点击显示 收起
1 From the Department of Medical and Surgical Critical Area, Center for the Study at Molecular and Clinical Level of Chronic, Degenerative and Neoplastic Diseases to Develop Novel Therapies, University of Florence, Florence, Italy (AMG, SF, AG, FS, GFG, and RA); the Laboratory of Clinical Epidemiology, Geriatric Department, National Institute of Research and Care on Aging, Florence, Italy (AMC and SB); the Division of Nutritional Sciences, Cornell University, Ithaca, NY (BB); and the Longitudinal Studies Section, Gerontology Research Center, National Institute on Aging, National Institutes of Health, Baltimore, MD (LF)
2 Supported as a "target project" by the Italian Ministry of Health. 3 Reprints not available. Address correspondence to AM Gori, Department of Medical and Surgical Critical Area, University of Florence, Section of Cardiology, Viale Morgagni, 85, 50134 Florence, Italy. E-mail: am.gori{at}dac.unifi.it.
ABSTRACT
Background: The mechanism by which high circulating homocysteine concentrations are a risk factor for atherothrombosis is incompletely understood. A proinflammatory state is related to atherosclerosis, and recent studies suggest that acute phase reactants correlate with circulating concentrations of homocysteine.
Objective: We determined whether high concentrations of inflammatory markers are associated with hyperhomocysteinemia independently of dietary vitamin intakes, vitamin concentrations, and cardiovascular disease risk factors in a large, representative sample of the general population.
Design: Five hundred eighty-six men and 734 women were randomly selected from the inhabitants of 2 small towns near Florence, Italy.
Results: After adjustment for multiple potential confounders, interleukin 1 receptor antagonist (IL-1ra) and interleukin 6 (IL-6) concentrations were significantly (P < 0.001) associated with plasma homocysteine concentrations in older (>65 y) populations. Compared with participants in the lowest IL-6 tertile, those in the highest tertile had a higher risk of having homocysteine concentrations that were high (>30 µmol/L; odds ratio: 2.6; 95% CI: 1.1, 5.6; P = 0.024) or in the intermediate range 15–30 µmol/L (odds ratio: 1.6; 95% CI: 1.2, 2.2; P = 0.0014). Sedentary state, intakes of vitamin B-6 and folic acid, and serum folate, vitamin B-12, vitamin B-6, and -tocopherol concentrations were significant independent correlates of homocysteine.
Conclusions: High circulating concentrations of IL-1ra and IL-6 are independent correlates of hyperhomocysteinemia and may explain, at least in part, the association between homocysteine and atherosclerosis.
Key Words: Homocysteine • inflammation • cytokines • macronutrients • micronutrients • vitamin concentrations • InCHIANTI Study
INTRODUCTION
Observational studies consistently show that hyperhomocysteinemia is frequent in persons with cardiovascular disease and is a strong, independent risk factor for new cardiovascular events (1–3). The molecular mechanism by which elevated plasma concentrations of homocysteine are related to the pathogenesis of atherothrombotic disease remains incompletely understood.
In vitro and in vivo studies suggest that homocysteine may accelerate the atherosclerotic process by promoting lipoprotein oxidation, exerting procoagulant activity, and enhancing collagen synthesis and smooth muscle cell proliferation (4, 5). More recently, researchers have focused on the relation between homocysteine and inflammation. Evidence shows that concentrations of acute phase reactants, such as fibrinogen, C-Reactive Protein (CRP), and -1 chymotrypsin, correlate with circulating concentrations of homocysteine (6–9). Preclinical studies indicate that interleukin 6 (IL-6) may interact with vitamin B-6 metabolism and compromise cystathionine ß-synthase activity, thereby rising plasma homocysteine concentrations (10). Interestingly, high circulating concentrations of proinflammatory cytokines are associated with a high risk of medical conditions that have also been associated with hyperhomocysteinemia, such as acute ischemic stroke, myocardial infarction, and, more recently, osteoporosis (11–15). Thus, it may be hypothesized that hyperhomocysteinemia and cardiovascular disease risk may be both mediated, in whole or in part, by a proinflammatory state.
The present study used data from a large population-based sample to determine whether circulating concentrations of inflammatory markers are associated with hyperhomocysteinemia, independently of dietary vitamin intakes, circulating vitamin concentrations, and cardiovascular disease risk factors.
SUBJECTS AND METHODS
Data are from the InCHIANTI Study, a population-based epidemiologic study conducted in the Chianti geographic area (Tuscany, Italy). The methodologic details of the study were described elsewhere (16).
Briefly, in August 1998, 1616 persons aged 20–102 y were selected from the population registry of Greve in Chianti (a rural area; 11709 inhabitants; 19.3% 65 y of age) and Bagno a Ripoli (Antella village, near Florence; 4704 inhabitants; 20.3% 65 y of age). The participation rate was 90% (1453 of 1616).
All subjects gave informed consent to participate in the study, which was approved by the Ethical Committee of the Italian National Institute of Research and Care of Aging. Blood samples were collected in the morning after the participants had fasted for 8 h and sat for 15 min. The analysis reported here is based on 1320 participants with complete data on homocysteine, inflammatory markers, and other important covariates.
Assays
Commercial enzymatic tests (Roche Diagnostics, GmbH, Mannheim, Germany) were used to measure serum total and HDL-cholesterol and triacylglycerol concentrations. Serum LDL-cholesterol concentrations were calculated by using the Friedewald formula. The interassay CV was <3.8% for total cholesterol, <5% for HDL cholesterol, and <2.5% for triacylglycerols. A low serum HDL-cholesterol concentration was defined as a value 40 mg/dL, a high serum LDL-cholesterol concentration as a value 160 mg/dL, and a high triacylglycerol concentration as a value 200 mg/dL.
Serum concentrations of interleukin 1ß (IL-1ß), IL-6, soluble IL-6 receptor (sIL-6r, 80 kDa), tumor necrosis factor (TNF-), and interleukin 1 receptor antagonist (IL-1ra) were quantified by high-sensitivity enzyme-linked immunoabsorbent assays with the use of commercial kits (BioSource Cytoscreen, Camarillo, CA). The minimum detectable concentrations were 0.01 pg/mL for IL-1ß, 0.10 pg/mL for IL-6, 8.0 pg/mL for sIL-6r, 0.09 pg/mL for TNF-, and 4.0 pg/mL for IL-1ra. The interassay CVs were 4.5% for IL-1ra and 7.0% for IL-6, sIL-6r, IL-1ß, and TNF-. Serum concentrations of IL-18 were detected in duplicate by using highly sensitive quantitative sandwich assays (Quantikine HS; R&D Systems, Minneapolis, MN). The minimum detectable concentrations were 0.7 pg/mL, and the interassay CV was 7%.
Serum CRP was measured in duplicate by using an enzyme-linked immunoabsorbent assay and a colorimetric competitive immunoassay that uses purified protein and polyclonal anti-CRP antibodies. The minimum detectable threshold was 0.03 mg/L, and the interassay CV was 5%.
For the measurement of homocysteine, whole venous blood was collected in tubes containing EDTA (0.17 mol/L), immediately placed in ice, and centrifuged within 30 min at 4°C (2000 x g for 15 min). The supernatant fluid was stored in aliquots at –80°C until assayed. Plasma concentrations of total homocysteine (free and protein bound) were measured with a fluorimetric polarized immunoassay method (IMX; Abbot Laboratories, Oslo, Norway). The sensitivity of the IMX homocysteine assay was 0.5 µmol/L, and the interassay CV was 4.1%=–0987
Sera for measuring folate, vitamin B-6, and B-12 were obtained by centrifuging blood collected in evacuated tubes without anticoagulant at 2000 x g for 10 min and were stored at –20°C. Vitamin B-6 was measured by HPLC (Immundiagnostik, Bensheim, Germany) and vitamin B-12 and folates by radioimmunoassay (ICN Pharmaceuticals, New York, NY). The minimum detectable concentrations were 0.6 ng/mL for folate, 0.2 ng/mL for vitamin B-6, and 75 pg/mL for vitamin B-12; the intraassay CVs were 4.1% for folate, 2.8% for vitamin B-6, and 11.2% for vitamin B-12; and the interassay CVs were 7.1% for folate, 4.1% for vitamin B-6, and 12.3% for vitamin B-12. Plasma vitamin E (-tocopherol) concentrations were measured by reversed-phase HPLC as previously described (17) and expressed in mmol/L. Triplicate analysis of the reference samples provided by the American Association for Laboratory Accreditation (Washington, DC) showed an intrabatch CV of 3% and an interbatch CV of 4.2%.
Covariates
Three consecutive measurements of blood pressure were made according to a standardized protocol after the subjects had rested for 10 min, and the average of the last 2 measurements was used to calculate systolic and diastolic blood pressure. Hypertension was defined as a systolic blood pressure >130 mm Hg and a diastolic blood pressure >80 mm Hg or current antihypertensive treatment. Diabetes mellitus was defined according to American Diabetes Association criteria (18).
Body mass index was calculated as weight (kg)/height2 (m) and was dichotomized in the analysis as >27 compared with <27. Waist circumference was dichotomized as > or <102 cm in men and 88 cm in women. The average daily intakes of alcohol (<30 compared with >30 g/d) and vitamins were estimated by administering the European Prospective Investigation into Cancer and Nutrition (EPIC) food-frequency questionnaire.
The EPIC food-frequency questionnaire provides a detailed assessment of food consumption during the previous year through a large number of structured and pre-coded questions. Originally, the questionnaire was conceived to be self-administered. However, in a pilot study we realized that in older subjects this method of administration provides ambiguous results, mainly because the questions are misunderstood. Thus, in the InCHIANTI Study, the EPIC questionnaire was administered by the interviewers. The information provided by the questionnaire was transformed into an average daily intake of macronutrients and micronutrients, including vitamins, by using custom software that used for reference a food-composition database for epidemiologic study in Italy, which was edited by the European Institute of Oncology in 1998 (19, 20).
On the basis of responses to multiple questions, a sedentary state was defined as being inactive or performing light-intensity physical activity (ie, walking, light housework) <1 h/wk. Smoking status was assessed by self-report. Pack-years, a measure of smoking exposure that combines intensity and duration, was calculated as packs smoked per day multiplied by years of smoking.
Statistical analysis
All analyses were performed with the use of the SAS statistical package (version 8.02; SAS Institute Inc, Cary, NC). P values <0.05 were considered statistically significant. Log-transformed values for homocysteine, IL-1ß, IL-1ra, IL-6, IL-6r, IL-18, TNF-, CRP, vitamin B-6, vitamin B-12, and folate were used in the analyses and back transformed for data presentation.
Differences in characteristics between men and women were determined by using a general linear model or a logistic regression analysis adjusted for age. To evaluate the relation between circulating homocysteine concentrations, vitamin intakes, circulating vitamin concentrations, and cardiovascular disease risk factors, we divided the study population into tertiles of circulating homocysteine (<12.2, 12.2–15.6, and >15.6 µmol/L). Differences between tertiles were evaluated by general linear models adjusted for sex, age, total energy intake, and serum creatinine.
Furthermore, in the final adjusted model of the multivariate linear regression analysis, we included sex x IL-6 and sex x IL-1ra as well as age x IL-6 and age x IL-1ra (age was coded as <65 y or 65 y) among the independent variables to test the interaction between inflammatory markers and sex and inflammatory markers and age, respectively.
In addition, we used multivariate linear regression analyses to test the independent association of inflammatory markers with plasma homocysteine concentrations, after adjusting for vitamin intakes, circulating vitamin concentrations, and cardiovascular disease risk factors. Variables that were not significantly and independently associated with homocysteine were removed from the final regression model by backward selection.
High concentrations of IL-1ra and IL-6 were defined as the highest tertile of the distribution of these cytokines. The odds ratio (OR) for having high (15–30 or > 30 µmol/L) compared with normal (<15 µmol/L) concentrations of homocysteine, according to cytokine tertiles, was estimated by multivariate logistic regression analysis with the use of homocysteine groups as a polychotomous dependent variable and high IL-6 and high IL-1ra as independent variables. Age; sex; serum creatinine; intakes of energy, folate, and vitamin B-6; circulating concentrations of vitamin B-6, vitamin B-12, -tocopherol, and folate; and sedentary state as covariates.
RESULTS
The characteristics of the InCHIANTI participants are shown in Table 1. The InChianti study participants consists of 734 females (55.6%) and 586 males (44.4%); significant differences by sex were found.
View this table:
TABLE 1. Characteristics of the study population
Mean (and 95% CI) plasma homocysteine concentrations were significantly higher (P < 0.0001) in men (15.3 µmol/L; 95% CI: 14.9, 15.7 µmol/L) than in women (13.5 µmol/L; 95% CI: 13.2, 13.8 µmol/L) after adjustment for age, serum creatinine, and total energy intake.
Adjusting for sex, serum creatinine and total energy intake, circulating concentrations of homocysteine were significantly (P < 0.0001) different across age groups (<65 y: mean 11.7 µmol/L, 95% CI 11.3–12.1 µmol/L, 65–74 y: 13.8 µmol/L, 95% CI 13.5–14.1 µmol/L, 75–84 y: 16.2 µmol/L, 95% CI 15.7–16.7 µmol/L and > 85 y: 18.3 µmol/L, 95% CI 17.4–19.3 µmol/L).
After adjustment for age, sex, total energy intake, and serum creatinine, significantly higher concentrations of IL-6, IL-1ra, TNF-, and IL-18 were found across homocysteine tertiles (Table 2).
View this table:
TABLE 2. Mean (and 95% CI) serum concentrations of inflammatory markers according to tertiles of circulating homocysteine concentrations1
After adjustment for age, sex, serum creatinine, and total energy intake, the average daily dietary intake of many vitamins (all B vitamins, ß-carotene, folic acid, and vitamins C, A, and E) and serum concentrations of folate, vitamin B-12, vitamin B-6, and -tocopherol were progressively and significantly lower according to homocysteine tertiles. After adjustment for the same covariates, participants with higher homocysteine concentrations were significantly more likely to be sedentary and to have low HDL-cholesterol concentrations (40 mg/dL) and high LDL-cholesterol concentrations (160 mg/dL) (Table 3). No significant interaction of sex and IL-6 or sex and IL-1ra was found. There was a significant interaction of age and IL-6, but not of age and IL-1ra.
View this table:
TABLE 3. Mean (and 95% CI) daily intakes and circulating concentrations of vitamins and cardiovascular disease risk factors according to tertiles of circulating homocysteine concentrations
In the multivariate linear regression analyses independent of age, sex, serum creatinine, and total energy intake, circulating concentrations of IL-6 and IL-1ra were strongly associated with homocysteine (models 1 and 2; Table 4) in the whole population as well as in older subject (>65 y) subgroups; the significant association was maintained in a model that included IL-6 and IL-1ra and adjusted for all the variables found to be associated with circulating homocysteine concentrations in the preliminary analyses (model 3; Table 4). When all variables not significantly and independently associated with circulating homocysteine were removed from this last model, IL-6 and IL-1ra concentrations as well as age, male sex, serum creatinine, total energy intake, sedentary state, vitamin B-6 and folate intakes, and circulating concentrations of vitamin B-12, vitamin B-6, folate, and -tocopherol were all significant, independent predictors of higher circulating homocysteine concentrations (model 4; Table 4) in the whole group as well as in older subjects.
View this table:
TABLE 4. Multiple linear regression models for markers of inflammation that influence homocysteine concentrations (log)1
In the whole group, after adjustment for these significant covariates and compared with participants in the lowest IL-6 tertile, those in the higher IL-6 tertiles were significantly more likely to have circulating homocysteine concentrations >30 µmol/L (OR: 2.6; 95% CI: 1.1, 5.6; P = 0.024) or homocysteine in the range 15–30 µmol/L (OR: 1.6; 95% CI: 1.2, 2.2; P = 0.0014) than homocysteinemia (<15 µmol/L), which was considered the reference condition (Figure 1). The association between high concentrations of IL-1ra (top tertile versus lowest tertiles) and an increased risk of having higher homocysteinemia (OR: 1.6; 95% CI: 0.7, 3.6; P = 0.231 for homocysteine > 30 µmol/L and OR: 1.3; 95% CI: 0.98, 1.7 for homocysteine 15–30 µmol/L; P = 0.070) was not statically significant.
FIGURE 1.. Odds ratios (and 95% CIs) for high (>30 µmol/L and intermediate (15–30 µmol/L) compared with normal (<15 µmol/L) homocysteine concentrations associated with circulating concentrations of interleukin 6 (IL-6) in the highest (3rd) compared with the 2 lowest tertiles (1st and 2nd). Values were estimated in a single polychotomous logistic regression model adjusted for age; sex; serum creatinine; sedentary state; energy, folic acid, and vitamin B-6 intakes; and circulating concentrations of vitamin B-12, vitamin B-6, folate, and -tocopherol. , Homocysteine concentrations >30 µmol/L (n = 45) versus <15 µmol/L (n = 799); P = 0.0024. •, Homocysteine concentrations of 15–30 µmol/L (n = 467) versus <15 µmol/L (n = 799); P = 0.0014.
DISCUSSION
This study provides original information about the complex relation between homocysteine, biomarkers of inflammation, vitamin dietary intakes, and circulating concentrations in a large population-based study, which includes >1300 Italian subjects dispersed over a wide age range. We found that, in older subjects, IL-6 and IL-1ra, but not other markers of inflammation (CRP, IL-1ß, IL-6r, IL-18, and TNF-), were independent predictors of homocysteine concentrations, independent of the dietary intake and circulating concentrations of the vitamins involved in the methionine cycle.
Our results show that subjects with IL-6 concentrations in the top tertile of the distribution had an increased risk of having circulating concentrations of homocysteine that, in clinical practice, are considered a marker of high cardiovascular disease risk. Thus, a common inflammatory factor may explain, at least in part, the association between high circulating concentrations of homocysteine and cardiovascular diseases described in many observational studies (1–3). In fact, a large number of experimental and clinical studies have provided convincing evidence of the crucial role of inflammation in the development and progression of atherosclerosis processes (21, 22); in particular, elevated IL-6 concentrations are closely related to an increased risk of future myocardial infarction (23) as well as clinical outcome in unstable angina (24, 25).
Interestingly, we found that IL-1ra, a naturally occurring antagonist of the pro-inflammatory cytokine IL-1ß, was also an independent correlate of homocysteinemia. Although, biologically, IL-1ra has antiinflammatory properties, because it competitively binds to IL-1ß membrane receptors, it is also an acute phase reactant produced by the liver in large quantities during an inflammatory state (26). Thus, high circulating concentrations of IL-1ra are generally considered to indicate a proinflammatory state. Interestingly, Biasucci et al (24) found that increasing concentrations of IL-1ra and IL-6 predict the risk of in-hospital coronary events in patients with unstable angina.
The tight relation between inflammation and homocysteinemia observed in our study suggest that homocysteine concentrations may be influenced by active inflammatory processes, as suggested by the higher concentrations of plasma homocysteine that are often observed in the days after an acute myocardial infarction (27) or stroke (28). Tissue damage accelerates specific remethylation reactions of DNA, RNA, and various proteins during tissue repair, with consequent generation of S-adenosylhomocysteine and release of homocysteine (29). However, given the cross-sectional nature of our study, a reverse causality, namely homocysteine directly stimulating proinflammatory signaling molecules, cannot be excluded. Studies have shown that homocysteine determines acute and chronic endothelial dysfunction by promoting the production of hydrogen peroxide and other highly reactive oxygen compounds, up-regulating cell adhesion molecules and inhibiting the release of nitric oxide (5, 30). Interestingly, homocysteine concentrations ranging from 10 to 100 µmol/L induce the monocyte expression and production of IL-8 and monocyte chemoattractant protein 1, chemokines that are essential in modulating the role of leukocytes in the inflammatory response to different types of vascular injuries. Additionally, folic acid supplementation of hyperhomocysteinemic patients reduces IL-8 release from peripheral blood mononuclear cells (31).
In our study, the association between inflammatory state and homocysteinemia was found in subjects aged 65 y but not in subjects aged <65 y. This finding is consistent with several lines of research, which have shown that a large percentage of old and very old persons are affected by a mild proinflammatory state (32, 33). The continuous inflammatory stimulus might determine an increased demand for folate that, if not compensated for by increased intakes, may cause hyperhomocysteinemia. The inflammatory process creates an oxidative stress condition. Interestingly, in our study, low concentrations of -tocopherol (circulating vitamin E) were significantly associated with higher homocysteinemia, which suggests that noncompensated oxidative stress may contribute to the increase in plasma homocysteine concentrations. The oxidative stress determines nucleic acid damages, including base modifications, double-base lesions, and stand breaks (34). Folate is required for DNA synthesis and repair (35). Interestingly, in animal models, activated macrophages overexpress high-affinity folate receptor (36). Thus, it is likely that in an inflammatory state vitamins involved in the methionine cycle may be mobilized from the liver and peripheral tissues to sites of inflammation (37).
Previous studies that addressed the relation between inflammation and homocysteine reported conflicting findings. A weak association between CRP and homocysteine concentrations was observed in the participants of the Framingham Heart Study (38) and in the apparently healthy middle-aged men enrolled in the Physicians’ Health Study (9). Conversely, inflammatory markers were not statistically associated with plasma homocysteine concentrations, either in 519 healthy middle-aged adults of the Atherosclerosis Risk in Communities Study (39) or in 373 healthy persons aged >65 y examined by Ravaglia et al (40).
An unexpected finding of our study was the significantly low prevalence of participants with LDL concentrations >160 mg/dL in the highest tertile of homocysteine. The reason for this finding remains unknown, but it is noteworthy that after adjustment for multiple potential confounders, LDL-cholesterol was no longer independently associated with homocysteine concentrations.
Our results provide solid evidence of a relation between inflammation and homocysteine, which has been suggested by previous investigations (6–9, 38). The fact that we found a strong independent association of IL-6 and IL-1ra but not of CRP with homocysteine is puzzling because IL-6 increases early in inflammation (41) and is the principal inductor of hepatocyte messenger RNA CRP expression (42, 43). Proinflammatory cytokines are involved in the transcriptional control of CRP production (44) and, additionally, the synthesis of CRP is strongly influenced by mechanisms of posttranscriptional regulation that are independent of IL-6 (45).
In conclusion, the results of our large population-based study showed a strong, independent association between inflammatory markers and homocysteine concentrations and identified the cytokines IL-6 and IL-1ra as major determinants involved in this association. This association, independent of the leading factor, may explain, at least in part, why subjects with high concentrations of homocysteine have a high risk of developing cardiovascular diseases. Future studies of homocysteine as a modifiable risk factor for health outcomes need to consider inflammation as a potential confounder in the analysis. The evaluation of homocysteine concentrations in addition to that of vitamin B-6, vitamin B-12, folate, and interleukins after 6 y of follow-up will allow us to further elucidate the relation between inflammation and homocysteine.
ACKNOWLEDGMENTS
LF, AMG, RA, and GFG were responsible for the study design. AMC, BB, SB, and FS were responsible for the collection of data. LF and SB were responsible for the statistical analysis. LF, SB, and FS were responsible for the clinical evaluation of the subjects. BB was responsible for the administration of the food-frequency questionnaire. AMG, SF, AMC, and AG were responsible for the laboratory investigations. AMG, LF, RA, and GFG were responsible for writing the manuscript. All authors disclose any affiliation with any organization with a financial interest, direct or indirect, in the subject matter or materials discussed in the manuscript that may affect the conduct or reporting of the work submitted
REFERENCES