- Email: [email protected]
Changes in plasma vitamin levels in the first 48 hours after onset of acute myocardial infarction
Changes in Plasma Vitamin Levels in the First 48 Hours After Onset of Acute Myocardial Infarction Robert Scragg, MBBS, PhD, Rodney Jackson, MB, ChB, Ian Holdaway, MD, FRACP, Gerald Woollard, MSc, PhD, and David Woollard, MSc, PhD
To establish whether plasma vitamin measuremsnts made after acute myocardial infarction (AMI) can be used in case-control studies of coronary artery disease, the short-term effect of AMI Ds, on plasma concentrations of 25-hydroxyvitamin beta-carotene, vitamin E and retinol was investigated. Gquentlal measures of these vitamins were made during the first 48 hours after AMI in 13 patients admitted to the hospital within 4 hours after the onset of symptoms. Plasma levels of 2!5hydroxyvltamin D did not change significantly during the first 12 hours after onset of symptoms. Ma-carotene levels increased signiffcantly (p
here is continuing interest in a possible role for vitamins in the etiology of coronary artery disease. Serum 25hydroxyvitamin D3 levels have been observed to be lower in acute myocardial infarction (AMI) patients than in control subjects.lT2Vitamin E inhibits platelet aggregation,3and blood vitamin E, A and carotenoid levelscorrelate with blood cholesterol,4-6 although a nonsignificant increase in serum vitamin A has been reported in heart diseasepatients compared to control subjects.’ Low blood levels of vitamins in winter8y9could be related to the known seasonalvariation in coronary artery disease mortality, which is typically 30 to 40% higher in winter than in summer.1oCase-controlstudies, which typically measureexposureat the time of disease occurrence, are potentially suitable for studying shortterm risk factors including associationsbetweenseasonal variations in blood vitamin levels and coronary artery diseaserisk, provided blood levels of vitamins are not affected by the acute phase reaction that occurs after AMI. This study establisheswhether plasma vitamin levels are affected by AM1 in the short term. Sequential blood sampleswere collected during the first 48 hours after the onset of symptomsof AMI. The vitamins chosenfor measurementwere retinol, beta-caroteneand vitamin E, becausethese are being measuredin the World Health Organization Monitoring of Trends and Determinants in Cardiovascular DiseaseStudy,” and 25-hydroxyvitamin D, becausethis vitamin may be related to the seasonal variation in coronary artery disease mortality12 and is probably the most useful vitamin D metabolite for evaluation of overall vitamin D status.13
T
METHODS
From the Department of Community Health, University of Auckland; the Departments of Endocrinology and Clinical Chemistry, Auckland Hospital; and the Lynfield Agricultural Center, Auckland, New Zealand. This study was supported in part by Roche Prcducta, New Zealand. Dr. Jackson is a National Heart Foundation of New Zealand Research Fellow. Manuscript received December 16, 1988; revised manuscript received and accepted July 13, 1989. No reprints will be available.
Nonfasting blood sampleswere collected immediately after admission to the hospital, at 8 hourly intervals during the subsequent24 hours and again at 48 hours, in 13 patients with proven AM1 admitted within 4 hours of onset of chest pain. The sampleswere all taken with patients in the supine position. Plasma samples were separatedwithin 4 hours of collection and stored at -20” C until analyzed. All patients were admitted to a coronary care unit in Auckland, New Zealand, between June 1985 and January 1986. AM1 was diagnosedon the basis of a history of prolonged chest pain, elevated cardiac enzymes and evolving electrocardiographic changes.14There were 11 men and 2 women,
THE AMERICAN JOURNAL OF CARDIOLOGY NOVEMBER 1,1989
971
-
aged 47 to 66 years (mean age 57). The study was approved by the Auckland Hospital Ethics Committee. Retinol, vitamin E and beta-carotenewere measured simultaneously from samples from all 13 subjects by high performance liquid chromatography. Although these 3 compoundshave differing chromatographic and spectral properties and in the past have been measured separately,careful choice of mobile phaseand the availability of programmable multiwavelength detectors allowed their measurement in a single injection sample using high performance liquid chromatography. Extraction of the 3 components from plasma followed an established procedure15 involving protein precipitation with methanol of samples plus internal standards and subsequentextraction into hexane. The hexane was then evaporatedunder nitrogen and the concentrated extract taken up in ethanol ready for injection. Chromatography was performed on a Novapak Cl8 radially compressed column. The mobile phase was composed of 45% acetonitrile, 45% methanol and 10%dichloromethane. This mixture allows good separation of all the measured componentsin a convenient time without the need for gradient elution and is a modification of a previously published method.I6 The individual components were detected in a single chromatographic run by the use of programmed multiple wavelength changes(325 nm for retinol, 292 nm for alpha-tocopherol and 465 nm for beta-carotene) against their respective internal stan-
dards. This detection method closely resemblesthat a Milne and Botnen.r7 For logistic reasons,2 different methods were used to measure25-hydroxyvitamin Ds. Method A was used for 8 patients (patients 1 to 8 in Figure l), while method B was used for 4 more patients (patients 9 to 12 in Figure 1). The same method was always used to measure vitamin D levels in all 5 samplescollected from any individual patient. Method A used extraction of plasma with ethanol’* and fractionation with sephadexLH-20. Assay was by competitive protein binding using human serum as binding protein and separation with dextrancharcoal. The sensitivity of the assay was 12.5 nmol/ liter and the within-assay and between-assaycoefficients of variation were f 9 and f 22% respectively. Method B involved dilution of replicate plasma samples (500 ~1) with an equal volume of water followed by addition of 25 &labeled 25-hydroxycholecalciferol to measurehormone recovery. Protein was removedby addition of acetonitrile, the supernatant was mixed with 500 ~1 of 0.4 M phosphate buffer (pH 10) and passed through a Cl8 Sep-pak cartridge (Waters Associates), discarding the eluate. After washing, 25-hydroxyvitamin D3 was eluted from the cartridge with 3 ml acetonitrile, evaporated and stored at -20’ C. The analysis involved reconstitution with 1 ml ethanol followed by competitive protein binding assay using bovine serum and separation with dextran-coated charcoal.19 The
Number
0’
0’8
16
24
Hours after admission 972
THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 64
1 48
ir i‘ TABLE 1 Vitamin
I Mean Plasma Vitamin Concentrations -
Mean 25.hydroxyvitamin D (nmol/liter) Mean of differences f SE* Percentage change* Beta-carotene (pmol/liter) mean Mean of differences f SE Percentage change Vitamin E (pmol/liter) mean Mean of differences f SE Percentage change Retinol (pmol/liter) mean Mean of differences f SE Percentage change
During the First 48 Hours After Admission for Acute Myocardial
Infarction
0 hours
8 hours
16 hours
24 hours
48 hours
103
103
98
89
89
-14i8 -13.6 0.532
-15f7 -14.6 0.448
0.545
36.9
2.03
Of3 0 0.574
-5*5 -4.9 0.556
0.029 f 0.012+ 5.3 34.4
0.012 f 0.028 2.2 33.4
-2.6 f 1.6 -7.0 1.88
-3.5 f 2.4 -9.5 1.78
-0.14 f 0.04’ -6.9
-0.24 f 0.06’ -11.8
Diierence between means is not always equal to mean of difference because of rounding errors. * Compared wrth concentration on admission (i.e.. 0 hours); 7 p <0.05; f p
-0.013 f 0.033 -2.4 31.1
-0.097 f 0.045 -17.8 27.4
-5.9 f 2.3’ -16.0 1.76
-9.6 f 3.6’ -26.0 1.53
-0.27 f 0.06* -13.3
-0.50 f 0.08% -24.6
of means.
the first 12 hours after the onset of symptoms of AM1 (i.e., over the first 8 hours after admission). In contrast, plasma beta-carotenelevels are elevatedduring the first 20 hours, and then decrease,whereasconcentrations of vitamin E and retinol progressivelydecreaseduring the first 48 hours after AMI. There appear to be no previous studies that have systematically examined changes RESULTS Mean plasma vitamin levels at each of the collection in blood vitamin levels during the first 48 hours after times are listed in Table I, along with the mean of the AMI, although a Danish study reported no change in difference from baseline and the percentage change levels of 25-hydroxyvitamin D in samples taken daily compared with baseline. Mean plasma levels of 25- for 4 days after AMI,* while a Swedish study found a hydroxyvitamin Ds at baselineand 8 hours were similar, significant decreasein vitamin E during the first week but then progressively decreased to 14.6% less than after AMI.** It is unlikely that these changes in plasma vitamin baseline at 48 hours. The trend in the means was similar for patients measured by each vitamin D method. levels are due to changes in hemoconcentration after The mean level of beta-carotene increased significantly AM1 becausethey did not vary in the same direction (5.3%, p <0.05) at 8 hours and thereafter progressively during the first 16 hours after admission. Neither does decreased.Mean level of vitamin E decreasedby 7.0% change in dietary intake after AM1 seema likely explaat 8 hours, and at 24 hours was 16.0%lessthan baseline nation for these results. While blood levels of vitamin (p <0.05). Retinol decreased significantly (6.9%, p E and beta-carotene are both strongly influenced by
THE AMERICAN JOURNAL OF CARDIOLOGY NOVEMBER I,1989
973
PLASMA VITAMINS AITER MYDCARDIAL INFARCTION
zc $ & 6
0.80 -
0.60
-
0’
a 0
8 Hours
16 after
24
I 48
admissions
be influenced by alterations in liver metabolism as may occur in the acute reaction to stress. This study showsthat plasma levels of beta-carotene, retinol and vitamin E change rapidly during the first 48 hours after AMI, which invalidates their measurement over this time period in casecontrol studies. By comparison, the similarity in 25-hydroxyvitamin D levels between the 2 samplestaken during the first 12 hours after onset of AM1 symptomssuggeststhat levels in samples taken during this time are likely to be similar to blood concentrations before AMI, and might be used as a proxy for levels before AM1 in case-control studies.
12. Scragg R. Seasonality of cardiovascular diseasemortality and the possible protective effect of ultra-violet radiation. Inr J Epidemiol 1981;10:337-341. 13. Bikle DD, Gee E, Halloran B, Kowalski MA, Ryzen E, Haddad JG. Assessment of the free fraction of 25-hydroxyvitamin D in serum and its regulation by albumin and the vitamin D binding protein. J Clin Endocrinol Metab 1986,63; 954-959. 14. World Health Organization working Party Report. IschaemicHeart Disease Registers. Part 2. Genes World Health Organization, 1969. 15. Bieri JG, Tolliver TJ, Catignani GL. Simultaneousdetermination of alphatocopheroland retinol in plasmaor red cells by high pressureliquid chromatography. Amer J Clin Nutr 1979;32:2143-2149. 16. Nelii HJCF, De LeenheerAP. Isocratic nonaqueousreversedphaseliquid chromatography of carotenoids.Anal Chem 1983;55:270-275. 17. Milne DB, BotnenJ. Retinol, alpha-tocopherol,lycopeneand alpha and betacarotene simultaneouslydetermined in plasma by isocratic liquid chromatography. Clin Chem 1986:32:874-876. 18. Morris JF, PeacockM. Assay of plasma 25-hydroxyvitamin D. Clin Chim Acta 1976;72:383-391.
REFERENCES 1. Lund b, BadskjaerJ, Lund B, SoerensenOH. Vitamin D and ischaemicheart disease.Harm Merab Res 1978:10:553-556. 2. Vik T, Try K, Thelle DS, Forde OH. Tromso Heart Study: vitamin D metabolism and myocardial infarction. Br Med J 1979;2:176. 3. Kanofsky JD, Kanofsky PB. Preventionof thromboembolicdiseaseby vitamin E. N Engl J Med 1981;305:173-174. 4. Ellii N, Lloyd B, Lloyd RS, Clayton BE. Seleniumand vitamin E in relation to risk factors for coronary heart disease.J Clin Parho 1984;37:200-206. 5. Flaim E, Williford WO, Mullen JL, Buzby GP, Crosby LO. The relationshipof serumcholesteroland vitamin A in hospitalizedpiitients with and without cancer. Am J Clin Nutr 1986;44:370-378.
6. Russel-Briefel R, Bates MW, Kuller LH. The relationship of plasma carotenoidsto health and biochemical factors in middle-agedmen. Am J Epidemiol 1985:122:741-749.
7. Stahelin HB, Buess E, Rose1F, Widmer LK, Brubacher G. Vitamin A, cardiovascular risk factors and mortality. Lnncet 1982:1:394-39X 8. DobsonHM, Muir MM, Hume R. The effect of ascorbicacid on the seasonal variations in serum cholesterol levels. ScottMed J 1984;29;176-182. 9. Stryd RF’, Gilbertson TJ, Brunden MN. A seasonalvariation study of 25hydroxyvitamin D, serum level in normal humans. J Clin Endowinol Metab 1979:48:771-775.
10. Dunn&n MG, Harland WA, Fyfe T. Seasonalincidenceand mortality of ischaemic heart disease.Lance?1970;2:793-797. 11. World Health Organization. Proposal for the Multinational Monitoring of Trends and Determinants in Cardiovascular Disease.Optional Study of Antioxidant Vitamins. Geneoa: World Health Organizarion, 1984. 974
THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 64
19. Hollis BW, Burton JH, Draper HH. A binding assayfor 25-hydroxycalcifer01sand 24.25dihydroxycalciferoI.susing bovine plasma globulin. Steroids 1977; 30:285-293. 20. Colton T. Statistics in Medicine. Boston Little, Brown, 1974. 21. Pelkonen R. Plasmavitamin A and E in the study of lipid and lipoprotein metabolismin coronary heart die. Arm Med &and 1963;174(suppl399):57. 22. Willett WC, Stampfer MJ, Underwood BA, Taylor JO, HennekensCH. Vitamins A, E and carotene:effectsof supplementationon their plasmalevels.Am J Clin Nut? 1983;38:559-566.
23. Willett WC, Stampfer MJ, Underwood BA, SampsonLA, HennekensCH, Wallingford JC, Cooper L, Hsieh C, SpeizerFE. Vitamin A supplementationand plasma retinol levels: a randomized trial among women. JNCI 1984;73:14451448.
24. Fraser DR. The physiologicaleconomyof vitamin D. Lancer 1983;1:969-971. 25. Mathews-Roth MM, Pathak MA, Fitzpatrick TB, Harber LC, Kass EH. Beta-caroteneas an oral photoprotectiveagent in erythropoietic protoporphyria. JAMA
1974:228:1004-1008.
26. Poskitt EME, Cole TJ, Lawson DEM. Diet, sunlight,and 25-hydroxyvitamin D in healthy children and adults. Br Med J 1979;1:221-223. 27. Holm B, Gcdal HC. Evidencethat changesin fibrinogen quality during acute phasereactions are of major importance for the amount of heparin precipitable fraction (HPF). Thrombosis Res 1985;39:449-458. 28. Bikle DD, Halloran BP, Gee E, Ryzen E, Haddad JG. Free 25-hydroxyvitamin D levels are normal in subjects with liver diseaseand reduced total 25hydroxyvitamin D levels. J C/in Itwesr 1986;78:748-752. 29. Peto R, Doll R, Buckley JD, SpoonMB. Can dietary beta-carotenematerially reduce human cancer rates? Nature 1981:290:201-208.
To establish whether plasma vitamin measuremsnts made after acute myocardial infarction (AMI) can be used in case-control studies of coronary artery disease, the short-term effect of AMI Ds, on plasma concentrations of 25-hydroxyvitamin beta-carotene, vitamin E and retinol was investigated. Gquentlal measures of these vitamins were made during the first 48 hours after AMI in 13 patients admitted to the hospital within 4 hours after the onset of symptoms. Plasma levels of 2!5hydroxyvltamin D did not change significantly during the first 12 hours after onset of symptoms. Ma-carotene levels increased signiffcantly (p
here is continuing interest in a possible role for vitamins in the etiology of coronary artery disease. Serum 25hydroxyvitamin D3 levels have been observed to be lower in acute myocardial infarction (AMI) patients than in control subjects.lT2Vitamin E inhibits platelet aggregation,3and blood vitamin E, A and carotenoid levelscorrelate with blood cholesterol,4-6 although a nonsignificant increase in serum vitamin A has been reported in heart diseasepatients compared to control subjects.’ Low blood levels of vitamins in winter8y9could be related to the known seasonalvariation in coronary artery disease mortality, which is typically 30 to 40% higher in winter than in summer.1oCase-controlstudies, which typically measureexposureat the time of disease occurrence, are potentially suitable for studying shortterm risk factors including associationsbetweenseasonal variations in blood vitamin levels and coronary artery diseaserisk, provided blood levels of vitamins are not affected by the acute phase reaction that occurs after AMI. This study establisheswhether plasma vitamin levels are affected by AM1 in the short term. Sequential blood sampleswere collected during the first 48 hours after the onset of symptomsof AMI. The vitamins chosenfor measurementwere retinol, beta-caroteneand vitamin E, becausethese are being measuredin the World Health Organization Monitoring of Trends and Determinants in Cardiovascular DiseaseStudy,” and 25-hydroxyvitamin D, becausethis vitamin may be related to the seasonal variation in coronary artery disease mortality12 and is probably the most useful vitamin D metabolite for evaluation of overall vitamin D status.13
T
METHODS
From the Department of Community Health, University of Auckland; the Departments of Endocrinology and Clinical Chemistry, Auckland Hospital; and the Lynfield Agricultural Center, Auckland, New Zealand. This study was supported in part by Roche Prcducta, New Zealand. Dr. Jackson is a National Heart Foundation of New Zealand Research Fellow. Manuscript received December 16, 1988; revised manuscript received and accepted July 13, 1989. No reprints will be available.
Nonfasting blood sampleswere collected immediately after admission to the hospital, at 8 hourly intervals during the subsequent24 hours and again at 48 hours, in 13 patients with proven AM1 admitted within 4 hours of onset of chest pain. The sampleswere all taken with patients in the supine position. Plasma samples were separatedwithin 4 hours of collection and stored at -20” C until analyzed. All patients were admitted to a coronary care unit in Auckland, New Zealand, between June 1985 and January 1986. AM1 was diagnosedon the basis of a history of prolonged chest pain, elevated cardiac enzymes and evolving electrocardiographic changes.14There were 11 men and 2 women,
THE AMERICAN JOURNAL OF CARDIOLOGY NOVEMBER 1,1989
971
-
aged 47 to 66 years (mean age 57). The study was approved by the Auckland Hospital Ethics Committee. Retinol, vitamin E and beta-carotenewere measured simultaneously from samples from all 13 subjects by high performance liquid chromatography. Although these 3 compoundshave differing chromatographic and spectral properties and in the past have been measured separately,careful choice of mobile phaseand the availability of programmable multiwavelength detectors allowed their measurement in a single injection sample using high performance liquid chromatography. Extraction of the 3 components from plasma followed an established procedure15 involving protein precipitation with methanol of samples plus internal standards and subsequentextraction into hexane. The hexane was then evaporatedunder nitrogen and the concentrated extract taken up in ethanol ready for injection. Chromatography was performed on a Novapak Cl8 radially compressed column. The mobile phase was composed of 45% acetonitrile, 45% methanol and 10%dichloromethane. This mixture allows good separation of all the measured componentsin a convenient time without the need for gradient elution and is a modification of a previously published method.I6 The individual components were detected in a single chromatographic run by the use of programmed multiple wavelength changes(325 nm for retinol, 292 nm for alpha-tocopherol and 465 nm for beta-carotene) against their respective internal stan-
dards. This detection method closely resemblesthat a Milne and Botnen.r7 For logistic reasons,2 different methods were used to measure25-hydroxyvitamin Ds. Method A was used for 8 patients (patients 1 to 8 in Figure l), while method B was used for 4 more patients (patients 9 to 12 in Figure 1). The same method was always used to measure vitamin D levels in all 5 samplescollected from any individual patient. Method A used extraction of plasma with ethanol’* and fractionation with sephadexLH-20. Assay was by competitive protein binding using human serum as binding protein and separation with dextrancharcoal. The sensitivity of the assay was 12.5 nmol/ liter and the within-assay and between-assaycoefficients of variation were f 9 and f 22% respectively. Method B involved dilution of replicate plasma samples (500 ~1) with an equal volume of water followed by addition of 25 &labeled 25-hydroxycholecalciferol to measurehormone recovery. Protein was removedby addition of acetonitrile, the supernatant was mixed with 500 ~1 of 0.4 M phosphate buffer (pH 10) and passed through a Cl8 Sep-pak cartridge (Waters Associates), discarding the eluate. After washing, 25-hydroxyvitamin D3 was eluted from the cartridge with 3 ml acetonitrile, evaporated and stored at -20’ C. The analysis involved reconstitution with 1 ml ethanol followed by competitive protein binding assay using bovine serum and separation with dextran-coated charcoal.19 The
Number
0’
0’8
16
24
Hours after admission 972
THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 64
1 48
ir i‘ TABLE 1 Vitamin
I Mean Plasma Vitamin Concentrations -
Mean 25.hydroxyvitamin D (nmol/liter) Mean of differences f SE* Percentage change* Beta-carotene (pmol/liter) mean Mean of differences f SE Percentage change Vitamin E (pmol/liter) mean Mean of differences f SE Percentage change Retinol (pmol/liter) mean Mean of differences f SE Percentage change
During the First 48 Hours After Admission for Acute Myocardial
Infarction
0 hours
8 hours
16 hours
24 hours
48 hours
103
103
98
89
89
-14i8 -13.6 0.532
-15f7 -14.6 0.448
0.545
36.9
2.03
Of3 0 0.574
-5*5 -4.9 0.556
0.029 f 0.012+ 5.3 34.4
0.012 f 0.028 2.2 33.4
-2.6 f 1.6 -7.0 1.88
-3.5 f 2.4 -9.5 1.78
-0.14 f 0.04’ -6.9
-0.24 f 0.06’ -11.8
Diierence between means is not always equal to mean of difference because of rounding errors. * Compared wrth concentration on admission (i.e.. 0 hours); 7 p <0.05; f p
-0.013 f 0.033 -2.4 31.1
-0.097 f 0.045 -17.8 27.4
-5.9 f 2.3’ -16.0 1.76
-9.6 f 3.6’ -26.0 1.53
-0.27 f 0.06* -13.3
-0.50 f 0.08% -24.6
of means.
the first 12 hours after the onset of symptoms of AM1 (i.e., over the first 8 hours after admission). In contrast, plasma beta-carotenelevels are elevatedduring the first 20 hours, and then decrease,whereasconcentrations of vitamin E and retinol progressivelydecreaseduring the first 48 hours after AMI. There appear to be no previous studies that have systematically examined changes RESULTS Mean plasma vitamin levels at each of the collection in blood vitamin levels during the first 48 hours after times are listed in Table I, along with the mean of the AMI, although a Danish study reported no change in difference from baseline and the percentage change levels of 25-hydroxyvitamin D in samples taken daily compared with baseline. Mean plasma levels of 25- for 4 days after AMI,* while a Swedish study found a hydroxyvitamin Ds at baselineand 8 hours were similar, significant decreasein vitamin E during the first week but then progressively decreased to 14.6% less than after AMI.** It is unlikely that these changes in plasma vitamin baseline at 48 hours. The trend in the means was similar for patients measured by each vitamin D method. levels are due to changes in hemoconcentration after The mean level of beta-carotene increased significantly AM1 becausethey did not vary in the same direction (5.3%, p <0.05) at 8 hours and thereafter progressively during the first 16 hours after admission. Neither does decreased.Mean level of vitamin E decreasedby 7.0% change in dietary intake after AM1 seema likely explaat 8 hours, and at 24 hours was 16.0%lessthan baseline nation for these results. While blood levels of vitamin (p <0.05). Retinol decreased significantly (6.9%, p E and beta-carotene are both strongly influenced by
THE AMERICAN JOURNAL OF CARDIOLOGY NOVEMBER I,1989
973
PLASMA VITAMINS AITER MYDCARDIAL INFARCTION
zc $ & 6
0.80 -
0.60
-
0’
a 0
8 Hours
16 after
24
I 48
admissions
be influenced by alterations in liver metabolism as may occur in the acute reaction to stress. This study showsthat plasma levels of beta-carotene, retinol and vitamin E change rapidly during the first 48 hours after AMI, which invalidates their measurement over this time period in casecontrol studies. By comparison, the similarity in 25-hydroxyvitamin D levels between the 2 samplestaken during the first 12 hours after onset of AM1 symptomssuggeststhat levels in samples taken during this time are likely to be similar to blood concentrations before AMI, and might be used as a proxy for levels before AM1 in case-control studies.
12. Scragg R. Seasonality of cardiovascular diseasemortality and the possible protective effect of ultra-violet radiation. Inr J Epidemiol 1981;10:337-341. 13. Bikle DD, Gee E, Halloran B, Kowalski MA, Ryzen E, Haddad JG. Assessment of the free fraction of 25-hydroxyvitamin D in serum and its regulation by albumin and the vitamin D binding protein. J Clin Endocrinol Metab 1986,63; 954-959. 14. World Health Organization working Party Report. IschaemicHeart Disease Registers. Part 2. Genes World Health Organization, 1969. 15. Bieri JG, Tolliver TJ, Catignani GL. Simultaneousdetermination of alphatocopheroland retinol in plasmaor red cells by high pressureliquid chromatography. Amer J Clin Nutr 1979;32:2143-2149. 16. Nelii HJCF, De LeenheerAP. Isocratic nonaqueousreversedphaseliquid chromatography of carotenoids.Anal Chem 1983;55:270-275. 17. Milne DB, BotnenJ. Retinol, alpha-tocopherol,lycopeneand alpha and betacarotene simultaneouslydetermined in plasma by isocratic liquid chromatography. Clin Chem 1986:32:874-876. 18. Morris JF, PeacockM. Assay of plasma 25-hydroxyvitamin D. Clin Chim Acta 1976;72:383-391.
REFERENCES 1. Lund b, BadskjaerJ, Lund B, SoerensenOH. Vitamin D and ischaemicheart disease.Harm Merab Res 1978:10:553-556. 2. Vik T, Try K, Thelle DS, Forde OH. Tromso Heart Study: vitamin D metabolism and myocardial infarction. Br Med J 1979;2:176. 3. Kanofsky JD, Kanofsky PB. Preventionof thromboembolicdiseaseby vitamin E. N Engl J Med 1981;305:173-174. 4. Ellii N, Lloyd B, Lloyd RS, Clayton BE. Seleniumand vitamin E in relation to risk factors for coronary heart disease.J Clin Parho 1984;37:200-206. 5. Flaim E, Williford WO, Mullen JL, Buzby GP, Crosby LO. The relationshipof serumcholesteroland vitamin A in hospitalizedpiitients with and without cancer. Am J Clin Nutr 1986;44:370-378.
6. Russel-Briefel R, Bates MW, Kuller LH. The relationship of plasma carotenoidsto health and biochemical factors in middle-agedmen. Am J Epidemiol 1985:122:741-749.
7. Stahelin HB, Buess E, Rose1F, Widmer LK, Brubacher G. Vitamin A, cardiovascular risk factors and mortality. Lnncet 1982:1:394-39X 8. DobsonHM, Muir MM, Hume R. The effect of ascorbicacid on the seasonal variations in serum cholesterol levels. ScottMed J 1984;29;176-182. 9. Stryd RF’, Gilbertson TJ, Brunden MN. A seasonalvariation study of 25hydroxyvitamin D, serum level in normal humans. J Clin Endowinol Metab 1979:48:771-775.
10. Dunn&n MG, Harland WA, Fyfe T. Seasonalincidenceand mortality of ischaemic heart disease.Lance?1970;2:793-797. 11. World Health Organization. Proposal for the Multinational Monitoring of Trends and Determinants in Cardiovascular Disease.Optional Study of Antioxidant Vitamins. Geneoa: World Health Organizarion, 1984. 974
THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 64
19. Hollis BW, Burton JH, Draper HH. A binding assayfor 25-hydroxycalcifer01sand 24.25dihydroxycalciferoI.susing bovine plasma globulin. Steroids 1977; 30:285-293. 20. Colton T. Statistics in Medicine. Boston Little, Brown, 1974. 21. Pelkonen R. Plasmavitamin A and E in the study of lipid and lipoprotein metabolismin coronary heart die. Arm Med &and 1963;174(suppl399):57. 22. Willett WC, Stampfer MJ, Underwood BA, Taylor JO, HennekensCH. Vitamins A, E and carotene:effectsof supplementationon their plasmalevels.Am J Clin Nut? 1983;38:559-566.
23. Willett WC, Stampfer MJ, Underwood BA, SampsonLA, HennekensCH, Wallingford JC, Cooper L, Hsieh C, SpeizerFE. Vitamin A supplementationand plasma retinol levels: a randomized trial among women. JNCI 1984;73:14451448.
24. Fraser DR. The physiologicaleconomyof vitamin D. Lancer 1983;1:969-971. 25. Mathews-Roth MM, Pathak MA, Fitzpatrick TB, Harber LC, Kass EH. Beta-caroteneas an oral photoprotectiveagent in erythropoietic protoporphyria. JAMA
1974:228:1004-1008.
26. Poskitt EME, Cole TJ, Lawson DEM. Diet, sunlight,and 25-hydroxyvitamin D in healthy children and adults. Br Med J 1979;1:221-223. 27. Holm B, Gcdal HC. Evidencethat changesin fibrinogen quality during acute phasereactions are of major importance for the amount of heparin precipitable fraction (HPF). Thrombosis Res 1985;39:449-458. 28. Bikle DD, Halloran BP, Gee E, Ryzen E, Haddad JG. Free 25-hydroxyvitamin D levels are normal in subjects with liver diseaseand reduced total 25hydroxyvitamin D levels. J C/in Itwesr 1986;78:748-752. 29. Peto R, Doll R, Buckley JD, SpoonMB. Can dietary beta-carotenematerially reduce human cancer rates? Nature 1981:290:201-208.