European Heart Journal (1992) 13, 1155-1163
High-density iipoprotein cholesterol and coronary, cardiovascular and all cause mortality among middle-aged Norwegian men and women I. STENSVOLD*, P. U R D A L I , H.
THURMER*,
A. TVERDAL*, P. G.
LUND-LARSEN* AND
O. P. Fossf
* National Health Screening Service, Oslo; ^Department of Clinical Chemistry, Ullev&l Hospital, Oslo, Norway KEY WORDS: Cholesterol, triglycerides, cardiovascular screening. From 1977 to 1982 screeningfor cardiovascular disease was performed in three Norwegian counties. All those aged between 40 and 54 years were invited, of whom 23 690 men and 23 425 women (90%) attended. Smoking habits and previous cardiovascular disease were recorded; total cholesterol, high-density Iipoprotein cholesterol (HDL cholesterol), triglycerides and blood pressure were measured. During subsequent follow-up (mean 6-8 years) 422 men and 54 women diedfrom coronary heart disease, 514 and 114 from all cardiovascular diseases and 983 and 404 from all causes, men and women respectively. For men, mortality decreased with increasing HDL cholesterol, to a minimum of around 1-5 mmol. I"' (58 mg. dl~'), whereafter mortality increased. This applies to coronary, cardiovascular and all causes of death, as well as to men with and without a history of disease. The association between mortality and HDL cholesterol in healthy men disappeared when total cholesterol was below 65 mmol. l~' (251 mg . dl~'). The inverse association between mortality and HDL cholesterol in women was somewhat stronger than in men, both for coronary and cardiovascular diseases. The relative risks of coronary death, associated with an increase in HDL cholesterol of 0-5 mmol. l~' (19 mg. dl~''), from the Cox proportional hazards regression, with other major cardiovascular risk factors as covariates, were 0-8 (95% confidence interval. 0-6,1-0) and 0-8 (0-7,1-0) for men with and without history of disease, respectively. Corresponding figures for women were 0-5 (0-3,0-9) and 0-7 (0-4,1-3). Introduction Epidemiological and other studies have clearly identified total serum cholesterol, cigarette smoking and blood pressure as independent risk factors in the development of coronary heart disease (CHD). An inverse relationship between high density Iipoprotein cholesterol (HDL cholesterol) and CHD has also been found in several epidemiological studies'1"31. In a re-examination of four American prospective studies, Gordon et a/.m concluded that the data supported an independent association between HDL cholesterol levels and CHD event rates. A recent study highlighted the prognostic importance of total, LDL, and HDL cholesterol in men with clinically evident cardiovascular disease141. The results from the British regional heart study showed, after re-examination, that although HDL cholesterol was an independent risk factor for ischaemic heart disease it was less important than total cholesterol'51. In the latter study, however, no attempt was made to exclude those with cardiovascular problems at baseline. In the second cardiovascular survey, which took place from 1977-82 in three counties in Norway, data on lipids and various other risk factors for development of coronary disease were obtained. The attending 47 115 men and women, aged 40-54 years, and selected only by Submitted for publication on 3 May 1991, and in final revised form 2 December 1991. Correspondence: Inger Stensvold, National Health Screening Service, P.O Box 8155, Dep, N-OO33 Oslo 1, Norway. 0195-668X/92/091155 + 09 $08.00/0
area of residence, have been followed throughout 1987 with respect to death. The purpose of this paper is to relate HDL cholesterol to coronary and cardiovascular death and to total mortality, both univariately, as well as when taking other major risk factors, including total cholesterol and triglycerides, into consideration. The association is evaluated both for persons reporting and not reporting cardiovascular disease, diabetes or symptoms of angina pectoris at baseline. Materials and methods Three cardiovascular screenings (screening 1, 2 and 3) were earned out by The National Health Screening Service in each of the three counties, Finnmark (1974, 1977, 1987), Sogn og Fjordane (1975, 1980, 1985) and Oppland (1976, 1981, 1986). Methods and results from screenings 1 and 2 have been reported'6"81. Screening 2 forms the basis of this follow-up study. Follow-up time was longest in Finnmark, the county with the highest coronary mortality rate. All residents in the three counties aged 40-54 years were invited to the screening, except for Finnmark, where only the age group 40-52 was invited. Attendance rates varied from 86% (Finnmark men) to 95% (Sogn og Fjordane women). HDL cholesterol measures from a total of 48 003 attenders, 23 690 men and 23 425 women are included in this study. A personal letter, including a questionnaire, invited subjects to the screening. Questions covered present and © 1992 The European Society of Cardiology
1156 I. Stensvoldet al.
1200-
0-5 (19) HDL cholesterol, mmoL I ' (mg dl~ Figure 1 Mortality related to HDL cholesterol in all men. Rate is adjusted for age and county. • = all causes (938 deaths), • — = all cardiovascular (514 deaths); • = all cancer (210 deaths); • • • x • • = all accidents and suicides (163 deaths).
800
•
s r eoo SOI
HDL cholesterol was measured enzymatically from the cholesterol remaining in the supernatant after precipitation of low density (LDL) and very low density lipoprotein (VLDL) particles by heparin-manganese reagent'121. The Ulleval Hospital analysed the Sogn og Fjordane and Oppland samples (n = 38 600), whereas the Finnmark sera (n = 8500) were analysed at the Institute of Medical Biology in Tromse'13', both laboratories using exactly the same HDL cholesterol method. In 16 out of 20 municipalities in Finnmark, serum was frozen and stored at - 2 0 CC for 12 months before HDL cholesterol determinations were made. Mean HDL cholesterol was 012mmol . 1 " ' lower in the frozen sera than in the fresh sera'131. To compensate for this, 0-12 mmol. 1"' was added to the HDL cholesterol values after analysis of the frozen sera. The stored samples constitute 18% of all the samples. The official causes of deaths are used, as coded by the Central Bureau of Statistics, according to the Norwegian version of the 8th (until 1986) and the 9th (1986 and after) revision of the International Classification of Diseases"4"1. Codes: 410-413, 414.0-414.1, 414.9, 798.1-798.2 (9th revision) define 'coronary heart disease deaths', codes: 390-459, 798.1-798.2 (9th revision) define 'cardiovascular deaths'. Approximately 50% of all cardiovascular deaths in the three counties were verified by
1400
V O. O
§ 400 o
-
\
V
per
previous diseases ('Have you now or in the past had: myocardial infarction, angina pectoris, diabetes mellitus, stroke or other cardiovascular diseases?', 'Are you being treated for hypertension?') and symptoms pointing to angina pectoris or peripheral vascular disease. Data on smoking habits and physical activity at leisure time were also recorded from the questionnaire. In this study the 'healthy group' consisted of those answering 'no' to all the disease and symptom questions. The questionnaire was checked at screening, and height (cm) and weight (kg) were measured. Systolic and diastolic blood pressure were measured twice, with an interval of 1-2 min, in the sitting position, the second values being used in this study. A non-fasting blood sample was drawn, centrifuged after clotting, and the chilled serum was then sent to the Central Laboratory, Ulleval Hospital, Oslo. Total serum cholesterol and tnglycendes were measured by enzymatic methods, according to the principles of Roschlau et a/.'9' and Eggstein et alP°\ except for the sera from Finnmark, where total cholesterol was measured by the LiebermannBurchard method'"1 and triglycerides by a fluorimetric method'"1. Parallel analyses over 4 months, employing both methods used for measuring cholesterol and triglycerides allowed previous total cholesterol and triglyceride results to be expressed in terms of the enzymatic method. The accuracy of the methods was ascertained by analysing certified materials from the National Bureau of Standards, U.S.A., and by participating in the cooperative cholesterol-trigJycerides standardization programme organized by the WHO regional lipid reference centre in Prague. Precision and accuracy were also controlled by repeated analysis of commercially available control sera.
a>
1—1
tJ
Q? 200 "
0
¥
•,
—.
5
High-density iipoprotein cholesterol and coronary, cardiovascular and all cause mortality among middle-aged Norwegian men and women I. STENSVOLD*, P. U R D A L I , H.
THURMER*,
A. TVERDAL*, P. G.
LUND-LARSEN* AND
O. P. Fossf
* National Health Screening Service, Oslo; ^Department of Clinical Chemistry, Ullev&l Hospital, Oslo, Norway KEY WORDS: Cholesterol, triglycerides, cardiovascular screening. From 1977 to 1982 screeningfor cardiovascular disease was performed in three Norwegian counties. All those aged between 40 and 54 years were invited, of whom 23 690 men and 23 425 women (90%) attended. Smoking habits and previous cardiovascular disease were recorded; total cholesterol, high-density Iipoprotein cholesterol (HDL cholesterol), triglycerides and blood pressure were measured. During subsequent follow-up (mean 6-8 years) 422 men and 54 women diedfrom coronary heart disease, 514 and 114 from all cardiovascular diseases and 983 and 404 from all causes, men and women respectively. For men, mortality decreased with increasing HDL cholesterol, to a minimum of around 1-5 mmol. I"' (58 mg. dl~'), whereafter mortality increased. This applies to coronary, cardiovascular and all causes of death, as well as to men with and without a history of disease. The association between mortality and HDL cholesterol in healthy men disappeared when total cholesterol was below 65 mmol. l~' (251 mg . dl~'). The inverse association between mortality and HDL cholesterol in women was somewhat stronger than in men, both for coronary and cardiovascular diseases. The relative risks of coronary death, associated with an increase in HDL cholesterol of 0-5 mmol. l~' (19 mg. dl~''), from the Cox proportional hazards regression, with other major cardiovascular risk factors as covariates, were 0-8 (95% confidence interval. 0-6,1-0) and 0-8 (0-7,1-0) for men with and without history of disease, respectively. Corresponding figures for women were 0-5 (0-3,0-9) and 0-7 (0-4,1-3). Introduction Epidemiological and other studies have clearly identified total serum cholesterol, cigarette smoking and blood pressure as independent risk factors in the development of coronary heart disease (CHD). An inverse relationship between high density Iipoprotein cholesterol (HDL cholesterol) and CHD has also been found in several epidemiological studies'1"31. In a re-examination of four American prospective studies, Gordon et a/.m concluded that the data supported an independent association between HDL cholesterol levels and CHD event rates. A recent study highlighted the prognostic importance of total, LDL, and HDL cholesterol in men with clinically evident cardiovascular disease141. The results from the British regional heart study showed, after re-examination, that although HDL cholesterol was an independent risk factor for ischaemic heart disease it was less important than total cholesterol'51. In the latter study, however, no attempt was made to exclude those with cardiovascular problems at baseline. In the second cardiovascular survey, which took place from 1977-82 in three counties in Norway, data on lipids and various other risk factors for development of coronary disease were obtained. The attending 47 115 men and women, aged 40-54 years, and selected only by Submitted for publication on 3 May 1991, and in final revised form 2 December 1991. Correspondence: Inger Stensvold, National Health Screening Service, P.O Box 8155, Dep, N-OO33 Oslo 1, Norway. 0195-668X/92/091155 + 09 $08.00/0
area of residence, have been followed throughout 1987 with respect to death. The purpose of this paper is to relate HDL cholesterol to coronary and cardiovascular death and to total mortality, both univariately, as well as when taking other major risk factors, including total cholesterol and triglycerides, into consideration. The association is evaluated both for persons reporting and not reporting cardiovascular disease, diabetes or symptoms of angina pectoris at baseline. Materials and methods Three cardiovascular screenings (screening 1, 2 and 3) were earned out by The National Health Screening Service in each of the three counties, Finnmark (1974, 1977, 1987), Sogn og Fjordane (1975, 1980, 1985) and Oppland (1976, 1981, 1986). Methods and results from screenings 1 and 2 have been reported'6"81. Screening 2 forms the basis of this follow-up study. Follow-up time was longest in Finnmark, the county with the highest coronary mortality rate. All residents in the three counties aged 40-54 years were invited to the screening, except for Finnmark, where only the age group 40-52 was invited. Attendance rates varied from 86% (Finnmark men) to 95% (Sogn og Fjordane women). HDL cholesterol measures from a total of 48 003 attenders, 23 690 men and 23 425 women are included in this study. A personal letter, including a questionnaire, invited subjects to the screening. Questions covered present and © 1992 The European Society of Cardiology
1156 I. Stensvoldet al.
1200-
0-5 (19) HDL cholesterol, mmoL I ' (mg dl~ Figure 1 Mortality related to HDL cholesterol in all men. Rate is adjusted for age and county. • = all causes (938 deaths), • — = all cardiovascular (514 deaths); • = all cancer (210 deaths); • • • x • • = all accidents and suicides (163 deaths).
800
•
s r eoo SOI
HDL cholesterol was measured enzymatically from the cholesterol remaining in the supernatant after precipitation of low density (LDL) and very low density lipoprotein (VLDL) particles by heparin-manganese reagent'121. The Ulleval Hospital analysed the Sogn og Fjordane and Oppland samples (n = 38 600), whereas the Finnmark sera (n = 8500) were analysed at the Institute of Medical Biology in Tromse'13', both laboratories using exactly the same HDL cholesterol method. In 16 out of 20 municipalities in Finnmark, serum was frozen and stored at - 2 0 CC for 12 months before HDL cholesterol determinations were made. Mean HDL cholesterol was 012mmol . 1 " ' lower in the frozen sera than in the fresh sera'131. To compensate for this, 0-12 mmol. 1"' was added to the HDL cholesterol values after analysis of the frozen sera. The stored samples constitute 18% of all the samples. The official causes of deaths are used, as coded by the Central Bureau of Statistics, according to the Norwegian version of the 8th (until 1986) and the 9th (1986 and after) revision of the International Classification of Diseases"4"1. Codes: 410-413, 414.0-414.1, 414.9, 798.1-798.2 (9th revision) define 'coronary heart disease deaths', codes: 390-459, 798.1-798.2 (9th revision) define 'cardiovascular deaths'. Approximately 50% of all cardiovascular deaths in the three counties were verified by
1400
V O. O
§ 400 o
-
\
V
per
previous diseases ('Have you now or in the past had: myocardial infarction, angina pectoris, diabetes mellitus, stroke or other cardiovascular diseases?', 'Are you being treated for hypertension?') and symptoms pointing to angina pectoris or peripheral vascular disease. Data on smoking habits and physical activity at leisure time were also recorded from the questionnaire. In this study the 'healthy group' consisted of those answering 'no' to all the disease and symptom questions. The questionnaire was checked at screening, and height (cm) and weight (kg) were measured. Systolic and diastolic blood pressure were measured twice, with an interval of 1-2 min, in the sitting position, the second values being used in this study. A non-fasting blood sample was drawn, centrifuged after clotting, and the chilled serum was then sent to the Central Laboratory, Ulleval Hospital, Oslo. Total serum cholesterol and tnglycendes were measured by enzymatic methods, according to the principles of Roschlau et a/.'9' and Eggstein et alP°\ except for the sera from Finnmark, where total cholesterol was measured by the LiebermannBurchard method'"1 and triglycerides by a fluorimetric method'"1. Parallel analyses over 4 months, employing both methods used for measuring cholesterol and triglycerides allowed previous total cholesterol and triglyceride results to be expressed in terms of the enzymatic method. The accuracy of the methods was ascertained by analysing certified materials from the National Bureau of Standards, U.S.A., and by participating in the cooperative cholesterol-trigJycerides standardization programme organized by the WHO regional lipid reference centre in Prague. Precision and accuracy were also controlled by repeated analysis of commercially available control sera.
a>
1—1
tJ
Q? 200 "
0
¥
•,
—.
5
