UNIVERSITY LIBRARI 111111111111111111111111111 11 1111 111111111 3 1293 017103 This is to certify that the dissertation entitled META-ANALYSIS OF TOTAL PLASMA HOMOCYSTEINE AND NUTRITIONAL STATUS OF B—VITAMINS OF HEALTHY AND DISEASED ADULTS presented by PRODROMOS A. PRODROMOU has been accepted towards fulfillment of the requirements for M. 3. degree in HUMAN NUTRITION JMf/ZA/ M2994 pm’lgssor Date February 17, 1998 MSU is an Afflrmatiw Action/Equal Opportunity Institution 0' 12771 LIBRARY Mlchigan State University PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. I DATE DUE DATE DUE DATE DUE MAY 1 3 1993 use mm.mu META-ANALYSIS OF TOTAL PLASMA HOMOCYSTEINE AND NUTRITIONAL STATUS OF B-VITAMINS OF HEALTHY AND DISEASED ADULTS BY Prodromos A. Prodromou A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Food Science and Human Nutrition 1998 ABSTRACT META-ANALYSIS OF TOTAL PLASMA HOMOCYSTEINE AND NUTRITIONAL STATUS OF B-VITAMINS OF HEALTHY AND DISEASED ADULTS By Prodromos A. Prodromou Meta-analysis was performed to determine the relationship between plasma concentration ([Hey]) and nutritional status of vitamins B6, B,2 and folate, to identify the reference value for plasma [Hey] of healthy adults, and to determine the differences in the plasma [Hey] between the healthy controls and diseased eases (CVD and CBVD) by age and gender. Data of 48 independent research studies (representing >30,000 subjects) were used to calculate weighted means, within-group variances and a fixed and random model. The mean [Hey] of cases included in this meta-analysis, 14 umol/l, was lower than the critical value cited by others (16 umol/l). Significant differences in mean (iSD) plasma [Hey] were observed between the cases (13913.9 umol/l) and controls (10.9il .5 umol/l); between healthy men and women (1 1.21:1 .6 vs 9.521: 3.9 umon); and between CVD or CBVD cases (12.6:t2.l vs 16.5: 3.3 umol/l). Plasma [Hey] was inversely associated with blood folate (B=-0.07, P<0.01), plasma vitamin 86 (B=—O.l9, P<0.01) and serum vitamin Bu (B=-o.17, P<0.01), dietary folate (3:031, P<0.01 ), dietary B,,((p=-9.51, P<0.01), but not with dietary vitamin 86 (excluded from the model). Findings on the predictability of plasma Hey values can help public health officials and health educators set up recommendations for the public to reduce the risk of CVD. To my mother, father and brother whom I love very much. ACKNOWLEDGMENTS I would like to thank Dr. Won Song and Dr. Wanda Chenoweth in the Department of Food Science and Human Nutrition for their guidance and support, Dr. Betsy Becker and Kyle Fahrbach in the Department of Counseling Educational Psychology and Special Education for all their help in the statistical analysis of this thesis. Special thanks to my family, labmates and friends who provided me with the mental strength to reach my goals. iv TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES CHAPTER ONE INTRODUCTION Introduction Objectives Hypotheses Significance CHAPTER TWO RELATED LITERATURE Homocysteine metabolism and hyperhomocysteinemia Hyperhomocysteinemia as a risk factor for cardiovascular diseases Prevalence of hyperhomocysteinemia among healthy individuals and in individuals with cardiovascular diseases Gender differences on plasma homocysteine concentration Vitamin supplementation on plasma homocysteine concentration Prevalence of suboptimal blood concentration of vitamins 85, Bu and folate in healthy individuals and individuals with hyperhomocysteinemia Folate, vitamin 86 and vitamin B12: dietary sources, deficiency, toxicity and reported intakes Mechanisms by which homocysteine may promote cardiovascular diseases: Protein C anticoagulant pathway; Heparin anthithrombin activity; Plasminogen activator function; Fibrin binding activity of lipoprotein(a); Tissue coagulation factor stimulation; V .“viii CDde ..14 “.15 ..16 ..17 ..18 ”.19 ..20 ..21 ”.26 ”.26 ”.29 ”.29 ”.30 ..30 Endothelial ADPase activity; Summary Meta analysis: purposes Meta analysis: procedures Problem formulation Data collection Data evaluation Data analysis and interpretation Presentation Strengths and limitations of meta-analysis CHAPTER THREE METHODOLOGY Data collection Coding of research reports Subject variables Study variables Statistical analysis CHAPTER FOUR RESULTS Research Reports and Subjects Differences in mean plasma Hcy concentration between vascular disease cases and healthy controls and between genders Differences in plasma Hcy concentration between men and women Normal range of plasma Hcy concentration of healthy adults and a cut-off point for hyperhomocysteinemia Difference in mean plasma Hcy concentration of CBVD and CVD cases The predictability of plasma Hcy concentration by blood B-vitamin concentration The predictability of plasma Hcy concentration by dietary intake of B-vitamins CHAPTER FIVE DISCUSSION, CONCLUSION Discussion Conclusion CHAPTER SIX ASSUMPTIONS, LIMITATIONS, IMPLICATIONS, RECOMMENDATIONS Assumptions Limitations vi “.31 ”.31 ”.32 “.33 ”.33 ”.34 ”.35 “.35 ”.36 ..36 ”.38 ”.39 ”.41 ”.42 ..42 ..52 ..62 ..64 ..69 ..70 ..73 ..76 ”.78 ..84 ”.85 ..85 Implications ”.86 Recommendations ”.87 APPENDICES Appendix A Coding sheet ”.89 Appendix B Coded variables ”.93 Appendix C Reported correlations of homocysteine and vitamins B6, 812 and folate ".98 Appendix D Reported values for coded variables of eight studies ”.100 BIBLIOGRAPHY “.103 vii Table Table Table Table Table Table Table Table Table Table Table Table Table 10. 11. 12. 13. LIST OF TABLES Classifications and clinical manifestations of hyperhomocysteinemia Risk factors for cardiovascular diseases Symptoms or signs of deficiency and toxicity of vitamins 85, BI; and folate Recommended Dietary Allowances for vitamins 85, B12 and folate Results of Medline search Included and excluded number of articles from the meta-analysis Study designs of the research reports included in the meta-analysis Characteristics and analytical methods of studies included in the meta-analysis Number of studies and subjects used for this meta-analysis Mean plasma Hcy concentration of sub- population groups Fixed effect sizes and homogeneity tests of mean plasma Hcy concentration Mean plasma Hcy concentration of healthy adults and those with vascular diseases Pearson correlation coefficients among plasma concentrations and dietary intakes of B-vitamins viii ..13 ..15 ..22 ..23 ..38 ..40 ..53 ..58 ..61 ..63 ..63 ..69 ..73 Table 14. Table 15. Simple and multiple regression of blood B-vitamin concentration to plasma Hcy concentration Multiple regression output of dietary B-vitamins predicting plasma Hcy ix ..74 ".77 Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure 10. 11. LIST OF FIGURES Pathways of homocysteine metabolism Diet and genetics in the development of cardiovascular diseases Blood clotting cascade in humans Weighted mean plasma Hcy concentration of cases and controls by age Weighted mean plasma Hcy concentration of controls by gender and age Weighted mean plasma Hcy concentration of cases by gender and age Weighted mean plasma Hcy concentration of CBVD & CVD by age Weighted mean plasma Hcy concentration by gender and disease Scatterplot of plasma Hcy concentration and plasma vitamin B; concentration Scatterplot of plasma Hcy concentration and serum vitamin Bu concentration Scatterplot of plasma Hcy concentration and blood folate concentration ..10 ..12 ..28 ..65 ..67 ..68 ..71 ..72 ..75 ”75 ”76 Chapter One INTRODUCTION Introduction Elevated plasma concentration of homocysteine (Hcy), a sulfur-containing amino acid, was initially reported in 1962 as a disease of genetic origin (Carlson & Neill, 1962). Inborn errors of the enzyme cystathionine b-synthetase (CBS) and/or methyltetrahydrofolate reductase (MTHFR), which are involved in Hcy metabolism, result in elevated Hcy concentration in plasma (hyperhomocysteinemia) and/or excretion in urine (homocysteinuria). Homozygous and heterozygous enzyme deficiency in humans results in high plasma concentration of Hcy, causing neurological abnormalities, mental retardation, fatal thrombosis and cardiovascular diseases (CVD) (Carlson & Neill, 1962). Thirty five years after the first reported incidence, hyperhomocysteinemia is now recognized as an independent risk factor for CVD (Taylor et al, 1991; Stampfer et al, 1992; Verhoef et al, 1994, 1996; Perry et al, 1995; Arnesen et al, 1995). Hyperhomocysteinemia is no longer addressed as a disorder of solely genetic origin (Selhub & Miller, 1992; 2 Jacques et al, 1996; Dudman et al, 1996). Hyperhomocysteinemia of nutritional origin, due to suboptimal blood concentrations of vitamins Be, Bugand folate, has in recent years received public attentions (Ubbink et al, 1993; Guttormsen et al, 1996; Selhub et al, 1996). Suboptimal plasma concentrations of vitamins Be, Bu and folate have been reported to cause a reduction in enzymaticactivity of CBS or MTHFR, and an elevated Hcy concentration in plasma by Dudman et al in 1996. Inadequate nutritional status of vitamins Beth; and folate with or without genetic defects have been reported to elevate plasma Hcy concentration (Jacques et al, 1996; Dudman et al, 1996; Guttormsen et al, 1996). The prevalence of hyperhomocysteinemia reported in the U.S vary widely among studies. It ranges from 20-30% in the elderly of 67-96 years of age to 40% in those of over 80 years of age (Selhub et al, 1996). Prevalence of suboptimal plasma concentration of the vitamins 85 (<30 nmol/l), Bug (<200 pmol/l) or folate (<5 nmol/l) was 67% in the elderly of 67-96 years of age (Selhub et al, 1993). Ubbink et al (1993) also reported that the prevalence of suboptimal plasma vitamins Bs(<30 nmol/l), Bu:(<200 pmol/l), and folate (<5 nmol/l) in men with hyperhomocysteinemia is 25%, 56% and 59%, respectively. 3 Most previous studies support the cause and effect relationship between suboptimal nutritional status of the three vitamins and elevated plasma Hcy concentrations (Selhub et al, 1993; Van den Berg et al, 1994; Pancharuniti et al, 1994; Verhoef et al, 1996). Many investigators (Pancharuniti et al, 1994; Van de Berg et al, 1994; Verhoef et al, 1996) reported significant inverse correlations between plasma Hcy concentration and plasma concentration of folate and vitamin Bu in both diseased with CVD (cases) and healthy individuals (controls). In contrast to these findings, Dalery et al (1995) reported no significant differences in blood vitamins Be, Bugand folate in 150 CAD cases, when compared to 584 healthy controls (age<60 years). Gender differences in plasma Hcy concentration of healthy people are also reported by different investigators. In 1992 Andersson et al reported lower plasma Hey concentration in 83 healthy women (age: 20-69 years) compared to 74 healthy men of similar age (9.6:3.0 vs 10.7:2.6 umol Hcy/l plasma). Cacan et al (1996) also reported lower plasma Hcy concentration in healthy women (age: 23-59 years) compared to healthy men (7.6:4.1 vs 9.714.9 umol Hcy/l plasma) of similar age. In relation to gender differences in CVD cases, Robinson et al (1995) reported higher plasma Hcy concentration in 103 female cases 4 when compared to 201 male cases (15.315.7 vs 13.9:4.5 umol Hcy/l plasma). Investigators have classified hyperhomocysteinemia by different criteria. In a cross sectional study of the elderly in the Framingham study, Selhub et al (1993) defined hyperhomocysteinemia as (>14.0 umol Hcy/l plasma), which is the cut-off point for plasma Hcy concentration of the 90th percentile in subjects of both genders (n=1160). The subjects had suboptimal plasma concentrations of three B- vitamins as defined by <70w‘percentiles: plasma pyridoxal- 5-phosphate<77.7 nmol/l, vitamin B12<389 pmol/l and folate<14.8 nmol/l. In 1995, Robinson et al reported (>13.5 umol Hcy/l plasma) as hyperhomocysteinemia based on above the 80“‘percentile cutoff point in male patients (n=304) with established coronary artery disease. Hopkins et a1 (1995) suggested >9 umol Hcy/l plasma as hyperhomocysteinemia that correspond to a progressive increase in CVD risk among men (n=162). In assessing the implications of this cause and effect relationship between nutrition and health in the public health arena, consistent definitions and reference values of the parameters are important. Health professionals do not know if the current Recommended Dietary Allowances (RDA) for vitamins 85, BR and folate are adequate to minimize the risk of 5 hyperhomocysteinemia. Information currently available on the relationships between hyperhomocysteinemia and dietary intakes of the three vitamins is scarce (Verhoef et al, 1995; Selhub et al, 1996). Individual studies (Andersson et al, 1992; Selhub et al, 1993; Nygard et al, 1995; Cacan et al, 1996; Riggs et al, 1996) identify age, disease, gender and nutritional status as possible risk factors for hyperhomocysteinemia. Findings of these studies are, however, inconsistent party due to sample variations. Individual studies identify different vitamins as the single most important nutrient in maintaining normal plasma Hcy concentration. Reference values used in determining hyperhomocysteinemia vary among the studies. No investigation has been made if hyperhomocysteinemia poses the same risk for CVD and cerebrovascular disease (CBVD). Findings of such an investigation will help to increase the public awareness for better health, and increase our knowledge to reduce the risk for CVD. I Meta-analysis offers the strength to combine systematically and quantitatively the previously reported research data on Hcy to draw a conclusion on nutritional status as a risk for hyperhomocysteinemia (Hedges & Olkin, 1985). Since many prior individual studies on Hcy included sample sizes too small to draw firm conclusions to apply in 6 the public health arena, the meta-analysis approach is proposed (Cooper & Hedges, 1994; Petitti, 1994; Hedges & Olkin, 1985). There has been only a single reported study of meta-analysis on homocysteine (Boushey et al, 1995). This meta—analysis determined the risk of hyperhomocysteinemia for arteriosclerotic vascular disease, estimated the potential reduction of CVD by increasing dietary or plasma folic acid. The present study is designed to address the nutritional effect of vitamins Ba Bug and folate on hyperhomocysteinemia. Identification of modifiable factors for hyperhomocysteinemia, a risk factor for CVD, is important in public health arena. Objectives of this study were: 1. To confirm that individuals with vascular diseases (cases) have a higher mean plasma Hcy concentration than healthy individuals (controls). 2. To determine if plasma Hcy concentration of men is higher than women for both vascular disease cases and healthy controls. 3. To establish reference values for plasma Hcy concentration of healthy controls and vascular disease cases. 4. To determine if plasma Hcy concentration differs between .7 cerebrovascular and cardiovascular disease cases. To estimate the predictability of plasma Hcy concentration by plasma vitamin Be, serum vitamin Bu and/or blood folate concentration. To estimate the predictability of plasma Hcy concentration by dietary intake of vitamin Be, vitamin Bu and/or folate. Hypotheses of the study were that: 1. 2. 3. 4. 5. Vascular disease cases have a higher mean plasma Hcy concentration than healthy controls. Men have a higher mean plasma Hcy concentration than women in both groups of vascular disease cases and healthy controls. Reference values are lower than those reported. Mean plasma Hcy concentration of healthy controls and vascular disease cases differ. Plasma Hcy concentration can be predicted by plasma concentration of vitamin Ba, vitamin Bu and/or folate. Plasma Hcy concentration can be predicted by dietary intake of vitamin B5, vitamin Bu and/or folate. Significance Cardiovascular diseases are currently at epidemic levels worldwide in all developed and developing countries (McCully, 1997). The emerging information reveals the necessity of maintaining nutritional adequacy for disease prevention. However, the amount of dietary vitamins 85, BR and folate that are needed to prevent hyperhomocysteinemia of nutritional origin in healthy adults is not known. Plasma Hcy concentration reference values for healthy adults and CVD patients vary among studies. Plasma levels of selected B-vitamins sufficient to maintain normal plasma Hcy levels have not been adequately addressed. Information on dietary and plasma levels of the B-vitamins in relation to Hcy is limited due to small numbers of studies and small sample sizes, which lead to limited generalizability of the findings to the whole population. Identification of unmodifiable risk factors (i.e. gender and age) and modifiable risk factors (i.e. nutritional status) for hyperhomocysteinemia is important. Findings on the predictability of plasma Hcy values based on plasma levels and dietary intake of vitamins B6, B12 and folate can help public health officials and health educators set up recommendations for the public to reduce the risk of CVD. Chapter Two RELATED LITERATURE Homocysteine metabolism.and.hyperhonocystein¢nia Homocysteine (Hcy) is an amino acid (HSCHZCHZCH(NH2)COOH) produced by demethylation of methionine, as an intermediate in the biosynthesis of cysteine from methionine via cystathionine in humans (Figure 1) (Dudman et al, 1996). Hcy is present in blood in two forms: free or protein-bound. The sum of the two forms represents total Hcy. Under normal physiological conditions more than 50% of the total Hey is bound to the protein. When plasma Hcy concentration is elevated percentage of free form in the blood is increased. In the biosynthesis of Hcy, methionine, an essential amino acid and a precursor of Hcy is enzymatically methylated to s-adenosylmethionine through the action of methionine adenosyltransferase. S-adenosylmethionine, the first metabolite of methionine, is then demethylated by transferases to s-adenosylhomocysteine. Enzymatic hydrolysis of s-adenosylhomocysteine by adenosylhomocysteine hydrolase 10 _ HEB—.30 :oooamzvmoflo-mm 1 Emfiaonmuoe ocfimumxooEon mo m>m32umm .H muomflm GE $555255 _ oaaeoi =2.er , . 50956:an omuaomnafiao 220552 52.55 _ /. :oooamzcmoflofiow Acme Sesame—E ~85on 380593.: . Quay «a» :8 E59: atom a. o - ”252. _ HEB—Em: _ oaaeaaseflsosa _HEEm>OoZo= _ mooRNmémommoEomflo A moooxfizvmofloamoéi ogsgfiazmoqaargaz > omfioumfib—zmoeooas own—26>: 055282 ofiogoofioammooopfim mEzoEeEaflmozm—QA _ Emegoozomimozmga mooREE:camoaoamovaiaofig v EOOOANEEUEUNIO-m-3mo=op< 11 yields the metabolite homocysteine, which can be metabolized via the irreversible trans-sulphuration pathway to cysteine. Major pathways of Hcy metabolism include first the remethylation of Hcy back to methionine. This remethylation reaction is catalyzed by methionine synthetase and 5— methyltetrahydrofolate reductase (MTHFR). MTHFR functions as the methyl group donor and requires folate as cofactor. The enzymatic conversion of Hcy to cystathionine is catalyzed by cystathionine b-synthetase, a vitamin B&.dependent enzyme. Final conversion of cystathionine to cysteine is by the action of cystathionase, a vitamin Badependent enzyme (Dudman et al, 1996). Genetic defects causing an enzymatic deficiency of cystathionine b-synthetase or MTHFR reductase result in an accumulation of Hcy in the blood, leading to intermediate or severe hyperhomocysteinemia as defined by >16 umol/l (Dudman et al, 1996). Current research also support that elevated plasma Hcy concentration may also be the result of nutritional inadequacies of vitamins 85, Bugand/or folate (Ubbink et al, 1993) (Figure 2). Remethylation of Hcy to methionine is catalyzed by MTHFR, a folate and vitamin Bu dependent enzyme, (Dudman et al, 1996). Conversion of Hcy to cystathionine, and then to cysteine, is also a vitamin Ba- dependent reactions. Cystathionine b-synthetase and 12 mommomflc unasomm>ofleumo mo ucoanHe>mc ecu CH weapocoo cam ucflo .N ousofim EEG 1 1 §820§ma_ > _ mEzoEE: :1 $520593 E395 .— mo 3ng «Swim no mean»: >555 Homamo ‘ 1 05.93 1 1 E52 Meme/mm“ q mzfifiS at. —. e «E .8 méeS .— mo 3ng Szmfia .8 82.9: >555 13 cystathionase are both enzymes which require vitamin 86 as a coenzyme. Hyperhomocysteinemia has been, for epidemiological purposes, defined by Kang et al (1996), as plasma Hcy concentration >16 umol/l whereas plasma Hcy concentration of healthy subjects as $16 pmol/l . When plasma Hcy concentration rises above 100 umol/l, Hcy is also found in the urine (homocysteinuria) (Carlson and Neil, 1962). Hyperhomocysteinemia is further classified in the literature as moderate, intermediate or severe, depending on plasma Hcy concentration (Table 1). Table 1. Classification and clinical manifestations of hyperhomocysteinemia Chufiwnmn IHummIky Chmahmmmwndmm Emmng (umol/l) Severe >100 "' neurological abnormalities homozygous and ‘ mental retardation heterozygous enzyme ‘ carotid thickening deficiencies Intermediate 31-100 ‘ premature cerebrovascular nutritional inadequacy peripheral and coronary with or without amaymmmw gummnkkm ‘ carotid thickening " thromboembolism Moderate 16-30 ‘ premature cerebrovascular nutritional inadequacy peripheral and coronary with or without artery disease genetic defect Normal <16 References: Kang S., 1996. 14 Hyperhomocysteinemia as a risk factor for cardiovascular diseases Hyperhomocysteinemia was identified initially as a risk factor for cardiovascular diseases (CVD) through case series, cross-sectional and case control studies (Swift et al, 1986; Olszewski et a1, 1991; Brattstrom et al, 1990; Clarke et al, 1991; Ubbink et a1, 1993; Pancharuniti et al, 1994; Fermo et al, 1995). These studies consistently suggest a strong positive relationship (Odds Ratio (OR), 95% CI = 1.0-6.7) between elevated plasma Hcy concentration and the risk for CVD as measured by the degree of stenosis of the coronary artery (Swift et al, 1986; Olszewski et al, 1991; Brattstrom et al, 1990; Clarke et al, 1991; Miller et al, 1992; Ubbink et al, 1993; Pancharuniti et al, 1994: Selhub et al, 1995; Fermo et al, 1995). To date, a limited number of prospective studies on plasma Hcy and CVD have been reported (Taylor et al, 1991; Stampfer et al, 1992; Verhoef et al, 1994; Perry et al, 1995; Verhoef et al, 1995; Arnesen et al, 1995). These studies involved large cohorts to verify hyperhomocysteinemia as independent risk factor for CVD (Table 2)(Taylor et al, 1991; Stampfer et al, 1992; Verhoef et a1, 1994; Perry et a1, 1995; Verhoef et al, 1995; Arnesen et al, 1995). 15 Table 2. Risk factors for cardiovascular diseases Dietary cholesterol " 7" 3 Plasma cholesterol 2' Dietary protein ' Dietary Fat ' Plasma Homocysteine " 2' 3 Agel Hypmumflmn' Snummmfl Oral contraceptives ‘ Sex| Alcoholism ' Snam' Emdrke' IdeMm' Boiyhkmshflex' References: (1) Gruberg et a1, 1981, (2) Ubbink et al, 1996, (”Pancharunti et a1, 1994. Prevalence of hyperhomocysteinemia among healthy individuals In 1992, Anderson et al reported that 12 of 169 apparently healthy individuals in the city of Malmo, Sweden aged 20-69 had hyperhomocysteinemia (>16 umol/l). Selhub et al (1994) also reported from the 20““biannual examination of the Framingham Heart Study (1989-90) that hyperhomocysteinemia (>90m‘percentile; plasma Hcy concentration >14.0 umol/l) acounted for 30% of the entire cohort and over 40% of individuals aged 80 years and older (n=418 men and 623 women). l6 Hyperhomocysteinemia has been reported in people of all ages. Genetic defects and/or nutritional inadequacy of B- vitamins have been used in explaining hyperhomocysteinemia. Prevalence of hyperhomocysteinemia in individuals with cardiovascular diseases In 1991, Clarke et al reported that hyperhomocysteinemia (>16 umol/l) was detected in 42% of 38 cerebrovascular disease patients, 28% of 25 peripheral vascular disease patients. After adjustments for the effects of conventional risk factors (i.e. hyperlipidemia, hypertension and cigarette smoking), patients with hyperhomocysteinemia had 3.2 OR for CVD. In 1992, Brattstrom et al reported that 40% of 142 stroke survivors and 6% of 66 control subjects had hyperhomocysteinemia. In 1995, Fermo et al also reported that moderate hyperhomocysteinemia (19.5-99.9 umol Hcy/l plasma for male; 15.0-99.9 umol Hcy/l plasma for female based on >95u‘percentile) was seen in 13.1% and 19.2% of patients with venous and arterial occlusive respectively. Dalery et al (1995) also reported a higher mean plasma Hcy concentration in coronary artery disease (CAD) patients (males and females) than healthy controls (men with CAD: 11.7i5.8 vs. controls: 9.7i4.9 nmol/ml; & women with CAD: 12.016.3 vs. controls: 7.6i4.1 nmol/ml, P<0.01 between cases and controls). The prOportion of patients with CAD having Hcy levels >90“‘percentile of 17 controls (>14.5 umol/l) was 18.1% for men and 44.4% for women (both P<0.01 compared to the controls). Reference range of plasma Hcy concentration for both healthy and CVD or CAD patients vary among studies. It ranges from 7.6 to 16.3 umol/l for healthy people, and from 13.5 umol/l to 33.3 umol/l for CVD or CAD patients. Many studies classified hyperhomocysteinemia as the >90th percentile of plasma Hcy concentration of healthy controls. With pooled data of numerous studies including a large number of population, reference values are hoped to be established in the proposed study. Gender differences on plasma homocysteine concentration In 1994, Jacobsen et al reported that 36 blood donors of apparently healthy men (XiSD: 34.4 i 9.4 yrs) had a significantly higher range of plasma Hcy concentration (9.26-12.30 umol/l) than 35 female blood donors (33.8 i 6.2 yrs; 7.85-10.34 umol/l). Brattstrom et al (1994) also reported that plasma Hcy concentration of 131 men was higher than that of 113 women with age ranging 35 to 95 years of age (meaniSD: 13.914.1 and 12.314.1 umol/l respectively; P<0.001) and increased markedly with age (r= 0.488; P<0.001). Gender differences in plasma Hcy concentration are not well understood biologically although most literature 18 supports the apparent gender differences. Gender needs to be more carefully examined in relation to the difference in plasma Hcy concentration in healthy and diseased stage. Vitamin supplementation on plasma homocysteine concentration Several studies reported effectiveness of megadose vitamin supplements in reducing plasma Hcy concentration. The doses used were several times higher than the RDA of vitamin B6(30 mg), vitamin By; (25 ug) and folate (800 ug). In 1990, Brattstrom et al reported that daily supplementation of 240 mg pyridoxine hydrochloride and 10 mg folate for four weeks in 20 patients (heterozygotes for cystathionine b-synthatase and with moderate hyperhomocysteinemia) resulted in a reduction of 53% in the mean plasma Hcy concentration (baseline: 23.1i19.2 umol/l; after supplementation: 10.917.2 umol/l). Naurath et al (1995) also reported a prospective double-blind controlled study in which intramuscular vitamin injections were given daily to the elderly eight times over a three-week period. The vitamin injections contained 1 mg vitamin Bu, 1 mg folate, and 5 mg vitamin B6,The elderly subjects in the study live at home (n=175; 65-96 years of age). The elderly who received the treatment (n=110), experienced 8% reduction in plasma Hcy concentration while those who received the placebo had no effect (Naurath et al, 1995). Ubbink et a1 (1995) also reported a reduction in plasma Hcy concentration 19 in white (n=18) and black (n=18) ranging from 19 to 24 years) when they were given 6 weeks daily oral vitamin supplementation of 1.0 mg folic acid, 400 ug vitamin BR and 10 mg vitamin Ba. The participants were from the annual recruitment of new police at Pretoria’s Police College. After the daily vitamin supplementation, fasting plasma Hcy concentration was reduced significantly from 9.6 i 3.5 to 7.2 i 1.6 umol/l in whites (P<0.05) and from 8.4 i 2.4 to 5.6 i 1.4 umol/l in blacks (P<0.01). The study of Ubbink et al (1995) also acknowledged the differences in plasma Hcy concentration between blacks and whites. Supplementation of vitamin B5, Ba and folate caused a significant reduction in mean plasma Hcy concentration in apparently healthy individuals. A greater reduction (53%) was achieved in individuals with hyperhomocysteinemia (Brattstorm et al, 1990). Thus nutritional adequacy of the B-vitamins may play a significant role in maintaining normal levels of plasma Hcy concentration. Prevalence of suboptimal blood concentration of vitamins 85, 8;; and folate in healthy individuals In 1992, Pennypacker et al reported that of 152 geriatric outpatients (65-99 yrs old) attending the Veteran’s Administration Medical Center Geriatrics Clinic in Denver, Colorado 14.5% had vitamin Buzdeficiency as diagnosed by S 300 pg/ml serum. Yao et al (1992) reported 20 that 16% of the elderly (N=l69) receiving primary care had vitamin Bu deficiency with < 200 pg/ml, and 21% had 201—299 pg/ml. In 1993, Joosten et al also reported that the free living elderly (>65 yrs old; n=145) and hospitalized elderly (>65 yrs old; n=135) had low serum concentration of the vitamin Bn:(6% and 5%, respectively), of folate (5% and 19%, respectively) and of vitamin B5(9% and 51%, respectively). Suboptimal plasma concentration of vitamins B5, Bu and folate are very common in the elderly. If increasing prevalence of hyperhomocysteinemia with age is related to the high prevalance of vitamin deficiencies in the subgroup should be further investigated. Prevalence of suboptimal blood concentration of vitamins 35, B12 and folate in individuals with hyperhomocysteinemia Ubbink et a1 (1993) reported that the prevalence of suboptimal concentration of plasma vitamin B5(<14 nmol/l), serum Bn3(<150 pmol/l) and plasma folate (<5 nmol/l) status in people with moderate hyperhomocysteinemia (>16.3 umol/l, n=44) was 25%, 56% and 59.1% respectively. In 1996, Selhub et al also reported these suboptimal plasma concentrations of one or more B-vitamins (85(12 nmol/l;IBu<140 pmol/l; folate<4 nmol/l) were detected in 67% of hyperhomocysteinemic subjects (216.3 umol/l plasma Hcy) included in the Framingham Heart Study. 21 The prevalence of suboptimal blood concentration of the B-vitamins in hyperhomocysteinemic patients is very high (25%-67%). Among individuals with hyperhomocysteinemia, deficiency of folate and BL; is higher than that of vitamin Ba. Folate, vitamin B; and vitamin B12 : sources, deficiency, toxicity and reported intakes Folate is a water-soluble vitamin found primarily in liver, yeast, leafy vegetables, legumes and some fruits (RDA, 1989; Combs, 1992). Folate was discovered in early 1930’s in India, where it was found as a cure for macrocytic anemia. A few years later, the vitamin was named folic acid, a term derived from the Latin word folium meaning leaf. Folate is unstable in heat during cooking (RDA, 1989; Combs, 1992). Serum folate concentration decreases when dietary intake of folate is inadequate. A serum folate concentration of <3 ng/ml indicates folate deficiency. Serum folate concentration, however, does not reflect depletion of body stores (Combs, 1992; Friedrich, 1988). Tissue depletion of folate is reflected by $140 ng/ml erythrocyte folate concentration. Tissue depletion of folate is also diagnosed by neutrophil hypersegmentation, a change in the morphology of the peripheral white blood cells. Hypersegmentation is caused by impaired DNA synthesis in white blood cells. The 22 end stage of folate deficiency is characterized by megaloblastic anemia, which is similar to vitamin Bu deficiency (Combs, 1992; Friedrich, 1988) (Table 3). Table 3. Symptoms or signs of deficiency and toxicity of vitamins Ba, BIZ and folate. Vitamin B5 Vitamin Bl2 F olate meme " peripheral neuropathies " macrocytie megaloblastic anemia " megaloblastic anemia 'dumnms ‘mmmnmhwmqumks ‘gmaflwmmmms " anemia “ memory loss " depression ‘ polyneuropathy " neural tube defects ‘pumdmmmmmm Toxiei : ‘emwmmmu ‘ummmmm ‘qflqmemmmmws ’smmmynmmmmmy (hmx) (mm) ‘amma ‘mmdhmmnqmy ‘ loss of small motor control (rats) Combs, F .G. The vitamins: Fundamental aspects in Nutrition. Academic Press; 1992. In a folate and vitamin andeficiency, the red blood cells are large in size (macrocytic) while concentration of hemoglobin remains normal (normochromic). In contrast, iron deficiency leads to small sized red blood cells (microcytic) with low concentrations of hemoglobin (hypochromic) (Combs, 1992; Friedrich, 1988). The RDA for folate is 200 ug for adult men, 180 pg for women, and 400 ug for pregnant women. For infants from birth to one year, the RDA for folate is set at 3.6 ug/kg body weight (Table 4). 23 Table 4. Recommended Dietary Allowances for vitamins 8,, B,2 and folate Vitamin B,5 Vitamin B,2 F olate (mg) (113) (pg) Males (15+ yrs) 2.0 2.0 200 Females (15+ yrs) 1.6 2.0 180 Pregnancy 2.2 2.2 400 Lactation 2.1 2.6 280 Source: Food and Nutrition Board. Recommended Dietary Allowances, 10‘” Ed. Washington, DC: National Academy Press;l989. Vitamin 86 is also a water soluble vitamin, found in meats, whole-grain products, vegetables, and nuts. It exists in various chemical forms in food (pyridoxine, pyridoxal, pyridoxamine). The various forms of the vitamin are converted in liver, in erythrocytes and in other tissues into pyridoxal phosphate (PLP) and pyridoxamine phosphate (PMP), which are the active forms of the vitamin (Combs, 1992; Friedrich, 1988; Bailey, 1990). PLP, serves as a coenzyme of transaminases and decarboxylases which are involved in the metabolism of amino acids; as a coenzyme for phosphorylases; in the biosynthesis of the neurotransmitters serotonin, epinephrine, and norepinephrine; coenzyme of glycogen phosphorylase (Combs, 1992; Friedrich, 1988; Bailey, 1990), and as a modulator of steroid hormone receptors. Severe deficiency of vitamin Be results in dermatologic and neurologic abnormalities (peripheral 24 neuropathies), weakness, cheilosis, glositis, and impaired cell-mediated immunity. The toxicity of vitamin B5 occurs rarely in humans. Massive doses (10XRDA) of the vitamin have produced convulsions in rats (Combs, 1992; Friedrich, 1988; Bailey, 1990). The requirement for vitamin Be varies depending on type of diet, health and some other factors. High intake of protein increases the dietary requirement for vitamin BS. (Combs, 1992; Friedrich, 1988; Bailey, 1990). Human RDA of vitamin B; is established based on the ratio 0.016 mg pyridoxine per gram protein intake(RDA,1989; Combs, 1992). The ratio is to prevent or eliminate the appearance of biochemical indicators of deficiency when daily protein intakes range from 54 to 165 g (RDA, 1989; Combs, 1992). The RDA for vitamin Bewas then established based on twice the RDA for protein (126 g/day for men, 100 g/day for women): 2.0 mg/day for men and 1.6 mg/day for women (RDA, 1989; Combs, 1992) (Table 4). Vitamin Bu:is synthesized exclusively by bacteria, and is found only in foods that have been bacterially fermented or derived from tissues of animals (Glusker, 1995; Combs, 1992; Friedrich, 1988; Bailey, 1990). Animal tissues that accumulate vitamin Bu (e.g. liver) are, therefore excellent food sources of the vitamin for humans. 25 Vitamin Bu is involved in the maintenance of nervous tissue, normal blood formation, and general growth. Biochemical functions of the vitamin include conversion of methyl malonyl-CoA to succinyl CoA, intracellular synthesis of polyglutamate forms of folate, methylation of homocysteine to methionine and in nucleic acid metabolism (Glusker, 1995; Combs, 1992; Friedrich, 1988; Bailey, 1990). Vitamin Buzdeficiency in humans causes delay of normal cell division, megaloblastic anemia, neurological abnormalities (neurological lesions), memory loss, and progressive nerve demyelination as part of a progressive neuropathy (Glusker, 1995; Combs, 1992; Bailey, 1990) (Table 3). The RDA for vitamin Bu is set at 2ug/day for adults, a value that is expected to prevent any signs of deficiency (Glusker, 1995; RDA, 1989; Combs, 1992) (Table 4). During the last 35 years vitamins Ba, Bu and folate have been recognized to play a significant role in Hcy metabolism. Nutritional deficiency of any one of the vitamins may lead to an elevation of the amino acid, in plasma, which is risk factor for thrombotic disorders and CVD. Nutritional adequacy of the three vitamins may be important in maintaining plasma Hcy concentration and to reduce the risk for CVD. 26 Mechanisms by which Hcy may promote cardiovascular diseases Hcy may promote cardiovascular diseases by any or combination of the following proposed mechanisms: a. Protein C anticoagulant pathway The blood clotting factors IX, X and prothrombin are synthesized in the liver in a process that requires vitamin K. Factor XI is activated via a process that requires calcium. Factor IX is then activated by the presence of the activated factor XI (XIa) and calcium (Dittman and Majerus, 1990; Rosendal et a1, 1995). Activated factor IX (IXa), and calciumtare required for the activation of factor X. During the final steps of the blood clotting cascade in humans, the coagulation factor prothrombin is converted to thrombin on the phospholipid membrane, in the presence of activated factor Xa, Va, VIIIa and calcium. Thrombin catalyzes the conversion of fibrinogen to soft clot fibrin. The soft clot fibrin is fragile and rapidly converted to a more stable hard clot in a reaction catalyzed by XIIIa (Dittman and Majerus, 1990; Rosendaal et al, 1995). Individuals deficient in various clotting factors have a pronounced tendency to bleed. A blood clot consists of arrays of cross-linked fibrin that forms an insoluble fibrous network. 27 Protein C is a vitamin K-dependent plasma protein which inhibits blood coagulation by enzymatic cleavage of the factors Va and VIIIa, and thus interferes with the regulation of intravascular clot formation (Dittman and Majerus, 1990; Rosendaal et al, 1995) (Figure 3). Thrombin is a proteinase that converts fibrinogen into fibrin, causing blood coagulation, by hydrolyzing peptides. Thrombin is bound to a platelet and endothelial cell membrane receptor called thrombomodulin. Thrombomodulin is an anticoagulant glycoprotein that serves as a cofactor for the activation of protein C. Binding of thrombin to thrombomodulin activates protein C. Activated protein C cleaves and thus inactivates factors Va and VIIIa (Figure 3)(Dittman and Majerus, 1990; Rosendaal et a1, 1995). By doing so, protein C causes an inhibition of the coagulant activity of the factors and helps maintaining blood fluidity (Svensson and Dahlback, 1994). Protein C deficient individuals often die in infancy of massive thrombotic complications. Homocysteine interferes with protein C by two possible mechanisms. First, homocysteine prevents thrombomodulin transport to the cell surface (Figure 3) (Lentz and Sadler, 1991). This prevents the formation of thrombin- thrombomodulin complex and causes a reduction in protein C activation, less inhibition of factors Va and VIIIa and 28 mamas: ea memOmmo meauuoao noose .m mucosa 9% Es area EEHE s A «Exaoaomm A EX 86mm 020 acmv :: _ A Ewes: _. . Pm a two . 5 3.353218 asses... . e .15 assessor SmoEEmmE A \ 88m ax in V m ; 95338::3 1r . A U £28m 83289: «:3 a> SQ 2X IwU 5 88mm eoaméoa qumEOo Xmmmzoo 5.6088805 + £2885. 25 86am + E 83am 29 promotion of blood coagulation. In the second mechanism homocysteine interferes the blood coagulation pathway by reducing the disulfide bonds between thrombomodulin and protein C. This decreases protein C activation and increases blood coagulation with thrombogenic events (Svensson et al, 1993; Rosendaal et a1, 1995). b. Heparin antithrombin activity There are numerous physiological mechanisms that limit clot formation. Antithrombin reduces all active clotting factors by binding to them. The presence of heparin enhances the activity of antithrombin. Heparin is released by injury, activates antithrombin and thereby prevents clot formation (Rosenberg and Rosenberg, 1984; Nishinaga et al, 1993). Heparin is the most frequently used anticoagulant administered before and after surgery to retard clot formation. Hcy causes a reduction in the binding affinity of endothelial cells for antithrombin. By doing so, less antithrombin binds to the endothelial cells and this decreases the anticoagulant properties of the factors (Rosenberg and Rosenberg, 1984; Nishinaga et al, 1993). c. Plasminogen activator function 30 Blood clots (fibrin) are normally eliminated as wound repairs progress. Fibrin is dissolved by plasmin, an enzyme which cleaves fibrin (Figure 3). Plasmin is formed from precursor plasminogen activator. Plasminogen activators have received considerable medical attention since they rapidly dissolve the blood clots responsible for heart attacks and strokes. Homocysteine causes a reduction of the cellular binding sites for the plasminogen activator (Hajjar et al, 1987; Hajjar et al, 1993). This causes accumulation of fibrin, reduces blood fluidity and increases the risk for thrombosis (Hajjar et al, 1987; Hajjar et al, 1993). d. Fflbrin binding activity of lipoprotein(a) Lipoprotein(a) is an atherogenic lipoprotein which is composed of low density lipoprotein (LDL) particles. Lipoprotein(a) which has a very similar structure to plasminogen, competes for the plasminogen activator. This causes a reduction in plasmin formation and promotes thrombotic events (Harpel et al, 1992). Homocysteine enhance the binding of Lipoprotein(a) to fibrin and promotes the thrombogenic events. e. Tissue coagulation factor stimulation 31 Tissue coagulation factor (Factor III) is a transmembrane glycoprotein that plays a major role in the initiation of blood coagulation. Cell surface factor III forms a complex with factor VIIa (Figure 3). This complex increases the activation of factors X, IX, which in the presence of VIIIa and Va convert prothrombin to thrombin. Plasma Hcy increases the factor III activity of endothelial cells resulting in increase of all other coagulation factors (Nemerson et al, 1992; Fryer et a1, 1993). f. Endothelial ADPase activity The endothelial cell membrane catabolizes ADP to AMP. The energy released is utilized to modulate platelet reactivity. Hcy reduces the ADPase activity (Broekman et al, 1994) and thus platelet reactivity, which lead to platelet aggregation and blood clotting. Summary Coagulation factors and protein C have a significant role in the maintenance of blood fluidity. The proposed mechanisms by which Hcy promotes atherogenic events are not fully understood. A number of other possible mechanisms have also been proposed. In summary, elevated plasma concentration of Hcy leads to inactivation of the anticoagulant factors and/or activation of the procoagulant 32 factors . Activation or inactivation of pro or anti- coagulant factors causes disruption of the blood fluidity which leads to thrombotic events. Meta analysis: purposes The scientific literature provides a large number of replicated studies on plasma Hcy concentration in relation to age, gender, disease stage and nutritional status of B- vitamins. Such replication is essential to enhance the validity and accuracy of previous experiments, and to further our knowledge of the same topic. Meta-analysis is a research method that combines previously reported research to arrive systematically and quantitatively at conclusions about the entire body of research on a selected topic. Meta-analysis attempts to integrate empirical research findings to derive generalizations (Cooper and Hedges, 1994). Meta-analysis, as a rigorous research synthesis method, attempts to make a sense of the rapidly expanding research literature (Glass, 1976). To date, reported meta-analytic research on plasma Hcy is limited to only one (Boushey et al, 1995). The meta- analytic assessment examined plasma Hcy as a risk factor for vascular disease and the potential reduction of CAD by increasing folate intake. 33 Meta-analysis was chosen for the current study because of inconsistency in the results of independent studies. Meta-analysis accumulates results across studies and help us to gain accurate knowledge on the relationship between age, gender, disease and B—vitamins status and plasma Hcy concentration in humans. Meta-analysis is the best approach to establish reference values for healthy and CVD patients in an effort to establish prognostic tools that can be used to reduce CVD. .Meta-analysis: procedures Meta-analysis follows five rigorous and systematic steps to arrive at conclusions: a) Problem formulation, b)Data collection, c) Data evaluation, d) Data analysis and interpretation and e) Presentation. a. Problem formulation Problem formulation is the stage to construct definitions that distinguish relevant studies from irrelevant studies, and to identify evidences to be included in the review. Formulation of a research problem should considers: a) the population to which generalizations are made, and b) the source of hypothesis. For the current meta— analysis, we assumed that the generalized population can be addressed by either fixed effects model (conditional), or 34 the random effects model (unconditional) (Cooper & Hedges, 1994). In the conditional model, the population to which generalization is applied derived from the examined population. The only source of error is therefore presumed to be the sampling of subjects in the studies. In the unconditional model, the current study sample is presumed to be a sample from a hypothetical collection of studies. The population to which generalizations are applied in the current study is based on the studies from which the study sample is drawn. The source of hypotheses in problem formulation is theory and previously reported primary research studies and meta-analyses. The strength of synthesis compared to primary research lies in its ability to examine information accrued over multiple replications. The results of meta-analysis can also help direct future research. b) Data collection This process is aimed at determining a) sources of relevant studies and 2) procedures to identify the relevant evidence from the potential sources. The data collection procedure is designed to yield studies that are representative of the intended population of studies. First, critical terms descriptive of the topic under investigation 35 are identified. Correct definition of terms can accurately describe the topic at the appropriate level of specificity. The right terms are then linked to selected library reference databases (i.e. Medline, Agricola etc.) to identify potential relevant studies. If the research criteria are consistent with the population definition, potential studies should be representative samples from the population of studies (Cooper & Hedges, 1994). Potential studies are then examined based on inclusion criteria, and relevant studies are accumulated. c) Data evaluation Data evaluation is the stage in which features of interest, subject characteristics, and correlations, and any other measures of association reported in the included reported study are coded (Cooper & Hedges, 1994: pg 10-11). This coding process allows the researcher to assess the quality and amount of information present in each reported study for the present meta—analytic study. d) Data analysis and interpretation At this step we apply statistical procedures to draw inferences about the research questions, and use statistical procedures to make inferences about the literature as a whole. The meta-analyst generally draws inferences using one 36 of the following types of information from research reports: a) data that can be used to calculate effect-size estimates (e.g., means, S.D, test statistic values), b) statistical significance of hypothesis tests, and c) direction of the outcomes. The statistical procedure is chosen based on the statistical format and available data of the research studies (Cooper & Hedges, 1994). Meta—analytic findings should be then carefully interpreted based on discussion of research reports included in the meta-analysis. e) Presentation At this stage the meta-analyst apply editorial criteria to identify and organize important information to be included in the scientific report (Cooper & Hedges, 1994). Information need to be displayed in simple ways that can enhance the clarity of the conclusion. The purpose of this stage is to present meta-analytic findings accurately in order to help the readers grasp the big picture quickly (Cooper & Hedges, 1994). Strengths and.limitations of meta-analysis Scientific literatures are cluttered with repeated studies of the same phenomena. Repetitive studies are sometimes because of inherent sampling error, investigator’s lack of knowledge or because they are skeptical about the 37 results of past studies. Improved analytical methods are necessary. Meta-analysis increases reliability and allows quantitative and qualitative data analysis by combining the results of previously reported individual studies The revolution of computer technology has greatly expanded the ability and speed of literature searches, making meta-analysis possible based on thousands of journals and volumes of research data (Cooper & Hedges, 1994). Meta- analysis increases the precision of estimates by combining related studies. The approach is valuable in formulating hypotheses that could not have been tested in primary research studies. Meta-analysis can uncover results that raise interesting questions about relationships among variables (e.g. CBVD vs. CVD). By accumulating results across studies, one can gain an accurate representation of relationship that exists in the population. Meta-analyses are sometimes subjected to criticisms for a lack of control or objectivity because the process involves comparisons or summaries of different studies. This criticism can be avoided if important distinctions are observed by various techniques, such as proper coding, to ensure that the researcher differentiates among studies. Lack of objectivity may be overcome by including variety of data including unpublished data such as theses or dissertations can reduce the bias. Chapter Three METHODOLOGY Data Collection Medline (1966-97) was searched for all relevant literature, including reviews, by using combinations of keywords: [plasma or blood] and [homocysteine], [plasma or blood or dietary] and [folate], [plasma or blood or dietary] and [B6 and/or pyridoxine], [plasma or blood or dietary] and [812 and/or cobalamin]. Repeated instances of research reports were excluded by combining previous searches with the term [or]. The medline search identified 182 articles as potentially relevant to the meta-analysis (Table 5). Table 5. Results of Medline search (1966-97) Keywords Articles (11) 1 [blood or plasma] and [homocysteine] 828 2. [plasma or blood or dietary] and [folate] 2263 3. [plasma or blood or dietary] and [8,5 and/or pyridoxine] 1 170 4. [plasma or blood or dietary] and [3,2 and/or cobalamin] 2471 5. #1 and #2 140 6. #1 and #3 37 7. #1and#4 86 8. #5 or #6 or #7 182 38 39 Seventeen more articles were identified from the references cited in these reports, for a total of 199 research reports. Articles were systematically reviewed and some were excluded specifically: 1) review articles, 2) articles focusing on the mechanisms by which homocysteine may promote CVD, 3) animal studies, 4) studies with Blacks, 5) studies with homozygous or heterozygous homocysteinemic patients, 6) studies on analytical methods for plasma Hcy, 7) studies with subjects deficient in vitamin B6 and/or vitamin B12 and/or folate, and 8) studies with supplementation of vitamin B" vitamin BR and folate. In all, 48 research reports met inclusion criteria while 151 were excluded from the meta-analysis (Table 6). Coding of Research Reports A standardized coding sheet (APPENDIX A) was developed to summarize data from each research report meeting the inclusion criteria. The coding sheet included four parts: a) characteristics of study and subjects, b) blood measurements, c) correlation matrix and d) relevant findings (Appendix A). The first part a) characteristics of study and subjects, included research-report reference characteristics (ID No, author, year, source and country), statistical 4 0 Table 6. Included and excluded number of articles from the meta-analysis. Articles (11) Articles for meta-analm'c inclusion 1. Medline search 182 2. Reference cited in the articles 17 Total 199 Exclusion Criteria 11.RefimwankM$ l9 2. Mechanisms of Hcy 25 3.1MMnmJMMdks 22 4..AfihmnlflnmhmnsNfieds 4 5.(inmficlhamfluummneumumna 21 6. Analytical methods for Hey 4 ‘1 suhmnmmflphwumliyfimmhm 32 ii BammnmnmmmkmumuMon 24 10441. 151 Tog included in the meta-analysis 48 methods, number of subjects of both genders of cases and controls, and disease type of the cases. The second part b) blood measurements included analytical methods and means and standard deviations for concentrations of plasma homocysteine and blood vitamins. The third part c) correlation matrix included reported correlations of plasma homocysteine and vitamins B5, Bu and folate, as well as correlations within the three vitamins for both cases and controls of both genders . The last part, d)relevant 41 findings, summarized relevant data or other information related to the examined variables.An example of a coded research report can be found in Appendix A. Subject.variables The subject variables were gender of the sample (coded as male, female or combined) and age (mean or range). The total population for the meta-analysis consisted of four groups: 1) cases of case-control studies, 2) controls of case-control studies, 3) healthy individuals included in cross-sectional studies and 4) diseased individuals included in cross-sectional studies. Cases were subjects diagnosed with pathological conditions of either a) cardiovascular disease (CVD), b) cerebrovascular disease (CBVD), or c) all other diseases (AOD). The CVD group included subjects with coronary artery disease, coronary heart disease, ischemic stroke, myocardial infarction, peripheral arterial occlusive disease, risk factors for CVD, transient ischemic attacks and subgroups with venous thrombosis. The CBVD group included subjects with acute stroke, stroke, cerebral bleeding, cerebral infarction and cerebral thrombosis. The third subgroup (AOD) included those subjects with asthma, dementia, end-stage renal disease, geriatric problems and alcoholism. Subjects falling within the AOD subgroup were excluded from the meta-analysis. Controls were subjects who 42 were free from any life—threatening diseases and matched for age and gender within each study with the cases. Study Variables Seven analytical methods of homocysteine measurement were reported in the 48 selected research reports. These included high performance liquid chromatography with fluorometric detection (HPLC-FD), electrochemical detection (HPLC-ED), radioenzymatic assay (HPLC-REA), gas chromatography—mass spectometry (HPLC-GCMS), amino acid analyzer (AAA), ion exchange chromatography (IEC) and non- specified high performance liquid chromatography (HPLC). Plasma Hcy concentrations were expressed in the literature in terms of the arithmetic mean (AM), geometric mean (GM) or median (MD). Statistical Analysis The Statistical Package for the Social Science (Windows version 7.5, 1997, SPSS Inc, Chicago, IL) was used for the statistical analyses of the present study. Procedures for the meta-analysis followed in this study are outlined in the Handbook of Research Synthesis (Cooper and Hedges, 1994). 43 Objective One: To determine whether vascular disease cases have a higher mean plasma Hcy concentration than healthy controls. All 48 studies reported mean plasma Hcy concentration of cases and controls in the same measurement unit (umol/L). Cross-sectional study subjects were not included in this analysis. Meta analysis for objective one followed steps by determining 1) A within-study mean difference in plasma Hcy concentration between the case and control groups (Ti: individual study effect size) was estimated by Formula 1. 2) The pooled variance within individual studies (ngmn) by Formula 2. 3) The variance of each individual study effect size (v;) by Formula 3. 4) The individual study weight (wg) by Formula 4. 5) The mean effect size (Tavuage) by Formula 5. The individual study effect size is: (Formula 1) T1 = Tn-Tu , where T; is the mean plasma [Hcy] difference of study i (effect size for study i), Tn is the mean plasma [Hcy] of the cases for study i, and Tu is the mean plasma [Hcy] of the controls for study it 44 The pooled variance within individual studies is: (nu-1) 3112+ (1121-1) 5212 (Formula 2) Sizpool = ' (n11+n21‘2) where Sizpool is the pooled variance in study i, n11 is the sample size of the cases for study i, n2; is the sample size of the controls for study i, 8211 is the variance of the cases for study 1, and $221 is the variance of the controls for study i; The variance of individual study effect size is: (Formula 3) v. = szpoaui/mul/nun . where v1 is the variance of T1 for study i, $129001 is the pooled variance for study i, n11 is the sample size of cases for study 1, and n21 is the sample size of controls for study i; The individual study weight is: (Formula 4) w; = 1/v1 , where w is the weight for study 1, and v1 is the variance of T1 for study i. The mean effect size is: (FormUJ-a 5) Taverage = 2(Wi*Ti)/2Wi I where Tavern. is the mean effect size, w; is the weight for study i, T; is the effect size for study i. A positive Twang,e indicates that mean plasma Hcy concentration of the cases is higher than that of the controls in the population. 45 A homogeneity test is then performed to test the null hypothesis that differences in mean plasma Hcy concentration between the cases and the controls groups in all of the studies derived from the same population [var(Ch)=0, where (Diis the population mean difference for study i]. The statistical test performed for homogeneity is given by Formula 6. The homogeneity test is: (Formula 6) Q: 2[ (Ti'Taverage) 2* (W1) 1 r where Q is a chi-square with k-l df, K is the number of studies, Tiis the effect size for study i, Tavenge is the mean effect size, and wi is the weight for study i. When all (91 are equal, the statistic Q is distributed as a chi-square with k-l degrees of freedom (df), with k as the number of studies. If the critical value for a chi- square with k-l df exceeds Q, the null hypothesis that all studies share a common mean difference is accepted. A fixed- effects model is then examined. If Q exceeds the critical value for a chi-square with k-1 df, the null hypothesis that all studies share a common mean difference is rejected. A random-effects model is then examined. 46 Fixed effects model: Ti=®i+ei where T1 is the effect size for study 1, Ch is the population mean for study i, and e1 is the sampling error for study i. Under the fixed effects model the data to be combined arise from a series of independent studies, in which the ith study reports one effect size T1, with population effect size (91, The fixed effects model is based in the assumption that ®1=...®k...=@, that all studies share a common effect size. Random effects model: Ti=®i+e1+u1, where T, is the effect size for study i, O; is the population mean for study 1, e, is the sampling error for study i, and u; is the random effect for study i. Under this model, CM is not fixed, that is all studies do not share a common effect. 91 is random and has its own distribution. A random effects variance of u; is estimated using Formula 7 below. The studies are re-weighted by Formula 8, and a new Taverage is calculated using Formula 5 with the new weights. 47 The random effects variance of uiis: (Formula 7) 029=Kl [Qe/(K-p‘1)] H2111 , where 0%.13 the random effects variance, k is the number of studies, Qe is the weighted residual sum of square of the regression, p is the number of parameters in the model, and 2m is the sum of individual studies weights; The new individual study weight is: (Formula 8) w1=1/(oze+vi) , where wi is the new weight for study i, (99 is the random effects variance, \h is the variance of Tifor study i. Onces fixed or random effects model is chosen based on the homogeneity test, confidence interval are calculated by Formulas 9 and 10. The variance of u is: (Formula 9) 02 = 1/2w1 , where 02 is the variance of u, and EM, is the sum of individual study weights; The confidence interval for u is: (Formula 10) 95%CI = Twat.” :l: 1.96 (o), where Tavemge is the mean effect size, and o is the standard deviation of the mean Tmmnwe. 48 If the 95% confidence interval does not contain zero, the null hypothesis that the mean population effect size between the vascular disease cases and healthy controls across studies is zero is rejected. Objective Two: To determine if plasma Hcy concentration of men is higher than that of women for both vascular disease cases and healthy controls. Data on mean plasma Hcy concentration of healthy subjects (males and females) from cross-sectional and case control studies were pooled together for gender differences analysis. Data were also pooled for mean plasma Hcy concentration differences among male and female subjects, diagnosed with CVD or CBVD (cases). Gender differences in mean plasma Hcy concentration within either the case or control group, individual study weights (wi), mean effect size (T1) and confidence intervals (CI) were calculated following formula 1-7 described in Objective One. Selection of a fixed or random effects model was based on a homogeneity test, similar to that in Objective One, which was to examine gender differences in mean plasma Hcy concentration in either the case or the control group. 49 Objective Three: To establish reference values for plasma Hcy concentration for vascular disease cases and healthy control groups. Reference values (mean1SD plasma Hcy concentration) were calculated for vascular disease cases and healthy controls (male or female) separately. Objective Four: To determine if plasma Hcy concentration differs between CBVD and CVD case groups. Data on mean plasma Hcy concentration of CBVD and CVD subjects (males and females) from cross-sectional and case control studies were pooled together for difference in mean plasma Hcy concentration. The within study comparison of the two disease groups was performed first with both male and female disease subjects in the same group. Subjects were then separated based on gender, and within study gender effect sizes were calculated as shown in Objectives One and Two. A homogeneity test was performed to determine whether a random effects model was appropriate, for determining if plasma Hcy concentration differs among case groups. If Q exceeded a chi-square test with k-l degrees of freedom then a random effects model is examined. A random effects variance is estimated using Formula 9 and the studies are re-weighted by Formula 10 as in Objective One. Based on the 50 new weights, a new Tmmnweand a new confidence interval were calculated, to test the null hypothesis that CBVD and CVD cases do not differ in mean plasma Hcy concentration. Gender difference within CBVD and CVD groups was also investigated. Objective Five: To estimate the predictability of plasma Hcy concentration by blood vitamin Bs, vitamin Bu:and/or folate concentration. A weighted multiple regression was to determine whether blood B—vitamin status of participants was related to their mean Hcy concentration. A homogeneity test was performed for fixed or random effects model as described before (Objective 3). The random effects model examined is following: Random effects model: Hcy = Bo + e; + U; + [31(861) + B2(B121) + [33(folatei)+ [34 (B61 x B121)+ [35 (8121 x folate,) + [is (B61 x folate,) + [37 (B61 x folate; x 8121) where Hcy is the study sample mean plasma Hcy concentration, Bo is the population mean plasma Hcy concentration, e, is the sampling error, Uiis the random effects for study i, Bra is the slope of the variable, 861 is the mean plasma vitamin B5 concentration in study i, B121 is the mean serum vitamin Bu concentration in study i, and folate; is the mean blood folate concentration in study 1. X is the interaction terms between the B-vitamins. 51 Objective Six: To estimate the predictability of plasma Hcy concentration by dietary intake of vitamin B5, vitamin Bu and folate. A weighted multiple regression was performed at the study level to determine whether dietary B-vitamins of controls and healthy cross-sectional subjects were related to their mean plasma Hcy concentrations. A homogeneity test was performed for fixed or random effects model as described before (Objective 3). The random effects model examined is the following: Random effects model: Hcy = Bo '1' 61 '1' U1 4' 61(861) + 32(3121) '1' B3(folatei)+ B4 (861 x BlZi)+ [35 (B121 x folatei) + [35 (B61 x folatei) + [37 (B61 x folate; x B121) where Hcy is the study sample mean plasma Hcy concentration, Bo is the population mean plasma Hcy concentration, e, is the sampling error, Uiis the random effects for study i, Bun is the slope of the variable, B61 is the dietary vitamin B5 intake in study i, 8121 is the dietary vitamin Bn:intake in study i, folate] is the dietary folate intake in study i. x is the interaction terms between the dietary B- vitamins. Chapter Four RESULTS Research Reports and Subjects Data included in this meta-analysis were from 48 primary research articles describing case-control studies (n =32) and cross-sectional studies (n=16), (Tables 7 and 8). Several articles reported data from multiple sub-studies with independent sub-group subjects. The 48 studies included in this meta—analysis (Tables 7-9) represent a total of 26,132 healthy subjects (controls), and 4,549 vascular diseased cases (CBVD or CVD). Mean ages of healthy controls and vascular disease cases were 56.3 i 11.0 yrs (range: 48.6 i 8.0 yrs from 25 studies for healthy controls) and 61.6 i 9.3 yrs (age range: 48.6 i 7.8 yrs from 27 studies), respectively for vascular disease cases. Differences in mean plasma ch concentration between vascular disease cases and healthy controls and.between genders. (Objective One and Two) 52 53 0.0.9.: 5. :. cocoon. 8:58 gee—.5938 Sofie u 6.5:: 1:288 880 a 8 038......“ 8m .. $2 .288 88 u 8 . omen: on: 8.2 a... 9...... 3.2.. 38...... can... a... 28.8.. 8 ..... 8....on as; 5.8.— 69.8 owe a s 9588...... as. 8.8.8 .885. weaves... ..28: «a... as... 2...... 8.2.. 38.... can... a... 38...... 8 38.. a 8 5882.. 8.. .3319.an Eur—3:38 .3 85.9, 83 5.63.2. .528 as... e 88.52... .598 a... see... 8 .8 .5828: ..o ”.888 .08.... 2.9.8 ..o 8:025 .868 28.... 0.5.8. 8.8.... .8 boa... . s... 25...... 8 .8 .088... as. .8 8:82.58 o: E: 55860:. e: £8..on «£528 $898 83:3 e .5383: to: .88... 85.8.3 .3 5.8.. a s 2520...... as. 8.8.8 ..28... .2825... ..28: 8.8.... meets. a 5.883... .8258... a... 28...... 8 82... .. 8 2882.: .8 .82..» 8.... $2 8 .8 9:. 3:.er 32>... 803 8:8... 5:89 ..o canoe... .18.. 32.6.:— a :38 .o 8.. 8.8.2.. 8.52.8... 2%... .252. a... 28.38 $2 8 8...... .. e 5.8.2: a... .28 .5: one .8 3:89: vs. 8.85:... ..o 0:... 05 e. 88:. .58 .85 ..o 8:o Ba: :32...» Rue 3 9:0 82...; 28. 2.3 5.8 .82.. 8.2.2.2. 8.8 as. 28.38 5.3 s82... a ..28 88.. 8.2.2.2. .. ea. 8.....8... 8 $2... a .8 88...... a 6on 8:3 82:... 2... So. uses... .528 8.. 2.... 28.2.... 8 .8 9.28 .o .2 ..o boa... 8: 8:8... 838.... 8:2. .3: So... .838... 8:3 8.8:... 2. s... .582... o... 2 Bass. 28.22.. 8.83852 82:... 2... .6. 582.... .228 2... 2... m28.2.... 8 .28.. a a .25. in $582.38 a... .o 538:8... o: xooe 688:. 8.38.. ..o 0.39.. .82 82.2.. 8 ..38 .8 Fe... 2. 2.8.. .88 a. 2.8.8. 5.0 $2 8 68.. a 8 88.5. a .0388 82.5... 88-88.88 a as. 8.8.8 .88 82.. 2.3. 8 . ..= 628.8 :8er 9.03 295:8 29:8 ..o 8.35:: Race 553 Bad Rae—8.8 83 EBB.» owe .3» 03. < 83:... 38.753232. e :8... .8828 .88 =3. 834.. 8 A28: _: .o 5554.. .e— m_os:ou 880 $3.89 82:3. .mflmxamcmnmuoe one :H cocoaocfl muuooou noumommu on» we mcmwmmp xeoum .n wanes 54 8.8 8.82 88.3.... 20.8.. 82.88.... .. 225...»... 8.8.8 9.888.... 08...... 8 .02.... .. 8 8:8... ..u .88.... 8.88.. ..o 8.. 2.8.2.. 8......an .8. 808.8858 2.. 8..8... 2882...... .. 2. .2. 88...... .28 ..8: 52.22.... 8...... 2.8.8.... 8... 98 2.8.8... 5.... 98...... 8 .02.... .. 8 .880 a. ..28... 8.8... 3.2 2 888 .888. 8...... . a. 988...... 8 ..8. a... 2.... 88.5.... 2...... 20.8.. .8”. 88.3.8 <...z 8 .38.. .. 8 50 0. .88... 8.008.. 0: 9.3.... 0:0 .9. ..0 0m. 0... 0.80.. 8.08... 3.80.000 8.8.82 2. 8.8.0... 3.88.. 2.. .8 28...... .3228... ..28: 2...... 8 .88.. .. 02.... . 5.... 08...... 8.22.... 8 .88.. .. 8 2...... 2 .33.... .28........... 5.3.. 0 :. 5:00.053: .0. 00.80.... 38:0» 0 500.... 0.000.. ..0..>:. 0.00:9. ..0 0,008.00 0.0.: ..0 03. 00.. 0.... 0...... 050...... 00 300.. .0 .0 .0...0= :0Q @— .00.: 8.5.0 :0..0_..m000..:.. .00.... 0800.8 0.0... .0». 0:. .8 .0. 005...: 888.0 80:0... .0 80...».0: ..28: 8.80082... 50.70000 ..0 008.00 .8... 00.. 0...... $5.... 8 .30.. .0 .0 .090: :00 m. £59000 3...... 3.0:. 0 .3 00.0.3.0 .8582 :. 80.38 M03530 0... 8...... 5.8.2.... .228. .2. 9.6 8...... 2. 8... 08...... ..28: 2.... 8.8.... 8 0...... 822...... . 2.... 88...... 9.6 8 .82... .. 8 0...: z 888 5.... 8. 2...“... 8. 8.2.... .0... ..28... .5... 882........ ..28: 83...... 2...... 2.. 8......8... 2.6 8 .28.. .. 8 220 n. 900.00... 900.. .00.... 8.88: 3.80.. 0:..00. .0. 0.0.058 60.0w $2 8 38.. .0 .0 0.5 n. .86 8 88...... 8.2.8.8.... .8 02.... 5.... 08...... 8 a: .3008. 0.80:8. 9.0.8:“... 5.... 8:03.... 00 A... .0... .2... 0.8... 0.88.... .82... 8.28. 882......» ..28: 8.80 8.82 . .8 8.88. 38...... 8...... 28... 8 .28.. .. 8 ..30 a: .882... 822...... 2.... 8.8.3 ..28... 22.22.... ..28: <2 8 ..22. .. 8 28.8.6 2 d>u .0 800.8..55 00m .0 30.5.... 0: 0:. 0.00.08... 0:00.505 0: 5.... 9.09000 3...... 8...... a .3 ..28—9:0 8.03.03 3:00.823 $2 8 600.. .0 .0 5000 a .88. 3.... 22.. w:.w:0. own 5.3 3.000. 5.8.0000 .000. 0... E0... 0800.8 b8025— <\z no in 88. .8... 2.... 8...... 8. 5.... 8.82 ..2........... .8... 2.. a... 8.8.... 22.2.... $2 8 .22... .. 8 22.8.2: .8 22.2.0 880 5.80 22...... .U.ucou. h manna 55 200.00... 0.50.8.2. .23 80......» 0200808002 8 dun .8000... .009. .0 2.000.. c0280.. .20... 2.0 ..0 .00.»... 0.. .23 02.0.2.2. 282... 0.2.0.8.. .23 02.0.2... 282.. 8 32... .0 .0 :00..Z inn 8.8.8... 0.8.8.8 5.... 88.38. 2.8.80.8. 8.8... 8 .20 88. 8 0.88.. 8:88.. .8 .0 .88... 2. .0... 88...... .28.. 0.8.8.. 2.8.8.. 80.802.205-82 8 .32. ... 8 89...: a... 8.88.98. 0.... 8.. 88...... ..2. 88...... 8.8.8.8 88988.88 28. .8... 88.3... .58.. 8.8.8. .8005. 80.58.... 8 .88 8.8.0.8....» 8 0.2. 8.5.8.. 8 .82. 8 8 0.80.0.2 00 .0Jw.0 ..0 0.00.. bx... :0... .200.» .808 20.0000. 0 ..0 b8. 05 20.... 200222.. ~..28... 8 .3" .0w0 ..0 0.00.. bx... 2 00.000 .0200... 0 50... .8288. .0000 .0 00.0. .8000 20.0000. 0.0.x...» 0... 0.0... 0.00220... ..28: 0000...... .28.... ..0 8:02..0 .00...._0 .23 200.3. 8 .30.. .0 .0 322.02 .00" .88 :8: 5.00.5... 8... .o .088... 00.2.5.0 Sz 8 80. 2.8....5 .3528 .0 8.298 8888.. 8.8... 0.5... .58.. S2 8 .32. 8 8 0.8088... .000 8.2...8 5.... 8.. ..8 .8. 8.. 8.88... 2.8.. .58.. .96 8.3.8.8.. 2.828.828... 8 .82. 8 8 ..38. .0 0.2.5.... ..0 8.0.... .0120. 0.. 2... 02.0.6 .0282. 2.0.5... ..0 00.... 0.8.3.. .2 8 $3.. .0 .0 00.33 0022.00 2.0.. .252. :0 .00 2.00 0. 030 2.0 080.. .0 9...... 8:02.00... 0 .3 8.00.0... 900......» 2.020 2200: 8 .amn 02.00.00... 0.. 2.0 00000... 28.500000 .30 0.0... 8......» .0: 2.. 2.3 00000... 0.50.8» 0...... 2.02... 2.0 .0020... 05 0. 2.9.0.0.. 0.000.. 9.00.. 2200: 00.2000 .23 0.2.0.3.. 82.5.32. 2.02.... 8 Ana: .0 .0 00.000. 0mm $2 8 .03 85.88.. 2923... ..28: $2 8 .32. .. 8 8.88. .000 ..2 0.000 ..0 002...:0 2.3 2.8.8.. 8 in” 8.8 5.3 8.88... 8.... 0..... 28... o... 0..... 2.8.. .58.. 9.2.. 0..........8. 8 82...... 2.... 88.8. 8 .82. 8 8 8.8.8.... .5 9.0.0 .0 .2 ..0 20.0... 0.. .23 03.08.. 00 002.0.8 00.20.. 8.5 30.. 00.0 0.00.. 0... ..0 88.0w mm ..0 0w0 0... 0.0.0.. .S. .2... ..0... .23 2.0.8.. 8 83.. .0 .0 0000.00.0— an 0.0.2... 2.0.3... 0. 8.00.0.2... 2.3 03...... 022.022.... 000...... .0020 2000.00 .0 magmas .2 0 002......» .o 888. 88... 8 00.0.5... 8.8.88 .880. . .8... 8.880 8.. 2.... 05 8.....8. ..8 .0... 88.3... 88.2.... 8 .32. ... 8 2.0.8.. .0 20.2.00 00000 5.000 0.2210. .8..coo. . 0.00. 56 82......8 38.82.88... 85.. 28 8......8 8822....» 28.8.. 88.8.. .58... 8 6%.. 8 8 .525 2 8...... 32. .8 825.38. .828 2.. 52. 3.88. 88.38 3.8: $2 8 G2... .. 8 ..23: 3 .9888. .88.... 2.. 8 8.8.2» 2.. 8. .88....8 8.8... .588... 8.. 2. 88.28 88.32.. <2 8 .82. 8 8 8.25 9 .ngEESSuxutseflgvfi .§%§§§ 8.882.. 5.8.. 8. 8.80 2.. .28.... 8238. .8259 8.. a... 8 ..22 ..s. 12.882 28. 82§_o> 8 $8: 8 8 «.3. «v 888... 8...... ..s. .88 8 8.8.82. 888 2.. 52... 8.8.8 €825. .28 5.8.. 28.28.... 2.. 22.. 88.8.. .2 8 .28.. 8 8 8.85m 3 82.8.5.8 8.58... 2.. 8. 888888.. 2.2 82.8 2.2.... 88: 558.88.. 2.. .8 8882.. Nazca <2 8 .28.. 8 8 22.8 3 .820 .o ow“ .o 88» cm .258 9.2.8 .58.... 05 .«c 8383. 9.3.3.20. $2 8 Jan .03 .6 83» 232. 3:8 :28 6:55.88... 05 mo Eon—Eon. 9.5.56 $2 8 dam 88 .o 88. 3-5. 2...... 22.8 8858.58... 2.. .o 8382.. «8.25m 32 8 £2... .8 8 2.58m .18 . 9.0 ..o 8.8...8 8m 2. .88.... 2. 5.3 582.... 8.828 5.8.. 8.888 .8 8.28.8 88.38 8.28880 .96 8.8.388 5.3 8.8.8.. 8 .38.. 8 8 2.8.8.. an .22.... .28. 8 8.8... 8.88.... .8 88...: .95”... 8.8... 88.88.8888 0..... 2. 5.3 228... M2.2.8.8 5.8.. 8.888 .8 2......28 88.3.5 5.3 28...... 85.8.. 8 8.88.8.8. .8 8.8.8.. 8 Ga: ... 8 :58... 8. 888.25... 2.8.. 2.. .8...8 88.... 2.2.2. ..a 2.. 22.28 2 8.... <2 8 Ga: .8 8 8...... on .95.» own 85. .56. .3 80:82: v5. 5 3 8.2.» ..o boa... 855.3 89.8 3898.332. 258. 62.. 3.8.8 5:8de 85:3 8.82m 8 $8: 1 .0 En mm .88 855.8 25.» 05 :. 92>: :2: mo 6.5.0 05 «a $0 039.53... 835:: 2.2.9.8 ..o 9.3: 320888 a 82.. ..28—8. 3535 .8 Egan. 8.28 .5382... 2.3 8.8.3.. 8 3: .8 8 2:82.05.— vn 8.3...» 5.3 mafia 888:. 85c 55. .o 8.39; .«o 8.3%“... 82.2.. 2. 5.2 88.8... 8.8.8.... 88.82 2.. 82.. .8828 $2 8 a8: 8 8 .8882 can .88» 3.9. 9.3.. 8.89.... .850 as. 3 2.38.3 no £858... 828... 2. 5.3458... 8.832... 88.82 2.. 22.. 8.8.8.8. $2 8 as: 8 8 .2822 2.9. .88» N73. main 3.88:. .850 as. .6 5.68., no 38.58:. 2.2.2.. 2. 5.3 88.8.. 8.8.2.8.. 8.8.82 2.. 5...... 8.8.8... . $2 8 $8.. 8 8 .2822 .5. 880 5.8a 22.5... 8.95.30 .8..cou. n 8.889 57 8:56 8:8 5 38203.3 0:3 «9:3? 03205qu no gag» .m§_%o._ 8:. 59a €82 09.28% 3~§qu8 .o 2E8 bSmE 3.5.. co 3%: 9:83 .3 0368 noun—anon Eon—ad 89¢ 8:13.803 803 2.3 3552; Q»:— c: S 25395 .3 boa:— o: as .23 =32 2:3 2.. s E. .9828. 2.. 5.5 38.3 258.5. .2 we a a? 853%.... 9893 8a .538 8 68: _. s~8§> 3 . 22:80 8.80 :33 flo5=< An.ucoov n magma 58 8-8 "82.23 $— . and-E: G8: a .o 5.8: 5n .2 3-8 88.22.: "$.28. 02 < a 30 <2 .5. .B 95.2: G8: a a 9.8 2 8-8 <2 .22 "2.23. .2 <2 .S— .E 9.6-E: 38: a a 9,80 m. was». ":22 £2.32: 88: .a .0 =5 2 8-8 "5.22 "2.28 >8. .0: 8-8 22.22 "2.22” <8 .22 8-8 "a _ .22 "2.22 m< 8.3.5 88: a .0 :80 .5 av "228 <9. 8.0-2: :8: a .0 x8885 2 8-8 88 .28... 2.: < 3 <2 .5. .9- 562: Ga: 3 3 585 a 3-8 50.28 ..2 8-3 232% <5. 3.5 $8: a 8 ace-=5 an 8-2 "28.2....- 8~.28 .5 NE; mar-.2: 812% m <2 cud-E: Ga: 3 .0 82.8% .2. 8-3" S23 "3.2” .8 .8 mmém "332$ «BEEN— Sh do on-..” 8~.2- "22.23 2 <2 .9 <<< 88: a .0 525.5 .3 8-8 “.22 a... 3.? min _ .an 8-8 ".2: <<< 3.3: a 6 again .8 3.: 28.288 812°: 96 95.2: G8: a .o 585< a 8-8 “5.28. ":22 mo .8- 38 h: _ .28 an _ .28 6 9.6-E: 88: a s 2.5 an 8-8 "5.2:- <2 .e 9.: £8: a s 83.5 n 8.8 “222:; 23.28 2. <3 m ..= 8-8 "222:; 28.2% «.2 < a .2 93-2: 83: a a 5&2 .2 » .om< 5m .2228 gum .830 3853 .6: 830 §a§> .3: ”553. 2 oz 2 oz 225202 $5.82 .8832 .mflm2HmcmnmumE mnu cfl vaSHUCH mmflvsum mo mvocume Hmoauxamcm Ucm mowumflumuomumco .m manna 59 G 2 am “3.23. "—2.28 02 \ 3 95m 5m 8-9:: 88: a a Sussex 5. 5-2 22. 5. {a EOE: E“ a... a E on $3. 22. 2:: m 932: G8: a .0 Eu mm 8-8 22: 2:: 93 <2 .42 5-3.5 $8: a a 255' :25. on Si. “—22.23: 08 3-9. “582wa Ann 918 “58.22% 3m: G8: a .0 can? an... E? 82.2% .fim 62. “5.22 $2.25 2n 5 .33. 28558.15— ”<2 8.8.220 83.82... 6. .83 8.3.8.3. “<5. 20.8... 9.5.30 H.5 .32 2.3. 5. 8.8.0 3:2 09% Em. ”Guam Egan—=20 5...... “UL 885x. “2: .33. 1.88.2 “<2 n.8aE2E. 15200 .8 ngoflEEfiV uufinuxm =2 ”Um. cognac—1.5050 H5 83.8555. ”<— m.mo..EE.—.r m=o5> .._.> Safe :3: 5:25 .95 2938.520 2.65 agate: 5.: 0......— §u§< 0.52.8. 8.25 US... 8.88... .828 .8 308.0% .92-...goasefi 26 “28 8.26. .m 8.8.: :8: F880 ”:5 8.632. 0.58.28: A... 385 3.338.an 8. 8.8.. .5. So: 9.2.... 3.3 .9. 3.8.2.55 8.85 ”E 8.83% 8.52.88»... 5m 8.3.5 2.3.80 .832 .888. Moos. 8:287. ...< “.322 8.53. ”<5. 5.15. 2% 8.5 .52. an «8:85.23 3-8 .52; 80.22. :5 m: 9.0:... 9.8.. ..w .0 =3 8 3.-.... 25¢ 28. m. 801...: $8 .. .a 38...; S iv “.52 . M. :828 .2 B 8-3:: 88.. .a s .885 8 .NA .22.. “.22.. < 5. 9.6:: 88.. .m 3 .58: w... 3.8. 28 5.0.: $8.. .m a 8.5%... 3. .5: 2.8 23. 5. .5. 5.3.... .88 c ..w .0 v.58: 9. 8-3 2.. 2. >9. 5. .5 SW: 3.3.. .a a £5 a. 3.-.... 2.8 2.8 .3. \ 3 .2 <2 8-3:: .82. .a s 585% .v 8-3. 83.2.. . 8 <2 .5. .E 9.6:: $8.. .a .0 85$ 3 +8 8.. . .28. 98.. .r. s 98.8 .8... 3-2. 88.2.: . 98.. .a .0 £58 5% :8 me. @288 3.. . 8 <2 .5. .E 9.6:”: 98.. ..W a 22.3 ..ah .. :8 ".323. ...8..2.8 were. o5 2653 0:33 >532. ”£9280 .9.on own .3 £95.80 28 880 .«o cougcoocoo so: «Ema—o :35 3230? .v 8.5mm MN mac» momma I 2223 I h. ..N 03. EN— 92:. fie —. Nam emu @060 mt. o3 om< «0.9m ENF no». wee 973 5.3. as": 8.8 oz 1 o u m a.» u 2. I 2. .. oN -mN (mourn) [(01.1 emseld 66 (p<0.05) (Table 11). A fixed effect model was subsequently used. Mean plasma Hcy concentration of healthy men (11.2:l.6 umol/l; range:9.3-l4.0 umol/l) was significantly higher (P<0.05) than that of healthy women (9.5:2.1 umol/l; range:7.6-13.2 umol/l) with an average difference of 1.7 u mol/l , (95%CI=1.3, 1.9). The gender difference persisted even when the effect of age was controlled (Figure 5). In all age groups, healthy men had a higher mean plasma Hcy concentration than healthy women. No data are available for comparison of healthy women to men for the age range of 50-59 years. Mean plasma Hcy concentration of the diseased men (14.2¢2.6 pmol/l; range:11.7-20.0 umol/l) was not statistically higher than that of diseased women (13.713.1 u mol/l; range:12.0-21.7 umol/l), with an average difference of 0.5umol/l, (95% CI=-0.6, 1.6) (Table 9). Male cases had a similar mean plasma Hcy concentration to female cases for the age groups of 50-59, 60-69, and 70-79 years (Figure 6). No data are available for women cases of 40-49 year range, and males of 80-89 year age range. 67 032:?“ 8% 0: ”Q2 .0333 wcmcoaaohcéz >5 365:5 Bacon >532. H2838 own 28 bucom 3 £833 has: he cougcoocoo mo: «Ema—c 52c puss—$03 .n oSME 26.20; I :62.— I 3. a. 2: 94. New m2 «8 won or. ow< Que. mace amém $8 carom H: "G (mourn) 59H emseId 68 CfiEO; I :0:— I mm s. ..N 03256 83 o: Hn—Z 33% 5:086:50 .8 33896523 53 038.“ ”mom—8 .owa ecu Epsom .3 830 .«o couabcoocoo zom «Ema—o 508 v8.30? 6 8:3...“ 8N 2 co 2.— own 2:. vmm 0?. 2.2. $0.8 anyom m.N—. H: "G l O r In I O F I m P l O N umN ('moum) [(01.1 emseld 69 Normal range of plasma Hcy concentration of healthy adults and a cut-off point for hyperhomocysteinemia. (Objective Three) We calculate the normal plasma Hcy concentration value (mean i SD) which can be used as a reference value of healthy adults. Normal range of plasma Hcy concentration of healthy adults were estimated based on data used for Objectives One, Two (Table 12). Table 12. Mean plasma Hey concentration of healthy adults and those with vascular dmamcm Group Gender Hcy (SD) Range (umol/l) (umol/l) vascular both 13.9 (3.9) 9.4-33.8 dmame cases men 14.2 (2.6) 11.7-20.0 women 13.7 (3.1) 12.0-21.7 healthy both 10.9 (1.5) 7.3-14.0 anunfls men 11.2 (1.6) 9.3-14.0 women 9.5 (2.1) 7.6-13.2 The large number of healthy men and women included in this meta-analysis had a mean(SD) plasma Hcy concentrations of 11.2(1.6) umolll and 9.5(2.1) umol/l, respectively. Mean(SD) plasma Hcy concentration of vascular disease men did not 7O differ from that of vascular disease women (Objective Four), but 13.9(3.9) ummVlbeen to indicate a critical value for hyperhomocysteinemia and increase risk for vascular diseases. Difference in mean plasma Hcy concentration of CBVD and CVD cases (Objective Four) Mean plasma Hcy concentration of CBVD cases (16.5:3.3 umol/l; range:13.7-21.7 umol/l) was significantly (P<0.05) higher than that of CVD cases (10.9:1.5 umol/l; range:7.3- 14.0 umol/l, Tavenge=2.6, 95%CI=2.42, 2.78). The weighted mean plasma Hcy concentration of CBVD patients was significantly higher than that of CVD patients even after age (Figure 7) and gender (Figure 8) were controlled. Insufficient studies and samples were available for CBVD in the age range of 40-49 and the CVD group for the 80-89 age range. A separate analysis of differences in mean plasma Hcy concentration among men and women within CBVD or CVD case groups, showed no significant differences within CBVD (CI=- O.4, 0.6) and CVD (CI=-0.2, 0.4) groups among men and women . 71 05232“ 8% 0: 52 A0336 Riots 120530.“ £2500qu 36.80an .0385 but.» 3:28 .wd v 0.00035 83036580 HQ>U E85585 E5200 6285:: 35800 5.50003 350.50 .0055 .m .0 V 0.00086 3300065300 HQ>mU .09“ 5 3:2» Q>U a. Emu co 5385:0280 >0: «Ema—d 50E “c8353 H 053"— Nov o3 one new .I. = MN 3 mm 2: u : 0m< 8.8 2.2 00.8 8.8 913 D>U I D>m0 I ('I/Ioum) 40H emseld 72 35:3 0385 33086630 HQ>U 0.80:3 0008:. 330865200 “Emu .0886 one 030» 5 880 do cob—350:3 >0: «Ema—c 508 82303 .w 05mm 03005 2: 22 3 man u: D>U D>m0 FEED; I COF— I (cmoum) 49H ewseld 73 The predictability of plasma Boy concentration by blood B- vitamin concentration. (Objective Five) Pearson's correlations were calculated to assess blood B-vitamin interactions (Table 12). All correlations within the B-vitamins were highly significant (P<.01). Blood folate had a correlation of 0.66 with plasma Baland a correlation of 0.42 with serum Bu. Serum Bugwas also highly correlated Table 13. Pearson correlations coefficients among plasma concentrations and dietary intakes of B-vitamins. bknd dkMHy serum B12 blood folate dietary B,2 dietary folate plasma B,s 0.58" 0.66“ dietary B,5 0.72" 0.28" serum B12 0.42" dietary B12 0.19“ " P< 0.01 with plasma Be (r=0.58). Simple linear regressions for predicting mean plasma Hcy concentration by individual mean blood B-vitamin concentration was performed (Table 14, Model 1-3). Data were pooled together from both vascular disease cases and healthy controls. All three blood B-vitamins had significant negative associations with plasma Hcy concentration (5mm = - 0.016; [33312 - -0.004,' and Bbfol = -0.033) (Table 14). The 74 negative association between the blood B-vitamins and mean plasma Hcy concentration is presented in Figures 9—11. Weighted multiple regression of plasma Hcy concentration of all three vitamins without interaction terms also regressed significance (3mm = -0.192; Bflnzs- 0.172; and Bbfifl = -0.068). When the weighted multiple regression model included all three blood B-vitamins plus their interactions (Table 14, Model 5), backward elimination retain serum 812 (83312 = — Table 14. Simple and multiple regression of blood B-vitamin concentration to plasma Hey concentration. godimknm Model B Std. Error t Sig 1 (Constant) 13.598 0.1 17 1 16.228 Vitamin B. -0.016 0.002 -9.255 0.000 2 (Constant) 14.125 0.225 62.913 Vitamin Bu -0.004 0.001 -5.941 0.000 3 (Constant) 29.665 1.550 19.142 folate -0.033 0.005 -7 .141 0.000 4 (Constant) 94.082 0.000 633.100 Vitamin B. -0.192 0.000 -114.000 0.000 Vitamin Bu -0.172 0.000 -413.000 0.000 folate -0.068 0.000 -5370.000 0.000 5 (Constant) 94.082 0.000 922.000 Vitamin Bu -0.152 0.000 -619.000 0.000 folate -0.067 0.000 -750.000 0.000 Vitamins B,*B,, -0.001 0.000 -235.000 0.000 75 22.? 18- 16- 14- ° 104 plasma [Hey] (umol/L) .2 a 20 4'0 6'0 8'0 100 120 plasma vitamin 86 (nmol/L) Figure 9. Scatterplot of plasma Hey concentration and plasma vitamin B,s concentration. 22 O 20- o 18- 16- § 14- ° no a O a o o E \u\_ g o I— 12" 0° 0 u o 5‘ a o to 101 E 8 Ti 8 . . 200 300 400 500 serum vitamin B12 (pmollL) Figure 10. Scatterplot of plasma Hey concentration and serum vitamin B12 concentration. 76 22 20- o 18- 16- 14- plasma [Hey] (umollL) 12 260 260 360 sic 330 360 380 blood folate (nmollL) Figure 11. Scatterplot of plasma Hcy concentration and blood folate concentration 0.152), blood folate (Benn = -0.067) and the interaction between plasma vitamin B; and serum vitamin B12 (836.312 = - 0.001). All three predictors showed significantly (P<0.001) negative association with mean plasma Hcy concentration. The predictability of plasma Hcy concentration by dietary intake of B-vitamins. (Objective Six) Pearson's correlations among intake of dietary vitamins 85,18” and folate of healthy control subjects (Table 12) were highly significant (P<0.01). Dietary folate correlated 77 with dietary vitamin B6 (r=0.28) and dietary vitamin Bu (r=0.l9). Dietary vitamin Bugwas also highly correlated with vitamin B6 (r=0.72, P<0.05). A weighted multiple regression model for dietary intakes of B-vitamins and their interactions (Table 15) Table 15. Multiple regression output of dietary B-vitamins predicting plasma Hey ffici ts Model B Std. Error t Sig 1 (Constant) 59.434 1.202 49.426 B,2 -9.515 0.353 -26.976 0.000 folate -0.307 0.005 -58.870 0.000 B,"‘B,2 -0.461 0.035 -l3.370 0.000 B,2*folate 0.064 0.002 41.466 0.000 Bffolate 0.009 0.001 1 1.632 0.000 resulted in significant negative regressions (Bmz = ~9.515; Bgolate - -0.307) . Dietary vitamin 85 was excluded from the model by backward elimination. Chapter Five DISCUSSION AND CONCLUSION Discussion Hyperhomocysteinemia has frequently been defined in literature by the 90-95u‘percentile of plasma Hcy concentration of healthy controls (Genest et al, 1990; Heijer et al, 1993; Selhub et al, 1993; Robinson et al, 1996). Reported critical values of plasma Hcy concentration of healthy adults range from 14.0 umol/l (Selhub et al, 1993) to 19.0 umol/l (Genest et al, 1990). This wide range of critical value for plasma Hcy concentration makes no distinction between genders, and does not reflect a reference range that can be generalized for public health. These critical values of individual studies, each of which was based on a small percent of the general population, have not been consistent across studies. Genest et a1 (1990) reported 90th and 95th percentile values of 15.0 and 19.0 umol/l for healthy controls and Heijer et al (1993) reported 95th percentile values of 18.5 umol/l among their controls. Moreover, Selhub et a1 (1993) 78 79 defined hyperhomocysteinemia as 14 umol/l, and Robinson et al (1996) reported 95"‘percentile values of 16.3 umol/l in healthy controls. This approach led scientists to adopt a wide range of plasma Hcy concentration for healthy adults based on individual study findings. Previous research (Boushey et al, 1995), reported that an increase in plasma Hcy concentration by 1 umol/l is associated with 10% higher risk for CVD, and a 5 umol/l increment in plasma Hcy may be associated with 250% increase in risk of vascular disease. The last ten years a substantial amount of data on plasma Hcy have been accumulated (normal ranges by age for men and women). Utilizing the unique statistical power of incorporating findings of individual studies to make generalization, this meta-analysis identified healthy and diseased reference ranges of plasma Hcy concentration for men and women. However, hyperhomocysteinemia a new independent risk factor for CVD (Clarke et al, 1991; Stampfer et al, 1991), can be linearly associated with the risk for CVD only in certain ranges. The need for plasma Hcy reference values that can be adopted as a screening tool to help maintain plasma Hcy levels within the normal range for disease prevention becomes an important issue. I confirmed, in this meta-analysis, a significantly higher mean plasma Hcy concentration of vascular disease cases than that of healthy controls (Malinow et al, 1989; 80 Stampfer et a1, 1992; Pancharuniti et al, 1994; Robinson et a1, 1995). The difference in mean plasma Hcy concentration of vascular disease cases and healthy controls in this meta- analysis was 3 umol/l. Verhoef et al in 1996 reported that a 3 umol/l increase in plasma Hcy was associated with 1.35 odds ratio for myocardial infarction. Among healthy subjects, we observed higher mean plasma Hcy concentration in men compared to women. The mean plasma Hcy difference of healthy men and women was about 2 umol/l. No significant differences were observed in mean plasma Hcy concentration among vascular disease men and women. Similar insignificant findings have also been reported by Dalery and colleagues in 1995; (11.7 i 5.8 umol/l and 12.0 i 6.3 umol/l). These findings suggest healthy reference ranges for men of 9.3- 14.0 pmol/l with a mean 1 SD of 11.2 i 1.6 umol/l and 7.6- 13.2 umol/l with a mean i SD of 9.5 i 2.1 umol/l for women. Mean plasma Hcy value of 214 umol/l was associated with vascular disease cases. Robinson et a1 (1995) reported that a plasma Hcy concentration of 14 umol/l conferred an odds ratio of coronary disease of 4.8. Nygard et al (1998) reported different reference ranges for apparently healthy participants of the Hordland Homocysteine Study as 6.2-18.7 umol/l and 5.1-16.5 umol/l for men and women, respectively. The reported findings of 81 Nygard et al, have similar lower reference ranges for both men and women with findings of this meta-analysis. Differences in the upper reference value may be attributed to the fact that researchers excluded previously diagnosed subjects of vascular diseases but no vascular assessment for possible development of the disease at the time of the study was performed. The normal health ranges, of plasma Hcy concentration established in this meta-analysis based on data of 2319 healthy and 3628 vascular disease subjects may be useful in identifying population risk groups for vascular diseases. Evaluation of mean plasma Hcy concentration of CBVD and CVD disease groups identify significant differences between the two groups. CBVD cases had higher mean plasma Hcy concentration than CVD cases. In my knowledge, this is the first time that separate evaluation for mean plasma Hcy concentration of CBVD and CVD disease groups is performed. There are no proposed mechanisms for the difference in mean plasma Hcy concentration among the disease groups. Further investigation is needed to establish reasoning. Studies (Ubbink et al, 1993; Jacques et al, 1996; Dudman et al, 1996), report suboptimal blood concentration of vitamins 85,18” and folate as modifiable risk factors for hyperhomocysteinemia. In this meta-analysis, vascular disease cases compared to the controls had lower 82 concentration of plasma vitamin 86 (47.4129.0; 65.9:38.7 nmo/l), lower blood folate (298.9:29.0; 339.2:22.1 nmol/l) but had comparable serum vitamin Bu (308.9177.9; 310.5:73.2 pmol/l ). Blood folate (B=-0.68), serum vitamin Bu (8:- 0.172), and plasma B5 (B=-0.192) exhibited significant (P<0.001) inverse associations with mean plasma Hcy concentration for both cases and controls, when controlling for interactions of the blood B-vitamins. Blood folate exhibited strong positive correlations with plasma vitamin 85 (r=0.66) and serum vitamin Bu;(r=0.42). Plasma Bg was also positively correlated with serum vitamin Bu (r=0.58). When plotting blood folate levels against plasma Hcy concentration we observed that suggested normal plasma Hcy concentration of 11.0 umol/l is associated with blood folate levels of >360 nmol/l. In relation to serum vitamin Bu, optimal plasma Hcy concentration can be achieved for serum concentrations of vitamin Bu greater than 350 pmol/l, while greater than 60 nmol/l of plasma B5 are needed for the same purpose. Based on these findings, I am suggesting the above critical values of blood vitamins B5,IBu and folate to maintain plasma Hcy concentration below the disease level (11.0 umol/l). Mean dietary intakes of folate and vitamin Bu showed significant inverse association (Bflflmm=-0.307; Bau=-9.515) 83 with mean plasma Hcy concentration for the healthy subjects, when dietary vitamin interactions where controlled. Dietary vitamin B5 showed no association with mean plasma Hcy concentration, and was excluded from the model by backward elimination. Univariate analysis of dietary vitamin B5also showed no association, although plasma vitamin B5 concentration exhibited a significant inverse association (8= -0.012) with mean plasma Hcy concentration. The conflicting findings on plasma as opposed to dietary vitamin B5I attribute to the limited number of studies which included information on dietary intakes of vitamin B5. Most studies reported data on dietary folate and dietary vitamin B”. Dietary data were only reported for the healthy control group, and no dietary B-vitamin comparison can be made with vascular disease cases. Dietary folate exhibited strong positive correlations with dietary vitamins B5 (r=0.28) and 812 (r=0.l9) . Dietary vitamin B5 also had a significant positive correlation with dietary vitamin Bu (r=0.72). Elderly people have higher mean plasma Hcy concentration than younger people, are frequently deficient in blood B-vitamins (Garry et al,1984; Joosten et al, 1993), and are in a higher risk for CVD. Suboptimal blood B-vitamin concentration may partially explain the increase risk for CVD in this group. Blood B-vitamin deficiency in the elderly may be attributed to socioeconomic status, intestinal 84 vitamin malabsorption or decreased appetite. These findings address the significance of nutritional adequacy of the B— vitamins for vascular disease prevention. Conclusion In conclusion, this meta-analysis confirmed that vascular disease cases have higher mean plasma Hcy than healthy controls. Men have higher mean plasma Hcy concentration than women for the healthy group, but similar mean plasma Hcy concentration for the vascular disease group. CBVD cases had higher mean plasma Hcy than CVD cases. Furthermore, this meta-analysis addressed the view that plasma Hcy concentration is inversely associated with blood or dietary B-vitamins. Adequate intake of B-vitamins for maintenance of optimal levels in the blood may be an important step for maintaining plasma Hcy concentration to normal levels, to reduce the risk for vascular diseases. Chapter Six ASSUMPTIONS, LIMITATIONS, IMPLICATIONS, AND RECOMMENDATIONS .Assumptions In conducting the present study, the following assumptions were made: 1. The errors of estimation eiare statistically independent, each with a mean of zero and estimation variance vi. 2. Each random effect U1 is independent with a mean of zero. 3. In the random effects model part of the variability of the true effect sizes is unexplained by the model. Limitations This study has some limitations that should be addressed: 1. The majority of the studies included in the meta—analysis represent case- control and cross-sectional study designs. No prospective studies were included in this meta-analysis. 85 86 2. The case-control studies included in this meta-analysis had relatively higher percentage of male subjects for both vascular disease and healthy groups when compared to female subjects. 3. The results of this meta-analysis are based on the group and not on the individual level. Certain limitations for generalization may apply. 4. Dietary data were limited to a very small number of studies. Implications Hyperhomocysteinemia has recently been identified as an independent risk factor for CVD. Reference values for healthy and vascular diseased people have not yet been established. This meta-analysis proposes reference values for healthy or diseased adult Caucasian men and women. These values are lower than the currently proposed in the literature. These reference values can be used as a screening tool for prevention, to help reduce the risk for CVD. We have shown that dietary and consequently blood B- vitamins are significant factors for predicting mean plasma Hcy concentration. Adequate dietary intakes of the B- vitamins may be needed to normalize levels of plasma Hcy and thereby lower vascular diseases. 87 Recommendations In this meta-analysis CBVD patients had a higher mean plasma Hcy concentration than CVD patients. No distinction was made by gender of the diseased due to limited data. More studies are accumulated and gender within diseased group should be further investigated. As individual research on plasma Hcy concentration and dietary intakes of vitamins B5, Bugand folate accumulates, a new meta-analysis is needed to establish dietary intakes of the B—vitamins, to help maintain normal plasma Hcy concentration. RDA are established for disease prevention. The new scientific findings on plasma Hcy concentration and the effect of the B-vitamins address the need to reconsider the RDA. Hyperhomocysteinemia is an independent risk factor for CVD. Monitoring of plasma Hcy concentration should be part of the standard blood analysis. This will help identify hyperhomocysteinemic people, help implement nutritional intervention with B-vitamin supplementation, and help reduce the risk of CVD. In our study, all three blood B-vitamins were significant determinants of mean plasma Hcy concentration. Blood folate concentration was the most significant factor. This relationship should be further examined by controlled supplementation studies of individuals and with combinations 88 of the B-vitamins. New blood B-vitamin reference values need to be established to help maintain normal plasma Hcy concentration in an effort to reduce the risk for CVD. APPENDICES APPENDIX A CODING SHEET '89 a 8 ”8033. a .80... .35....» 0228.82 .6 8.58% so: 2.; 2.358 8388 .820. 0 EB. 880.00 00.8 . mm. .02 2 8.062 3 33E 036 .96 a8 02 2.3.... .050 08 .80. a 6...; c. 8550 Be. 35.8 .96. 808.0 r82 :38 800 . we. .89 N... 0.0.050. on. 8.2. .228 3:30... 5.0.593 8:08:50. 8:0. .69 x Boa .363: <>ooz< 58202 038. cause .39.. .040 08.. 0858. .8350 0.: 5.58 .33 .99 <90 .69 8505. 82.58.80. .3 00582. 2.58... 2.85 m 08> .E .2 8.80 845...: 883056 0.3: 8:825. 6.... .3 3.885 8.3385... 82 30> 3 2060.030 >505 ..m .0 0:28: .653 on 02 o. ammmm QZHQOU .fl xHszmm¢ c292 90 8.2.0. no.3. nab—.8 x 8569 83E «38 $53. mmams. mmawux :23... $235. :83... «B 39...: 2m .35 NE SE 0223 x 35.8 5&8 3:89 22..» 8.2.9 9:88 $.33 8:5 38.: 8.2. 28:8 x 0.235” 58:” Sufi 38a. 8.39 35.8 3.58 933. n.3,? no.2. "88 $59.. $59.. mmamfi mwamfl ..soes 2.8.3 goes £25. 320. Stan—L Nvfl wanna—A an gnaw—l >0: 8.9200. .fl xHszmmd 91 £0ch n o «ammo u o o.o.o. .ooow 22o. 5.20 «E 39.20 «E .88 «E .3 8 now $05 on .25 8 85 .2 2.8x. end no «Fm «Ema... «no no «No no o. to no 8 2.8x. cud no on... no and. no 8.... no cud. no of 93. no Rd. no .No- no _o. 99.20 .2 gm _o. 85 NE .5955 NE 96 «5 $6 8 96 +55 8 8m 8 .3 _o. snow. NE £8... 8 Ema... .o.ezou. .¢ xHozmmmc 92 .5060 3.09 ”300.0000 .5200 on - 0.0 0.05. .0000: ”0.0.0.. 00.00... .5900 0.00.03 ”0.05. .0000: “mum 0800.0 ”000000.000 .0. 000.9 0.00.0.0 00.2.2.0. 00. 000: .053. 002 .228. 9.0.20 .3 8080.80. 8 080028000 .220... 0820... a 0.2.6. ..8 0.00.800 .. .mEmEmcooE 00.0. >0 0.020.000.050 0.00050 0. 0.00000 0.500.0.m>000.0£00>... .00. 00.0.0 80.3.. 00 .. I00. .02 09.0 .00 ..20 00.0.00 .09... not... .0. .050 .0. E05033 .000 0.00 0. 0000.00. 0.0.. to 00 0...... 00.0.0800 00.... >0... 0. 0000.00. $080-0. < 4.00.: m 0500 000. >0... 00.00.... 00 000.03 000 00E 0.00 0000.0 9.0 .0. x00 0. 0000.00. 0208.020 0 003 0.0.... 00.0.0.0 ..00. 0. 0.0.0. .0 0.052030100005090 000 96 0...... 0.00.35 d0... 0E3... 00...... 3.8200. .H xHDzmmm¢ APPENDIX B CODED VARIABLES APPENDIX B. sequence variable 1. 3. id# first author study design \omqmmAwNH code 1-48 93 CODED VARIABLES explanation Coding number of journal article Alfthan Andersson Araki Arnesen Brattstrom Cacan Chadefaux Coull Chu Cravo Dalery Den Heijer Fermo Gary Genest Herzlich Hopkins Israelsson Jacobsen Joosten Landgren Lewis Lindenbaum Malinow Moolgard Nilsson Nygard Pancharuniti Perry Riggs Robinson Selhub Stampfer Swift Ubbink Verhoef Wu “v 3.0- ‘A cases-control cross-sectional APPENDIX B. (cont’d) sequence variable 4. disease [cases] 5. hcymeth [analytical methods for HCY] 6. gender 7. agecas 8. sdagca 94 code \meawmw NH explanation cerebrovascular diseases acute stroke stroke cerebral bleeding cerebral infraction cerebral thrombosis cardiovascular diseases coronary artery disease coronary heart disease ischemic stroke myocardial infraction peripheral arterial occlusive disease risk factors for CVD transient ischemic attacks venous thrombosis all other diseases asthma alcoholics dementia and stage renal disease geriatric problems Intermittent HPLC-FD HPLC-ED HPLC-REA HPLC-GCMS AAA HPLC IEC male female both age for cases standard deviation of age for cases APPENDIX B. (cont'd) sequence variable 9. ageco 10. sdagco 11. cases 12. controls 13. typme 14. mehcy 15. mhca 16. sdhcas 17. mhcon 18. sdhco 19. meb6 20. mb6ca 21. sdb6cas 22. mb6co 23. sdb6co 24. mb12ca 25. sdb12ca 95 code waH wNH H ALUNH H |-' explanation age for controls standard deviation of age for controls number of cases number of controls Arithmetic Mean Geometric Mean Median plasma serum blood red cell mean plasma homocysteine for cases standard deviation mean plasma homocysteine for controls standard deviation plasma serum blood red cell mean vitamin B6 for cases standard deviation mean vitamin B6 for controls standard deviation mean vitamin 812 for cases standard deviation APPENDIX B. (cont’d) sequence variable 26. mb12co 27. sdb12co 28. mefol 29. mfolca 30. sdmfolca 31. mfolco 32. sdfolco 33. Kcalca 34. deca 35. Kcalco 36. dec0 37. dieb6ca 38. sddb6ca 39. dieb6co 40. sddb6co 41. dib12ca 42. sddb12ca 43. dib12co 44. sddb12co 96 code H .5me explanation mean vitamin 812 for controls standard deviation plasma serum blood red cell mean folate for cases standard deviation mean folate for controls standard deviation daily Calorie intake for cases standard deviation daily calorie intake for controls standard deviation dietary b6 for cases standard deviation dietary b6 for controls standard deviation dietary b12 for cases standard deviation dietary b12 for controls standard deviation 97 APPENDIX B. (cont'd) seguence variable code 45. difolca 1 46. sddfolca 1 47. difolco 1 48. sddfoco 1 explanation dietary folate for cases standard deviation dietary folate for controls standard deviation APPENDIX C REPORTED CORRELATIONS OF HOMOCYSTEINE AND VITAMINS BS, 312 AND FOLATE 98 mN.0 00.0 0n.0 nv.0 0v.0 Hv.0 mu.0 0m.0 0v.0 00.0 mv.0 H¢.0 nn.0 vN.0 0n.0 «0.0 hn.0 0n.0 wn.0 on.0 an.0 u¢.0 nn.0 0H.0 5H.0 0n.0 0N.0 vn.0 «v.0 MN.0 wn.0 mN.0 v0.0 AOhuhbm mumHoM cam mam 00.H 00.H 00.H 00.H 00.H 00.H 00.H 00.H 00.H 00.H 00.m 00.m 00.m 00.m 00.a 00.N 00.H 00.H 00.H 00.H 00.a 00.H 00.N 00.H 00.H 00.a 00.H 00.m 00.m 00.m 00.m 00.m 00.m mmdmz fin.0 00.H md.0 00.H mn.0 00.H hn.0 00.fi m«.0 00.H 0n.0 00.H mn.0 00.H 0n.0 00.H «v.0 00.H nu.0 00.H Hn.0 00.N MN.0 00.N un.0 00.N 0«.0 00.N bu.0 00.fi 0m.0 00.N 0n.0 00.H bu.0 00.H 00.0 00.H 0m.0 00.H nn.0 00.H bfl.0 00.H mn.0 00.N hH.0 00.H 0H.0 00.H MH.0 00.H 00.0 00.H 0n.0 00.N mn.0 00.N md.0 00.N 0H.0 00.N Hm.0 00.N NN.0 00.N NHfluHUN mmdwz 0d.0 nu.0 NN.0 0N.0 0N.0 on.0 Hv.0 0N.0 mH.0 0H.0 mH.0 0N.0 HN.0 0n.0 on.0 mH.0 om.0 wfluNUN 00.H 00.H 00.H 00.H 00.H 00.a 00.H 00.H 00.H 00.H 00.H 00.H 00.H 00.H 00.H 00.H 00.H fimflmz 00.N 00.N 00.N 00.H 00.H 00.N 00.N 00.H 00.H 00.H 00.N 00.N 00.H 00.H 00.N 00.N 00.fi 00.H 00.H 00.H 00." 00.N 00.N 00.N 00.H 00.H 00.fl 00.H 00.H 00.H 00.H 00.H 00.H 00.H 00. H 00.H 00.N 00.H 00.H 00.N 00.N 00.H 00.N 00.H 00." 00." 00.H 00.N 00.H 00.N 00.H 00.N 00.H 00.N 00.N 00." 00.N 00.H 00.H 00.N 00.N 00.H 00.H 00.N 00.N 00.N 00.H 00.N 00. N 00.0HH 00.HbN 00.Hmm 00.HMN 00.00 00.00H 00.an 00.m¢ 00.0HH 00.00 00.mm 00.0n 00.mmH 00.00N 00.0mm 00.Nm 00.00 00.00 00.nm 00. «F 00.00H 00.0MH 00.vv 00.th 00.¢0n 00.HOH 00.nmn 00.¢wH 00.00 00.00 00.an 00.bmm 00.5N 00.nNH 00.05 00.H 00.m 00.m 00.m 00.H 00.H 00.m 00.m 00.m 00.m 00.m 00.H 00.H 00.m 00.m 00.m 00.m 00.H 00.m 00.m 00.m 00.N 00.H 00.m 00.m 00.m 00.N 00.H 00.N 00.H 00.N 00.H 00.m 00.m 00.N 00. H 00.hm 00.mm 00.0m 00.mm 00.mm 00.Hm 00.Hm 00.Hm 00.Hm 00.0m 00.0w 00.0w 00.0w 00.0w 00.0w 00.0w 00.mN 00.nN 00.HN 00.0N 00.0H 00.mH 00.5H 00.bH 00.0H 00.HH 00.HH 00.HH 00.HH 00.m 00.m 00.m 00.m 00.N 00. N mmgdmo ADONU £82802 mmmsoz .BdumHQ ~59sz ~5me 00.0¢ 00.0v 00.00 00.m¢ 00.H¢ 00.0m 00.0m 00.hm 00.nm 00.mm 00.¢m 00.0m 00.Nm 00.Nm 00.Hm 00.Hm 00.0m 00.0w 00.0w 00.mN 00.0w 00.vN 00.HN 00.HN 00.0w 00.0H 00.¢H 00.¢H 00.¢H 00.0 00.0 00.b 00.0 00.N 00.N .25 .mm mcflEmufl> new mcfimumxuosoz mo mcoflumHmnuou omuHOQmm .o xHQmem¢ 99 00.0 «0.0 00.0 0H.0 QOhuNHm H 00.0 H 00.0 00.H H HN.0 H H H H 00.0 H 5H.0 00.H H 00.0 H 00.0 00.H 004m: .HOhu0m 004m: «Hfln0n 3052 00.0 00.0 00.0 00.H 00.H 00.N 00.0 00.H 00.H 00.H 00.N 00.0 00.H 00.H 00.0 00.N 00.H 00.H 00.H 00.H 00.0 00.0 00.0 00.0 00.H 00.H 00.H 00.H 00.H 00.H 00.H .00.H 00.H 00.H 00.H mmwfiqmo ADOMU “0.20002 mmmduoz 00.H 00.0 00.H 00.H 00.0 00.0 00.H 00.0 00.H 00.0 00.0 00.H 00.0 00.H 00.0 00.H 00.0 00.H 00.N 00.0 00.0 00.0 00.H 00.H 00.0 00.0 00.H 00.H 00.0 00.0 00.N 00.H 00.0 00.0 00.0HH 00.H5N 00.HON 00.H00 00.05 00.00H 00.00H 00.00 00.5HH 00.00 00.00 00.00 00.00H 00.¢0N 00.000 00.00 00.00 00.00 00.00 00.05 00.05H 00.00H 00.0w 00.05H 00.000 00.HOH 00.00N 00.00H 00.05 00.00 00.NOH 00.500 00.50 00.0NH 00.05 00.H BdUmHQ 00.0 00.0 00.0 00.H 00.H 00.0 00.0 00.0 00.0 00.0 00.H 00.H 00.0 00.0 00.0 00.0 00.H 00.0 00.0 00.0 00.0 00.H 00.0 00.0 00.0 00.0 00.H 00.0 00.H 00.N 00.H 00.0 00.0 00.0 00.H mmnzmo AU.UCOUV 00.50 00.00 00.00 00.0w 00.00 00.0v 00.00 00.00 00.00 00.H0 00.H0 00.00 00.H0 00.00 00.H0 00.50 00.H0 00.50 00.00 00.00 00.00 00.00 00.00 00.00 00.0w 00.00 00.0w 00.00 00.0w 00.H0 00.0w 00.H0 00.0N 00.00 00.00 00.0N 00.HN 00.00 00.0N 00.0N 00.0H 00.0w 00.0H 00.0w 00.5H 00.HN 00.5H 00.H0 00.0H 00.00 00.HH 00.0H 00.HH 00.0H 00.HH 00.VH 00.HH 00.0H 00.0 00.0 00.0 00.0 00.0 00.5 00.0 00.5 00.0 00.0 00.N 00.0 momBQd #DH .0 xHQmem< APPENDIX D REPORTED VALUES FOR CODED VARIABLES OF EIGHT STUDIES 100 00.H0 00.N 00.0 00.0 00.N 00.0 00.0 00.00 00.H 00.H 00.0 00.0 00.0 00.0 00.00 00.0 00.N 00.0 00.0 00.0 00.0 00.00 00.H 00.H 00.0 00.0 00.0 00.0 00.00 00.N 00.0 00.00 00.0 00.N 00.H 00.N 00.0 00.5 00.0w 00.H 0N.0 00.55 00.H 00.N 00.H 00.N 00.0 00.5 00.00 00.0H 00.0 00.0 00.H0 0N.0H 00.00 00.0 00.N 00.H 00.H 00.0 00.5 oo.¢m 00.vm 00.H 00.v 00.H0 05.0 00.00 00.H 00.0 00.H 00.H 00.0 00.5 00.0 00.0 00.H0 00.0 00.0 00.H 00.H 00.0 00.0 00.0 00.0 00.N 00.0 00.H 00.H 00.0 00.0 00.0 00.H 00.H 00.0 00.H 00.H 00.0 00.0 00.0 00.0 00.H0 00.0 00.0 00.N 00.H 00.0 00.0 00.0 00.0 00.N 00.0 00.0 00.H 00.0 00.0 00.0H 00.H 00.H 00.0 00.N 00.H 00.0 00.0 00.0 00.5 00.00 00.0 0N.Nm 00.0 00.0 00.0 00.H 00.0 00.0 00.00 00.0H 00.0 00.0 00.0 00.0 00.H 00.0 00.0 oo.NN 00.H0 00.H 00.H 00.0 00.0 00.H 00.0 00.0 00.0H 00.0 00.H0 00.0 00.0 00.0 00.0 00.0 00.0H 00.0 00.H0 00.0 00.0 00.0 00.0 00.0 oo.vH 00.0 00.H0 00.0 00.0 00.0 00.0 00.0 00.0 00.5 00.H0 00.5 00.H0 00.0 00.H 00.0 00.H 00.0 00.v 00.0v 00.0H 00.0 00.N 00.H 00.0 00.H 00.0 00.v 00.00v 00.0HH 00.H 00.H 00.H 00.0 00.H 00.v 00.v 00.0 05.0 00.00 00.0 00.H 00.H 00.H 00.0 00.0 00.5 00.0 00.N 00.H 00.H 00.H 00.0 00.0 00.0H 00.H 00.H 00.H 00.H 00.H 00.0 00.0 00.0 00.0H 00.00 00.0H 00.00 00.0 00.H 00.H 00.H 00.0 00.0 00.0H 00.0H 00.N 00.0 00.H 00.H 00.H 00.0 00.0 00.00 00.00 00.H 00.H 00.H 00.H 00.H 00.0 00.0 00.0 00.0 00.5 00.N 00.N 00.0 00.00 00.0 00.N 00.5 00.0 00.0 00.N 00.05 00.H 00.H 00.5 00.0 00.N 00.N 00.00 00.0 00.0 00.H 00.H 00.H 00.H 00.H 00.Nv 00.H 00.H 00.H 00.H 00.H 00.H 00.H 00.0NH 00.00 00.0 00.0 00.H 00.0 00.H 00.H 00.H 00.HVH 00.00 00.H , 00.H 00.H 00.0 00.H 00.H 00.H 00.20002 mm 04002 “09200 OUU¢Qm oumwd Umnoo ~00 mdem> Uwuuoamm .Q xHQmedd 101 00.0 00.0H 00.H 00.H 00.0 00.0 00.0 00.0H 00.0 00.H 00.0 00.0 00.0 00.0H 00.H 00.H 00.0 00.0 00." 00.0H 00.H 00.H 00.0 00.0 00.HH 00.H" 00.H 00.0 00.H0 00.H 00.H 00.0 00.0 00.0H 00.0” 00.0 00.0 00.0w 00.H 00.H 00.0 00.0 00.00 00.00 00.00 00.H0 00.H 00.n 00.HH 00.0 00.0H 00.0 00.H 00.0 00.0 0N.NH 00.0” 00.00 00.00 00.0 00.0 00.0H 00.0 00.0H 00.H 00.H 00.0 00.0 00.0 00.0H 00.H 00.0 00.~H 00.H 00.H 00.0 00.0 00.H 00.H 00.0 00.0 00.H 00.H 00.0 00.0 0H.0 00.00 00.H 00.0 00.0H 00.H 00.H 00.0 00.0 00.H 00.H 00.0 00.0 00.H 00.H 00.m 00.0 00.00 00.0m 00.0 00.0H 00.H 00.0 00.HH 00.00 00.0H 00.H 00.H 00.0 00.0 00.H 00.0 00.0 00.0 00.H 00.H 00.0 00.0 00.H 00.0H 00.H 00.H 00.0 00.0 00.0 00.0H 00.0 00.H 00.0 00.0 00. 00.0H 00.H 00.H 00.0 00.0 00.n 00.HH 00.0 00.0H 00.~ 00.H 00.0 00.0 00.~ 00.H 00.0 00.0 00.m 00.H 00.0 00.0 00.0 00.0 00.H 00.H 00.0 00.0 . 00.H 00.“ 00.m 00.0 00.H 00.H 00.0 00.0 00.~ 0m.0 00.0 0H.mH 00.H 00.H 00.0 00.0 00.H 00.H 00.m 00.0 00.0 00.H 00.m 00.m 00.00 00.00 00.H 00.H 00.H 00.0 00.0 00.0 00.H 00.H 00.0 00.0 00.m 00.0H 00.H 00.H 00.0 00.0 00.0 0H.0H 00.0 00.H 00.H 00.H «0.0 00.0H 00.0 00.H 00.H 00.H 00.0 00.0 00.0 00.0 00.m 00.H 00.H 00.H 0H.m «0.0 00.0 m0.0 00.0 00.0 00.H 00.H 80QO 00002 «00000 <03: 000: 8000 zoom: «.0QO 5m: 00mm... .020? 0052 03 Guy .0 xHQmemd 102 00.HOH 00.00H 00.00H 00.00H 00.00 00.00 00.H0 00.00 00.000 00.000 00.000 00.500 00.000 00.500 00.000 00.500 0000000 000002 00.00H 00.5HH 00.00 00.00H 00.50H 00.VHH 00.00H 040320 4040000 00.050 00.000 00.000 00.H50 00.500 00.000 00.000 ”0002# £04002 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00002 00.0HH 00.00H 00.H0 00.00H 00.00H 00.05 00.00H 00.HHH 000HQO 00.000 00.000 00.050 00.050 00.000 00.v00 00.000 00.050 000Hm2 00.00H 00.00H 00.55 00.05 00.00H 00.00H 00.00H 00.500 00.000 00.000 00.500 00.000 00.H50 00.000 <00Hmom 400Hmz 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 NHmmZ 10.00000 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.m 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.H 00.0 00.0 00.H 00.H 00.H 00.0 00.H momas< 00H .0 XHszmmd BIBLIOGRAPHY 103 BIBLIOGRAPHY Alfihan G, Pekkanen J, Jauhiainen M, et a1. Relation of serum homocysteine and lipoprotein(a) concentrations to atherosclerotic disease in a prospective Finnish population based study. Atherosclerosis 1994;106:9-19. Andersson A, Brattstrom L, Israelsson B, Isaksson A, Hamfelt A, Hultberg B. Plasma homocysteine before and after methionine loading with regard to age, gender, and menopausal status. Eur J Clin Invest 1992;22:79-87. Araki A, Sako Y, Fukushima Y, Matsumoto M, Asada T, Kita T. Plasma sulfhydryl- containing amino acids in patients with cerebral infarction and in hypertensive patients. Atherosclerosis 1989;79:139-46. Amadottir M, Brattstrom L, Simonsen O, et al. The effect of high-dose pyridoxine and folic acid supplementation on serum lipid and plasma homocysteine concentrations in dialysis patients. Clin Nephrol 1993;40:236-40. Arnesen E, Refsum H, Bonaa KH, Ueland PM, Forde OH, Nordrehaug JE. Serum total homocysteine and coronary heart disease. Int J Epidemiol 1995;24:704-9. Aronson DC, Onkenhout W, Raben AM, Oudenhoven LP, Brommer EJ, Van Bockel JH. Impaired homocysteine metabolism: a risk factor in young adults with atherosclerotic arterial occlusive disease of the leg. Br J Surg 1994;81:1114-8. Bergmark C, Mansoor MA, Swedenborg J, De Faire U, Svardal AM, Ueland PM. Hyperhomocysteinemia in patients operated for lower extremity ischaemia below the age of 50 - effect of smoking and extent of disease. Eur J Vasc Surg 1993;7z391-6. Bostom AG, Yanek L, Hume AL, Eaton CB, McQuade W, Nadeau M, Perrone G, Jacques PF, Selhub J. High dose ascorbate supplementation fails to affect plasma homocysteine levels in patients with coronary heart disease. Atherosclerosis 1994;111: 267-70. Boushey JC, Beresford AAS, Omenn SG, Motulsky GA. A quantitative assessment of plasma homocysteine as a risk factor for vascular disease. JAMA 1995;274: 1049-57. Bowers C. Folate and neural tube defects. Nutr Rev 1995;53: S33-8. Brattstrom L, Lindgren A. Hyperhomocysteinemia as a risk factor for stroke. Neurol Res 1992;14:81-4. Brattstrom L, Israelsson B, Hultberg B. Plasma homocysteine and methionine tolerance in early- onset vascular disease. Haemostasis 1989;19: Suppl 1:35-44. 104 Brattstrom L, Israelsson B, J eppsson JO. Folic acid - an innocuous means to reduce plasma homocysteine. Scand J Clin Lab Invest 1988;48:215-21. Brattstrom L, Israelsson B, Lindgarde F, Hultberg B. Higher total plasma homocysteine in vitamin B,2 deficiency than in heterozygosity for homocysteinuria due to cystathionine beta- synthase deficiency. Metabolism 1988;37:175-8. Brattstrom L, Israelsson B, Norrving B, et al. Impaired homocysteine metabolism in early-onset cerebral and peripheral occlusive arterial disease. Effects of pyridoxine and folic acid treatment. Atherosclerosis 1990;81:51-60. Brattstrom L, Lindgren A, Israelsson B, Andersson A, Hultberg B. Homocysteine and cysteine: determinants of plasma levels in middle-aged and elderly subjects. J Intern Med 1994;236:633-41. Brattstrom L, Lindgren A, Israelsson B, et al. Hyperhomocysteinaemia in stroke: prevalence, cause, and relationships to type of stroke and stroke risk factors. Eur J Clin Invest 1992;22:214-21. Brattstrom L. Vitamins as homocysteine-lowering agents. J Nutr 1996;26:12768-1280S. Cacan SL, Xhignesse M, Piolot A, Selhub J, Davignon J, Genest J. Plasma total homocysteine in healthy subjects; sex-specific relation with biological traits. Am J Clin Nutr 1996;64:587-93. Broekman MJ, Hajjar KA, Marcus AJ, et al. Homocysteine inhibits ecto-ADPase activity of human umbilical vein endothelial cells. (Abstr) J Clin Invest 1994;67:77a. Cacan SL, Xhignesse M, Piolot A, Selhub J, Davignon J, Genest J. Plasma total homocysteine in healthy subjects; sex-specific relation with biological traits. Am J Clin Nutr 1996;64:587-93. Carlson NAJ, Neill DW. Metabolic abnormalities detected in a survey of mentally backward individuals in Northern Ireland. Archives of Disease in Childhood 1962;37:505-9. Chadefaux B, Cooper BA, Gilfix BM, et al. Homocysteine: relationship to serum cobalamin, serum folate, erythrocyte folate, and lobation of neutrophils. Clin Invest Med 1994;17:540-50. Chasan Taber L, Selhub J, Rosenberg 1H, et al. A prospective study of folate and vitamin B, and risk of myocardial infarction in US physicians. J Am Coll Nutr 1996;15:136-43. 105 Chauveau P, Chadefaux B, Coude M, Aupetit J, Kamoun P, Jungers P. Long-tenn folic acid (but not pyridoxine) supplementation lowers elevated plasma homocysteine level in chronic renal failure. Miner Electrolyte Metab 1996;22:106-9. Chauveau P, Chadefaux B, Coude M, et a1. Hyperhomocysteinemia, a risk factor for atherosclerosis in chronic uremic patients. Kidney Int Suppl 1993;41:872-7. Chu RC, Hall CA. The total semm homocysteine as an indicator of vitamin Bl2 and folate status. Am J Clin Pathol 1988;90:446-9. Clarke R, Daly L, Robinson K, Naughten E, Cahalane S, Fowler B, Graham I. Hyperhomocysteinemia; an independent risk factor for vascular disease. N Engl J Med 1991;324:1149-55. Combs F G. The Vitamins: Fundamental aspects in nutrition and health. New York: Academic Press, 1992. Cooper H, Hedges L. The Handbook of Research Synthesis. New York: Russell Sage Foundation, 1994. Coull BM, Malinow MR, Beamer N, Sexton G, Nordt F, Garmo PD. Elevated plasma homocyst(e)ine concentration as a possible independent risk factor for stroke. Stroke 1990;21 :572-6. Cravo ML, Gloria LM, Selhub J, et al. Hyperhomocysteinemia in chronic alcoholism: correlation with folate, vitamin Ba, and vitamin B, status. Am J Clin Nutr 1996;63:220- 4. Curtis D, Sparrow R, Brennan L, VanderWeyden MB. Elevated serum homocysteine as a predictor for vitamin B,2 or folate deficiency. Eur J Haematol 1994;52:227-32. Dalery K, Lussier Cacan S, Selhub J, Davignon J, Latour Y, Genest J Jr. Homocysteine and coronary artery disease in French Canadian subjects: relation with vitamins 8.2. B6, pyridoxal phosphate, and folate. Am J Cardiol 1995;75:1107-11. Dittrnan WA, Majerus WP. Structure and function of thrombomodulin: a natural anticoagulant. Blood 1990;75:329-36. Dudman BPN, Guo X, Gordon BR, Dawson AP, Wilcken LED. Human homocysteine catabolism: three major pathways and their relevance to development of arterial occlusive disease. J Nutr 1996;126:12958-3028. 106 Dudman BPN, Wilcken DE, Wang J, Lynch JF, Macey D, Lundberg P. Disordered methionine/homocysteine metabolism in premature vascular disease. Its occurrence, cofactor therapy, and enzymology. Arterioscler Thromb 1993;13: 1253-60. Ellis JM, McCully KS. Prevention of myocardial infarction by vitamin 8,. Res Commun Mol Pathol Pharrnacol 1995;89:208-20. Fermo I, D’Angelo SV, Paroni R, Mazzola G, Calori G, D’Angelo A. Prevalence of moderate hyperhomocysteinemia in patients with early-onset venous and arterial occlusive disease. Ann Intern Med 1995;123:747-53. Food and Nutrition Board. Recommended Dietary Allowances. 10th ed. Washington, DC: National Academy Press, 1989. Franken DG, Boers GH, Blom HJ, Trijbels JM. Effect of various regimens of vitamin B, and folic acid on mild hyperhomocysteinaemia in vascular patients. J Inherit Metab Dis 1994;17:159-62. Friedrich, W. Vitamins. Berlin, New York: St Martin’s Press, 1988. Garry PJ, Goodwin JS, Hunt WC. Folate and vitamin B,2 status in a healthy elderly population. JAGA 1984;32:719-26. Genest JJ, McNamara JR, Salem DN, et al. Plasma homocyst(e)ine levels in men with premature coronary artery disease. J Am Coll Cardiol 1990;16:1114-9. Giles WH, Kittner SJ, Anda RF, Croft JB, Casper ML Serum folate and risk for ischemic stroke. First National Health and Nutrition Examination Survey Epidemiologic Follow-up Study. Stroke 1995;26:1166-70. Glass GV. Primary, Secondary, and Meta-analysis of Research. Educational Research, 1976. Glusker PJ. Vitamin BI2 and the B,2 Coenzymes. In Vitamins and Hormones. New York: Academic Press, 1995. Gruberg RE, Raymond AS. Beyond Cholesterol. New York: St Martin’s Press, 1981. Guttormsen AB, Mansoor AM, Fiskerstrand T, Ueland PM, Refsum H. Kinetics of plasma homocysteine in healthy subjects after peroral homocysteine loading. Clin Chem 1993;39:1390-7. 107 Guttormsen AB, Schneede J, Fiskerstrand T, Ueland PM, Refsum HM. Plasma concentrations of homocysteine and other aminothiols compounds are related to food intake in healthy human subjects. J Nutr 1994;124: 1934-41. Guttormsen AB, Schneede J, Ueland PM, Refsum H. Kinetics of total plasma homocysteine in subjects with hyperhomocysteinemia due to folate or cobalamin deficiency. Am J Clin Nutr 1996;63:194-202. Hajjar KA. Homocysteine-induced modulation of tissue plasminogen activator binding to its endothelial cell membrane receptor. J Clin Invest 1993;91:2873-9. Hajjar KA. The endothelial cell tissue plasminogen activator receptor. J Biol Chem 1991;266:21962-70. Hajjar KA, Hamel MN, Harpel CP, Nachman LR Binding of tissue plasminogen activator to cultured human endothelial cells. J Clin Invest 1987;80:1712-9. Hall CA, Chu RC. Serum homocysteine in routine evaluation of potential vitamin B,2 and folate deficiency. Eur J Haematol 1990;45:143-9. Harpel CP, Chang TV, Borth W. Homocysteine and other sulfhydryl compounds enhance the binding of lipoprotein(a) to fibrin: A potential biochemical link between thrombosis, atherogenesis, and sulthydryl compound metabolism. Proc Natl Acad Sci 1992;89:10193- 7. Harpel CP, Zhang X., Borth W. Homocysteine and hemostasis: Pathogenetic mechanisms predisposing to thrombosis. J Nutr 1996;126:12858-98. Hedges VL, Olkin I. Statistical Methods for Meta-Analysis. New York: Academic Press, 1985. Heijer MD, Blom HJ, Gerrits WB, et al. Is hyperhomocysteinemia a risk factor for recurrent venous thrombosis? Lancet 1995;345:882-5. Heijer MD, Koster T, Blom HJ. Hyperhomocysteinemia as a risk factor for deep-vein thrombosis. N Engl J Med 1996;334:759-62. Herzlich BC, Lichstein E, Schulhoff N, et a1. Relationship among homocyst(e)ine, vitamin Bl2 and cardiac disease in the elderly: association between vitamin B,2 deficiency and decreased left ventricular ejection fraction. J Nutr 1996;126:1249S-53S. Hopkins PN, Wu LL, Wu J, et al. Higher plasma homocyst(e)ine and increased susceptibility to adverse effects of low folate in early familial coronary artery disease. Arterioscler Thromb Vasc Biol 1995;15:1314-20. 108 Hultberg B, Berglund M, Anderson A., Frank A. Elevated plasma homocysteine in alcoholics. Alcohol Clin Exp Res 1993;17:687-9. Israelsson B, Brattstrom L, Refsum H. Homocysteine in frozen plasma samples. A short cut to establish hyperhomocysteinemia as a risk factor for arteriosclerosis? Scand J Clin Invest 1993;53:465-9. Israelsson B, Brattstrom LE, Hultberg BL. Homocysteine and myocardial infarction. Atherosclerosis 1988;71:227-33. Jacob RA, Wu MM, Henning SM, Swendseid ME. Homocysteine increases as folate decreases in plasma of healthy men during short-term dietary folate and methyl group restriction. J Nutr 1994;124:1072-80. Jacobsen DW, Gatautis VJ, Green R, et a]. Rapid HPLC determination of total homocysteine and other thiols in serum and plasma: sex difi'erences and correlation with cobalamin and folate concentrations in healthy subjects. Clin Chem 1994;40:873-81. Jacques PF, Bostom AG, Williams RR, et al. Relation between folate status, a common mutation in methyltetrahydrofolate reductase, and plasma homocysteine concentrations. Circulation 1996;93:7-9. J oosten E, Pelemans W, Devos P, et a1. Cobalamin absorption and serum homocysteine and methylmalonic acid in elderly subjects with low serum cobalamin. Eur J Haematol 1993;51:25-30. Joosten E, Van den Berg A, Riezler R, et a1. Metabolic evidence that deficiencies of vitamin B,2 (cobalamin), folate, and vitamin B, occur commonly in elderly people. Am J Clin Nutr 1993;58:468-76. Kang S. Treatment of hyperhomocysteinemia: Physiological basis. J Nutr 1996;126: 1273S-5S. Kang SS, Wong PW, Norusis M. Homoeysteinemia due to folate deficiency. Metabolism 1987;36:458-62. Landgren F, Israelsson B, Lindgren A, Hultberg B, Andersson A, Brattstrom L. Plasma homocysteine-lowering effect of folic acid. J Intern Med 1995;237:381-8. Lentz SR, Sadler EJ. Inhibition of thrombomodulin surface expression and protein C activation by the thrombogenic agent homocysteine. J Clin Invest 1991 ;88:1906-1914. Lewis CA, Pancharuniti N, Sauberlich HE. Plasma folate adequacy as determined by homocysteine level. Ann N Y Acad Sci 1992;669:360-2. 109 Lindenbaum J, Rosenberg IH, Wilson PWF, Stabler SP, Allen RH. Prevalence of cobalamin deficiency in the Framingham elderly population. Am J Clin Nutr 1994;60:2- 11. Loehrer FM, Haefeli WE, Angst CP, Browne G, Frick G, Fowler B. Effect of methionine loading on 5-methyltetrahydrofolate, S-adenosylrnethionine and S-adenosylhomocysteine in plasma of healthy humans. Clin Sci Colch 1996;91:79-86. Lussier Cacan S, Xhignesse M, Piolot A, Selhub J, Davignon J, Genest J Jr. Plasma total homocysteine in healthy subjects: sex-specific relation with biological traits. Am J Clin Nutr 1996;64:587-93. Malinow MR, Kang SS, Taylor LM, et al. Prevalence of hyperhomocyst(e)inemia in patients with peripheral arterial occlusive disease. Circulation 1989;79:1180-8. Malinow RM. Plasma homocysteine: A risk factor for arterial occlusive diseases. J Nutr 1996;126:12388-438. Manssor AM, Ueland MP, Aarsland A,Svarda1 MA. Redox status and protein binding of plasma homocysteine and other aminothiols in patients with homocysteinuria. Metabolism 1993;42:1481-5. Mason JB, Miller JW. The effects of vitamins B”, B,, and folate on blood homocysteine levels. Ann. N Y Acad Sci 1992;669:197-204. Matthews JH. Cobalamin and folate deficiency in the elderly. Baillieres Clin Haematol 1995;8:679-97. MeCully KS. Chemical pathology of homocysteine. Atherogenesis. Ann Clin Lab Sci 1993;23:477-93. Miller WJ, Ribaya Mercado DJ, Russel MR, et al. Effect of vitamin B, deficiency on fasting plasma homocysteine concentrations. Am J Clin Nutr 1992;55:1154-60. Molgaard J, Malinow MR, Lassvik C, Helm AC, Upson B, Olsson AG. Hyperhomocysteinemia: an independent risk factor for intermittent claudication. J Int Med 1992;231:273-9. Naurath HJ, Joosten E, Riezler R, Stabler SP, Allen RH, Lindenbaum J. Effects of vitamin Bu. folate, and vitamin B, supplements in elderly people with normal serum vitamin concentrations. Lancet 1995;346:85-9. 110 Nilsson K, Gustafson L, Faldt R, Andersson A, Hultberg B. Plasma homocysteine in relation to serum cobalamin and blood folate in a psychogeriatric population. Eur J Clin Invest 1994;24:600-6. Nilsson K, Gustafson L, Faldt R, et al. Hyperhomocysteinemia - a common finding in a psychogeriatric population. Eur J Clin Invest 1996;26:853-9. Nishinaga M, Ozawa T, Shimada K. Homocysteine, a thrombogenic agent, suppresses anticoagulant heparan sulfate expression in cultured porcine aortic endothelial cells. J Clin Invest 1993;92:1381-6. Nygard O, Refsum H, Ueland MP, Vollset SE. Major determinants of plasma total homocysteine distribution: the Hordland homocysteine study. Am J Clin Nutr 1998;67:263-70. Nygard O, Vollset SE, Refsum H, et al. Total plasma homocysteine and cardiovascular risk profile. JAMA 1995;274:1526-33. Olszewski AJ, Szostak WB, Bialkowska M, Rudnicki S, McCully KS. Reduction of plasma lipid and homocysteine levels by pyridoxine, folate, cobalamin, choline, riboflavin, and troxerutin in atherosclerosis. Atherosclerosis 1991;88:978-82. Pancharuniti N, Lewis CA, Sauberlich HE, et al. Plasma homocyst(e)ine, folate, and vitamin B,2 concentrations and risk for early-onset coronary artery disease. Am J Clin Nutr 1994;59:940-8. Pennypacker LC, Allen RH, Kelly JP, et al. High prevalence of cobalamin deficiency in elderly outpatients. J Am Geriatr Soc 1992;40:1197-204. Perry U, Refsum H, Morris RW, Ebrahim SB, Ueland PM, Shaper AG. Prospective study of serum total homocysteine concentration and risk of stroke in middle-aged British men. Lancet 1995;346:1395-99. Refsum H, Helland S, Ueland PM. Fasting plasma homocysteine as a sensitive parameter of antifolate effect: a study of psoriasis patients receiving low-dose methotrexate treatment. Clin Pharrnacol Ther 1989;46:510-20. Refsum H, Nygard O, Kvale G, Ueland MP, Vollset ES. The Hordland homocysteine study: The opposite tails odds ratios reveal differential effects of gender and intake of vitamin supplements at high and low plasma total homocysteine concentrations. J Nutr 1996; 126: 12448-88. Petitti BD. Meta Analysis, Decision Analysis and Cost-effectiveness Analysis. London: Oxford University Press, 1994. 111 Riggs KM, Spiro A III, Tucker K, Rush D. Relations of vitamin B”, vitamin B,, folate, and homocysteine to cognitive performance in the Normative Aging Study. Am J Clin Nutr 1996;63:306-14. Rimm BE, Willet CW, Hu BF. Folate and vitamin B, from diet and supplements in relation to risk of coronary heart disease among women. JAMA 1998;279:359-64. Robinson K, Gupta A, Dennis V, et al. Hyperhomocysteinemia confers an independent increased risk of atherosclerosis in end-stage renal disease and is closely linked to plasma folate and pyridoxine concentrations. Circulation 1996;94:2743-8. Robinson K, Mayer EL, Miller DP, et al. Hyperhomocysteinemia and low pyridoxal phosphate. Common and independent reversible risk factors for coronary artery disease. Circulation 1995;92:2825-30. Robinson K, Mayers E,Jacobsen DW. Homocysteine and coronary artery disease. Clin J Med 1994;61:438-50. Rosenberg DR, Rosenberg SJ. Natural anticoagulant mechanisms. J Clin Invest 1984;74:1-6. Rosenberg IH. Homocysteine, vitamins and arterial occlusive disease: An overview. J Nutr 1996;126:12358-78. Rosendaal FR, Koster T, Vandenbrouke JP, Reitsma PH. High risk of thrombosis in patients homozygous for factor V Leiden (Activated Protein C resistance). Blood 1995;85:1504-8. Selhub J, Jacques PF, Bostom AG, et al. Association between plasma homocysteine concentrations and extracranial carotid-artery stenosis. N Engl J Med 1995;332:286-91. Selhub J, J acques PF, Bostom AG, et al. Relationship between plasma homocysteine, vitamin status and extracranial carotid-artery stenosis in the Framingham Study population. J Nutr 1996;126:12588-658. Selhub J, Jacques PF, Wilson PW, Rush D, Rosenberg IH. Vitamin status and intake as primary determinants of homocysteinemia in an elderly population . J AMA 1993;270: 2693-8. Selhub J ,Miller WJ. The pathogenesis of homocysteinemia: interruption of the coordinate regulation by S-adenosylmethionine of the remethylation and transsulfuration of homocysteine. Am J Clin Nutr 1992;55:131-8. 112 Stabler SP, Lindenbaum J, Allen R. The use of homocysteine and other metabolites in the specific diagnosis of vitamin B,2 deficiency. J Nutr 1996;126:1266S-728. Stabler SP, Lindenbaum J, Savage DG, Allen RH. Elevation of serum cystathionine levels in patients with cobalamin and folate deficiency. Blood 1993;81:3404-13. Stabler SP, Marcell PD, Podell ER, Allen RH, Savage DG, Lindenbaum J. Elevation of total homocysteine in the serum of patients with cobalamin or folate deficiency detected by capillary gas chromatography-mass spectrometry. J Clin Invest 1988;81:466-74. Stampfer MJ, Malinow MR, Willet WC, et al. A prospective study of plasma homocysteine and risk of myocardial infarction in US physicians. JAMA 1992;7z877-80. Steegers Theunissen RP, Boers GH, Trijbels FJ, et al. Maternal hyperhomocysteinemia: a risk factor for neural-tube defects? Metabolism 1994;43:1475-80. Sumner AE, Chin MM, Abrahm JL, et al. Elevated methylmalonic acid and total homocysteine levels show high prevalence of vitamin B,2 deficiency after gastric surgery. Ann Intern Med 1996;124:469-76. Svensson PJ, Dahlback B. Resistance to activated protein C as a basis for venous thrombosis. N Eng J Med 1993;330:517-22. Swifi ME, Shultz TD. Relationship of vitamins B, and B,2 to homocysteine levels: risk for coronary heart disease. Nutr Reports Inter 1986;34:1-14. Taylor LM Jr., DeFrang RD, Harris EJ Jr., Porter JM. The association of elevated plasma homocysteine with progression of symptomatic peripheral arterial disease. J Vasc Surg 1991;13:128-36. Ubbink JB, Delport R, Verrnaak H. Plasma homocysteine concentrations in a population with low coronary heart disease prevalence. J Nutr 1996;126:12548-7S. Ubbink JB, VanderMerwe A, Delport R, Allen RH, Stabler SP, Riezler R, Verrnaak WJ. The effect of a subnormal vitamin B, status on homocysteine metabolism. J Clin Invest 1996;98:177-84. Ubbink JB, VanderMerwe A, Verrnaak WJ, Delport R. Hyperhomocysteinemia and the response to vitamin supplementation. Clin Investig 1993;71:993-8. Ubbink JB, Verrnaak WJ, Delport R, VanderMerwe A, Becker PJ, Potgieter H. Effective homocysteine metabolism may protect South African blacks against coronary heart disease. Am J Clin Nutr 1995;62: 802-8. 113 Ubbink JB, Verrnaak WJ, VanderMerwe A, Becker PJ, Delport R, Potgieter HC. Vitamin requirements for the treatment of hyperhomocysteinemia in humans. J Nutr 1994;124: 1 927-3 3. Ubbink JB, Verrnaak WJ, VanderMerwe A, Becker PJ. Vitamin B,,, vitamin B,, and folate nutritional status in men with hyperhomocysteinemia. Am J Clin Nutr 1993;57:47- 53. Ubbink JB. Vitamin nutrition status and homocysteine: an atherogenic risk factor. Nutr Rev 1994;52:383-7. Ueland MP, Mansoor AM, Guttormsen BA, Muller F, Aukrust P, Refsum H, Svardal MA. Reduced, oxidized and protein-bound forms of homocysteine and other aminothiols in plasma - comprise the redox thiol status - a possible element of the extracellular antioxidant defense system. J Nutr l996;26:128lS-4S. Upchurch RG, Welch NG, Loscalzo J. Homocysteine, EDRF and endothelial function. J Nutr 1996;126:12908-4S. VandenBerg M, Franken DG, Boers GH, et al. Combined vitamin B, plus folic acid therapy in young patients with arteriosclerosis and hyperhomocysteinemia. J Vasc Surg 1994;20:933-40. Verhoef P, Hennekens CH, Allen RH, Stabler SP, Willet WC, Stampfer MJ. Plasma homocysteine and risk of angina pectoris: results from a prospective study. Presented at the International Conference on Homocysteine Metabolism: From Basic Science to Clinical Medicine. Abstract. Irish J Med Sci 1995;164:26. Verhoef P, Hennekens CH, Malinow MR, Kok FJ, Willet WC, Stampfer MJ. A prospective study of plasma homocysteine and risk of ischemic stroke. Stroke 1994;25:1924-30. Verhoef P, Stampfer MJ, Buring JE, et al. Homocysteine metabolism and risk of myocardial infarction: relation with vitamins B,, B,,, and folate. Am J Epidemiol 1996;143:845-59. Verhoef P, Stampfer MJ. Prospective studies of homocysteine and cardiovascular disease. Nutr Rev 1995;53:283-8. Wacher WK, Straf LM. The future of Meta-Analysis. New York: St Martin’s Press, 1990. Wilcken B, Turner B. Homocysteinuria. Reduced folate levels during pyridoxine treatment. Arch Dis Child 1973;48: 58-62. 114 Wilcken DE, Dudman NP, Tyrrell PA, Robertson MR. Folic acid lowers elevated plasma homocysteine in chronic renal insufficiency: possible implications for prevention of vascular disease. Metabolism 1988;37:697-701. Wolf MF. Meta-Analysis: Quantitative Methods for Research Synthesis. California: Sage Publications, 1986. Woutcrs MG, Boers GH, Blom HJ, et al. Hyperhomocysteinemia: a risk factor in women with unexplained recurrent early pregnancy loss. Fertil Steril 1993;60:820-5. Wu LL, Wu J, Hunt SC, et a1. Plasma homocysteine as a risk factor for early familial coronary artery disease. Clin Chem 1994;40:552-61. Yao Y, Yao SL, Yao SS, Yao G, Lou W. Prevalence of vitamin B,2 deficiency among geriatric outpatients. J Fam Pract 1992;35:524-8.