PATHOPHYSIOLOGY AND NATURAL HISTORY
CORONARY ARTERY DISEASE
Plasma and lipoprotein cholesterol and triglyceride
concentrations in the Pima Indians: distributions
differing from those of Caucasians
BARBARA V. HOWARD, PH.D., MICHAEL P. DAVIS, M.S., DAVID J. PETrITT, M.D.,
WILLIAM C. KNOWLER, M.D., DR. P.H., AND PETER H. BENNETT, M.B., M.R.C.P.
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
ABSTRACT Plasma and lipoprotein cholesterol and triglyceride concentrations were measured in
Pima Indians, a genetically homogeneous population with a high prevalence of obesity and diabetes
and a low frequency of coronary heart disease. The data for nondiabetic Pima Indians were compared
with those of U.S. Caucasians measured during the Lipid Research Clinics Prevalence Study. Plasma
total and low-density lipoprotein (LDL) cholesterol levels in Pima men older than 30 years of age and in
women over 25 were lower than those of U.S Caucasians, and in men the total and LDL cholesterol
levels did not increase with age. In Pima male subjects of all ages high-density lipoprotein (HDL)
cholesterol levels were lower than that in Caucasian male subjects, and in Pima female subjects HDL
cholesterol levels averaged 11 mg/dl lower than those in female Caucasian subjects. On the other hand,
plasma triglyceride levels in Pima male subjects up to 35 years of age and in female subjects up to 55
years old were higher than those of Caucasians. Comparison of plasma cholesterol and triglyceride
levels in Pima male and female subjects showed the same relationships as those in Caucasians (i.e.,
higher in male subjects until the older age groups), but Pima male and female subjects had similar HDL
cholesterol levels. There was a strong negative correlation between obesity and HDL cholesterol levels
in Pimas. Obesity appeared to account for the lower HDL cholesterol levels in Pima male subjects
compared with that in Caucasian male subjects, but obesity could not explain the lack of differences
between the sexes in HDL levels in the Pimas. Triglyceride levels correlated positively with obesity in
the Pimas, but the relationship was not as strong as that reported for Caucasians. The results indicate
that there are differences in lipoprotein distributions between Pimas and U.S. Caucasians; the relationship between lipoprotein levels and the incidence of coronary disease in the Pimas must be determined.
Circulation 68, No. 4, 714724, 1983.
.
EPIDEMIOLOGIC STUDIES have established an association of total and low-density lipoprotein (LDL)
cholesterol levels with an increased risk of arteriosclerotic heart disease,' 3 a possible association between
total and very low-density lipoprotein (VLDL) triglyceride levels and coronary heart disease,-- and a strong
negative relationship between high-density lipoprotein
(HDL) cholesterol levels and coronary heart disease.7-9
To understand these associations it is necessary to determine the extent to which plasma lipoprotein levels
are influenced by diet and other environmental factors
and the extent to which they are genetically controlled.
One approach to understanding factors that control
plasma lipoprotein levels is to compare different racial
From the Phoenix Clinical Research Section and Southwestern Field
Studies Section, National Institute of Arthritis, Diabetes, and Digestive
and Kidney Diseases, National Institutes of Health, Phoenix.
Address for correspondence: Dr. Barbara V. Howard, Phoenix Clinical Research Section, NIH-NIADDK, 4212 North 16th St., Phoenix,
AZ 85016.
Received March 7, 1983; revision accepted June 9, 1983.
714
groups. The largest epidemiologic study of plasma
lipoprotein levels has been the Lipid Research Clinics
(LRC) Prevalence Study,'0 which has measured plasma cholesterol concentrations and triglyceride and lipoprotein cholesterol concentrations in a large number
of U.S. residents. Comparison of distributions of plasma triglyceride and plasma and lipoprotein cholesterol
levels between blacks and Caucasians who were examined during the LRC survey suggests that there are
differences between U.S. blacks and Caucasians both
in plasma triglyceride levels and in HDL cholesterol
concentrations. I I
There are a few other studies that compare plasma
lipid or lipoprotein levels, which were measured with
similar methods in different racial groups. These include a comparison of an African black and European
Caucasian community'2 and a study of New Zealand
Maoris and other New Zealand residents.'3 Both of
these comparative studies report racial differences in
plasma lipid and lipoprotein distributions. In addition,
CIRCULATION
PATHOPHYSIOLOGY AND NATURAL HISTORY-CORONARY ARTERY DISEASE
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
Rhoads et al. 14 published data suggesting that plasma
cholesterol levels are lower in Japanese men residing
in Hawaii than in other residents of Hawaii; Connor et
al." observed extremely low cholesterol levels in the
Tarahumara Indians.
This report presents data on plasma lipoprotein cholesterol and triglyceride levels in the Pima Indians.
This genetically homogeneous, 16 well-characterized
population'7 has a low frequency of coronary heart
disease,'8 and previous studies have suggested that total plasma cholesterol concentrations are low in this
group.'9 We have used methods that were carefully
standardized to be comparable to those used by the
LRC so that direct comparison could be made with the
Caucasian populations in the LRC group. The high
prevalence of obesity20 in the Pima population has allowed examination of the relationships between obesity and plasma lipoprotein cholesterol and triglyceride
concentrations over a wide range of obesity.
Methods
Study subjects. The study subjects were Pima Indians participating in a longitudinal population study in the Gila River
Indian Community of Arizona. Samples were collected for lipoprotein quantification from April 1979 through May 1982 from
all fasting subjects who were at least 15 years of age and of at
least half Pima ancestry. The Pima study population over 15
years of age at that time consisted of approximately 1500 male
and 1550 female subjects. Fasting blood samples were obtained
from 506 male (approximately 34%) and 885 female subjects
(approximately 57%). Women who were pregnant or taking oral
contraceptives were analyzed sep'arately.
Samples for lipoprotein determinations were obtained at a
routine biennial examination for the longitudinal study of diabetes. This examination included measurements of height,
weight, triceps skin-fold thickness, and chest and arm girths. A
detailed medical history, an electrocardiogram, and data on
smoking (if past or currently; if currently, then less than or more
than one pack per day) and on alcohol and coffee consumption
(numbers of drinks per day) were obtained. Percentage of desirable weight was computed with National Research Council Tables2'; additional criteria for obesity were weight (kg)/height
(m) (wt/ht), the body mass index (BMI) in kg/i2, and weight
(kg)/height (i3) (wt/ht3).
The Pimas have a high prevalence of diabetes mellitus,'7 a
factor known to influence plasma lipoprotein levels. For the
purpose of comparison with the LRC Prevalence Study, diabetics (179 male and 324 female subjects) were excluded from
this analysis. All subjects had an oral glucose tolerance test.
Diabetes was diagnosed if the 2 hr afterload (75 g carbohydrate)
plasma glucose level was at least 200 mg/dl22 at any survey
examination, or if the Indian Health Service Hospital- serving
the community found a fasting, postprandial, or 2 hr afterload
glucose concentration of at least 200 mg/dl. It should be noted
that the nondiabetic group included individuals with impaired
glucose tolerance as indicated by a 2 hr afterload glucose concentration of at least 140 mg/dl but less than 200 mg/dl (57 male
and 125 female subjects). Exclusion of these individuals had no
influence on the conclusions of this study.
In interpreting the data acquired from the Pimas, it is important to emphasize that the conclusions may not be representative
Vol. 68, No. 4, October 1983
of the total population, since not all members were sampled and
since diabetics were excluded. However, it has been previously
established that the selection methods for the longitudinal population study did not lead to errors in the estimation of diabetes
incidence. 23 The influence of diabetes and impaired glucose
tolerance on plasma lipids and lipoproteins in this population
will be examined in a separate report.
Lipid and lipoprotein measurements. Venous blood samples were collected in EDTA after an overnight fast. Plasma was
separated after centrifugation at approximately 700 g for 15 min
at 100 C. None of the samples contained evidence of chylomicra
after an overnight incubation at 40 C. Plasma was sampled for
measurement of total cholesterol and triglyceride concentrations. Lipoproteins were then isolated with procedures described by Havel et al.24 and modified by Marsh.25 Five milliliters of plasma was overlaid with 2 ml of 0. 16M NaCl, 1 mM
EDTA (EDTA-saline), d = 1.006 g/ml, and VLDL were isolated by ultracentrifugation for 16 hr at 40,000 rpm in a Beckman
ultracentrifuge with a type 40 rotor. It has been shown that
VLDL preparations isolated by migration through overlaid saline contain less than 10% albumin and other proteins.26 After
VLDL were removed, the plasma was adjusted to d = 1.063 g/
ml with NaBr (d = 1.35 g/ml), and LDL were isolated by
centrifugation for 20 hr at 40,000 rpm in a type 40.3 rotor. The
HDL fraction was the infranatant after removal of the LDL by
tube slicing. Efficiency of ultracentrifugal lipoprotein isolation
was monitored by electrophoresis on agarose gels according to
the method of Noble.27 Recovery of cholesterol in the lipoprotein fractions isolated by ultracentrifugation averaged 94%.
HDL subfractions were isolated from HDL in a portion of the
samples according to the method of Anderson et al.28 Samples
for HDL' subfractions were not selected by clinical status but
were determined largely by laboratory workload. Four milliliters of the HDL fraction was adjusted to d = 1. 105 g/ml and
overlaid with 2 ml of EDTA-saline (d = 1.100 g/ml), and
HDL2b were isolated after centrifugation for 26 hr at 50,000 rpm
in a 50-Ti rotor. After 1.5 ml of the supernatant that contained
HDL,h was removed, the infranatant was adjusted to d = 1.130
g/ml and overlaid with EDTA-saline (d = 1.125 g/ml), and
HDL2a were isolated by centrifugation for 48 hr at 50,000 rpm.
The infranatant after removal of HDL2a (1.5 ml) was assumed to
be'HDL3. Recovery of cholesterol in the HDL subfractions
averaged 96%.
Triglyceride and cholesterol concentrations in plasma and
isolated lipoproteins were quantified on an' AutoAnalyzer II
(Technicon) with the cholesterol extract method of Rush et al.29
and the triglyceride enzymatic method of Bucolo and Davis.30
For quantification of cholesterol levels in HDL subfractions the
sensitivity of the assay was increased by'diluting the sample
1:10 rather than 1:20 in isopropanol. For this procedure the
cholesterol standard was prepared in 90% rather than 95% isopropanol. The triglyceride and cholesterol assays and HDL isolation procedure were standardized with control plasma calibration pools supplied by the Lipid Standardization Laboratory,
Centers for Disease Control, Atlanta. The use of these controls
resulted in stabilization of the population measurements over
time and enabled direct comparison of the Pima data with the
published LRC norms. The coefficient of variation for the measurement of serum poo1s provided by the Centers for Disease
Control was 2.8% for high cholesterol concentrations, 3.6% for
low cholesterol concentrations, and 4.8% for triglyceride concentrations. Since our methods for lipoprotein isolation differed
from those of the LRC in that LDL in our study were isolated by
sequential ultracentrifugation, a series of 24 samples was quantified by both procedures. LDL cholesterol concentrations determined by ultracentrifugation according to the methods described above were comparable to those computed by the
715
HOWARD et al.
difference between LDL plus HDL and HDL precipitated by
heparin manganese (127 + 5.3 vs 127 ± 5.5 mg/dl; r = .93).
As an additional control to compare our methods with those of
the LRC, HDL isolated by ultracentrifugation were compared
periodically with that isolated by precipitation with heparin and
manganese, as performed by the LRC.31 The two methods were
comparable (47.1 ± 1.7 vs 48.3 + 2.2 mg/dl, r = .96).
Data analysis. Statistical analyses were performed with the
Statistical Analysis System, Cary, NC. Significance of differences between means was assessed by Student's t test. Relationships between two variables were assessed with Pearson correlation coefficients; for more than two variables partial
correlation analysis was used.
TABLE I
Characteristics of Pima nondiabetics
Age (yr)
n
Wt (kg)
PDWA
BMIB
Males
15-19
67
20-24
42
25-29
61
30-34
46
Results
35-44
47
Total and LDL cholesterol. Weight and weight-height
parameters of the Pima study subjects by age and sex
45-54
32
55-64
14
80C
(44-169)
89
(47-154)
103
(58-178)
100
(55- 160)
93
(62-136)
82
(56-118)
79
(48-108)
70
(44-103)
121
(73-258)
128
(76-210)
147
(83-235)
148
(88-232)
142
(104-197)
126
(89-173)
124
(82-145)
115
(75-166)
28
(17-59)
30
(18-48)
35
(19-54)
34
(20-53)
32
(24-45)
29
(20-40)
28
(19-36)
72
135
(85-219)
149
(90-236)
158
(85-232)
161
(103-257)
159
(96-228)
161
(98-242)
164
(91-240)
139
(99-182)
28
(18-46)
31
(19-49)
33
(18-49)
34
(22-54)
33
(20-48)
34
(20-52)
35
(20-51)
shown in table 1. Numbers of subjects of each sex
and age were proportional to the nondiabetic population census (data not shown). There is a high prevalence of obesity in this population,20 and thus there was
a wide range of weight in each age group. Means and
SEM for plasma and lipoprotein cholesterol and triglyceride concentrations for male and female subjects
by age group are shown in table 2.
Plasma cholesterol levels in Pimas are compared
with those for Caucasians in the LRC study in figure 1
(diabetics have not been excluded from the LRC population data, but less than 1% of that population was
judged to be diabetic either by fasting glucose levels or
medical history*). Plasma cholesterol levels in the
Pima men (figure 1) were significantly lower than
those of Caucasians for all ages above 30. In Pima
male subjects, total cholesterol levels were lower in
subjects 15 to 19 years old, but after age 20, in contrast
to the pattern in Caucasians, there was little change in
total plasma cholesterol levels with age. Total plasma
cholesterol levels in Pima female subjects increased
with age throughout the age ranges studied. No Pima
male subject and only one Pima female subject had a
plasma cholesterol concentration greater than 300 mg/
dl. When compared with Caucasian female subjects
(figure 1), all Pima female subjects except those 20 to
24 years old had significantly lower mean plasma cholesterol levels.
LDL cholesterol values for both male and female
Pimas followed patterns with age similar to those for
total cholesterol levels (table 2) and were significantly
lower than those of Caucasians in both men and women over 35 years old (data not shown). LDL cholesterol
levels for two age groups are shown in figure 2, which
compares the cumulative frequency distribution of
LDL cholesterol levels in Pima male and female subare
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
*Segal P: Personal communication.
716
3
65
Females
15-19
18
96
(45-137)
20-24
82
25-29
88
30-34
67
35-44
79
45-54
57
55-64
22
¢a 65
16
80
(45-137)
85
(44-143)
86
(54-134)
85
(51-119)
85
(50-126)
85
(44-125)
69
(48-100)
26
(17-40)
30
(21-38)
APercent desirable weight as determined by National Research Council tables.
BBody mass index measured as weight (kg)/height (m2).
CTop number is the mean with the range in parentheses.
jects with those of Caucasians in the LRC study. Differences between Pimas and Caucasians are greater
above the 50th percentile of LDL cholesterol levels.
HDL cholesterol. HDL cholesterol concentrations in
Pima male and female subjects also differed from those
of Caucasians in the LRC study (figure 3). Male Pimas
had lower HDL cholesterol levels than Caucasians, a
difference that was significant at ages 25 to 29 and 35
to 44 years. Differences between HDL cholesterol levels in Pimas and Caucasians were much greater in
female subjects (figure 3 right); mean HDL cholesterol
levels in Pima female subjects of all ages were significantly and substantially lower than those of the Caucasians (mean difference = 11 mg/dl).
When plasma total and HDL cholesterol levels of
CIRCULATION
PATHOPHYSIOLOGY AND NATURAL HISTORY-CORONARY ARTERY DISEASE
TABLE 2
Lipoprotein cholesterol and triglyceride levels (mg/dl) in Pima nondiabetics (mean ± SEM)
Cholesterol
Age (yr)
Triglyceride
Total
VLDL
LDL
HDL
Total
VLDL
LDL
155+3
179±5
181 ±4
183±4
185±5
190±6
184±8
182±9
lo+ 1
100±3
121 ±4
123±4
122±4
124±4
125±5
94±6
129± 11
134±8
153± 12
159± 14
61 ±5
84±8
124,±8
45 ±1
43±2
42± 1
43±2
41 ± 1
44±2
44±3
119±9
51 ±2
105± 14
29± !
38±3
39±2
46±3
47±4
42±3
44±4
34±3
99±2
103±2
44± 1
45 ±1
85±4
101 ±4
107±2
45 ±1
41 ± 1
109±5
121 ±6
41 ± 1
45 ±1
46±3
52±3
120±7
130±7
131±13
125± 16
Males
15-19
20-24
25-29
30-34
35-44
45-54
55-64
65
15±1
18± 1
19±2
21 ±2
21 ±2
17±3
12±2
150± 11
130±14
90+8
103 10
105 10
100± 10
80±11
55± 11
Females
10±o1
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
153±2
159±2
165±3
166±4
169±4
15-19
20-24
25-29
30-34
35-44
45-54
55-64
65
12± 1
13± 1
14± 1
182±4
188±9
207 o10
15± 1
16± 1
17±2
110±3
113±3
121 ±3
125±7
18±3
137+10
Males
_-_ Pimas
27± 1
60±4
65±4
73±5
68±5
75±6
78±11
72 14
32±
34± 1
38±2
39±2
42±2
44±3
43±4
study, the ratios of HDL/LDL and HDL/total cholesterol in Pima men were significantly lower than those
of Caucasian men between 20 to 29 years old, similar
to those of Caucasian men in the middle-aged groups,
and significantly greater in those over 65 years old. On
the other'hand, Pima female subjects had significantly
lower ratios of HDL/LDL and HDL/total cholesterol
than did Caucasian female subjects at each age group
below 55 years.
HDL subfractions of Pima men and women between
35 and 65 years old are shown in table 4. HDL3 accounted forr 67% to 70% of HDL in male subjects and
63% to 67% in female subjects. HDL2b, HDL2a, and
HDL3 concentrations were not significantly different in
Pima male and female subjects were compared (figure
4), women had significantly lower total cholesterol
levels than men between ages 20'and 44, but total
cholesterol levels in women exceeded those of men
after age 55. HDL cholesterol concentrations in Pima
male and female subjects were not significantly different except at 25 to 29 years old. LDL cholesterol
concentrations relationships between Pima male and
female subjects were similar to those for total cholesterol concentrations (table, 2).
The distribution of cholesterol among lipoprotein
classes was examined by computing ratios of HDL/
LDL and HDL/total cholesterol (table 3). When compared with ratios computed for Caucasians in the LRC
230
49±3
Femnales
0-0 LRC
220
210
200
compared with those of U.S. Caucasian populations
measured in the LRC Prevalence Study. Values are
mean ± SEM. *p < .05, **p < .01. ***p < .001.
The numbers of Pima individuals in each group are
shown in table 1.
I
190
1
FIGURE 1. Plasma total cholesterol levels in Pimas
Ii
A-
5
.
..
.
I
.5
t2-25-
0
19 24 29 U
35-
--I
45-
55-
54
64
I
>65
Age (yeavs)
Vol. 68, No. 4, October 1983
717
HOWARD et al.
55-64 yrs
Females
55-64 yrs
8060 40-
t)
t7 20-
l-
it11
100°°
'
-I
35-44 yrs
9806040-
*
20
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
r X *.w
Males
35-44 yrs
100 -
50
200
150
100
150
100
50
200
LDL Cholesterol (mg/dl)
FIGURE 2. Cumulative frequency distribution of LDL cholesterol
levels in Pimas vs Caucasians in the LRC study.
male and female subjects at any of the three decades
examined, indicating that Pima male and female subjects were similar in distribution of HDL subfractions
as well as in the total amount of HDL.
Triglyceride. Plasma triglyceride levels in Pimas are
compared with means for Caucasians in the LRC study
in figure 5. In Pima male subjects triglyceride levels
were significantly higher than those in Caucasian male
subjects in the lower age groups, but they were similar
to those for Caucasians after age 35. Plasma triglyceride levels of Pima female subjects were significantly
higher than the means for Caucasians in the LRC study
for all age groups except for those over 55 years.
Plasma triglyceride levels in both Pima male and female subjects increased with age up to 30. After age 30
triglyceride levels had little relationship with age in
women; in men over age 55 triglyceride levels were
lower with increasing age. No female subjects and
only two Pima male subjects had plasma triglyceride
concentrations over 400 mg/dl, and only eight male
and five female subjects had concentrations above 300
mg/dl. When plasma triglyceride levels of Pima male
and female subjects were compared (figure 4, C),
women had significantly lower triglyceride levels than
men between the ages of 20 and 45 years. VLDL
triglyceride levels in Pimas followed a relationship
with age and sex similar to that described for total
triglyceride levels (table 2).
Interrelationships among lipoprotein classes were
assessed by computing simple correlation coefficients.
Total triglyceride levels were strongly correlated with
VLDL triglyceride levels (r = .95, p < .0001). Total
cholesterol levels were strongly correlated with LDL
cholesterol levels (r - .92, p < .0001); HDL cholesterol levels were also related to total cholesterol levels
(r = .24, p < .0001), and negatively correlated with
plasma triglyceride levels (r - 19, p < .0001). The
relationship between HDL levels and total cholesterol
levels was significant even when corrected for triglyceride levels, age, and BMI (partial r
.32, p <
.0001).
Obesity. Since there is a high prevalence of obesity in
the Pimas,20 the relationships between obesity and the
various lipoprotein classes were examined. Simple
correlation coefficients between indices of obesity and
Males
60
-
so
Females
Pimas
o--c LRC
-
'O 55-
o 50
va
00O45
4
-C
a 40
z
-
Ir-
*
I
I
I
I
3544
4554
I
I
5564
>65
,
-
j~~~~..
46
*
I
15- 20- 25- 3019 24 29 34
~~
i~~~~~f-
,~~
I
I
I
I
15- 20- 25- 3019 242934
I
3544
I
1
5564
4554
>A
Age (years)
FIGURE 3. Plasma HDL cholesterol levels in Pimas vs Caucasians in the LRC study. Values are mean
< .01, ***p < .001.
718
+
SEM. *p < .05, **p
CIRCULATION
PATHOPHYSIOLOGY AND NATURAL HISTORY-CORONARY ARTERY DISEASE
lipoprotein distributions are shown in table 5. Neither
total cholesterol levels nor LDL cholesterol levels was
related to any of the indices of obesity in Pima male or
female subjects. On the other hand, HDL cholesterol
levels in both Pima male and female subjects were
significantly negatively correlated with all of the indices of obesity. This negative relationship between
HDL cholesterol levels and obesity (as measured by
BMI) was also significant when simultaneously adjusted for age, smoking, alcohol consumption, and VLDL
triglyceride concentration (partial r = - .22, p <
*-* Males
Females
210
A
o-c-
X
,I
200
I
E
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
E
190-
-6L..
;>
180-
170o
160
150Io
55
la
-
a-
0
co
45 -
0
40
0CS
_I
I
I
I
180 -
C
160 S
140-
X
120-
E
>. 100-
._
80
C1
I
I
I
I
15- 20- 25- 3019 24 29 34
I
I
I
3544
4554
5564
>65
Age (years)
FIGURE 4. A, Comparison of plasma total cholesterol levels. B, Comparison of HDL cholesterol levels. C, Comparison of plasma total
triglyceride levels. The numbers of subjects at each group are shown in
table 1. Values are mean + SEM. *p < .05, **p < .01, ***p < .001.
These comparisons are between Pima male and female subjects.
Vol. 68, No. 4, October 1983
.0001 for male subjects and r = - .33, p < .0001 for
female subjects). Plasma and VLDL triglyceride levels
in both sexes were positively correlated with obesity,
although the relationship between triglyceride levels
and obesity was not as strong as that between HDL
levels and obesity. The positive relationship between
VLDL triglyceride levels and BMI was significant in
male but not in female subjects after simultaneous adjustment for age, smoking, and alcohol consumption
(partial r = .19, p < .0007 for male subjects and r =
.06, NS for female subjects). Evidence that the various
anthropometric measurements are reflective of adiposity in the Pimas was obtained from studies of the
relationships among various indices of obesity in a
sample of 33 male and female nondiabetic Pima subjects admitted to a metabolic ward. BMI, wt/ht3, and
triceps skin-fold thickness were all closely related to
percent adiposity as determined by deuterium oxide
dilution (r = .67, .56, and .73, respectively). Wt/ht
and the sum of arm and chest girths were less related to
adiposity (r = .59 and r = .38, respectively). In most
cases the lipid parameters in table 5 followed the same
pattern. The exception was the lower correlations observed with triceps skin-fold thicknesses, especially in
men; this was probably because the skin-fold measurements were made by multiple examiners.
To assess whether obesity accounted for the differences in HDL levels between Pimas and Caucasians in
the LRC survey, a regression line was computed for
the relationship between HDL cholesterol levels and
BMI in Pimas. A decrease of approximately 0.4 mg/dl
was observed for each increment in BMI of 1.0 kg/m2
(y = - .42x + 56.83 in male subjects; y = - .44x +
58.5 in female subjects). A change of similar magnitude (i.e. , approximately 0.5 mg/dl per unit increment
in BMI - reported in ref. 32 as 0.1 [kg/cm2 x 1000]
increment in Quetelet Index) was observed in the Caucasians in the LRC study.32 When the values for HDL
cholesterol levels in Pimas and Caucasians in the LRC
study are compared (figure 3), it appears that if the
values for HDL cholesterol levels in Pimas were adjusted to median BMI values for the Caucasians in the
LRC group or vice versa, there would be minimal
differences between Pima and Caucasian male subjects; on the other hand, values for Pima female subjects would still be much lower than those for Caucasians. Thus, obesity may account for most of the
decreased HDL levels in Pima male subjects but not in
Pima female subjects. Similarly, obesity would not
appear to account for the lack of differences between
the sexes in plasma HDL cholesterol levels in Pimas
(see figure 4 and table 1).
719
HOWARD
et
al.
TABLE 3
Distribution of lipoprotein cholesterol in Pimas vs Caucasians in the LRC study (mean ± SEM)
HDL/LDL
Age (yr)
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
Males
15-19
20-24
25-29
30-34
35-44
45-54
55-64
> 65
Females
15-19
20-24
25-29
30-34
35-44
45-54
55-64
> 65
HDL/total
Pima
Caucasian
Pima
Caucasian
0.485+0.022
0.373 ± 0.020c
0.356±0.018B
0.414 ±0.023
0.531 ±0.015
0.478 ±0.019
0.419±0.012
0.386 ± 0.008
0.349 ± 0.006
0.330±0.005
0.365 ±0.009
0.368 ± 0.013
0.298±0.009
0.243 ± 0.01Oc
0.231 +0.009B
0.237 ±0.012
0.225 0.009
0.238±0.013
0.242 0.020
0.290 0.019B
0.307±0.005
0.287 ± 0.007
0.259±0.005
0.242 ±0.003
0.222 ± 0.002
0.214±0.002
0.234±0.004
0.239 + 0.005
0.595 ± 0.013
0.563±0.020
0.570±0.016
0.548 ± 0.013
0.495 ± 0.011
0.464±0.006
O.-289 0.006c
0.285±0.008c
0.275 ±0.007c
0.254 0.007c
0.249 0.007c
0.254 +0.008B
0.251 ±0.015
0.262+0.021
0.334 ±0.005
0.324±0.008
0.329±0.006
0.326 ± 0.005
0.297 +0.004
0.281 ±0.004
0.262±0.005
0.259±0.005
0.347 ± 0.017
0.380±0.030
0.371 ± 0.037
0.470± 0.044A
0.459 ± 0.014c
0.459±0.018c
0.436±0.015C
0.391 ± 0.014c
0.388 ± 0.017c
0.392 + 0.016c
0.390+0.030
0.414±0.042
0.414+0.010
0.404±0.013
Statistical comparisons (t test for Pimas vs Caucasians): Ap < .05; Bp < .01; Cp < .001.
Sex hormone use and pregnancy. Since sex hormone
use has been previously shown to influence plasma
lipoprotein distributions, '0 plasma lipoproteins in Pima
female subjects currently taking oral contraceptives
were compared with those of nonpregnant female subjects of similar ages who were not taking contraceptives (table 6). BMIs were not significantly different in
these two groups. Pima female subjects taking oral
contraceptives had higher HDL cholesterol levels. In
subjects 15 to 24 years old the group on oral contraceptives also had higher triglyceride levels and somewhat,
but not significantly, higher total cholesterol levels.
TABLE 4
HDL subfractions (mg/dl) in Pima men and women (mean
Age
(yr)
Males
35-44
(n = 23)
45-54
(n = 15)
55-64
(n
=
+
SEM)
HDL2b
HDL2a
HDL3
HDL3 as
% of HDL
4.6±1.1
7.7 ± 0.9
28.6±1.3
70%
4.6±+1.0
8.2±+1.5
29.8± 1.5
70%
5.3±+ 1.3
7.9±+ 1.6
26.5 ± 1.2
67%
5.9 ± 0.6
8.9 ± 0.5
24.7±+ 0.8
63%
5.0+0.6
8.8+0.6
26.4± 1.0
66%
4.8+0.4
9.1±0.4
28.3±+1.2
67%
9)
Females
3544
(n = 43)
45-54
(n = 39)
55-64
(n = 16)
720
The influence of pregnancy in plasma lipid levels in
this group was also examined (table 6). Pregnant Pima
female subjects had higher total and HDL cholesterol
levels and higher total and VLDL triglyceride levels
than nonpregnant female subjects of comparable ages.
Discussion
Total and LDL cholesterol. The Pima population has
afforded an opportunity for the examination of plasma
lipoprotein cholesterol and triglyceride levels in a
well-characterized homogeneous racial group. The results indicate that there are distinct characteristics of
their lipoprotein distributions. Despite the high prevalence of obesity among the Pimas,20 plasma total and
LDL cholesterol concentrations in men over 30 years
old and in women over 25 years old were lower than
those of Caucasians in the LRC study. This lower
plasma cholesterol level had been suggested previously when serum cholesterol levels in the Pimas were
compared with those of the population of Tecumseh,
MI.L8 The present study compares the Pima total and
lipoprotein cholesterol concentrations with those of
Caucasians in the LRC study by use of similarly standardized methods for measurement, and it also establishes that the lower total cholesterol level is mainly
the result of lower levels of plasma LDL. The lower
cholesterol levels in adult Pima men compared with
those in adult Caucasian men appear to be related to the
observation that there is little change with age after age
20 in Pimas, whereas in Caucasian male subjects choCIRCULATION
PATHOPHYSIOLOGY AND NATURAL HISTORY-CORONARY ARTERY DISEASE
Males
Females
180
*-@ Pimas
o--e LRC
0
, 160-
E
140
-
0
._
o 120-
'a
1._
100-
E
X 80E60J
*
I
*** ***
I
I
**
*
I
15- 20- 25- 3019 24 29 34
I
I
3544
4554
I
I
>65
5564
I
*
-
-
I
I
15- 20- 25- 3019 24 29 34
I
3544
I
4554
I
5564
1
>65
Age (years)
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
FIGURE 5. Plasma total triglyceride levels in Pimas compared with those of Caucasians in the LRC study. Values are mean ±
SEM. *p < .05, **p < .01, ***p < .001.
lesterol levels are positively associated with age. In
contrast, in female subjects cholesterol levels increase
with age in both racial groups, but in each age group
the levels are lower in the Pimas.
HDL cholesterol. Another difference of lipoproteins
in the Pimas compared with those of Caucasians in the
LRC study is that male and female subjects had similar
concentrations of HDL cholesterol. In male Pimas,
HDL cholesterol levels were slightly lower than those
of Caucasians; in female Pimas, HDL cholesterol levels averaged more than 11 mg/dl lower than the LRC
means for female Caucasians. Although HDL cholesterol levels were negatively associated with obesity in
both male and female Pimas, the lack of differences
between the sexes in HDL cholesterol levels did not
appear to be attributable to obesity (see below). Distributions of HDL subfractions were also similar in male
and female Pimas, with no increase in HDL2 levels in
female subjects of the magnitude reported in other
studies.33 34
There have been other reports of HDL cholesterol
levels in different populations, which suggest that differences between the sexes in HDL cholesterol levels
may be most prominent in Caucasians. Data from the
studies of blacks in the LRC populations11 and in Africa'2 indicate that HDL cholesterol concentrations in
black male subjects are higher than in Caucasian male
subjects and similar to those of black female subjects.
Studies of Polynesian groups",'1 have shown plasma
cholesterol values that resemble those of the Pimas
with relatively low HDL cholesterol levels and similar
values for HDL in male and female subjects. Finally,
the Tarahumara Indians, who have plasma cholesterol
Vol. 68, No. 4, October 1983
levels even lower than those of the Pimas, also have
no differences between the sexes in HDL cholesterol
levels. 15
-Total and VLDL triglycerides. Although total and LDL
cholesterol levels were low, total and VLDL triglyceride concentrations in Pimas were not decreased. Recent studies of VLDL and LDL turnover in this population have shown that approximately 50% of the VLDL
apoprotein B is metabolized through an alternate pathway and is not converted to LDL.36 This suggests a
metabolic mechanism for the maintenance of low LDL
in this population despite the presence of adequate or
even elevated amounts of its precursor.
Factors contributing to plasma lipid concentrations in
Pimas
Obesity. The Pimas, because of the high prevalence
of obesity, afforded us an unusual opportunity to assess
the relationships among obesity and plasma lipoproteins. There was a strong negative correlation between
obesity and HDL cholesterol levels, with decreases of
approximately 0.4 mg/dl of HDL cholesterol with each
kg/m2 increment of BMI. This inverse relationship between HDL levels and obesity has been observed previously,32 37 38 although this is the first report in a population with such a high prevalence and wide range of
obesity. Triglyceride levels were positively correlated
with obesity in the Pimas, but not as strongly as reported previously in other groups;37-39 increases in triglyceride levels with obesity in the Pimas were small (I to 2
mg/dl per kg/m2 increase in BMI), and there was little
evidence of the hypertriglyceridemia often associated
with obesity in other populations. Total and LDL cho721
HOWARD et al.
TABLE 5
Pearson (simple) correlation coefficients between lipoproteins and indices of obesity in the Pima
Males (n = 329)
Cholesterol
Total
LDL
HDL
HDL2b (n
=
67)
HDL,a (n
=
67)
HDL3 (n = 67)
Triglyceride
BMIA
Wt/Ht
Wt/Ht3
TricepsB
Girthc
.06
.10
-.25
(.0001)D
-.25
(.04)
-.32
(.08)
.08
.05
.09
-.25
(.0001)
- .26
(.04)
-.33
(.06)
-.07
.08
.11
-.25
- .01
.05
.08
-.24
(.0001)
- .23
.17
(.002)
.19
(.0007)
(.008)
.16
(.004)
(.0004)
.21
(.0001)
-.04
.04
-.30
(.0001)
-.27
-.02
.06
-.29
(.0001)
-.26
(.002)
(.004)
-.22
(.01)
-.23
-.22
-.09
Total
VLDL
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
Females (n = 507)
Cholesterol
Total
LDL
HDL
HDL2b (n
HDL2a (n
=
=
- .03
129)
.05
-.30
(.0001)
-.27
(.003)
-.23
129)
-.22
129)
.15
-.24
(.009)
HDL3 (n
=
Triglyceride
ABoy mass index
(.01)
(.01)
.17
.16
(.0002)
Total
VLDL
=
(.007)
-.22
(.0003)
(.0001)
.00
-.14
(.02)
-.24
(.05)
-.08
-.30
(.01)
.04
-.28
(.03)
.08
-.22
- .10
.20
.06
.17
.08
(.003)
.17
(.005)
-.00
.06
-.27
(.0001)
-.21
(.02)
(.01)
- .02
.04
-.27
(.0001)
-.22
(.02)
-.23
(.01)
-.15
(.01)
.17
(.0001)
.11
.11
.12
(.01)
(.02)
(.009)
.18
(.0001)
.15
(.002)
.18
(.0001)
.13
(.005)
wt/ht2.
"Triceps skin-fold thickness.
cEqual to the sum of arm and chest circumferences.
0p value for those of significance
< .05.
lesterol levels had little relationship with obesity in the
Pimas, in contrast to those in other studies.37 39 The
data on the Pima may help to distinguish among
mechanisms of changes in lipoproteins induced by
obesity. It has been suggested that obesity initially
leads to increased levels of VLDL; this in turn leads to
increased levels of LDL, the catabolic product of
VLDL, and to lower levels of HDL because of the
inverse relationship of HDL with VLDL, which was
found in the Pimas and in other populations. In the
Pimas, however, VLDL triglyceride levels were only
slightly increased with increasing obesity, and LDL
cholesterol levels were not related to obesity. Previous
studies of VLDL metabolism in the Pimas confirm a
lack of significant overproduction of VLDL triglycerides with increasing weight.40 Thus, the data from the
722
Pima suggest that lower HDL concentrations in obesity
may be independent of increases in VLDL concentrations or metabolism; this is supported by the significant
relationship determined by partial correlation analysis
that included VLDL triglyceride levels.
Diet. It is difficult to determine whether differences
in lipoprotein patterns observed in the Pimas as compared with those of Caucasians in the LRC study may
be accounted for by environmental factors such as diet.
The high triglyceride levels in the younger Pima
groups and the somewhat lower HDL cholesterol levels in Pima male subjects may be related to increased
caloric intake or reduced physical activity. The lower
total and LDL cholesterol levels, however, do not appear to be related to dietary factors. A study of eating
habits in women of this population 12 years ago indiCIRCULATION
PATHOPHYSIOLOGY AND NATURAL HISTORY-CORONARY
TABLE 6
Effect of oral contraceptives and pregnancy on plasma lipoproteins
in Pima female subjects (mean + SEM)
Sex hormone use/pregnant
15-24 yr
N
Age
BMI (kg/M2)
Total cholesterol
(mg/dl)
No/no
Yes/no
No/yes
184
19.5 ± 0.2
29.8 ± 0.4
15
20.6 ± 0.7
29.0 ± 2.0
11
19.9 ± 0.8
27.9 ± 1.7
156± 2
169 ± 7
197 ± 12c
HDL cholesterol
(mg/dl)
Plasma triglyceride
(mg/dl)
VLDL triglyceride
(mg/dl)
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
25-34 yr
N
Age
BMI (kg/m2)
Total cholesterol
(mg/dl)
HDL cholesterol
(mg/dl)
44 ± 1
50 ± 2B
55 ± 4B
93 ± 3
144 + 17B
157 ± 17C
55 ± 2
78 ± I 1A
79±lOA
155
29.5 + 0.2
33.5+0.6
9
29.5 ± 0.8
31.6±2.2
11
29.2 ± 0.6
34.3±1.0
165+2
162±1
195±1B
43 ± 1
50±4A
49 ± 3A
114±4
112± 10
189± 17C
68 ± 3
52 ± 6
105 ± 13c
Plasma triglyceride
(mg/dl)
VLDL triglyceride
(mg/dl)
Statistical comparisons (t test vs group of women neither pregnant nor
taking sex hormones): Ap < .5; Bp < .01; Cp < .001.
cated that caloric distribution, daily cholesterol intake,
and amount of polyunsaturated fat consumption were
similar to those of the general U.S. population." More
recent inquiries of diet habits of the Pima population
confirm the earlier findings of consumption of large
amounts of saturated fat.*
Racial differences. Although some aspects of plasma
lipoprotein distributions in the Pimas may be attributable to obesity or diet, others do not appear to be
related to identifiable environmental factors. Although
diet habits in the Pimas are not strongly different from
other U.S. groups, there are many other social, environmental, and lifestyle factors that were not measured
in this study, which may account for the observed
lipoprotein distributions. Alternatively, there may be
genetic factors that control lipoprotein cholesterol and
triglyceride levels. Further comparisons of lipoprotein
levels in various racial groups coupled with studies of
families within these groups will be necessary to understand how genetic and environmental factors interact to influence plasma lipoprotein levels.
*Hendricks E: Unpublished data.
Vol. 68, No. 4, October 1983
ARTERY DISEASE
Sex hormones and pregnancy. Although the number of
Pima women using oral contraceptives was small, the
data from the Pima corresponded to those acquired
from other populations on the influence of sex hormone use on plasma lipoprotein levels. '0 Pima female
subjects taking oral contraceptives had higher HDL
cholesterol levels; higher mean values for total cholesterol and triglyceride levels were also observed. The
influence of pregnancy was also similar to that reported previously,42' 3 with large increases in total and
HDL cholesterol levels and triglyceride levels.
Cardiovascular disease. The occurrence of cardiovascular disease in relationship to lipoprotein levels is an
interesting question for future study in this population.
Myocardial infarctions are infrequent in Southwestern
Indians' and data from the Pimas indicate that the
prevalence of ECG abnormalities (as defined by the
Tecumseh criteria) is approximately half of that for
Caucasians in Tecumseh. 18 The relatively low total and
LDL cholesterol levels in the Pimas are consistent with
the lower prevalence of cardiovascular disease; on the
other hand, HDL cholesterol levels in Pimas are low
The observed lack of differences between the sexes in
HDL cholesterol levels in Pimas and other groups also
suggest that increased HDL levels may not be a major
explanation for the lower prevalence of cardiovascular
disease in female subjects, since the prevalence of
cardiovascular disease is lower in female Pima subjects compared with that of male Pima subjects,'8 as
well as in other groups. An analysis of cardiovascular
disease risk factors and a prospective study of the relationship between plasma lipoprotein levels and cardiovascular disease are planned in this population.
We are grateful for the cooperation of the members of the
Gila River Indian Community, for the efforts of Dr. J. R. Lisse,
Dr. M. J. Carraher, and the laboratory and secretarial staffs of
the Southwestern Field Studies Section and Phoenix Clinical
Research Section,- for the expert technical assistance of Kim
Sizemore and Annette Kennedy, and for the secretarial assistance of Verna Kuwanhoyioma. We also appreciate the advice
of Dr. P. Segal and Dr. B. Rifkind in the preparation of the
manuscript, and we are grateful to Dr. G. Heiss for supplying
unpublished LRC population data on lipoprotein cholesterol
ratios and to Dr. R. Daniels for unpublished data on adiposity in
the Pimas. We also appreciate the advice of Dr. D. Anderson on
techniques of HDL subfractionation.
References
1. Kannel WB, Castelli WP, Gordon T, McNamara PM: Serum cholesterol, lipoproteins, and the risk of coronary heart disease. The
Framingham study. Ann Intern Med 74: 1, 1971
2. Pooling Project Research Group: Relationship of blood pressure,
serum cholesterol, smoking habit, relative weight and ECG abnormalities to incidence of major coronary events: final report of the
pooling project. J Chronic Dis 31: 201, 1978
3. Keys A: Seven countries: a multivariate analysis of death and
723
HOWARD et al.
4.
5.
6.
7.
8.
9.
10.
11.
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
724
coronary heart disease. Cambridge, MA, 1980, Harvard University Press
Albrink MJ, Man EB: Serum triglycerides in coronary artery disease. Arch Intern Med 103: 4, 1959
Hulley SB, Rosenman RH, Bawol RD, Brand RJ: Epidemiology as
a guide to clinical decisions. The association between triglyceride
and coronary heart disease. N Engl J Med 302: 1383, 1980
Lippel K, Tyroler H, Eder H, Gotto A, Vahoriny G: Relationship
of hypertriglyceridemia to atherosclerosis. Arteriosclerosis 1: 406,
1981
Barr DP, Russ EM, Eder HA: Protein-lipid relationships in human
plasma. II. In Atherosclerosis and related conditions. Am J Med
11: 480, 1951
Miller GJ, Miller NE: Plasma high density lipoprotein concentration and development of ischaemic heart disease. Lancet 1: 16,
1975
Heiss G, Tyroler H: Epidemiology of high density lipoprotein a
review. In Lippel K, editor: Report of the High Density Lipoprotein Methodology Workshop. Bethesda, 1979, NIH publication
No. 79-1661, pp 7-28
The Lipid Research Clinics Population Studies Data Book. Volume
I. The prevalence study. Bethesda, 1980, NIH publication No. 801527
Tyroler HA, Glueck CJ, Christensen B. Keviterovic PD: Plasma
high density lipoprotein cholesterol. Comparisons in black and
white populations. Circulation 62 (suppl IV): IV-99, 1980
Ononogbu IC: Comparison of high density lipoprotein and serum
cholesterol levels in a European and an African community.
Atherosclerosis 34: 49, 1979
Stanhope JM, Sampson VM: High-density-lipoprotein cholesterol
and other serum lipids in a New Zealand biracial adolescent sample. The Wairoa College Survey. Lancet 1: 968, 1977
Rhoads GG, Gulbrandsen CL, Kagan A: Serum lipoproteins and
coronary heart disease in a population study of Hawaii Japanese
men. N Engl J Med 294: 293, 1976
Connor WE, Cequeira MT, ConnorRW, Wallace RB, Malinow R,
Casdorph HR: The plasma lipids, lipoproteins and diet of the Tarahumara Indians of Mexico. Am J Clin Nutr 31: 1131, 1978
Williams RC, Knowler WC, Butler WJ, Pettitt DJ, Lisse JR, Bennett PH, Mann DL, Johnson AH, Terasaki PI: HLA-A2 and Type 2
(insulin-independent) diabetes mellitus in Pima Indians: an association of allele frequency with age. Diabetologia 21: 460, 1981
Bennett PH, Rushforth NB, Miller M, LeCompte P: Epidemiological studies of diabetes in the Pima Indians. Recent Prog Horm-Res
32: 333, 1976
Ingelfinger JA, Bennett PH, Liebow IM, Miller M: Coronary heart
disease in the Pima Indians. Electrocardiographic findings and
postmortem evidence of myocardial infarction in a population with
a high prevalence of diabetes mellitus. Diabetes 25: 561, 1976
Savage PJ, Hamman RF, Bartha G, Dippe SE, Miller M, Bennett
PH: Serum cholesterol levels in American (Pima) Indian children
and adolescents. Pediatrics 52: 274, 1976
Knowler WC, Pettitt DJ, Savage PJ, Bennett PH: Diabetes incidence in Pima Indians: contributions of obesity and parental diabetes. Am J Epidemiol 113: 144, 1981
National Research Council: National Academy of Sciences Publication No. 1146, Washington DC, 1964, p 4
Second Report.
WHO Expert Committee on Diabetes Mellitus
Geneva, 1980, World Health Organization Technical Report Series
No. 646, pp 8-14
Knowler WC, Bennett PH, Hamman RF, Miller M: Diabetes incidence and prevalence in Pima Indians: a 19-fold greater incidence
than in Rochester, Minnesota. Am J Epidemiol 108: 497, 1978
Havel RJ, Eder HA, Bragden JH: Distribution and chemical composition of ultracentrifugally separated lipoproteins in human serum. J Clin Invest 34: 1345, 1955
25. Marsh JB: Lipoproteins in a nonrecirculating perfusate of rat liver.
J Lipid Res 15: 544, 1974
26. Schonfeld G, Birg G, Miller JP, Kessler G, Santiago J: Apolipoprotein B levels and altered lipoprotein composition in diabetes.
Diabetes 23: 827, 1974
27. Noble RP: Electrophoretic separation of plasma lipoproteins in
agarose gel. J Lipid Res 9: 693, 1968
28. Anderson DW, Nichols AV, Pan SS, Lindgren FT: High density
lipoprotein distribution. Resolution and determination of three major components in a normal population sample. Atherosclerosis 29:
161, 1978
29. Rush RL, Leon L, Turrell J: Automated simultaneous cholesterol
and triglyceride determination on the Auto Analyzer instrument. In
Advances in Automated Analyses. Miami, 1970, Thurman, pp 503
30. Bucolo G, Davis H: Quantitative determination of serum triglycerides by use of enzymes. Clin Chem 19: 475, 1973
31. Warmick GR, Albers JJ: A comprehensive evaluation of the
heparin-manganese precipitation procedure for estimating high
density lipoprotein cholesterol. J Lipid Res 19: 65, 1978
32. Glueck CJ, Taylor HL, Jacobs D, Morrison JA, Beaglehole R,
Williams OD: Plasma high-density lipoprotein cholesterol: association with measurements of body mass. The Lipid Research Clinics
Program Prevalence Study. Circulation 62 (suppl IV): IV-62, 1980
33. Eder HA: The rationale for measurement of HDL subclasses. In
Lippel K, editor: Report of the high density lipoprotein methodology workshop. Bethesda, 1979, NIH publication no. 97-1661, pp
279
34. Cheung MC, Albers JJ: The measurement of apolipoprotein A-I
and A-Il levels in men and women by immunoassay. J Clin Invest
60: 43, 1977
35. Nestel PJ, Zimmet P: HDL levels in Pacific Islanders. Atherosclerosis 40: 257, 1981
36. Egusa G, Beltz WF, Howard BV: Incomplete conversion of very
low density lipoprotein apolipoprotein B to low density lipoprotein
apolipoprotein B: a mechanism for the maintenance of low plasma
cholesterol. Circulation 66: 961, 1982 (abst)
37. Albrink MJ, Krauss RM, Lindgren FT, Von Der Groeben J, PanS,
Wood PD: Intercorrelations among plasma high density lipoprotein, obesity and triglycerides in a normal population. Lipids 15:
668, 1980
38. Phillips NR, Havel RJ, Kane JP: Levels and interrelationships of
serum and lipoprotein cholesterol and triglycerides. Association
with adiposity and the consumption of ethanol, tobacco, and beverages containing caffeine. Arteriosclerosis 1: 13, 1981
39. Carlson LA, Ericsson M: Quantitative and qualitative serum lipoprotein analysis. I. Studies in healthy men and women. Atherosclerosis 21: 417, 1975
40. Howard BV, Zech L, Davis M, Bennion LJ, Savage PJ, Nagulesparan M, Bilheimer D, Bennett PH, Grundy SM: Studies of very
low density lipoprotein triglyceride metabolism in an obese population with low plasma lipids: lack of influence of body weight or
plasma insulin. J Lipid Res 21: 1032, 1980
41. Reid JM, Fullmer SD, Pettigrew KD, Burch TA, Bennett PH,
Miller M, Whedon GD: Nutrient intake of Pima Indian women:
relationships to diabetes mellitus and gallbladder disease. Am J
Clin Nutr 24: 1281, 1971
42. Gillett MPT, Silva TB, Besterman EMM: Plasma lipids and HDL
cholesterol during pregnancy. Atherosclerosis 33: 495, 1979
43. Knoop RH, Warth MR, Carrol CJ: Lipid metabolism in pregnancy.
I. Changes in lipoprotein triglyceride and cholesterol in normal
pregnancy and the effects of diabetes mellitus. J Reprod Med 10:
95, 1973
44. Sievers ML: Myocardial infarction among Southwestern American
Indians. Ann Intern Med 67: 800, 1967
CIRCULATION
Plasma and lipoprotein cholesterol and triglyceride concentrations in the Pima Indians:
distributions differing from those of Caucasians.
B V Howard, M P Davis, D J Pettitt, W C Knowler and P H Bennett
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
Circulation. 1983;68:714-724
doi: 10.1161/01.CIR.68.4.714
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1983 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/68/4/714
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/