Current Infection With Helicobacter pylori, but Not Seropositivity to Chlamydia pneumoniae or Cytomegalovirus, Is Associated With an Atherogenic, Modified Lipid Profile
Abstract—Infectious agents may be involved in atherothrombogenesis. The potential pathogenic pathway, however, remains unclear. We investigated the association between various infectious agents and lipoproteins known to have an atherogenic effect. We recruited 470 healthy blood donors and 238 patients with angiographically proven coronary heart disease (CHD), aged 40 to 68 years. Seropositivity to Chlamydia pneumoniae (CP), chlamydial lipopolysaccharide, and cytomegalovirus (CMV) was determined; infection with Helicobacter pylori (HP) was assessed by using the [13C]urea breath test. In all subjects, total cholesterol, high density lipoprotein (HDL) cholesterol, lipoprotein(a), and various apolipoproteins (apos) were determined. In unadjusted analysis, mean HDL cholesterol concentration was significantly decreased in HP-positive healthy subjects (1.36 vs 1.44 mmol/L, P=0.006) compared with HP-negative subjects. The HDL cholesterol to total cholesterol ratio was significantly decreased in HP-positive (0.259 vs 0.276, P=0.01) and CP-seropositive (0.266 vs 0.280, P=0.04) healthy subjects compared with (sero)negatives. Mean apoAI levels were significantly lower in HP-positive healthy subjects (1.46 vs 1.51 g/L, P=0.03) and in CMV-positive healthy subjects (1.47 vs 1.52 g/L, P=0.01) compared with (sero)negative subjects. After multivariable adjustment by means of linear regression analysis, only the association between HP infection and decreased HDL cholesterol (P=0.002), decreased HDL cholesterol to total cholesterol ratio (P=0.005), decreased apoAI (P=0.02), and increased apoB (P=0.02) persisted and remained significant. There was no independent association between other lipoproteins and serological markers of CP or CMV infection. Current infection with HP, but not seropositivity to CP or CMV, was associated with an atherogenic, modified lipid profile. These lipid alterations could explain, at least in part, the reported weak association between chronic HP infection and atherosclerotic diseases.
This study was supported by grants from the Medical Faculty of the University of Ulm, Ulm; Medac Diagnostica, Wedel; and Astra Chemicals, Wedel, Germany.
- Received July 4, 2000.
- Accepted August 22, 2000.
The role of infection in the pathogenesis of atherosclerosis is still a matter of debate. The underlying processes that might explain the potential association between infectious agents and atherosclerotic disease also remain unclear. Various potential pathomechanisms have been postulated, eg, increased production of cytokines and acute-phase reactants, local or systemic disturbance of fibrinolysis and blood coagulation, direct infection of the arterial wall via macrophages and alteration of vascular cell function, and an immunological response (cross-reaction) due to bacterial heat shock protein.1 2 3 Because it is well known that acute infections are able to modify serum lipids, some authors have also suggested that alterations of the lipid metabolism due to chronic infections could represent an atherogenic link.4 5 6 7 8 9 10
An adversely altered lipid profile, ie, increase of total and LDL cholesterol, a decrease of HDL cholesterol, or an increase in lipoprotein(a) [Lp(a)], is known as a major risk factor for cardiovascular disease. Elevated levels of total cholesterol and triglycerides and decreased HDL cholesterol concentration were reported for subjects with seropositivity to Chlamydia pneumoniae (CP)4 5 6 or Helicobacter pylori (HP),7 8 and elevated levels of Lp(a) were found in subjects with seropositivity to cytomegalovirus (CMV)9 or, questionably, to CP.10
To examine this hypothesis, we measured a variety of lipoproteins and assessed their association with markers of CP, HP, and CMV infection in a large sample of apparently healthy subjects as well as in patients with coronary heart disease (CHD). Multivariate analysis was performed to adjust for various potential confounders.
The study population comprised healthy subjects and patients with angiographically documented CHD, 40 to 68 years of age, who were enrolled for this study between October 1996 and November 1997. The response rate among patients was 78% and among healthy subjects, 84%. Patients with CHD were recruited at the University Hospital of Ulm and showed at least 1 coronary stenosis >50% at catheterization within the last 6 months before entry into the study. Patients with acute coronary syndromes during the last 4 weeks before the study were excluded. Healthy subjects were voluntary blood donors from the associated Red Cross blood bank, who were without manifested heart disease or a history of angina pectoris according to the Rose angina questionnaire. Exclusion criteria for all study participants were use of lipid-lowering drugs, acute infection, acute state of a chronic infectious or inflammatory disease, severe liver or renal disease, neoplasm, and hematological disorders.
All participants underwent blood sampling and a [13C]urea breath test and were interviewed by a standardized questionnaire administered by the same trained team of interviewers. Informed consent was obtained from all participants. The study was approved by the ethics committee of the University of Ulm.
Blood was taken under standardized conditions, centrifuged at 3000g for 10 minutes, and immediately divided into aliquots. Plasma and serum specimens were frozen and stored at −70°C until analysis. All laboratory determinations were done in a blinded fashion.
Antibodies (IgG, IgA, and IgM) against Chlamydia species (chlamydial lipopolysaccharide, cLPS) were measured by a recombinant antigen-based ELISA (Medac). The antigen contained an epitope shared by all Chlamydia species.11 An IgG titer ≥1:100 (IgA ≥1:50) was defined as seropositive according to the manufacturer’s instructions. Determinations of specific antibodies (IgG and IgA) against CP were performed by a microimmunofluorescence method as described elsewhere.12 An IgG titer ≥1:64 (IgA ≥1:32) was considered to be seropositive. For analysis, we compared subjects with both elevated IgG and IgA antibodies versus all others.
For measurement of antibodies (IgG, IgA, and IgM) against CMV, we used an established ELISA (Medac). The respective test-specific quality controls were considered. The cutoff was defined as the mean optical density of negative control, +0.14 for IgG and IgM and +0.09 for IgA. The “gray zone” was defined as the cutoff ±10%. Current HP infection was determined in all study participants by means of a modified [13C]urea breath test as described elsewhere.13
Total cholesterol was measured enzymatically with reagents from Wako Chemicals GmbH. HDL cholesterol was measured in the supernatant after precipitation of apoB-containing lipoproteins with phosphotungstate acid and MgCl2 obtained from Roche Diagnostics. ApoAI, B, CIII, and E and Lp(a) were determined by immunoturbidimetry with antisera from Greiner Biochemicals. Apo AII and CII were measured by immunoturbidimetry with antisera from Roche Diagnostics and Kamiya Biomedical Company, respectively. The apolipoprotein assays were calibrated with the following reference sera: N apolipoprotein standard (Behring) for apo AI, AII, B, and E; reference standard from Immuno and multicalibrator set from Kamiya for apoCII and CIII. All analyses were performed on a Wako R-30 automated analyzer in a laboratory setting certified according to ISO 9001.
Baseline characteristics of the study participants are shown in a descriptive way. Lipoprotein and apolipoprotein concentrations are reported as age- and sex-adjusted arithmetic means [geometric mean for Lp(a) values because of their skewed distribution]. To evaluate differences between (sero)positives and (sero)negatives, we calculated age- and sex-adjusted probability values for differences in means of the respective markers. Multivariate linear regression was performed to assess the independent association between markers of infection and lipoproteins; age, sex, pack-years of smoking, alcohol consumption, physical activity, case-control status, and years of school education were controlled for. Data were analyzed by using the SAS statistical software package (version 6.12).
Characteristics of the Study Population
In this study, 708 subjects (238 patients with CHD and 470 voluntary blood donors) were enrolled. Table 1⇓ shows demographic characteristics for both patients with CHD and apparently healthy blood donors, classified according to those with and without signs of infection. These descriptive data show a slightly elevated prevalence of CHD and associated use of cardiac drugs in those subjects with signs of any infection.
HP and Lipoproteins
As shown in Table 2⇓, mean HDL cholesterol and apoAI concentrations, as well as the HDL cholesterol to total cholesterol ratio, were significantly lower in healthy subjects with current HP infection compared with those without. A similar trend was seen in patients with CHD, although the differences were not statistically significant in this group. Other lipoprotein concentrations were not appreciably different between subjects with and without HP infection.
CP and Lipoproteins
The HDL cholesterol to total cholesterol ratio was significantly decreased in healthy subjects with combined positivity to IgA and IgG against CP, compared with all others (negatives and positives to either IgA or IgG), but not in patients with CHD (Table 3⇓). Other lipoprotein concentrations were similar in seropositive and seronegative subjects. This was true even when high antibody titers were considered as the cutoff, eg, HDL cholesterol 1.17 mmol/L (IgG <64), 1.18 mmol/L (IgG 64 to 511), and 1.20 mmol/L (IgG ≥512), or 1.19 mmol/L (IgA <32), 1.18 mmol/L (IgA 32 to 127), and 1.18 mmol/L (IgA ≥128). Additional analysis of IgA and IgG antibodies (in combination and considering only high antibody titers, ie, IgG >400 and IgA >200) against cLPS also did not show any statistically significantly differences for lipoprotein concentrations between seropositive and seronegative subjects (data not shown).
CMV and Lipoproteins
ApoAI and apoAII levels were slightly decreased in CMV-positive healthy subjects but not in CMV-positive patients with CHD, compared with CMV-negative subjects (Table 4⇓). Other lipoprotein concentrations were similar in subjects with and without seropositivity to CMV.
Multivariate Linear Regression Analysis
Multivariate analysis was performed in the present sample of patients with CHD and healthy subjects (Table 5⇓). In fully adjusted analyses, only subjects with current HP infection showed a statistically significantly altered lipid profile (decreased HDL cholesterol, HDL cholesterol to total cholesterol ratio, and apoAI concentration and an increased apoB concentration). No independent association was found between seropositivity to CP, cLPS, CMV, and lipoproteins, except for apoE concentration (significantly increased in CMV-positives). These results remained unchanged after exclusion of those individuals who were taking potentially lipid-modifying drugs, like β-blockers, diuretics, or angiotensin-converting enzyme inhibitors (data not shown).
The present study shows a statistically significantly modified atherogenic lipid profile in patients with CHD and apparently healthy subjects with current HP infection, which persisted after control for potential confounders. No such pattern was found in CP- or CMV-positive subjects. This alteration of the lipid profile may explain to some extent the reported weak association between atherosclerosis and chronic HP infection.
Experimental and clinical studies have shown that the acute-phase response induced by acute infections may modify lipid metabolism,14 15 16 17 18 19 resulting in decreased HDL cholesterol, total cholesterol, and apoAI concentrations and an increase in Lp(a) and triglycerides. Furthermore, experimental investigations demonstrated that tumor necrosis factor-α and other proinflammatory cytokines (interleukin-1, interleukin-6, and interferon-α) are able to affect different key sites in lipid metabolism20 : They decrease the activation of adipose tissue lipoprotein lipase, stimulate hepatic fatty acid synthesis, and influence lipolysis. These separate processes can alter the lipid profile in an atherogenic way. Therefore, it was hypothesized that chronic infections due to various bacterial and viral agents may lead to an atherogenic modification of lipid metabolism.
HP and Lipoproteins
In 1996, Niemelä et al7 reported significantly elevated triglyceride levels in 62 HP-positive subjects compared with 54 HP-negative healthy controls (1.50 vs 1.17 mmol/L). The authors also found a trend toward decreased HDL cholesterol concentrations in these 62 HP-positive healthy controls as well as in 74 HP-positive patients with CHD. Laurila et al8 found significantly increased triglyceride (1.17 vs 1.00 mmol/L) and total cholesterol (6.34 vs 5.87 mmol/L) levels in 460 HP-positive subjects compared with 269 HP-negative nonsmoking reindeer herders after adjustment for age, body mass index, and social class, but HDL cholesterol concentrations were similar in both groups. In a meta-analysis of risk factors for CHD and chronic infection with HP, Danesh and Peto21 found slightly but significantly decreased HDL cholesterol concentrations (−0.032 mmol/L, SEM 0.008) in seropositive subjects compared with seronegative controls.
The present study is the first to use a [13C]urea breath test to assess HP infection in this context. This test is more suitable for detecting current gastric HP infection than are serological tests.13 Because of interactions between moderate alcohol consumption and lipid metabolism, as well as between alcohol consumption and the prevalence of HP infection, we also controlled for this variable by means of linear regression analysis, and similar to the above-mentioned studies, we also found an atherogenic lipid profile: a decrease of HDL cholesterol, the HDL cholesterol to total cholesterol ratio, and apoAI concentration and an increase of apoB in HP-positive subjects.
CP and Lipoproteins
Dahlén et al10 observed markedly elevated Lp(a) concentrations in CP-positive men and suggested an autoimmune process to explain this finding. Subsequent publications22 23 from this group showed controversial results: In a nested case-control study from northern Sweden, they did not find an interactive effect between high Lp(a) levels and CP IgG titers, nor a predictive value of these markers for future ischemic cerebral infarctions.22 However, elevated levels of Lp(a) combined with IgG antibodies against CP in plasma and in circulating immune complexes were able to predict myocardial infarction in the same study.23 In our study population, geometric means of Lp(a) were not elevated in either CP- or cLPS-positive individuals.
Laurila et al4 reported significantly elevated triglyceride levels and an adversely modified HDL cholesterol to total cholesterol ratio (0.20 vs 0.22) in 83 reindeer herders with persistent CP infection compared with 62 controls without seropositivity to CP measured twice over 3 years. Analyses from the same study5 (reindeer herders health survey 1986 to 1989, Finland) confirmed these results in nonsmokers (n=506) but not in smokers (n=542). In 199 male participants of the World Health Organization–MONICA (MONItering trends and determinants in CArdiovascular disease) population survey in northern Ireland, Murray et al6 found significantly elevated total cholesterol (+0.05 mmol/L) and decreased HDL cholesterol (−0.11 mmol/L) concentrations in CP-positives compared with CP-negatives after multivariable adjustment.
In the present study, we found a significantly decreased HDL cholesterol to total cholesterol ratio in healthy subjects with combined seropositivity (IgG and IgA) against CP but not in patients with CHD, and after multivariable adjustment, this association diminished and became statistically nonsignificant. Thus, our data do not confirm findings of prior studies.4 5 6 This remained true even when high antibody titers were considered as cutoff points. Furthermore, the present study is the first in which IgG and IgA antibodies against cLPS were also determined in this context. This test is not specific for CP but is more objective than the microimmunofluorescence method. In particular, the LPS structure of Gram-negative bacteria has been claimed to be a potent endotoxin and to induce production of cytokines,2 with consecutive atherogenic lipid alterations.24 However, we did not find a significant association between antibodies against cLPS and any of the lipoproteins.
CMV and Lipoproteins
In the Atherosclerosis Risk In Communities study,9 mean Lp(a) levels were not different between 539 CMV-positive and 141 CMV-negative subjects; however, the proportion of markedly elevated Lp(a) levels (≥148 μg/mL) was higher in seropositives. Furthermore, a second analysis of prospective data by the same investigators showed an increased risk for carotid atherosclerosis, in particular for CMV-positive subjects with elevated Lp(a) concentrations. The authors concluded that a CMV-induced increase in Lp(a) might result in a hypercoagulable state.
To our knowledge, our study is the first in which a variety of lipid parameters were examined and related to CMV infection. With the exception of significantly elevated apoE concentrations in CMV-positive individuals, none of these [including Lp(a)] were associated with CMV serostatus after adjustment for various potential confounders.
Current infection with HP is associated with an atherogenic lipid profile. This finding might be explained by the fact that HP is known to cause a chronic gastric infection, which may influence lipid metabolism similar to the lipid alterations during any acute infection. However, we found no evidence that seropositivity to CP or CMV is associated with an atherogenic modification of lipoproteins in this study population, as suggested by others.4 5 6 Although we analyzed subjects with combined seroprevalence of IgG and IgA antibodies (against CP and cLPS), it is not yet known whether CP or CMV is able to cause chronic infection in immunocompetent individuals. Our results do not exclude the possibility of short-term lipid alterations during an acute infection, reinfection, or reactivation with CP or CMV; however, such modification of lipid metabolism might not explain an increased cardiovascular risk. The true role of these infectious agents in atherosclerotic diseases is still unclear: CMV and CP, but not HP, have been repeatedly detected in atheromatous lesions. Therefore, indirect atherogenic effects due to an infection with HP, like the reported adverse lipid profile, might be important. Lipid alterations might explain in part the reported weak association between atherosclerotic diseases and current HP infection and might also explain an underestimation of this association when the influence of lipids is controlled for in multivariate analysis.
- ↵Libby P, Egan D, Skarlatos S. Roles of infectious agents in atherosclerosis and restenosis: an assessment of the evidence and need for future research. Circulation. 1997;96:4095–4103.
- ↵Laurila A, Bloigu A, Näyhä S, Hassi J, Leinonen M, Saikku P. Chronic Chlamydia pneumoniae infection is associated with a serum lipid profile known to be a risk factor for atherosclerosis. Arterioscler Thromb Vasc Biol. 1997;17:2910–2913.
- ↵Laurila A, Bloigu A, Näyhä S, Hassi J, Leinonen M, Saikku P. Chlamydia pneumoniae antibodies and serum lipids in Finnish men: cross sectional study. BMJ. 1997;314:1456–1457.
- ↵Murray LJ, O’Reilly DPJ, Ong GML, Evans AE, Bamford KB. Chlamydia pneumoniae antibodies are associated with an atherogenic lipid profile. Heart. 1999;81:239–244.
- ↵Niemelä S, Karttunen T, Korhonen T, Läärä E, Karttunen R, Ikäheimo M, Kesäniemi YA. Could Helicobacter pylori infection increase the risk of coronary heart disease by modifying serum lipid concentrations. Heart. 1996;75:573–575.
- ↵Nieto FJ, Sorlie P, Comstock GW, Wu K, Adam E, Melnick JL, Szklo M. Cytomegalovirus infection, lipoprotein (a), and hypercoagulability: an atherogenic link? Arterioscler Thromb Vasc Biol. 1997;17:1780–1785.
- ↵Brade L, Brunnemann H, Ernst M, Fu Y, Holst O, Kosma P, Näher H, Persson K, Brade H. Occurrence of antibodies against chlamydial lipopolysaccharide in human sera as measured by ELISA using an artificial glycoconjugate antigen. FEMS Immun Med Microbiol. 1994;8:27–41.
- ↵Alvarez C, Ramos A. Lipids, lipoproteins, and apoproteins in serum during infection. Clin Chem. 1986;32:142–145.
- ↵Cabana VG, Siegel JN, Sabesin SM. Effects of the acute phase response on the concentration and density distribution of plasma lipids and apolipoproteins. J Lipid Res. 1989;30:39–49.
- ↵Jacobs DR, Hebert B, Schreiner PJ, Sidney S, Iribarren C, Hulley S. Reduced cholesterol is associated with recent minor illness: the CARDIA Study. Am J Epidemiol. 1997;146:558–564.
- ↵Pesonen E, Rapola J, Viikari J, Turtinen J, Åkerblom HK. Altered serum lipid profile after systemic infection in children: risk factor for CHD? Eur Heart J. 1993;14(suppl K):7–11.
- ↵Feingold KR, Grunfeld C. Role of cytokines in inducing hyperlipidemia. Diabetes. 1992;41(suppl 2):97–101.
- ↵Danesh J, Peto R. Risk factors for coronary heart disease and infection with Helicobacter pylori. BMJ. 1998;316:1130–1132.
- ↵Glader CA, Stegmayr B, Boman J, Stenlund H, Weinehall L, Hallmans G, Dahlén GH. Chlamydia pneumoniae antibodies and high lipoprotein (a) levels do not predict ischemic cerebral infarctions: results from a nested case-control study in northern Sweden. Stroke. 1999;30:2013–2018.
- ↵Glader CA, Boman J, Saikku P, Stenlund H, Weinehall L, Hallmanns G, Dahlén GH. The proatherogenic properties of lipoprotein (a) may be enhanced through the formation of circulating immune complexes containing Chlamydia pneumoniae-specific IgG antibodies. Eur Heart J. 2000;21:639–646.
- ↵Lopes-Virella MF. Interactions between bacterial lipopolysaccharides and serum lipoproteins and their possible role in coronary heart disease. Eur Heart J. 1993;14(suppl K):118–124.