|
|
||||||||
1 LOCUS for Homocysteine and Related Vitamins and
2 Department of Pharmacology, University of Bergen, N-5021 Bergen, Norway.
3 Department of Pharmacology, University of Oxford, Mansfield Rd., Oxford OX1 3QT, United Kingdom.
aAddress correspondence to this author at: Section for Medical Statistics, Department of public Health and Primary Health Care, University of Bergen, Armauer Hansens Hus, N-5021 Bergen, Norway. Fax 47-55-974964; e-mail Lina.El-Khairy{at}smis.uib.no.
| Abstract |
|---|
|
|
|---|
Methods: We measured plasma tCys in blood samples from 12 595 men and women 4042 years of age and from 4766 men and women 6567 years of age, collected as part of the Hordaland Homocysteine Study in the year 19921993. Follow-up data on mortality were collected through 1999. Data on CVD hospitalizations were collected from hospital records up to May 31, 1998.
Results: After a follow-up time of 6.67.6 years, there were a total of 610 deaths, of which 243 were cardiovascular deaths and 367 were noncardiovascular deaths. There was no association between tCys and all-cause, cardiovascular, or noncardiovascular mortality. When we used tCys values <247.6 µmol/L (lowest quartile) as the reference category, the adjusted mortality ratio (MR) for all-cause mortality at tCys concentrations of 247.6270.79, 270.8295.79, and
295.8 µmol/L (highest quartile) were 1.0, 0.9, and 1.0, respectively. The adjusted MRs for cardiovascular mortality were 1.0, 1.1, and 1.1, respectively. There were no associations between tCys and 1275 CVD hospitalizations, except that tCys was significantly associated with hospitalizations from coronary artery bypass grafting.
Conclusion: Plasma tCys is not associated with mortality or CVD hospitalizations.
| Introduction |
|---|
|
|
|---|
Cysteine is another sulfhydryl-containing amino acid that is structurally similar to homocysteine. Furthermore, experimental studies indicate that cysteine could be atherogenic (10)(11). Cysteine supports superoxide modification of LDL and has a cytotoxic effect in vitro against several cell types (12)(13).
Few studies have evaluated the association between plasma total cysteine (tCys) and CVD (14)(15)(16)(17)(18). These studies found higher concentrations of tCys in cases than in controls. In contrast, van den Brandhof et al. (19) found no relationship between tCys and the risk of coronary heart disease. We have previously evaluated the association between tCys and the risk of CVD, using the European Concerted Action Project data. In this case-control study, tCys had a U-shaped relationship with peripheral and cerebrovascular disease and a weak positive relationship with coronary heart disease (20).
Using baseline data from the Hordaland Homocysteine Study, we have previously evaluated the relationship between tCys and several lifestyle and CVD risk factors. The strongest determinants of tCys were age, sex, body mass index (BMI), cholesterol, and diastolic blood pressure (21). In the present work, we investigated the relationship between tCys measured at baseline and the risk of mortality and CVD hospitalizations in this population.
| Participants and Methods |
|---|
|
|
|---|
mortality data
Follow-up data on mortality were collected for the three age groups, but we included only the youngest age group (4042 years) and the oldest age group (6567 years) in this study. Causes of death were coded by Statistics Norway (Oslo) and were obtained from the death certificates of the 610 deaths that had occurred through 1999. Categories were constructed using the underlying cause of death according to the ninth revision of the International Classification of Disease (ICD-9; for deaths from 1992 through 1995) and ICD-10 (for deaths from 1996 through 1999).
hospitalization data
No follow-up data on hospitalizations were collected for the 4364 years group; therefore, these individuals were not included in the study. Data on hospitalizations were obtained from computerized records containing discharge diagnoses from six hospitals in Hordaland County between the time of baseline data collection and May 31, 1998. These records were searched for CVD codes or surgical procedures. Disease categories were constructed according to ICD-9. More details on the data collection have been published previously (25).
statistical methods
The relationship between tCys and mortality/CVD hospitalizations was studied with the Cox proportional hazards model. For the total study population, tCys quartiles were used, and tCys concentrations <247.6 µmol/L (lowest quartile) were used as the reference category.
To have a more balanced distribution of individuals in each tCys category, we used age-specific tCys quartiles in each age group. All associations were adjusted for sex and age and then were further adjusted for CVD risk factors and factors associated with the tCys concentration, including tHcy, BMI, cholesterol, smoking, and diastolic blood pressure.
The analyses were also performed for different combinations of tCys and tHcy. Plasma tCys tertiles were used to define the low, medium, and high tCys concentrations, and tHcy concentrations <12 µmol/L and
12 µmol/L were used to define low and high tHcy. All statistical analyses were performed with SAS software (release 8.2 for Windows).
| Results |
|---|
|
|
|---|
|
There was a significant correlation between tCys and tHcy in this population (r = 0.22; P <0.0001).
TCYS and mortality
After a follow-up time of 6.67.6 years, there were a total of 610 deaths, of which 243 were cardiovascular and 367 were noncardiovascular.
We investigated all-cause mortality, cardiovascular and noncardiovascular mortality, cancer mortality, and noncardiovascular noncancer mortality according to tCys quartiles. All associations were adjusted for sex and age and further adjusted for BMI, cholesterol, tHcy, smoking, and diastolic blood pressure.
In the total population, the crude relationship between tCys and all-cause mortality was strong [mortality ratios (MRs) with 95% confidence intervals (95% CIs) across the tCys quartiles were 1.5 (1.22.2), 2.5 (1.83.3), and 4.8 (3.76.3), respectively; P <0.0001 for trend]. However, adjusting for age abolished this effect. After adjustment, there was no relationship between tCys and all-cause mortality or CVD mortality (Table 2
). When we used the lowest quartile of tCys as the reference, there was no association between tCys and noncardiovascular mortality across the tCys quartiles (P = 0.71 for trend) or cancer mortality (P = 0.38 for trend). However, there was a weak and statistically nonsignificant association between tCys and noncardiovascular noncancer mortality [Adjusted MRs (95% CIs), 1.3 (0.62.5), 1.0 (0.52.0), and 1.8 (0.93.5); P = 0.07 for trend]. All relationships were evaluated in the younger and older groups separately, and similar results were obtained for both groups (Table 2
). All relationships were further evaluated at high tCys concentrations (>95th percentile) compared with lower tCys concentrations, and similar results were obtained (results not shown).
|
The relationship between tCys and specific CVD mortality was evaluated, and there was no association between tCys and cerebrovascular or coronary heart disease mortality (results not shown).
To test whether tCys might be involved in the promotion of existing disease, we evaluated the relationship between tCys and all-cause, CVD, and noncardiovascular mortality in individuals with baseline CVD or hypertension or diabetes and in those free of CVD at baseline. We observed stronger but statistically nonsignificant associations in those with preexisting CVD, whereas there was no association between tCys and CVD risk among those free of CVD at baseline. Furthermore, to evaluate whether the association between tCys and mortality is influenced by the length of follow-up time, we evaluated the associations for those who died during the first 4 years of follow-up but found no association between tCys and mortality (results not shown).
mortality risk at various combinations of TCYS and THCY
We investigated the risk of all-cause and CVD mortality in various combinations of tCys and tHcy concentrations (Table 3
). A combination of low tCys (<255.1 µmol/L) and low tHcy (<12 µmol/L) was used as the reference. There was no association with mortality at low tHcy combined with medium (255.1286.09 µmol/L) or high tCys (
286.1 µmol/L), but there was an association with all-cause and CVD mortality at high tHcy (
12 µmol/L) combined with low, medium, or high tCys. The results indicate a pattern suggestive of higher all-cause and cardiovascular mortality for the combination of high tHcy and low tCys (Table 3
). In addition, we evaluated the association between tCys and mortality separately in those with high tHcy. There was a statistically nonsignificant inverse trend toward lower risk as tCys increased (results not shown).
|
TCYS and cvd hospitalizations
We assessed the relationship between tCys and the risk of hospitalizations for CVD for the whole study population and for the younger and older age groups separately (Table 4
). There was a trend of increasing sex- and age-adjusted hospitalization rate with increasing tCys in the total study population and in the young age group. This weak association disappeared after adjustment for BMI, cholesterol, tHcy, smoking, and diastolic blood pressure.
|
We evaluated the association between tCys and the risk of hospitalization for different disease categories (acute myocardial infarction, coronary heart disease, arterial occlusive disease, cerebrovascular disease, and venous thrombosis) and surgical procedures [percutaneous coronary intervention and coronary artery bypass grafting (CABG)] separately. There were no associations between tCys and specific CVD hospitalizations, but there was a positive association between tCys and hospitalizations for CABG. When we used tCys concentrations <247.6 µmol/L (lowest quartile) as the reference category, the adjusted hospitalization ratio for CABG at tCys concentrations of 247.6270.79, 270.8295.79, and
295.8 µmol/L (highest quartile) were 0.9 (0.32.6), 1.6 (0.64.2); and 2.0 (0.85.1); P = 0.02 for trend. The associations between tCys and the different CVD hospitalizations were slightly strengthened when the analysis was confined to the older age group and to individuals with baseline CVD or hypertension.
| Discussion |
|---|
|
|
|---|
The relationship between tHcy and CVD and non-CVD mortality and CVD morbidity was studied recently in this population (4)(25). The results showed that tHcy is a strong predictor of both CVD and non-CVD mortality and of CVD morbidity in elderly individuals. The median follow-up time for the mortality study was 4.1 years, and only individuals belonging to the oldest age group (6567 years of age at baseline) were included. In the present study, individuals were followed up for >7 years; therefore, the lack of association between tCys and mortality might be attributable to the longer follow-up time. In fact, a few studies have shown that associations between tHcy and risk of disease/mortality are weakened by a longer follow-up time (3)(26). However, when analyses of the tCys-mortality relationship were confined to those who died during the first 4 years of follow-up, similar results were obtained. Therefore, the longer follow-up time could not explain our findings.
To study the interaction between tCys and tHcy in relation to mortality, we investigated the risk of all-cause and CVD mortality in various combinations of tCys and tHcy concentrations. The highest MR was for the combination of low tCys and high tHcy. In agreement with this, using data from the European Concerted Action Project, we have previously found that the highest risk for CVD was for the combination of low tCys and high tHcy (20).
In the present study we found no association between tCys and CVD mortality or CVD hospitalizations. Few case-control studies have evaluated the relationship between tCys and the risk of CVD (14)(15)(17)(19)(20). Two of these studies reported higher tCys concentrations in cardiovascular patients than in controls (14)(15). After adjusting for several factors, including age, sex, smoking status, and creatinine, Jacob et al. (17) found a positive association between tCys and the risk of CVD. However, the authors did not adjust for tHcy, so the possibility remains that tCys merely reflected the association between tHcy and disease. Alternatively, the positive associations between tCys and the risk of CVD in these case-control studies might be attributable to the effect of the disease itself on tCys concentrations because tCys was measured in these patients after the occurrence of disease. A renal mechanism should be considered because there are preliminary data showing increased tCys in renal patients (27).
We have previously found in a case-control study a U-shaped relationship with cerebrovascular and peripheral vascular disease and a weak positive association between tCys and coronary heart disease (20). The lack of association in the present study should carry weight. Our study was prospective and included a large number of events that have been shown to be related to tHcy.
Many case-control and prospective studies have shown that homocysteine is associated with CVD (1)(2). The absence of a relationship between tCys and CVD mortality/morbidity in the present study may be regarded as supportive of the homocysteine theory of CVD (28)(29)(30). It has been suggested that confounding by renal function might explain the tHcy-CVD relationship (31). Renal failure has been shown to cause increased concentrations of both tHcy and tCys (32). In contrast to tCys, tHcy showed a strong association with CVD mortality/morbidity in the same population (4)(25), which suggests that the renal mechanism does not fully explain the relationship between tHcy and CVD because one would then expect an artificial relationship with tCys and later morbidity/mortality. Furthermore, Demuth et al. (33) have recently shown that tHcy is independently and positively associated with carotid artery internal diameter and intima media thickness, both of which are measures of preclinical vascular disease, but that tCys was not. This adds credence to tHcy as an independent cardiovascular risk factor.
Some authors have reported a strong association between tHcy and noncardiovascular noncancer mortality (3)(4). In the present study, tCys was weakly associated with noncardiovascular noncancer mortality. This is a diverse group, with the underlying cause of death ranging from respiratory diseases to accidents. It is therefore difficult to speculate on the nature of the borderline significant association between tCys and mortality in this group.
In conclusion, unlike tHcy, tCys was not associated with all-cause mortality, CVD or non-CVD mortality, and CVD morbidity. These findings indicate that although cysteine shares some of the determinants of tHcy and is metabolically and structurally linked to homocysteine in plasma, tCys is not a predictor of mortality or morbidity in the general population.
| Acknowledgments |
|---|
| Footnotes |
|---|
| References |
|---|
|
|
|---|
T polymorphism, and risk of ischemic stroke: results of a meta-analysis. Neurology 2002;59:529-536.The following articles in journals at HighWire Press have cited this article:
![]() |
Q. Han, M. Xu, L. Tang, X. Sun, N. Zhang, X. Tan, X. Tan, Y. Tan, and R. M. Hoffman Homogeneous Enzymatic Colorimetric Assay for Total Cysteine Clin. Chem., July 1, 2004; 50(7): 1229 - 1231. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |