Scand J Soc Med, Vol. 20, No.4
Coronary risk factors in middle-aged men as related to smoking, coffee intake and physical activity Arne T. Hostmark', John Berg', Stein Brudal', Steinar R. Berge", Peter Kierulf and Tor Bjerkedal' From the I Department of Preventive Medicine, University of Oslo, Occupational Health Service for lSTK-Alcatel. and JOslo Transport. "Research Group, Central Laboratory, Ulleval Hospital, Oslo
Coronary risk [actors ill middle-aged men as related to smoking. coffee intake and physical activity. Hostmark, A. T., Berg. J" Brudal, S., Berge, S., Kierulf, P. and Bjerkedal, T. (Department of Preventive Medicine, University of Oslo, Occupational Health Service for STK-Alcatel, and Oslo Transport. Research Group, Central Laboratory, UIleval Hospital, Oslo). Scand J Soc Med 1992, 4 (196-203), The relationship between lifestyle and coronary risk factors in blood was investigated in 165 middle-aged men. Plasma fibrinogen, serum triacylglycerols (TG), and apolipoprotein B (apo B) were higher, and high density lipoprotein cholesterol (HOLe) lower in smokers (n = 69) than in non-smokers (II = 96). By linear regression analysis there was a significant positive association between degree of smoking and either total cholesterol (TC), apo B, or plasma fibrinogen, whereas smoking was inversely related to HDLc. Smoking and coffee intake were positively correlated with an atherogenic index, reflecting the balance between low and high density lipoproteins. Sedentary men (n = 59) had higher mean TC, apo B, and atherogenic index than physically active men (n = 104). Smokers used salt more often than non-smokers, Physically active men used vegetables more often than sedentary men. Combining several habits into a "badhabit" score gave a high level of significance for its association with TC, apo B, and the atherogenic index. The results indicate that lifestyle may influence several blood factors involved in atherosclerosis development. Key words: lifestyle factors, smoking, coffee, physical activity, fibrinogen, triacylglycerols, apolipoproteins
It is generally accepted that lifestyle may influence
the incidence of coronary heart disease (1). The relationship between lifestyle factors (e.g. diet, smoking, coffee intake, physical activity) and frequency of coronary artery disease can at least in part be attributed to an effect of various habits on risk factors in blood.
Serum total (TC) and high density lipoprotein (HOLe) cholesterol, apolipoprotein A and B (apoA, apo B), triacylglycerols and plasma fibrinogen concentration are important blood risk factors for coronary artery disease (2-7). This raises the question to what extent various habits may be related to each of these blood risk factors. It is well established that a high intake of saturated fat increases serum cholesterol concentration (8). Also coffee consumption has been implicated as a cholesterol-raising factor (9, 10). Physical activity of the endurance type has, on the other hand, a favourable influence on serum lipids and lipoproteins (1112). The association between smoking and incidence of ischaemic heart disease (and peripheral atherosclerosis) has been attributed to carbon monoxide and nicotine poisoning, and increased plasma fibrinogen concentration (1, 13-15). More recently it has been observed that smokers have an increased level of lipid peroxides, and an increased macrophage uptake of modified low density lipoproteins, indicative of an increased tendency to develop atherosclerosis (16). We reported earlier (7) that an "atherogenic index", reflecting the balance between low (LOL) and high (HOL) density lipoproteins, provided a better estimate than individual lipoprotein components to separate groups of persons according to the extent of coronary atherosclerosis (assessed by arteriography). The present work was carried out to study the extent to which various individual coronary risk factors in blood (and the "atherogenic index") might be related to some lifestyle factors (smoking, coffee intake, physical activity, diet), keeping in mind that statistical relationships may not be causal ones.
Scand J Soc Med 20
Downloaded from sjp.sagepub.com at UCSF LIBRARY & CKM on March 8, 2015
Coronary risk factors in middle-aged men
197
Table I. Corollary risk factors ill smokers and non-smokers among 165 healthy men, 4D-fJO years of age living in Oslo 1988. Non-smokers (n = 96)
Smokers (n = 69) TC (mmol/l) Apo B (gil) Apo A (gil) TG (mmol/l) HDLc (mmolll) ATH-index FlBR (gil) SBP (mm Hg) OBP (mm Hg) BW (kg) H (m) BW/W
6.49 1.71 2.25 1.55 1.21 3.61 3.13 127.1 81.7 77.7 1.77 24.7
± ± ± ± ± ± ± ± ± ± ± ±
0.14 0.05 O.O-t 0.09 0.03 0.26 0.07 1.8 1.1 1.3 0.01 0.3
6.17 1.56 2.41 1.34 1.31 2.74 2.80 128.3 83.7 81.0 1.78 25.4
± 0.10 ± O.Ool ± 0.03 ± 0.08 ± 0.03 ± 0.16 ± 0.07 ± 1.3 ± 0.9 ± 1.0 ± 0.01 ± 0.3
(p = 0.02)*
(P < (P < (P = (P =
0.05) 0.02) O.OOol) 0.001)
(P < 0.05)
TC = total cholesterol; apo B = apolipoprotein B; apo A = apolipoprotein A; TG = triacylglycerols; HOLe = high density lipoprotein cholesterol; ATH = atherogenic index = (TC-HDLc» (apo B) I (apo A» (HOLe). FlBR = fibrinogen; SBP = systolic blood pressure; DBP = diastolic blood pressure; BW = body weight; H = height. Mean values ± SEM. a) Significance level vs the other group.
SUBJECTS AND METHODS Subjects The study population consisted of 165 healthy men, aged 4Q-60 years. Persons with any kind of disease, or using drugs were excluded from the study. The subjects gave their voluntary consent to participate in the investigation. The protocol was approved by the Department of Preventive Medicine, University of Oslo. The subjects underwent a complete physical examination before entering into the study. Age, body weight and height were 49.2 ± 0.5 years, 79.6 ± 0.8 kg, and 177.9 ± 0.5 em (mean ± SEM), respectively. Questionnaire In a questionnaire, each respondent answered questions about smoking (number of cigarettes per day, years of smoking, previous smoker), coffee drinking (number of cups per day, type of coffee), physical activity (vide infra), kind of fat in the diet, the use of salt, fruit and vegetables. A "badhabit" score was calculated on the basis of coffee intake, number of cigarettes smoked per day, years of cigarette smoking, and whether or not physical activity was carried out. Coffee intake was given values 1 to 4: Value 1 = no coffee, value 2 = 1-4 cups/day, value 3 = 5-8 cups/ day and value 4 = 8 cups/day. Smoking groups were given values 1 (never smoked), 2 (exsmoker), 3 « 10 cigarettes/ day), 4 (11-20 cigarettes/day), and 5 (> 20 cigarettes/day). Smoking duration was given values 1 (0 years), 2 « 15 years), 3 (1&-23years), and 4 (> 23 years). Physical activity was given values 1 and 2 according to answer yes or no to the question: Do you perform any kind of physical activity, e.g, walking, jogging, cross-country skiing, gymnastics, bicycling, swimming, football, tennis etc.? The "badhabit" levels arc quartiles of the expression: coffee intake> smoking group> smoking duration> physical activity.
Blood sampling and determination of risk factors in blood Blood samples were taken after an overnight fast, and serum and plasma were kept at -20°C prior to analysis of risk factors in blood. Total serum cholesterol, Hfrl.cand triacylglycerols were determfned enzymatically using Boehringer kit reagents. Apolipoprotein A (total A-I + A.II) and apolipoprotein B were estimated by single radial immunodiffusion on M-partigen plates (Behringwerke AG, W. Germany) containing antibodies against the apolipoproteins. A standard serum obtained from the manufacturer of these plates was included. Fibrinogen was determined in citra ted plasma as thrombin coagulation time (16). The "atherogenic index" was calculated as (TC-HOLc). (apoB)1 (HOLe» (apoA). Statistical analysis Results are presented as mean ± SEM. Differences between two groups were evaluated with the independent Student's r-test, and between several groups by one-way analysis of variance, using Scheffc's method for multiple comparison. This analysis, and that of univariate and multiple linear regression, was carried out with SPSS, and logistic regression with SYSTAT, using a personal computer. The significance level was p < 0.05 unless otherwise specified.
RESULTS Smoking and coronary risk factors The serum level of apo B, triacylglycerols (TO), atherogenic index, and plasma fibrinogen was higher,whereas the serum level of HDLc was lower in smokers than in non-smokers (Table I). Total cholesterol (TC), apo A, systolic (SBP) and diastolic Scand J Soc M~d 20
Downloaded from sjp.sagepub.com at UCSF LIBRARY & CKM on March 8, 2015
A. T. Hostmark et al.
198 lC
Apa B (gIl)
(mmal/I)
~-h-i
6
2
~
4 2
(p,0.05)
°
(p·O.Q1)
°
3 4 5
1 2
1
HOLe (mmal/I) 1.2
3
~
]
(p·O.Ol) 1
2 3 4 5
t-+-t--+--.
6
4 2
°
~ (p,(\.006) '
1
~
,
2 3
4 5
~
3
(p·o.ooll
(NS)
2 3 4 5
FIBR (all)
ATH
2
1
4 5
2 3
1
4
t
0.8
°
(p.006)
Apa A (gill
~
0.4
4 5
2 3
:In J (mmal/l)
lG
°
2 3
1
4 5
Fig. 1. Smoking level and coronary risk factors in healthy Oslo-men, 40-60 years of age in 1988. Serum total cholesterol (TC), apolipoprotein B (apo B), triacylglycerols (TG), plasma fibrinogen (FIBR), serum high density lipoprotein cholesterol (HOLe), apolipoprotein A (a po A), and atherogenic index (ATH, see text) in healthy, middle-aged men grouped according to level of smoking (see text). Mean values ± SEM. Number of men in groups 1-5 were 44, 50, 14, 44, and 10. Of the 165 men entering the study, three were excluded due to missing questionnaire and/or laboratory data. a) p 9 vs 0 cups/day
1.30" 1.23" 0.96
2.48" 1.30 0.72
0.25" 0.32"
0.95 0.77' 0.68
1.52* 1.06 6.77*
"Odds ratios calculated from logistic regression, hDefined in text. 'Fibr = fibrinogen. "Frequencies of using vegetables, varying from 1 to 7 days per week, were compared. e Addition of salt to diet: 1 = never, 2 = occationally, 3 = nearly always. "Wald-test significant at the P < 0.05 level.
logistic regression analysis (Table II) it was found that the level of the atherogenic index increased by 30%, and plasma fibrinogen concentration by 148% when going from non-smokers plus exsmokers to smokers. Also, smokers tended to use salt more often than non-smokers (Table II). Physical activity, dietary habits and corollary risk factors The group of men (II = 104) who answered that they were physically active, had a significantly lower mean level of TC, apo B, and AI'l-l-index than the sedentary group (II = 59, Table Ill). By Wald-test (Table II), a 23% reduction in physical activity was associated with an increase in ATH-index by one unit. Apo A, TG, fibrinogen, HDLc, SBP, DBP, body weight, and relative weight were similar in the two activity groups. Number of cigarettes smoked per day was significantly (p = 0.007) lower in phys-
ically active than in sedentary men, i.e, 10.5 ± 1.0 against 15.1 ± 1.4 (mean ± SEM). Intake of vegetables one more day per week corresponded with a 23% increase in physical activity (Table II). Going from one smoking group to the next higher group was associated with a 75% reduction in physical activity. Coffee intake and corollary risk factors Type of coffee Filtered coffee was used by 105, instant coffee by 31 and boiled coffee by 17 men. Five men did not use coffee. Information on type of coffee was missing in 7 men. By Scheffe multiple comparison test there were no significant group differences in coronary risk factors, but mean values for TC, apo B, fibrinogen, and the atherogenic index were all higher in the group using boiled coffee than corresponding values in any of the other groups (Table IV).
Table Ill. Corollary risk factors ill physically active and sedentary Oslo-men, 40--60 years of age ill 1988. Physically active (n = 1(4) TC" (mmolll) Apo B (gil) Apo A (gil) TG (mmollI) HOLe (mmolll) ATH-index FIBR (gil) SBP (mm Hg) DBP (mm Hg) BW (kg) H (m) BW/W
6.07 1.54 2.36 1.38 1.27 2.80 2.88 127.7 82.5 79.3 1.78 25.0
± ± ± ± ± ± ±
±
± ± ± ±
0.10 0.04 0.03 0.08 0.03 0.16 0.06 1.2 0.8 1.0 0.01 0.3
Sedentary (n = 59) 6.72 1.78 2.42 1.51 1.27 3.64 3.04 128.4 83.6 79.7 1.77 25.3
± 0.10 (p ± 0.06 (p ± 0.04
= O.ooo)b = 0.000)
± 0.09 ± ± ± ± ± ± ± ±
0.03 0.29 (p 0.08 2.0 1.2 1.4 0.01 0.4
= 0.006)
a) Abbreviations are as in Table 1. Mean values ± SEM. b) Significance level vs the other group. Of the 165 men entering the study, two were excluded due to missing questionnaire and/or laboratory data. Scand J Soc Med 20
Downloaded from sjp.sagepub.com at UCSF LIBRARY & CKM on March 8, 2015
200
A. T. Hostmark et al.
Table IV. Corollary risk factors according to type of coffee consumed by healthy Oslo-men, 40-60 years of age ill.1988.
TCa (mmol/l) Apo B (gil) Apo A (gil) TG (mmolll) HDLc (mmol/l) ATH-index FlBR (gil) SBP (mm Hg) DBP (mm Hg) BW (kg) H (m) BW/H!
Boiled (n = 17)
Filtered (n = 105)
Instant (n = 31)
No coffee (n = 5)
6.9 1.7 2.4 1.8 1.3 4.1 3.2 130
6.2 ± 0.1 1.6 ± 0.0 2.4 ± 0.0 1.4 ± 0.1 1.3 ± 0.0 3.0 ± 0.2 2.9 ± 0.1 126 ± I 82 ± I 80.5 ± 1.0 1.79 ± 0.01 25.1 ± 0.3
6.4 ± 0.2 1.7 ± 0.1 2.5 ± 0.1 1.4 ± 0.1 1.3 ± 0.1 3.0 ± 0.4 3.0 ± 0.1 130 ±2 81 ± I 78.4 ± 1.5 1.76 ± 0.01 25.2 ± 0.4
6.0 ± 0.4 1.6 ± 0.2 2.3 ± 0.1 1.2 ± 0.2 1.2 ± 0.1 3.0 ± 0.6 2.5 ± 0.1 142 ±6 90 ±4 79.2 ± 3.7 1.74 ± 0.03 26.3 ± 1.9
± 0.3 ± 0.1 ± 0.1 ± 0.0 ± 0.1 ± 0.7 ± 0.2 ±3 ±2 8~ 78.3 ± 2.6 1.76 ± 0.01 25.2 ± 0.8
Mean values ± SEM. Abbreviations arc as in Table I. Of the 165 men entering the study, seven were excluded due to missing questionnaire and/or laboratory data.
u
Level of coffee intake The level of coronary risk factors seemed to vary with amount of coffee intake, as observed for smoking, but significant relationships were only noted for HDLc, and ATH-index (Fig. 3). By Wald-test (Table II), salt in the diet was positively associated with coffee intake. Distribution of coffee intake in smokers was shifted towards higher values than the distribution in non-smokers (results not shown). In smokTC (mmolll)
6
~
Apo 8 (gil) 2
2 0
(NS)
,-
(NS)
4 -
t-.~"'-; 2
(NS)
0
1
3 4
234
FI8R (gil)
6
~
0 2
"Badhabit score" and coronary risk factors There was a significant positive association between increasing "badhabit" quartiles and TC (r = 0.228, P = 0.004, Fig. 4), apo B (r = 0.260,p = 0.001), TG (r
TG (mmolll) 2
~
4
ers drinking < 4 cups of coffee per day (II = 18) mean plasma fibrinogen concentration was significantly lower than in smokers drinking> 4 cups/day (II = 51), i.e. 2.87 + 0.10 gil vs 3.22 ± 0.08 gil.
1
234
(NS)
0 1 234
Fig. 3. Coffee intake and coronary
1.2
~
4
3
0.8
ATH
Apo A (gil)
HDL e (mmol/I)
2
~
3 2
0.4
~ (p,0.03)
(NS)
(p-O.OS)
0
0
1 234
0
1 234
1
2
3
4
COFFEE INTAKE (groups) Scand J Soc Med 20
Downloaded from sjp.sagepub.com at UCSF LIBRARY & CKM on March 8, 2015
risk factors in healthy Oslo-men, 40-60 years of age in 1988. Coronary risk factors in healthy, middle-aged men grouped according to level of coffee intake (see text). Abbreviations are defined in Figure I and text. Mean values±SEM. Number of men in groups 1-4 were 9, 62, 82, and II. Of the 165 men entering the study, one man was excluded due to missing questionnaire and/or laboratory data. Significance levels for linear regression are shown.
Coronary risk factors in middle-aged men Fig. 4. "Badhabit" quartiles and coronary risk factors in healthy Oslo-men. 4Q-60 years of age in 1988. Coronary risk factors in healthy. middle-aged men grouped according to quartiles of a "bad habit score" (see text). Abbreviations are as in Figure 1. Mean values ± SEM. Number of men in groups 1-1 were 40. 41, 45. and 36. Of the 165 men entering the study, three were excluded due to missing questionnaire and/or laboratory data. a) p < 0.05 vs group 1; b) P < 0.05 vs group 2 (Scheffe multiple comparison test). Significance levels for linear regression are shown.
Te (mmolfO 6
~
0
0 234
0.8 0.4
(p,O.03)
0
1
2 3
4
~ 0
J
2 3
4
-----
b
(p·O.OIl
1
2
3
4
ATH
3
2
/' (p,O.OOD3)
(NS)
1 2 3 4
2 0
1
4
~
4
(p·O.04)
Apo A (gil)
:r
~
2 3 4
1
FIBR (gil)
6
(p·O.DO II
HOLe (mmolfl)
1.2
2
~
(p,O.OO4)
1
a
2
4 2
TG (mmolll)
Apo B (gil)
201
0
1
234
"BADHABIT" QUARTILE
= 0.166, P = 0.04), fibrinogen (r = 0.205, p = 0.01), the atherogenic index (r = 0.285, p = 0.0003), whereas "badhabit" quartiles were inversely related to HOLc (r = -0.169, p = 0.03). Multiple regression By multiple regression analysis it was found that smoking level was significantly associated with plasma fibrinogen (p < 0.01), HOLc (p < 0.007) and the atherogenic index (p < 0.001), smoking duration with apo B (p = 0.05) and TG (p = 0.01), and physical activity with TC (p < 0.001) and apo B (p = 0.001). DISCUSSION These results would be in support of the contention that lifestyle has an influence on the development of atherosclerotic diseases, since the concentration of several blood risk factors for such diseases was statistically related to lifestyle factors. It should be kept in mind, however, that statistical associations may not imply causal relationships. Until now, knowledge about blood chemistry cannot fully explain the increased risk of atherosclerotic diseases in smokers. Presumably, by further exploring biochemical mechanisms involved, and increasing the number of smoking-related blood variables, a higher proportion of the increased atherosclerosis
risk in smokers can be explained. Of particular interest in this context is the increased serum level of lipid peroxides (and increased macrophage uptake of oxidized low density lipoproteins) in smokers, indicative of an increased tendency to develop atherosclerosis (16). The results confirm the previously reported positive association between smoking and plasma fibrinogen concentration (13, 14), which is a major coronary risk factor (13). As suggested by the present study smoking may also promote atherosclerosis via several other risk factors, such as total cholesterol, apo B, triacylglycerols, and low HOL-cholesterol. That smoking influences coronary risk factors is supported by the observation that not only number of cigarettes smoked per day, but also years of smoking were related to several risk factors in blood. An explanation for the association between smoking and serum lipoprotein components was recently provided by Cade et al (18) who observed that smokers had a more unfavourable diet than non-smokers. The finding of only a few significant associations between diet habits and coronary risk factors in this study can probably be attributed to the lack of quantitative information on diet habits provided in the answers to the questions asked. It is noteworthy, though, that heavy smoking, little physical activity, and extra salt in the diet combine in the same persons. Scand J Soc Meel 20
Downloaded from sjp.sagepub.com at UCSF LIBRARY & CKM on March 8, 2015
202
A. T. Hostmark et al.
Questions aboutalcohol consumption were omitted from the questionnaire, since it is difficult to obtain reliable answers on alcohol intake. Alcohol intake can affect lipoprotein metabolism (19), and variation in alcohol intake might have influenced the results. As expected, there was a tight coupling between smoking and coffee consumption. In fact, among smokers there was not one person who did not also drink coffee. In non-smokers there was no association between coffee intake and blood risk factors (not shown). Multiple regression analysis suggests that smoking is independently associated with several blood risk factors and the atherogenic index. However, more extensive studies and statistical evaluations are required to determine the relative importance of the several lifestyle factors for atherosclerosis risk factors in blood. Boiled coffee has been implicated as a cholesterolraising factor (9, 10), but the harmful effect of coffee on coronary heart diseases has been disputed by others (20, 21). In the present study only a small number of men used boiled coffee. Although mean values for TC, apo B, fibrinogen, TG and the atherogenic index were higher in the 18 men drinking boiled coffee than in any of the other coffee groups, statistical significance was not attained, leaving the question unanswered whether or not boiled coffee has a negative effect on atherosclerosis-related risk factors in blood. The finding that the atherogenic index was influenced by smoking is of special interest. This index was previously found to give a closer association than individual lipoprotein components with extent of coronary atherosclerosis, as assessed by number of stenosed arteries (7). Accordingly, the observation {hat the magnitude of this index was appreciably increased in smokers compared with non-smokers strongly supports that smoking promotes atherosclerosis. The improved separation between patients with and without coronary atherosclerosis provided by the index is based on the observation that the relative amount of apo B and cholesterol in LDL can be different in cases and controls (22), and on the assumption that also the relative amounts of apo A and cholesterol in HDL might differ. As shown by Sniderman (22), a proportion of patients prone to develop atherosclerosis had hyperbetalipoproteinemia with fairly normal levels of total cholesterol.
Conceivably, this condition would be more common in a group of patients subjected to coronary angiography than in the present group of normolipemic and clinically healthy, middle-aged men. The negative association between physical activity and incidence of coronary heart disease can be largely explained by a beneficial influence of physical activity on serum lipid concentration and lipoprotein distribution, i.e. lowered serum level of triacylglycerols, and possibly serum total cholesterol, and increased level of HDLe and HDL2 (11, 12). It would appear that the lowered concentration of TC, apo B and atherogenic index in physically active men of the present study supports the hypothesis that physical activity has an anti atherogenic effect. It should be kept in mind, however, that number of cigarettes smoked per day was higher in sedentary than in physically active men, probably contributing to explain the apparent negative influence of sedentary lifestyle on coronary risk factors. The biochemical mechanisms involved in the effects of smoking on lipoprotein metabolism is poorly understood. It was recently shown that smokers had a reduced dietary response of lipoprotein lipase, a key enzyme in lipoprotein catabolism (23). Reduced lipoprotein lipase activity in smokers might contribute to their increased serum lipoprotein levels, which might also be caused by a more unfavourable dietary pattern (18). Increased levels of low density lipoproteins, especially if peroxidized (such as in smokers, 16), would favour atherosclerosis. In addition, smoking would have negative cardiovascular effects by increasing the plasma fibrinogen concentration, and by carbon monoxide and nicotine poisoning of cells in the vessel wall.
ACKNOWLEDGMENTS The authors thank Norsk Hydro a.s., Peter Moller a.s., and Kristiania Brodfabrik a. s. Legat for naringsmiddelforskning for financial support. The technical assistance of Eva Kristensen and Ellen Loken is gratefully acknowledged.
REFERENCES 1. Prevention of coronary heart disease. Report of a WHO Expert Committee. WHO Technical Report Series, No. 678, 1982. 2. Castelli WP, Doyle JT, Gordon T et al. HDL cholesterol and other lipids in coronary heart disease. The co-operative lipoprotein phenotyping study. Lancet 1977; I: 965-70.
Scand J Soc Med 20 Downloaded from sjp.sagepub.com at UCSF LIBRARY & CKM on March 8, 2015
Coronary risk factors ill middle-aged men 3. Miller GJ, Miller NE. Plasma high density lipoprotein concentration and development of ischaemic heart disease. Lancet 1975; I: 16-9. 4. Heiss G. Johnson NJ, Reiland S et al. The epidemiology of high density lipoprotein cholesterol levels. The Lipid Research Clinics Programme Prevalence Study. Circulation 1980; 62: Suppl. IV: 1741-4. 5. Kottke BA. Zinsmeister AR. Holmes Jr DR et al. Apolipoproteins and coronary artery disease. Mayo Clin Proc 1986; 61: 313-20. 6. Mahlcy RW, Inncararity TL, Rail Jr SWC et al, Plasma lipoproteins: apolipoprotein structure and function. J Lipid Res 1984; 25: 1277-94. 7. Hostmark AT, Osland A, Simonsen S, Levorstad K. Lipoprotein-related coronary risk factors in patients with angiographically defined coronary artery disease: relation to number of stenosed arteries. J Int Med 1990; 228: 317-21. 8. Norum KR. Berg T, Helgerud P, Drevon C. Transport of cholesterol. Physiol Rev 1983; 63: 1343-418. 9. Forde OH, Fonnebo Knutsen S, Arnesen E, Theile OS. The Trornso heart study: coffee consumption and serum lipid concentrations in men with hypercholesterolemia: a randomized intervention study. Br Mcd J 1985; 290: 893-5. 10. Bak AAA, Grobbee DE. The effect on serum cholesterol levels of coffee brewed by filtering or boiling. N Eng J Med 1989; 321: 1432-7. II. Hostmark AT. Physical activity and plasma lipids. Scand J Soc Med 1982; 10, Suppl 29: 83-91. 12. Paffenbarger RS, Hyde RT. Exercise in prevention of coronary heart disease. Prev Med 1984; 13: 3-22. 13. Kannel WB, Wolf PA, Castelli WP, D'Agostino RB. Fibrinogen and risk of cardiovascular disease. The Framingham study. lAMA 1987; 258: 1183-6. 14. Hostmark AT, Bjerkedal T, Kierulf P et al, Fish oil and plasma fibrinogen. Br Med J 1988; 297' 180-1.
203
15. Fowkes FGR. Aetiology of peripheral atherOsclerosis. Smoking seems especially important. Br Med J 1989; 298: 405-6. 16. Harats D, Dabach Y, Helander G. et al. Fish oil ingestion in smokers and non-smokers enhances pcroxidation of plasma lipoproteins. Atherosclerosis 1991; 90: 127-39. 17. Clauss A. Gerinnerungsphysiologische Schnellmetode zur Bestimmung des Fibrinogens. Acta Haematol (Basel) 1957; 17: 237-46. 18. Cade JE, Margetts BM. Relationship between diet and smoking - Is the diet of smokers different? Epidemiol Community Health 1991; 45: 270-2. 19. Nilsen 0, Forde OH, Brenn T. The Trornso heart study. Distribution and population determinants of gamma glutamyltransfcrasc, Am J Epidemiol 1990; 132: 31&--26. 20. Haffner SM Knapp JA, Stern MP et al. Coffee consumption, diet, and lipids. Am J Epidemiol 1985; 122: 1-12. 21. Grobbee DE, Rimm E, Giovannucci E et al, Coffee, caffeine, and cardiovascular disease in men. N Engl J Med 1990; 323: 1026-32. 22. Sniderman A, Shapiro S, Marpole D ct al. Association of coronary atherosclerosis with hyperapobctallpoproteinemia (increased protein but normal cholesterol levels in human plasma low density (beta) lipoproteins. Proc Natl Acad Sci (USA) 1980; 77: 60-t-8. 23. Chajek-Shaul T. Smoking depresses adipose lipoprotein lipase response to oral glucose. Eur J Clin Invest 1990; 20: 299-304.
Address for offprints: Arne T. Hostmark Gydas vel 8 0363 Oslo
Scand J Soc Med 20 Downloaded from sjp.sagepub.com at UCSF LIBRARY & CKM on March 8, 2015
Coronary risk factors in middle-aged men as related to smoking, coffee intake and physical activity Arne T. Hostmark', John Berg', Stein Brudal', Steinar R. Berge", Peter Kierulf and Tor Bjerkedal' From the I Department of Preventive Medicine, University of Oslo, Occupational Health Service for lSTK-Alcatel. and JOslo Transport. "Research Group, Central Laboratory, Ulleval Hospital, Oslo
Coronary risk [actors ill middle-aged men as related to smoking. coffee intake and physical activity. Hostmark, A. T., Berg. J" Brudal, S., Berge, S., Kierulf, P. and Bjerkedal, T. (Department of Preventive Medicine, University of Oslo, Occupational Health Service for STK-Alcatel, and Oslo Transport. Research Group, Central Laboratory, UIleval Hospital, Oslo). Scand J Soc Med 1992, 4 (196-203), The relationship between lifestyle and coronary risk factors in blood was investigated in 165 middle-aged men. Plasma fibrinogen, serum triacylglycerols (TG), and apolipoprotein B (apo B) were higher, and high density lipoprotein cholesterol (HOLe) lower in smokers (n = 69) than in non-smokers (II = 96). By linear regression analysis there was a significant positive association between degree of smoking and either total cholesterol (TC), apo B, or plasma fibrinogen, whereas smoking was inversely related to HDLc. Smoking and coffee intake were positively correlated with an atherogenic index, reflecting the balance between low and high density lipoproteins. Sedentary men (n = 59) had higher mean TC, apo B, and atherogenic index than physically active men (n = 104). Smokers used salt more often than non-smokers, Physically active men used vegetables more often than sedentary men. Combining several habits into a "badhabit" score gave a high level of significance for its association with TC, apo B, and the atherogenic index. The results indicate that lifestyle may influence several blood factors involved in atherosclerosis development. Key words: lifestyle factors, smoking, coffee, physical activity, fibrinogen, triacylglycerols, apolipoproteins
It is generally accepted that lifestyle may influence
the incidence of coronary heart disease (1). The relationship between lifestyle factors (e.g. diet, smoking, coffee intake, physical activity) and frequency of coronary artery disease can at least in part be attributed to an effect of various habits on risk factors in blood.
Serum total (TC) and high density lipoprotein (HOLe) cholesterol, apolipoprotein A and B (apoA, apo B), triacylglycerols and plasma fibrinogen concentration are important blood risk factors for coronary artery disease (2-7). This raises the question to what extent various habits may be related to each of these blood risk factors. It is well established that a high intake of saturated fat increases serum cholesterol concentration (8). Also coffee consumption has been implicated as a cholesterol-raising factor (9, 10). Physical activity of the endurance type has, on the other hand, a favourable influence on serum lipids and lipoproteins (1112). The association between smoking and incidence of ischaemic heart disease (and peripheral atherosclerosis) has been attributed to carbon monoxide and nicotine poisoning, and increased plasma fibrinogen concentration (1, 13-15). More recently it has been observed that smokers have an increased level of lipid peroxides, and an increased macrophage uptake of modified low density lipoproteins, indicative of an increased tendency to develop atherosclerosis (16). We reported earlier (7) that an "atherogenic index", reflecting the balance between low (LOL) and high (HOL) density lipoproteins, provided a better estimate than individual lipoprotein components to separate groups of persons according to the extent of coronary atherosclerosis (assessed by arteriography). The present work was carried out to study the extent to which various individual coronary risk factors in blood (and the "atherogenic index") might be related to some lifestyle factors (smoking, coffee intake, physical activity, diet), keeping in mind that statistical relationships may not be causal ones.
Scand J Soc Med 20
Downloaded from sjp.sagepub.com at UCSF LIBRARY & CKM on March 8, 2015
Coronary risk factors in middle-aged men
197
Table I. Corollary risk factors ill smokers and non-smokers among 165 healthy men, 4D-fJO years of age living in Oslo 1988. Non-smokers (n = 96)
Smokers (n = 69) TC (mmol/l) Apo B (gil) Apo A (gil) TG (mmol/l) HDLc (mmolll) ATH-index FlBR (gil) SBP (mm Hg) OBP (mm Hg) BW (kg) H (m) BW/W
6.49 1.71 2.25 1.55 1.21 3.61 3.13 127.1 81.7 77.7 1.77 24.7
± ± ± ± ± ± ± ± ± ± ± ±
0.14 0.05 O.O-t 0.09 0.03 0.26 0.07 1.8 1.1 1.3 0.01 0.3
6.17 1.56 2.41 1.34 1.31 2.74 2.80 128.3 83.7 81.0 1.78 25.4
± 0.10 ± O.Ool ± 0.03 ± 0.08 ± 0.03 ± 0.16 ± 0.07 ± 1.3 ± 0.9 ± 1.0 ± 0.01 ± 0.3
(p = 0.02)*
(P < (P < (P = (P =
0.05) 0.02) O.OOol) 0.001)
(P < 0.05)
TC = total cholesterol; apo B = apolipoprotein B; apo A = apolipoprotein A; TG = triacylglycerols; HOLe = high density lipoprotein cholesterol; ATH = atherogenic index = (TC-HDLc» (apo B) I (apo A» (HOLe). FlBR = fibrinogen; SBP = systolic blood pressure; DBP = diastolic blood pressure; BW = body weight; H = height. Mean values ± SEM. a) Significance level vs the other group.
SUBJECTS AND METHODS Subjects The study population consisted of 165 healthy men, aged 4Q-60 years. Persons with any kind of disease, or using drugs were excluded from the study. The subjects gave their voluntary consent to participate in the investigation. The protocol was approved by the Department of Preventive Medicine, University of Oslo. The subjects underwent a complete physical examination before entering into the study. Age, body weight and height were 49.2 ± 0.5 years, 79.6 ± 0.8 kg, and 177.9 ± 0.5 em (mean ± SEM), respectively. Questionnaire In a questionnaire, each respondent answered questions about smoking (number of cigarettes per day, years of smoking, previous smoker), coffee drinking (number of cups per day, type of coffee), physical activity (vide infra), kind of fat in the diet, the use of salt, fruit and vegetables. A "badhabit" score was calculated on the basis of coffee intake, number of cigarettes smoked per day, years of cigarette smoking, and whether or not physical activity was carried out. Coffee intake was given values 1 to 4: Value 1 = no coffee, value 2 = 1-4 cups/day, value 3 = 5-8 cups/ day and value 4 = 8 cups/day. Smoking groups were given values 1 (never smoked), 2 (exsmoker), 3 « 10 cigarettes/ day), 4 (11-20 cigarettes/day), and 5 (> 20 cigarettes/day). Smoking duration was given values 1 (0 years), 2 « 15 years), 3 (1&-23years), and 4 (> 23 years). Physical activity was given values 1 and 2 according to answer yes or no to the question: Do you perform any kind of physical activity, e.g, walking, jogging, cross-country skiing, gymnastics, bicycling, swimming, football, tennis etc.? The "badhabit" levels arc quartiles of the expression: coffee intake> smoking group> smoking duration> physical activity.
Blood sampling and determination of risk factors in blood Blood samples were taken after an overnight fast, and serum and plasma were kept at -20°C prior to analysis of risk factors in blood. Total serum cholesterol, Hfrl.cand triacylglycerols were determfned enzymatically using Boehringer kit reagents. Apolipoprotein A (total A-I + A.II) and apolipoprotein B were estimated by single radial immunodiffusion on M-partigen plates (Behringwerke AG, W. Germany) containing antibodies against the apolipoproteins. A standard serum obtained from the manufacturer of these plates was included. Fibrinogen was determined in citra ted plasma as thrombin coagulation time (16). The "atherogenic index" was calculated as (TC-HOLc). (apoB)1 (HOLe» (apoA). Statistical analysis Results are presented as mean ± SEM. Differences between two groups were evaluated with the independent Student's r-test, and between several groups by one-way analysis of variance, using Scheffc's method for multiple comparison. This analysis, and that of univariate and multiple linear regression, was carried out with SPSS, and logistic regression with SYSTAT, using a personal computer. The significance level was p < 0.05 unless otherwise specified.
RESULTS Smoking and coronary risk factors The serum level of apo B, triacylglycerols (TO), atherogenic index, and plasma fibrinogen was higher,whereas the serum level of HDLc was lower in smokers than in non-smokers (Table I). Total cholesterol (TC), apo A, systolic (SBP) and diastolic Scand J Soc M~d 20
Downloaded from sjp.sagepub.com at UCSF LIBRARY & CKM on March 8, 2015
A. T. Hostmark et al.
198 lC
Apa B (gIl)
(mmal/I)
~-h-i
6
2
~
4 2
(p,0.05)
°
(p·O.Q1)
°
3 4 5
1 2
1
HOLe (mmal/I) 1.2
3
~
]
(p·O.Ol) 1
2 3 4 5
t-+-t--+--.
6
4 2
°
~ (p,(\.006) '
1
~
,
2 3
4 5
~
3
(p·o.ooll
(NS)
2 3 4 5
FIBR (all)
ATH
2
1
4 5
2 3
1
4
t
0.8
°
(p.006)
Apa A (gill
~
0.4
4 5
2 3
:In J (mmal/l)
lG
°
2 3
1
4 5
Fig. 1. Smoking level and coronary risk factors in healthy Oslo-men, 40-60 years of age in 1988. Serum total cholesterol (TC), apolipoprotein B (apo B), triacylglycerols (TG), plasma fibrinogen (FIBR), serum high density lipoprotein cholesterol (HOLe), apolipoprotein A (a po A), and atherogenic index (ATH, see text) in healthy, middle-aged men grouped according to level of smoking (see text). Mean values ± SEM. Number of men in groups 1-5 were 44, 50, 14, 44, and 10. Of the 165 men entering the study, three were excluded due to missing questionnaire and/or laboratory data. a) p 9 vs 0 cups/day
1.30" 1.23" 0.96
2.48" 1.30 0.72
0.25" 0.32"
0.95 0.77' 0.68
1.52* 1.06 6.77*
"Odds ratios calculated from logistic regression, hDefined in text. 'Fibr = fibrinogen. "Frequencies of using vegetables, varying from 1 to 7 days per week, were compared. e Addition of salt to diet: 1 = never, 2 = occationally, 3 = nearly always. "Wald-test significant at the P < 0.05 level.
logistic regression analysis (Table II) it was found that the level of the atherogenic index increased by 30%, and plasma fibrinogen concentration by 148% when going from non-smokers plus exsmokers to smokers. Also, smokers tended to use salt more often than non-smokers (Table II). Physical activity, dietary habits and corollary risk factors The group of men (II = 104) who answered that they were physically active, had a significantly lower mean level of TC, apo B, and AI'l-l-index than the sedentary group (II = 59, Table Ill). By Wald-test (Table II), a 23% reduction in physical activity was associated with an increase in ATH-index by one unit. Apo A, TG, fibrinogen, HDLc, SBP, DBP, body weight, and relative weight were similar in the two activity groups. Number of cigarettes smoked per day was significantly (p = 0.007) lower in phys-
ically active than in sedentary men, i.e, 10.5 ± 1.0 against 15.1 ± 1.4 (mean ± SEM). Intake of vegetables one more day per week corresponded with a 23% increase in physical activity (Table II). Going from one smoking group to the next higher group was associated with a 75% reduction in physical activity. Coffee intake and corollary risk factors Type of coffee Filtered coffee was used by 105, instant coffee by 31 and boiled coffee by 17 men. Five men did not use coffee. Information on type of coffee was missing in 7 men. By Scheffe multiple comparison test there were no significant group differences in coronary risk factors, but mean values for TC, apo B, fibrinogen, and the atherogenic index were all higher in the group using boiled coffee than corresponding values in any of the other groups (Table IV).
Table Ill. Corollary risk factors ill physically active and sedentary Oslo-men, 40--60 years of age ill 1988. Physically active (n = 1(4) TC" (mmolll) Apo B (gil) Apo A (gil) TG (mmollI) HOLe (mmolll) ATH-index FIBR (gil) SBP (mm Hg) DBP (mm Hg) BW (kg) H (m) BW/W
6.07 1.54 2.36 1.38 1.27 2.80 2.88 127.7 82.5 79.3 1.78 25.0
± ± ± ± ± ± ±
±
± ± ± ±
0.10 0.04 0.03 0.08 0.03 0.16 0.06 1.2 0.8 1.0 0.01 0.3
Sedentary (n = 59) 6.72 1.78 2.42 1.51 1.27 3.64 3.04 128.4 83.6 79.7 1.77 25.3
± 0.10 (p ± 0.06 (p ± 0.04
= O.ooo)b = 0.000)
± 0.09 ± ± ± ± ± ± ± ±
0.03 0.29 (p 0.08 2.0 1.2 1.4 0.01 0.4
= 0.006)
a) Abbreviations are as in Table 1. Mean values ± SEM. b) Significance level vs the other group. Of the 165 men entering the study, two were excluded due to missing questionnaire and/or laboratory data. Scand J Soc Med 20
Downloaded from sjp.sagepub.com at UCSF LIBRARY & CKM on March 8, 2015
200
A. T. Hostmark et al.
Table IV. Corollary risk factors according to type of coffee consumed by healthy Oslo-men, 40-60 years of age ill.1988.
TCa (mmol/l) Apo B (gil) Apo A (gil) TG (mmolll) HDLc (mmol/l) ATH-index FlBR (gil) SBP (mm Hg) DBP (mm Hg) BW (kg) H (m) BW/H!
Boiled (n = 17)
Filtered (n = 105)
Instant (n = 31)
No coffee (n = 5)
6.9 1.7 2.4 1.8 1.3 4.1 3.2 130
6.2 ± 0.1 1.6 ± 0.0 2.4 ± 0.0 1.4 ± 0.1 1.3 ± 0.0 3.0 ± 0.2 2.9 ± 0.1 126 ± I 82 ± I 80.5 ± 1.0 1.79 ± 0.01 25.1 ± 0.3
6.4 ± 0.2 1.7 ± 0.1 2.5 ± 0.1 1.4 ± 0.1 1.3 ± 0.1 3.0 ± 0.4 3.0 ± 0.1 130 ±2 81 ± I 78.4 ± 1.5 1.76 ± 0.01 25.2 ± 0.4
6.0 ± 0.4 1.6 ± 0.2 2.3 ± 0.1 1.2 ± 0.2 1.2 ± 0.1 3.0 ± 0.6 2.5 ± 0.1 142 ±6 90 ±4 79.2 ± 3.7 1.74 ± 0.03 26.3 ± 1.9
± 0.3 ± 0.1 ± 0.1 ± 0.0 ± 0.1 ± 0.7 ± 0.2 ±3 ±2 8~ 78.3 ± 2.6 1.76 ± 0.01 25.2 ± 0.8
Mean values ± SEM. Abbreviations arc as in Table I. Of the 165 men entering the study, seven were excluded due to missing questionnaire and/or laboratory data.
u
Level of coffee intake The level of coronary risk factors seemed to vary with amount of coffee intake, as observed for smoking, but significant relationships were only noted for HDLc, and ATH-index (Fig. 3). By Wald-test (Table II), salt in the diet was positively associated with coffee intake. Distribution of coffee intake in smokers was shifted towards higher values than the distribution in non-smokers (results not shown). In smokTC (mmolll)
6
~
Apo 8 (gil) 2
2 0
(NS)
,-
(NS)
4 -
t-.~"'-; 2
(NS)
0
1
3 4
234
FI8R (gil)
6
~
0 2
"Badhabit score" and coronary risk factors There was a significant positive association between increasing "badhabit" quartiles and TC (r = 0.228, P = 0.004, Fig. 4), apo B (r = 0.260,p = 0.001), TG (r
TG (mmolll) 2
~
4
ers drinking < 4 cups of coffee per day (II = 18) mean plasma fibrinogen concentration was significantly lower than in smokers drinking> 4 cups/day (II = 51), i.e. 2.87 + 0.10 gil vs 3.22 ± 0.08 gil.
1
234
(NS)
0 1 234
Fig. 3. Coffee intake and coronary
1.2
~
4
3
0.8
ATH
Apo A (gil)
HDL e (mmol/I)
2
~
3 2
0.4
~ (p,0.03)
(NS)
(p-O.OS)
0
0
1 234
0
1 234
1
2
3
4
COFFEE INTAKE (groups) Scand J Soc Med 20
Downloaded from sjp.sagepub.com at UCSF LIBRARY & CKM on March 8, 2015
risk factors in healthy Oslo-men, 40-60 years of age in 1988. Coronary risk factors in healthy, middle-aged men grouped according to level of coffee intake (see text). Abbreviations are defined in Figure I and text. Mean values±SEM. Number of men in groups 1-4 were 9, 62, 82, and II. Of the 165 men entering the study, one man was excluded due to missing questionnaire and/or laboratory data. Significance levels for linear regression are shown.
Coronary risk factors in middle-aged men Fig. 4. "Badhabit" quartiles and coronary risk factors in healthy Oslo-men. 4Q-60 years of age in 1988. Coronary risk factors in healthy. middle-aged men grouped according to quartiles of a "bad habit score" (see text). Abbreviations are as in Figure 1. Mean values ± SEM. Number of men in groups 1-1 were 40. 41, 45. and 36. Of the 165 men entering the study, three were excluded due to missing questionnaire and/or laboratory data. a) p < 0.05 vs group 1; b) P < 0.05 vs group 2 (Scheffe multiple comparison test). Significance levels for linear regression are shown.
Te (mmolfO 6
~
0
0 234
0.8 0.4
(p,O.03)
0
1
2 3
4
~ 0
J
2 3
4
-----
b
(p·O.OIl
1
2
3
4
ATH
3
2
/' (p,O.OOD3)
(NS)
1 2 3 4
2 0
1
4
~
4
(p·O.04)
Apo A (gil)
:r
~
2 3 4
1
FIBR (gil)
6
(p·O.DO II
HOLe (mmolfl)
1.2
2
~
(p,O.OO4)
1
a
2
4 2
TG (mmolll)
Apo B (gil)
201
0
1
234
"BADHABIT" QUARTILE
= 0.166, P = 0.04), fibrinogen (r = 0.205, p = 0.01), the atherogenic index (r = 0.285, p = 0.0003), whereas "badhabit" quartiles were inversely related to HOLc (r = -0.169, p = 0.03). Multiple regression By multiple regression analysis it was found that smoking level was significantly associated with plasma fibrinogen (p < 0.01), HOLc (p < 0.007) and the atherogenic index (p < 0.001), smoking duration with apo B (p = 0.05) and TG (p = 0.01), and physical activity with TC (p < 0.001) and apo B (p = 0.001). DISCUSSION These results would be in support of the contention that lifestyle has an influence on the development of atherosclerotic diseases, since the concentration of several blood risk factors for such diseases was statistically related to lifestyle factors. It should be kept in mind, however, that statistical associations may not imply causal relationships. Until now, knowledge about blood chemistry cannot fully explain the increased risk of atherosclerotic diseases in smokers. Presumably, by further exploring biochemical mechanisms involved, and increasing the number of smoking-related blood variables, a higher proportion of the increased atherosclerosis
risk in smokers can be explained. Of particular interest in this context is the increased serum level of lipid peroxides (and increased macrophage uptake of oxidized low density lipoproteins) in smokers, indicative of an increased tendency to develop atherosclerosis (16). The results confirm the previously reported positive association between smoking and plasma fibrinogen concentration (13, 14), which is a major coronary risk factor (13). As suggested by the present study smoking may also promote atherosclerosis via several other risk factors, such as total cholesterol, apo B, triacylglycerols, and low HOL-cholesterol. That smoking influences coronary risk factors is supported by the observation that not only number of cigarettes smoked per day, but also years of smoking were related to several risk factors in blood. An explanation for the association between smoking and serum lipoprotein components was recently provided by Cade et al (18) who observed that smokers had a more unfavourable diet than non-smokers. The finding of only a few significant associations between diet habits and coronary risk factors in this study can probably be attributed to the lack of quantitative information on diet habits provided in the answers to the questions asked. It is noteworthy, though, that heavy smoking, little physical activity, and extra salt in the diet combine in the same persons. Scand J Soc Meel 20
Downloaded from sjp.sagepub.com at UCSF LIBRARY & CKM on March 8, 2015
202
A. T. Hostmark et al.
Questions aboutalcohol consumption were omitted from the questionnaire, since it is difficult to obtain reliable answers on alcohol intake. Alcohol intake can affect lipoprotein metabolism (19), and variation in alcohol intake might have influenced the results. As expected, there was a tight coupling between smoking and coffee consumption. In fact, among smokers there was not one person who did not also drink coffee. In non-smokers there was no association between coffee intake and blood risk factors (not shown). Multiple regression analysis suggests that smoking is independently associated with several blood risk factors and the atherogenic index. However, more extensive studies and statistical evaluations are required to determine the relative importance of the several lifestyle factors for atherosclerosis risk factors in blood. Boiled coffee has been implicated as a cholesterolraising factor (9, 10), but the harmful effect of coffee on coronary heart diseases has been disputed by others (20, 21). In the present study only a small number of men used boiled coffee. Although mean values for TC, apo B, fibrinogen, TG and the atherogenic index were higher in the 18 men drinking boiled coffee than in any of the other coffee groups, statistical significance was not attained, leaving the question unanswered whether or not boiled coffee has a negative effect on atherosclerosis-related risk factors in blood. The finding that the atherogenic index was influenced by smoking is of special interest. This index was previously found to give a closer association than individual lipoprotein components with extent of coronary atherosclerosis, as assessed by number of stenosed arteries (7). Accordingly, the observation {hat the magnitude of this index was appreciably increased in smokers compared with non-smokers strongly supports that smoking promotes atherosclerosis. The improved separation between patients with and without coronary atherosclerosis provided by the index is based on the observation that the relative amount of apo B and cholesterol in LDL can be different in cases and controls (22), and on the assumption that also the relative amounts of apo A and cholesterol in HDL might differ. As shown by Sniderman (22), a proportion of patients prone to develop atherosclerosis had hyperbetalipoproteinemia with fairly normal levels of total cholesterol.
Conceivably, this condition would be more common in a group of patients subjected to coronary angiography than in the present group of normolipemic and clinically healthy, middle-aged men. The negative association between physical activity and incidence of coronary heart disease can be largely explained by a beneficial influence of physical activity on serum lipid concentration and lipoprotein distribution, i.e. lowered serum level of triacylglycerols, and possibly serum total cholesterol, and increased level of HDLe and HDL2 (11, 12). It would appear that the lowered concentration of TC, apo B and atherogenic index in physically active men of the present study supports the hypothesis that physical activity has an anti atherogenic effect. It should be kept in mind, however, that number of cigarettes smoked per day was higher in sedentary than in physically active men, probably contributing to explain the apparent negative influence of sedentary lifestyle on coronary risk factors. The biochemical mechanisms involved in the effects of smoking on lipoprotein metabolism is poorly understood. It was recently shown that smokers had a reduced dietary response of lipoprotein lipase, a key enzyme in lipoprotein catabolism (23). Reduced lipoprotein lipase activity in smokers might contribute to their increased serum lipoprotein levels, which might also be caused by a more unfavourable dietary pattern (18). Increased levels of low density lipoproteins, especially if peroxidized (such as in smokers, 16), would favour atherosclerosis. In addition, smoking would have negative cardiovascular effects by increasing the plasma fibrinogen concentration, and by carbon monoxide and nicotine poisoning of cells in the vessel wall.
ACKNOWLEDGMENTS The authors thank Norsk Hydro a.s., Peter Moller a.s., and Kristiania Brodfabrik a. s. Legat for naringsmiddelforskning for financial support. The technical assistance of Eva Kristensen and Ellen Loken is gratefully acknowledged.
REFERENCES 1. Prevention of coronary heart disease. Report of a WHO Expert Committee. WHO Technical Report Series, No. 678, 1982. 2. Castelli WP, Doyle JT, Gordon T et al. HDL cholesterol and other lipids in coronary heart disease. The co-operative lipoprotein phenotyping study. Lancet 1977; I: 965-70.
Scand J Soc Med 20 Downloaded from sjp.sagepub.com at UCSF LIBRARY & CKM on March 8, 2015
Coronary risk factors ill middle-aged men 3. Miller GJ, Miller NE. Plasma high density lipoprotein concentration and development of ischaemic heart disease. Lancet 1975; I: 16-9. 4. Heiss G. Johnson NJ, Reiland S et al. The epidemiology of high density lipoprotein cholesterol levels. The Lipid Research Clinics Programme Prevalence Study. Circulation 1980; 62: Suppl. IV: 1741-4. 5. Kottke BA. Zinsmeister AR. Holmes Jr DR et al. Apolipoproteins and coronary artery disease. Mayo Clin Proc 1986; 61: 313-20. 6. Mahlcy RW, Inncararity TL, Rail Jr SWC et al, Plasma lipoproteins: apolipoprotein structure and function. J Lipid Res 1984; 25: 1277-94. 7. Hostmark AT, Osland A, Simonsen S, Levorstad K. Lipoprotein-related coronary risk factors in patients with angiographically defined coronary artery disease: relation to number of stenosed arteries. J Int Med 1990; 228: 317-21. 8. Norum KR. Berg T, Helgerud P, Drevon C. Transport of cholesterol. Physiol Rev 1983; 63: 1343-418. 9. Forde OH, Fonnebo Knutsen S, Arnesen E, Theile OS. The Trornso heart study: coffee consumption and serum lipid concentrations in men with hypercholesterolemia: a randomized intervention study. Br Mcd J 1985; 290: 893-5. 10. Bak AAA, Grobbee DE. The effect on serum cholesterol levels of coffee brewed by filtering or boiling. N Eng J Med 1989; 321: 1432-7. II. Hostmark AT. Physical activity and plasma lipids. Scand J Soc Med 1982; 10, Suppl 29: 83-91. 12. Paffenbarger RS, Hyde RT. Exercise in prevention of coronary heart disease. Prev Med 1984; 13: 3-22. 13. Kannel WB, Wolf PA, Castelli WP, D'Agostino RB. Fibrinogen and risk of cardiovascular disease. The Framingham study. lAMA 1987; 258: 1183-6. 14. Hostmark AT, Bjerkedal T, Kierulf P et al, Fish oil and plasma fibrinogen. Br Med J 1988; 297' 180-1.
203
15. Fowkes FGR. Aetiology of peripheral atherOsclerosis. Smoking seems especially important. Br Med J 1989; 298: 405-6. 16. Harats D, Dabach Y, Helander G. et al. Fish oil ingestion in smokers and non-smokers enhances pcroxidation of plasma lipoproteins. Atherosclerosis 1991; 90: 127-39. 17. Clauss A. Gerinnerungsphysiologische Schnellmetode zur Bestimmung des Fibrinogens. Acta Haematol (Basel) 1957; 17: 237-46. 18. Cade JE, Margetts BM. Relationship between diet and smoking - Is the diet of smokers different? Epidemiol Community Health 1991; 45: 270-2. 19. Nilsen 0, Forde OH, Brenn T. The Trornso heart study. Distribution and population determinants of gamma glutamyltransfcrasc, Am J Epidemiol 1990; 132: 31&--26. 20. Haffner SM Knapp JA, Stern MP et al. Coffee consumption, diet, and lipids. Am J Epidemiol 1985; 122: 1-12. 21. Grobbee DE, Rimm E, Giovannucci E et al, Coffee, caffeine, and cardiovascular disease in men. N Engl J Med 1990; 323: 1026-32. 22. Sniderman A, Shapiro S, Marpole D ct al. Association of coronary atherosclerosis with hyperapobctallpoproteinemia (increased protein but normal cholesterol levels in human plasma low density (beta) lipoproteins. Proc Natl Acad Sci (USA) 1980; 77: 60-t-8. 23. Chajek-Shaul T. Smoking depresses adipose lipoprotein lipase response to oral glucose. Eur J Clin Invest 1990; 20: 299-304.
Address for offprints: Arne T. Hostmark Gydas vel 8 0363 Oslo
Scand J Soc Med 20 Downloaded from sjp.sagepub.com at UCSF LIBRARY & CKM on March 8, 2015
