Atherosclerosis, 21 (1975) 61-76
0 Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands
61
EFFECT OF CIGARETTE SMOKING ON LIPIDS, LIPOPROTEINS, BLOOD COAGULATION, FIBRINOLYSIS AND CELLULAR COMPONENTS OF HUMAN BLOOD
J. D. BILLIMORIA,
HELEN POZNER,
B. METSELAAR,
F. W. BEST*
AND
D. C. 0. JAMES
Department of Chemical Pathology, Westminster Medical SchooI, Research Laboratory, Udall Street, and the Blood Transfusion Unit, Westminster Hospital, London S WI (Great Britain) (Received July lst, 1974) (Accepted August 6th, 1974)
SUMMARY
A group of 187 volunteers aged 30-60 years were divided into heavy and light smokers and non-smokers. Heavy smokers smoked over 20 cigarettes per day and light smokers between 5-15, averaging about 7 cigarettes per day. (1) Compared with male non-smokers (N.S.), the male heavy smokers (H.S.) had a higher fasting serum turbidity, higher levels of cholesterol (mainly confined to the ester cholesterol fraction) and higher levels of serum phospholipids and triglycerides. In this H.S. group, the esterified fatty acid index (EFI) of beta and pre-beta lipoprotein were also raised as shown by paper and cellulose acetate electrophoresis. The Stypven clotting times were also shorter than for the N.S. group. (2) Increased levels of triglycerides, pre-beta EFI, longer fibrinolysis times and shorter Stypven times were recorded in the group of female heavy smokers. Changes in cholesterol levels, beta EFI, phospholipids and fasting serum turbidity were not seen in this group. (3) Again, in the male heavy smokers, haematocrit and haemoglobin levels and mean corpuscular volumes were raised. The white cell count (WBC) was very significantly raised and a differential count indicated rises in neutrophils and lymphocytes. These changes were not significant in the female H.S. group. (4) No significant changes were found in the group of light smokers but a trend was usually seen that suggested that these changes were influenced by the number of cigarettes smoked. We are indebted to the Tobacco Research Council and The Westminster Hospital Endowment Fund for generous financial assistance. * Medical Officer of I.C.I., Administrative Headquarters, I.C.I. House, Milbank, London SWl, Great Britain.
62
J. D. BILLIMORIA,
H. POZNER,
B. METSELAAR,
F. W. BEST, D. C. 0. JAMES
Key words: Cell counts - Cigarette smoking - Fasting serum turbidity - Fibrinolysis Lipids - Lipoprotein ester$edfatty
acid index - Stypven coagulation
INTRODUCTION
Hammond and Horn1 correlated mortality data and the smoking histories of 187,783 men and showed that overall death rates and lung cancer death rates were higher in smokers than non-smokers. Deaths due to coronary heart disease were more than twice as high for regular cigarette smokers who smoked over 20 cigarettes per day. Similar findings have been reported in the U.S. Public Health Service and in the combined Albany and Framingham studies 213.Numerous prospective studies-7 have revealed highly consistent results implicating smoking as a high risk factor in coronary disease. There is relatively less experimental evidence to explain this vast epidemiological data. Higher serum cholesterol levels have been demonstrated by several authors+1s in cigarette users and Gofman et al. l1, Bronte-Stewart12 and Caganova et al.13 have reported lipoprotein patterns in smokers which were similar to those shown in patients with ischaemic heart disease and significantly different from non-smokers. Recently, acute effects of cigarette smoking have been reported although conflicting views exist on the significance of such effects. An acute rise in serum free fatty acid levels was demonstrated by Kershbaum et al.14, a greater response being shown in patients with ischaemic heart diseaseIs. Konttinen and Rajalsalmils reported a postprandial rise in serum triglycerides which was less in a smoking than a nonsmoking group. No change was found by some authorsl7J* in serum cholesterol, triglyceride and phospholipid levels after smoking. Some reports also exist over the acute effect of smoking on blood coagulation 1sp2s.Chronic effects (with the exception of our own work) have yet to receive attention. Carbon monoxide levels, cited as an atherogenic factor, have been shown to fall rapidly to zero on abstinence from smoking21. METHODS
Volunteers
In the present work, 180 apparently healthy volunteers between the ages 30-70 years have been studied with respect to changes in their blood coagulation, serum lipids and their lipoprotein levels resulting from cigarette smoking. Studies on fibrinolysis and blood cellular components were also carried out in most volunteers. Exclusions. Twelve volunteers were excluded from the study as they were found to have either ECG or haematological abnormalities. Twenty subjects had their blood analysis carried out but were rejected for the final statistical evaluation as they underwent drug therapy during the course of the study. A further 25 subjects who were
SMOKING,
LIPIDS
AND FIBRINOLYSIS
63
previous smokers fell into small groups with variable smoking habits and were not statistically assessed. Where difficulty was encountered in obtaining the full volumes of blood required, only haematological tests were carried out. Eventually, the heavy smoking groups consisted of 36 subjects (23 males and 13 females) who smoked more than 20 cigarettes per day; a light smoking group consisting of 27 subjects (9 males and 18 females) who smoked between 5-15 cigarettes per day and the non-smoking group of 58 subjects (34 males and 24 females). Blood sampling. Volunteers were asked to fast (12 h) and were rested for 0.5 h before withdrawing 40 ml blood from a cubital vein by using a two syringe technique. Plasma from the 1st syringe titrated blood (1 vol. 3.8 % aq.NasCeH507. 2H20 w/v: 9 ~01s. blood) was removed and, after centrifugation for 10 min at 3500 g, it was used for lipid estimation and turbidity measurements. Plasma from the 2nd syringe titrated blood, similarly obtained, was used for Stypven coagulation, euglobulin lysis and fibrinogen determinations. Serum from the 1st syringe blood was used for lipoprotein electrophoresis, triglyceride and esterified fatty acid determination; 2nd syringe blood (EDTA 1 mg/ml) was used for haemoglobin, white cell counts and blood film; 2nd syringe blood [NaF, 1 mg/ml, (COOK)2 3 mg/ml] was used for blood sugar estimation. Whole blood fibrinolytic activity was determined on 4.5 ml 2nd syringe blood diluted with phosphate buffer (500 ml) and 0.5 ml aq. 3.8 % NasC6Hs07. 2H20 w/v at pH 7.4, contained in a siliconed conical flask cooled in crushed ice. Analytical methods Total cholesterolaa, free cholesterol23 and triglyceridesa* were measured by the methods indicated. Esterified fatty acids were estimated by our modification25 of the method of Rapport and Alonzoae. Fibrinogen levels27 and plasma euglobulin lysis times2s, platelet counts2Q and platelet adhesiveness? were measured by established methods. Stypven clotting timesal. A mixture of 2.5 ml N/l0 NaCl and 2.5 ml of M/40 CaCla was added to a 5 ml vial of Russell Viper venom. Citrated plasma (0.1 ml) and 0.1 ml physiological saline were warmed to 37°C in a water bath. The venom/CaCls mixture (0.1 ml prewarmed to 37°C) was added and the stop watch was started. The time for the first appearance of a clot was noted. Whole blood$brinolysis was measured by a time lapse photographic technique which represents a modification of the methods of Fearnley et aZ.32and Lackner and Gossens3. In principle the diluted blood was clotted with thrombin and, after refrigeration to preserve the heat-labile factor, the clot was retracted on a platinum gauze (+ cm sq.) fused to a glass rod and held in a fibrinolysis tube so that the gauze with the retracted clot was below the surface of the defibrinated blood. The tubes were incubated at 37°C in a thermostatically controlled air cabinet and hourly photographs of the gauzes were automatically recorded. After the first hour the free red cells settled to the bottom of the tube and the progress of the lysis of the red clot on the gauze was photographed through the clear serum. On printing the negatives the end point of the lysis was recorded when the clear gauze was observed free from any clots.
64
J. D. BILLIMORIA,
H. POZNER,
B. METSELAAR,
F. W. BEST, D. C. 0. JAMES
Serum turbidity was measured in a nephelometer (Evans Electroselenium, Middlesex) as described previously31. Lipoproteins. Paper electrophoresis was carried out by the method of Lees and Hatchs4 with minor modifications as reported by us previouslyz5. Cellulose acetate electrophoresis was carried out on Sepraphore III; the cellulose strips were stained with oil red 0, bleached with 1% w/v NaOCl soln. for 1 min and cleared by immersion in methanol containing 10% v/v glacial acetic acid. They were mounted on glass slides and the band dye concentration obtained by transmission scanning using the Gelscan Scanner (Gelman Instrument Co., Ann Arbor, Michigan 48106). The relative dye uptake (RDU) of each band was calculated as a percentage of the total dye uptake. The RDU/lOO x fatty acid mg/lOO ml of the particular serum gave the esterified fatty index (EFI) of each lipoprotein band25. A similar quantitation was used with paper electrophoresis, but reflectance scanning of the paper strip was used to determine the band positions and the bands were cut out, eluted with 5 % acetic acid and the eluates read at 540 rnp in an absorption photometer. The RDU was calculated and converted to EFI units as above. Cell counts. Red cell and total white cell counts, haemoglobin (Hb), haematocrit (Hc), mean corpuscular volume (MCV), mean corpuscular haemoglobin (MCH) and mean corpuscular haemoglobin concentration (MCHC) were determined by automated techniques using a Coulter counter, Model S. Blood films for differential counts were stained with May-Griinwald/Giemsa using a staining machine. The differential count was only worked out if the film appeared to be normal. Blood sugar was estimated on an Autoanalyser using a ferricyanide reduction method. Euglobulin lysis times were measured by a modification of the method of Buckel12s. Thrombin (50 ,ug/ml) was used for clotting instead of CaCls and a phosphate buffer (pH 7.3) was used instead of the borate buffer (pH 9.0). The water used was glass distilled. RESULTS
Cholesterol
Fig. 1 shows the levels of total serum cholesterol, in 35 male and female heavy smokers compared with 49 male and female non-smokers. The overall mean cholesterol value was greater for heavy smokers 236 f 8 mg/lOO ml than for non-smokers (mean value) 215 f 6 mg/lOO ml but was not significantly different. When males and females were analysed separately, significant differences were found. As shown in Fig. 1 the mean value for female heavy smokers is almost the same as for female nonsmokers being 221 mg/lOO ml and 215 mg/lOO ml respectively. Comparison of the 23 male heavy smokers (243 f 11 mg/lOO ml) with the 27 male non-smokers (215 & 6), showed differences which were significant (P < 0.025). The results of mean cholesterol values for male light smokers (231 f 16) did not differ significantly from the non-smokers (215 f 6) but showed a distinct trend which
65
SMOKING, LlPIDS AND FIBRINOLYSIS .m Heavy Smokers
~mg/lOOml) Non-Smokers
*
40Q
400_
100 (mean ,,,,,,...,..,.,,,) Female o (mean mean -
100. Male . OVerall
0 - _ _)
Fig. 1. Serum cholesterol (mg/lOO ml) of heavy smokers and non-smokers. Mean age: males, nonsmokers(N.S.)44.6& 7.7;heavysmokers(H.S.)44.7 & 7.2;females(N.S.)43.5 +9.3;(H.S.)44.8* 8.4. For standard deviation and P-values see text.
suggested that the level of serum cholesterol in men increased in relation to the number of cigarettes smoked. Serum free and ester cholesterol were also measured in the above volunteers. No differences in free cholesterol levels were found and ester cholesterol values were raised only in the male heavy smokers: 205 f 11 mg/lOO ml compared with 171 f 8 in the corresponding non-smokers (P < 0.01). Triglycerides
Fig. 2 shows the results of the long term effect of smoking on serum triglycerides in 28 male and female heavy smokers compared with 49 male and female non-smokers. TRlGLYCERlDES (mg/lOOml) Non-Smokers
Heavy
Smokers
280.
.
.
.
0 2OQ
0
2T 200.
.
0
0
.
0 _ - _ -. _ - - -.l0Qo
“. o
ca 0
a
.
.
- - . *.
:
0 .
Female o (mean Overall mean __
*.
0
* *
- - -)
.
. .
1oao . 0 ,., *t .._*.,,.....................~.... ,, .o,,.,,,.,,.. +9_Q_.o.._._,p._ 0.01, and in male heavy smokers mean triglycerides (134 f 15) were also raised compared with male nonsmokers (82 f 7), P > 0.0025. Further inspection of Fig. 2 showed that values above 180 mg/lOO ml serum triglycerides occurred only in heavy smokers whereas values below 50 mg/lOO ml were found only in non-smokers. Our upper normal limit (mean f SD.) in this laboratory without taking into account the smoking habit was 175 mg/lOO ml. The mean triglyceride level of the 16 light smokers (males and females) was 93 f 8 mg/lOO ml and that of the corresponding non-smokers 78 f 5 mg/lOO ml: this difference was not statistically significant, but again indicated that the rise in triglyceride depended on the number of cigarettes smoked. Effect of smoking on turbidity Fig. 3 shows differences found in plasma turbidity between male and female heavy smokers and non-smokers. The 36 heavy smokers were found to have a higher mean value, 26 * 2 nephelometric units (N.U.) compared with 20 f 1 N.U. for the 48 non-smokers (P < 0.025). Nearly all the higher values were attributable to male heavy smokers (mean 30 f 3.0); the non-smokers had a mean of 21 f 1.2 (P < 0.0025). Light smokers did not show any differences from the non-smokers even though the males and females were separately analysed. The mean turbidity value for heavy-smoking females did not differ from that of non-smokers.
Fig. 3. Serum turbidity (nephelometric units) of heavy smokers and non-smokers. Mean age: males, non-smokers (N.S.) 43.7 i: 6.9; heavy smokers (H.S.) 44.7 A 7.2; females (N.S.) 44.0 f 9.4; (H.S.) 45.2 + 8.2 For standard deviation and P-values see text.
SMOKING, LIPIDS AND FIBRINOLYSIS
67
EJSect of smoking on total phospholipids
Male heavy smokers had higher total plasma phospholipid levels of 241 j, 12 mg/lOO ml compared with 215 f 9.5 mg/lOO ml for the corresponding non-smokers (P < 0.05). This difference was not exhibited by females and there was no statistical difference observed between smokers and non-smokers when male and female results were combined. E#ect of smoking on lipoproteins (using paper electrophoresis) (a) fi-Lipoprotein
Fig. 4 shows the result of smoking on the esterified fatty acid index (EFI) of @lipoproteins in 38 male and female heavy smokers compared with 49 male and female non-smokers. The lipoproteins were separated by paper electrophoresis and quantitated as indicated above. The overall mean value for males and females is raised in the heavy smoking group (159 f 7.5 EFI units) compared with the non-smoking group (137 f 4.5 EFI units). As shown in Fig. 4, the mean value for female heavy smokers is almost equal to the mean value for female non-smokers, being 132 f 9 and 134 f 11 EFI units respectively. The mean EFI for ,%lipoprotein in male heavy smokers was 173 f 9 compared with 140 f 5 for non-smokers. This large difference was highly significant (P < 0.0025). These lipoprotein results obtained on division of data according to the smoking habit correlate well with the results of cholesterol measurements (shown above). As ,&lipoprotein contains most of the ester cholesterol, the results confirm the increased
BETA
(paper
electrophoresis)
Fig. 4. /?-Lipoproteins (EFI units, paper electrophoresis) of heavy smokers and non-smokers. Mean age: males, non-smokers (N.S.) 45.1 5 7.8; heavy smokers (H.S.) 44.1 + 6.7; females (N.S.) 44.0 f 9.2; (H.S.) 44.5 f 8.8. For standard deviation and P-values see text.
68
J. D. BILLIMORIA,
H. POZNER,
B. METSELAAR,
F. W. BEST, D. C. 0. JAMES
level of this lipid in male heavy smokers. Light smokers did not differ from nonsmokers in their cholesterol levels, although the mean cholesterol value in the male group lay between the mean values for non-smokers and heavy smokers. Exactly the same situation arises in the case of B-lipoprotein EFI values: the mean value for male light smokers is 155 & 23 EFI units which lies between the corresponding heavy smokers and non-smokers values of 173 & 9 and 140 f 5 respectively. (b) Pre-/l-lipoprotein
Fig. 5 shows the results of the long-term effect of smoking on EFI values for pre/I-lipoprotein, obtained by paper electrophoresis, for 33 male and female heavy smokers compared with 49 male and female non-smokers. The mean values of pre-/Ilipoprotein EFI of the heavy smokers were all raised, and approximately equally for males and females. The mean EFI for pre-j3-lipoprotein in the group of heavy smokers was 61 f 7 EFI units compared with non-smokers 25 & 3 EFI units (0.0005 < P > 0.0025). As shown in Fig. 5, in the non-smoking group, 13 cases (6 males, 7 females) or 27% did not show any pre-@-lipoprotein bands, whereas in the heavy smoking group only 2 cases (1 male, 1 female) out of 33 (i.e. 6 % of the cases) showed no pre-P band. Results of males and females, separately analysed with respect to smoking and pre-/?-lipoprotein levels, showed similar differences: male non-smokers and female non-smokers had mean pre-ig-lipoprotein values that were approximately the same, while male heavy smokers and female heavy smokers had equally raised means. It was found that the pre-/l-lipoprotein values for light smokers showed very significant differences when compared with non-smokers. Male light smokers as well as female light smokers showed the following raised mean pre-@-lipoprotein values: 56 f 8 and 58 & 6 respectively for light smokers compared with 28 f 4 and 25 f 3 for non-smokers (in each case P < 0.0025). EFI
( paper electrophoresis) Non-Smokers
Heavy Smokers 2OQ
20Q
0
.
.. *,,,,,,,,,,,,.,. ,**.l*.l.l.~. ..” ...***. *.*.** .“.I,,..,.,*.,.., ~._._~.L._.._.;._.-._._.; ___ . . . . O_ Female
0
( mean
---.---)
a. . .* 0 0 00 1)11 11#1*1 111*,*, O&,y*l*l.ll.llll CR I.I 1111k.1 r..‘ ,,##, :. * 1 l.#l*ll*l. 4**l.lll*ll.l*llll,, o-‘-‘-‘0 -’ fb.-‘o.- ._._.-.-.r.-‘-__‘_ O& 0 Male . ( mean ,,,,,,.,,..,,.,,)Overa 11mean_
Fig. 5. Pre-&lipoproteins (EFI units, paper electrophoresis) of heavy smokers and non-smokers. Mean age: males, non-smokers (N.S.) 45.1 & 7.8; heavy smokers (H.S.) 44.1 f 6.7; females (N.S.) 44.0 ;t 9.2; (H.S.) 44.5 & 8.8. For standard deviation and P-values see text.
SMOKING, LIPIDS AND FIBRINOLYSIS
69
Separation of lipoproteins by cellulose acetate electrophoresis
During winter months we picked 43 normal subjects randomly from the 30-60 year old group, to check on any seasonal variation in cholesterol and triglyceride values. Electrophoresis was carried out on the sera of these subjects using polycellulose acetate (Sepraphore III) as the separating and support medium. The subjects consisted of 20 heavy smokers (11 females and 9 males) and 23 non-smokers (9 females and 14 males). Samples were not taken from light smokers. (a)
B-Lipoprotein
Fig. 6 shows the results of @-lipoprotein EFI for the above volunteers and the differences found when they were divided into heavy smokers and non-smokers. The mean EFI value for male and female heavy smokers (191 f 8) was greater than that obtained for male and female non-smokers (168 f 6), P < 0.0125. The difference in means between heavy smokers and non-smokers was 22 EFI units using Sepraphore electrophoresis, which compared favourably with a difference of 23 units with paper electrophoresis. When the actual figures from the two electrophoretic procedures were compared it could be seen that EFI on paper were approximately 30 units less than those on cellulose acetate. When males and females were separately analysed the differences between smokers and non-smokers found by Sepraphore electrophoresis also corresponded to those found by the paper electrophoresis technique. /?-Lipoprotein EFI in 11 female heavy smokers compared with 9 non-smokers were not significantly different. In contrast, a large increase was found in the mean value for 9 males who were heavy smokers compared with 14 male non-smokers.
BETA-
(cellulose acetate electrophoresis)
Heavy Smokers 3T----l
Female
Non-Smokers 300{_
o ( mean-.-.-.) Male Overall mean -
. (mean ..............)
Fig. 6. B-Lipoproteins (EFI units, cellulose acetate electrophoresis) of heavy smokers and nonsmokers. Mean age: males, non-smokers (N.S.) 42.9 & 8.7; heavy smokers (H.S.) 47.8 i 6.8; females (N.S.) 42.9 =t 8.7; (H.S.) 47.1 f 8.0. For standard deviation and P values see text.
J. D. BILLIMORIA, H. POZNER, B. METSELAAR, F. W. BEST, D. C. 0. JAMES PRF-EFTA
Female
o
F F ~(celluloseacetateelectrophoresis)
( -.-.Overall
)
Male mean -
. ( mean ,,., ,,,..,..,,. )
Fig. 7. Pre-p-lipoproteins (EFI units, cellulose acetate electrophoresis) of heavy smokers and nonsmokers. Mean age: males, non-smokers (N.S.) 42.9 & 8.7; heavy smokers (H.S.) 47.8 & 6.8; females (N.S.) 50.7 ?t 0.1; (H.S.) 47.1 f 8.0. For standard deviation and P values see text.
(b) Pre-/I-lipoprotein
Fig. 7 shows the results of pre-p-lipoprotein EFI for 43 subjects and the differences between the mean values for heavy smokers and non-smokers. The overall mean value for male and female heavy smokers (91 EFI) is greater than that for male and female non-smokers (61 EFI, P < 0.001). These results (Fig. 7) compared with those of paper electrophoresis (Fig. 5) show that a difference of 30 EFI units was observed on cellulose acetate and 36 EFI units on paper electrophoresis between the pre-b-lipoproteins of smokers and nonsmokers. This slight discrepancy will be clear when Figs. 5 and 7 are compared. On paper, the separation of pre-/3 and ,~5is incomplete when only small amounts of pre-p are present, as the pre-p is retained in the p band. Thus it is seen in Fig. 5 that with those non-smokers where only traces of pre-p are present, these are not detected and are recorded as zero. On cellulose acetate (Fig. 7) even the traces are detected and have a finite value. The small loss of pre-B from the non-smokers on paper electrophoresis accounts for the greater difference of pre-Slipoproteins observed between smokers and non-smokers on this media. Euglobulin lysis time and Stypven clotting time
These tests are shown in Fig. 8. Euglobulin lysis times were significantly longer for both male (286 f 21) and female (315 f 20) groups of heavy smokers compared with non-smokers (respectively 236 f 18 and 193 f 17; P < 0.05 for men and P < 0.0005 for women). The difference was more marked in the female group where even light smokers (240 f 30) showed a marked and significant decrease in fibrinolytic activity. Stypven clotting times (SCT) of heavy smokers were significantly shortened in
71
SMOKING, LIPIDS AND FIBRINOLYSIS NON-S
M OKERS
HEAVY
SMOKERS
Fig. 8. Euglobulin lysis times (min) and Stypven clotting times (set) of heavy smokers and nonsmokers. Mean age: males, non-smokers (N.S.) 44.3 + 7.9; heavy smokers (H.S.) 45.2 j, 7.3; females (N.S.) 43.0 & 9.3; (H.S.) 43.7 f 7.6. For standard deviation and P values see text.
both males and females from 39.5 f 1.3 and 41 f 1.4 to 33 f 0.9 and 33 f 1.6 respectively (P < 0.0005 for both sexes). Although the SCT of both sexes in the light smokers was also shorter than that of non-smokers, the differences were not significant. Haematocrit and haemoglobin levels
These are shown in Fig. 9. Male heavy smokers gave an haematocrit of 47.4% compared with 44.9 % for non-smokers (P < 0.0025). No differences were observed in the female groups and changes in male light smokers were not significant. HEAVY
NON-SMOKERS
SMOKERS
haematocrit(“,,)
I
I I
I
Fig. 9. Haematocrit (%) and haemoglobin (g/ml) of heavy smokers and non-smokers. Mean age: males, non-smokers (N.S.) 44.3 + 7.9; heavy smokers (H.S.) 45.2 i 7.3; females (N.S.) 43.0 :t 9.3; (H.S.) 43.7 i 7.6. For standard deviation and P values see text.
72
J. D. BILLIMORIA, H. POZNER, B. METSELAAR, F. W. BEST, D. C. 0. JAMES
The haemoglobin of male heavy smokers 15.4 g/ml was significantly higher than non-smokers 14.6 g/ml (P < 0.0025). No significant changes were found in the groups of females. Mean corpuscular volumes (MCV)
and haemoglobin (MCH)
A small increase in MCV was found in male heavy smokers (P < 0.05). No significant changes were found in the women. The MCH of male heavy smokers (29.7 pg) was slightly increased and significantly greater than for non-smokers (28.6 pg; P < 0.005). No significant changes were observed in women. The mean corpuscular haemoglobin concentration (MCHC) showed no changes accountable to smoking. White cell ( WBC) and diflerential counts
The results are shown in Fig. 10. The WBC in male heavy smokers was significantly increased (8.5 & 0.5) x 10s compared with (5.5 f 0.2) x 103 in non-smokers (P < 0.005). There was a small increase in WBC in the female heavy smokers but it was not significant (0.05 < P < 0.10). The neutrophils showed a significant increase in heavy smokers (H.S.) of both sexes compared with non-smokers (N.S.), viz. male H.S. 5.0 x 103, male N.S. 3.1 x 10s (P < 0.0005); female H.S. 4.5 x 103, female N.S. 3.5 x 10s (P < 0.05). A similar
*.*. L-H
0.
6.
.
*
.
.
*
4.
*
*
2r_”
. ..*
.*:...
..
.*
. .
* **t * ** I * ** * 1
*
.
.
*
*.: **.
.
*
. ** **** ****** *
*
*
*
I
I
lymphocytes
(x103)
II
Fig. 10. Total white cell counts, neutrophils and lymphocyte counts of heavy smokers and nonsmokers. Mean age: males, non-smokers (N.S.) 44.3 k 7.9; heavy smokers (H.S.) 45.2 & 7.3; females (N.S.) 43.0 % 9.3; (H.S.) 43.7 f 7.6. For standard deviation and P values see text.
SMOKING,
LIPIDS AND FIBRINOLYSIS
73
trend was seen in light smokers but the difference was not statistically significant. The results for lymphocyte counts for heavy (H.S.) and light (L.S.) male smokers were higher than those for non-smokers (N.S.). Thus for men H.S. = 3.0 x 103, N.S. = 2.2 x 10s (P < 0.025), L.S. = 2.6 x 10s (0.10 > P > 0.05). Changes were not apparent in the female groups. Eosinophils: Male H.S. gave values of 264 x lo3 compared with N.S. 143 x lo3 (P < 0.05) and L.S. 189 x 10s (not significant). Female H.S. gave values of 144 x lo3 compared with N.S. 78 x 10s (P -C 0.05) and L.S. 152 x 10s (P < 0.05). Busophils: Whereas basophils were rarely noted in films from non-smokers they were always present in blood films from smokers, especially for the male heavy smokers. No changes due to smoking were observed in blood sugar levels, blood red cell counts, fibrinogen levels or platelet adhesiveness properties.
DISCUSSION
A significant difference was shown in the Stypven coagulation test (SCT), where a shortening of plasma clotting time was demonstrated in heavy smokers even after they had abstained from smoking for 12 h. Previously only the acute changes in coagulation have been studied immediately after smoking I or 2 cigarettes; such changes were probably a result of increased noradrenalin. As light smokers also show a shortened SCT - although to a much lesser extent - our observation is likely to relate to hypercoagulability due to a chronic effect of cigarette smoking. Increase in serum total cholesterol levels were confined to male heavy smokers : an increased index of p-lipoproteins and the fact that the increase was found in the ester cholesterol fraction is additional confirmation of the above finding. However, we have observed a substantial seasonal variation in cholesterol levels and, as this variation is as large as the difference between smokers and non-smokers, it is important - as here - to test the two groups during the same season. Significantly increased levels of triglyceride have been found in heavy smokers of both sexes and this triglyceride is contained in the pre-/?-lipoprotein fraction as shown by the increased EFI index of this lipoprotein. Increased pre-/3-lipoproteins have have shown been reported by Besterman35, while both Albrinkss and Carlson that triglycerides (TG) are elevated in ischaemic heart-disease (IHD). We have also observed during routine tests carried out on patients over the last 5 years that a majority of IHD patients show an increased TG level and a raised pre-8 EFI index. In fact a larger number of IHD patients have triglyceride levels in excess of 175 mg/lOO ml than cholesterol levels over 300 mg/lOO ml. (The above levels were the upper limits (mean f 2 S.D.) of triglycerides and cholesterol observed for normal subjects aged 40-70 years in our laboratory.)
74
J. D. BILLIMORIA, H. POZNER, B. METSELAAR, F. W. BEST, D. C. 0. JAMES
In our experience nephelometry of fasting visible lipaemia in patients with IHD invariably shows a small fasting serum turbidity of 30-40 units, compared with our range of lo-26 units in normal subjects (see refs.38939).The present results show a very similar small rise in turbidity in male heavy smokers. Fibrinolytic activity tested by both whole blood lysis and euglobulin lysis tests is reduced in heavy smokers; a similar decrease has also been observed in patients with IHD. The raised haematocrit value found in the male heavy smokers is not accompanied by a corresponding rise in red cell count. This is partly explained by a corresponding rise in the mean cell volume of this group, with a possible contribution from a reduction of plasma volume caused by the vasoconstrictive and diuretic effects of nicotine. The raised carboxyhaemoglobin content of the blood following chronic tobacco smoking would result in a degree of anoxia which, in turn, would stimulate erythropoesis and thus at least partly explain the raised Hb level. The increased total white count is reflected in nearly all the white cell groups and is most evident in the heavy smokers. These observations agree with those of Corre et ~1.40.Although no explanation is apparent, it was noted that basophils were invariably present in the blood films of smokers. It has been suggested that the living style of smokers may differ from that of nonsmokers, and that the increased cholesterol level in heavy smokers may be genetically associated with but not directly related to the cause of the urge to smoke. “Styles of living” are difficult to assess in experimental studies and, taking into account the numerous changes other than in the cholesterol level observed in both male and female smoker&, we have come to the operational conclusion that heavy smokers are a “high risk” group for coronary disease. It is doubtful that much is gained in this instance by consideration of nebulous cause-and-effect relationships. In most laboratories, including our own, the upper normal limits for serum lipids have been recorded without reference to the smoking habit; it now appears that this factor would have to be taken into account in determining a state of hyperlipidaemia. This step would undoubtedly cause the upper limit to be reduced in normal subjects who are non-smokers. ACKNOWLEDGEMENTS
We thank Mr. K. L. E. Ling for skilled technical assistance. We are also grateful to all volunteers from I.C.I. who participated in this work; to Mrs. Joyce Porter (Med. Dept., I.C.I.) who supervised the arrangements: to Mrs. D. V. Winter and Dr. J. Thurston (Cardiac Dept., Westminster Hospital) for respectively preparing and interpreting the ECGs. REFERENCES 1 HAMMOND, E. G. AND HORN, D., The relationship between human smoking habits and death rates (a follow-up study of 187,783 men), J. Amer. Med. Ass., 155 (1954) 1316.
SMOKING, LIPIDS AND FIBRINOLYSIS
75
2 DOYLE, J. T., DAWBER, T. R., KANNEL, W. B., KINCH, S. H. AND KAHN, M. A., The relationship of cigarette smoking to coronary heart disease, J. Amer. Med. Ass., 190 (1964) 886. 3 DOYLE, J. T., DAWBER, T. R., KANNEL, W. B., HESLIN,A. S. AND KAHN, M. A., Cigarette smoking and coronary heart disease, New Engl. J. Med., 266 (1962) 796. 4 PAUL, O., LEPPER,M. H., PHELAN, W. H., DEPERTUIS,G. W., MACMILLAN, A., MCKEAN, H. AND PARK, H., A longitudinal study of coronary heart disease, Circulation, 28 (1963) 20. 5 BORHANI,N. O., HECHTER,H. H. AND BRESLAW,L., Report of a ten year follow-up study of the San Francisco longshoremen. Mortality from coronary heart disease and from all causes, J. Chron. Dis., 16 (1963) 1251. 6 HAMMOND,E. C., Smoking in relation to mortality and morbidity. Findings in first thirty-four months of follow-up in a prospective study started in 1959, Cancer Inst., 32 (1964) 1161. 7 SHAPIRO,S., WEINBLATT,E., FRANK, C. W. AND SAGER, R. V., The H.I.P. study of incidence and prognosis of coronary heart disease: preliminary findings on incidence of myocardial infarction and angina, J. Chron. Dis., 18 (1965) 527. 8 KARVONEN,M. J., ORMA, E., KEYS, A., FIDANZA, F. AND BROZEK, J. Cigarette smoking, serum cholesterol, blood pressure and body fatness, Lancer, 1 (1959) 492. 9 THOMAS, C. B., Familial and epidemiological aspects of coronary disease and hypertension, J. Chron. Dis., 7 (1958) 198. 10 THOMAS, C. B., Familial patterns in hypertension and coronary heart disease, Circulation, 20
(1959) 25. 11 GOFMAN, J. W., LINDREN, F. T., STRISOWER,B., DE LALLA, O., GLAZIER, F. AND TAMPLIN, A., Cigarette smoking, serum lipoproteins and coronary disease, Geriatrics, 10 (1955) 349. 12 BRONTE-STEWART,B., Cigarette smoking and ischaemic heart disease, Brit. Med. J., 1 (1969) 379. 13 CAGANOVA, A., CAGAN, S. AND SIMKO, V., Bratisl. Lek. Listy, 50 (3) (1968) 387. 14 KERSHBAUM,A., BILLET,S., DICKSTEIN,I. R. AND FEINBERG,L. J., Effect of cigarette smoking and nicotine on serum free fatty acids. Based on a study in the human subject and the experimental animal, Circ. Res., 9 (1961) 631. 15 KERSHBAUM,A. AND BILLET, S., Cigarette smoking and blood lipids, J. Amer. Med. Ass., 187 (1964) 32. 16 KONTTINEN,A. AND RAJALSALMI,M., Effect of heavy cigarette smoking on postprandial triglycerides, free fatty acids and cholesterol, Brit. Med. J., 1 (1963) 850. 17 SHORT, J. J. AND JOHNSON,H. J., A direct comparison of the reaction of the human system to tobacco smoke and adrenalin, J. Lab. Clin. Med., 24 (1939) 590. 18 PAGE, I. H., LEWIS, L. A. AND MOINUDOLIN,M., Effect of cigarette smoking on serum cholesterol and lipoprotein concentrations, J. Amer. Med. Ass., 171 (1959) 1500. 19 BLACKBURN,H., ORMA, E., HASTEL,G. AND PUNSAR,S., Tobacco smoking and blood coagulation. Acute effect on plasma Stypven time, Amer. J. Med. Sci., 238 (1959) 448. 20 AMBRUS, J. L. AND MINK, I. B., The effect of cigarette smoking on blood coagulation, Clin.
Pharmacol. Therap., 5 (1964) 428. 21 ROSENBERG,A., Fall in carbon monoxide blood levels after stopping smoking, Lancet, 1 (1972) 593. 22 BILLIMORIA,J. D. AND JAMES,D. C., Two improved methods for serum cholesterol estimation,
Clin. Chim. Acra, 5 (1960) 644. 23 SPERRY, W. M. AND WEBB, M., Revision of Schoenheimer-Sperry mination, J. Biof. Chem., 187 (1950) 97.
method for cholesterol deter-
24 VAN HANDEL, E. AND ZILVERSMIT,D. B., Micromethod for the determination of serum triglycerides, J. Lab. Clin. Med., 50 (1957) 152. 25 BILLIMORIA,J. D., FAHMY, M. F., JEPSON,E. M. AND MACLAGAN, N. F., The use of the esterified fatty acid index in the classification and quantitation of hyperlipoproteinaemia, Atherosclerosis, 14 (1971) 359. 26 RAPPORT, M. M. AND ALONZO, N., Photometric determination of fatty acid groups in phospholipids, J. Biol. Chem., 217 (1955) 193. 27 INGRAM,C. I., The determination of plasma fibrinogen by the clot weight method, Biochem. J., 51 (1952) 583. 28 BUCKELL, M., The effect of citrate on euglobulin methods of estimating fibrinolytic activity, J. Clin. Pathol., 11 (1950) 403. 29 BRICKER, C. AND CRONKITE,E. P., Morphology and enumeration of human blood platelets, J. Appl. Physiol., 3 (1950) 365.
76
J. D.
BILLIMORIA, H. POZNER, B. METSELAAR, F. W. BEST, D. C. 0. JAMES
30 HELLEM,A. J., The Adhesiveness of Human Blood Platelets in vitro (Norwegian Monographs on Medical Science), Oslo University Press, Oslo, 1960. 31 MACLAGEN,N. F., BILLIMORIA,J. D. AND CURTIS,C., Lipaemia and blood coagulation with special reference to the stypven technique, Lancer, 2 (1958) 865. 32 FEARNLEY,G. R., BALMFORTH,G. V., AND FEARNLEY,E., Evidence of a diurnal fibrinolytic rhythm, with a simple method of measuring natural fibrinolysis, Clin. Sci., 16 (1957) 645. 33 LACKNER,H. AND GOOSEN, C. C., Fibrinolytic activity of blood: photographic determination of clot lysis time, Acta Haematol., 22 (1959) 58. 34 LEES, R. S. AND HATCH, F. T., Sharper separation of lipoproteins by paper electrophoresis in albumin containing buffer, J. Lab. Clin. Med., 61 (1963) 518. 35 BESTERMAN,E. M., Lipoproteins in coronary artery disease, Brit. Heart J., 19 (1957) 503. 36 ALBRINK,M. J. AND MAN, E. B., Serum triglycerides incoronary artery disease, Arch. Intern. Med.,
103 (1959) 4. 37 CARLSON, L. A.
AND BOTTIGEN,L. E., Stockholm prospective study. lschaemic heart disease in relation to fasting values of plasma triglycerides and cholesterol, Lancer, 1 (1972) 865. 38 MITCHELL, J. R. A. AND BRONTE-STEWART,B., Alimentary lipaemia and heparin clearing in ischaemic heart-disease, Lancer, 1 (1959) 167. 39 JAMES,D. C., DRYSDALE,J., BILLIMORIA,J. D., WHEATLEY,D., GAVEY, C. J. AND MACLAGAN, N. F., Lipaemia and blood-coagulation defects in relation to ischaemic heart-disease, Lancer, 2
(1961) 798. 40 CORRE, F., LELLOUCH,J. AND SCHWARTZ, D., Smoking and leucocyte-counts. miological survey, Lancer, 1 (1971) 632.
Results of an epide-
41 POZNER,H. AND BILLIMORIA,J. D., Effect of smoking on blood clotting and lipid and lipoprotein levels, Lancer, 1 (1970) 1318.
0 Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands
61
EFFECT OF CIGARETTE SMOKING ON LIPIDS, LIPOPROTEINS, BLOOD COAGULATION, FIBRINOLYSIS AND CELLULAR COMPONENTS OF HUMAN BLOOD
J. D. BILLIMORIA,
HELEN POZNER,
B. METSELAAR,
F. W. BEST*
AND
D. C. 0. JAMES
Department of Chemical Pathology, Westminster Medical SchooI, Research Laboratory, Udall Street, and the Blood Transfusion Unit, Westminster Hospital, London S WI (Great Britain) (Received July lst, 1974) (Accepted August 6th, 1974)
SUMMARY
A group of 187 volunteers aged 30-60 years were divided into heavy and light smokers and non-smokers. Heavy smokers smoked over 20 cigarettes per day and light smokers between 5-15, averaging about 7 cigarettes per day. (1) Compared with male non-smokers (N.S.), the male heavy smokers (H.S.) had a higher fasting serum turbidity, higher levels of cholesterol (mainly confined to the ester cholesterol fraction) and higher levels of serum phospholipids and triglycerides. In this H.S. group, the esterified fatty acid index (EFI) of beta and pre-beta lipoprotein were also raised as shown by paper and cellulose acetate electrophoresis. The Stypven clotting times were also shorter than for the N.S. group. (2) Increased levels of triglycerides, pre-beta EFI, longer fibrinolysis times and shorter Stypven times were recorded in the group of female heavy smokers. Changes in cholesterol levels, beta EFI, phospholipids and fasting serum turbidity were not seen in this group. (3) Again, in the male heavy smokers, haematocrit and haemoglobin levels and mean corpuscular volumes were raised. The white cell count (WBC) was very significantly raised and a differential count indicated rises in neutrophils and lymphocytes. These changes were not significant in the female H.S. group. (4) No significant changes were found in the group of light smokers but a trend was usually seen that suggested that these changes were influenced by the number of cigarettes smoked. We are indebted to the Tobacco Research Council and The Westminster Hospital Endowment Fund for generous financial assistance. * Medical Officer of I.C.I., Administrative Headquarters, I.C.I. House, Milbank, London SWl, Great Britain.
62
J. D. BILLIMORIA,
H. POZNER,
B. METSELAAR,
F. W. BEST, D. C. 0. JAMES
Key words: Cell counts - Cigarette smoking - Fasting serum turbidity - Fibrinolysis Lipids - Lipoprotein ester$edfatty
acid index - Stypven coagulation
INTRODUCTION
Hammond and Horn1 correlated mortality data and the smoking histories of 187,783 men and showed that overall death rates and lung cancer death rates were higher in smokers than non-smokers. Deaths due to coronary heart disease were more than twice as high for regular cigarette smokers who smoked over 20 cigarettes per day. Similar findings have been reported in the U.S. Public Health Service and in the combined Albany and Framingham studies 213.Numerous prospective studies-7 have revealed highly consistent results implicating smoking as a high risk factor in coronary disease. There is relatively less experimental evidence to explain this vast epidemiological data. Higher serum cholesterol levels have been demonstrated by several authors+1s in cigarette users and Gofman et al. l1, Bronte-Stewart12 and Caganova et al.13 have reported lipoprotein patterns in smokers which were similar to those shown in patients with ischaemic heart disease and significantly different from non-smokers. Recently, acute effects of cigarette smoking have been reported although conflicting views exist on the significance of such effects. An acute rise in serum free fatty acid levels was demonstrated by Kershbaum et al.14, a greater response being shown in patients with ischaemic heart diseaseIs. Konttinen and Rajalsalmils reported a postprandial rise in serum triglycerides which was less in a smoking than a nonsmoking group. No change was found by some authorsl7J* in serum cholesterol, triglyceride and phospholipid levels after smoking. Some reports also exist over the acute effect of smoking on blood coagulation 1sp2s.Chronic effects (with the exception of our own work) have yet to receive attention. Carbon monoxide levels, cited as an atherogenic factor, have been shown to fall rapidly to zero on abstinence from smoking21. METHODS
Volunteers
In the present work, 180 apparently healthy volunteers between the ages 30-70 years have been studied with respect to changes in their blood coagulation, serum lipids and their lipoprotein levels resulting from cigarette smoking. Studies on fibrinolysis and blood cellular components were also carried out in most volunteers. Exclusions. Twelve volunteers were excluded from the study as they were found to have either ECG or haematological abnormalities. Twenty subjects had their blood analysis carried out but were rejected for the final statistical evaluation as they underwent drug therapy during the course of the study. A further 25 subjects who were
SMOKING,
LIPIDS
AND FIBRINOLYSIS
63
previous smokers fell into small groups with variable smoking habits and were not statistically assessed. Where difficulty was encountered in obtaining the full volumes of blood required, only haematological tests were carried out. Eventually, the heavy smoking groups consisted of 36 subjects (23 males and 13 females) who smoked more than 20 cigarettes per day; a light smoking group consisting of 27 subjects (9 males and 18 females) who smoked between 5-15 cigarettes per day and the non-smoking group of 58 subjects (34 males and 24 females). Blood sampling. Volunteers were asked to fast (12 h) and were rested for 0.5 h before withdrawing 40 ml blood from a cubital vein by using a two syringe technique. Plasma from the 1st syringe titrated blood (1 vol. 3.8 % aq.NasCeH507. 2H20 w/v: 9 ~01s. blood) was removed and, after centrifugation for 10 min at 3500 g, it was used for lipid estimation and turbidity measurements. Plasma from the 2nd syringe titrated blood, similarly obtained, was used for Stypven coagulation, euglobulin lysis and fibrinogen determinations. Serum from the 1st syringe blood was used for lipoprotein electrophoresis, triglyceride and esterified fatty acid determination; 2nd syringe blood (EDTA 1 mg/ml) was used for haemoglobin, white cell counts and blood film; 2nd syringe blood [NaF, 1 mg/ml, (COOK)2 3 mg/ml] was used for blood sugar estimation. Whole blood fibrinolytic activity was determined on 4.5 ml 2nd syringe blood diluted with phosphate buffer (500 ml) and 0.5 ml aq. 3.8 % NasC6Hs07. 2H20 w/v at pH 7.4, contained in a siliconed conical flask cooled in crushed ice. Analytical methods Total cholesterolaa, free cholesterol23 and triglyceridesa* were measured by the methods indicated. Esterified fatty acids were estimated by our modification25 of the method of Rapport and Alonzoae. Fibrinogen levels27 and plasma euglobulin lysis times2s, platelet counts2Q and platelet adhesiveness? were measured by established methods. Stypven clotting timesal. A mixture of 2.5 ml N/l0 NaCl and 2.5 ml of M/40 CaCla was added to a 5 ml vial of Russell Viper venom. Citrated plasma (0.1 ml) and 0.1 ml physiological saline were warmed to 37°C in a water bath. The venom/CaCls mixture (0.1 ml prewarmed to 37°C) was added and the stop watch was started. The time for the first appearance of a clot was noted. Whole blood$brinolysis was measured by a time lapse photographic technique which represents a modification of the methods of Fearnley et aZ.32and Lackner and Gossens3. In principle the diluted blood was clotted with thrombin and, after refrigeration to preserve the heat-labile factor, the clot was retracted on a platinum gauze (+ cm sq.) fused to a glass rod and held in a fibrinolysis tube so that the gauze with the retracted clot was below the surface of the defibrinated blood. The tubes were incubated at 37°C in a thermostatically controlled air cabinet and hourly photographs of the gauzes were automatically recorded. After the first hour the free red cells settled to the bottom of the tube and the progress of the lysis of the red clot on the gauze was photographed through the clear serum. On printing the negatives the end point of the lysis was recorded when the clear gauze was observed free from any clots.
64
J. D. BILLIMORIA,
H. POZNER,
B. METSELAAR,
F. W. BEST, D. C. 0. JAMES
Serum turbidity was measured in a nephelometer (Evans Electroselenium, Middlesex) as described previously31. Lipoproteins. Paper electrophoresis was carried out by the method of Lees and Hatchs4 with minor modifications as reported by us previouslyz5. Cellulose acetate electrophoresis was carried out on Sepraphore III; the cellulose strips were stained with oil red 0, bleached with 1% w/v NaOCl soln. for 1 min and cleared by immersion in methanol containing 10% v/v glacial acetic acid. They were mounted on glass slides and the band dye concentration obtained by transmission scanning using the Gelscan Scanner (Gelman Instrument Co., Ann Arbor, Michigan 48106). The relative dye uptake (RDU) of each band was calculated as a percentage of the total dye uptake. The RDU/lOO x fatty acid mg/lOO ml of the particular serum gave the esterified fatty index (EFI) of each lipoprotein band25. A similar quantitation was used with paper electrophoresis, but reflectance scanning of the paper strip was used to determine the band positions and the bands were cut out, eluted with 5 % acetic acid and the eluates read at 540 rnp in an absorption photometer. The RDU was calculated and converted to EFI units as above. Cell counts. Red cell and total white cell counts, haemoglobin (Hb), haematocrit (Hc), mean corpuscular volume (MCV), mean corpuscular haemoglobin (MCH) and mean corpuscular haemoglobin concentration (MCHC) were determined by automated techniques using a Coulter counter, Model S. Blood films for differential counts were stained with May-Griinwald/Giemsa using a staining machine. The differential count was only worked out if the film appeared to be normal. Blood sugar was estimated on an Autoanalyser using a ferricyanide reduction method. Euglobulin lysis times were measured by a modification of the method of Buckel12s. Thrombin (50 ,ug/ml) was used for clotting instead of CaCls and a phosphate buffer (pH 7.3) was used instead of the borate buffer (pH 9.0). The water used was glass distilled. RESULTS
Cholesterol
Fig. 1 shows the levels of total serum cholesterol, in 35 male and female heavy smokers compared with 49 male and female non-smokers. The overall mean cholesterol value was greater for heavy smokers 236 f 8 mg/lOO ml than for non-smokers (mean value) 215 f 6 mg/lOO ml but was not significantly different. When males and females were analysed separately, significant differences were found. As shown in Fig. 1 the mean value for female heavy smokers is almost the same as for female nonsmokers being 221 mg/lOO ml and 215 mg/lOO ml respectively. Comparison of the 23 male heavy smokers (243 f 11 mg/lOO ml) with the 27 male non-smokers (215 & 6), showed differences which were significant (P < 0.025). The results of mean cholesterol values for male light smokers (231 f 16) did not differ significantly from the non-smokers (215 f 6) but showed a distinct trend which
65
SMOKING, LlPIDS AND FIBRINOLYSIS .m Heavy Smokers
~mg/lOOml) Non-Smokers
*
40Q
400_
100 (mean ,,,,,,...,..,.,,,) Female o (mean mean -
100. Male . OVerall
0 - _ _)
Fig. 1. Serum cholesterol (mg/lOO ml) of heavy smokers and non-smokers. Mean age: males, nonsmokers(N.S.)44.6& 7.7;heavysmokers(H.S.)44.7 & 7.2;females(N.S.)43.5 +9.3;(H.S.)44.8* 8.4. For standard deviation and P-values see text.
suggested that the level of serum cholesterol in men increased in relation to the number of cigarettes smoked. Serum free and ester cholesterol were also measured in the above volunteers. No differences in free cholesterol levels were found and ester cholesterol values were raised only in the male heavy smokers: 205 f 11 mg/lOO ml compared with 171 f 8 in the corresponding non-smokers (P < 0.01). Triglycerides
Fig. 2 shows the results of the long term effect of smoking on serum triglycerides in 28 male and female heavy smokers compared with 49 male and female non-smokers. TRlGLYCERlDES (mg/lOOml) Non-Smokers
Heavy
Smokers
280.
.
.
.
0 2OQ
0
2T 200.
.
0
0
.
0 _ - _ -. _ - - -.l0Qo
“. o
ca 0
a
.
.
- - . *.
:
0 .
Female o (mean Overall mean __
*.
0
* *
- - -)
.
. .
1oao . 0 ,., *t .._*.,,.....................~.... ,, .o,,.,,,.,,.. +9_Q_.o.._._,p._ 0.01, and in male heavy smokers mean triglycerides (134 f 15) were also raised compared with male nonsmokers (82 f 7), P > 0.0025. Further inspection of Fig. 2 showed that values above 180 mg/lOO ml serum triglycerides occurred only in heavy smokers whereas values below 50 mg/lOO ml were found only in non-smokers. Our upper normal limit (mean f SD.) in this laboratory without taking into account the smoking habit was 175 mg/lOO ml. The mean triglyceride level of the 16 light smokers (males and females) was 93 f 8 mg/lOO ml and that of the corresponding non-smokers 78 f 5 mg/lOO ml: this difference was not statistically significant, but again indicated that the rise in triglyceride depended on the number of cigarettes smoked. Effect of smoking on turbidity Fig. 3 shows differences found in plasma turbidity between male and female heavy smokers and non-smokers. The 36 heavy smokers were found to have a higher mean value, 26 * 2 nephelometric units (N.U.) compared with 20 f 1 N.U. for the 48 non-smokers (P < 0.025). Nearly all the higher values were attributable to male heavy smokers (mean 30 f 3.0); the non-smokers had a mean of 21 f 1.2 (P < 0.0025). Light smokers did not show any differences from the non-smokers even though the males and females were separately analysed. The mean turbidity value for heavy-smoking females did not differ from that of non-smokers.
Fig. 3. Serum turbidity (nephelometric units) of heavy smokers and non-smokers. Mean age: males, non-smokers (N.S.) 43.7 i: 6.9; heavy smokers (H.S.) 44.7 A 7.2; females (N.S.) 44.0 f 9.4; (H.S.) 45.2 + 8.2 For standard deviation and P-values see text.
SMOKING, LIPIDS AND FIBRINOLYSIS
67
EJSect of smoking on total phospholipids
Male heavy smokers had higher total plasma phospholipid levels of 241 j, 12 mg/lOO ml compared with 215 f 9.5 mg/lOO ml for the corresponding non-smokers (P < 0.05). This difference was not exhibited by females and there was no statistical difference observed between smokers and non-smokers when male and female results were combined. E#ect of smoking on lipoproteins (using paper electrophoresis) (a) fi-Lipoprotein
Fig. 4 shows the result of smoking on the esterified fatty acid index (EFI) of @lipoproteins in 38 male and female heavy smokers compared with 49 male and female non-smokers. The lipoproteins were separated by paper electrophoresis and quantitated as indicated above. The overall mean value for males and females is raised in the heavy smoking group (159 f 7.5 EFI units) compared with the non-smoking group (137 f 4.5 EFI units). As shown in Fig. 4, the mean value for female heavy smokers is almost equal to the mean value for female non-smokers, being 132 f 9 and 134 f 11 EFI units respectively. The mean EFI for ,%lipoprotein in male heavy smokers was 173 f 9 compared with 140 f 5 for non-smokers. This large difference was highly significant (P < 0.0025). These lipoprotein results obtained on division of data according to the smoking habit correlate well with the results of cholesterol measurements (shown above). As ,&lipoprotein contains most of the ester cholesterol, the results confirm the increased
BETA
(paper
electrophoresis)
Fig. 4. /?-Lipoproteins (EFI units, paper electrophoresis) of heavy smokers and non-smokers. Mean age: males, non-smokers (N.S.) 45.1 5 7.8; heavy smokers (H.S.) 44.1 + 6.7; females (N.S.) 44.0 f 9.2; (H.S.) 44.5 f 8.8. For standard deviation and P-values see text.
68
J. D. BILLIMORIA,
H. POZNER,
B. METSELAAR,
F. W. BEST, D. C. 0. JAMES
level of this lipid in male heavy smokers. Light smokers did not differ from nonsmokers in their cholesterol levels, although the mean cholesterol value in the male group lay between the mean values for non-smokers and heavy smokers. Exactly the same situation arises in the case of B-lipoprotein EFI values: the mean value for male light smokers is 155 & 23 EFI units which lies between the corresponding heavy smokers and non-smokers values of 173 & 9 and 140 f 5 respectively. (b) Pre-/l-lipoprotein
Fig. 5 shows the results of the long-term effect of smoking on EFI values for pre/I-lipoprotein, obtained by paper electrophoresis, for 33 male and female heavy smokers compared with 49 male and female non-smokers. The mean values of pre-/Ilipoprotein EFI of the heavy smokers were all raised, and approximately equally for males and females. The mean EFI for pre-j3-lipoprotein in the group of heavy smokers was 61 f 7 EFI units compared with non-smokers 25 & 3 EFI units (0.0005 < P > 0.0025). As shown in Fig. 5, in the non-smoking group, 13 cases (6 males, 7 females) or 27% did not show any pre-@-lipoprotein bands, whereas in the heavy smoking group only 2 cases (1 male, 1 female) out of 33 (i.e. 6 % of the cases) showed no pre-P band. Results of males and females, separately analysed with respect to smoking and pre-/?-lipoprotein levels, showed similar differences: male non-smokers and female non-smokers had mean pre-ig-lipoprotein values that were approximately the same, while male heavy smokers and female heavy smokers had equally raised means. It was found that the pre-/l-lipoprotein values for light smokers showed very significant differences when compared with non-smokers. Male light smokers as well as female light smokers showed the following raised mean pre-@-lipoprotein values: 56 f 8 and 58 & 6 respectively for light smokers compared with 28 f 4 and 25 f 3 for non-smokers (in each case P < 0.0025). EFI
( paper electrophoresis) Non-Smokers
Heavy Smokers 2OQ
20Q
0
.
.. *,,,,,,,,,,,,.,. ,**.l*.l.l.~. ..” ...***. *.*.** .“.I,,..,.,*.,.., ~._._~.L._.._.;._.-._._.; ___ . . . . O_ Female
0
( mean
---.---)
a. . .* 0 0 00 1)11 11#1*1 111*,*, O&,y*l*l.ll.llll CR I.I 1111k.1 r..‘ ,,##, :. * 1 l.#l*ll*l. 4**l.lll*ll.l*llll,, o-‘-‘-‘0 -’ fb.-‘o.- ._._.-.-.r.-‘-__‘_ O& 0 Male . ( mean ,,,,,,.,,..,,.,,)Overa 11mean_
Fig. 5. Pre-&lipoproteins (EFI units, paper electrophoresis) of heavy smokers and non-smokers. Mean age: males, non-smokers (N.S.) 45.1 & 7.8; heavy smokers (H.S.) 44.1 f 6.7; females (N.S.) 44.0 ;t 9.2; (H.S.) 44.5 & 8.8. For standard deviation and P-values see text.
SMOKING, LIPIDS AND FIBRINOLYSIS
69
Separation of lipoproteins by cellulose acetate electrophoresis
During winter months we picked 43 normal subjects randomly from the 30-60 year old group, to check on any seasonal variation in cholesterol and triglyceride values. Electrophoresis was carried out on the sera of these subjects using polycellulose acetate (Sepraphore III) as the separating and support medium. The subjects consisted of 20 heavy smokers (11 females and 9 males) and 23 non-smokers (9 females and 14 males). Samples were not taken from light smokers. (a)
B-Lipoprotein
Fig. 6 shows the results of @-lipoprotein EFI for the above volunteers and the differences found when they were divided into heavy smokers and non-smokers. The mean EFI value for male and female heavy smokers (191 f 8) was greater than that obtained for male and female non-smokers (168 f 6), P < 0.0125. The difference in means between heavy smokers and non-smokers was 22 EFI units using Sepraphore electrophoresis, which compared favourably with a difference of 23 units with paper electrophoresis. When the actual figures from the two electrophoretic procedures were compared it could be seen that EFI on paper were approximately 30 units less than those on cellulose acetate. When males and females were separately analysed the differences between smokers and non-smokers found by Sepraphore electrophoresis also corresponded to those found by the paper electrophoresis technique. /?-Lipoprotein EFI in 11 female heavy smokers compared with 9 non-smokers were not significantly different. In contrast, a large increase was found in the mean value for 9 males who were heavy smokers compared with 14 male non-smokers.
BETA-
(cellulose acetate electrophoresis)
Heavy Smokers 3T----l
Female
Non-Smokers 300{_
o ( mean-.-.-.) Male Overall mean -
. (mean ..............)
Fig. 6. B-Lipoproteins (EFI units, cellulose acetate electrophoresis) of heavy smokers and nonsmokers. Mean age: males, non-smokers (N.S.) 42.9 & 8.7; heavy smokers (H.S.) 47.8 i 6.8; females (N.S.) 42.9 =t 8.7; (H.S.) 47.1 f 8.0. For standard deviation and P values see text.
J. D. BILLIMORIA, H. POZNER, B. METSELAAR, F. W. BEST, D. C. 0. JAMES PRF-EFTA
Female
o
F F ~(celluloseacetateelectrophoresis)
( -.-.Overall
)
Male mean -
. ( mean ,,., ,,,..,..,,. )
Fig. 7. Pre-p-lipoproteins (EFI units, cellulose acetate electrophoresis) of heavy smokers and nonsmokers. Mean age: males, non-smokers (N.S.) 42.9 & 8.7; heavy smokers (H.S.) 47.8 & 6.8; females (N.S.) 50.7 ?t 0.1; (H.S.) 47.1 f 8.0. For standard deviation and P values see text.
(b) Pre-/I-lipoprotein
Fig. 7 shows the results of pre-p-lipoprotein EFI for 43 subjects and the differences between the mean values for heavy smokers and non-smokers. The overall mean value for male and female heavy smokers (91 EFI) is greater than that for male and female non-smokers (61 EFI, P < 0.001). These results (Fig. 7) compared with those of paper electrophoresis (Fig. 5) show that a difference of 30 EFI units was observed on cellulose acetate and 36 EFI units on paper electrophoresis between the pre-b-lipoproteins of smokers and nonsmokers. This slight discrepancy will be clear when Figs. 5 and 7 are compared. On paper, the separation of pre-/3 and ,~5is incomplete when only small amounts of pre-p are present, as the pre-p is retained in the p band. Thus it is seen in Fig. 5 that with those non-smokers where only traces of pre-p are present, these are not detected and are recorded as zero. On cellulose acetate (Fig. 7) even the traces are detected and have a finite value. The small loss of pre-B from the non-smokers on paper electrophoresis accounts for the greater difference of pre-Slipoproteins observed between smokers and non-smokers on this media. Euglobulin lysis time and Stypven clotting time
These tests are shown in Fig. 8. Euglobulin lysis times were significantly longer for both male (286 f 21) and female (315 f 20) groups of heavy smokers compared with non-smokers (respectively 236 f 18 and 193 f 17; P < 0.05 for men and P < 0.0005 for women). The difference was more marked in the female group where even light smokers (240 f 30) showed a marked and significant decrease in fibrinolytic activity. Stypven clotting times (SCT) of heavy smokers were significantly shortened in
71
SMOKING, LIPIDS AND FIBRINOLYSIS NON-S
M OKERS
HEAVY
SMOKERS
Fig. 8. Euglobulin lysis times (min) and Stypven clotting times (set) of heavy smokers and nonsmokers. Mean age: males, non-smokers (N.S.) 44.3 + 7.9; heavy smokers (H.S.) 45.2 j, 7.3; females (N.S.) 43.0 & 9.3; (H.S.) 43.7 f 7.6. For standard deviation and P values see text.
both males and females from 39.5 f 1.3 and 41 f 1.4 to 33 f 0.9 and 33 f 1.6 respectively (P < 0.0005 for both sexes). Although the SCT of both sexes in the light smokers was also shorter than that of non-smokers, the differences were not significant. Haematocrit and haemoglobin levels
These are shown in Fig. 9. Male heavy smokers gave an haematocrit of 47.4% compared with 44.9 % for non-smokers (P < 0.0025). No differences were observed in the female groups and changes in male light smokers were not significant. HEAVY
NON-SMOKERS
SMOKERS
haematocrit(“,,)
I
I I
I
Fig. 9. Haematocrit (%) and haemoglobin (g/ml) of heavy smokers and non-smokers. Mean age: males, non-smokers (N.S.) 44.3 + 7.9; heavy smokers (H.S.) 45.2 i 7.3; females (N.S.) 43.0 :t 9.3; (H.S.) 43.7 i 7.6. For standard deviation and P values see text.
72
J. D. BILLIMORIA, H. POZNER, B. METSELAAR, F. W. BEST, D. C. 0. JAMES
The haemoglobin of male heavy smokers 15.4 g/ml was significantly higher than non-smokers 14.6 g/ml (P < 0.0025). No significant changes were found in the groups of females. Mean corpuscular volumes (MCV)
and haemoglobin (MCH)
A small increase in MCV was found in male heavy smokers (P < 0.05). No significant changes were found in the women. The MCH of male heavy smokers (29.7 pg) was slightly increased and significantly greater than for non-smokers (28.6 pg; P < 0.005). No significant changes were observed in women. The mean corpuscular haemoglobin concentration (MCHC) showed no changes accountable to smoking. White cell ( WBC) and diflerential counts
The results are shown in Fig. 10. The WBC in male heavy smokers was significantly increased (8.5 & 0.5) x 10s compared with (5.5 f 0.2) x 103 in non-smokers (P < 0.005). There was a small increase in WBC in the female heavy smokers but it was not significant (0.05 < P < 0.10). The neutrophils showed a significant increase in heavy smokers (H.S.) of both sexes compared with non-smokers (N.S.), viz. male H.S. 5.0 x 103, male N.S. 3.1 x 10s (P < 0.0005); female H.S. 4.5 x 103, female N.S. 3.5 x 10s (P < 0.05). A similar
*.*. L-H
0.
6.
.
*
.
.
*
4.
*
*
2r_”
. ..*
.*:...
..
.*
. .
* **t * ** I * ** * 1
*
.
.
*
*.: **.
.
*
. ** **** ****** *
*
*
*
I
I
lymphocytes
(x103)
II
Fig. 10. Total white cell counts, neutrophils and lymphocyte counts of heavy smokers and nonsmokers. Mean age: males, non-smokers (N.S.) 44.3 k 7.9; heavy smokers (H.S.) 45.2 & 7.3; females (N.S.) 43.0 % 9.3; (H.S.) 43.7 f 7.6. For standard deviation and P values see text.
SMOKING,
LIPIDS AND FIBRINOLYSIS
73
trend was seen in light smokers but the difference was not statistically significant. The results for lymphocyte counts for heavy (H.S.) and light (L.S.) male smokers were higher than those for non-smokers (N.S.). Thus for men H.S. = 3.0 x 103, N.S. = 2.2 x 10s (P < 0.025), L.S. = 2.6 x 10s (0.10 > P > 0.05). Changes were not apparent in the female groups. Eosinophils: Male H.S. gave values of 264 x lo3 compared with N.S. 143 x lo3 (P < 0.05) and L.S. 189 x 10s (not significant). Female H.S. gave values of 144 x lo3 compared with N.S. 78 x 10s (P -C 0.05) and L.S. 152 x 10s (P < 0.05). Busophils: Whereas basophils were rarely noted in films from non-smokers they were always present in blood films from smokers, especially for the male heavy smokers. No changes due to smoking were observed in blood sugar levels, blood red cell counts, fibrinogen levels or platelet adhesiveness properties.
DISCUSSION
A significant difference was shown in the Stypven coagulation test (SCT), where a shortening of plasma clotting time was demonstrated in heavy smokers even after they had abstained from smoking for 12 h. Previously only the acute changes in coagulation have been studied immediately after smoking I or 2 cigarettes; such changes were probably a result of increased noradrenalin. As light smokers also show a shortened SCT - although to a much lesser extent - our observation is likely to relate to hypercoagulability due to a chronic effect of cigarette smoking. Increase in serum total cholesterol levels were confined to male heavy smokers : an increased index of p-lipoproteins and the fact that the increase was found in the ester cholesterol fraction is additional confirmation of the above finding. However, we have observed a substantial seasonal variation in cholesterol levels and, as this variation is as large as the difference between smokers and non-smokers, it is important - as here - to test the two groups during the same season. Significantly increased levels of triglyceride have been found in heavy smokers of both sexes and this triglyceride is contained in the pre-/?-lipoprotein fraction as shown by the increased EFI index of this lipoprotein. Increased pre-/3-lipoproteins have have shown been reported by Besterman35, while both Albrinkss and Carlson that triglycerides (TG) are elevated in ischaemic heart-disease (IHD). We have also observed during routine tests carried out on patients over the last 5 years that a majority of IHD patients show an increased TG level and a raised pre-8 EFI index. In fact a larger number of IHD patients have triglyceride levels in excess of 175 mg/lOO ml than cholesterol levels over 300 mg/lOO ml. (The above levels were the upper limits (mean f 2 S.D.) of triglycerides and cholesterol observed for normal subjects aged 40-70 years in our laboratory.)
74
J. D. BILLIMORIA, H. POZNER, B. METSELAAR, F. W. BEST, D. C. 0. JAMES
In our experience nephelometry of fasting visible lipaemia in patients with IHD invariably shows a small fasting serum turbidity of 30-40 units, compared with our range of lo-26 units in normal subjects (see refs.38939).The present results show a very similar small rise in turbidity in male heavy smokers. Fibrinolytic activity tested by both whole blood lysis and euglobulin lysis tests is reduced in heavy smokers; a similar decrease has also been observed in patients with IHD. The raised haematocrit value found in the male heavy smokers is not accompanied by a corresponding rise in red cell count. This is partly explained by a corresponding rise in the mean cell volume of this group, with a possible contribution from a reduction of plasma volume caused by the vasoconstrictive and diuretic effects of nicotine. The raised carboxyhaemoglobin content of the blood following chronic tobacco smoking would result in a degree of anoxia which, in turn, would stimulate erythropoesis and thus at least partly explain the raised Hb level. The increased total white count is reflected in nearly all the white cell groups and is most evident in the heavy smokers. These observations agree with those of Corre et ~1.40.Although no explanation is apparent, it was noted that basophils were invariably present in the blood films of smokers. It has been suggested that the living style of smokers may differ from that of nonsmokers, and that the increased cholesterol level in heavy smokers may be genetically associated with but not directly related to the cause of the urge to smoke. “Styles of living” are difficult to assess in experimental studies and, taking into account the numerous changes other than in the cholesterol level observed in both male and female smoker&, we have come to the operational conclusion that heavy smokers are a “high risk” group for coronary disease. It is doubtful that much is gained in this instance by consideration of nebulous cause-and-effect relationships. In most laboratories, including our own, the upper normal limits for serum lipids have been recorded without reference to the smoking habit; it now appears that this factor would have to be taken into account in determining a state of hyperlipidaemia. This step would undoubtedly cause the upper limit to be reduced in normal subjects who are non-smokers. ACKNOWLEDGEMENTS
We thank Mr. K. L. E. Ling for skilled technical assistance. We are also grateful to all volunteers from I.C.I. who participated in this work; to Mrs. Joyce Porter (Med. Dept., I.C.I.) who supervised the arrangements: to Mrs. D. V. Winter and Dr. J. Thurston (Cardiac Dept., Westminster Hospital) for respectively preparing and interpreting the ECGs. REFERENCES 1 HAMMOND, E. G. AND HORN, D., The relationship between human smoking habits and death rates (a follow-up study of 187,783 men), J. Amer. Med. Ass., 155 (1954) 1316.
SMOKING, LIPIDS AND FIBRINOLYSIS
75
2 DOYLE, J. T., DAWBER, T. R., KANNEL, W. B., KINCH, S. H. AND KAHN, M. A., The relationship of cigarette smoking to coronary heart disease, J. Amer. Med. Ass., 190 (1964) 886. 3 DOYLE, J. T., DAWBER, T. R., KANNEL, W. B., HESLIN,A. S. AND KAHN, M. A., Cigarette smoking and coronary heart disease, New Engl. J. Med., 266 (1962) 796. 4 PAUL, O., LEPPER,M. H., PHELAN, W. H., DEPERTUIS,G. W., MACMILLAN, A., MCKEAN, H. AND PARK, H., A longitudinal study of coronary heart disease, Circulation, 28 (1963) 20. 5 BORHANI,N. O., HECHTER,H. H. AND BRESLAW,L., Report of a ten year follow-up study of the San Francisco longshoremen. Mortality from coronary heart disease and from all causes, J. Chron. Dis., 16 (1963) 1251. 6 HAMMOND,E. C., Smoking in relation to mortality and morbidity. Findings in first thirty-four months of follow-up in a prospective study started in 1959, Cancer Inst., 32 (1964) 1161. 7 SHAPIRO,S., WEINBLATT,E., FRANK, C. W. AND SAGER, R. V., The H.I.P. study of incidence and prognosis of coronary heart disease: preliminary findings on incidence of myocardial infarction and angina, J. Chron. Dis., 18 (1965) 527. 8 KARVONEN,M. J., ORMA, E., KEYS, A., FIDANZA, F. AND BROZEK, J. Cigarette smoking, serum cholesterol, blood pressure and body fatness, Lancer, 1 (1959) 492. 9 THOMAS, C. B., Familial and epidemiological aspects of coronary disease and hypertension, J. Chron. Dis., 7 (1958) 198. 10 THOMAS, C. B., Familial patterns in hypertension and coronary heart disease, Circulation, 20
(1959) 25. 11 GOFMAN, J. W., LINDREN, F. T., STRISOWER,B., DE LALLA, O., GLAZIER, F. AND TAMPLIN, A., Cigarette smoking, serum lipoproteins and coronary disease, Geriatrics, 10 (1955) 349. 12 BRONTE-STEWART,B., Cigarette smoking and ischaemic heart disease, Brit. Med. J., 1 (1969) 379. 13 CAGANOVA, A., CAGAN, S. AND SIMKO, V., Bratisl. Lek. Listy, 50 (3) (1968) 387. 14 KERSHBAUM,A., BILLET,S., DICKSTEIN,I. R. AND FEINBERG,L. J., Effect of cigarette smoking and nicotine on serum free fatty acids. Based on a study in the human subject and the experimental animal, Circ. Res., 9 (1961) 631. 15 KERSHBAUM,A. AND BILLET, S., Cigarette smoking and blood lipids, J. Amer. Med. Ass., 187 (1964) 32. 16 KONTTINEN,A. AND RAJALSALMI,M., Effect of heavy cigarette smoking on postprandial triglycerides, free fatty acids and cholesterol, Brit. Med. J., 1 (1963) 850. 17 SHORT, J. J. AND JOHNSON,H. J., A direct comparison of the reaction of the human system to tobacco smoke and adrenalin, J. Lab. Clin. Med., 24 (1939) 590. 18 PAGE, I. H., LEWIS, L. A. AND MOINUDOLIN,M., Effect of cigarette smoking on serum cholesterol and lipoprotein concentrations, J. Amer. Med. Ass., 171 (1959) 1500. 19 BLACKBURN,H., ORMA, E., HASTEL,G. AND PUNSAR,S., Tobacco smoking and blood coagulation. Acute effect on plasma Stypven time, Amer. J. Med. Sci., 238 (1959) 448. 20 AMBRUS, J. L. AND MINK, I. B., The effect of cigarette smoking on blood coagulation, Clin.
Pharmacol. Therap., 5 (1964) 428. 21 ROSENBERG,A., Fall in carbon monoxide blood levels after stopping smoking, Lancet, 1 (1972) 593. 22 BILLIMORIA,J. D. AND JAMES,D. C., Two improved methods for serum cholesterol estimation,
Clin. Chim. Acra, 5 (1960) 644. 23 SPERRY, W. M. AND WEBB, M., Revision of Schoenheimer-Sperry mination, J. Biof. Chem., 187 (1950) 97.
method for cholesterol deter-
24 VAN HANDEL, E. AND ZILVERSMIT,D. B., Micromethod for the determination of serum triglycerides, J. Lab. Clin. Med., 50 (1957) 152. 25 BILLIMORIA,J. D., FAHMY, M. F., JEPSON,E. M. AND MACLAGAN, N. F., The use of the esterified fatty acid index in the classification and quantitation of hyperlipoproteinaemia, Atherosclerosis, 14 (1971) 359. 26 RAPPORT, M. M. AND ALONZO, N., Photometric determination of fatty acid groups in phospholipids, J. Biol. Chem., 217 (1955) 193. 27 INGRAM,C. I., The determination of plasma fibrinogen by the clot weight method, Biochem. J., 51 (1952) 583. 28 BUCKELL, M., The effect of citrate on euglobulin methods of estimating fibrinolytic activity, J. Clin. Pathol., 11 (1950) 403. 29 BRICKER, C. AND CRONKITE,E. P., Morphology and enumeration of human blood platelets, J. Appl. Physiol., 3 (1950) 365.
76
J. D.
BILLIMORIA, H. POZNER, B. METSELAAR, F. W. BEST, D. C. 0. JAMES
30 HELLEM,A. J., The Adhesiveness of Human Blood Platelets in vitro (Norwegian Monographs on Medical Science), Oslo University Press, Oslo, 1960. 31 MACLAGEN,N. F., BILLIMORIA,J. D. AND CURTIS,C., Lipaemia and blood coagulation with special reference to the stypven technique, Lancer, 2 (1958) 865. 32 FEARNLEY,G. R., BALMFORTH,G. V., AND FEARNLEY,E., Evidence of a diurnal fibrinolytic rhythm, with a simple method of measuring natural fibrinolysis, Clin. Sci., 16 (1957) 645. 33 LACKNER,H. AND GOOSEN, C. C., Fibrinolytic activity of blood: photographic determination of clot lysis time, Acta Haematol., 22 (1959) 58. 34 LEES, R. S. AND HATCH, F. T., Sharper separation of lipoproteins by paper electrophoresis in albumin containing buffer, J. Lab. Clin. Med., 61 (1963) 518. 35 BESTERMAN,E. M., Lipoproteins in coronary artery disease, Brit. Heart J., 19 (1957) 503. 36 ALBRINK,M. J. AND MAN, E. B., Serum triglycerides incoronary artery disease, Arch. Intern. Med.,
103 (1959) 4. 37 CARLSON, L. A.
AND BOTTIGEN,L. E., Stockholm prospective study. lschaemic heart disease in relation to fasting values of plasma triglycerides and cholesterol, Lancer, 1 (1972) 865. 38 MITCHELL, J. R. A. AND BRONTE-STEWART,B., Alimentary lipaemia and heparin clearing in ischaemic heart-disease, Lancer, 1 (1959) 167. 39 JAMES,D. C., DRYSDALE,J., BILLIMORIA,J. D., WHEATLEY,D., GAVEY, C. J. AND MACLAGAN, N. F., Lipaemia and blood-coagulation defects in relation to ischaemic heart-disease, Lancer, 2
(1961) 798. 40 CORRE, F., LELLOUCH,J. AND SCHWARTZ, D., Smoking and leucocyte-counts. miological survey, Lancer, 1 (1971) 632.
Results of an epide-
41 POZNER,H. AND BILLIMORIA,J. D., Effect of smoking on blood clotting and lipid and lipoprotein levels, Lancer, 1 (1970) 1318.
