1293
Changes in Lipoprotein Subfractions During
Diet-Induced and Exercise-Induced Weight
Loss in Moderately Overweight Men
Paul T. Williams, PhD, Ronald M. Krauss, MD,
Karen M. Vranizan, MA, and Peter D.S. Wood, DSc
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
We studied separately the effects of weight loss by calorie restriction (dieting) and by calorie
expenditure (primarily, running) on lipoprotein subfraction concentrations in sedentary,
moderately overweight men assigned at random into three groups as follows: exercise without
calorie restriction (n=46), calorie restriction without exercise (n=42), and control (n=42).
Plasma lipoprotein mass concentrations were measured by analytic ultracentrifugation for
lotation rates (F,'20, S') within high density lipoprotein (HDL) (F,'20 0-9), low densit
lipoprotein (LDL) (Sf 0-12), intermediate density lipoprotein (IDL) (S? 12-20), and very low
density lipoprotein (VLDL) (Sf 20-400) particle distributions. Particle diameter and flotation
rate of the most abundant LDL species were determined by nondenaturing polyacrylamide
gradient gel electrophoresis and analytic ultracentrifugation, respectively. During the 1-year
trial, the exercisers ran (mean±+SD) 15.6±9.1 km/wk, and the dieters ate 340±71 fewer kilocalories per day than at baseline. Total body weight was reduced significantly more in dieters
(-7.2±4.1 kg) and exercisers (-4.0±3.9 kg) than controls (0.6+3.7 kg). As compared with
mean changes in controls, the exercisers and dieters significantly increased HDL2 mass (48.6%
and 47.1%, respectively), decreased VLDL mass (-23.9% and -25.5%), and increased LDL
peak particle diameter (2.4 and 3.2 A). When adjusted to an equivalent change in body mass
index by analysis of covariance, 1) exercise-induced and diet-induced weight loss produced
comparable mean changes in the mass of small LDL and VLDL, and in LDL peak particle
diameter; 2) the exercisers versus control group difference in HDL2 was attributed to the
exercisers' reduced body mass index; and 3) HDL2 increased significantly less in dieters than
in exercisers. In dieters, low calorie intake might mitigate the effects of weight loss on HDL2.
(Circulation 1990;81:1293-1304)
who are at low risk of coronary heart
disease have low serum mass concentrav
tions of smaller, less-buoyant low density
lipoproteins (LDL) and very low density lipoproteins
(VLDL), and high concentrations of two high density
lipoprotein (HDL) subfractions, HDL2 and HDL3.1-3
Their LDL particles tend to have a high peak flotation rate and large peak particle diameter.4 EndurM
en
From the Research Medicine and Radiation Biology Division,
Lawrence Berkeley Laboratory, Berkeley, and Stanford Center for
Research in Disease Prevention, Stanford University School of
Medicine, Stanford, California.
Supported in part by grants HL-24462, HL-02183, and HL18574 from the National Heart, Lung, and Blood Institute of the
National Institutes of Health and conducted at the Lawrence
Berkeley Laboratory (Department of Energy DE-AC0376SF00098 to the University of California).
Address for reprints: Paul T. Williams, PhD, Research Medicine and Radiation Biology Division, Lawrence Berkeley Laboratory, Bldg. 934, 1 Cyclotron Road, Berkeley, CA 94720.
Received March 31, 1989; revision accepted December 13, 1989.
ance exercise can produce physiological changes that
promote these lipoprotein characteristics. As compared with sedentary men, long-distance runners
See p 1428
have higher HDL2 concentrations, lower small LDL
and VLDL concentrations, and higher LDL peak
flotation rates.56 Moreover, sedentary men who begin
running show changes in these lipoprotein subfractions that correlate with training level and weight
loss.7-9 These observations, however, are not conclusive proof that endurance exercise causes these lipoprotein changes. Self-selection might contribute to the
lipoprotein differences between runners and sedentary men.10 Strong proof of a cause-and-effect relation
requires the demonstration of significant lipoprotein
differences between exercise and control groups in a
randomized intervention trial.
The reasons for lipoprotein changes in runners are
the subject of controversy.1 -16 Some attribute the
1294
Circulation Vol 81, No 4, April 1990
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
changes to the muscle adaptations to running.14
Alternatively, runners are leaner than sedentary
men, and we have proposed that reduced adiposity
might explain some of the lipoprotein differences
between runners and sedentary men.7-9,15,16 It is
unclear from previous studies whether, in men, weight
loss by calorie restriction alone has the same effects on
lipoprotein metabolism as weight loss by exercise. Seldom controlled and seldom restricted to men, previous
diet studies have produced conflicting results for HDL
cholesterol,11,12,17-20 triglycerides,11,18-23 and LDL
cholesterol.1218"19,21-24 These inconsistencies might
relate, in part, to study differences in dietary composition, rapidity of the weight losses, or measurement
protocol (i.e., whether during active weight loss or
when weight has stabilized17). Although we have
found that the increase in HDL cholesterol can be
greater for exercise-induced weight loss than dietinduced weight loss,7 these results were obtained
from post-hoc correlation analyses and require confirmation. To our knowledge, the effects of dietinduced and exercise-induced weight loss on lipoprotein subfractions have not been previously compared
in a controlled clinical trial.
We, therefore, performed a randomized controlled
trial to compare the 1-year changes in lipoprotein
subfractions in men assigned to one of the following
three experimental conditions: Weight loss by exercise (primarily, running), weight loss by calorie
restriction, and control. Two main hypotheses were
tested. First, we tested the separate effects of
exercise-induced and diet-induced weight loss by
contrasting the mean changes of the diet and the
exercise groups with those of the controls. Second,
we tested for differences between exercise-induced
and diet-induced weight loss by contrasting the mean
changes in the exercise group with those of the diet
group. The changes in lipoprotein-cholesterol measurements in this study have been reported by Wood
et al.25 This report extends these findings to 1)
changes in the mass concentrations of subfractions
within the LDL, IDL, and VLDL regions, 2) changes
in the size and buoyancy of the predominant LDL
peak, and 3) changes in the total mass concentrations
of the HDL2 and HDL3 subfractions.
Methods
and
Subjects
Laboratory Measurements
We recruited 155 sedentary men, aged 30-59
years, 20-60% over Metropolitan ideal weight,26 who
were nonsmokers, not on medication that might
affect lipid metabolism, and nonhypertensive (blood
pressure, < 160/100 mm Hg). Their plasma total cholesterol concentrations were below 320 mg/dl, and
their plasma triglyceride concentrations were below
500 mg/dl. After their baseline evaluation, these men
were assigned at random into one of the following
three experimental conditions: Diet (calorie restriction without increasing exercise), exercise (physical
activity increase, primarily running, with no change in
diet), and control (no change in diet or exercise).25
The diet and exercise programs were each targeted to
reduce the men's body fat by one third over a
9-month period. We asked the men in the diet group
to reduce total calorie intake without changing the
proportions of fat, carbohydrates, protein, or alcohol
consumed. Their diets were individually prescribed,
assuming that a 7,762-kcal reduction in energy intake
would produce a 1-kg fat loss. To achieve a one third
body fat loss in the exercise group, the men were
asked to begin calisthenics and to walk, jog, or run for
25 minutes, three times per week at 60-80% of
maximal heart rate. The periods of continuous jogging were increased to 40-50 minutes, 5 days per
week. The controls were asked to remain sedentary
and to not change their diets. During the last 6 weeks
of the trial, the dieters attempted to stabilize their
weight loss by adjusting energy intake, and the exercisers attempted to stabilize their weight loss by
adjusting exercise level while keeping energy intake
constant.
At baseline, 7 months, and 1 year, the men
reported to our clinic in the morning, after having
abstained for 12-16 hours from all food and any
vigorous activity. We estimated body compositions by
hydrostatic weighing and maximal oxygen uptakes in
ml/kg/min (Vo2max) and 1/min (VoL) by recording
gas exchange during treadmill tests to exhaustion.25
Energy intakes were estimated by computer analysis
of food diaries maintained by the participants over a
7-day diet period.27 Self-report physical activity level
was estimated from a 7-day physical activity
questionnaire.28 Additionally, the runners recorded
exercise duration and frequency in diaries. These
entries were verified by the training staff. Blood
samples were collected in EDTA (1 mg/l) after an
overnight fast at baseline, 7 months, and 1 year.
Lipoprotein containing fractions were prepared and
studied by analytic ultracentrifugation as previously
described.29'30 Concentrations of total lipoprotein
mass were estimated by using computer techniques
for 15 HDL flotation intervals between F,?20 0-9
(half-integer increments from 0 to 6 and integer
increments, thereafter), 11 LDL flotation intervals
between SO 0-12 (integer increments between S0
0-10 and, then, S? 10-12), four IDL flotation intervals between S °12-20 (two-unit increments), and 14
VLDL flotation intervals between S O 20-400 (increments of 10 units below S? 100 and increments of 50
units, thereafter).29,30 Results are also presented for
HDL2 (Fl,20 3.5-9), HDL3 (Fl,20 0-3.5), small LDL (S f
0-7), large LDL (Sf 7-12), IDL (S 12-20), and
VLDL (S? 20-400) mass concentrations, and LDL
peak flotation (So) rates (i.e., the mode of the
distribution of LDL particles).29'30 Particle diameters
of the most abundant LDL species were determined
from nondenaturing polyacrylamide gradient gel
electrophoresis of the d less than or equal to 1.063
plasma fractions on 2-16% gels, stained for protein,
as previously described.31,32
Williams et al Weight Loss and Lipoprotein Subfractions
1295
TABLE 1. Body Mass Index, Calorie Intake, and Treadmill-Test Performance in Exercisers, Dieters, and Control Subjects
Control (mean±SD) Exercise-control difference Diet-control difference Exercise-diet difference
Body mass index (kg/m2)
- 1.04±0.48*
Baseline
29.95±2.38
-0.54±0.50
-0.50+0.46
A 7 mo
0.05±0.78
-2.46±0.22t
-0.98±0.18t
1.47±0.23t
A 1 yr
0.18±1.20
-1.41±0.261:
-2.45±0.27t
1.04±0.27*
Calorie intake (kcal/day)
Baseline
2,522.8±572.5
104.4±117.4
-26.3±123.1
130.7±110.2
A 7 mo
-98.4±503.3
14.5±105.4
347.2± 104.9i
-332.8±113.5*
A 1 yr
-103.2±493.2
-104.7±119.6
-237.0±106.2*
132.3± 114.5
Maximum aerobic capacity
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
(ml/kg/min)
Baseline
33.62±4.45
1.73± 1.00
0.28±0.93
1.45±0.92
A 1 yr
-2.45±3.24
2.48±1.00*
6.64±0.81*
4.15± 1.07t
VoL (I/min)
Baseline
3.20±0.46
0.12±0.09
-0.03±0.10
0.15±0.09
A 1 yr
-0.21±0.28
-0.02±0.06
0.42±0.08t
0.44±0.08t
Treadmill test duration (min)
Baseline
11.26±1.39
0.22±0.32
0.15±0.32
0.07±0.32
A 1 yr
-1.64±0.23
0.85+0.37*
1.10±0.36t
1.95±0.37t
Reported physical activity
(kcal/kg/day)
Baseline
35.17±3.38
-0.36±0.63
-0.49±0.64
0.13±0.50
A 1 yr
0.54+4.44
3.26±1.25t
0.42+0.89
2.84±1.18t
Values are mean±SEM, unless otherwise noted.
Forty-two controls, 42 dieters, and 46 exercisers had complete data on body mass index and reported physical activity; 39 controls, 40
dieters, and 46 exercisers had complete data on calorie intake; and 40 controls, 41 dieters, and 45 exercisers had complete data on Vo2max,
VoL, and treadmill-test duration.
*p<0.05, tp<0.O1, *p<O.OOl are significance levels for Wilcoxon two-sample test.
Statistics
The tables present the mean (+ 1 SD) for lipoprotein levels and other variables at baseline, and mean
changes in these variables between baseline, 7 months,
and 1 year. The effects of the diet and exercise
interventions are estimated by subtracting the mean
change scores of the control group from those of the
diet and exercise groups. The net change is then
presented ±+1 SEM. The significance of these differences are evaluated from the Wilcoxon two-sample
test. Pearson correlation coefficients and linear regression describe the pairwise associations between lipoprotein mass concentrations, weight loss, distance run
per week, maximum aerobic capacity (Vo2max), and
calorie intake. Analysis of covariance was used to
adjust changes in lipoproteins for changes in nutrient
intake and body mass index. This procedure uses
parallel regression lines to describe the relation
between dependent variable and the covariate. Separate intercepts are fitted to the regression lines of the
three groups, and the distances between the parallel
lines are used to test for significant group differences.
The analysis assumes that the relation between the
dependent variable and covariate is the same within
each group. The equality of the regression slopes was
tested before adjustment. The analyses include only
those subjects with complete data on lipoprotein subfractions and other variables, as required (see footnotes to Table 1). Because the assumption of bivariate
normality might not hold for serum lipoprotein mass
concentrations, adiposity, distance run, and energy
intake, we verified the standard significance levels by
permutation tests.33
Results
Six of the 52 exercisers, seven of the 51 dieters, and
eight of the 52 controls were omitted from the
analyses because their data were incomplete for
lipoprotein subfractions, and two additional controls
and two additional dieters were omitted because
their body composition measurements were incomplete. The remaining 46 exercisers, 42 dieters, and 42
controls seem well matched at baseline for body mass
index, calorie intake, Vo2max, treadmill test duration
(Table 1), plasma lipoprotein mass concentrations
(Table 2), and body composition.25
The controls' fitness decreased slightly, but their
body mass index, energy intake, and reported physical activity remained relatively constant. With one
exception, HDL3, the controls' lipoproteins showed
little mean change (Table 2). This suggests that
secular trends and experimental artifacts affecting
lipoproteins were mostly minor. In the analyses to
follow, the average change in the controls is subtracted from the average changes in the dieters and
exercisers. We assume these net differences estimate
the direct effects of the two weight-loss interventions.
1296
Circulation Vol 81, No 4, April 1990
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
TABLE 2. Lipoprotein Subfraction Concentrations in Exercisers, Dieters, and Control Subjects
Control (mean+SD) Exercise-control difference Diet-control difference Exercise-diet difference
HDL2 mass (mg/dl)
0.7±5.6
1.5 ±5.4
23.4+24.4
2.2+5.5
Baseline
0.1+7.2
16.4+3.7*
16.5 + 7.1*
0.4±16.2
A 7 mo
0.8+5.5
11.9±5.1t
12.7+5.4*
2.7±22.9
A 1 yr
HDL3 mass (mg/dl)
-3.0+8.2
-2.5±7.5
-5.5±8.2
215.3±34.0
Baseline
-1.7+6.9
12.8+7.4
11.1+6.6
-12.4+32.9
A 7 mo
19.3±7.1*
-2.7±6.8
- 17.2±31.9
A 1 yr
16.7±6.8t
Small LDL mass (mg/dl)
- 1.4± 13.9
11.5±15.5
10.1+14.4
216.2±58.1
Baseline
- 12.7+20.0
-35.3+10.0*
A 7 mo
-3.9±41.9
-22.5±11.3t
- 16.2±9.3
11.0±10.7
-27.2±+11.1*
-8.3±46.7
.A 1 yr
Large LDL mass (mg/dl)
- 19.3 ±8.5t
-10.6±9.1
120.9±46.0
-8.8±7.3
Baseline
4.1±35.8
14.9+7.5
12.8±8.3
2.1±8.0
A 7 mo
-1.9±7.6
5.4±7.4
8.7±36.9
7.3±8.2
A 1 yr
IDL mass (mg/dl)
-2.4±3.3
1.0±3.5
38.2±+14.3
-1.4±3.4
Baseline
-4.9+7.5
3.9±3.4
-1.8±9.7
-1.0±2.7
A 7 mo
0.8±12.2
-0.9±2.8
-5.6±2.9
4.7±3.0
A 1 yr
VLDL mass (mg/dl)
7.1±15.6
5.9±14.6
1.2± 16.6
109.1±62.6
Baseline
A 7 mo
2.8±48.9
-24.9±12.7
6.4±13.6
-31.3±11.7*
- 1.7± 12.9
A 1 yr
6.5±56.5
-29.0+ 13.3t
-27.3±11.4t
LDL peak flotation (S') rate
-0.23±0.25
5.69±+1.12
-0.12+0.23
0.11±0.25
Baseline
-0.02±0.62
0.01±0.16
0.46+0.15*
0.47±0.15*
A 7 mo
A 1 yr
0.11±0.74
0.19±0.14
0.29+0.18
-0.10±0.16
LDL peak particle diameter (A)
259.01+8.21
-1.08±1.73
0.62±1.84
-1.70±1.78
Baseline
A 7 mo
0.08±6.46
3.69+ 1.43*
-0.25± 1.42
3.43± 1.40t
A 1 yr
1.60±7.26
2.36± 1.33t
3.20± 1.68t
-0.83±1.44
Values are mean ±SEM, unless otherwise noted. HDL, high density lipoprotein; LDL, low density lipoprotein; IDL, intermediate density
lipoprotein; VLDL, very low density lipoprotein.
*p<0.01, tp<0.05 are significance levels for Wilcoxon two-sample test.
Table 1 shows that the intervention goals were
partially achieved. The exercisers ran (mean+SD)
15.6±9.1 km/wk during the year (18.9±13.1 km/wk
between the fifth and 12th months) and reported
higher physical activity levels while not significantly
decreasing their mean energy intake. Their fitness
increased by the end of the trial, that is, their
treadmill test lengthened, VoL increased, and
Vo2max increased. In contrast, the dieters
decreased total calorie intake without increasing
either VoL or self-reported physical activity. Body
mass index decreased significantly in both experimental groups, significantly less in exercisers than
dieters, despite our efforts to achieve similar losses
in both groups. The exercisers lost almost exclusively fat (-4.15±3.70 kg) with little lean body mass
change (+ 0.11 ± 2.22 kg), whereas dieters lost lean body
mass (-1.31±2.55 kg) as well as fat (-5.93±4.14 kg).
The exercisers lost 49% of their -8.53 kg fat-loss goal,
and the dieters lost 70% of their -8.53 kg fat-loss goal.
There were only modest changes in weight in dieters
(0.01±+ 1.08 kg) and exercisers (-0.3±+ 1.17 kg) during
the 6-week weight-stabilization period.
Effects of Exercise-Induced and Diet-Induced Weight
Loss on Lipoprotein Subfiactions
Table 2 shows that the exercise-induced and dietinduced weight loss groups each increased mean
plasma HDL2-mass concentrations, increased LDL
peak particle diameter, and decreased mean plasma
VLDL-mass concentrations. These results are significant for both the 7-month and 1-year measurements.
Exercise-induced and diet-induced weight loss also
each significantly reduced small LDL-mass concentrations and significantly increased mean LDL peak
flotation rate after 7 months. The significant reduction in small LDL concentrations was sustained in
dieters after 1 year but only marginally in exercisers
Williams et al Weight Loss and Lipoprotein Subfractions
1297
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
TABLE 3. Adjusted Mean Differences in the Lipoprotein Subfraction Changes of Exercisers, Dieters, and Controls
Adjusted for changes in body mass index
Adjusted for changes in nutrient intake*
ExercisersExercisersExercisersDietersExercisersDietersdieters
dieters
controls
controls
controls
controls
HDL2 mass (mg/dl)
-0.7±7.0
6.6±6.8
-8.4±7.8
7 mo-baseline
19.1+6.9t
19.8±7.2t
14.9±7.1t
2.2±5.6
11.1±5.Ot
-1.2±5.1
17.4±5.6t
15.2±5.9t
1 yr-baseline
-12.2±5.5it
HDL3 mass (mg/dl)
11.7±7.3
-4.2±7.4
-4.7±8.2
13.1±7.5
17.8±10.2
7 mo-baseline
16.0±7.6t
25.7±7.5§
-4.1±7.1
17.3±9.0
-1.8±7.3
15.5 ±7.6t
1 yr-baseline
21.6±7.1t
Small LDL mass (mg/dl)
-26.1 ±11.9t
-7.2±12.5
13.8±12.1
-9.3±12.1
-2.1±13.8
7 mo-baseline
-39.9±12.4§
-18.3±10.7
-25.2 ±11.2t
-2.6±10.6
6.9±10.6
2.2±10.8
4.9±11.7
1 yr-baseline
Large LDL mass (mg/dl)
10.4±9.2
10.1±8.2
9.4±8.5
-1.0±11.5
7 mo-baseline
18.8±8.14
8.7±8.4
3.4±8.0
1 yr-baseline
14.3±7.7
-0.3±7.6
-1.9±8.4
-5.3±9.9
14.7±8.1
IDL mass (mg/dl)
-3.4±3.3
7 mo-baseline
-0.5±3.2
-5.6±3.3
5.1±3.2
3.8±3.2
7.2±3.6
1.5±3.0
-3.5±3.2
4.9±3.0
2.1±3.1
-0.5±3.4
2.5±3.1
1 yr-baseline
VLDL mass (mg/dl)
11.9± 13.8
-8.1± 14.5
-24.4± 13.6
-15.3± 14.1
-7.2±16.0
7 mo-baseline
-36.2±14.2t
-23.1±12.9
3.2±12.8
-26.2±13.4
-27.1±14.5
0.9±13.2
1 yr-baseline
-26.3±13.5t
LDL peak flotation (S') rate
0.35±0.16:
0.08±0.16
0.24±0.16
-0.11±0.18
0.54±0.16§
0.46±0.16t
7 mo-baseline
0.28±0.16
-0.05±0.16
-0.17±0.16
-0.34±0.17
0.17±0.16
1 yr-baseline
0.34±0.17t
LDL peak particle diameter (A)
4.05±1.53§
-0.03±1.49
1.74±1.64
2.10+1.50
0.37±2.04
4.02±1.47t
7 mo-baseline
-0.80±1.47
0.77±1.49
1 yr-baseline
0.19±1.56
-0.58±1.85
3.07±1.48t
3.88±1.56t
Values are mean±SEM. HDL, high density lipoprotein; LDL, low density lipoprotein; IDL, intermediate density lipoprotein; VLDL, very
low density lipoprotein.
*Simultaneous adjustment for changes in animal protein, plant protein, carbohydrate, saturated fat, monounsaturated fat, polyunsaturated fat, alcohol, and fiber and cholesterol intake (per MJ energy intake) by analysis of covariance.
tp<0.01, tp<0.05, §p<0.001 are significance levels for analysis of covariance results.
(p .0.10). Neither intervention program significantly
influenced mean concentrations of large LDL or IDL
particle mass concentrations. Table 3 shows that
altered nutrient intake did not contribute significantly to differences in lipoprotein change among
exercisers, dieters, and controls.
Most of the mean differences in lipoprotein changes
between exercisers and controls and between dieters
and controls can be attributed to weight loss (Table 3).
Adjustment for body mass index change by analysis of
covariance eliminated the significance of the exercisers' HDL2 increase at 7 months (p=0.32) and 1 year
(p=0.82), their increase in LDL peak particle diameter after 7 months (p=0.16) and 1 year (p=0.90),
their 7-month small LDL decrease (p=0.43), and their
7-month increase in LDL peak flotation rate
(p=0.12). Adjustment for body mass index change
also eliminated the significance of the dieters' 7month HDL2 increase (p=0.82), their 7-month and
1-year increases in LDL peak particle diameter
(p=0.86 andp=0.75), their 7-month VLDL decrease
(p=0.70), their 7-month and 1-year small LDL
decreases (p=0.90 and p=0.71, respectively), and
their 7-month increase in LDL peak flotation rate
(p=0.59). Adjustment had little effect on the significance of the 1-year VLDL decrease in exercisers
(p=0.06) and dieters (p=0.10).
Exercise-Induced Versus Diet-Induced Weight Loss
Table 2 shows that exercise-induced weight loss
and diet-induced weight loss programs produced
comparable mean changes in lipoprotein subfraction
concentrations. The dieters, however, lost more
weight than the exercisers. This difference in weight
loss could mask important differences in lipoprotein
change. Two procedures were used to test whether
equivalent weight loss by diet and by exercise produced equivalent lipoprotein changes.
First, we used analysis of covariance to compare
the exercisers' and dieters' average change in plasma
lipoproteins at an equivalent mean change in body
mass index (Table 3 and Figure 1). This procedure
fits parallel regression lines to the relation between
change in body mass index and change in HDL2, and
then tests whether the distances between the lines
are significantly different from zero. At equivalent
weight loss, exercise and diet produced comparable
changes in the mass of small LDL, large LDL, IDL,
Circulation Vol 81, No 4, April 1990
1298
20
Unadjusted differences
15
E
10
-j
5
C]
Differences adjusted for
change In body mass index
N.S.
m
0
0
c
D
N.S.
.5.
ast
-10
c
Exercise---1 Contrnol
-.
Diet-
Exercli
r-ntrnl Diet
-
Exercisetl^ntrxfl
G(Initr911
Exercise
rnl M;f
--Liomroi-Dimr
_
CD
Unadjusted differences
Differences adjusted for
change In body mass index
10
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
N.S.
p|5
-
30-
0
a
Exercise-
.40, Control
Diet-
Exercli
Control Diet
Exercis-
Diet-
ExerciseDi
FIGURE 1. Bar graphs showing mean lipoprotein changes of
exercisers, dieters, and controls compared before (left) and
after (right) adjustment for change in body mass index. Before
adjustment, exercisers and dieters each increased HDL2 and
decreased small LDL concentrations relative to controls, and
there were no significant differences between exercisers and
dieters. Adjustment eliminates all small LDL differences
between groups and exerciser-control differences in HDL2.
Dieters' HDL2 increase, however, was less than expected,
considering their mean weight loss. When adjusted to equivalent weight loss, HDL2 increased significantly more in exercisers than dieters.
VLDL, and LDL peak particle diameter. The analyses of covariance, however, suggest that when
adjusted to equivalent change in body mass index,
HDL2 increased significantly more after 1 year in
exercisers than dieters (Figure 1). This is because the
dieters' mean HDL2 increased significantly less than
statistically expected given their mean weight loss,
that is, less than the change predicted from the
regression coefficient between changes in HDL2 and
changes in body mass index. The exercisers' adjusted
7-month increases in HDL2 and LDL peak flotation
rate were also greater than those of the dieters.
These results were also obtained when exercisers and
dieters were adjusted to equivalent changes in lean
and fat body mass; adjusted HDL2 increased
12.9+7.3 mg/dl (mean+SEM) more after 7 months
(p=0.08) and 9.9+5.1 mg/dl more after 1 year
(p=0.05) in exercisers than dieters, and LDL peak
flotation rate increased 0.35 mg/dl more after 7
months in exercisers than dieters (p=0.04).
Second, we restricted the comparison to a subset of
the exercisers and dieters who were within an overlapping range of weight loss, between -3.24 and
-0.11 kg/m2 at 7 months and between -3.52 and
-0.45 kg/m2 at 1 year. After 7 months, there were
only minor differences (±+SEM) in HDL3 (exercisediet, -7.5±7.7 mg/dl), small LDL (-4.0±13.0 mg/
dl), large LDL (7.5+9.0 mg/dl), IDL (1.2±4.0 mg/
dl), VLDL (5.2±13.9 mg/dl), and LDL peak particle
diameter (-0.73±1.64 A) in this subset. Similarly,
after 1 year, the exercisers and dieters showed little
difference (±SEM) in HDL3 (-5.0±7.9 mg/dl),
small LDL (-0.5±12.6 mg/dl), large LDL (-5.0±7.9
mg/dl), IDL (3.7±3.5 mg/dl), VLDL (3.4±13.6 mg/
dl), LDL peak flotation rate (0.14±0.18 mg/dl), and
LDL peak particle diameter (-0.54±1.64 A). When
restricted to a common weight-loss range, however,
the exercisers had marginally greater increases
(±SEM) in HDL2 after 7 months (4.6±7.7 mg/dl),
HDL2 after 1 year (8.5±6.0 mg/dl), and LDL peak
flotation rate after 7 months (0.25±0.17 Sf).
Although more direct and involving fewer assumptions, the second approach has less statistical power
to detect significant exercise group versus diet group
differences because the sample sizes are reduced.
Forty-one exercisers and 31 dieters are compared at
7 months, and 34 exercisers and 36 dieters are
compared at 1 year.
Correlational Analyses
Change in body mass index correlated significantly
and negatively with the 7-month and 1-year changes
in HDL2 and LDL peak flotation rate in all three
groups (Table 4). Weight loss also correlated positively and significantly with small LDL changes in
exercisers, dieters, and controls after 1 year, and in
exercisers after 7 months. Changes in IDL and body
mass index were positively correlated but these were
not always significant. Alteration in nutrient intake
generally did not account for the relation between
changes in lipoproteins and body mass index. The
lipoprotein changes generally correlated more
strongly with change in body mass index than change
in percentage of body fat.
The significant correlations between the exercisers'
lipoprotein changes, their distance run, and their
increased fitness can be largely ascribed to weight
loss. Distance run correlated with changes in HDL2
(r=0.30, pcO.05) and LDL peak flotation rate
(r=0.28,p=0.06) after 7 months, and with changes in
HDL2 (r=0.44,p.0.01) and LDL peak particle diameter (r=0.31,p<0.05) after 1 year. When adjusted for
change in body mass index, however, distance run
was no longer significantly correlated with 7-month
changes in HDL2 (r=0.16) and LDL peak flotation
rate (r=0.16) or with 1-year changes in HDL2
(r=0.25) and LDL peak particle diameter (r=0.12).
Adjustment for body mass index change also eliminated the significant relation between changes in
Williams et al- Weight Loss and Lipoprotein Subfractions
1299
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
TABLE 4. Correlation Coefficients Between Changes in Body Mass Index (kg/rm2) and Changes in Lipoprotein Mass Concentrations, Low
Density Lipoprotein Peak Flotation Rate, and Peak Particle Diameter in Exercisers, Dieters, and Control Subjects Who Participated in a
1-Year Study of Diet and Exercise
Correlations with ABMI
Unadjusted correlation with
Correlations with A percent
A body mass index
adjusted for A nutrient intake*
body fat
Exercisers Dieters Controls Exercisers Dieters Controls Exercisers Dieters Controls
HDL2 mass
-0.30
7 mo-baseline
-0.27
-0.55§
-0.26
-0.38t
-0.40t -0.29t
-0.40t -0.39t
-0.54§
1 yr-baseline
-0.374:
-0.57§
-0.42t
-0.56§
-0.25
-0.49t
-0.51t
-0.40t
HDL3 mass
7 mo-baseline
0.23
0.03
-0.07
0.13
0.07
-0.07
0.04
-0.19
0.05
1 yr-baseline
-0.11
0.11
-0.09
-0.09
-0.20
-0.04
0.12
0.09
0.46t
Small LDL mass
0.23
7 mo-baseline
0.304:
0.22
0.27
0.304:
0.13
-0.06
-0.07
0.33t
1 yr-baseline
0.334:
0.344
0.354
0.15
0.27
0.09
-0.02
0.31t
0.33t
Large LDL mass
7 mo-baseline
0.06
-0.13
-0.354:
0.06
-0.08
-0.26
0.01
0.06
-0.46t
1 yr-baseline
-0.21
-0.14
-0.18
-0.16
-0.32
-0.334
-0.12
0.00
-0.12
IDL mass
7 mo-baseline
0.23
0.27
0.07
0.34*
0.14
0.14
0.32*
0.42t
0.26t
1 yr-baseline
0.11
0.15
0.19
0.17
0.09
0.13
0.15
0.15
0.32t
VLDL mass
7 mo-baseline
0.16
0.18
0.14
0.17
0.14
0.18
0.10
0.11
0.17
1 yr-baseline
-0.12
0.16
-0.03
-0.07
0.21
-0.19
-0.22
0.00
-0.18
LDL peak flotation rate (S')
7 mo-baseline
-0.29
-0.354
-0.21
-0.09
-0.20
-0.21
-0.35t
-0.33:t -0.27t
1 yr-baseline
-0.53§
-0.41t -0.35*
-0.19
-0.53t
-0.17
-0.33t
-0.46t -0.40t
LDL peak particle diameter
7 mo-baseline
-0.26
-0.22
-0.09
-0.14
-0.13
-0.06
-0.05
-0.28
-0.24
1 yr-baseline
-0.27
-0.19
-0.27
-0.31
-0.51*
-0.22
-0.09
-0.11
-0.43t
BMI, body mass index; HDL, high density lipoprotein; LDL, low density lipoprotein; IDL, intermediate density lipoprotein; VLDL, very
low density lipoprotein.
*Simultaneous adjustment for changes in animal protein, plant protein, carbohydrate, saturated fat, monounsaturated fat, polyunsaturated fat, alcohol, and fiber and cholesterol intake (per MJ energy intake) by partial correlation.
tp<O.Ol, .p<0.05, §p<0.001 are significance levels for Pearson correlations.
HDL2 and treadmill test duration (r=0.36,ps0.01, to
r=0.22) and between changes in LDL peak flotation
rate and Vo2max (r=0.33,p<0.05, to r=-0.02), and
treadmill test duration (r=0.41, pO.O1, to r=0.26).
This adjustment had little effect on the correlation of
1-year change in IDL versus distance run (r= -0.30,
pcO.05, to r=-0.29) and treadmill test duration
(r= -0.30, pcO.05, to r= -0.27). Adjustment for distance run does not eliminate the exercisers' significant correlations between change in body mass index
and changes in HDL2-mass (r=-0.30 at 7 months,
r=-0.35 after 1 year), small LDL (r=0.30 at 7
months, r=0.36 after 1 year), LDL peak flotation
(r=-0.49 after 1 year), and LDL peak particle
diameter (r= -0.33 after 1 year).
Results for Individual Flotation Intervals
Figures 2 and 3 present the mean changes in HDL
mass and LDL mass concentrations by flotation rate
for exercisers, dieters, and controls. After 7 months,
HDL mass of F,'20 2.5-3.5 had increased significantly
more in dieters than exercisers; however, by 1 year,
this distinction between diet-induced and exercise-
induced weight loss ceased to exist. Other analyses (not
displayed) showed that adjustment for nutrient intake
had little effect on the range of HDL and LDL flotation
interval, showing significant treatment versus control
group differences at 7 months or 1 year. Reductions in
VLDL mass concentrations were significant within S 0f
50-250 in dieters and within S 0 100-200 in exercisers
after 7 months, and within S 0 100-200 in dieters and
within S ' 40-200 in exercisers after 1 year.
As shown by the correlations of Table 5, 7-month
and 1-year changes in body mass index were inversely
related to change in F?°20 3.5-8 HDL mass concentrations for exercisers, dieters, and controls. In exercisers, 1-year changes in HDL mass showed comparable correlations with distance run and with change
in body mass index throughout the range Fl?20 3.5-8.
Changes in Vo2max and treadmill test duration correlated with narrower ranges of HDL mass change.
Discussion
We have shown in this controlled randomized
experiment that weight loss by exercise alone or diet
alone significantly increases plasma HDL2 mass con-
Circulation Vol 81, No 4, April 1990
1300
10
8
6
4.
2
-2
-4.
E
-6
-8'
-10
-12
-14
-16
-18
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
-2
E
-4
-6
-8'
-10
-12
-14'
-
-16
-
Exercise
Diet
p<O.05 for Exercise-Control
p<O.05 for Diet-Control
Control
|
2
4
HDL flottation rate
6
-
|
8
(F1.20)
FIGURE 2. Graphs showing mean changes in HDL mass
concentrations by flotation rate in exercisers, dieters, and
controls. Cross-hatched areas designate significant differences
between exercisers and controls and between dieters and
controls.
FIGURE 3. Graphs showing mean changes in LDL mass
concentrations by flotation rate in exercisers, dieters, and
controls. Cross-hatched areas designate significant differences
between exercisers and controls and between dieters and
controls.
centrations, LDL peak flotation rate, and LDL particle diameter and decreases both small LDL and
VLDL plasma mass concentrations. It is likely that
processes associated with weight loss cause the
changes in HDL2, small LDL, and LDL peak flotation rate because, in addition to the controlled
results, the changes correlated significantly with the
amount of weight lost, separately in each of the three
experimental groups. Moreover, when adjusted to
equivalent change in body mass index, the exercise
minus control group differences in HDL2 and small
LDL changes were no longer significant, and there
were no significant differences in the exercisers' and
dieters' unadjusted mean change in small LDL,
VLDL, and LDL peak particle diameter (Table 3).
Although changes in LDL peak particle diameter
correlated with distance run, and HDL- change correlated significantly with distance run and changes in
Vo2max and treadmill test duration in exercisers,
these correlations were not significant when adjusted
for body mass index change.
The increase in HDL2 mass (+SEM) is consistent
with the increase in plasma HDL2 cholesterol
reported by Wood et a125 in these men after 7 months
(exercise-control, 2.71±1.11 mg/dl; diet-control,
2.50±0.75 mg/dl) and 1 year (2.60±0.83 and
2.59±0.75 mg/dl, respectively). The decrease in
VLDL mass concurs with Wood et al's25 finding that
both weight-loss modalities decreased plasma concentrations of triglycerides (exercise control,
-
-21.79±10.66 mg/dl; diet-control, -35.05 10.87
mg/dl after 7 months; -22.12±10.67 and
-31.17±12.81 mg/dl, respectively, after 1 year).
Wood et a125 found no significant change in plasma
concentrations of LDL cholesterol after 7 months
(exercise-control, 1.39±4.12 mg/dl; diet-control,
-4.59±4.41 mg/dl) or 1 year (-1.95±5.29 and
-3.98±5.53 mg/dl, respectively). The LDL cholesterol measurement might be insensitive to changes in
±
Williams et al Weight Loss and Lipoprotein Subfractions
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
TABLE 5. Correlations Between Changes in High Density Lipoprotein Mass Concentrations
Distance Run, and Fitness
Exercise
Exercise
1-Yr A
1-Yr A
treadmill-test
7-Mo
1-Yr
duration
ABMI
ABMI
Interval (F120) 1-Yr distance run Vo2max
0.14
0.00
-0.25
0.25
-0.36*
8.9-9.0
0.19
0.06
0.32*
-0.36*
-0.36*
7.0-8.0
0.27
0.19
6.0-7.0
0.38t
-0.37t
-0.45t
0.32*
0.29
-0.481
5.5-6.0
-0.38t
0.39t
0.33*
0.36*
-0.491
5.0-5.5
-0.39t
0.41t
0.35*
4.5-5.0
0.44t
0.42t
-0.38t
-0.501
0.33*
4.0-4.5
-0.34*
0.45t
-0.481
0.45t
0.20
3.5-4.0
0.371
0.38t
-0.31*
-0.39t
0.24
0.04
0.25
3.0-3.5
-0.28
-0.26
-0.13
-0.02
0.11
2.5-3.0
-0.20
-0.18
0.06
0.00
-0.06
2.0-2.5
0.00
-0.11
0.09
0.00
-0.21
0.27
1.5-2.0
-0.01
0.16
-0.22
0.01
-0.04
1.0-1.5
0.41t
0.18
-0.12
-0.03
0.35*
0.04
0.5-1.0
0.22
0.12
0.04
0.01
0.03
0.0-0.5
1301
by Flotation Interval and Body Mass Index,
Diet
7-Mo
ABMI
-0.28
-0.38t
-0.41t
-0.40t
-0.35*
-0.34*
-0.37*
-0.34*
-0.19
-0.05
0.04
0.10
0.10
0.12
0.07
Control
1-Yr
ABMI
-0.09
-0.22
-0.35*
-0.41t
-0.42t
-0.44t
-0.441
-0.38*
-0.27
-0.14
-0.07
-0.06
0.00
0.12
0.14
7-Mo
ABMI
0.07
0.03
-0.12
-0.26
-0.31*
-0.34*
-0.34*
-0.27
-0.20
-0.17
-0.13
-0.05
0.07
0.07
0.13
1-Yr
ABMI
-0.34*
-0.514
-0.53t
-0.541
-0.54:
-0.514
-0.50:
-0.48t
-0.451
-0.37*
-0.25
-0.04
0.24
0.37*
0.25
BMI, body mass index (kg/M2).
*p<0 05, tp<0.01, tp<O.OOl are significance levels for Pearson correlations.
LDL distribution because of its nonspecificity, that is,
it encompasses small LDL, large LDL, and IDL
cholesterol.30,31 Variations in the relative concentrations of small, dense and larger, more-buoyant LDL
species might contribute to variations in LDL peak
flotation rate and particle diameter as assessed by
analytic ultracentrifugation and gradient gel electrophoresis, respectively. The increases in LDL peak
flotation rate and particle diameter associated with
weight loss could reflect an increase in the ratio of
large to small LDL, as well as increased size and
density of the major LDL component.
Although some investigators have ascribed the
lipoprotein changes during diet-induced weight loss
to altered nutrition, and cross-sectional surveys and
experimentation do suggest that dietary composition
might affect lipoprotein subfraction concentrations
or distributions,34,35 perhaps through adiposity
changes,36 the present study has shown that adjustment for changes in dietary composition generally
did not affect the significance of the group differences or the correlations between changes in body
mass index and lipoproteins. Although this suggests
that the lipoprotein changes we observed during
weight loss were not because of changes in dietary
composition, it should be recognized that these
adjustments are limited by the precision of the 7-day
food-record estimates of nutrient and total calorie
intake. For example, weight loss in dieters and exercisers were less than predicted from their reported
calorie intakes and expenditures.
Exercise-Induced Weight Loss
The hypothesis that exercise might elevate HDL
cholesterol and affect other lipoproteins through
weight loss is controversial. Williams'5 studied the
relation of reduced adiposity to HDL cholesterol
concentrations in 23 published cross-sectional comparisons of long-distance runners and sedentary men.
He showed that the runners' and sedentary men's
mean HDL cholesterol differences were largely
explained by their adiposity differences (r=0.80,
across studies). Yet, mean training distances were
unrelated to the runners' and nonrunners' HDL
cholesterol differences in these studies. From these
results, he proposed the following theory: Longdistance runners have the lipoprotein metabolism of
men who are below their usual weight (their purported sedentary set-point weight) and not the
metabolism of equivalently lean men who are neither
exercising nor dieting.
We propose the following mechanism for the lipoprotein changes in exercisers: Lipoprotein lipase
activities are increased in the muscle and adipose
tissues of long-distance runners.37 Although the
importance of the runners' higher lipoprotein lipase
activity in muscles is generally emphasized,13"4 published data suggest that the increased lipoprotein
lipase activity in the adipose tissue of runners
predominates,37 with adipose activity being 79% of
the whole body lipoprotein lipase activity in
runners.37 If the probability of a lipolytic reaction is a
function of lipoprotein lipase activity, then the site of
chylomicron and VLDL lipolysis is more likely to
occur in adipose tissue than in muscle tissue. A
depletion of adipocyte triglyceride stores with
exercise-induced weight loss could, therefore, induce
increased adipocyte lipoprotein lipase activity, which
could in turn affect lipoprotein levels. Cross-sectional
studies of runners suggest that formerly obese mar-
1302
Circulation Vol 81, No 4, April 1990
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
athon runners have smaller fat cells than other
runners,38 and adipose tissue lipoprotein lipase is
inversely related to runners' fat cell diameter.39 Thus,
runners who have lost the most weight since starting
to run (i.e., those farthest below their weight setpoint) have the greatest increases in adipose tissue
lipoprotein lipase activity, and might also have the
greatest increases in HDL2, LDL peak flotation rate,
and particle diameter, and the greatest decreases in
small LDL and VLDL concentrations. This same
mechanism could explain the changes in lipoproteins
seen in weight-losing dieters. Previously obese, sedentary subjects who are weight-stable at a reduced
weight have increased lipoprotein lipase activity in
adipose tissue.40
The low triglyceride and VLDL concentrations of
runners and dieters might be explained, in part, by
high lipoprotein lipase activity causing chylomicron
and VLDL particles to be catabolized and cleared
more rapidly. Fasting VLDL might also reflect a
reduction in the postprandial triglyceride-rich lipoprotein pool.41 Some attribute elevated HDL2 of
exercisers to more rapid transfer of free cholesterol
and phospholipids to HDL during the accelerated
catabolism VLDL and chylomicrons.42,43 The
reduced concentrations of triglyceride-rich lipoproteins in exercisers and dieters might also reduce
cholesteryl ester-triglyceride exchange between lipoprotein subclasses.44 Running and weight loss each
decrease postprandial lipemia (the chylomicron
pool) and fasting VLDL.45-48 This, in turn, could
result in the accumulation of cholesteryl-esterenriched HDL2 and large LDL subspecies, and
reduced formation of HDL and LDL particles with
triglyceride-enriched cores that could be hydrolyzed
to smaller, denser HDL and LDL subspecies.49,50
Weight loss by exercise or dieting might also
reduce hepatic lipase activity.6,5152 Hepatic lipase
hydrolyzes the HDL phospholipids.53-55 Furthermore, the rate of cholesterol transfer from HDL to
hepatocytes is reported to decrease as the phospholipid/cholesterol ratio of the HDL particle is
increased.56-58 Thus, reduced hepatic lipase activity
in runners and dieters might result in the accumulation of HDL2 particles that have high phospholipid
content relative to their cholesterol content and,
therefore, less rapid transfer of HDL2 cholesteryl
esters to hepatocytes.
Exercise-Induced Weight Loss Versus Diet-Induced
Weight Loss
Exercise- and diet-induced weight losses each
increase lipoprotein lipase activity and each decrease
hepatic lipase activity.6,40,51,52,59 In dieters, however,
low caloric intake might mitigate the effects of these
lipase changes on HDL2 and other lipoproteins.
Nikkila et a143 calculate that fivefold to 10-fold more
surface phospholipids and cholesterol are transferred
to HDL from chylomicrons than from endogenous
VLDL.43 Therefore, despite similar increases in lipoprotein lipase, dieters might increase HDL2 and
HDL cholesterol concentrations less than exercisers
because dieting reduces the amount of cholesterol,
phospholipids, and apolipoproteins available for
transfer to HDL. Caloric restriction might also cause
less accumulation of HDL cholesterol and HDL
phospholipids when hepatic lipase activity is
reduced,56 and less accumulation of HDL cholesteryl
ester when cholesteryl ester-triglyceride exchange is
reduced.44 Alternatively, in exercisers, high-energy
intake might accentuate the HDL2 increase. Higher
caloric intake might increase the uptake of cholesterol and phospholipid by HDL particles during
lipolysis, accentuate HDL2 accumulation when low
hepatic lipase activity reduces HDL phospholipid
hydrolysis and HDL cholesterol efflux, and accentuate HDL2 increase when less HDL cholesteryl ester
is transferred to triglyceride-rich lipoproteins.
These interpretations are consistent with the correlations presented by Stefanick et a151 in these men.
Weight loss produced the same reductions in hepatic
lipase activity whether it was achieved by exercise
(r=0.50) or dieting (r=0.51). Weight loss also produced similar increases in postheparin plasma lipoprotein lipase activity in exercisers (r= -0.36) and
dieters (r= -0.29). Changes in HDL2 mass, however,
were more strongly correlated with change in hepatic
lipase activity in men who lost weight by running
(r= -0.40) than men who lost weight by dieting
(r= -0.24). The increase in lipoprotein lipase was
also associated with greater increases in HDL2 mass
in runners (r=0.38) than dieters (r=0.00).
Our findings also coincide with other published
data. For equivalent losses of body weight, the
increases in HDL cholesterol for most weightloss-by-diet studies are less than the HDL cholesterol
differences between lean runners and heavier sedentary men. They are also less frequently significant.
Failure to measure HDL cholesterol in steady-state
conditions might account for some of the inconsistent
findings of diet studies. Reductions in adipocyte
triglyceride stores do not necessarily increase adipose
tissue lipoprotein lipase activity in the hypocaloric
state.59,60 Differences in energy flux could be an
important difference between exercise- and dietinduced weight losses. Men who are below their
usual weight by dieting generally sustain their fat loss
only by continuing to restrict their energy intake. The
increase in adipose tissue lipoprotein lipase activity
in those men might primarily serve to return adipose
mass and fat-cell size to that specified by the postulated set-point for body weight.61 Men who have lost
weight by long-distance running, however, are able to
sustain the reductions in fat-cell size on unrestricted
diets that often contain 40-60% more calories than
sedentary men who are at stable weight.62
Acknowledgments
Analytic ultracentrifuge measurements were made
by the staff of the Research Medicine and Radiation
Biology Division under the direction of Dr. Frank
Lindgren. We wish to thank Dr. Marcia Stefanick,
Williams et al Weight Loss and Lipoprotein Subfractions
Mr. Richard Terry, Ms. Darlene Dreon, Ms. Barbara
Frey-Hewitt, and Ms. Nancy Ellsworth for their help
in completing the study.
References
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
1. Krauss RM, Lindgren FT, Williams PT, Kelsey SF, Brensike J,
Vranizan K, Detre KM, Levy RI: Intermediate-density lipoproteins and progression of coronary artery disease in hypercholesterolaemic men. Lancet 1987;2:62-66
2. Ballantyne FC, Clark RS, Simpson HS, Ballantyne D: High
density and low density lipoprotein subfractions in survivors of
myocardial infarction and in control subjects. Metabolism
1982;31:433-437
3. Gofman JW, Young W, Tandy R: Ischemic heart disease,
atherosclerosis, and longevity. Circulation 1966;34:679-697
4. Austin MA, Breslow JL, Hennekens CH, Buring JE, Willett
WC, Krauss RM: Low density lipoprotein subclass patterns
and risk of myocardial infarction. JAMA 1988;260:1917-1921
5. Wood PD, Haskell WL, Klein H, Lewis S, Stern MP, Farquhar
JW: The distribution of plasma lipoproteins in middle-aged
male runners. Metabolism 1976;25:1249-1257
6. Williams PT, Krauss RM, Wood PD, Lindgren FT, Giotas C,
Vranizan KM: Lipoprotein subfractions of runners and sedentary men. Metabolism 1986;35:45-52
7. Williams PT, Wood PD, Krauss RM, Haskell WL, Vranizan
KM, Blair SN, Terry R, Farquhar JW: Does weight loss cause
the exercise induced increase in plasma high-density lipoproteins? Atherosclerosis 1983;47:173-185
8. Williams PT, Krauss RM, Vranizan KM, Albers JJ, Terry RB,
Wood PDS: The effects of long-distance running and weight
loss on plasma low-density-lipoprotein subfraction concentrations in men. Arteriosclerosis 1989;9:623-632
9. Wood PD, Haskell WL, Blair SN, Williams PT, Krauss RM,
Lindgren FT, Albers JJ, Ho PH, Farquhar JW: Increased
exercise level and plasma lipoprotein concentrations: A oneyear randomized study in sedentary middle-aged men. Metabolism 1983;32:31-39
10. Williams PT, Wood PD, Haskell WL, Vranizan KM: The
effects of running mileage and duration on plasma lipoprotein
concentrations. JAMA4 1982;247:2674-2679
11. Schwartz RS: The independent effects of dietary weight loss
and aerobic training on high density lipoproteins and apolipoprotein A-I concentrations in obese men. Metabolism 1987;
36:165-171
12. Sopko G, Leon AS, Jacobs DR, Foster N, Moy J, Kuba K,
Anderson JT, Casal D, McNallt C, Frantz I: The effects of
exercise and weight loss on plasma lipids in young obese men.
Metabolism 1985;34:227-236
13. Nikkili EA: Role of lipoprotein lipase in metabolic adaptation
to exercise training, in Borensztajn J (ed): Lipoprotein Lipase.
Chicago, Evener Publishers, 1987, pp 187-199
14. Kiens B, Lithell H: Lipoprotein metabolism influenced by
training-induced changes in human skeletal muscle. J Clin
Invest 1989;83:558-564
15. Williams PT: Weight-set point and the high-density lipoprotein
concentrations of long-distance runners. Metabolism (in press)
16. Williams PT: Weight set-point theory predicts HDLcholesterol levels in previously-obese long-distance runners.
Int J Obesity (in press)
17. Friedman CI, Falko JM, Patel ST, Kim MH, Newman HAI,
Barrows H: Serum lipoprotein responses during active and
stable weight reduction in reproductive obese females. J Clin
Endocrinol Metab 1982;55:258-262
18. Brownell KD, Stunkard AJ: Differential changes in plasma
high density lipoprotein cholesterol levels in obese men and
women during weight reduction. Arch Intern Med 1981;
141:1142-1146
19. Streja DA, Boyko E, Rabkin SW: Changes in plasma high
density lipoprotein cholesterol concentration after weight
reduction in grossly obese subjects. BrMed J 1980;281:770-772
20. Thompson PD, Jeffery RW, Wing RR, Wood PD: Unexpected
decrease in plasma high density lipoprotein cholesterol with
weight loss. Am J Clin Nutr 1979;32:2016-2021
1303
21. Zimmerman J, Kaufmann NA, Fainaru M, Eisenberg S,
Oschry Y, Friedlander Y, Stein Y: Effects of weight loss in
moderate obesity on plasma lipoprotein and apolipoprotein
levels and on high density lipoprotein composition. Atherosclerosis 1984;4:115-123
22. Weltman A, Matter S, Stamford BA: Caloric restriction and/or
mild exercise: Effects on serum lipids and body composition.
Am J Clin Nutr 1980;33:1002-1009
23. Wolf RN, Grundy SM: Influences of weight reduction on
plasma lipoproteins of obese patients. Arteriosclerosis 1983;
3:160-169
24. Walker WJ, Lawry EY, Love DE, Mann GE, Levine SA, Stare
FJ: Effects of weight reduction and caloric balance on serum
lipoprotein and cholesterol levels. Am JMed 1953;14:654-664
25. Wood PD, Stefanick ML, Dreon D, Frey-Hewitt B, Garay SC,
Williams PT, Superko HR, Fortmann SP, Albers JJ, Vranizan
KM, Ellsworth NM, Terry RB, Haskell WL: Changes in
plasma lipids and lipoproteins in overweight men during
weight loss through dieting as compared with exercise. N Engl
J Med 1988;319:1173-1179
26. Metropolitan Life Insurance Co: New weight standards for
men and women. Statistical Bulletin. New York, Nov-Dec. 1959
27. Nutrition Coordinating Center: Reference Food Table, Version
10. Minneapolis, Minn, University of Minnesota, 1984
28. Blair SN, Haskell WL, Ho P, Paffenbarger RS Jr, Vranizan
KM, Farquhar JW, Wood PD: Assessment of habitual physical
activity by a seven-day recall in a community survey and
controlled experiments. Am J Epidemiol 1985;122:794-804
29. Lindgren FT, Jensen LC, Hatch FT: The isolation and quantitative analysis of lipoproteins, in Nelson GJ (ed): Blood
Lipids and Lipoproteins: Quantitation, Composition and Metabolism. New York, Wiley-Interscience, 1972, pp 181-274
30. Lindgren FT, Jensen LC, Wills RD, Freeman RD: Flotation
rates, molecular weights and hydrated densities of the lowdensity lipoproteins. Lipids 1969;4:337-344
31. Krauss RM, Burke DJ: Identification of multiple subclasses of
plasma low density lipoproteins in normal humans. J Lipid Res
1982;23:97-104
32. Nichols AV, Krauss RM, Musliner TA: Nondenaturing polyacrylamide gradient gel electrophoresis, in Segrest JP, Albers
JJ (eds): Methods in Enzymology. Plasma Lipoproteins. Panl A.
Preparation, Structure, and Molecular Biology. Orlando, Fla,
Academic Press, Inc, vol 128, 1986, pp 417-431
33. Edgington ES: Randomization Tests. New York, Marcel Dekker, 1980, pp 1-287
34. Williams PT, Krauss RM, Wood PD, Albers JJ, Dreon D,
Ellsworth N: Associations of diet and alcohol intake with high
density lipoprotein subclasses. Metabolism 1985;34:524-530
35. Williams PT, Krauss RM, Wood PD, Kindel-Joyce S, Dreon
DM, Vranizan KM, Wood PD: Relationship of dietary fat,
protein, cholesterol and fiber intake to atherogenic lipoproteins in men. Am J Clin Nutr 1985;44:788-797
36. Dreon DM, Frey-Hewitt B, Ellsworth N, Williams PT, Terry
RB, Wood PD: Associations of diet and obesity in middleaged men. Am J Clin Nutr 1988;47:995-1000
37. Nikkili EA, Taskinen MR, Rehunen S, Harkonen M: Lipoprotein lipase activity in adipose tissue and skeletal muscle of
runners: Relation to serum lipoproteins. Metabolism 1978;
27:1661-1671
38. Tremblay A, Despr6s JP, Bouchard C: Adipose tissue characteristics of ex-obese long-distance runners. Int J Obes 1984;
8:641-648
39. Savard R, Despr6s JP, Deshaies Y, Marcotte M, Bouchard C:
Adipose tissue lipid accumulation pathways in marathon runners. Int J Sports Med 1985;6:287-291
40. Schwartz RS, Brunzell JD: Increase of adipose tissue lipoprotein lipase activity with weight loss. J Clin Invest 1981;
67:1425-1430
41. Nestel PJ: Relationship between plasma triglycerides and
removal of chylomicrons. J Clin Invest 1964;43:943-949
42. Patsch JR, Gotto AM, Olivecrona T, Eisenberg S: Formation
of high density lipoprotein2-like particles during lipolysis of
very low density lipoproteins in vitro. Proc Natl Acad Sci USA
1978;75:4519-4523
1304
Circulation Vol 81, No 4, April 1990
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
43. Nikkili EA: HDL in relation to the metabolism of triglyceriderich lipoproteins, in Miller NE, Miller GJ (eds): Clinical and
Metabolic Aspects of High-Density Lipoproteins. Amsterdam,
Elsevier Science Publishers BV, 1984, pp 217-245
44. Tall AR: Plasma lipid transfer proteins. J Lipid Res 1988;
27:361-367
45. Nikkila EA, Konttinen A: Effects of physical activity on
postprandial levels of fats in serum. Lancet 1962;2:1151-1154
46. Sady SP, Thompson PD, Cullinane EM, Kantor MA, Domagala E, Herbert PN: Prolonged exercise augments plasma
triglyceride clearance. JAMA 1986;250:2552-2555
47. Vihko V, Sarviharju J, Souminen H: Effects of endurance
training on concentrations of individual plasma free fatty acids
in young men at rest and after moderate bicycle ergometer
exercise. Ann Med Exp Biol Fenniae 1973;51:112-117
48. Bray GA: Metabolic effects of corpulence, in Howard IA (ed):
Recent Advances in Obesity Research. London, Newman Publishing, 1975, pp 56-65
49. Krauss RM: Physical heterogeneity of apolipoprotein Bcontaining lipoproteins, in Lippel K (ed): Proceedings of the
Workshop on Lipoprotein Heterogeneity. NIH Publication No.
87-2646. U.S. Department of Health and Human Services,
1987
50. Deckelbaum RJ: Coupled lipid transfer and lipolysis in intravascular processing/remodeling of apolipoprotein B containing lipoproteins, in Lippel K (ed): Proceedings of the Workshop
on Lipoprotein Heterogeneity. NIH Publication no. 87-2646, US
Dept of Health and Human Services, 1987, pp 57-65
51. Stefanick ML, Terry RB, Haskell WL, Wood PD: Relationships of changes in post-heparin hepatic and lipoprotein lipase
activity to HDL-cholesterol changes following weight loss
achieved by dieting versus exercise, in Gallo L (ed): Cardiovascular Disease: Molecular and Cellular Mechanisms, Prevention, and Treatment. Plenum Press, 1987, pp 61-68
52. Marniemi J, Peltonen P, Vuori I, Hietanen E: Lipoprotein
lipase of human postheparin plasma and adipose tissue in
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
relation to physical training. Acta Physiol Scand 1980;
110:131-135
Groot PHE, Scheek LM, Jansen H: Liver lipase and high
density lipoprotein. Lipoprotein changes after incubation of
human serum with rat liver lipase. Biochim Biophys Acta
1983;75:393-400
Grosser J, Schrecker 0, Greten H: Function of hepatic
triglyceride lipase in lipoprotein metabolism. J Lipid Res
1981;22:437-442
Jansen H, Van Tol A, Hulsmann WC: On the metabolic
function of heparin releasable liver lipase. Biochem Biophys
Res Commun 1980;92:53-59
Miller NE: Current concepts of the role of HDL in reverse
cholesterol transport, in Miller NE, Miller GJ (eds): Clinical
and Metabolic Aspects of High-density Lipoproteins. Amsterdam, Elsevier Science Publishers BV, 1984, pp 187-216
Rothblat GH, Arbogast LY, Ray EK: Stimulation of esterified
cholesterol accumulation in tissue culture cells exposed to high
density lipoproteins enriched in free cholesterol. J Lipid Res
1978;19:350-358
Bamberger M, Glick JM, Rothblat GH: Hepatic lipase stimulates the uptake of high density lipoprotein cholesterol by
hepatoma cells. J Lipid Res 1983;24:869-876
Taskinen M-R, Nikkila EA: Effects of caloric restriction on
lipid metabolism in man. Atherosclerosis 1979;32:289-299
Guy-Grand B, Bigorie B: Effect of fat cell size, restrictive diet
and diabetes on lipoprotein lipase release by human adipose
tissue. Horm Metab Res 1975;7:471-475
Keesey RE: A set point analysis of the regulation of body
weight, in Stunkard AJ (ed): Obesity. Philadelphia, WB Saunders, 1980, pp 144-165
Blair SN, Ellsworth NM, Haskell WL, Stern MP, Farquhar
JW, Wood PD: Comparison of the nutrient intake in middleaged men and women runners and controls. Med Sci Sports
Exerc 1981;13:310-315
KEY WORDS * lipoproteins * weight loss
-
exercise
Changes in lipoprotein subfractions during diet-induced and exercise-induced weight loss
in moderately overweight men.
P T Williams, R M Krauss, K M Vranizan and P D Wood
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
Circulation. 1990;81:1293-1304
doi: 10.1161/01.CIR.81.4.1293
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1990 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539
The online version of this article, along with updated information and services, is located on
the World Wide Web at:
http://circ.ahajournals.org/content/81/4/1293
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally
published in Circulation can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further
information about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation is online at:
http://circ.ahajournals.org//subscriptions/