Atherosclerosis 234 (2014) 17e22
Contents lists available at ScienceDirect
Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis
S-adenosylhomocysteine is associated with subclinical atherosclerosis and renal function in a cardiovascular low-risk population Adam M. Zawada a,1, Kyrill S. Rogacev a,1, Björn Hummel b, c, 2, Judith T. Berg a, Annika Friedrich a, Heinz J. Roth d, Rima Obeid c, Jürgen Geisel c, Danilo Fliser a, Gunnar H. Heine a, * a
Department of Internal Medicine IV, Saarland University Medical Center, Homburg, Germany Department of Clinical Hemostaseology and Transfusion Medicine, Saarland University Medical Center, Homburg, Germany Clinical Chemistry and Laboratory Medicine/Central Laboratory, Saarland University Medical Center, Homburg, Germany d Labor Dr. Limbach und Kollegen, Medizinisches Versorgungszentrum, Heidelberg, Germany b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 28 November 2013 Received in revised form 22 January 2014 Accepted 5 February 2014 Available online 18 February 2014
Objective: Although homocysteine has been proposed as a cardiovascular risk factor, interventional trials lowering homocysteine have not consistently demonstrated clinical benefit. Recent evidence proposed the homocysteine metabolite S-adenosylhomocysteine (SAH) rather than homocysteine itself as the real culprit in cardiovascular disease. Of note, SAH is predominantly excreted by the kidneys, and cannot be lowered by vitamin supplementation. Due to its cumbersome measurement, data from large studies on the association between SAH, kidney function and cardiovascular disease are not available. Methods: We recruited 420 apparently healthy subjects into our I Like HOMe FU study. Among all study participants, we assessed parameters of C1 metabolism (homocysteine, SAH and S-adenosylmethionine), renal function (estimated glomerular filtration rate [eGFR]) and subclinical atherosclerosis (common carotid intima-media-thickness [IMT]). eGFR was estimated by the CKD-EPIcreat-cys equation. Results: Traditional cardiovascular risk factors and subclinical atherosclerosis were associated with SAH, but not with homocysteine (IMT vs SAH: r ¼ 0.129; p ¼ 0.010; IMT vs homocysteine: r ¼ 0.009; p ¼ 0.853). Moreover, renal function was more closely correlated with SAH than with homocysteine (eGFR vs SAH: r ¼ 0.335; p < 0.001; eGFR vs homocysteine: r ¼ 0.250; p < 0.001). The association between eGFR and SAH remained significant after adjustment for traditional cardiovascular risk factors. Conclusion: In summary, cardiovascular risk factors, subclinical atherosclerosis and eGFR are more strongly associated with SAH than with homocysteine in apparently healthy subjects. Thus, SAH might represent a more promising target to prevent cardiovascular disease than homocysteine. Ó 2014 Elsevier Ireland Ltd. All rights reserved.
Keywords: Homocysteine S-adenosylhomocysteine S-adenosylmethionine Kidney Arteriosclerosis
1. Introduction Hyperhomocysteinemia has independently predicted cardiovascular disease in multiple prospective cohort studies [1e3]. In line with this, recent post hoc analyses of the MESA (Multi-Ethic
* Corresponding author. Tel.: þ49 6841 1623527; fax: þ49 6841 1623545. E-mail addresses: [email protected] (A.M. Zawada), kyrill.rogacev@uks. eu (K.S. Rogacev), [email protected] (B. Hummel), judith_berg@hotmail. de (J.T. Berg), [email protected] (A. Friedrich), [email protected] (H.J. Roth), [email protected] (R. Obeid), [email protected] (J. Geisel), danilo.fl[email protected] (D. Fliser), [email protected] (G.H. Heine). 1 Both authors contributed equally. 2 Present address: Department of Clinical Hemostaseology and Transfusion Medicine, Germany. http://dx.doi.org/10.1016/j.atherosclerosis.2014.02.002 0021-9150/Ó 2014 Elsevier Ireland Ltd. All rights reserved.
Study of Atherosclerosis) and the NHANES III (National Health and Nutrition Examination Survey III) cohorts demonstrated that addition of homocysteine to the Framingham Risk Score improved risk prediction considerably [4]. However, a causal role for homocysteine in cardiovascular disease remains controversial as large-scale interventional trials have failed to show positive results by lowering homocysteine via vitamin supplementation [5e8]. Recently, a more comprehensive acknowledgment of the metabolism of homocysteine and its derivatives, referred to as “one carbon metabolism (C1 metabolism)”, allowed a possible explanation of these seemingly controversial findings [9,10]. In C1 metabolism homocysteine is remethylated to S-Adenosylmethionine (SAM) which serves as universal methyl group donor. After transfer of its methyl group, SAM is converted to S-Adenosylhomocysteine (SAH)
18
A.M. Zawada et al. / Atherosclerosis 234 (2014) 17e22
and subsequently to homocysteine (Figure S1). Despite this strong interrelationship between homocysteine and SAH, vitamin supplementation, which lowers plasma homocysteine levels, does not affect SAH levels [11,12]. In keeping with this finding, recent pioneering clinical and experimental studies suggested SAH rather than homocysteine as the real culprit in cardiovascular disease [13e15]. SAH is a strong inhibitor of DNA methyltransferases and thus directly involved in epigenetic gene regulation. Interestingly, small pilot studies have found renal excretory function to determine plasma SAH [16,17]; in turn, elevated SAH concentrations have been linked to disturbed regulation of atherosclerosis-related genes in hemodialysis patients, which may potentially contribute to their high cardiovascular disease burden [18]. We hypothesized that an interrelationship between renal function, C1 metabolism and subclinical atherosclerosis may exist even among apparently healthy subjects from the general population. For this purpose we analyzed plasma homocysteine, SAH, and SAM in 420 individuals with low cardiovascular risk. 2. Methods 2.1. Study population The I Like HOMe (Inflammation, Lipoprotein Metabolism and Kidney Damage in early atherogenesis e The Homburg Evaluation) study project analyzes the relationship between inflammatory and metabolic risk factors and subclinical atherosclerosis among health-care workers without overt cardiovascular disease. The I Like HOMe study initially comprised 620 subjects, who were recruited between 2004 and 2005. Participants represented a cardiovascular low-risk population, and characteristics of the initial I Like HOMe study cohort have been previously published [19]. Between 2009 and 2011 I Like HOMe participants were re-invited for a follow-up examination (henceforth referred to as ‘I Like HOMe FU’), in which parameters for C1 metabolism, renal function and subclinical atherosclerosis were assessed. 369 participants from the initial I Like HOMe study accepted our invitation, with a further 51 healthy individuals who wished to participate added to the I Like HOMe FU cohort, yielding a total cohort size of 420 subjects. All study participants provided written consent. The study protocol was approved by the local Ethics Committee. A history of smoking, diabetes, current medication intake, cardiovascular comorbidity was obtained via a standardized questionnaire. All participants who had stopped smoking at least one month before study entry were categorized as former smokers. Diabetes was confirmed in subjects with self-reported diabetes mellitus, via a non-fasting plasma glucose 200 mg/dl, a fasting plasma glucose 126 mg/dl, or with current use of hypoglycaemic medication. Body mass index (BMI) was calculated as individual’s body weight divided by the square of their height. Systolic blood pressure (SBP), diastolic blood pressure (DBP), and heart rate were measured after 5 min of inactivity via an automated blood pressure recording apparatus (GE Carescape DINAMAP V100; GE Healthcare). Mean arterial blood pressure was calculated as DBP þ [(SBP DBP)/3], and pulse pressure was calculated as SBP DBP. 2.2. Assessment of intima-media-thickness Intima-media-thickness (IMT) was measured in a supine position with the head turned slightly to the opposite direction. IMT was recorded at the far-wall of the common carotid artery proximal to the bulbus. The Acuson Sequoia C512 Ultrasound Unit (Acuson, Thousand Oaks, CA) equipped with a linear transducer (model 15L8w, Acuson; 8e15 MHz) and automatic measurement software were used.
2.3. Clinical laboratory analysis Blood samples were obtained after overnight fasting. Routine clinical laboratory parameters were measured in the Central Laboratory of the Saarland University Medical Center. Renal function was assessed by estimated glomerular filtration rate according to the CKD-EPIcreatcys equation [20]. 2.4. Analysis of parameters of C1 metabolism For the quantification of S-adenosylmethionine (SAM) and Sadenosylhomocysteine (SAH), EDTA samples were immediately placed on ice and centrifuged within 60 min for 10 min at 2000 g. Before storing the samples at 70 C, 1000 ml of EDTA-plasma was acidified with 100 ml 1 N acetic acid to prevent SAM degradation. The HPLC-MS/MS detection of SAM and SAH concentrations in EDTA-plasma was performed using a Waters 2795 alliance HT, coupled to a Quattro Micro API tandem mass spectrometer (Waters Corporation) according to the method described by Kirsch et al. [21]. Total plasma homocysteine concentrations were determined by using a fluorescence polarization immunoassay on the Abbott AxSym system (Abbott Laboratories, North Chicago, IL). 2.5. Statistical analysis Data management and statistical analysis were performed with SPSS 20.0 (IBM, New York City, NY). Unless indicated otherwise, continuous data are expressed as mean standard deviation and compared by Student’s t-test for independent variables. Categorical variables are presented as percentage of participants. Correlation coefficients were calculated by Pearson test. Subsequently, multivariate linear regression analysis was calculated (method of variable entry: stepwise), which included gender and age (Model 1), traditional cardiovascular risk factors (body mass index, systolic BP, HDL-cholesterol, triglycerides, and blood glucose (Model 2)), and eGFR and albuminuria (Model 3) as independent parameters, and plasma SAH as dependent variable. Two-sided P-values < 0.05 were considered significant. 3. Results Of 420 I Like HOMe FU study participants, 402 were included in the present analysis. The remaining 18 subjects were excluded, as one or more of the following variables were missing: homocysteine, SAH, SAM, albuminuria, cystatin C and/or creatinine. 3.1. Traditional cardiovascular risk factors, GFR and subclinical atherosclerosis Among the 402 I Like HOMe FU participants, 270 were female (67.2%), 107 were current smokers (26.6%), 22 had diabetes mellitus (5.5%), 86 were on antihypertensive medication (21.4%) and 94 had a family history of premature-onset cardiovascular disease (23.4%), respectively. Further baseline characteristics are displayed in Table 1; overall I Like HOMe FU participants represented a low risk study population, and accordingly 337 subjects had a Framingham Risk Score (FRS) below 10% (median FRS: 1%; IQR 1%e5%). Mean (SD) intima-media-thickness among all I Like HOMe FU participants was 0.628 0.127 mm. IMT values showed significant correlation with classical cardiovascular risk factors, as listed in Table S1. Furthermore, IMT values differed between female (0.616 0.118 mm) and male participants (0.652 0.142 mm; p ¼ 0.009), but not between subjects with or without albuminuria (0.640 0.154 mm and 0.627 0.127 mm, respectively; p ¼ 0.719),
A.M. Zawada et al. / Atherosclerosis 234 (2014) 17e22
3.3. C1 parameters and kidney function
Table 1 Baseline characteristics of I Like HOMe FU participants (n ¼ 402). Total cohort (n ¼ 402)
SAH 10.75 SAH > 10.75 p nmol/l (n ¼ 201) nmol/l (n ¼ 201)
Age (years) 48 7 47 7 Weight (kg) 77 17 71 16 Height (cm) 169 9 167 8 Body mass index 27 5 26 5 2 (kg/m ) Waist-to-Hip Ratio 0.85 0.09 0.82 0.08 137 18 133 17 Systolic BP (mmHg) Diastolic BP 89 12 87 10 (mmHg) Mean BP 105 13 103 11 (mmHg) Pulse pressure 48 12 46 12 (mmHg) 101 12 105 10 eGFR (ml/min/1.73 m2) Plasma creatinine 0.83 0.15 0.77 0.10 (mg/dl) Cystatin C (mg/l) 0.74 0.10 0.70 0.09 IMT (mm) 0.628 0.127 0.610 0.122 Total cholesterol 215 37 214 38 (mg/dl) HDL-cholesterol 64 17 67 17 (mg/dl) Triglycerides 114 67 102 51 (mg/dl) Blood glucose 97 16 94 13 (mg/dl) C-reactive protein 2.1 2.3 2.0 2.2 (mg/l)
50 82 172 28
7 17 10 5
19
0.001
Contents lists available at ScienceDirect
Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis
S-adenosylhomocysteine is associated with subclinical atherosclerosis and renal function in a cardiovascular low-risk population Adam M. Zawada a,1, Kyrill S. Rogacev a,1, Björn Hummel b, c, 2, Judith T. Berg a, Annika Friedrich a, Heinz J. Roth d, Rima Obeid c, Jürgen Geisel c, Danilo Fliser a, Gunnar H. Heine a, * a
Department of Internal Medicine IV, Saarland University Medical Center, Homburg, Germany Department of Clinical Hemostaseology and Transfusion Medicine, Saarland University Medical Center, Homburg, Germany Clinical Chemistry and Laboratory Medicine/Central Laboratory, Saarland University Medical Center, Homburg, Germany d Labor Dr. Limbach und Kollegen, Medizinisches Versorgungszentrum, Heidelberg, Germany b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 28 November 2013 Received in revised form 22 January 2014 Accepted 5 February 2014 Available online 18 February 2014
Objective: Although homocysteine has been proposed as a cardiovascular risk factor, interventional trials lowering homocysteine have not consistently demonstrated clinical benefit. Recent evidence proposed the homocysteine metabolite S-adenosylhomocysteine (SAH) rather than homocysteine itself as the real culprit in cardiovascular disease. Of note, SAH is predominantly excreted by the kidneys, and cannot be lowered by vitamin supplementation. Due to its cumbersome measurement, data from large studies on the association between SAH, kidney function and cardiovascular disease are not available. Methods: We recruited 420 apparently healthy subjects into our I Like HOMe FU study. Among all study participants, we assessed parameters of C1 metabolism (homocysteine, SAH and S-adenosylmethionine), renal function (estimated glomerular filtration rate [eGFR]) and subclinical atherosclerosis (common carotid intima-media-thickness [IMT]). eGFR was estimated by the CKD-EPIcreat-cys equation. Results: Traditional cardiovascular risk factors and subclinical atherosclerosis were associated with SAH, but not with homocysteine (IMT vs SAH: r ¼ 0.129; p ¼ 0.010; IMT vs homocysteine: r ¼ 0.009; p ¼ 0.853). Moreover, renal function was more closely correlated with SAH than with homocysteine (eGFR vs SAH: r ¼ 0.335; p < 0.001; eGFR vs homocysteine: r ¼ 0.250; p < 0.001). The association between eGFR and SAH remained significant after adjustment for traditional cardiovascular risk factors. Conclusion: In summary, cardiovascular risk factors, subclinical atherosclerosis and eGFR are more strongly associated with SAH than with homocysteine in apparently healthy subjects. Thus, SAH might represent a more promising target to prevent cardiovascular disease than homocysteine. Ó 2014 Elsevier Ireland Ltd. All rights reserved.
Keywords: Homocysteine S-adenosylhomocysteine S-adenosylmethionine Kidney Arteriosclerosis
1. Introduction Hyperhomocysteinemia has independently predicted cardiovascular disease in multiple prospective cohort studies [1e3]. In line with this, recent post hoc analyses of the MESA (Multi-Ethic
* Corresponding author. Tel.: þ49 6841 1623527; fax: þ49 6841 1623545. E-mail addresses: [email protected] (A.M. Zawada), kyrill.rogacev@uks. eu (K.S. Rogacev), [email protected] (B. Hummel), judith_berg@hotmail. de (J.T. Berg), [email protected] (A. Friedrich), [email protected] (H.J. Roth), [email protected] (R. Obeid), [email protected] (J. Geisel), danilo.fl[email protected] (D. Fliser), [email protected] (G.H. Heine). 1 Both authors contributed equally. 2 Present address: Department of Clinical Hemostaseology and Transfusion Medicine, Germany. http://dx.doi.org/10.1016/j.atherosclerosis.2014.02.002 0021-9150/Ó 2014 Elsevier Ireland Ltd. All rights reserved.
Study of Atherosclerosis) and the NHANES III (National Health and Nutrition Examination Survey III) cohorts demonstrated that addition of homocysteine to the Framingham Risk Score improved risk prediction considerably [4]. However, a causal role for homocysteine in cardiovascular disease remains controversial as large-scale interventional trials have failed to show positive results by lowering homocysteine via vitamin supplementation [5e8]. Recently, a more comprehensive acknowledgment of the metabolism of homocysteine and its derivatives, referred to as “one carbon metabolism (C1 metabolism)”, allowed a possible explanation of these seemingly controversial findings [9,10]. In C1 metabolism homocysteine is remethylated to S-Adenosylmethionine (SAM) which serves as universal methyl group donor. After transfer of its methyl group, SAM is converted to S-Adenosylhomocysteine (SAH)
18
A.M. Zawada et al. / Atherosclerosis 234 (2014) 17e22
and subsequently to homocysteine (Figure S1). Despite this strong interrelationship between homocysteine and SAH, vitamin supplementation, which lowers plasma homocysteine levels, does not affect SAH levels [11,12]. In keeping with this finding, recent pioneering clinical and experimental studies suggested SAH rather than homocysteine as the real culprit in cardiovascular disease [13e15]. SAH is a strong inhibitor of DNA methyltransferases and thus directly involved in epigenetic gene regulation. Interestingly, small pilot studies have found renal excretory function to determine plasma SAH [16,17]; in turn, elevated SAH concentrations have been linked to disturbed regulation of atherosclerosis-related genes in hemodialysis patients, which may potentially contribute to their high cardiovascular disease burden [18]. We hypothesized that an interrelationship between renal function, C1 metabolism and subclinical atherosclerosis may exist even among apparently healthy subjects from the general population. For this purpose we analyzed plasma homocysteine, SAH, and SAM in 420 individuals with low cardiovascular risk. 2. Methods 2.1. Study population The I Like HOMe (Inflammation, Lipoprotein Metabolism and Kidney Damage in early atherogenesis e The Homburg Evaluation) study project analyzes the relationship between inflammatory and metabolic risk factors and subclinical atherosclerosis among health-care workers without overt cardiovascular disease. The I Like HOMe study initially comprised 620 subjects, who were recruited between 2004 and 2005. Participants represented a cardiovascular low-risk population, and characteristics of the initial I Like HOMe study cohort have been previously published [19]. Between 2009 and 2011 I Like HOMe participants were re-invited for a follow-up examination (henceforth referred to as ‘I Like HOMe FU’), in which parameters for C1 metabolism, renal function and subclinical atherosclerosis were assessed. 369 participants from the initial I Like HOMe study accepted our invitation, with a further 51 healthy individuals who wished to participate added to the I Like HOMe FU cohort, yielding a total cohort size of 420 subjects. All study participants provided written consent. The study protocol was approved by the local Ethics Committee. A history of smoking, diabetes, current medication intake, cardiovascular comorbidity was obtained via a standardized questionnaire. All participants who had stopped smoking at least one month before study entry were categorized as former smokers. Diabetes was confirmed in subjects with self-reported diabetes mellitus, via a non-fasting plasma glucose 200 mg/dl, a fasting plasma glucose 126 mg/dl, or with current use of hypoglycaemic medication. Body mass index (BMI) was calculated as individual’s body weight divided by the square of their height. Systolic blood pressure (SBP), diastolic blood pressure (DBP), and heart rate were measured after 5 min of inactivity via an automated blood pressure recording apparatus (GE Carescape DINAMAP V100; GE Healthcare). Mean arterial blood pressure was calculated as DBP þ [(SBP DBP)/3], and pulse pressure was calculated as SBP DBP. 2.2. Assessment of intima-media-thickness Intima-media-thickness (IMT) was measured in a supine position with the head turned slightly to the opposite direction. IMT was recorded at the far-wall of the common carotid artery proximal to the bulbus. The Acuson Sequoia C512 Ultrasound Unit (Acuson, Thousand Oaks, CA) equipped with a linear transducer (model 15L8w, Acuson; 8e15 MHz) and automatic measurement software were used.
2.3. Clinical laboratory analysis Blood samples were obtained after overnight fasting. Routine clinical laboratory parameters were measured in the Central Laboratory of the Saarland University Medical Center. Renal function was assessed by estimated glomerular filtration rate according to the CKD-EPIcreatcys equation [20]. 2.4. Analysis of parameters of C1 metabolism For the quantification of S-adenosylmethionine (SAM) and Sadenosylhomocysteine (SAH), EDTA samples were immediately placed on ice and centrifuged within 60 min for 10 min at 2000 g. Before storing the samples at 70 C, 1000 ml of EDTA-plasma was acidified with 100 ml 1 N acetic acid to prevent SAM degradation. The HPLC-MS/MS detection of SAM and SAH concentrations in EDTA-plasma was performed using a Waters 2795 alliance HT, coupled to a Quattro Micro API tandem mass spectrometer (Waters Corporation) according to the method described by Kirsch et al. [21]. Total plasma homocysteine concentrations were determined by using a fluorescence polarization immunoassay on the Abbott AxSym system (Abbott Laboratories, North Chicago, IL). 2.5. Statistical analysis Data management and statistical analysis were performed with SPSS 20.0 (IBM, New York City, NY). Unless indicated otherwise, continuous data are expressed as mean standard deviation and compared by Student’s t-test for independent variables. Categorical variables are presented as percentage of participants. Correlation coefficients were calculated by Pearson test. Subsequently, multivariate linear regression analysis was calculated (method of variable entry: stepwise), which included gender and age (Model 1), traditional cardiovascular risk factors (body mass index, systolic BP, HDL-cholesterol, triglycerides, and blood glucose (Model 2)), and eGFR and albuminuria (Model 3) as independent parameters, and plasma SAH as dependent variable. Two-sided P-values < 0.05 were considered significant. 3. Results Of 420 I Like HOMe FU study participants, 402 were included in the present analysis. The remaining 18 subjects were excluded, as one or more of the following variables were missing: homocysteine, SAH, SAM, albuminuria, cystatin C and/or creatinine. 3.1. Traditional cardiovascular risk factors, GFR and subclinical atherosclerosis Among the 402 I Like HOMe FU participants, 270 were female (67.2%), 107 were current smokers (26.6%), 22 had diabetes mellitus (5.5%), 86 were on antihypertensive medication (21.4%) and 94 had a family history of premature-onset cardiovascular disease (23.4%), respectively. Further baseline characteristics are displayed in Table 1; overall I Like HOMe FU participants represented a low risk study population, and accordingly 337 subjects had a Framingham Risk Score (FRS) below 10% (median FRS: 1%; IQR 1%e5%). Mean (SD) intima-media-thickness among all I Like HOMe FU participants was 0.628 0.127 mm. IMT values showed significant correlation with classical cardiovascular risk factors, as listed in Table S1. Furthermore, IMT values differed between female (0.616 0.118 mm) and male participants (0.652 0.142 mm; p ¼ 0.009), but not between subjects with or without albuminuria (0.640 0.154 mm and 0.627 0.127 mm, respectively; p ¼ 0.719),
A.M. Zawada et al. / Atherosclerosis 234 (2014) 17e22
3.3. C1 parameters and kidney function
Table 1 Baseline characteristics of I Like HOMe FU participants (n ¼ 402). Total cohort (n ¼ 402)
SAH 10.75 SAH > 10.75 p nmol/l (n ¼ 201) nmol/l (n ¼ 201)
Age (years) 48 7 47 7 Weight (kg) 77 17 71 16 Height (cm) 169 9 167 8 Body mass index 27 5 26 5 2 (kg/m ) Waist-to-Hip Ratio 0.85 0.09 0.82 0.08 137 18 133 17 Systolic BP (mmHg) Diastolic BP 89 12 87 10 (mmHg) Mean BP 105 13 103 11 (mmHg) Pulse pressure 48 12 46 12 (mmHg) 101 12 105 10 eGFR (ml/min/1.73 m2) Plasma creatinine 0.83 0.15 0.77 0.10 (mg/dl) Cystatin C (mg/l) 0.74 0.10 0.70 0.09 IMT (mm) 0.628 0.127 0.610 0.122 Total cholesterol 215 37 214 38 (mg/dl) HDL-cholesterol 64 17 67 17 (mg/dl) Triglycerides 114 67 102 51 (mg/dl) Blood glucose 97 16 94 13 (mg/dl) C-reactive protein 2.1 2.3 2.0 2.2 (mg/l)
50 82 172 28
7 17 10 5
19
0.001
