|Home | About | Journals | Submit | Contact Us | Français|
Higher plasma concentrations of the endogenous nitric oxides synthase (NOS) inhibitor asymmetric dimethylarginine (ADMA) are associated with increased risk of cardio- and cerebrovascular events and death, presumably by promoting endothelial dysfunction and subclinical atherosclerosis. We hypothesized that plasma ADMA concentrations are positively related to common carotid artery intimal media thickness (CCA-IMT) and to internal carotid (ICA)/bulb-IMT.
We investigated the cross-sectional relations of plasma ADMA with CCA-IMT and ICA/bulb-IMT in 2958 Framingham Heart Study participants (mean age 58 years, 55% women).
In unadjusted analyses, ADMA was positively related to both CCA-IMT (β per SD increment 0.012, p<0.001) and ICA/bulb IMT (β per SD increment 0.059, p<0.001). In multivariable analyses (adjusting for age, sex, systolic blood pressure, antihypertensive treatment, smoking status, diabetes, body mass index (BMI), Total to HDL cholesterol ratio, log C-reactive protein, and serum creatinine), plasma ADMA was not associated with CCA-IMT (p=0.991), but remained significantly and positively related to ICA/bulb IMT (β per SD increment 0.0246, p=0.002).
In our large community-based sample, we observed that higher plasma ADMA concentrations were associated with greater ICA/bulb-IMT but not with CCA-IMT. These data are consistent with the notion that ADMA promotes subclinical atherosclerosis in a site-specific manner, with a greater proatherogenic influence at known vulnerable sites in the arterial tree.
Carotid artery intimal media thickness (IMT) is a widely accepted indicator of subclinical atherosclerosis burden, with higher values being associated with an adverse cardiovascular prognosis. Consequently, carotid IMT has also been proposed as a surrogate endpoint for therapeutic interventions directed at lowering atherosclerotic burden.1, 2
Accumulating scientific evidence links asymmetric dimethylarginine (ADMA, an endogenous inhibitor of all major isoforms of nitric oxide synthase [NOS]), to cardio- and cerebrovascular disease.3, 4 Higher plasma ADMA concentrations are associated with increased risk of myocardial infarction, stroke, and total mortality in a broad spectrum of people in the general population5-7 and in patients with prevalent coronary heart disease,8, 9 septic shock10 and renal failure.11 It is widely assumed that the association of ADMA and adverse clinical events in these diverse samples is largely related to the attenuation of the vasculoprotective effects of NO, leading to endothelial dysfunction and subsequent atherosclerosis.4,12-15
Few clinical studies have directly related endogenous ADMA concentrations to the extent of atherosclerosis in community-based samples; thus ADMA has been related to CT coronary artery calcification in young adults.16 In 1999 Miyazaki et al. reported a strong positive correlation of ADMA with common carotid artery (CCA) intimal media thickness (IMT) in 122 clinically healthy Japanese,17 a finding that has been replicated in another Japanese sample18 and in smaller studies investigating patients with prevalent disease.19-21 Another recent report noted an inverse association of ADMA concentrations with IMT, thereby rendering the issue unclear.22 These prior studies were limited by modest sample sizes and referral biases. More recently ADMA was related to baseline CCA-IMT and to progression in CCA-IMT over 6 years in two community-based cohorts23, 24 of native American and Japanese persons. However, no study has assessed the association of ADMA with IMT of the carotid bulb and the proximal part of the internal carotid artery (ICA), a surprising omission given abundant clinical and experimental data suggesting that this vascular region may be especially vulnerable to attenuation of NO synthesis.25
We hypothesized that higher ADMA concentrations are associated with greater IMT in the CCA as well as in the ICA/bulb, and tested this hypothesis in the large, community-based Framingham Offspring Study cohort.
We have detailed the selection criteria and design of the Framingham Heart Study elsewhere.26 The Framingham Offspring Study began in 1971 with the enrolment of 5124 participants, who were either the children of the original cohort participants or spouses of these children. The study protocol was approved by the Boston Medical Center Institutional Review Board, and all participants provided written informed consent.
Offspring cohort participants undergo routine examinations at the Heart Study clinic approximately once every four years. At each Heart Study visit, attendees undergo anthropometric measurements, medical history and physical examination, and laboratory assessment of cardiovascular risk factors. Of 3532 attendees at the sixth examination cycle (1995-1998), 3453 (98%) participants had plasma ADMA concentration measured. We excluded 413 individuals from the present investigation because of prevalent CVD including stroke, 124 were excluded for missing carotid information, 29 were excluded for missing ADMA, and 8 were excluded for creatinine greater than 2 mg/dL. After these exclusions, 2958 individuals (1622 women) remained eligible for the present investigation. Participants with prevalent cardiovascular disease (CVD) were excluded to avoid confounding by prevalent disease status. Also, if ADMA is associated with prevalent CVD and prevalent CVD is associated with greater IMT, we could find a positive association between ADMA and IMT that is simply an epiphenomenon.
At the sixth examination cycle, carotid ultrasonography was performed and images were analyzed according to a standardized protocol by a single trained sonographer using a single ultrasound machine (Toshiba Medical Systems, Tustin, California).27, 28 A high-resolution 7.5-MHz transducer was used for imaging the CCA, whereas a 5.0-MHz transducer was used for the carotid bulb and the ICA. The trained sonographer was blinded to the clinical information of the participants, and made IMT measurements, which were overread by 1 of the investigators (JP). Carotid IMT measurements were obtained from 2 diastolic images at 3 sites in each (right and left) carotid artery. These sites were at the level of the distal CCA, at the carotid artery bulb, and at the proximal ICA (defined as the first 2 cms). At each site the maximal IMT was assessed in the right near and far walls, and in the left near and far walls, thus giving 4 wall-segment measurements at each site. These 12 measurements were grouped into two sets. The mean of the 4 measurements recorded in the distal CCA was estimated as the mean maximal CCA-IMT. Thus the mean maximum wall thickness of the CCA was estimated as the mean of the maximum wall thicknesses for near and far wall on both the left and right sides: (mLNW+mLFW+mRNW+mRFW)/4. The mean of the 8 measurements recorded at the carotid bulb and proximal ICA sites was estimated as the mean maximal carotid bulb/ICA-IMT. We have previously reported good reproducibility for our measurement protocols, with intraclass correlation coefficients for mean maximal CCA-IMT and mean maximal ICA/bulb-IMT of 0.86 and 0.74, respectively.27
ADMA was measured as detailed elsewhere) using a fully validated commercially available high-throughput liquid chromatography-tandem mass spectrometry assay (DLD Diagnostika, Hamburg, Germany).29, 30 The intra-assay and the inter-assay coefficients of variation (CV) were 3.2% and <5%, respectively.
Means and standard deviations are presented to summarize continuous clinical, imaging and biochemical characteristics, and percentages are presented for categorical characteristics. We evaluated the distributional properties of the variables graphically, and log-transformed variables that were skewed. Pearson's correlation coefficient was calculated to assess correlations among the variables. T-tests were used for the baseline comparisons of continuous data.
It has previously been shown that the CCA and the ICA segments may differ with regard to their associations with risk common vascular risk factors.1,25 We, therefore, defined two vascular targets to be assessed, i.e, CCA-IMT and the ICA/bulb-IMT. The corresponding null hypothesis to be tested was that in a multivariable-adjusted model neither CCA-IMT nor the ICA/bulb-IMT is related to continuous ADMA concentrations. To account for the inflation of the alpha error due to analysis of two carotid IMT measurements, we applied a Bonferroni correction and defined a p <0.025 (0.05 divided by 2) as indicating statistical significance, and as the threshold required to reject the null hypothesis. All additional analyses were deemed exploratory, and a p value of <0.05 was considered statistically significant for these analyses.
We used multiple linear regression models to relate plasma ADMA concentrations (independent variable) to CCA-IMT and ICA/bulb-IMT (dependent variables), both sets of measurements being modelled as continuous untransformed variables, given their normal distributions. We did not observe effect modification by sex upon formal statistical testing for an interaction between ADMA and sex for CCA-IMT or ICA/bulb IMT measurements. Accordingly, all analyses were performed for pooled sexes.
We constructed two sets of multivariable models: a. Model 1 adjusted for age, sex and body mass index (BMI); b. Model 2 adjusted for age, sex, systolic blood pressure, antihypertensive treatment, smoking status, diabetes, body mass index (BMI), total to HDL cholesterol ratio, C-reactive protein (log-transformed), and serum creatinine (fully-adjusted model). We used a general linear model for analysis of covariance to estimate adjusted least squares means of the IMT measures across quartiles of plasma ADMA. We further evaluated the effect modification of ADMA by age (dichotomized at the median for the sample), diabetes (yes/no), smoking, and body mass index (≥30 kg/m2; yes/no) by incorporating corresponding interaction terms into the multivariable model 2; none of these interactions were statistically significant. We also evaluated effect modification by a composite measure of vascular risk, the Framingham risk score.31
All analyses were performed with SAS statistical software version 9.1 (SAS Institute 2002).
The baseline clinical and biochemical characteristics as well as distributions of carotid IMT measures of the 2958 participants (55% women) in the present investigation are displayed in Table 1.
In unadjusted analyses, ADMA was positively related to both CCA-IMT (β per SD increment 0.012, p<0.001) and ICA/bulb IMT (β per SD increment 0.059, p<0.001). Upon adjustment for other risk factors, ADMA remained positively related to ICA/bulb-IMT (p=0.002), but not to CCA-IMT (p=0.99). Table 2 displays the results of the multivariable-adjusted regression analyses relating ADMA concentrations (modeled as a continuous variable and as quartiles) to CCA-IMT and ICA/bulb-IMT. Correspondingly, we observed an increase of the ICA/bulb-IMT from the first to the fourth quartile of ADMA (p for trend=0.029), while there was no significant trend for CCA-IMT (p=0.39), as shown in Tables 2 and and33.
Using GLM procedures we also assessed whether the ADMA concentration retains an independent association with ICA/bulb-IMT when accounting for the individual cardiovascular risk based on the Framingham Risk Score. The interaction between ADMA quartiles and Framingham Risk Score groups was not found significant (p=0.32), therefore the association of ADMA with ICA/bulb-IMT does not depend on the Framingham Risk Score. Subsequently, in a multivariable model controlling for Framingham Risk Score group, ADMA was associated with ICA/bulb-IMT, (p<0.001).
In our cross-sectional study of a large community-based sample, we observed a positive association of plasma ADMA concentrations with ICA/bulb-IMT but not with CCA-IMT.
The L-arginine-NO pathway is crucial for the regulation of vascular tone and protection of vascular integrity by preventing endothelial activation.32, 33 Premature atherosclerosis and spontaneous myocardial infarction were observed in a new mouse model deficient in all three NOS isoforms, which elegantly verifies the concept that failure of NO synthesis can promote generalized atherosclerosis.34 In a concentration range matching that observed in vivo inside cells, ADMA has discernable inhibitory effects on NO synthesis.35 Mice with partial ablation of the ADMA degrading enzyme dimethylarginine dimethylaminohydrolase 1 (DDAH1) also have only moderately (30-50%) elevated plasma ADMA concentrations, but demonstrate endothelial dysfunction, increased vascular resistance, elevated blood pressure, and reduced cardiac output.36 In contrast, mice over-expressing DDAH1 or DDAH2 appear to be protected from vascular damage.37, 38 This is further supplemented by several clinical studies.3, 5, 8-12 Thus, our cross-sectional observations are consistent with the known association of ADMA with subclinical atherosclerosis in some prior reports.
An alternative explanation for the association of ADMA and ICA IMT would be that in patients with subclinical atherosclerosis higher ADMA concentrations are simply a marker of associated processes such as oxidative stress or concomitant renal dysfunction that impairs its generation, metabolism or excretion.39 We controlled for renal function by excluding individuals with overt renal dysfunction at baseline and adjusting for serum creatinine in all multivariable analyses.
Several investigations have documented the differential responses of different anatomical vascular sites to systemic risk factors.40, 41 The varying association of ADMA with atherosclerotic lesions in carotid artery locations that are only centimetres apart is intriguing. This finding should however be interpreted with caution, since some prior studies did relate ADMA to CCA-IMT;18, 24 the absence of an associaation between ADMA and CCA-IMT in the present study may therefore be due to chance or may reflect differences in the stage of the disease at both sites as well as ethnicity and other differences between our sample and the prior cohort studies. Yet, for the carotid artery, it has frequently been reported that the region of the bulb and the adjacent part of the ICA appear to be more sensitive to circulating risk factors for atherosclerosis than the common carotid artery.42, 43, 44, 45, 46 A major physiological distinction between the CCA and the ICA/bulb are differences in flow patterns (see25 for an excellent review). For anatomical reasons, laminar flow in the carotid bulb is frequently disturbed, leading to insufficient stimulation of NOS expression and activation. Thus, it is conceivable that this location may be especially sensitive to further impairment of NO synthesis by higher ADMA concentrations. The differential association of ADMA with CCA- and ICA-IMT might also reflect a higher prevalence of atherosclerosis and plaque formation in the ICA-IMT and could suggest that ADMA is associated with more severe lesions.
The strengths of our investigation include the large, community-based and well-characterized sample, and the determination of ADMA blinded to IMT measurements. Key limitations of our study include its cross-sectional design that precludes any causal inferences, and the predominantly white sample, which limits the generalizability of our observations to other ethnicities.
In our large community-based sample, we observed that higher plasma concentrations of the endogenous NOS inhibitor ADMA were associated with greater ICA/bulb-IMT but not with greater CCA-IMT. These findings are consistent with the promotion of subclinical atherosclerosis in a site-specific manner (i.e., at sites of arterial vulnerability) by circulating ADMA. Further, ADMA may serve as a biomarker for carotid disease and additional exploration of this pathway could further our understanding of the pathophysiology underlying the development of carotid atherosclerosis.
Grant Support: This work was supported by National Institutes of Health/National Heart, Lung, and Blood Institute Contract N01-HC-25195, N01HV28178, and 2K24HL04334 (Dr Vasan), the National Institute on Aging ( R01 AG16495; AG08122, Dr. Wolf), the National Institute of Neurological Disorders and Stroke ( R01 NS17950, Dr. Wolf) and the Deutsche Forschungsgemeinschaft grant Bo1431/4-1 Dr. Böger). The content is solely the responsibility of the authors and does not necessarily represent the official views of NINDS, NHLBI, NIA or NIH.
Conflict of Interest: Drs. Böger, Schwedhelm, and Maas are named as inventors on patents relating to analytical assays for methylarginines and receive modest royalties from these.
Financial Disclosures: No other authors reported financial disclosures.