Clinical Science (1983)65,669-612
669
SHORT COMMUNICATION
Effects of smoking on oral fat tolerance and high density
lipoprotein cholesterol
R.
s. ELKELES, s. R.
KHAN,
v. C H O W D H U R Y A N D
M. B. SWALLOW
Departments of Medicine and Human Metabolism, St Mary’s Hospital, R a e d Street, London
(Received 19 May 1983; accepted 18 JuZy 1983)
Summary
1. Changes in serum triglyceride and high
density lipoprotein (HDL) cholesterol after a fatty
meal have been studied in smokers and nonsmokers.
2. Average serum triglyceride during the study
was higher in smokers than in non-smokers.
3 . In non-smokers there was a rise in the H D L /
H D L cholesterol ratio after oral fat, but not in
smokers.
4. These frndings are compatible with the hypothesis that smoking interferes with the lipolysis of
triglyceride rich lipoproteins and the conversion
of H D b into HDL.
Key words: cholesterol, HDL, H D L , H D L ,
lipoprotein, serum triglyceride, smoking.
Abbreviation: HDL, high density lipoprotein.
Introduction
Smoking is a well established risk factor for both
coronary heart disease and peripheral vascular
disease, as is a low level of high density lipoprotein
(HDL). Smoking is associated with low circulating
levels of HDL [11. It has been proposed that one of
the mechanisms whereby smoking could accelerate
atherosclerosis is by interfering with the metabolism
of chylomicrons and very low density lipoproteins
(VLDL) [2]. HDL comprises at least two subfractions, H D L and H D L . During lipolysis, by
the enzyme lipoprotein lipase, of VLDL in vifro,
there is conversion of HDL into HDI+ [3].
Correspondence: Dr R. S. Elkeles, St Mary’s
Hospital, Praed Street, London W2 1NY.
Similar changes have been observed in vivo in man
after a normal meal [4,5]. We have tried to test
the hypothesis that smoking interferes with
intravascular lipolysis in vivo by studying the
changes occurring in serum triglyceride, HDL and
its subfractions after a fatty meal in healthy
smokers and non-smokers.
Subjects and methods
We studied 14 healthy male medical students, all
less than 100% of ideal body weight. Their mean
age was 20 (range 18-24) years. Seven were nonsmokers and seven were regular smokers, mean 15
cigarettes per day (range 8-30). The mean percentage ideal body weight of the non-smokers was 95%
(range 87-100) and that of the smokers was 100%
(range 92-log), and the difference between the
two was not significant. Of the non-smokers, two
played football, one rugby, regularly, and four did
not participate in sport. Of the smokers, one
played volley ball, one badminton, one water-polo,
regularly, and four did not participate in sport.
Subjects in both groups consumed similar
amounts of alcohol.
After an overnight fast, an indwelling cannula
was inserted into an antecubital vein. After a basal
blood sample had been taken subjects ate a breakfast consisting of 25 g of cornflakes with 2.25 ml
of double cream per kilogram body weight. In
addition, they ate two eggs fried in butter and two
slices of brown toast with 40 g of butter and one
cup of black coffee. A further meal was given 3 h
later consisting only of a weighed amount of
lettuce, cucumber, tomato, beetroot, green pepper
and an orange. Water was allowed freely. Smokers
were also given one cigarette per hour to smoke
during the test. After the initial blood sample
. Serum cholesterol, triglyceride. total HDL cholesterol. HDLa HDL3 cholesterol and HDLzIHDLp cholesterol ratio in smokers (S) and non-smokers (NS)b
fatty meal
els (0h)
Average results f SE of differences, on log scale, are shown.Values in mmol/l for each time are given in parentheses.
Total cholesterol
S
NS
0.694
(4.94)
0.643
(4.40)
Diff.
Total triglyceride
S
NS
Diff.
0.022 -0.153
(0.70)
(1.05)
Total HDL cholesterol
S
NS
0.138
(1.37)
0.208
( I .62)
Diff.
HDL, cholesterol
S
NS
-0.336
(0.46)
-0.291
Diff.
(0.5 1 )
* P = 0.04; **P = 0.02.
HDL, cholesterol
HDL
S
NS
-0.050
(0.89)
0.002
(1.00)
-0.286
(0.52)
-0.026
(0.49)
-0.029
(0.48)
-0.054
(0.46)
(0.46)
-0.067
(0.44)
-0.063
(0.45)
-0.327
(0.47)
Diff.
S
ver 0 h at
vel
tO.027
(5.25)
+ 0.022
(5.20)
+ 0.006
(5.01)
t0.003
(4.98)
to.009
(5.05)
t0.021
(5.19)
0.706
(5.09)
tO.010 t0.017
(4.50) i 0.015
tO.008 +0.014
(4.48) * 0.020
to.011 -0.005
(4.51) i 0.017
t0.012 -0.009
(4.52) f 0.913
t0.017 -0.008
(4.57)
t 0.014
+0.014 t0.007
(4.54) t 0.010
0.653
(4.50)
+0.053
0.056
f
t0.447
(2.94)
t0.476
(3.14)
tO.309
(2.14)
t0.136
( 1.44)
t0.292
( I .38)
t0.337
(1.53)
to.155
0.082
+0.139
t 0.091
t 0.069
t 0.082
t
tO.058
(1.20)
-0.099
(0.84)
tO.240
(1.22)
t0.131 + 0.005
(0.95) i 0.087
t 0.043 to.015
(0.78) f 0.083
t 0.095 -0.004
(0.56) t 0.077
0.2 I 1
.(1.63)
-0.018
(0.96)
t0.229.
1.100
t
0.000 -0.011
(1.58)
-0.010 -0.015
(1.34)
(1.56)
t0.004
-0.017
(1.63)
(1.32)
t0.034
t 0.02 I
(1.75)
(1.44)
+0.026 t0.036
( I ,461 (1.76)
+0.038 t0.038
(1.76)
(1.50)
to.011
0.017
to.005
i 0.014
-0.021
2 0.019
-0.013
t 0.026
-0.010
t 0.012
0.000
i 0.014
t0.012
-0.019
(0.53)
(0.44)
t 0.03 1 t0.057
(0.58)
(0.43)
-0.036 t0.087
(0.63)
(0.42)
-0.025 -0.136
(0.70)
(0.44)
t0.102
-0.014
(0.65)
(0.45)
+0.085
-0.004
(0.46)
(0.62)
-0.089
f 0.0.55
t0.028
+0.007 -0.021
(0.96) t 0.018
(0.90)
-0.001 -0.042
tO.041
(0.91) t 0.024
(0.89)
tO.018 -0.028 -0.046
(0.94) f 0.035
(0.93)
tO.054
t0.026 -0.031
(0.94) t 0.032
(0.94)
tO.015 +0.038
-0.053
(1.00)
(1.04) f 0.036
tO.059
0.000 to.059
(1.00) t 0.036
(1.02)
0.221
(1.66)
-0.075
t 0.040
-0.355
(0.44)
-0.133
f 0.090
-0.027
(0.94)
(1 37)
0.146
( 1.40)
t
-0.222
(0.60)
-0.031
i 0.060
-0.088
t 0.061
-0.123
f 0.067
--0.161**
0.059
-0.116
t
f
0.081
-0.013
(0.97)
-0.014
f 0.034
-0.05 1
Smoking and high density lipoprotein cholesterol
further venous blood samples were taken at 2, 4,
6,7,8 and 9 h. HDL cholesterol was measured by a
precipitation method [6]. HDL, and HDLBcholesterol was measured by a further precipitation
procedure [7], verified in our laboratory [8].
Triglyceride and cholesterol were measured by
Auto-analyzer enzymatic methods. The coefficient
of variation for total HDL cholesterol was 2% and
for HDL, and H D L cholesterol was 3%respectively.
For statistical analysis the data were logarithmically transformed. Changes occurring at each
time from the initial value were compared in the
two groups by the unpaired t-test.
Results
There were no significant differences in initial
values between smokers and non-smokers. Mean
serum triglyceride between 0 and 9 h was greater
in smokers than in non-smokers (P= 0.04) (Table
1). There was a tendency for serum triglyceride to
rise higher and remain elevated for longer in the
smokers than in non-smokers.
There was a rise in H D L cholesterol in nonsmokers but not in smokers and this difference
was significant between the groups at 7 h (P=
0.02).
There were no significant differences in changes
in HDLB cholesterol between the two groups. The
HDL,/HDL cholesterol ratio rose in non-smokers,
but not in smokers, and this difference was significant at 7 h (P= 0.02) (Table 1).
Discussion
We have shown, we believe for the first time,
differences in the handling of oral fat between
smokers and non-smokers. It is unlikely that the
differences observed were due to differences in
gastric emptying between the groups. No consistent
changes of smoking on gastric emptying have been
observed [9].
However, information on this topic has been
conflicting [9], so that an effect of smoking on
gastric emptying cannot be completely ruled out
as a factor in our results. The reciprocal changes in
HDL, and H D L cholesterol occurring after a meal
are thought to be physiological and reflect the
intravascular lipolysis of VLDL [3-51. During this
process there is conversion of H D b into HDL
[3]. The lack of increase in the HDL.JHDL3
cholesterol ratio in smokers suggests that this
process fails to occur to the same extent in smokers
as in non-smokers. It is therefore possible that
smoking interferes with the normal production of
HDL,. The two groups studied were similar in all
67 1
respects except their smoking habits, so that it
seems reasonable to ascribe the differences found
to smoking. Of the two HDL subfractions, HDL,
has been shown to be the more important in
protecting against coronary heart disease [ 10,111.
Furthermore, impairment of the clearance of
chylomicrons and VLDL could lead to the accumulation of remnant lipoproteins, which may also
contribute to atherogenesis [12], From our data
we are not able to say whether the effects described
are due to the acute or chronic exposure to
cigarette smoke. We are also not able to say which
component of cigarette smoking is responsible for
the changes. These points require further study.
However, our findings do provide one possible
mechanism for the increased atherogenesis associated with smoking.
Acknowledgments
S.R.K. was supported by the Special Trustees of
St Mary’s Hospital. We thank Professor Victor
Wynn for provision of laboratory facilities. We also
thank the volunteers who participated in the study,
and staff of the Metabolic Ward. We are greatly
indebted to Professor M. J. R. Healy, London
School of Hygiene and Tropical Medicine, for
carrying out the statistical analysis.
References
1. Williams, P., Robinson, D. & Bailey, A. (1979) High
density lipoprotein and coronary risk factors in normal
men. Lancet, i, 72-75.
2.Topping, D.L., Dwyer, T. & Weller, R.A. (1977)
Peripheral vascular disease in cigarette smokers and
impaired hepatic metabolism of lipoprotein remnants.
Luncet, ii, 1327-1328.
3. Patsch, J.R., Gotto, A.M., Olivecrona, T. & Eisenberg,
S. (1978) Formation of high density lipoprotein-like
particles during lipolysis of very low density lipoproteins in vitro. Proceedings of the National Academy of
Sciences U.S.A., 75,4519-4523.
4. Mattock, M., Salter, A., Omer, T. & Keen, H. (1979)
Changes in high density lipoprotein subfraction
distribution during alimentary lipaemia. Clinical
Science, 58, 16P.
5 . Baggio, G., Fellin, R., Baiocchi, M.R., Martini, S.,
Baldo, G., Manzato, E. & Crepaldi,G. (1980) Relationship between triglyceride-rich lipoprotein (chylomicrons and VLDL) and HDL, and HDL, in the
post-prandial phase in humans. Atherosclerosis, 37,
271-276.
6. Burstein, M., Scholnick, H.R. & Morfin, R. (1970)
Rapid method for isolation of lipoproteins from
human serum by precipitation with polyanions.
Journal of Lipid Research, 11,583-595.
7. Gidez, L.I., Miller, G.J.,Burstein, M. & Eder, H.A.
(1979) Analysis of plasma high density lipoprotein
subclasses by a precipitation procedure - correlation
with preparative and analytical ultracentrifugation.
672
R. S. Elkeles et al.
In: Report of the High Density Lipoprotein Methodology Workshop, pp. 328-342. Ed. Lippel, K.DHEW
NIH Publication no. 79-1661.NIH, Bethesda.
8. Khan, S.R., Elkeles, R.S., Niththyananthan, R., Seed,
M. & Wynn, V. (1982) Analysis of high density lipoprotein subclasses by a precipitation procedure:
comparison with preparative ultracentrifuge. Clinica
Chimica Acta, 120,4945.
9.Bennett, J.R. (1972)Smoking and the gastro-intestinal
tract. Gut, 13,658-665.
lO.Gofman, J.W., Young, W. & Tandy, R. (1966)
Ischaemic heart disease, atherosclerosis and longevity.
Circulation, 34, 679-697.
11.Miller, N.E., Hammett, F., Saltissi, S., Rao, S., Zeller,
H.V., Coltart, J. & Lewis, B. (1981) Relation of
angiographically defied coronary artery disease to
plasma lipoprotein subfractions and apolipoproteins.
British Medical Journal, ii, 1741-1744.
12.Zilversmit, D.B. (1979) Atherogenesis - a postprandial phenomenon. Circulation, 60,473-485.