant associations have been reported between CMV seropositivity and CVD risk [6,7], and also between high CMV antibody titres and increased CVD risk [8,9]. However, negative associations have also been reported [6,10,11]. Few relatively small studies exist that have shown an association between CMV and early atherosclerotic changes; none the less, CMV has been shown to be associated with endothelial function [12,13] and increased carotid intima-media thickness (IMT) [14]. Not all reports are unanimous, however [15,16]. Interestingly, recent publication has shown that essential hypertension is associated with increased CMV DNA copy number and CMV-encoded microRNA expression [17]. Also, CMV infection has been linked with the pathogenesis of increased arterial blood pressure via stimulation of renin and cytokine production and contributes to increased blood pressure values in mice [18].

Recent studies have linked high cytomegalovirus antibody titres with cardiovascular disease and total mortality among older people [19–21]. Therefore, we hypothesized that CMV antibody titres, rather than seropositivity, might be more relevant markers of deleterious effects of CMV infection in young adults. In this study, we sought to identify if intense humoral CMV-specific immunity, measured by CMV antibody titres, is associated with early vascular changes, IMT, carotid artery distensibility (Cdist), brachial artery flow-mediated dilation (FMD), blood pressure values or other traditional CVD risk factors in young healthy adults.

Materials and methods

The study population consisted of participants in the Cardiovascular Risk in Young Finns study, which is an ongoing multi-centre follow-up study in five university hospitals in cities with medical schools in Finland. The study began in 1980, when 3596 participants between the ages of 3 and 18 years were selected randomly from the national population registers. The study design has been presented in more detail elsewhere [22,23]. The 21-year follow-up was conducted in 2001, when the participants were between 24 and 39 years of age. Cardiovascular risk factor measurements, including body mass index (BMI), waist circumference, serum lipids, blood pressure values, levels of C-reactive protein (CRP), SES (occupation and education), alcohol consumption and smoking habits, were recorded during this follow-up. The study complies with the Declaration of Helsinki. The study was approved by local ethics committees and subjects gave informed consent.

Blood pressure measurements were taken using a random zero sphygmomanometer (Hawksley & Sons Ltd, Lancin, UK), and the mean of three measurements was used in the analysis. BMI was calculated from the measured height and weight. Waist circumference and hip circumference were measured, and the waist–hip ratio was calculated. Data regarding occupation, education, smoking habits, alcohol consumption and physical activity were gathered via questionnaires. Measurements of plasma lipids, glucose and insulin were performed by fasting plasma, as described previously [22]. CRP concentrations were analysed with a high-sensitivity latex turbidimetric immunoassay (Wako Chemicals GmbH, Neuss, Germany) with a detection limit of 0·06 mg/l. CMV IgG antibody titres were analysed using a commercial enzyme immunoassay (Enzygnost Anti-CMV/IgG; Siemens Healthcare Diagnostics Products GmbH, Marburg, Germany). According to the manufacturer, the test sensitivity was 99·3% and the specificity was 98·2%. Seropositivity for CMV was defined as a serum anti-CMV immunoglobulin (Ig)G titre of ≥230.

Carotid ultrasound measurements were performed using a Sequoia 512 high-resolution ultrasound system (Acuson, Mountain View, CA, USA). Subjects were instructed to avoid smoking, high-calorie meals, coffee and other caffeine drinks on the day of the ultrasound measurements. Subjects also were instructed to avoid vigorous exercise and alcohol consumption on the previous evening of the measurements. Cdist, which depicts the ability of the large arteries to expand under cardiac pulse pressure, was assessed from the formula Cdist = ([Ds–Dd]/Dd)/(Ps–Pd), where Ds is the systolic diameter, Dd is the diastolic diameter, Ps is the systolic blood pressure and Pd is the diastolic blood pressure [24]. Mean IMT was derived from a minimum of four measurements of the posterior wall of the left carotid artery (at ∼10 mm proximal to the bifurcation) [24]. To assess brachial FMD, the left brachial artery diameter was measured both at rest and during reactive hyperaemia. Increased flow was induced by inflation of a pneumatic tourniquet placed around the forearm to a pressure of 250 mmHg for 4·5 min, followed by release. Three measurements of arterial diameter were performed at end-diastole at a fixed distance from an anatomic marker at rest and at 40, 60 and 80 s after cuff release. The vessel diameter in scans after reactive hyperaemia was expressed as the percentage relative to the resting scan. The average of three measurements at each time-point was used to derive the maximum FMD (the greatest value between 40 and 80 s) [25].

CMV antibody titres were measured successfully in 2133 subjects. We found that 648 subjects were seronegative (<230 antibody titres) and that 1485 subjects were seropositive (≥230). Subjects with missing data on smoking, education, BMI, waist circumference, waist–hip ratio, blood pressure, insulin, alcohol consumption, IMT and/or Cdist values were not included in the analysis. Therefore, the total number of subjects included in the study was 1931 subjects (1074 women and 857 men). The subjects were then divided into four quartiles according to CMV antibody titres. CMV antibody titres were parameterized with a dummy variable comparing the highest quartile (Q4), with the bottom three quartiles combined (Q1–Q3). A t-test was used for normally distributed variables and the Mann–Whitney U-test was used for skewed variables (triglycerides, insulin and CRP). The χ2 test was used for categorized variables. A multivariable linear regression model was used to assess CMV significance as an independent determinant for blood pressure values and FMD. Log10-transformed values were used for skewed variables. All analyses were also repeated in a subcohort consisting of only CMV seropositive subjects.

Results

In our study cohort, there were significantly (P = 0·003) more seropositive women (72·3%) than seropositive men (65·9%). Additionally, there were significantly (P < 0·001) more women with CMV antibody levels over 14 000 titres compared to men (72·5% versus 27·5%). Because of the strong sex difference in our study sample, all analyses were performed separately in men and women, including the quartile division.

Characteristics of the study subjects according to sex and CMV antibody titre quartiles are presented in Table 1. In men, high CMV antibody titres were associated significantly with diastolic blood pressure (P = 0·002) and were associated borderline significantly with systolic blood pressure (P = 0·053) compared to all other men. In men, age was also associated with the highest CMV antibody titre quartile (P < 0·001). There were no significant associations between other traditional CVD risk factors (anthropometric, lipid, metabolic, inflammatory or lifestyle factors) and high CMV antibody titres. In men, subjects with high CMV antibody titres had significantly lower FMD values (P = 0·013). IMT and Cdist values did not differ significantly between CMV antibody titres. In women, there were no significant associations between CMV antibody titres and early markers of atherosclerosis. However, there was a trend between high CMV antibody titre and higher FMD values (P = 0·070).

Table 1(a).  Baseline characteristics of women according to cytomegalovirus immunoglobulin (Ig)G antibody titrers.
CharacteristicsCytomegalovirus IgG antibody titre quartiles*P-value
1234
<230 titre (n = 297)290–9 300 titre (n = 234)9 400–15 000 titre (n = 289)16 000–46 000 titre (n = 254)
  1. *Mean values and standard deviation, t-test for difference between highest CMV antibody titrers compared to other groups. Median values and interquartile range (IQR), Mann–Whitney U-test for difference between highest CMV antibody titres compared to other groups. Percentages and χ2 test for difference between highest CMV antibody titres compared to other groups. §Data missing for 136 participants. ||Data missing for 231 participants. #Data missing for 62 participants. HDL: high-density lipoprotein; LDL: low-density lipoprotein; IMT: carotid intima-media thickness; Cdist: carotid artery distensibility; FMD: flow-mediated dilation.

Age (years)*31·2 (4·9)30·8 (5·0)32·4 (5·0)32·1 (4·7)0·120
Smokers, % daily16·220·320·421·30·413
Occupational status, % manual§21·025·921·725·00·424
Education, % comprehensive school6·77·810·48·30·990
Body mass index (kg/m2)*23·9 (4·1)24·0 (4·4)24·4 (4·3)24·5 (4·6)0·200
Waist circumference (cm)*77·7 (10·3)77·8 (10·6)79·9 (11·2)79·6 (11·7)0·166
Waist–hip ratio*0·79 (0·06)0·78 (0·06)0·80 (0·06)0·80 (0·06)0·145
Systolic blood pressure (mmHg)*115·3 (12·2)116·4 (12·2)116·9 (11·6)115·2 (12·8)0·288
Diastolic blood pressure (mmHg)*71·3 (8·2)71·4 (9·1)72·4 (8·1)71·2 (8·9)0·360
Total cholesterol (mmol/l)*5·00 (0·84)5·03 (0·88)5·07 (0·91)5·05 (0·96)0·780
HDL cholesterol (mmol/l)*1·40 (0·29)1·44 (0·29)1·38 (0·30)1·38 (0·30)0·302
LDL cholesterol (mmol/l)*3·09 (0·73)3·08 (0·75)3·18 (0·75)3·16 (0·82)0·453
Triglycerides (mmol/l)1·00 (0·80–1·30)1·00 (0·80–1·30)1·00 (0·75–1·40)1·00 (0·78–1·30)0·591
Glucose (mmol/l)*4·8 (0·5)4·8 (0·6)4·9 (0·8)4·8 (1·1)0·453
Insulin (mU/l)6·0 (5·0–8·0)6·0 (4·0–8·0)7·0 (5·0–9·0)6·0 (4·0–9·0)0·766
C-reactive protein (mg/l)0·77 (0·30–1·90)1·00 (0·38–2·50)0·84 (0·33–2·22)0·80 (0·34–2·29)0·941
Physical activity index*||17·7 (15·9)16·2 (13·6)15·8 (15·6)16·4 (14·5)0·886
Alcohol (no. drinks per week)*3·4 (4·3)4·2 (5·5)3·7 (4·9)3·9 (7·9)0·763
IMT (mm)*0·56 (0·08)0·57 (0·09)0·58 (0·09)0·58 (0·08)0·221
Cdist (%/10 mmHg)*2·38 (0·81)2·32 (0·71)2·30 (0·76)2·33 (0·76)0·996
FMD (%)*#8·49 (4·55)8·73 (4·43)8·72 (4·60)9·24 (4·52)0·070
Table 1(b).  Baseline characteristics of men according to cytomegalovirus (CMV) immunoglobulin (Ig)G antibody titres.
CharacteristicsCytomegalovirus IgG antibody titre quartilesP-value
1234
<230 titre (n = 292)240–5 400 titre (n = 139)5500–10 000 titre (n = 206)11 000–35 000 titre (n = 220)
  • *

    Mean values and standard deviation, t-test for difference between highest CMV antibody titrers compared to other groups. Median values and interquartile range (IQR), Mann–Whitney U-test for difference between highest CMV antibody titres compared to other groups. Percentages and χ2 test for difference between highest CMV antibody titres compared to other groups. **Data missing for 126 participants. ††Data missing for 91 participants. ‡‡Data missing for 79 participants. HDL: high-density lipoprotein; LDL: low-density lipoprotein; IMT: carotid intima-media thickness; Cdist: carotid artery distensibility; FMD: flow-mediated dilation.

Age (years)*30·8 (4·8)31·6 (4·9)31·5 (5·0)33·0 (5·1)<0·001
Smokers, % daily23·439·630·628·20·796
Occupational status, % manual**42·543·345·843·50·934
Education, % comprehensive school7·211·55·35·90·451
Body mass index (kg/m2)*25·8 (4·1)25·5 (3·8)25·3 (3·6)26·1 (4·0)0·126
Waist circumference (cm)*89·9 (11·4)89·6 (10·5)88·7 (10·0)90·2 (10·7)0·374
Waist–hip ratio*0·89 (0·06)0·90 (0·07)0·89 (0·06)0·90 (0·06)0·347
Systolic blood pressure (mmHg)*129·2 (13·2)127·5 (12·1)128·1 (11·9)130·8 (16·4)0·053
Diastolic blood pressure (mmHg)*74·5 (8·8)73·9 (8·1)74·0 (7·8)76·8 (10·8)0·002
Total cholesterol (mmol/l)*5·12 (0·99)5·27 (1·11)5·26 (0·96)5·25 (1·00)0·536
HDL cholesterol (mmol/l)*1·16 (0·27)1·17 (0·28)1·17 (0·28)1·15 (0·27)0·282
LDL cholesterol (mmol/l)*3·34 (0·90)3·45 (0·97)3·45 (0·88)3·45 (0·88)0·478
Triglycerides (mmol/l)1·20 (0·90–1·70)1·30 (0·90–1·90)1·20 (0·90–1·80)1·30 (0·90–1·80)0·460
Glucose (mmol/l)*5·2 (0·5)5·3 (1·6)5·2 (0·5)5·2 (0·5)0·850
Insulin (mU/l)6·0 (4·0–9·0)6·0 (4·0–9·0)6·0 (4·0–9·0)7·0 (4·3–9·0)0·329
C-reactive protein (mg/l)0·56 (0·29–1·25)0·70 (0·26–1·51)0·52 (0·27–1·33)0·62 (0·30–1·43)0·236
Physical activity index*,††18·3 (18·3)13·1 (15·0)15·3 (16·9)15·1 (16·7)0·436
Alcohol (no. drinks per week)*8·2 (9·3)10·6 (12·9)8·5 (8·3)7·6 (10·0)0·123
IMT (mm)*0·59 (0·10)0·59 (0·09)0·60 (0·10)0·59 (0·11)0·811
Cdist (%/10 mmHg)*2·04 (0·67)1·99 (0·64)2·03 (0·65)1·94 (0·66)0·086
FMD (%)*‡‡7·17 (3·96)6·96 (4·03)7·23 (4·42)6·36 (3·69)0·013

A multivariable regression model was used to analyse whether CMV antibody titres are independent determinants for blood pressure and FMD. In men, high CMV antibody titres were associated independently with systolic blood pressure (P = 0·029) and diastolic blood pressure (P = 0·004) after adjusting for age, BMI, HDL cholesterol, triglycerides, insulin, CRP, smoking, occupational status and alcohol consumption (Table 2). In men, high CMV antibody titres were also associated independently and inversely with FMD (P = 0·014) after adjusting for age, baseline brachial diameter, BMI, systolic blood pressure, HDL cholesterol, triglycerides, insulin, CRP, smoking, occupational status and alcohol consumption (Table 3). In women, high CMV antibody titres were not associated independently with FMD (P = 0·112).

Table 2.  Determinants of systolic and diastolic blood pressure in a multivariate linear regression model in men (n = 730).
Blood pressureRisk variableβ ± s.e.P-value
  1. High CMV antibody (titre); CMV antibody titres were parameterized with a dummy variable comparing the highest quartile with the bottom three quartiles combined. HDL: high-density lipoprotein; CMV: cytomegalovirus.

SystolicAge (years)0·026 ± 0·1010·800
Body mass index (kg/m2)0·868 ± 0·161<0·001
HDL cholesterol (mmol/l)3·215 ± 2·0270·113
Triglycerides (mmol/l)7·840 ± 2·9550·008
Insulin (mU/l)6·898 ± 2·5530·007
C-reactive protein (mg/l)1·452 ± 1·0920·184
Smoking (daily)−3·147 ± 1·0980·004
Non-manual occupation−2·438 ± 0·9910·014
Alcohol (no. drinks per week)0·087 ± 0·0490·075
High CMV antibody (titre)2·367 ± 1·0840·029
 R2 = 0·181 
DiastolicAge (years)0·385 ± 0·063<0·001
Body mass index (kg/m2)0·431 ± 0·099<0·001
HDL cholesterol (mmol/l)3·895 ± 1·2520·002
Triglycerides (mmol/l)8·742 ± 1·825<0·001
Insulin (mU/l)5·387 ± 1·5770·001
C-reactive protein (mg/l)0·928 ± 0·6740·169
Smoking (daily)−2·114 ± 0·6780·002
Non-manual occupation−2·151 ± 0·612<0·001
Alcohol (no. drinks per week)0·095 ± 0·0300·002
High CMV antibody (titre)1·927 ± 0·6690·004
 R2 = 0·275 
Table 3.  Determinants of flow-mediated dilation in a multivariate linear regression model in men (n = 657).
Risk variableβ ± s.e.P-value
  1. High CMV antibody (titre); CMV antibody titres were parameterized with a dummy variable comparing the highest quartile with the bottom three quartiles combined. HDL: high-density lipoprotein; s.e.: standard error.

Age (years)0·004 ± 0·0320·929
Baseline brachial diameter (mm)−2·828 ± 0·361<0·001
Body mass index (kg/m2)0·236 ± 0·054<0·001
Systolic blood pressure (mmHg)−0·025 ± 0·0130·050
HDL cholesterol (mmol/l)0·355 ± 0·6410·580
Triglycerides (mmol/l)−0·098 ± 0·9450·917
Insulin (mU/l)−0·209 ± 0·8190·799
C-reactive protein (mg/l)−0·093 ± 0·3460·787
Smoking (daily)−0·273 ± 0·3510·436
Non-manual occupation−0·386 ± 0·3130·219
Alcohol (no. drinks per week)−0·019 ± 0·0160·231
High CMV antibody (titre) −0·844 ± 0·3430·014
 R2 = 0·118 

We also carried out a subanalysis in CMV seropositive subjects. In men, high CMV antibody titres were associated with age (P = 0·001), systolic blood pressure (P = 0·020), diastolic blood pressure (P = 0·001), alcohol consumption (P = 0·050) and FMD (P = 0·033). High CMV antibody titres remained a significant determinant for systolic blood pressure (P = 0·021), diastolic blood pressure (P = 0·005) and FMD (P = 0·022) in the multivariable model. In seropositive women, no significant associations were found between CMV antibody titres and the risk factors or markers of early atherosclerosis.

Discussion

In this study, we showed that high CMV antibody titres are associated independently with blood pressure values and associated inversely with FMD in young men. To our knowledge, there are no prior studies demonstrating an association between CMV antibody titres and blood pressure in humans. Recently, in a Chinese cohort, Li et al. have shown that plasma CMV DNA copy number is associated with hypertension [17]. Additionally, they showed that CMV-encoded microRNA, hcmv-miR-UL112, was highly expressed in hypertensive patients. Further they showed that hcmv-miR-UL112 could target interferon regulatory factor 1, which is related to up-regulation of angiotensin II type 2 receptor [26]. This pathway is one plausible pathological mechanism between CMV infection and increased blood pressure. An association between CMV infection and increased arterial pressure has also been shown in mice [18]. Cheng et al. showed that CMV infection induced renin expression in a dose-dependent manner in mouse and human cells and that increased angiotensin-II, interleukin (IL)-6, tumour necrosis factor (TNF)-α and monocyte chemotactic protein-1 (MCP-1) levels were identified in mouse serum. Both IL-6 and TNF-α levels have been shown to correlate with increased blood pressure values [27,28]. The association shown in this study between high CMV antibody titres and with blood pressure values in young men supports the possible relationship between CMV infection and blood pressure. Because renin, angiotensin-II, IL-6, TNF-α and MCP-1 levels were not measured in this study, we were not able to evaluate whether this association depends on these factors. However, CRP concentrations, which are considered to be indicators of systemic low-grade inflammation, were available from the present cohort. We found that CRP concentrations did not differ between CMV antibody titre quartiles and that the association between CMV antibody titres and blood pressure values was not attenuated after adjustment with CRP. However, it is possible that CMV may mediate other inflammatory pathways than CRP. Based on the works of Li et al. and Cheng et al., it can be speculated that the activation of the renin–angiotensin system (RAS) is possibly an underlying mediator behind these findings. It is possible that high CMV antibody titres are indicators of frequent reactivation of CMV or reinfection with new strains of CMV, leading to stronger immunity in these individuals. CMV activity may lead to an increased RAS activation, leading to arterial constriction via the influence of angiotensin-II, a mechanism that possibly explains the observed increased blood pressure values in this study.

The association between CMV infection and endothelial function has been reported previously in relatively small study populations, but the findings have not been consistent. In a population consisting of paediatric heart transplant patients (n = 50), Simmond et al. demonstrated that decreased FMD is associated with CMV replication after transplantation [12]. In a middle-aged population (mean age 38 years), which consisted of diabetic and non-diabetic subjects (n = 157), Grahame-Clarke et al. showed that CMV seropositivity was associated with impaired vascular function measured by venous occlusion plethysmography with bradykinin and glyceryl trinitrate [13]. Additionally, CMV infection has been shown to cause arterial dysfunction in a mouse model [29]. However, negative findings have also been reported in small cohorts, which consisted of young Japanese men (n = 81) [16] and young Canadian men (n = 65) [15]. In our study we demonstrated for the first time, in a large cohort of young men (n = 657), that high CMV antibody titres were associated independently with endothelial function. In vitro, CMV infection has been shown to activate the P38-mitogen-activated protein kinase (MAPK) signalling pathway and up-regulate phosphatase and tensin homologue [30,31]. Via this pathway, endothelial nitric oxide synthase is inhibited and reduced nitric oxide (NO) production leads to endothelial dysfunction [32,33]. The association between CMV and endothelial dysfunction has been hypothesized further to be one possible pathological mechanism of hypertension [33]. This mechanism is a possible explanation for the demonstrated association between CMV and reduced FMD in our study, and it is also a possible alternative explanation for the demonstrated association between CMV and increased blood pressure. In this study, high CMV antibody titres were not associated with IMT or Cdist. Previously, in a cohort consisting of middle-aged subjects, high CMV antibody titres were shown to be a risk factor for increased IMT [14]. One possible explanation for this discrepancy between the results could be the age difference between the cohorts. Possible deleterious effects of CMV immune response on IMT may not be currently detectable in young adult ages.

In this study, CMV antibody titres were significantly higher among women, an observation that has also been seen in other populations [2]. Interestingly, high CMV antibody titres were not associated with decreased FMD values or with increased blood pressure values in women. Thus, it might be possible that CMV infection leads to dissimilar immune responses and subclinical manifestations in women and men. Zhu et al. found that the association of an immune response to CMV and the risk of coronary artery disease differed between the sexes [34]. In their study, coronary artery disease risk was increased among women who had humoral antibodies against CMV but not in those women who had only a cell-mediated immune response to CMV. In men, there was no difference between different immune responses and coronary artery disease risk. Our work also supports the hypothesis that immune responses to CMV may differ between the sexes. Zhu et al. proposed that the differences in the findings between the sexes may be due to the association found between CMV and CRP in men, but not in women. In our study, CMV antibody titres were not associated with CRP. All in all, we did not find any significant implications for high CMV antibody titres in young women. Based on our current knowledge, we do not have an explanation for the observed sex-related differences in our results.

The role of CMV as a risk factor for CVD is controversial. A great deal of previous studies investigated the role of CMV seropositivity as a risk factor for CVD indices [6,7,10]. Few studies have investigated the role of CMV antibody titres [8,9,11]. Regardless of the approach, both negative and positive associations have been reported. Numerous mechanisms have been reported on how CMV infection may lead to the development and exacerbation of atherosclerosis. CMV has been shown to infect human endothelial cells [35] and leucocytes [36]. Infection of these cells leads to neutrophil transendothelial migration [37], smooth muscle cell migration [38], intracellular adhesion molecule expression and leucocyte adhesion [39]. Conversely, it has been demonstrated that CMV-independent TNF-α production can induce CMV reactivation [40]. Thus, the alternative explanation could also be the reverse causality. High CMV antibody titres might be a result of immunoresponses stimulated by pathological changes independent of CMV.

In this study, the subjects with high CMV antibody titres were compared to those with moderate, low or seronegative CMV antibody titres, whereas the traditional set-up has been to compare seropositive and seronegative subjects. The latter set-up may be problematic due to the high prevalence of the virus in humans. More than half of humans have serological evidence of CMV infection and it is also possible that there are some seronegative cases that have had prior CMV infection without leaving detectable serological evidence. This methodology raises the question of how an almost ubiquitous infection could be a sole risk factor. Along these lines, we used a set-up where subjects with high CMV antibody titres were compared to other subjects. This strong antibody-specific immunity against CMV could be the result of various factors: severe primary infection, recent infection, reinfections of CMV with a different strain, frequent reactivations of the virus in the body or non-viral-related individual differences in immunity. The advantage of this set-up is that these subjects certainly have a strong CMV-specific humoral immune response, but conversely, we do not know what mechanism may have caused it. Further studies are needed to differentiate the aetiology of strong CMV-specific immune response. Because our study cohort was ethnically homogeneous, the generalizability of our results is limited to white European subjects. It is also important to remember that observational studies cannot establish causality.

In conclusion, this study showed that high CMV antibody titres associate directly with blood pressure and inversely with FMD in young men. These associations were not found in women. Our results support the idea that this common virus could be a risk factor for unfavourable changes in the cardiovascular system at an early age.

Acknowledgements

The authors wish to thank Sinikka Repo-Koskinen and Maritta Virtanen for their skilful technical assistance. The expert technical assistance in the statistical analyses by Irina Lisinen and Ville Aalto are gratefully acknowledged. The Young Finns Study has been financially supported by the Academy of Finland (grant nos 117797, 117941, 126925, 121584, 124282), the Social Insurance Institution of Finland, the Turku University Foundation, the Finnish Cultural Foundation, the Yrjö Jahnsson Foundation, the Emil Aaltonen Foundation (T.L.), the Medical Research Fund of Tampere University Hospital, Turku University Hospital Medical Fund, Kuopio University Hospital Medical Fund, the Juho Vainio Foundation, the Finnish Foundation for Cardiovascular Research and the Tampere Tuberculosis Foundation.

Disclosure

The authors declare that there is no conflict of interest.

Ancillary