American Journal of Epidemiology
Copyright © 2000 by The Johns Hopkins University School of Hygiene and Public Health
All rights reserved
Vol. 151, No. 9
Printed in USA.
Relation of Abdominal Height to Cardiovascular Risk Factors in Young Adults
The Bogalusa Heart Study
Jeanette Gustat, Abdalla Elkasabany, Sathanur Srinivasan, and Gerald S. Berenson
Obesity and fat patterns are important predictors of coronary heart disease risk. The relations of abdominal
height (sagittal diameter) and various obesity measures to coronary heart disease risk factors were examined
in a community-based sample of 409 Blacks and 1,011 Whites aged 20-38 years in Bogalusa, Louisiana
(1995-1996). Obesity measures used included weight, waist circumference, waist:hip ratio, waist:height ratio,
abdominal height, triceps and subscapular skinfold thicknesses, body mass index, and conicity index. Abdominal
height was highly correlated with other obesity measures, especially waist circumference (0.937-0.944, p <
0.001), and was least correlated with height. In multivariate analysis, abdominal height was an independent
predictor of levels of total cholesterol, triglycerides, very low density lipoprotein cholesterol, low density
lipoprotein cholesterol, high density lipoprotein cholesterol, glucose, and insulin and of systolic and diastolic
blood pressures (p < 0.05 to p < 0.001), with total R2 values ranging from 0.13 to 0.52. Abdominal height
contributed more to the prediction of blood pressure than did other measures of central obesity. In canonical
analysis, abdominal height was correlated more strongly with the coronary disease risk factor variables as a
group than were other obesity measures. These results suggest that abdominal height adds another dimension
to measures of obesity in that it may help to assess a component of visceral fat that other measures miss. Am
J Epidemiol 2000;151:885-91.
abdomen; adipose tissue; body mass index; obesity; risk factors
Obesity and the distribution of body fat are important
predictors of coronary heart disease morbidity and
mortality (1-6). Recently, obesity was identified as an
independent risk factor for coronary disease by the
American Heart Association (7). Obesity, especially
central body fatness, is associated with coronary disease risk factors, including dyslipidemia, hyperinsulinemia/insulin resistance, and hypertension (8-11). In
epidemiologic studies, various anthropometric measures, including body mass index (weight (kg)/height
(m)2), subcutaneous skinfold thicknesses, and body circumferential measurements, are used to assess adiposity and fat distribution and their relation to coronary
disease risk factors. Besides waist:hip ratio, abdominal
height (sagittal diameter) and conicity index (abdominal
girth (m)/[0.109Vweight (kg)/height (m)]) are considered indirect measures of visceral fat distribution
(12-14). The present community-based study examined
the relation between abdominal height (as opposed to
other obesity measures) and cardiovascular risk factors.
MATERIALS AND METHODS
Population
The Bogalusa Heart Study is a long term epidemiologic study of cardiovascular risk factors from birth
through age 38 years in a biracial population (65 percent White and 35 percent Black) in Bogalusa,
Louisiana (15, 16). During 1995-1996, 1,420 subjects
aged 20-38 years were examined as part of the posthigh school cohort study. The study population was 71
percent White and 61 percent female.
General examinations and anthropometry
Trained examiners collected all data under rigid
protocols (15). Subjects were instructed to fast for 12
hours before venipuncture, and compliance was
assessed by interview on the morning of the screening. Screening consisted of venipuncture followed by
replicate measurements of abdominal height, height,
weight, triceps and subscapular skinfold thicknesses,
and waist and hip circumferences. Replicate measures
were conducted by the same examiner, and the average of the measurements was used. Abdominal
height, defined as the thickness of the abdomen at
waist level, was measured with a portable sliding-
Received for publication January 11,1999, and accepted for publication May 6, 1999.
Abbreviation: CAT, computerized axial tomography.
From the Tulane Center for Cardiovascular Health, Tulane
University Medical Center, New Orleans, LA.
885
886
Gustat et al.
beam abdominal caliper while the subject lay supine
on an examining table (Holtain-Kahn Abdominal
Caliper; Holtain Ltd., Dyfed, Wales). With the base of
the caliper placed under the subject's back, the
caliper's sliding beam was brought down to a point
midway between the iliac crests. The subject was
asked to inhale and exhale naturally; the sliding beam
was then moved so that it slightly touched the subject's abdomen without compression. In the supine
position, the body's visceral fat projects the abdomen
in a sagittal direction, and gravity moves the subcutaneous fat to the sides (17).
Replicate systolic and diastolic (fourth and fifth
Korotkoff phases) blood pressure levels were measured. The fifth phase measurement was used in this
analysis. The average of six readings, three each taken
by two trained observers, was used for analysis.
Laboratory analyses
Serum total cholesterol and triglyceride levels were
measured by enzymatic procedures in an Abbott VP
Analyzer (Abbott Laboratories, North Chicago,
Illinois) (18, 19). Lipoprotein cholesterol levels were
measured by a combination of heparin-calcium precipitation and agar-agarose gel electrophoresis (20).
Measurements of total cholesterol, triglycerides, and
high density lipoprotein cholesterol in the Bogalusa
Heart Study are being continuously monitored through
a surveillance program of the Centers for Disease
Control and Prevention (Atlanta, Georgia).
Plasma glucose level was measured as part of a multiple chemistry profile (SMA 20) by a glucose oxidase
method; plasma immunoreactive insulin level was
measured using a radioimmunoassay procedure, with
the Phadebas insulin kit (Pharmacia Diagnostics,
Piscataway, New Jersey).
Statistical analyses
All analyses were conducted using SAS software
(21). Descriptive statistics for measured variables
were calculated for each race-sex group. The logtransformed values of the triglyceride and insulin
measures were used in all but the descriptive statistics. Analysis of variance was used to test for differences and interactions between race-sex groups.
Pearson correlations were used to examine the associations between abdominal height and other measures
of obesity, as well as the associations between measures of central obesity (waist circumference, conicity
index, waistrhip ratio, and waist:height ratio) and the
cardiovascular risk factor variables. Fisher's Z transformation and the z test were used to test for differences between correlations.
To determine the contribution of abdominal height
to the variance of cardiovascular risk factor variables,
we developed linear multivariate prediction models
using PROC GLM in SAS. Variables were forced into
the model. Models were developed for each cardiovascular risk factor with serum total cholesterol, fasting
serum triglycerides, fasting very low density lipoprotein cholesterol, low density lipoprotein cholesterol,
high density lipoprotein cholesterol, systolic and diastolic blood pressure, plasma glucose, and insulin each
used as a dependent variable. Partial R2's were calculated for the contribution of the central obesity variables as a unit and the contribution of abdominal
height. The full models included race, sex, age,
abdominal height, body mass index, conicity index,
waist:hip ratio, waist:height ratio, waist circumference, triceps and subscapular skinfold thicknesses, and
weight as the independent variables. In addition to the
full model, two reduced models were used. The
reduced model for estimating the partial R2's for
abdominal height included all variables but abdominal
height. The reduced model for estimating the partial
/?2's for the central obesity measures included all variables but the central obesity measures. The partial R2
values are equal to the increase in R2 from the reduced
model to the full model relative to the proportion of
variance unexplained by the reduced model.
Canonical correlation analysis, a multivariate technique, was used to examine the relations between the
obesity measures and each of three sets of risk variables: lipoprotein variables, blood pressure variables,
and glucose and insulin levels. This procedure allows
the simultaneous examination of predictor and criterion (outcome) variables (22). The analysis produces a
"canonical variable" expressed as a correlation. With
canonical analysis, it is also possible to examine the
correlation of each of the variables within the group
(predictor variables, such as the obesity measures) to
the grouped criterion variables (outcome variables,
such as the lipoprotein variables). The hypothesis that
the canonical correlations are zero in the population is
tested. Only the correlations with the first significant
canonical variable were examined.
RESULTS
As table 1 shows, Blacks had a higher body weight,
body mass index, abdominal height, subscapular skinfold thickness, and hip circumference than did Whites
(p < 0.05). Females, overall, were slightly younger
and shorter than males and had a higher triceps skinfold thickness than males (p < 0.05). There were no
race or sex interactions with the obesity measures.
Pearson correlations between abdominal height and
other obesity measures are shown in figure 1 for the
Am J Epidemiol Vol. 151, No. 9, 2000
Abdominal Height and Cardiovascular Risk Factors
887
TABLE 1. Mean ages and levels of obesity measures, by race and sex: Bogalusa Heart Study,
1995-1996*
Age (years)t
Height (cm)t
Weight (kg)§
Triceps skinfold (mm)t
Subscapular skinfold (mm)§
Waist circumference (cm)
Hip circumference (cm)§
Abdominal height (cm)§
Body mass indexH,§
Waist:hip ratio
Waist:height ratio
Conicity index
White males
(n = 398)
Black males
(n=152)
White females
(n = 613)
Black females
(n = 257)
30.2 (5.0)$
177.1 (6.5)
86.9(17.3)
18.3(9.8)
21.5 (10.7)
94.2 (14.2)
105.7(10.1)
21.7(3.5)
27.7 (5.3)
0.89 (0.08)
0.53 (0.08)
1.23(0.10)
30.0 (5.4)
176.1 (6.4)
86.7(24.1)
15.6(12.0)
21.2(14.0)
90.4 (17.9)
105.7(14.0)
22.0 (4.5)
27.9 (7.4)
0.85 (0.07)
0.51 (0.10)
1.19(0.09)
29.5 (5.0)
163.1 (6.4)
69.6(18.5)
27.7(10.6)
24.3(12.4)
80.4 (14.8)
104.4(13.4)
19.5 (3.9)
26.2 (6.8)
0.77 (0.06)
0.49 (0.09)
1.13(0.08)
28.8 (5.2)
162.9(6.9)
81.2 (24.2)
30.2 (13.3)
30.1 (13.9)
89.5(17.9)
111.6(15.4)
22.8 (4.5)
30.5 (8.6)
0.80 (0.07)
0.55(0.11)
1.17(0.09)
* Tests of differences with the obesity measures were adjusted for age.
t Sex difference: p < 0.05.
i Numbers in parentheses, standard deviation.
H Weight (kg)/height (m)2.
§ Race difference: p < 0.05.
different race-sex groups. All correlations except height
were highly significant (p < 0.001). Waist circumference was the strongest correlate of abdominal height in
every race-sex grouping. Correlations ranged from
0.937 for White males to 0.944 for White females. The
next highest correlations were noted for body mass
index and waist:height ratio. Correlations for body
mass index ranged from 0.91 in White males to 0.94 in
Black females. For waist:height ratio, correlations
ranged from 0.92 in White males to 0.93 in White
W10
•4-1
c
>
>
>
>
>
>
Q)
'o
£0.8
O
O
-I
1
>
o
C 0.6
O
;
I
WA White Males (n=393)
^m Black Males (n=152)
I
1 White Females (n=606)
P?^3 Black Females (n=254)
>
>
>
>
>
>
>
>
'•+*
JH
0)
£ 0.4
o
>
>
>
>
>
>
>
>
)
)
>
o
c
O 0.2
12
(0
a>
a.
0.0
r
Q.
I
if
u.
w
m.
Q
a
,o>
u.
to
ara
'o
w
O
u
3
(0
.•£*
E
o
-c
,?
«
5
>
1
£
,9
aj
X
S
a
tt
if
5s
w
i
C
•
<
<
<
(
(
\
/
>
<
<
<
/
/
'
>
/
'
T
i
(o
;j
FIGURE 1. Pearson correlations between abdominal height and other obesity measures, by race and sex: Bogalusa Heart Study, 1995-1996.
All correlations except height: p < 0.001. Height for Black females: p < 0.05; height for all others: p > 0.05. SF, skinfold; BMI, body mass index.
Am J Epidemiol Vol. 151, No. 9, 2000
888
Gustat et al.
females. Height showed the weakest relation with
abdominal height in every grouping, with correlations
ranging from 0.05 in White females to 0.16 in Black
females. The correlation between height and abdominal
height was significant for Black females (p < 0.05) but
not for the other race-sex groups (p > 0.05).
Pearson correlations between the circumference
measures of central obesity (abdominal height, waist
circumference, conicity index, waist:hip ratio, and
waist:height ratio) and cardiovascular risk factor
variables for the overall samples are presented in figure 2. These variables were chosen because they are
measures of central obesity. The measures correlated
significantly with serum total cholesterol, triglycerides, very low density lipoprotein cholesterol, low
density lipoprotein cholesterol, high density lipoprotein cholesterol (inverse association), plasma glucose and insulin levels, and systolic and diastolic
blood pressures (p < 0.0001). Correlations were
highest with insulin; both abdominal height and
waist:height ratio showed higher correlations than
the other measures {p < 0.001) and were not found to
be significantly different (p > 0.05). Waist:hip ratio
was least correlated with insulin level (p < 0.001)
and significantly less correlated than the other measures of central obesity (p < 0.05). With respect to
high density lipoprotein cholesterol, abdominal
height showed weaker correlations than waist circumference and waist:hip ratio (p < 0.05). With very
low density lipoprotein cholesterol, abdominal
height was significantly less correlated than
waist:hip ratio (p < 0.05). Abdominal height also
showed a lower correlation (p < 0.05) than conicity
index with triglyceride level.
The correlations with insulin were further explored
for each race-sex group (data not shown). The correlations for abdominal height, waist:height ratio, and
waist circumference were similar and were not found
to be significantly different {p > 0.05) for the different
groups. However, abdominal height, waist:height
ratio, and waist circumference showed higher correlations than waist:hip ratio in all race-sex groups (p <
0.05). Conicity index showed a significantly smaller
correlation (p < 0.01) than abdominal height, waist
circumference, and waist:height ratio with insulin in
White males, White females, and Black females.
In a multivariate analysis which included waist circumference, triceps and subscapular skinfold thick-
0.8
n=1420
0)
"o
0)
o
o
c
o
0.6
0.4
0.2
is
2
0.0
o
-0.2
o
c
o
(5
Abdominal Height
• • • Waist
I
i Conicity
I
1 Waist/Hip
Rx^n Waist/Height
-0.4
Q)
Q.
-0.6
o
p
9
o
Q
Q
I
Q.
CO
CO
Q.
m
CO
IS
o
o
o
CO
o
FIGURE 2. Relation of central obesity measures (abdominal height, waist circumference, conicity index, waisthip ratio, and waist:height ratio)
to cardiovascular risk factor variables: Bogalusa Heart Study, 1995-1996. *Comparison of correlations with abdominal height—triglycerides (TG):
abdominal height was lower than conicity index (p < 0.05); very low density lipoprotein cholesterol (VLDL-C): abdominal height was lower than
waist:hip ratio (p < 0.05); high density lipoprotein cholesterol (HDL-C): abdominal height was lower than waist circumference and waist:hip ratio
(p < 0.05); insulin: abdominal height was greater than all other measures (p < 0.05) except waist:height ratio (p > 0.05); all other differences: p >
0.055. TC, total cholesterol; LDL-C, low density lipoprotein cholesterol; Sys. BP, systolic blood pressure; Dias. BP, diastolic blood pressure.
Am J Epidemiol
Vol. 151, No. 9, 2000
Abdominal Height and Cardiovascular Risk Factors 889
nesses, conicity index, weight, abdominal height, body
mass index, waist:hip ratio, and waist:height ratio in
addition to race, sex, age, and their interactions,
abdominal height was independently associated with
all of the measured risk factor variables (p < 0.05 to
p < 0.001) (table 2). Other central obesity measures
(waist circumference, conicity index, waist:hip ratio,
and waist:height ratio) were also independently related
to risk factor variables (p < 0.05 to p < 0.001), except
systolic and diastolic blood pressure. The total R2 values ranged from 0.130 to 0.520, with abdominal height
adding 0.3-3.2 percent of the variation. The greatest
improvement was noted in the models predicting
triglyceride (/?2's = 0.279-0.301) and insulin (R2's =
0.504-0.520) levels.
TABLE 2. Models predicting levels of cardiovascular
disease risk factors and partial R2's for abdominal height and
other central obesity variables: Bogalusa Heart Study,
1995-1996 (n=1,420)t
Partial Ft*
Dependent
variable
Abdominal
height
Other
central
obesity
measures§
0.136" *
0.301** *
0.006**
0.03**
0.020***
0.032***
0.212** •
0.130** *
0.236** •
0.152** *
0.520** •
0.265** *
0.241** •
0.009***
0.005*
0.004*
0.008***
0.032***
0.003*
0.003*
0.037***
0.014**
0.010**
0.009*
0.016***
0.004 (NSH)
0.005 (NS)
Total
R'X
Total cholesterol
Triglyceride (fasting only)
VLDLH cholesterol
(fasting only)
LDLH cholesterol
HDL11 cholesterol
Glucose
Insulin
Systolic blood pressure
Diastolic blood pressure
* p < 0.05; * • p < 0.01; *** p < 0.001.
t Sample size for fasting triglycerides and fasting VLDL cholesterol: n =
1,315.
t Analysis included waist circumference, triceps and subscapular skinfold thicknesses, conicity index, weight, abdominal height, body mass index,
waist:hip ratio, race, sex, and age.
§ Waist circumference, conicity index, waistihip ratio, and waist:height
ratio.
Tl VLDL, very low density lipoprotein; LDL, low density lipoprotein; HDL,
high density lipoprotein; NS, not significant.
The results of the canonical analysis of obesity measures, serum lipoprotein variables, blood pressure, and
plasma glucose and insulin levels are presented in
table 3 by race and sex. All canonical correlations were
significant (p < 0.001). Obesity measures (predictor
variables) as a whole correlated most strongly with the
canonical variable for glucose and insulin combined
(criterion variables) when compared with the canonical variables for lipoproteins and blood pressure. The
canonical correlations for the obesity measures and
glucose and insulin grouped together ranged from 0.68
for White males to 0.75 for White females. Among the
obesity measures, abdominal height showed stronger
relations with the canonical variables for lipoproteins,
blood pressure, and glucose and insulin levels than
nearly all of the other obesity measures. These trends
were similar for each race-sex group.
DISCUSSION
Although sophisticated and expensive imaging techniques such as computerized axial tomography (CAT)
and magnetic resonance imaging are accurate measures of abdominal fat deposition, inexpensive and
readily available methods are needed for epidemiologic studies. Understanding simplified methods is
important, since central obesity may be associated
with increased cardiovascular risk even in the presence
of slight overweight or normal weight (23). In this
study, abdominal height was found to correlate highly
with several of the obesity measures. It contributed
significantly (up to 3 percent of the variation) to the
prediction of each of the modeled cardiovascular risk
factors, suggesting that abdominal height is an additional parameter of risk. Abdominal height contributed
the most to the prediction of fasting triglyceride and
insulin levels and was important in predicting blood
pressure outcomes when other measures of central
obesity were not. Canonical correlation analysis
TABLE 3. Canonical analysis of obesity measures, serum lipoprotein (LP) variables, blood pressure (BP) variables, and plasma
glucose and insulin (Gl), by race and sex: Bogalusa Heart Study, 1995-1996
White males
Obesity
measure
All obesity measures
Abdominal height
Triceps skinfold
Subscapular skinfold
Waist
Conicity index
Waist/hip
Waist/height
Weight
Body mass indext
LP
BP
Gl
LP
BP
0.56***
0.51
0.25
0.39
0.50
0.49
0.54
0.50
0.40
0.43
0.45***
0.40
0.22
0.32
0.42
0.40
0.40
0.42
0.35
0.37
0.68***
0.68
0.45
0.55
0.65
0.54
0.56
0.64
0.61
0.63
0.55***
0.50
0.39
0.46
0.47
0.37
0.45
0.47
0.45
0.47
0.54***
0.42
0.30
0.40
0.41
0.44
0.39
0.41
0.35
0.34
* * * p < 0.001.
t Weight (kg)/height (mf.
Am J Epidemiol
Vol. 151, No. 9, 2000
White females
Black males
Gl
0.69***
0.67
0.62
0.67
0.65
0.55
0.51
0.65
0.61
0.63
Black females
LP
BP
Gl
LP
BP
0.58***
0.53
0.38
0.47
0.51
0.49
0.51
0.53
0.43
0.46
0.40***
0.37
0.29
0.35
0.37
0.27
0.25
0.36
0.38
0.38
0.75***
0.72
0.62
0.70
0.72
0.59
0.56
0.71
0.67
0.68
0.55***
0.46
0.36
0.37
0.49
0.51
0.46
0.46
0.43
0.39
0.48***
0.37
0.20
0.29
0.34
0.24
0.27
0.32
0.36
0.35
Gl
0.70***
0.68
0.58
0.63
0.68
0.52
0.50
0.67
0.66
0.67
890
Gustat et al.
showed that abdominal height was generally more
highly correlated with the canonical variables and was
therefore a slightly better predictor of serum lipoprotein levels, blood pressure, and plasma glucose and
insulin levels than the other obesity measures were.
Abdominal height has been shown to be highly correlated with visceral adipose tissue (24—26). Van der
Kooy et al. (25) reported that abdominal height may
better assess changes in visceral fat measured by magnetic resonance imaging than waist circumference and
waist:hip ratio. Kvist et al. (26), on the basis of CAT
scan measurements, suggested that visceral adipose
tissue be predicted from abdominal height rather than
from measurements of other areas of the body.
Additionally, waist circumference and abdominal
height were shown to be related to the metabolic and
blood pressure variables more than waist:hip ratio
(12). The multivariate analysis of the present study
suggested that the independent association of abdominal height with cardiovascular risk factors may be due
to its ability to measure a component of visceral obesity that the other indirect obesity measures miss.
Assessment of central obesity may contribute to the
evaluation of an individual's risk for coronary heart
disease (2, 3, 6). Kannel et al. (27) reported that total
risk for cardiovascular events increased with the
degree of central obesity, although they did not find
that one measure of central obesity was better than the
others in predicting coronary disease. Various measures of central obesity are often used in epidemiologic
studies. Waist circumference, either by itself or in relation to hip circumference or height, assesses abdominal fat and has been found to be highly correlated with
cardiovascular risk factors. It is interesting that
abdominal obesity measured by waist:height ratio was
found to be a better predictor of coronary disease risk
factors than waist:hip ratio (28). This ratio was also
shown to be a strong predictor of abdominal obesity
(29). Another measure used to assess abdominal adiposity is the conicity index (30). This is a standardized
index suggested to measure the circumference of the
body as it departs from a cylindrical shape, and it indicates the proportion of total body fat that may be found
in the midsection of the body. However, questions
have arisen as to whether abdominal girth should be
used in the formula at all (31). Kahn (31) suggests that
abdominal height would be preferable.
It has been reported that neither waist girth nor
abdominal height alone is sufficient to distinguish
coronary disease outcomes in a case-control study
(13). However, the ratio of abdominal height to
midthigh girth was associated more highly with coronary disease outcomes than was waist:hip ratio in that
study. Pouliet et al. (12) showed that abdominal height
measured by CAT scan and waist circumference are
better than waist:hip ratio in assessing central visceral
obesity. However, they did not measure abdominal
height manually and compare it with measurements
derived from CAT scans. The present study shows that
manually measured abdominal height significantly
contributes to the prediction of cardiovascular risk factor levels after adjustment for other central obesity
measures, including waist:hip ratio and waist circumference (table 2).
In summary, abdominal height adds another dimension to central obesity measurement. It contributes to
the prediction of cardiovascular risk factor levels similarly to other measures of obesity but may help to
assess a component of visceral fat deposition that the
other measures miss. The preciseness needed for measurement of body fatness and the availability of instrumentation beyond that used for weight, height, waist,
and skinfold measurements should determine the practicality of measurements used.
ACKNOWLEDGMENTS
This research was supported by grant HL38844 from the
National Heart, Lung, and Blood Institute.
The Bogalusa Heart Study is a joint effort of many investigators and staff members whose work is gratefully
acknowledged. The authors thank the Bogalusa Heart Study
field staff and the Core Laboratory staff.
REFERENCES
1. Barrett-Connor EL. Obesity, atherosclerosis, and coronary
artery disease. Ann Intern Med 1985;103:1010-19.
2. Peiris AN, Sothmann MS, Hoffman RG, et al. Adiposity, fat
distribution, and cardiovascular risk. Ann Intern Med 1989;
110:867-72.
3. Rimm EB, Stampfer MJ, Giovannucci E, et al. Body size and
fat distribution as predictors of coronary heart disease among
middle-aged and older US men. Am J Epidemiol 1995;141:
1117-27.
4. Hubert HB, Feinleib M, McNamara PM, et al. Obesity as an
independent risk factor for cardiovascular disease: a 26-year
follow-up of participants in the Framingham Heart Study.
Circulation 1983;67:968-77.
5. Must A, Jacques PF, Dallal GE, et al. Long-term morbidity and
mortality of overweight adolescents: a follow-up of the
Harvard Growth Study of 1922 to 1935. N Engl J Med 1992;
327:1350-5.
6. Rexrode KM, Carey VJ, Hennekens CH, et al. Abdominal adiposity and coronary heart disease in women. JAMA 1998;
280:1843-8.
7. Eckel RH, Krauss RM. American Heart Association call to
action: obesity as a major risk factor for coronary heart disease. AHA Nutrition Committee. Circulation 1998;97:
2099-2100.
8. Despr6s JP, Moorjani S, Lupien PJ, et al. Regional distribution
of body fat, plasma lipoproteins, and cardiovascular disease.
Am J Epidemiol Vol. 151, No. 9, 2000
Abdominal Height and Cardiovascular Risk Factors
Arteriosclerosis 1990;10:497-511.
9. Kissebah AH, Vydelingum N, Murray R, et al. Relation of
body fat distribution to metabolic complications of obesity. J
Clin Endocrinol Metab 1982;54:254-60.
10. Segal KR, Dunaif A, Gutin B, et al. Body composition, not
body weight, is related to cardiovascular disease risk factors
and sex hormone levels in men. J Clin Invest 1987;80:1050-5.
11. Lapidus L, Bengtsson C, Larsson B, et al. Distribution of adipose tissue and risk of cardiovascular disease and death: a 12
year follow up of participants in the Population Study of
Women in Gothenberg, Sweden. Br Med J (Clin Res Ed) 1984;
289:1257-61.
12. Pouliot MC, Despres JP, Lemieux S, et al. Waist circumference
and abdominal sagittal diameter: best simple anthropometric
indexes of abdominal visceral adipose tissue accumulation and
related cardiovascular risk in men and women. Am J Cardiol
1994;73:460-8.
13. Kahn HS, Austin H, Williamson DF, et al. Simple anthropometric indices associated with ischemic heart disease. J Clin
Epidemiol 1996;49:1017-24.
14. Valdez R. A simple model-based index of abdominal adiposity.
J Clin Epidemiol 1991;44:955-6.
15. Berenson, GS, McMahan CA, Voors AW, et al. Cardiovascular
risk factors in children: the early natural history of atherosclerosis and essential hypertension. New York, NY: Oxford
University Press, 1980.
16. Berenson GS, ed. Causation of cardiovascular risk factors in
children: perspectives on cardiovascular risk in early life. New
York, NY: Raven Press, 1986.
17. Sjostrom L. A computer-tomography based multicompartment
body composition technique and anthropometric predictions of
lean body mass, total and subcutaneous adipose tissue. Int J
Obes 1991;15(suppl 2): 19-30.
18. Allain CC, Poon LS, Chan CS, et al. Enzymatic determination
of total serum cholesterol. Clin Chem 1974;20:470-5.
19. Bucolo G, David H. Quantitative determination of serum
triglycerides by the use of enzymes. Clin Chem 1973;19:
476-82.
Am J Epidemiol
Vol. 151, No. 9, 2000
891
20. Srinivasan SR, Berenson GS. Serum lipoproteins in children
and methods for study. In: Lewis LA, ed. CRC handbook of
electrophoresis, vol 3. Lipoprotein methodology and human
studies. Boca Raton, FL: CRC Press, 1983:185-204.
21. SAS Institute, Inc. SAS/STAT user's guide, version 6. 4th ed.
Cary, NC: SAS Institute, Inc, 1989.
22. Dillion WR, Goldstein M. Multivariate analysis: methods and
applications. New York, NY: John Wiley and Sons, Inc, 1984.
23. Bouchard C, Bray GA, Hubbard VS. Basic and clinical aspects
of regional fat distribution. Am J Clin Nutr 1990;52:946-50.
24. Zamboni M, Turcato E, Armellini F, et al. Sagittal abdominal
diameter as a practical predictor of visceral fat. Int J Obes
Relat Metab Disord 1998;22:655-60.
25. van der Kooy K, Leenen R, Seidell JC, et al. Abdominal diameters as indicators of visceral fat: comparison between magnetic resonance imaging and anthropometry. Br J Nutr 1993;
70:47-58.
26. Kvist H, Chowdhury B, Grangard U, et al. Total and visceral
adipose-tissue volumes derived from measurements with computed tomography in adult men and women: predictive equations. Am J Clin Nutr 1988;48:1351-61.
27. Kannel WB, Cupples LA, Ramaswami R, et al. Regional obesity and risk of cardiovascular disease: The Framingham
Study. J Clin Epidemiol 1991 ;44:183-90.
28. Hsieh SD, Yoshinaga H. Abdominal fat distribution and coronary heart disease risk factors in men—waist/height ratio as a
simple and useful predictor. Int J Obes Relat Metab Disord
1995; 19:585-9.
29. Ashwell M, Cole TJ, Dixon AK. Ratio of waist circumference
to height is strong predictor of intra-abdominal fat. (Letter).
BMJ 1996;313:559-60.
30. Valdez R, Seidell JC, Ahn YI, et al. A new index of abdominal
adiposity as an indicator of risk for cardiovascular disease: a
cross-population study. Int J Obes Relat Metab Disord 1993;
17:77-82.
31. Kahn HS. Choosing an index for abdominal obesity: an opportunity for epidemiologic clarification. J Clin Epidemiol 1993;
46:491^.